Child Health and
the Environment
DONALD T. WIGLE, MD, PhD, MPH
OXFORD UNIVERSITY PRESS
Child Health
and the
Environment
This page intentionally left blank
Child Health
and the
Environment
DONALD T. WIGLE, MD, PhD, MPH
R. Samuel McLaughlin Centre for Population
Health Risk Assessment
Institute of Population Health
University of Ottawa
Ottawa, Canada
1
2003
1
Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai
Nairobi São Paulo Shanghai Taipei Tokyo Toronto
Copyright © 2003 by Oxford University Press, Inc.
Published by Oxford University Press, Inc.
198 Madison Avenue, New York, New York, 10016
http://www.oup-usa.org
Oxford is a registered trademark of Oxford University Press
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.
Library of Congress Cataloging-in-Publication Data
Wigle, D. T.
Child health and the environment / Donald T. Wigle.
p. cm.
Includes bibliographical references and index.
ISBN 0-19-513559-8
1. Pediatric toxicology. 2. Environmentally induced diseases in children.
3. Children—Health risk assessment. I. Title.
RA1225.W545 2003
615.9'0083—dc21 2002026944
987654321
Printed in the United States of America
on acid-free paper
To Beth,
children everywhere,
and Garreth,
who bravely fought non-Hodgkin’s lymphoma
from age 10 to 19 years.
This page intentionally left blank
Acknowledgments
I thank all who contributed to this book including the anonymous reviewers. In particular, I thank John Last for his encouragement from the
time this book was just an idea, and Jeffrey House for his many helpful
suggestions, support, and patience. Special thanks go to colleagues who
reviewed chapters and provided many helpful suggestions: Tye Arbuckle,
Rick Burnett, Bob Dales, Eric Dewailly, Warren Foster, Rick Gallagher,
Howard Morrison, Dieter Riedel, Pat Rasmussen, Will Robertson, Bob
Spasoff, Paul Villeneuve, Slavica Vlahovich, and Mike Wade. I also thank
Edith Barry, Lynda Crawford, and other Oxford staff for their help in improving the manuscript.
vii
This page intentionally left blank
Preface
The public health goal of collectively assuring the conditions in which
people can be healthy1 is particularly relevant to children as they are vulnerable to environmental hazards but have little or no control over their
environmental conditions. Children differ profoundly from adults with
respect to physiology, metabolism, growth, development, and behavior.
By interfering with child growth and development during critical time
periods, environmental hazards may cause structural and functional
deficits and lifelong disability. The long life expectancy of children carries the potential for relatively high cumulative exposures and time to develop delayed adverse health outcomes; for instance, intense sun exposure during childhood is a major determinant of adult melanoma risk.
This book explores potential health outcomes of prenatal and childhood exposure to environmental hazards, particularly anthropogenic contaminants. Among the overarching themes are the susceptibility of the
rapidly developing fetus and infant to early-life toxic exposures and the
importance of modifying factors (e.g., poverty, genetic traits, nutrition)
and timely intervention. Public health policy development in this field
1 National
Academy of Sciences. 1988. The future of public health. Washington, D.C.:
National Academy Press.
ix
x
PREFACE
must respond to high public concern about the safety and well-being of
children but is complicated by the multiplicity of environmental contaminants, major knowledge gaps, the limitations of toxicologic and epidemiologic studies, and a lack of scientific consensus on causal relationships. Under the precautionary principle, lack of full scientific certainty
does not justify postponement of cost-effective measures to prevent significant potential public health risks. This book documents several historic examples of environment-related child health disasters resulting
from failures to apply the precautionary principle.
Chapters 1 to 3 present overviews of key children’s environmental
health issues and the role of environmental epidemiology and risk assessment in child health protection. Chapter 1 shows that international,
national, and other bodies have identified asthma, air pollution (indoor
and outdoor), lead, pesticides, water contaminants (chemical and microbial), climate change, hormonally active agents, and environmental
tobacco smoke as important environmental health issues for children.
Epidemiologic studies have identified adverse health effects during gestation, childhood, and adulthood arising from early-life exposure to diverse environmental toxicants such as ionizing radiation, lead, methylmercury, polychlorinated biphenyls (PCBs), environmental tobacco
smoke, and outdoor air pollutants. Nevertheless, knowledge about the
proportions of prenatal, childhood, and adult adverse health outcomes
that are attributable to prenatal and childhood environmental exposures
is very limited. Chapter 2 illustrates epidemiologic strengths with published examples and discusses their limitations such as problems in quantifying exposures, assessing delayed effects, and limited ability to detect
relatively small risks.
When there is evidence that a particular environmental factor poses
a threat to human health, regulatory authorities face the challenge of deciding how much population exposure should be permitted. Chapter 3
describes the role of the U.S. National Academy of Sciences risk assessment framework in quantifying health risks for the purpose of setting exposure limits protective of human health. Among the issues covered are
processes used by national and international agencies to assess causal relationships, the assessment of dose–response relationships, uncertainties
surrounding recommended exposure limits, and the need for improved
premarket testing of commercial chemicals for early-life toxicity. The
chapter also addresses key issues related to risk assessments of carcinogens, reproductive toxicants, developmental toxicants, and neurotoxins.
Lead and mercury are potent neurotoxins and the developing fetus
and infant are especially sensitive to their effects. The role of social factors (especially poverty) in childhood lead exposure; the importance of
Preface
xi
physiologic, nutritional, and developmental factors unique to childhood;
possible developmental effects of prenatal parental lead exposure; the apparent absence of a blood lead threshold for hearing and cognitive deficits;
and the persistent effects of childhood lead exposure on adolescent and
adult cognitive performance are all discussed in Chapters 4 and 5. Major
issues include controversies surrounding the efficacy of blood lead screening and lead abatement interventions, the possible neurotoxic effects of
relatively low-level dietary methylmercury, and the need to further reduce the levels of population exposure to these toxicants. Sources and potential health effects of inorganic mercury, elemental mercury, arsenic, cadmium, and manganese are also discussed.
The disastrous health effects (intrauterine growth retardation, developmental delays, cognitive deficits, and chloracne) among infants of
women highly exposed to PCBs are discussed in Chapter 6. Also covered
are the possible health effects of early-life exposure to background levels
of PCBs and related organochlorine compounds that share a common
mechanism of toxicity, that is, activation of the aryl hydrocarbon receptor. This chapter documents the global dispersion and bioaccumulation
of organochlorine compounds in aquatic and terrestrial food chains, and
it flags such issues as uncertainties around the potential neurotoxic effects of relatively low-level lactational exposure to PCBs and related compounds, and the need to reduce human exposures.
About 3 million tons of conventional pesticide active ingredient
chemicals are used annually worldwide, inevitably exposing the developing fetus and child to at least trace levels of currently used and persistent agents (e.g., DDT). Chapter 7 examines the known and potential
health effects of pesticides, including acute poisonings, developmental effects, reproductive effects, neurotoxicity, and cancer. It addresses their
potential to disrupt fetal and childhood growth and developmental processes and the inadequacy of premarket toxicity testing, as well as the
potential role of pesticides in childhood cancer and the need to monitor
population exposure levels.
Human experience with the drug diethylstilbestrol (DES) showed
that prenatal exposure to this potent synthetic estrogen could cause reproductive tract abnormalities and vaginal cancer in offspring. Several
environmental contaminants modulate endocrine function in experimental animals but their possible roles in human fetal and child development
and health are unknown as very few epidemiologic studies have addressed these issues, and almost none have measured internal doses.
Chapter 8 covers the potential roles of hormonally active environmental
contaminants in the apparent trends toward reduced average age at
menarche, reduced sperm quality, increased male reproductive tract birth
xii
PREFACE
defects, and increased cancer incidence rates. The importance of monitoring population exposures to hormonally active contaminants and
tiered toxicity testing of high-production volume chemicals for hormonal
activity are noted.
Children prenatally exposed to atomic bomb radiation had substantially increased risks of microcephaly and severe mental retardation while
those exposed as young children during the 1986 Chernobyl nuclear accident had increased thyroid cancer risks. Relatively low-level prenatal exposure to medical X-rays appears to increase the risk of childhood leukemia but the possible role of low-level environmental radiation in adverse
developmental outcomes (birth defects, intrauterine growth retardation)
and childhood cancer is poorly understood. These issues and those related
to other types of electromagnetic radiation are presented in Chapter 9. After describing the mixed evidence of a link between power-frequency magnetic fields and childhood cancer, this chapter notes widespread exposures
and uncertainties, and calls for precautionary measures such as minimizing exposure during pregnancy and childhood. This chapter documents
the role of intense childhood sun exposure in malignant melanoma and
its possible links to cataracts and immune suppression. Also addressed are
uncertainties related to the efficacy of sunscreen use in preventing melanoma, the importance of multiple interventions in protecting children from
intense sun exposure, and the need to monitor their exposures and sunprotective behaviors (including those of their guardians).
Environmental tobacco smoke (ETS) is a major indoor air contaminant with known adverse effects on child health, notably respiratory and
middle ear infections, lung function deficits, and asthma. In addition, ETS
appears to cause sudden infant death syndrome and is a possible cause
of childhood and adult cancers. Biologic agents, toxic gases, and volatile
organic carbons (VOCs) comprise the other indoor air pollutants reviewed
in Chapter 10. The chapter addresses the roles of house-dust mite antigens and other aeroallergens in the onset or exacerbation of asthma, carbon monoxide poisoning from indoor combustion sources, the possible
role of VOCs and nitrogen dioxide in childhood respiratory disease and
cancer, and the virtual absence of national comprehensive prevention programs for indoor air health hazards.
Children are vulnerable to outdoor air pollution because they often
engage in physical activities or play outdoors; also, they have relatively
high air intake compared to adults. Chapter 11 describes the role of major outdoor air pollutants in adverse developmental outcomes (intrauterine growth retardation, preterm birth), respiratory tract inflammation and
hyperreactivity, lung function deficits, and respiratory illnesses such as
incident asthma. The need to minimize exposure of children and preg-
Preface
xiii
nant women to ambient air pollutant levels from major sources, especially
motor vehicles and industry is noted. Evaluation of progress in this field
requires monitoring personal exposures and prevalence/incidence rates
of potential respiratory health effects.
An adequate supply of safe drinking water is an elusive goal in economically disadvantaged regions globally. Chapter 12 addresses waterborne microbial and chemical hazards, the latter including chlorination
disinfection by-products (DBPs) and toxic chemical contaminants from
hazardous waste disposal and other sources. Emerging evidence of developmental effects (spontaneous abortions, stillbirths, intrauterine
growth retardation, and certain birth defects) related to first trimester maternal DBP exposure, the susceptibility of surface waters to high DBP levels, and economic barriers to DBP abatement, particularly in small water
systems, are discussed. Hazardous waste disposal and storage often contaminate groundwater but the few existing epidemiologic studies preclude an adequate assessment of this child health hazard. Needs to protect source waters, reduce DBP levels, control hazardous waste disposal,
and monitor water quality are noted.
Children develop enteric infections by ingesting fecally contaminated
water and other substances (e.g., formula mixed with contaminated
water) and by engaging in hand-mouth behavior. Although such infections are a leading cause of childhood deaths in economically disadvantaged countries globally, Chapter 12 focuses mainly on microbial threats
in developed countries. Escherichia coli 0157:H7, a major cause of acute renal failure in children, can be transmitted through contaminated drinking water. Water appears to be a source of Helicobacter pylori infection during childhood; this organism causes chronic gastrointestinal infection and
increased risks of peptic ulcers and stomach cancer during adulthood.
The need to address both microbial and disinfection by-product hazards
is a major risk management challenge.
This book will interest professionals and graduate students in the
fields of public health, pediatrics, environmental health, epidemiology,
and toxicology. The introductory and concluding segments of each chapter should interest a wider audience including health policy analysts in
voluntary and governmental agencies. The final chapter summarizes the
associations between environmental exposures and child health outcomes
described in the previous chapters and calls for measures to create the evidence needed to enable public health decisions protective of child health.
The five tables in this chapter are unique in that they summarize available information on the burden of child health adverse outcomes and the
potential role of environmental hazards together with the level of epidemiologic evidence.
This page intentionally left blank
Contents
1. Environmental Threats to Child Health: Overview, 1
Environmental Health Concerns About Children, 2
Children’s Vulnerabilities, 7
Risk Management Issues, 14
2. Environmental Epidemiology, 27
Epidemiology, 27
Study Types, 34
3. Risk Assessment, 47
Risk Assessment Framework, 47
Selected Risk Assessment Practices, 58
4. Metals—Lead, 71
Health Effects, 72
Exposures, 79
Risk Management, 82
xv
xvi
CONTENTS
5. Metals—Mercury, Arsenic, Cadmium, and Manganese, 99
I. Mercury, 100
Methylmercury, 100
Elemental and Inorganic Mercury, 113
II. Other Metals and Metalloids, 117
Arsenic, 117
Cadmium, 121
Manganese, 125
6. PCBs, Dioxins, and Related Compounds, 136
Health Effects, 137
Exposures, 149
Risk Management, 150
7. Pesticides, 162
Health Effects, 163
Exposures, 171
Risk Management, 173
8. Hormonally Active Agents, 189
Normal Endocrine Function, 192
Mechanisms of Environmental Hormonally Active Agents, 201
Health Effects, 208
Exposures, 214
Risk Management, 215
9. Radiation, 229
I. Ionizing Radiation, 230
Health Effects, 231
Exposures, 239
Risk Management, 240
II. Power Frequency Electric and Magnetic Fields
and Radiofrequency Radiation, 243
Health Effects, 244
Exposures, 252
Risk Management, 253
III. Sunlight, 256
Health Effects, 257
Exposure, 260
Risk Management, 260
10. Indoor Air, 270
Asthma, 272
Environmental Tobacco Smoke, 275
Contents
Biologic Agents, 281
Volatile Organic Chemicals and Gases, 288
11. Outdoor Air, 300
Health Effects, 301
Exposures, 312
Risk Management, 317
12. Water, 334
I. Chemical Contaminants, 334
Health Effects, 335
Exposures, 342
Risk Management, 345
II. Waterborne Infections, 350
Bacteria, 351
Protozoa, 356
Viruses, 358
Risk Management, 359
13. Conclusion, 366
Environmental Threats to Child Health, 366
Knowledge Development Policy Issues, 375
Epilogue, 380
Index, 383
xvii
This page intentionally left blank
Child Health
and the
Environment
This page intentionally left blank
1
Environmental Threats to
Child Health: Overview
Control of childhood infections through sanitation, immunization, improved nutrition and housing, and antibiotics during the twentieth century greatly increased life expectancy at birth and dramatically changed
patterns of childhood illnesses in developed countries. But there is growing evidence that global changes in atmosphere, terrestrial ecosystems,
and climate, driven by population increase and consumption, pose threats
to current and future human health. Children are especially vulnerable
because they have no control over their prenatal and postnatal environments, including the quality of the air they breathe, the water they drink,
the food they eat, and their place of residence. Exposure to environmental toxicants during prenatal and early childhood periods can disrupt developmental processes, causing structural and functional abnormalities
that range from subtle to obvious, immediate to delayed, and transient to
permanent. The leading health conditions that result in illness, disability,
and death among children now include asthma, unintentional injuries,
cancer, low birth weight, neurodevelopmental deficits, and birth defects.
Apart from injuries, the proportions of these conditions attributable to environmental hazards are uncertain or unknown. By using a Delphi process and other sources to estimate attributable risks and economic impacts, a recent study concluded that 100% of lead poisoning, 30% of
1
2
CHILD HEALTH
AND THE
ENVIRONMENT
asthma, 5% of cancer, and 10% of neurobehavioral disorders among children in United States are caused by environmental pollutants and impose
an economic burden of about $55 billion annually (Landrigan et al., 2002).
Enteric and related infections caused by use of fecally contaminated
water and respiratory conditions related to indoor and outdoor air pollution cause about 13% of all disability-adjusted life years (DALYs) lost
globally, with considerably higher proportions in economically disadvantaged regions. Indigenous groups dependent on traditional foods may
have increased risks of exposure to environmental hazards such as
methylmercury in fish, organochlorine compounds in whale blubber, and
lead in waterfowl. Even in developed countries, children in disadvantaged groups have higher levels of exposure to environmental hazards
such as lead, environmental tobacco smoke (ETS), cockroach antigen, and
outdoor air pollution and lower access to protective interventions such
as sunscreens.
This book addresses the impacts of chemical, radiologic, and biologic
environmental contaminants on child health and development from conception to early adulthood. Chapters 2 and 3, respectively, describe the
roles of epidemiology and environmental risk assessment in providing an
evidence base for public health policy and program decisions. Succeeding chapters review current evidence on major environmental hazards including lead, mercury and other heavy metals, dioxins, polychlorinated
biphenyls, pesticides, radiation, hormonally active agents, indoor air, outdoor air, and drinking water. The final chapter summarizes known and
suspected environmental threats to child health and policy issues arising
from knowledge gaps. The present chapter deals with important issues
related to child health, environmental hazards, vulnerabilities of the developing fetus and child, and the prevention and control of environmental hazards.
Environmental Health Concerns About Children
Leading Adverse Child Health Outcomes
Major pregnancy and child health outcomes (events per year in the United
States) and their known or suspected environmental links are presented
in Table 1–1. The very large annual number of fetal deaths (almost 1 million recognized events) and low-birth-weight infants (about 300,000) indicate the major impact of these conditions on population health. The
reported number of fetal deaths is likely to have been substantially underestimated; longitudinal studies of women using biomarkers to detect
Environmental Threats to Child Health: Overview
3
TABLE 1–1. Adverse Fetal, Infant, and Child Health Outcomes, United States
(a) Adverse pregnancy outcomes
Outcome
Number of Events
Fetal deaths a
Low birth weight b
1,500 g
1500–2499 g
Total
983,000
57,477
243,706
301,183
(b) Childhood diseases
Hospitalizations e
Deaths c,d
Age 1 Yr
Perinatal conditions (low
birth weight, complications
of pregnancy, other)
Birth defects
SIDS
Respiratory disease—total
(asthma)
Cardiovascular diseases
Gastrointestinal diseases
Cancer and other neoplasms
Nervous system diseases
Certain infectious diseases
Injuries (including poisonings)
Other
Total
a Miscarriages
14,084
5,473
2,648
687
(5)
667
500
126
441
562
1,285
1,464
27,937
Age 1–14 Yr
135
977
—
644
(153)
591
249
1,594
877
493
6,163
1,121
12,844
Age 15 Yr
Age 15 Yr
14,219
170,000
6,450
2,648
1,331
(158)
1,258
749
1,720
1,318
1,055
7,448
2,585
40,781
130,000
—
741,000
(190,000)
25,000
221,000
36,000
89,000
173,000
228,000
645,000
2,458,000
and stillbirths, USA, 1996. Source: Ventura et al. (1999).
b USA,
1999. Ventura et al. (2001).
c USA,
1999. Hoyert et al. (2001).
d USA,
1999. Anderson (2001).
e USA,
1999. Popovic (2001).
early pregnancy have shown that 20%–40% of conceptions end in fetal
death before 20 weeks’ gestation but only a quarter to a half are clinically
recognized. Among human early fetal deaths, about 10%–20% have
autosomal chromosome aneuploidy and 10%–20% have congenital heart
defects.
Leading causes of infant deaths include perinatal conditions (e.g., conditions related to complications of pregnancy, labor, or delivery, preterm
birth, intrauterine growth retardation, birth trauma, or respiratory distress), birth defects, and sudden infant death syndrome (SIDS). Among
4
CHILD HEALTH
AND THE
ENVIRONMENT
older children, the main fatal conditions are injuries, cancer, and birth defects. The main causes of childhood hospitalization are respiratory diseases
(infections, asthma), injuries, and gastrointestinal conditions (infections,
other). All of these conditions have known or suspected environmental
links. Reported increases of birth defects and cancers among children and
young adults and their known or suspected links to preconceptual, prenatal, and childhood exposures have raised public concern.
During the period from 1974 to about 1984 in Canada, incidence rates
of overall childhood cancer increased by about 15%; since then, incidence
rates of total and specific childhood cancers (leukemia, brain cancer,
Hodgkin’s disease, and non-Hodgkin’s lymphoma) have been relatively
constant. Childhood leukemia incidence rates increased in the United
States during 1974–1991 but appear to have decreased slightly thereafter.
Childhood brain cancer incidence rate increases have been reported in
several countries; the U.S. increases occurred mainly during 1983–1986,
possibly due to improved detection of low-grade cerebral and brainstem
gliomas after the introduction of magnetic resonance imaging (MRI)
(Smith et al., 1998). In sum, there is little convincing evidence of childhood cancer incidence rate increases since the late 1980s.
Incidence rates for several types of cancer have increased among
young adults in Canada and some other countries during recent decades:
melanoma, thyroid cancer (especially among females), testicular cancer,
and non-Hodgkin’s lymphoma. Intense sun exposure during childhood
appears to explain most of the increase in melanoma. Some of the increase
in non-Hodgkin’s lymphoma among men was likely caused by human
immunodeficiency virus (HIV) infection, but this cannot explain the striking global increases in this cancer across gender and age groups and beginning before the HIV epidemic. Infection with simian virus 40 was recently identified as a possible cause of non-Hodgkin’s lymphoma among
persons who received contaminated polio vaccines during 1955–1963
(Shivapurkar et al., 2002; Vilchez et al., 2002). Exposure to medical X-rays
during childhood and youth may partially explain the increased thyroid
cancer rates. Unexplained are testicular cancer incidence rates that increased twofold or more during the past three or four decades in several
geographic regions, especially among more recent birth cohorts (see, e.g.,
Liu et al., 1999).
Recognized Children’s Environmental Health Issues
There is considerable consistency in the children’s environmental health
issues identified as important by international and national agencies
(Table 1–2). Six or more of the ten agencies acknowledged asthma, air (indoor and outdoor), lead, pesticides, and water contaminants (chemical
Environmental Threats to Child Health: Overview
5
TABLE 1–2. Children’s Environmental Health Issues Identified by International,
National, and Other Organizations
Issue
Air—outdoor
Water—infectious agents
Lead
Asthma
Air—indoor
Pesticides
Water—chemical contaminants
Environmental tobacco smoke
Poverty
Hormonally active agents
Hazardous waste sites
Radiation (sunlight/
ultraviolet radiation)
Birth defects
Cancer
Climate change
Number
of Agencies
9
9
8
7
7
7
6
5
4
4
4
4
3
3
3
Issue
Other metals (mercury,
cadmium)
Persistent organic pollutants
Acute respiratory infections
Developmental disabilities
(cerebral palsy, autism,
learning disabilities,
hearing loss)
Injuries
Radiation (ionizing)
Acute poisonings
Diarrheal diseases
Food—contaminants
Genetically modified
organisms
Vectors of parasitic diseases
(malaria)
Number
of Agencies
3
3
2
2
2
2
1
1
1
1
1
Sources: World Health Organization (2001), European Centre for Environment and Health (1999), Lvovsky
(2001) (note—the survey was not targeted to children, but the issues identified are strongly linked to
child health), Pan American Center for Sanitary Engineering and Environmental Science (2001), G8 countries (1998), Commission for Environmental Cooperation (2000), U.S. Environmental Protection Agency
(1996), Centers for Disease Control and Prevention (2000), U.S. Department of Health and Human Services (2000) (note: these are objectives for the entire U.S. population but relate directly to children’s health),
Council of State and Territorial Epidemiologists (2001) (note: these are objectives for the entire U.S. population but relate directly to children’s health).
and microbial) as important issues directly relevant to children. The World
Health Organization (WHO) Europe and the G8 nations (Canada, France,
Germany, Italy, Japan, Russia, United Kingdom, United States of America) added three other issues—climate change, hormonally active agents
(HAAs), and environmental tobacco smoke. An Environmental Protection
Agency (EPA) advisory committee noted five areas where children’s needs
were not adequately addressed by existing EPA regulations: mercury
emissions, farm worker protection, triazine pesticides (atrazine in drinking water), organophosphate and carbamate insecticides (neurotoxicity),
and indoor and outdoor air quality and asthma.
Knowledge of Environmental Threats to Child Health
The relative importance of children’s environmental health issues could
be approached by measuring the frequency of adverse health conditions
among children, identifying their causal factors, measuring exposure to
6
CHILD HEALTH
AND THE
ENVIRONMENT
causal factors, and estimating attributable risks. In practice, this is only
partially feasible because of major gaps in understanding the relationships
between prenatal and childhood environmental exposures and adverse
health outcomes. On the one hand, up to 60% of SIDS deaths may be caused
by prenatal and postnatal tobacco smoke exposure. On the other hand,
several epidemiologic studies have shown associations between childhood
cancer and pesticide exposure indices, but causality is uncertain because
of the lack of strong epidemiologic evidence, positive or negative.
Cancer incidence patterns among identical and nonidentical twins indicate that about 80% of cancers that occur commonly during childhood
are attributable to environmental factors or gene–environment interactions, but these remain largely unknown. The U.S. National Academy
of Sciences estimated that toxic chemical and physical agents cause about
3% of all developmental disorders and that a combination of genetic and
nongenetic factors (including infections, tobacco, alcohol, and environmental hazards) may cause about 25% of these disorders (National Academy of Sciences, 2000). It is important to recognize that very few epidemiologic studies of adverse developmental outcomes with large sample
sizes and rigorous exposure assessment have been conducted. It seems
likely that future studies will reveal higher attributable risks of environmental factors than current estimates indicate. Although the impacts of
environmental toxicants on child health have not been quantified systematically, succeeding chapters describe many important known links.
Some environmental hazards have been assessed in a least a few
high-quality epidemiologic studies (e.g., childhood leukemia and powerfrequency magnetic field exposures), but uncertainties related to exposure
indices, mixed exposures, low frequency of highly exposed subjects, and
inconsistent findings have complicated the interpretation of study results.
Although there have been many studies of pesticides and childhood cancer, few have had strong statistical power and exposure assessment. Despite extensive research on the exacerbation of asthma by environmental
exposures, there have been relatively few studies of causal factors for incident asthma. In addition to adverse health outcomes during childhood
and youth, early-life exposure to environmental hazards may cause cancer and other adverse health effects during adulthood.
The childhood equivalent of the Framingham Heart Study or the U.S.
Nurses’ Health Study would be large longitudinal studies with intensive
environmental exposure assessments beginning before conception or during early pregnancy with prolonged follow-up to identify health outcomes
during pregnancy, infancy, childhood, adolescence, and adulthood. Such
studies, initiated in Europe and at the planning stage in the United States,
promise to provide much needed information on a wide variety of po-
Environmental Threats to Child Health: Overview
7
tential health outcome and environmental exposure relationships (Golding et al., 2001; National Institute of Child Health and Human Development 2001).
Children’s Vulnerabilities
The genome controls prenatal and postnatal growth and function but, as
documented throughout this book, genes and the many molecular processes they control can be disrupted by environmental hazards. Inherited
mutations and a wide range of social, behavioral, or other factors that increase exposure to environmental hazards can all increase a child’s vulnerability. Sociodemographic subgroups of children may have both higher
exposures related to older or deficient housing, residence in areas with
high outdoor air pollution, dependence on contaminated drinking water,
consumption of traditional foods, and the presence of household smokers and increased vulnerability because of maternal and childhood dietary deficiency and other risk factors.
The National Research Council report Pesticides in the Diets of Infants
and Children documented age-related population heterogeneity with respect to exposure levels and toxicity (National Academy of Sciences, 1993).
This report noted several unique aspects of children: (1) rapid body
growth and development—the underlying molecular and cellular processes are vulnerable to disruption by toxicants, causing irreversible adverse effects on body structure (birth defects, reduced growth rates) and
function, (2) the potential for relatively high exposures related to children’s diet, behavior, and physiologic/metabolic differences from adults,
(3) immature detoxification systems, and (4) inadequate toxicity testing
of chemicals for developmental, neurobehavioral, immunologic, and reproductive system effects of perinatal exposures.
Disruption of Growth and Developmental Processes
Development has been described as evolution’s foremost accomplishment
in gene regulation, involving a complex orchestration of genes activated
in specific cells at specific times (National Academy of Sciences, 2000).
The cascades of genetically controlled molecular processes that underlie
growth and development from fertilized egg to mature youth create
periods during which toxic exposures can cause irreversible structural
and/or functional abnormalities. Periods of vulnerability for adverse developmental outcomes depend on the mechanism of action of a given tox-
8
CHILD HEALTH
AND THE
ENVIRONMENT
icant, the dose of toxicant taken up by the target tissue, the developmental timetable of the target tissue, and the age at evaluation of outcomes.
Birth Defects
Although there are known risk factors for birth defects (e.g., maternal
smoking and alcohol consumption during pregnancy, relative folic acid
deficiency, and use of certain pharmaceuticals), the attributable risks are
generally low and the causes of most birth defects remain unknown. Research in this field is complicated by the fact that spontaneous abortion
during the first trimester is quite common (20%–40% of all conceptions
and about 10% of recognized pregnancies), and a high proportion of affected fetuses have birth defects and/or chromosomal abnormalities.
Studies of birth defects among infants therefore include only a fraction of
incident cases, that is, those prevalent at birth. Although molecular mechanisms of teratogens are poorly understood, rodent models indicate that
many embryotoxins are proteratogens that are activated in vivo by enzymes including P450 cytochromes and peroxidases to electrophiles or
free radicals that may damage DNA directly or indirectly through the formation of reactive oxygen species such as hydroxyl radicals. The embryo
may be vulnerable to reactive intermediates because of immature detoxification systems. See later chapters for further discussion of potential environmental causes of birth defects.
Nervous System
The adult brain, a complex network of about 1011 neurons and 1014 synaptic connections, has a high metabolic rate, consumes about a fifth of the
body’s oxygen uptake, and is almost entirely dependent on glucose for
energy. The development of the nervous system from the embryonic
through the adolescent periods depends on genes and chemical messengers that guide a complex series of processes that occur at specific points
in time and space. Development proceeds faster in some brain regions
than others; for example, the growth rates of the human diencephalon
and cerebellum, respectively, peak at birth and age 7 months. Although
the neuronal population is complete by age 2 years, synapse formation
and apoptosis continue until about age 5 years, and myelination continues through childhood and adolescence.
Periods of vulnerability during nervous system development include
(Rice and Barone, 2000)
• Neural tube closure during early gestation
• Neuron proliferation, migration, synaptogenesis, gliogenesis, myelinogenesis, and apoptosis during gestation and infancy
• Brain remodeling during adolescence
Environmental Threats to Child Health: Overview
9
The vulnerability of the developing brain to neurotoxins depends on access of the active agent to the nervous system and the timing of exposure
in relation to developmental changes. The so-called blood–brain barrier
is not fully developed until about age 6 months and, even then, it only
partially protects the brain from environmental toxicants, especially lipidsoluble agents. Perinatal exposure to neurotoxins can disrupt subsequent
cascades of developmental processes, greatly amplifying adverse effects,
but later exposures may have little or no effect. Radiotherapy of brain tumors before age 4 years disrupts neuron proliferation and synapse formation and causes substantial cognitive deficits; treatment at age 4 to 7
years or later, respectively, causes mild or no detectable cognitive deficits.
Until the 1970s, concern about the impact of neurotoxins such as lead,
mercury, and alcohol was almost entirely limited to adults. Frank mercury poisoning among infants (“pink disease” or acrodynia) was once
thought to be an infectious disease; use of mercurous chloride in teething
powder was not recognized as the actual cause until 1947, partly because
the clinical signs differed from those of adult mercury poisoning. Perinatal exposure to methylmercury in Iraq and Japan during the 1950s and
1960s caused severe neurobehavioral deficits and deaths among offspring
at exposure levels that caused minimal or no maternal toxicity. The use
of lead in gasoline, paint, and other products caused widespread exposure of children and adverse effects ranging from subtle neurobehavioral
deficits to severe and occasionally fatal childhood poisonings during
much of the twentieth century. During the 1950s and 1960s, many newborn infants were washed daily with a 3% suspension of hexachlorophene;
this practice was discontinued after discovery of a link to vacuolar encephalopathy of the brainstem reticular formation in preterm infants.
Preterm human infants and young rats are far more susceptible than adults
are to myelin degeneration caused by dermally absorbed hexachlorophene,
a lipid-soluble substance with a very high affinity for myelin.
Animal studies have identified neurotoxic mechanisms relevant to humans. Ethanol and certain drugs (e.g., barbiturates) interfere with neurotransmitter activity at N-methyl-D-aspartate (NMDA) and ␥-aminobutyric
acid type A (GABAA) receptors, the most ubiquitous receptor systems in
the developing brain; exposure of rodents to such neurotoxicants during
the neonatal brain growth spurt period causes widespread apoptosis of
developing neurons. Neonatal exposure of rodents to pesticides that target neurotransmitter systems (e.g., chlorpyrifos) also disrupts brain development processes and triggers apoptosis (see Chapter 7, Pesticides).
Further research is needed to assess the known and hypothesized links of
environmental neurotoxins to schizophrenia, dyslexia, epilepsy, autism,
mental retardation, attention deficit hyperactivity disorder (ADHD), learning disorders, and adult neurologic diseases.
10
CHILD HEALTH
AND THE
ENVIRONMENT
Immune System
Known or suspected environmental immunosuppressants in humans
include ultraviolet light (inhibits natural killer cell activity and contact
hypersensitivity in adults), high-dose ionizing radiation, and 2,3,7,8tetrachlorodibenzo--dioxin (TCDD). Rodent studies have shown that
immune system development can be disrupted by perinatal exposure to
relatively low doses of various toxicants (e.g., dioxin or dioxin-like
organochlorines, polycyclic aromatic hydrocarbons, certain pesticides,
heavy metals), with resultant persistent immunosuppression (Holladay
and Smialowicz, 2000). These toxicants may interfere with hematopoietic
cell proliferation, differentiation, and migration, expansion of lineagecommitted stem cells, colonization of postnatal lymphopoietic compartments, cell–cell interactions, and maturation to immunocompetence.
There is limited evidence that perinatal exposure of genetically predisposed rodents to immunotoxicants increases the risk of hypersensitivity
responses and autoimmune diseases but little evidence from human studies. Adult humans exposed to contaminants in cooking oil and tryptophan supplements developed autoimmune connective tissue disorders,
but the role of perinatal environmental toxicant exposures in autoimmune
disease in humans is unknown.
Respiratory System
Development of the human respiratory system involves the differentiation, proliferation, and organization of multiple cell types into a complex
system with over 300 million alveoli, the terminal gas exchange sacs
(Pinkerton and Joad, 2000). Lung development begins at about 4 weeks’
gestation, but alveolarization does not occur until the third trimester and
the number of alveoli in a newborn’s lungs is only 20% that of adults.
Airway outgrowth, branching, and alveolarization continue until about
age 18–20 years under the control of substances such as epidermal growth
factor, transforming growth factor-␣, and retinoic acid. Factors contributing to the susceptibility of the developing human respiratory system to
environmental toxicants include the following:
• Several lung enzyme systems responsible for detoxification of xenobiotics are not fully developed at birth.
• Postnatal lung growth and development continues from birth until late
adolescence, a period of 16–18 years during which children are exposed
to airborne toxicants and aeroallergens.
• Polymorphisms in any of several candidate genes may increase susceptibility to asthma.
Environmental Threats to Child Health: Overview
11
Perinatal exposure to ETS is associated with lung function growth deficits
and incident asthma. The infant lung appears to be susceptible to idiopathic pulmonary hemosiderosis caused by combined exposure to the toxicogenic fungus Stachybotrys chartarum and ETS. Spores of S. chartarum are
respirable and slowly release toxins that cause capillary fragility and suppress immune function; the ability of fungal toxins to inhibit protein synthesis in rapidly growing lungs may partially explain the susceptibility
of infants to this disease.
Reproductive System
Experimental animal studies revealed vulnerable periods of exposure during reproductive system development including
• Spermatogenesis—preconceptual exposure of males to genotoxins can
damage sperm DNA and cause early embryo death and birth defects
• Male reproductive development
° Prenatal exposure of the male rat to androgen receptor antagonists
(e.g., the pesticides vinclozolin, procymidone, linuron, and dichlorodiphenyltrichloroethane [DDT]) causes reduced anogenital distance
and induces areolas at relatively low doses, hypospadias, agenesis of
reproductive accessory tissues, and retained nipples at intermediate
doses, and undescended testes and epididymal agenesis at high
doses.
° Immature and pubertal rats appear to be more sensitive than
adults to testicular toxicity of phthalate esters and the pesticide, 1,2dibromo-3-chloropropane.
• Ovarian development—neonatal exposure of female rats to androgens
causes delayed puberty, irregular ovarian cycles, lower numbers of
ovarian follicles, and premature cessation of ovulation.
• Puberty—exposure of experimental animals to certain neurotoxins
(heavy metals, solvents, or pesticides) may accelerate or delay puberty.
Exposure
Behavior and Diet
Breast milk is a potentially important source of polychlorinated biphenyls
(PCBs) and other fat-soluble contaminants for infants, especially those
whose mothers consumed large amounts of contaminated fish or other
foods. Infants and toddlers frequently mouthe or lick objects or surfaces;
young children showed about 10 hand–mouth contacts per hour when
videotaped while playing. Children often sit on floors or grass/soil while
watching television, playing, or eating snacks, thus being exposed to tox-
CHILD HEALTH
12
AND THE
ENVIRONMENT
icants in house dust, carpets, and soil via skin contact, ingestion, or inhalation. Compared to adults, 1-year-old infants consume (per unit body
weight per day) twice as much tap water, total vegetables, and total citrus fruits and 10–20 times as many pears, apples, and total dairy products (Table 1–3); children aged 3–5 years consume 2–3 times as much tap
water, total vegetables, and total citrus fruits and 7–8 times as many apples and total dairy products. These habits increase the risk of exposure
to pesticide residues on citrus fruits and vegetables and to fat-soluble
organochlorine compounds in dairy products.
Genetic Characteristics
Certain genetic traits interact with infectious, chemical, physical, nutritional, and other factors to cause adverse health effects. Diseases caused
by single gene mutations may be aggravated by environmental contaminants; for example, cystic fibrosis is exacerbated by ETS. Genetic factors
are also important in relatively common childhood conditions and diseases including birth defects, cancer, and asthma. Polymorphisms involve
two or more distinct alleles at one genetic locus at stable frequencies in
the population too large (usually defined as 1%) to be explained solely
by recurrent mutation—the average heterozygosity per nucleotide site in
humans is about 1:1000. Although persons with polymorphisms usually
have no obvious health problems, they may be more susceptible to environmental and other hazards (see the examples in succeeding chapters).
As the function of the human genome becomes better understood, the
TABLE 1–3. Ratio of Childhood to Adult Intakes (Amounts per
Kilogram of Body Weight per Day) of Air, Water, and Selected Foods
Substance
Total tap water
Air—inhalation rates at rest
Total vegetables
Citrus fruits
Apples
Bananas
Peaches
Pears
Peas
Tomatoes
Total meats
Total dairy products
Age 1 Year
Age 3–5 Years
2.1
3.4
1.8
2.2
14.2
6.0
9.5
20.7
3.5
1.7
1.7
20.3
2.4
2.8
1.9
3.0
8.4
2.1
3.1
2.3
2.4
2.5
2.3
6.8
Source: U.S. Environmental Protection Agency (1997).
Environmental Threats to Child Health: Overview
13
role of polymorphisms in susceptibility to environmental hazards will
likely have major implications for disease prevention and control policies
and programs. Existing environmental standards invoke uncertainty factors intended to protect susceptible subgroups; improved knowledge may
show that current standards are inadequately protective of susceptible
subgroups.
Physiology
Certain physiologic and metabolic characteristics during prenatal and
postnatal development may increase the risk of adverse health effects
from environmental toxicants. Compared to an adult, an infant has about
twice the surface area per unit body weight and a correspondingly higher
metabolic rate, a threefold higher intake of air per unit body weight per
day, and an immature blood–brain barrier. The term blood–brain barrier encompasses multiple mechanisms that control access of blood components
to the brain; fetal and neonatal blood–brain barriers are relatively impermeable to protein but are more permeable than adult barriers to small
lipophilic molecules such as unconjugated bilirubin.
Metabolism
To varying degrees, toxicants absorbed into the body are detoxified in the
liver, kidneys, and other tissues by xenobiotic1 metabolizing systems. The
metabolism of many lipophilic xenobiotics involves two phases: phase I—
mainly oxidative reactions and phase II—conjugation with water-soluble
moieties, a step that facilitates excretion. Phase I involves mixed-function
oxidases (P450 cytochromes) that can (1) inactivate xenobiotics to less toxic
derivatives amenable to conjugation and excretion or (2) activate them to
strong electrophiles or unstable compounds that generate highly reactive
free radicals. Phase I products may be conjugated during phase II with
glucuronide, sulfate, acetate, glutathione, or other conjugating agents, reactions catalyzed by specific enzymes (e.g., glucuronyl transferase,
N-acetyltransferase, glutathione S-transferase). Glutathione also scavenges electrophilic xenobiotics, thus protecting RNA, DNA, and other potential targets.
Humans display wide variations in susceptibility to xenobiotics, in
part due to genetic polymorphisms in P450 cytochromes and other phase
I and II enzymes. The genes that encode cytochromes are divided into
1A
xenobiotic is any chemical not produced in vivo.
14
CHILD HEALTH
AND THE
ENVIRONMENT
families and subfamilies, each with some degree of specificity for certain
xenobiotics, including (1) CYP1—the CYP1A subfamily includes CYP1A1
(encodes a cytochrome active in metabolism of benzo(a)pyrene; occurs in
the liver, lung, and kidney), CYP1A2 (encodes a cytochrome that metabolizes acetanilide; found mainly in the liver) and (2) CYP2, which includes
several subfamilies—CYP2A1 and CYP2A2 encode cytochromes that are
active in metabolism of sex steroids (testosterone, progesterone, and androstenedione) and xenobiotics; CYP2E1 is restricted to mammals and appears to encode a cytochrome that activates benzene, nitrosamines, and
certain other xenobiotics, thereby contributing to their carcinogenicity.
Children with polymorphisms for CYP1A1, CYP2E1, and other xenobiotic metabolizing enzymes appear to have an increased risk of developing acute lymphoblastic leukemia (Infante-Rivard et al., 1999; Krajinovic
et al., 2002).
Immature Detoxification Systems
Pharmacokinetic studies of drugs used to treat newborn infants indicate
that they can metabolize xenobiotics, but clearance is generally slow (Gow
et al., 2001). Liver enzymes develop at different rates postnatally; for example, levels of glycine N-acyltransferase, involved in detoxification of
drugs and other xenobiotics, are very low at birth and do not reach adult
levels until about age 18 months. In a population exposed to air pollution, levels of polycyclic aromatic hydrocarbon (PAH)-DNA adducts,
total aromatic-DNA adducts, and cotinine in cord blood were higher than
those in maternal blood, suggesting reduced fetal detoxification capacity
(especially since maternal exposure to PAHs exceeds fetal exposure)
(Whyatt et al., 2001).
Risk Management Issues
Historical Perspective
The former belief that the placenta protects the fetus from toxic chemicals was shattered by repeated events during the mid-twentieth century
involving serious and sometimes fatal effects of prenatal maternal exposures including ionizing radiation, methylmercury, diethylstilbestrol
(DES), and thalidomide. Thalidomide, an antinausea drug once widely
prescribed during pregnancy, caused severe birth defects in about 7000
infants during 1958–1962. This experience stimulated developmental tox-
Environmental Threats to Child Health: Overview
15
icity testing of new commercial chemicals and birth defect monitoring in
many countries, and it showed that:
• A chemical that was virtually nontoxic in mice and adult humans
caused a markedly increased risk of severe birth defects when consumed during pregnancy.
• New commercial chemicals should be adequately screened for developmental toxicity in diverse experimental animals before humans are
exposed (e.g., thalidomide is teratogenic in rabbits and primates, but
rodents are generally resistant).
• International premarketing regulatory practices varied widely. The
drug was available across the counter in Germany, the country with
the highest number of affected infants; because of case reports of peripheral neuropathy in adult users, the U.S. Food and Drug Administration (FDA) restricted its use to clinical trials, saving many infants
from devastating birth defects
Human Exposure Assessment
Children may be exposed to environmental contaminants in air, water,
soil, dust, and food by ingestion, inhalation, or dermal contact. The potential anthropogenic sources of environmental contaminants include fossil fuel combustion, manufacturing processes, various uses of commercial
products (pesticides, building materials, solvents), human activities (e.g.,
smoking indoors), waste disposal (hazardous waste disposal sites, incineration), and accidents. One of the major drivers is the vast and rapidly
growing number of commercial chemicals. Over 70,000 commercial chemicals are registered for use in the United States, and the EPA receives about
1500 petitions annually to approve new chemicals or new uses of existing chemicals (U.S. Environmental Protection Agency, 2001a).
Few countries have assessed children’s exposures to environmental
contaminants through population-based biomonitoring. The United
States has included children in the U.S. National Health and Nutrition
Examination Survey (NHANES) surveys over the past 25 years and has
assessed their exposure to contaminants including lead, other metals, ETS,
phthalates, and organophosphate pesticides. The German Environmental
Survey has been conducted three times since 1985–1986 and has included
analyses of blood, urine, and scalp hair samples from children aged 6–
14 years and environmental samples (house dust, drinking water, indoor
and personal air, food) for metals, volatile organic chemicals (VOCs), and
organochlorine compounds.
16
CHILD HEALTH
AND THE
ENVIRONMENT
Recent Progress
Although much remains to be done to reduce children’s exposures to
known environmental hazards and to define the links between environmental factors and child health, there have been significant achievements:
• Blood lead levels declined sharply among children in all sex, age,
ethnic, and income subgroups immediately after the introduction of
lead-free gasoline in 1976.
• Population exposure to ETS has been reduced; median serum cotinine
levels among nonsmokers in the United States decreased from 0.2 to
0.05 ng/mL between 1988–1991 and 1999.
• The U.S. 1996 Food Quality Protection Act requires that the unique exposures and susceptibilities of children be considered in pesticide risk
assessment.
• Some countries have banned or restricted uses of a few pesticides
mainly to protect children, such as daminozide and chlorpyrifos.
• Breast milk concentrations of PCBs, DDT/DDE (dichlorodiphenyldichloroethylene), and several other organochlorine compounds have
decreased to varying degrees during recent decades.
Not all of these actions were targeted solely to children; the conversion
to lead-free gasoline followed the introduction of catalytic converters by
car manufacturers to reduce emissions under the 1970 Clean Air Act; lead
inactivated the platinum catalyst in catalytic converters, necessitating the
use of lead-free fuel.
Toxicity Testing of Commercial Chemicals
Volume. Regulatory agencies rely mainly on industry to conduct toxicity tests of new commercial chemicals. Extensive testing of pharmaceuticals, including clinical trials, has been required for many years and has
shown that children’s reactions often vary quantitatively and/or qualitatively from those of adults. The EPA estimated that up to a quarter of the
approximately 70,000 commercial chemicals have neurotoxic potential,
but only about 10% (excluding pharmaceuticals) have been tested for neurotoxicity. Most high production volume (HPV) chemicals (those produced in or imported into the United States in amounts of at least 500
tons per year) have not been subjected to the six basic toxicity screening
tests prescribed by the Organization for Economic Cooperation and Development (OECD) for its 30 member countries. The U.S. National Toxicology Program, established in 1978 to coordinate toxicity testing on behalf of federal agencies, biomedical science communities, and the public,
Environmental Threats to Child Health: Overview
17
is probably the largest chemical toxicity testing program in the world, but
even it can provide complete toxicologic evaluations on only 10–20 chemicals per year.
Developmental toxicity. In addition to the sheer number of commercial chemicals, routinely required toxicity tests may not detect important
developmental effects. Relatively little toxicity testing has been directed
to early embryogenesis, that is, the period between fertilization and gastrulation. In experimental animals, the effects of mutagens vary during
early embryogenesis (Rutledge, 1997):
• Germ cells and early multipotential embryonic cells—mutagens affect
many cells and organ systems, causing pre- and peri-implantation
deaths, balanced chromosomal translocations (causing sterility or reduced fertility of offspring), growth retardation, and moderately increased rates of structural anomalies (mainly anencephaly, cleft palate,
and skeletal anomalies).
• Zygote—exposure to mutagens at this stage causes early, middle, and
late gestation deaths, as well as high rates of a restricted range of structural anomalies associated with chromosomal breaks and other cytogenetic abnormalities (mainly skeletal, eye, and abdominal wall closure
defects, but no increase in anencephaly, spina bifida, or heart or urinary tract malformations).
• Preimplantation conceptus—mutagens cause growth deficits, high
rates of structural anomalies (anencephaly, skeletal abnormalities, cleft
palate), and late fetal death.
Nonmutagenic teratogens produce a spectrum of structural anomalies
specific to time periods as brief as 1–2 days. In a review of the mechanisms of developmental defects, the U.S. National Academy of Sciences
concluded that (National Academy of Sciences, 2000) (1) the mechanism
of developmental toxicity is partially understood for a few toxicants (e.g.,
retinoic acid, diethylstilbestrol, and TCDD) but is not completely known
for any compound; (2) testing should be done across the entire developmental period, including early fetal loss; and (3) there is potential to rapidly and inexpensively screen many chemicals for the ability to disrupt
signaling pathways central to normal development.
Neurotoxicity. The EPA-designated toxicants that require developmental neurotoxicity testing include (in descending order) central nervous system teratogens and their structural analogues, adult neurotoxins,
adult neuroactive agents, hormonally active compounds, and develop-
18
CHILD HEALTH
AND THE
ENVIRONMENT
mental toxicants. Much of the baseline evidence for triggers assumes that
developmental and reproductive toxicity studies have been done, but the
latter are required by the EPA only for registration of food-use pesticides
and not for other pesticides or commercial chemicals (Claudio et al., 2000).
Most of the 140 pesticides considered to be neurotoxic by the EPA have
not been tested for developmental neurotoxicity, indicating that historic
practices generally failed to trigger neurodevelopmental toxicity testing
of known adult neurotoxins. An EPA advisory group recommended routine developmental neurotoxicity testing for registration of all food-use
pesticides. Recommendations for neurotoxicity testing that address the
need to protect child health include (International Programme on Chemical Safety, 2001) the following:
• Develop standardized test methods and norms to evaluate neurotoxicity in infants and children.
• Increase testing in animals involving perinatal exposure to chemicals
and/or mixtures of chemicals to define the relative sensitivity of the
developing nervous system to neurotoxic injury.
• Develop and validate efficient animal tests for developmental neurotoxicity for use in international collaborative studies.
Carcinogenicity testing. The Carcinogenic Potency Database contains
the results of over 5000 carcinogenic bioassays on about 1300 chemical
entities (Gold, 2001); in other words, only a fraction of the commercial
chemicals have been tested for carcinogenicity in animals. An EPA review
of animal carcinogenesis bioassay studies relevant to the issue of perinatal exposures found that lifelong exposure of animals beginning neonatally often produces a higher incidence of tumors with shorter latencies
but seldom produces types of tumors not found in the standard bioassay;
perinatal exposure alone to three known carcinogens did not consistently
cause an increased incidence of cancer. The EPA review concluded that
there is insufficient evidence to justify routine inclusion of a perinatal exposure component in the standard carcinogenesis bioassay, a conclusion
endorsed by the EPA Scientific Advisory Panel. Notwithstanding this decision, carcinogens that are more potent or that cause unique types of cancer in rodents after perinatal exposure include cycasin (brain and jejunal
tumors), DES (vaginal cancer and uterine adenocarcinomas in exposed
females and testicular tumors among exposed males and the male offspring of perinatally exposed females), genistein (a natural phytoestrogen present in soy—uterine adenocarcinomas), and N-ethyl-N-nitrosourea
(nephroblastoma and brain gliomas).
Environmental Threats to Child Health: Overview
19
Toxicity testing priorities. The EPA Scientific Advisory Panel recommended that priorities for testing commercial chemicals for child health
hazards should be based on criteria such as potential for children’s exposures to exceed those of adults and should include those chemicals to
which children are uniquely exposed or for which children have unique
susceptibility rather than production volume alone. After excluding lowvolume chemicals (less than 5 tons/yr) and high molecular weight, poorly
absorbable polymers, there were about 15,000 chemicals produced or imported at levels above 5 tons/yr, including about 2800 HPV commercial
chemicals. Since 1979, about 540 of the chemicals in the 15,000 chemical
subset above have been tested within the EPA’s Existing Chemicals Testing Program; these were mainly HPV chemicals. In 1990, the EPA developed a Master Testing List of over 500 existing chemicals (i.e., already
used commercially) based on toxicity testing priorities of U.S. federal
agencies and the OECD; as of 2001, testing actions were underway for almost 300 entities.
Despite this progress, only about 7% of HPV chemicals have been
adequately tested; the remainder are missing one or more of the OECD
Screening Information Data Set (SIDS) tests, and 43% are missing all SIDS
tests (U.S. Environmental Protection Agency, 1998). Although the EPA is
pursuing toxicity testing of HPV chemicals through voluntary agreements, it can require chemical manufacturers and processors to test existing chemicals that pose unreasonable risks to human health or the environment. Among the 491 HPV commercial chemicals to which children
are likely exposed, only 25% have been adequately tested (U.S. Environmental Protection Agency, 2000). Under a voluntary EPA program, companies that manufacture or import 23 targeted chemicals are asked to
sponsor a three-tier toxicity-testing program.
Strengthening Prevention
The Precautionary Principle
Modern use of the precautionary principle in environmental health can
be traced to the 1992 United Nations Conference on Environment and
Development that promulgated the Rio Declaration. Principle 15 of the
Declaration states, “In order to protect the environment, the precautionary approach shall be widely applied by States according to their capabilities. Where there are threats of serious or irreversible damage, lack of
full scientific certainty shall not be used as a reason for postponing costeffective measures to prevent environmental degradation” (United Nations, 1992). Concern about the slow pace of efforts to address climate
20
CHILD HEALTH
AND THE
ENVIRONMENT
change, ecosystem degradation, and resource depletion was a major
driver of the Rio Declaration.
Policymakers generally encounter a high level of uncertainty about
children’s environmental health risks because of knowledge gaps concerning relevant exposures and dose–response relationships for individual toxicants and mixtures. In the face of such uncertainty, a requirement
for scientific consensus on causality is not necessarily appropriate for
management of children’s environmental health risks. As stated elsewhere, “If exposure is widespread and the consequences [are] serious, a
need for primary prevention may suggest that even a moderate degree
of evidence justifies regulatory action. On the other hand, if the probability of human exposure is low and the adverse health effects [are] uncertain, then the best policy may be collection of improved data” (HertzPicciotto, 1995). Scientific uncertainty about child health and the
environment relates to several factors:
• Absent or inadequate evidence—gaps in knowledge of the toxicology
and epidemiology of potential environmental hazards
• Inconsistent results of toxicologic studies—use of different test animal
species or strains, differences in purity, dose, and route of exposure of
the test substance, small sample size
• Inconsistent results of epidemiologic studies—limited statistical power,
inadequate exposure estimates, uncontrolled confounders
• Uncertainty about the shape of the dose–response curve at doses below those observed
• Doubt about the adequacy of uncertainty factors used in quantitative
risk assessments to compensate for knowledge gaps including extrapolation of results from animal studies to humans and the distribution
of exposures and susceptibility factors among children in the general
population
To obtain an improved evidence base for future policy and program development in this field, current efforts in two areas require strengthening:
(1) research to better define environmental hazards, susceptible populations, and dose–response relationships and (2) tracking systems to monitor population exposure levels. The EPA and the National Institute of
Environmental Health Sciences (NIEHS) have funded 12 children’s environmental health research centers to address priority issues including
child health effects of various toxicants (lead, mercury, PCBs, pesticides)
and the role of environmental exposures in cognitive deficits, autism,
learning disabilities, ADHD, and asthma (U.S. Environmental Protection
Agency 2002).
Environmental Threats to Child Health: Overview
21
Tracking Systems
Systems that track the occurrence of health conditions and their determinants have greatly aided prioritization, planning, and evaluation of
disease control and prevention programs in specific fields (infectious diseases, unintentional injuries, disability, occupation-related diseases) and
in nationwide health objectives (see, e.g., Thacker and Stroup, 1994). A
framework for environmental tracking systems might include hazards,
exposures, health outcomes, and children’s environmental health policies.
Tracking systems for child-related aspects of outdoor air pollution, for instance, might address hazard occurrence (e.g., concentrations of priority
contaminants in outdoor air), exposure levels (e.g., proportion of children
exposed to ambient air contaminant levels above current air quality standards), health outcomes (e.g., incidence of emergency room visits for
childhood asthma), and health policies (e.g., clean air policies, standards,
guidelines). It appears that no jurisdiction has implemented comprehensive tracking systems to monitor needs and progress related to children’s
environmental health, but some elements exist:
• Hazards
° Outdoor air—many developed countries have national and/or regional monitoring networks to measure ambient air contaminants including particulate matter and ozone.
° Drinking water—drinking water facilities in developed countries
generally monitor microbial contaminants and may monitor chemical contaminants such as chlorination disinfection by-products (total
trihalomethanes) and lead.
° Indoor air—some national health surveys and many occasional surveys have included questions on the smoking habits of household
members; census data may include information on housing characteristics such as fuels used for cooking and heating.
° Food contaminants—some countries have conducted limited sampling of foods for pesticide residues and other contaminants; the
enormous variety and volume of foods and the practice of testing
batches rather than individual portions preclude detection of lowfrequency, high-pesticide residue levels before distribution and consumption. This issue is further complicated by sparse or absent data
on food consumption patterns within narrow age ranges among children (needed because of the large variation in diet as an infant or
child ages).
• Exposures and doses—exposure generally refers to the level of a contaminant in environmental media (air, water, food, soil) to which a person is directly exposed, while the internal dose is the amount of con-
22
CHILD HEALTH
AND THE
ENVIRONMENT
taminant taken up by an individual from all sources and is usually estimated by measuring contaminant levels in body tissues or fluids.
Tracking systems relevant to estimation of children’s environmental
hazard exposures or doses include
° Surveys of children and other household members including activities, smoking habits, pesticide use, and diet
° Biomonitoring—surveys of children and mothers that collect biologic
samples such as blood, urine, breast milk, and hair and analyze selected contaminant levels
• There have been few population surveys of children’s diet and very
limited tracking of children’s doses of environmental contaminants, the
recent effort of the U.S. Centers for Disease Control and Prevention being a notable exception (Centers for Disease Control and Prevention,
2001). Although a few countries have conducted breast milk contaminant monitoring for over 20 years, most monitoring has been confined
to sporadic small samples of convenience (Hooper, 1999).
• Health outcomes—developed countries generally collect data on vital
statistics (stillbirths, low birth weight, birth defects), hospitalizations/
physician services (e.g., spontaneous abortion, birth defects, asthma,
acute respiratory infections, gastrointestinal infections), and health status (e.g., health surveys that include items on parent-reported child
health conditions such as physician-diagnosed asthma).
• Children’s environmental health policies—some nongovernmental organizations track policies relevant to children’s environmental health,
but little organized information is published in the scientific literature.
This is a promising area for coordinated, ongoing tracking and communications to better inform the public about unmet needs.
Conclusion
Children’s environmental health is increasingly recognized as a global
public health issue of great importance. International, national, and
other bodies have identified asthma, air pollution (indoor and outdoor), lead, pesticides, water contaminants (chemical and microbial),
climate change, HAAs, and ETS as important children’s environmental health issues. Given our current limited knowledge, assumptions
about thresholds for noncarcinogenic effects of certain perinatal toxicant exposures may prove inappropriate; for example, the apparent
threshold for lead toxicity has decreased substantially during recent
decades because of improved research methods. Knowledge is also
Environmental Threats to Child Health: Overview
23
very limited concerning the proportions of adverse health outcomes
during pregnancy, childhood, and adulthood that are attributable to
prenatal and childhood environmental exposures. Answers to this
question will require substantial targeted epidemiologic and toxicologic research to measure relative risks and track environmental contaminant exposures.
The developing fetus and child are unusually susceptible to environmental hazards because of unique growth and developmental processes, immature metabolic systems, physiology, and behaviors. In developed countries, regulatory agencies have recently begun to adjust their
risk assessment processes to better protect children. Children remain,
however, continually exposed to multiple low-level environmental contaminants beginning in utero. The recent nascent efforts to undertake large
longitudinal research studies of children with a major focus on assessment of exposures beginning in early gestation and detection of neurobehavioral and other adverse health outcomes should be encouraged
and expanded.
The reader is referred to these selected websites for further information on children’s environmental health:
• World Health Organization, The Gateway to Children’s Environmental Health. Available at http://www.who.int/peh/ceh/index.htm
• U.S. Environmental Protection Agency, Office of Children’s Health
Protection. Available at http://yosemite.epa.gov/ochp/ochpweb.nsf/
homepage
• U.S. National Institute of Child Health and Development, The Longitudinal Cohort Study of Environmental Effects on Child Health and
Development. Available at http://nationalchildrensstudy.gov/
References
Anderson RN. (2001). Deaths: Leading causes for 1999. National Vital Statistics
Reports 49:88 pp.
Centers for Disease Control and Prevention. (2000). Strategic plan 1999–2003. Atlanta: Centers for Disease Control and Prevention.
Centers for Disease Control and Prevention. (2001). National report on human exposure to environmental chemicals. Atlanta: Centers for Disease Control and
Prevention.
Claudio L, Kwa WC, Russell AL, Wallinga D. (2000). Testing methods for developmental neurotoxicity of environmental chemicals. Toxicol Appl Pharmacol
164:1–14.
24
CHILD HEALTH
AND THE
ENVIRONMENT
Commission for Environmental Cooperation. (2000). Background paper for the
Symposium on North American Children’s Health and the Environment,
Toronto, Canada, 10 May 2000. Montreal: Commission for Environmental Cooperation.
Council of State and Territorial Epidemiologists. (2001). Development of environmental public health indicators. Located at http://www.cste.org/ps/2001/
2001-env-02.htm.
European Centre for Environment and Health, WHO. (1999). Children’s health
and the environment (EUR/ICP/EHCO 02 02 05/16). Rome: European Centre for Environment and Health.
G8 countries. (1998). 1997 Declaration of the environment leaders of the Eight on
children’s environmental health. Can J Public Health 89(Suppl 1):S5–S8.
Gold LS. (2001). The carcinogenic potency project. Located at http://potency.
berkeley.edu/cpdb.html.
Golding J, Pembrey M, Jones R. (2001). ALSPAC—the Avon Longitudinal Study
of Parents and Children. I. Study methodology. Paediatr Perinat Epidemiol
15:74–87.
Gow PJ, Ghabrial H, Smallwood RA, Morgan DJ, Ching MS. (2001). Neonatal hepatic drug elimination. Pharmacol Toxicol 88:3–15.
Hertz-Picciotto I. (1995). Epidemiology and quantitative risk assessment: a bridge
from science to policy. Am J Public Health 85:484–91.
Holladay SD, Smialowicz RJ. (2000). Development of the murine and human immune system: differential effects of immunotoxicants depend on time of exposure. Environ Health Perspect 108(Suppl 3):463–73.
Hooper K. (1999). Breast milk monitoring programs (BMMPs): world-wide early
warning system for polyhalogenated POPs and for targeting studies in children’s environmental health. Environ Health Perspect 107:429–30.
Hoyert DL, Arias E, Smith BL, Murphy SL, Kochanek KD. (2001). Deaths: final
data for 1999. National Vital Statistics Rep 49:114 pp.
Infante-Rivard C, Labuda D, Krajinovic M, Sinnett D. (1999). Risk of childhood
leukemia associated with exposure to pesticides and with gene polymorphisms. Epidemiology 10:481–7.
International Programme on Chemical Safety. (2001). Environmental health criteria No. 223. Integrated approach to the assessment of neurotoxicity of chemicals. Geneva: International Programme on Chemical Safety.
Krajinovic M, Sinnett H, Richer C, Labuda D, Sinnett D. (2002). Role of NQO1,
MPO and CYP2E1 genetic polymorphisms in the susceptibility to childhood
acute lymphoblastic leukemia. Int J Cancer 97:230–6.
Landrigan PJ, Schechter CB, Lipton JM, Fahs MC, Schwartz J. (2002). Environmental pollutants and disease in American children: estimates of morbidity,
mortality, and costs for lead poisoning, asthma, cancer, and developmental disabilities. Environ Health Perspect 110:721–8.
Liu S, Wen SW, Mao Y, Mery L, Rouleau J. (1999). Birth cohort effects underlying
the increasing testicular cancer incidence in Canada. Can J Public Health 90:
176–80.
Lvovsky K. (2001). Environment strategy papers, No. 1. Health and environment.
Washington, D.C.: The World Bank.
National Academy of Sciences. (1993). Pesticides in the diets of infants and children. Washington, D.C.: National Academy Press.
Environmental Threats to Child Health: Overview
25
National Academy of Sciences. (2000). Scientific frontiers in developmental toxicology and risk assessment. Washington, D.C.: National Academy Press.
National Institute of Child Health and Human Development. (2002). The National
Children’s Study. Located at http://nationalchildrensstudy.gov/about/
042002.cfm.
Pan American Center for Sanitary Engineering and Environmental Science. (2001).
Children’s health and the environment. Main themes. Located at http://www.
cepis.ops-oms.org/indexeng.html.
Pinkerton KE, Joad JP. (2000). The mammalian respiratory system and critical windows of exposure for children’s health. Environ Health Perspect 108(Suppl 3):
457–62.
Popovic JR. (2001). 1999 national hospital discharge survey: annual summary with
detailed diagnosis and procedure data. Vital Health Stat 13:214 pp.
Rice D, Barone S Jr. (2000). Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models. Environ Health Perspect 108(Suppl 3):511–33.
Rutledge JC. (1997). Developmental toxicity induced during early stages of mammalian embryogenesis. Mutat Res 396:113–27.
Shivapurkar N, Harada K, Reddy J, Scheuermann RH, Xu Y, McKenna RW, Milchgrub S, Kroft SH, Feng Z, Gazdar AF. (2002). Presence of simian virus 40 DNA
sequences in human lymphomas. Lancet 359:851–2.
Smith MA, Freidlin B, Ries LA, Simon R. (1998). Trends in reported incidence of
primary malignant brain tumors in children in the United States. J Natl Cancer Inst 90:1269–77.
Thacker SB, Stroup DF. (1994). Future directions for comprehensive public health
surveillance and health information systems in the United States. Am J Epidemiol 140:383–97.
United Nations. (1992). Report of the United Nations Conference on Environment
and Development, Rio de Janeiro, 1992, Annex 1. Rio declaration on environment and development. New York: United Nations.
U.S. Department of Health and Human Services. (2000). Healthy people 2010 (S/N
017-001-00547-9). Washington, D.C.: U.S. Government Printing Office.
U.S. Environmental Protection Agency. (1996). Environmental health threats to
children (EPA 175-F-96-001). Washington, D.C.
U.S. Environmental Protection Agency. (1997). Exposure factors handbook
(EPA/600/P-95/002Fa). Washington, D.C.
U.S. Environmental Protection Agency. (1998). Chemical hazard data availability
study. Located at http://www.epa.gov/opptintr/chemtest/hazchem.pdf.
U.S. Environmental Protection Agency. (2000). Voluntary children’s chemical evaluation program. Located at http://www.epa.gov/chemrtk/vccep/childhlt.
htm.
U.S. Environmental Protection Agency. (2001a). Chemical testing and information
home page. Located at http://www.epa.gov/opptintr/chemtest/index.htm.
U.S. Environmental Protection Agency. (2002). Centers for Children’s Environmental Health and Disease Prevention Research. Located at http://es.epa.
gov/ncer/centers/cecehdpr/01/.
Ventura SJ, Martin JA, Curtin SC, Menacker F, Hamilton BE. (2001). National Vital Statistics Reports 49:99 pp.
Ventura SJ, Mosher WD, Curtin SC, Abma JC, Henshaw S. (1999). Highlights of
26
CHILD HEALTH
AND THE
ENVIRONMENT
trends in pregnancies and pregnancy rates by outcome: estimates for the
United States, 1976–96. Natl Vital Stat Rep 47:1–9.
Vilchez RA, Madden CR, Kozinetz CA, Halvorson SJ, White ZS, Jorgensen JL,
Finch CJ, Butel JS. (2002). Association between simian virus 40 and nonHodgkin lymphoma. Lancet 359:817–23.
Whyatt RM, Jedrychowski W, Hemminki K, Santella RM, Tsai WY, Yang K, Perera FP. (2001). Biomarkers of polycyclic aromatic hydrocarbon-DNA damage
and cigarette smoke exposures in paired maternal and newborn blood samples as a measure of differential susceptibility. Cancer Epidemiol Biomarkers
Prev 10:581–8.
World Health Organization. (2001). The gateway to children’s environmental health.
Setting priorities. Located at http://www.who.int/peh/ceh/taskforce.htm.
2
Environmental Epidemiology
Epidemiology is the study of the distribution and determinants of health
conditions in human populations as a basis for preventive and other interventions. Unique to environmental epidemiology is its focus on environmental factors to which humans are unwittingly exposed. The scope
of environmental epidemiology addressed in this book includes studies
of adverse health outcomes in human populations and their relationship
to prenatal and childhood exposures to environmental hazards. The purpose of this chapter is to describe the types of environmental epidemiologic studies, their strengths and weaknesses, their role in describing population distributions of exposures and health outcomes, and their use to
generate and test hypotheses about environmental threats to child health.
The discussion includes important issues underlying the limited ability
of epidemiologic studies to identify hazards when population exposures
are low and health outcomes are subtle or delayed.
Epidemiology
Strengths
The major strength of epidemiologic studies is their ability to assess relationships between environmental exposures and health outcomes directly
27
28
CHILD HEALTH
AND THE
ENVIRONMENT
in humans under real-life conditions. This avoids uncertainties related to
extrapolations between species and from the high doses used in animal
studies to the generally lower doses observed in humans. Other strengths
include the ability to
• Evaluate risks in large populations with diverse exposures and genetic
traits
• Exploit administrative databases with high-quality diagnostic information on many adverse health outcomes
• Assess interactions between environmental exposures and other factors
unique to humans, such as alcohol, ETS, and genetic factors
Limitations
Timeliness
Epidemiologic studies can identify hazards only after humans have already been exposed and have developed adverse health effects. Toxicants
that cause delayed health effects such as cancer may be recognized only
after populations have been exposed for many years.
Nonexperimental Design
Environmental epidemiology usually involves observational studies of
humans inadvertently exposed to environmental toxicants (see the exceptions noted below). Observational studies are susceptible to biases related to selection of subjects, inadequate control of potential confounders,
and measurement of exposures and health outcomes. Scientific consensus on the significance of observational study results may be elusive and
may delay decisions to act on potential environmental hazards, especially
if the biologic mechanisms are unknown (see discussion of the precautionary principle near the end of Chapter 1).
Exposure Assessment
Exposure assessment is as a major obstacle in environmental epidemiologic studies, and the absence of quantitative exposure data is the main
reason for not using epidemiologic studies in quantitative risk assessments (Rothman, 1993). Exposure indices used in epidemiologic studies
must be practical, affordable, and able to capture information on exposures that occur at low average levels but may fluctuate substantially over
time. Potential components of exposure indices include dose (intensity,
pattern), timing (frequency, relationship to critical developmental periods, duration, time since first exposure), and exposure route (inhaled, ingested, dermal). When the biologic mechanism of an exposure–risk relationship is not understood, there is considerable uncertainty concerning
Environmental Epidemiology
29
the exposure index most likely to increase the risk of a given health outcome. This is an issue, for instance, in epidemiologic studies of powerfrequency magnetic fields and childhood cancer (see Chapter 9, Radiation). In such studies, exposure assessments have included time-weighted
average exposures above a threshold value, cumulative exposures, personal or residential exposure levels, proximity to power lines, and electric appliance use, but there is no biologic basis to define the best exposure indices.
Precision and Validity
Precision is the inverse of the variance of an estimate arising from random
measurement error, and increasing the sample size is the main method
used to improve precision. Measurement validity is the degree to which a
measurement method actually measures a given characteristic (e.g., how
well does motor vehicle traffic density reflect differences in ambient carbon monoxide exposure?). Internal and external validity, respectively, reflect the degree to which inferences about the study population and general populations are valid (Rothman and Greenland, 1998). Internal
validity implies accurate effect estimates (apart from random errors)
within a study; three main types of bias may reduce internal validity: selection, confounding, and information bias.
Selection bias. The relation between exposure and health outcome
varies between study participants and persons in the base population1
and may arise, for instance, through self-selection or diagnostic bias.
Confounding. The apparent relationship between an exposure and a
health outcome actually results from two or more effects, including the
exposure of interest. A confounder is a variable that can cause or prevent
the outcome of interest, is not an intermediate variable, and is associated
with the factor of interest in the base population. Methods used in epidemiologic studies to avoid or adjust for confounders include stratified
and multivariate analyses.
Information bias. This arises from misclassification, that is, errors in
measurement of exposures, other individual characteristics, or health outcomes. Differential misclassification occurs when measurement errors depend on the exposure or health outcome of interest; for example, in a casecontrol study of childhood cancer, the mothers of cases may put more
effort into recalling past exposures than the mothers of healthy controls,
1 Base
population refers to the population sampling frame from which the study sample was drawn.
30
CHILD HEALTH
AND THE
ENVIRONMENT
thereby creating spurious associations. Nondifferential misclassification
happens when measurement error is independent of exposure or health
outcome status. When nondifferential misclassification of a dichotomous
exposure occurs independent of other errors, the effect estimate is biased
toward the null, that is, the true association will be underestimated. Epidemiologic studies occasionally include direct estimates of misclassification; for example, parental recall bias for information on proximity to
power lines and exposure to prenatal X-rays was assessed in a casecontrol study of childhood leukemia (Infante-Rivard and Jacques, 2000).
There are circumstances under which nondifferential misclassification can
bias the effect estimate away from the null. For information on this and
other sources of bias, the reader is referred to other sources (Rothman,
2002; Rothman and Greenland, 1998).
Extrapolation of Dose–Response Relationships
Neither epidemiologic nor experimental animal studies can readily measure lifetime excess risks of interest to regulators, that is, 105 or 106. Results in the observed exposure range must be extrapolated to substantially lower exposures with no direct evidence of the shape of the dose–
response curve at such levels. The degree to which exposure metrics used
in epidemiologic studies correlate with actual doses is usually unknown,
contributing uncertainty to dose–response relationships and their extrapolation. In humans or animals, exposure to high doses of a toxicant may
cause health outcomes that mask low-dose effects; for example, high doses
of ionizing radiation can kill cells, while lower doses can cause cell transformation leading to cancer.
Other Issues
Ethical, cultural, privacy, response-burden, and related issues constrain
the design and conduct of epidemiologic investigations. Study protocols
are generally assessed for adherence to ethical principles including autonomy (respecting individual rights and freedoms), beneficence (doing
good), and nonmaleficence (doing no harm). In some cases, the ethics
principle of justice (fair and equitable distribution of resources) may be
assessed, for example, ensuring that high-risk disadvantaged groups are
included in a cohort or intervention study. Ethical issues increase with the
sensitivity of information collected and can arise at various points in etiologic epidemiologic studies including:
• Study design and conduct—selection of the study population, protocol
development, subject recruitment, solicitation of informed consent, and
adequacy of quality control and follow-up procedures
Environmental Epidemiology
31
• Interpretation and communication of results to individual participants
and others, including health authorities and the public
• Banked biologic specimens—stewardship issues include secure storage,
controlled access, and use for unanticipated research objectives not
communicated to subjects at the time informed consent is obtained
The increasing opportunity for epidemiologists to conduct research involving the linkage of computerized databases is countered by public concern about privacy. Study protocols must include methods to protect privacy and provide evidence of potential benefits of the proposed research.
The American College of Epidemiology noted deficiencies in existing
ethics guidelines for epidemiologists in the areas of education, policy, and
advocacy and called for dynamic guidelines that emphasize core values,
obligations, and virtues (Weed and Coughlin, 1999). Although the risk to
child health is minimal in observational epidemiologic studies, the age at
which children can give truly informed consent is not well defined; for
example, longitudinal studies need parent/guardian consent when the
child is young but may require shared or direct consent as the child ages.
Exposure Assessment
Epidemiologists assess exposure to environmental hazards by several
methods including (1) self-reported information captured with the aid
of standardized questionnaires, (2) proximity to sources of contaminants,
(3) environmental contaminant measurements coupled with information
on residential history or activities, and (4) biomarkers of the internal dose.
It is important to note that there may be relatively strong or very weak
correlations between contaminant levels in environmental media (air,
water, soil, dust, foods) and the dose actually taken up by a child or parent. The biologically relevant dose is generally the concentration of a contaminant or its biologically active derivative in target tissues, but peak or
cumulated doses over time may be etiologically important for delayed
health effects. Although some tissue contaminant levels can be measured
(e.g., lead in deciduous teeth), they provide only limited information on
exposure patterns, especially over long time periods; for instance, concentrations of organophosphate pesticide urinary metabolites reflect recent exposure.
Accurate assessment of exposure to environmental and other factors,
including potential confounders, is essential for detecting health effects
of low-level environmental contaminants. Bias can arise not only from
misclassification of the exposure of interest, but also from misclassification of confounders, potentially biasing the effect estimates toward or
32
CHILD HEALTH
AND THE
ENVIRONMENT
away from the null (Greenland, 1980). Validated exposure methods,
whether questionnaire, administrative record, or laboratory based, help
reduce misclassification of exposure. Other challenges include evaluation
of complex exposures (e.g., mixtures of toxicants with variable compositions and toxicity) and assessment of exposures during critical perinatal
time periods when susceptibility to developmental, reproductive, and
neurotoxic effects is high.
Estimation of Past Exposures
Epidemiologic studies dependent on recall of exposures are susceptible
to misclassified exposure status, thereby reducing statistical power and
increasing the likelihood of biased risk estimates. Valid exposure recall is
facilitated when exposure factors are well defined and relatively stable
over time (e.g., self-reported information on smoking habits, occupation,
and place of residence). Historic administrative records documenting information on environmental contaminant levels are not subject to recall
bias, but older data may be limited by subsequent changes in the scope,
methods, and sensitivity of contaminant analyses. For instance, chlorinated disinfection by-products in drinking water were not routinely measured until recent decades; even today, historic data are usually limited
to total trihalomethanes, compounds that represent only a few of the
several hundred disinfection by-products and only a fraction of the total
mutagenic activity in chlorinated water. Similarly, historic ambient air pollution data are generally limited to a few major contaminants; ambient
air VOCs, for instance, have been excluded from routine monitoring in
the United States except in certain centers with major summer ozone pollution problems.
Some epidemiologic studies have collected information on individual exposure for certain factors (e.g., ETS, residential history) but have
used geographically based data to assess exposure to environmental contaminants (e.g., data on ambient air quality and drinking water quality).
Individual exposure levels tend to be misclassified when group-based exposure indices are used because of individual differences in activity patterns, exposure rates, metabolism, and excretion (Hatch and Thomas,
1993). Personal dosimetry studies, for instance, have shown that ambient
environmental contaminant data explain only 2%–25% of personal exposures for most toxic and carcinogenic VOCs and pesticides (Wallace, 1993).
Environmental Contaminant Level and Internal Dose Estimation
Methods for assessing environmental exposures of children and parents
range from simple questionnaires to personal monitoring devices and lab-
Environmental Epidemiology
33
oratory analyses of contaminants in blood and other biologic samples. Potential methods and information sources include:
Environmental contaminant levels
• Administrative records
° Occupation—exposure information on workplace records may be
limited to job title but may include environmental exposure levels,
for example, for persons exposed to airborne toxicants in underground mines.
° Air and water quality—government agencies may collect data on air
and water quality; such data are generally limited to larger population centers even in economically advantaged countries.
• Self-reported information
° Young children—information may be collected from parents with the
aid of standardized questions on parental, child, and home characteristics; in panel studies, parents may keep daily diaries of children’s
activities.
° Older children—these children may complete questionnaires and/or
keep a daily diary of activities.
• Contaminant measurements
° Environmental contaminants may be measured in samples of media
including air, water, house dust, soil, and food, such as, house dust
mite antigens in dust samples from a child’s mattress.
° Personal sampling—children may be monitored using devices attached to the clothing to measure air contaminant levels, magnetic
field strength, and ultraviolet light intensity as the child moves
through various microenvironments over the course of a few hours,
a day, or several days.
• Models
° Environmental contaminant levels may be estimated by modeling of
contaminant emission data and supplementary information.
° Internal doses may be estimated by modeling of data on contaminant
levels in environmental media, consumption rates (air, water, food, soil,
dust), and activity patterns (e.g., percentage of time spent indoors).
Biomonitoring. Biomarkers of internal doses include contaminant levels (e.g., heavy metals, dioxins, PCBs) in blood, breast milk, urine, hair,
tissue, or other samples from children and/or their mothers. Certain molecular or cellular effects of toxicants are proxies for the internal dose,
such as, 4-aminodiphenyl hemoglobin adducts (tobacco smoke) and persistent chromosome translocations in peripheral lymphocytes (ionizing
34
CHILD HEALTH
AND THE
ENVIRONMENT
radiation). For infants and young children, in particular, minimally invasive samples and microanalytic techniques are essential (e.g., to enable
multiple contaminant analyses on finger prick or heel prick blood samples). Biomarkers of exposure and susceptibility will become more important in epidemiologic studies to the extent that they represent the biologically relevant dose and reduce misclassification of exposure status.
At present, however, many biomarkers reflect only recent exposures (e.g.,
urinary levels of organophosphate insecticide metabolites), a serious limitation for investigation of health outcomes with latent periods substantially longer than the tissue half-life of the toxicant(s) of interest.
Study Power
Power, defined as the likelihood that an epidemiologic study will demonstrate an association if one exists, depends on several factors including
the strength of the association, the observed number of health outcomes,
and the range and distribution of exposures in the study sample (Prentice and Thomas, 1993). Inadequate power is a major source of inconsistency between epidemiologic studies, especially when point estimates of
risk are compared. Random exposure measurement error and low exposure levels can greatly reduce study power. Differences between studies
in distribution of exposure levels can outweigh other factors as a cause
of heterogeneous effect measures of the same association; small studies
with high exposures may be more powerful than larger studies with lower
exposures (Hertz-Picciotto and Neutra, 1994).
Interactions between environmental and other factors are potentially
of great public health importance, but the power of epidemiologic studies to detect interactions is generally an order of magnitude less than their
power to detect main effects (Greenland, 1993). Constraints related to
power are not limited to epidemiologic studies; animal experiments compensate for small sample sizes by exposing animals to very high doses of
toxicants, raising uncertainty about exposure–risk relationships at low
doses. For a more complete coverage of design issues, the reader is referred to other sources (Elwood, 1998; Gordis, 2000; Rothman and Greenland, 1998).
Study Types
Research on environmental threats to child health has included case reports and ecologic, cross-sectional, case-control, cohort, and experimental epidemiologic studies.
Environmental Epidemiology
35
Descriptive Studies
Descriptive epidemiologic studies illustrate the population-based variation in the risk of health outcomes by person, place, and time by generating age-specific and age-standardized incidence, prevalence, or mortality rates by sex, health outcome, geographic area, and time period. When
the number of health outcome events in the study population is small,
analyses may include indirect age-standardized ratios of observed to expected events. Descriptive data serve to (1) assess needs for and progress
in disease prevention and control, (2) generate etiologic hypotheses,
(3) assess time trends, (4) identify emerging health problems, (5) raise
awareness of the burden of various health conditions, and (6) provide a
baseline for forecasting future disease burdens.
Many governmental and voluntary agencies produce and disseminate the results of descriptive studies and promote their use in development and evaluation of disease prevention and control policies and programs. For examples, the reader can access websites for surveillance of
birth defects (International Clearinghouse for Birth Defects 2002) and cancer (American Cancer Society 2002; Health Canada 2002). Other sources
for additional details on descriptive epidemiologic studies include Stroup
and Teutsch (1998) and Teutsch and Churchill (2000).
Acute effects such as pesticide poisonings may be attributed to a causal
agent based on well-documented physician reports of one or a few cases.
Children poisoned by organophosphate or carbamate insecticides, for instance, can be diagnosed through clinical signs and symptoms and reduced
plasma butyrylcholinesterase or red blood cell acetylcholinesterase levels.
Systematic reporting of such events through surveillance systems serves to
assess preventive programs.
Although case reports alone rarely provide evidence of cause–effect
relationships for delayed health effects, they are valuable for generating
hypotheses and stimulating investigations that are more definitive. The
causal role of soot in scrotal cancer among young chimney sweeps was
suspected 150 years before the identification of benzo(a)pyrene and other
carcinogens in soot (Pott, 1775). During the late nineteenth century, a
physician attributed childhood lead poisoning cases to hand–mouth behavior and residence in homes with lead-based paint (Gibson, 1892).
Cluster Investigation
A cluster is defined as an aggregation of relatively uncommon events in
space and/or time in amounts believed or perceived to be greater than
expected by chance. Generally, a cluster is reported to public health au-
36
CHILD HEALTH
AND THE
ENVIRONMENT
thorities when concerned persons perceive an apparently excessive incidence of conditions such as birth defects or childhood cancer in their residential area. There have been many reported clusters of childhood leukemia with hypothesized links to environmental hazards including
ionizing radiation and hazardous waste sites. The Centers for Disease
Control and Prevention (CDC) investigated 108 cancer clusters between
1961 and 1990 and did not find a clear cause for any of them (Caldwell,
1990). Although the likelihood of identifying a causal agent from investigation of a cluster is very low, timely investigation helps to reassure the
public that reasonable measures have been taken to detect potential
causes. Investigators generally confront a poorly defined base population
from which the cases arose, however, increasing the chance of overestimating the disease rate through boundary shrinkage (Olsen et al., 1996).
There is also potential for recall bias related to heightened awareness of
potential causes among case families.
Proposed criteria for identifying clusters where investigation is more
likely to be fruitful include (Neutra, 1990)
• Cases—there are at least five cases of a disease that has known causes,
or the cluster is an acute condition or an endemic condition (i.e., persistent high occurrence).
• Exposure—the suspected causal agent persists in the environment or
in vivo and can be measured, exposure levels vary in the target population or there is the potential to study several exposed and unexposed
communities, people can recall their exposure accurately, or it can be
reconstructed from records
• Other—high relative risk exists, and obvious artifacts have been
eliminated.
In view of the low likelihood of detecting causes of clusters and limited
public health resources, the CDC recommended a four-phase approach
comprising the initial response, assessment, a major feasibility study, and
an etiologic investigation (Centers for Disease Control and Prevention,
1990). The protocol includes communications with regional authorities
and the public, study design and methods, criteria for confirmed and
probable cases, questionnaires for systematic information collection, specified samples and laboratory tests for diagnostic and exposure indicators,
and report format.
The first evidence of the extreme toxicity of in utero methylmercury
exposure to the developing fetus came from an investigation of a cluster
of a strange polio-like disease among inhabitants of Minimata, a small
fishing village in Japan; ultimately, over 2000 patients were diagnosed including 40 infants who generally appeared normal at birth but exhibited
Environmental Epidemiology
37
severe neurotoxicity within a few months. This and related events are described in more detail in Chapter 5 (Metals—Mercury, Arsenic, Cadmium,
and Manganese).
Ecologic Studies
In pure ecologic studies, the unit of observation is a group of people and
analyses assess the relation between group exposures (e.g., an average
level or prevalence rate of a given factor) and health outcomes (e.g., a disease incidence or mortality rate) (Morgenstern and Thomas, 1993). Because this design does not capture information on exposures and health
outcomes at the individual level, there are major uncertainties in the interpretation of risk estimates (see below). Further discussion uses a recent
categorization of ecologic studies (Rothman and Greenland, 1998).
Multiple-Group Designs
Exploratory study: compares disease rates among population subgroups during the same time period, without measurement of environmental exposures at the individual or group level; analyses assess the degree of spatial autocorrelation or clustering suggestive of environmental
effects.
Analytic study: assesses the association between average exposure
levels and disease rates among multiple population subgroups; analyses
include ordinary least squares and linear or exponential relative rate
models.
Time-Trend Designs
Exploratory study: compares disease rates over time in a defined population and may be used to forecast future trends; analyses range from
simple graphical displays to autoregressive integrated moving averages
(ARIMA) and age-period-cohort models.
Analytic study: assesses the association between average exposure
levels and disease rates over time; analyses are as noted above.
Mixed Designs
Exploratory study. This study combines multiple-group and timetrend features; analyses include ARIMA and age-period-cohort models.
Analytic study. This study examines the association between differences in average exposure levels and disease rates over time and between
multiple population subgroups.
38
CHILD HEALTH
AND THE
ENVIRONMENT
Reasons for choosing an ecologic study design include low cost, convenience, impracticality of measuring individual exposures for many individuals, uniform exposures within regions necessitating comparison of
several regions, and simplicity of analyses and presentation (Rothman
and Greenland, 1998). Ecologic studies generally take advantage of routinely available information and provide relatively rapid results that may
justify more definitive investigations. Because they are based on information about groups, privacy and confidentiality issues are minimal or
nonexistent. Ecologic studies are appropriate for assessing the association
between relatively rapid changes in illness rates and average exposure
levels, such as daily emergency room visits for asthma and daily (or more
frequent) air pollution levels, applying variable lag periods to allow time
for adverse health outcomes to occur.
The major weakness of ecologic studies is their reliance on data for
groups rather than individuals; associations between exposures and
health outcomes at the group level may not reflect those at the individual level (Morgenstern, 1982; Morgenstern and Thomas, 1993). Given the
absence of information on disease risks among exposed and unexposed
individuals, such studies cannot directly estimate exposure effects. This
weakness is exacerbated when the disease of interest occurs only after
long or highly variable latent periods; under such conditions, a time-trend
analytic study assesses disease rates for a population of constantly changing individuals as persons enter or leave the base population over time.
Ecologic studies are also subject to biases related to confounder misclassification, multicolinearity of risk factors, and temporal ambiguity of
cause and effect. See other sources for more detailed discussion of these
and other limitations of ecologic studies (Morgenstern and Thomas, 1993).
Examples
• A multiple-group analytic study in Sweden revealed a substantially
higher risk of acute lymphatic leukemia among children living in areas
with medium or high average radon levels (Kohli et al., 2000)
• A time-trend analytic study showed that daily hospital admissions of infants for respiratory diseases during the summer high-ozone season were
associated with daily 1-hour maximum ozone levels independent of
other ambient air pollutants and climatic conditions (Burnett et al., 2001)
Hypothesis-Testing Studies
Cross-Sectional Studies
A cross-sectional study for investigating disease etiology generally surveys a population (or a sample of a population) at a point in time and
Environmental Epidemiology
39
captures exposure information using standardized questionnaires, sometimes supplemented by other measures (physical examination, functional
tests, collection and analysis of blood, urine, or other biologic samples).
Certain demographic characteristics (e.g., age, sex, ethnicity, education),
environmental exposures (e.g., presence of pets or smokers in the home),
and health outcomes (e.g., physician-diagnosed childhood asthma) can
be assessed reliably using self- or parent-administered questionnaires, enabling the use of large sample sizes at relatively low cost. National health
surveys conducted for general health planning and policy purposes have
been used occasionally to capture environment-related exposure and
health outcome information; as documented below and in later chapters,
NHANES is a noteworthy example.
Although such studies can identify the prevalent cases in a population at a given point in time, they cannot identify the base population
from which the cases arose. Studies of health conditions with long latent
periods that only collect point-in-time information are prone to temporal
ambiguity, that is, current exposures may not reflect those during the etiologically relevant period. Parents of an asthmatic child, for example, may
have modified their smoking habits, pet ownership status, or home environment since onset of the disease. Cross-sectional studies are also subject to length-biased sampling, that is, overrepresentation of cases with
long-duration diseases and underrepresentation of those with shortduration diseases. Nevertheless, cross-sectional studies that collect accurate information on exposures during etiologically relevant time periods
can provide valid exposure-specific risk estimates.
Examples. A cross-sectional study of children provided the first strong
evidence of neurotoxicity from moderate lead exposure of children. Lead
levels in tooth dentine were measured; these reflect cumulated lead exposure since birth, thus avoiding the need to rely on current exposure status or recall of past exposures (Needleman et al., 1979). The NHANES has
been used repeatedly to assess important child health and environmental issues including
• Blood lead levels in relation to declining use of leaded gasoline in the
late 1970s (Annest et al., 1983), long-term blood lead concentration
trends (Centers for Disease Control and Prevention, 2000), and the relation between lead exposure and cognitive function (Lanphear et al.,
2000)
• Lung function and outdoor air pollution (Schwartz, 1989)
• Risk factors for childhood asthma (Lanphear et al., 2001)
40
CHILD HEALTH
AND THE
ENVIRONMENT
Case-Control Studies
Well-conducted population-based case-control studies are generally relatively inexpensive, produce reasonable risk estimates, are very appropriate for identification of environmental or other hazards, and are often the
only practical design for investigation of rare diseases such as childhood
cancer or the less common birth defects. Case-control studies generally
include samples of cases (persons with a specific health condition) and
controls (healthy persons or persons with other health conditions believed
to be unrelated to the exposure of interest). Increasingly, researchers try to
recruit samples representative of a defined population, usually defined by
geographic area of residence; this is not always possible, and some studies use other samples (e.g., hospital-based). Most studies of childhood cancer have recruited newly diagnosed (i.e., incident) cases, while investigations of birth defects and asthma have usually included prevalent cases.
Selection of controls representative of the base population from which
the cases developed is a major challenge. Equally demanding is the collection of unbiased retrospective exposure information, especially for delayed health outcomes; cases and controls may differentially recall their
exposures. Similar to cross-sectional studies, case-control investigations
of prevalent cases may identify associations with factors that influence
survival or severity rather than etiology. Unless the sampling fractions
are known (as in nested case-control studies), case-control assessments
can estimate relative risk but not absolute risk among exposed subjects.
Case-control studies are inefficient for studies of rare exposures unless
they are conducted within a high-exposure cohort. See other sources for
further information about case-control studies (Gordis, 2000; Rothman
and Greenland, 1998).
Examples. The role of environmental factors in spontaneous abortion,
stillbirths, birth defects, SIDS, childhood cancer, and other health outcomes has been assessed in numerous case-control studies. For instance:
• A nested case-control study of childhood cancer within a cohort of over
32,000 twins born in Connecticut during 1930–1969 demonstrated an
association between leukemia and in utero exposure to X-rays (Harvey
et al., 1985)
• The Baltimore-Washington Infant Study showed that a specific congenital heart defect (transposition of the great arteries) was associated with
maternal first trimester exposure to pesticides (Loffredo et al., 2001)
Cohort Studies
A cohort study generally begins with a group of healthy persons with heterogeneous exposures, or with two or more groups of healthy persons
Environmental Epidemiology
41
with contrasting exposures, and follows them over time to detect adverse
health outcomes. Information may be collected prospectively, retrospectively, or both. In countries with computerized population-based health
care and/or vital statistics records, follow-up of cohorts may be achieved
through computerized record linkage (see, e.g., Fair et al., 2000). Cohort
studies have several strengths including:
• Avoidance of temporal ambiguity, that is, virtual certainty that exposure precedes the health outcome of interest
• Lack of selection bias related to disease status, that is, subjects are recruited while healthy
• Ability to assess the relation between exposures of interest and multiple health outcomes including specific diseases and overall survival
• Ability to assess risks of rare exposures, such as the future risk of adult
cancers among a cohort of children exposed to high-dose ionizing
radiation
Potential weaknesses of cohort studies include loss of subjects during
follow-up (if substantial, this can cause loss of statistical power and introduction of bias), inefficiency for investigation of rare health outcomes,
reliance on baseline exposure assessment, and high cost. Exposure status
may change for persons followed for long periods to detect delayed health
effects; in principle, this problem can be addressed by repeat exposure assessments, but prospective cohorts are generally expensive, especially if
they require large samples and extensive exposure measurements.
Examples
• A record-based cohort study of over 140,000 births showed that women
who consumed chlorinated dark water had a twofold increased risk of
giving birth to children with urinary tract defects (note: certain dark
surface waters are high in humic acids and generate high levels of chlorinated by-products when chlorinated) (Magnus et al., 1999)
• A longitudinal study of over 7000 pregnancies showed an association
between paternal gonadal exposure to diagnostic X-rays during the
year before conception and reduced birth weight, independent of maternal smoking and other potential confounders (Shea and Little, 1997)
Experimental Studies
The simplest experimental study design involves random assignment of
individuals or groups to receive or not receive an intervention and follow-up to detect health outcomes. Children are unable to give informed
consent, and their participation in experimental studies is almost always
limited to potentially beneficial interventions. Older children have occa-
42
CHILD HEALTH
AND THE
ENVIRONMENT
sionally been included in experimental studies of air pollutants, for example, to test the effect of short controlled exposures to ozone on respiratory function and symptoms. The main advantage of experiments is the
greatly increased chance that observed differences between subjects exposed or not exposed to the intervention are not caused by unknown or
uncontrolled confounders. The relatively high cost of most experimental
intervention studies usually limits them to relatively small samples sizes
and limited statistical power. Randomization at the group level reduces
statistical power and precision because of within-group dependence of
the outcome variable (Donner et al., 1981). Nonrandom loss of subjects to
follow-up can introduce bias, and noncompliance with the intervention
can cause underestimation of the true effect.
Examples. Examples of the relatively few experimental studies that
have been done include field trials of sunscreen use to prevent benign
nevi in children (Gallagher et al., 2000), house dust mite antigen reduction to reduce childhood asthma severity (Weeks et al., 1995), and house
dust control to reduce childhood lead levels (Lanphear et al., 1996).
Conclusion
Epidemiologic studies provide valuable information on the relationships
between environmental exposures and the risk of adverse health outcomes in humans under real-life conditions. They have identified adverse
health effects during gestation, childhood, and adulthood arising from
early life exposure to diverse environmental toxicants including ionizing
radiation, lead, methylmercury, PCBs, ETS, and outdoor air pollutants. In
addition, epidemiologic studies have revealed limited evidence of health
effects from environmental factors including pesticides, chlorinated disinfection by-products in drinking water, and extremely low frequency
electromagnetic fields. Their major limitation is their post hoc nature, that
is, they can detect hazards only after humans have been exposed and developed adverse health effects. Proposed criteria for selecting epidemiologic studies appropriate for quantitative risk assessments are shown in
Table 2–1 (Hertz-Picciotto, 1995). Epidemiologic studies of many children’s environmental health issues conducted to date failed to satisfy all
five criteria because:
• Most children have low exposures to many environmental contaminants; any associations with adverse health effects are likely to be weak
unless sufficient numbers of children across a range of exposures are
available to assess dose–response relationships.
TABLE 2–1. Criteria for Selecting Epidemiologic Studies for Risk Assessments
Necessity of Criterion Being Met To:
Criteria
Serve as a Basis
for Dose–Response
Extrapolation
Check Plausibility of an
Animal-Based Risk Assessment
Contribute to the Weight of
Evidence for a Suspected Health Hazard
1. Moderate to strong
positive association
Necessary
Not necessary; often this criterion
is not met
If met, adds to weight of evidence for a
hazard
2. Strong biases ruled out
or unlikely
Necessary
Should be met, at least partially
If met, strengthens evidence that agent is
or is not a hazard
3. Confounding controlled,
or likely to be limited
Necessary
Should be met, at least partially,
or limits on confounding should
be estimated
If met, strengthens evidence that agent is
or is not a hazard
4. Quantification of exposures
linked to individuals
Necessary
Some quantification of exposures
is needed, even if based on
data external to study site
Usually not met
5. Montonic dose–response
relationship
Not necessary but
adds certainty to
risk estimates
Criteria 1–4 should
be met
Not necessary
May or may not be met
Two of criteria 1–3 should be met
All other studies
Summary of requirements
Source: Hertz-Picciotto (1995).
44
CHILD HEALTH
AND THE
ENVIRONMENT
• Quantification of environmental exposures is generally difficult, especially when only individual recall information is available; even if appropriate exposure biomarkers exist, there may be problems including
access to adequate biologic specimens (e.g., blood), reliance on indices
that can only measure recent exposures (e.g., organophosphate pesticide metabolites in urine), and the high cost of laboratory analyses.
• Monotonic dose–response relationships are more likely to be observed
when the true association is moderate or strong and in studies with
good exposure data and power; these conditions have been met in some
but not most epidemiologic studies.
References
American Cancer Society. (2002). Facts and figures 2002. Located at http://www.
cancer.org
Annest JL, Pirkle JL, Makuc D, Neese JW, Bayse DD, Kovar MG. (1983). Chronological trend in blood lead levels between 1976 and 1980. N Engl J Med 308:
1373–7.
Burnett RT, Smith-Doiron M, Stieb D, Raizenne ME, Brook JR, Dales RE, Leech
JA, Cakmak S, Krewski D. (2001). Association between ozone and hospitalization for acute respiratory diseases in children less than 2 years of age. Am
J Epidemiol 153:444–52.
Caldwell GG. (1990). Twenty-two years of cancer cluster investigations at the Centers for Disease Control. Am J Epidemiol 132:S43–7.
Centers for Disease Control and Prevention. (1990). Guidelines for investigating
clusters of health events. MMWR 39:1–23.
Centers for Disease Control and Prevention. (2000). Blood lead levels in young
children—United States and selected states, 1996–1999. MMWR 49:1133–7.
Donner A, Birkett N, Buck C. (1981). Randomization by cluster. Sample size requirements and analysis. Am J Epidemiol 114:906–14.
Elwood JM. (1998). Critical appraisal of epidemiological studies and clinical trials. New York: Oxford University Press.
Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC. (2000). An assessment of the validity of a computer system for probabilistic record linkage of
birth and infant death records in Canada. The Fetal and Infant Health Study
Group. Chronic Dis Can 21:8–13.
Gallagher RP, Rivers JK, Lee TK, Bajdik CD, McLean DI, Coldman AJ. (2000).
Broad-spectrum sunscreen use and the development of new nevi in white children: A randomized controlled trial. JAMA 283:2955–60.
Gibson JL. (1892). Notes on lead-poisoning as observed among children in Brisbane. Proc Intercolonial Med Congr Aust 3:76–83.
Gordis L. (2000). Epidemiology. Philadelphia: W.B. Saunders Company.
Greenland S. (1980). The effect of misclassification in the presence of covariates.
Am J Epidemiol 112:564–9.
Greenland S. (1993). Basic problems in interaction assessment. Environ Health Perspect 101(Suppl 4):59–66.
Environmental Epidemiology
45
Harvey EB, Boice JD, Honeyman M, Flannery JT. (1985). Prenatal x-ray exposure
and childhood cancer in twins. N Engl J Med 312:541–5.
Hatch M, Thomas D. 1993. Measurement issues in environmental epidemiology.
Environ Health Perspect 101(Suppl 4):49–57.
Health Canada. (2002). Cancer surveillance on-line. Located at http://cythera.
ic.gc.ca/dsol/cancer/.
Hertz-Picciotto I. (1995). Epidemiology and quantitative risk assessment: a bridge
from science to policy. Am J Public Health 85:484–91.
Hertz-Picciotto I, Neutra RR. (1994). Resolving discrepancies among studies: the
influence of dose on effect size. Epidemiology 5:156–63.
Infante-Rivard C, Jacques L. (2000). Empirical study of parental recall bias. Am J
Epidemiol 152:480–6.
International Clearinghouse for Birth Defects. (2002). International Clearinghouse for
Birth Defect monitoring systems. Located at http://www.icbd.org/index.html.
Kohli S, Noorlind Brage H, Lofman O. (2000). Childhood leukaemia in areas with
different radon levels: a spatial and temporal analysis using GIS. J Epidemiol
Commun Health 54:822–6.
Lanphear BP, Aligne CA, Auinger P, Weitzman M, Byrd RS. (2001). Residential exposures associated with asthma in US children. Pediatrics 107:505–11.
Lanphear BP, Dietrich K, Auinger P, Cox C. (2000). Cognitive deficits associated
with blood lead concentrations 10 microg/dL in U.S. children and adolescents. Public Health Rep 115:521–9.
Lanphear BP, Winter NL, Apetz L, Eberly S, Weitzman M. (1996). A randomized
trial of the effect of dust control on children’s blood lead levels. Pediatrics
98:35–40.
Loffredo CA, Silbergeld EK, Ferencz C, Zhang J. (2001). Association of transposition of the great arteries in infants with maternal exposures to herbicides and
rodenticides. Am J Epidemiol 153:529–36.
Magnus P, Jaakkola JJ, Skrondal A, Alexander J, Becher G, Krogh T, Dybing E.
(1999). Water chlorination and birth defects. Epidemiology 10:513–7.
Morgenstern H. (1982). Uses of ecologic analysis in epidemiologic research. Am J
Public Health 72:1336–44.
Morgenstern H, Thomas D. (1993). Principles of study design in environmental
epidemiology. Environ Health Perspect 101(Suppl 4):23–38.
Needleman HL, Gunnoe C, Leviton A, Reed R, Peresie H, Maher C, Barrett P.
(1979). Deficits in psychologic and classroom performance of children with elevated dentine lead levels. N Engl J Med 300:689–95.
Neutra RR. (1990). Counterpoint from a cluster buster. Am J Epidemiol 132:1–8.
Olsen SF, Martuzzi M, Elliott P. (1996). Cluster analysis and disease mapping—
why, when, and how? A step by step guide. BMJ 313:863–6.
Pott P. (1775). Cancer scroti. In Chirurgical observations relative to the cataract, the
polypus of the nose, the cancer of the scrotum, the different kinds of ruptures,
and the mortification of the toes and feet, pp. 63–68. London: T.J. Carnegy.
Prentice RL, Thomas D. (1993). Methodologic research needs in environmental
epidemiology: data analysis. Environ Health Perspect 101(Suppl 4):39–48.
Rothman KJ. (1993). Methodologic frontiers in environmental epidemiology. Environ Health Perspect 101(Suppl 4):19–21.
Rothman KJ. (2002). Epidemiology: an introduction. New York: Oxford University Press.
46
CHILD HEALTH
AND THE
ENVIRONMENT
Rothman KJ, Greenland S. (1998). Modern epidemiology. Philadelphia: LippincottRaven.
Schwartz J. (1989). Lung function and chronic exposure to air pollution: a crosssectional analysis of NHANES II. Environ Res 50:309–21.
Shea KM, Little RE. (1997). Is there an association between preconception paternal x-ray exposure and birth outcome? The ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. Am J Epidemiol 145:546–51.
Stroup DF, Teutsch SM. (1998). Statistics in public health: quantitative approaches
to public health problems. New York: Oxford University Press.
Teutsch SM, Churchill RE. (2000). Principles and practice of public health surveillance. New York: Oxford University Press.
Wallace L. (1993). A decade of studies of human exposure: what have we learned?
Risk Anal 13:135–9.
Weed DL, Coughlin SS. (1999). New ethics guidelines for epidemiology: background and rationale. Ann Epidemiol 9:277–80.
Weeks J, Oliver J, Birmingham K, Crewes A, Carswell F. (1995). A combined approach to reduce mite allergen in the bedroom. Clin Exp Allergy 25:1179–83.
3
Risk Assessment
Risk assessment has been described as a bridge between science and policy (Hertz-Picciotto, 1995). More formally, it is the use of a factual base to
define the health effects of exposure of individuals or populations to hazardous materials and situations (National Academy of Sciences, 1983). The
goal is to provide an evidence base for decisions on policies and programs
to protect the public from identified health hazards. The process involves
the synthesis of epidemiologic, toxicologic, and related research findings
and judgments on causal relationships and exposure–risk relationships.
This chapter discusses general issues related to hazard identification,
dose–response assessment, and risk characterization. Topics include processes used by national and international agencies to identify causal relationships and health-protective exposure limits followed by discussion
of key issues related to risk assessments of carcinogens, reproductive toxicants, developmental toxicants, and neurotoxins.
Risk Assessment Framework
The model most frequently used for quantitative risk assessment is that
proposed by the U.S. National Academy of Sciences in its 1983 landmark
report on risk assessment (National Academy of Sciences, 1983):
47
48
CHILD HEALTH
AND THE
ENVIRONMENT
• Hazard identification—determination of whether a particular chemical
is a causal factor1 for particular health outcomes
• Dose–response assessment—quantification of the relationship between
the dose (or exposure) and the probability of adverse health outcomes
• Exposure assessment—quantification of the extent of human exposure
before or after application of regulatory controls
• Risk characterization—description of the nature and magnitude of the
human risk including attendant uncertainty
The National Academy of Sciences report defined risk assessment as a
scientific process distinct from socioeconomic and other political considerations. The sparseness and uncertainty of relevant scientific knowledge
were the main problems identified for assessment. In a departure from
that 1983 report, the recently revised U.S. risk assessment framework emphasized economic costs (The Presidential/Congressional Commission on
Risk Assessment and Risk Management, 1997a, 1997b). It also recommended early and active stakeholder involvement, testing of chemical
mixtures, inclusion of microbiologic and radiation hazards (in addition to
chemicals), comparable risk metrics for carcinogens and other hazards,
increased use of mode of action information, and realistic exposure scenarios. After comparing chemical risk assessment procedures, assumptions, and policies across four federal agencies, the U.S General Accounting Office concluded that incomplete scientific information on human
health effects and exposure to hazards continues to be a major source of
uncertainty (U.S. General Accounting Office, 2001).
Hazard Identification
Hazard identification is the determination of whether a particular agent
is or is not causally linked to particular health effects (National Academy
of Sciences, 1983). Toxicologic and epidemiologic studies provide essential information for hazard identification and dose–response assessment.
Only animal studies can identify new hazards before human exposure occurs, and only epidemiologic studies can directly demonstrate human
health risks related to existing environmental hazards.
Reasonable evidence of causality is a prerequisite for a meaningful
risk assessment. In assessing the impact of tobacco use on human health,
the 1964 U.S. Surgeon General’s report grouped the available scientific
evidence into two main categories (U.S. Surgeon General, 1964): (1) ex1 A causal factor of an event has been defined as a factor whose operation increases the
frequency of the event (Elwood, 1998).
Risk Assessment
49
perimental studies of animals exposed to tobacco smoke, tars, and specific chemical components—these showed that tobacco smoke contains
several carcinogens, promoters, and toxic gases that cause adverse health
effects in tissues and cells similar to those in heavy smokers—and (2) clinical and epidemiologic studies of humans—these demonstrated multiple
pathologic effects in cells and tissues of smokers and strong associations
between smoking and lung cancer, heart disease, emphysema, and several other diseases. The Surgeon General’s report used the criteria illustrated below for lung cancer to evaluate the causal significance of associations between smoking and health effects:
• Consistency—all 36 reasonably designed epidemiologic studies found
associations between smoking and lung cancer.
• Strength—the 16 studies that measured relative risk all found similar
high values (9–10 for average smokers, 20 or higher for heavy smokers).
• Specificity—although smoking causes more than one adverse health
outcome, tobacco smoke is a complex mixture capable of causing more
than one disease and even single toxicants can cause multiple health
effects.
• Temporal relationship—the reported age at onset of smoking occurred
well before any detectable health effects.
• Coherence—a causal association was consistent with evidence of increasing lung cancer death rates, dose–response relations, anatomic site
of cancer, and death rate differentials between men and women, urban
and rural populations, and socioeconomic subgroups.
Hill (1965) concluded that none of the Surgeon General’s criteria provide
indisputable evidence for or against causality. The absence of appropriate temporality, however, would be strong evidence against causality, assuming no ambiguity about time relations (Rothman and Greenland,
1998). Hill stated that strong associations are more likely to be causal because a potential confounder would have to be even more closely associated with the health outcome and likely to be already evident. A strong
association, however, is neither necessary nor sufficient to prove causality and only rules out modest effects from confounders or other sources
of bias (Rothman and Greenland, 1998).
Studies may produce inconsistent evidence if some populations lack
factors required for the effect of a putative causal factor; thus, one can assess consistency only when a causal mechanism is understood in detail.
Other factors contributing to inconsistency include random variation in
small-sample studies and measurement errors. Specificity is not a discriminating criterion; as stated succinctly elsewhere, “there is no biolog-
50
CHILD HEALTH
AND THE
ENVIRONMENT
ical justification for specificity and it is useless and misleading” (Rothman and Greenland, 1998).
Hill noted the importance of a monotonic dose–response relationship; this appears to be true for some hazards (e.g., genotoxic carcinogens), but others have apparent thresholds. Because biologic plausibility
and coherence of evidence must be judged within the context of current
knowledge, violation of these criteria may simply reflect knowledge gaps.
Positive biologic plausibility—for example, evidence from experimental
animal studies—may be supportive of causality, but negative experimental evidence can be misleading if it arises from a weak intervention
or low statistical power. Given these limitations of causal criteria, researchers could adopt causal values to encourage debate on uncertainty
and scientific values (Kaufman and Poole, 2000).
Dose–Response Assessment
This phase of risk assessment determines the relation between exposure
magnitude and the probability of particular health effects using two main
approaches, one for carcinogens and the other for noncarcinogens. The
main difference is the default assumption for carcinogens of low-dose linearity with no threshold, compared to an assumption of low-dose nonlinearity with a threshold for noncarcinogens. Regulatory agencies have
used lifetime excess cancer risks of 106 to 105 as minimal risk targets2
in setting exposure limits for the general population to carcinogens and
certain other toxicants with no apparent threshold. Dose–response relationships observed in animal or human subjects are generally based on
exposures that cause far higher lifetime risks; for example, 50% or more
of the animals in the highest dose category of carcinogenesis bioassays
may develop cancer.
The challenge is to use observed dose–response data to estimate risks
at exposures several orders of magnitude below the observed range. Ideally, epidemiologic estimates of disease incidence rates by sex, age, and
exposure level should be available to support low-dose extrapolations and
other risk estimates. Toxicants may have linear, sigmoidal, U-shaped, or
inverted U-shaped dose–response relationships. If the observed data do
not include the full range necessary to detect a true non-linear dose–
response relationship, low-dose extrapolation could be quite misleading.
Simulation modeling indicates that point estimates of low-dose risks may
differ from the actual risks by three orders of magnitude or more, depending
2 An excess risk of 106 means that among 1 million persons exposed for a lifetime at
the exposure limit level, one extra cancer would occur.
Risk Assessment
51
on the statistical model used and the effects of competing risks, background
response, latency, and experimental design (Krewski et al., 1983).
Exposure Assessment
This phase of risk assessment estimates the extent of human exposure under various intervention scenarios. Data on the prevalence of exposure by
intensity, sex, and age (together with the related risk data mentioned
above) greatly facilitate risk estimates for various exposure scenarios. The
main environmental exposure indices are measures of environmental contaminant concentrations and internal dose levels. Population-based data
on internal dose levels are rarely available and, in their absence, risk assessors often rely on modeled estimates dependent on external contaminant levels (air, water, food, and soil), human behavior (consumption data,
activities), bioavailability, and route-specific absorption factors (inhalation, oral, dermal). Given the uncertainties surrounding these factors, recent efforts to develop population-based estimates of internal dose promise to provide more precise exposure estimates, at least for those hazards
for which valid biomarkers of internal dose exist. Biomarkers of internal
dose reflect exposures from all sources and pathways over times that vary
by the in vivo half-lives of contaminants.
The 1996 Food Quality Protection Act (FQPA) mandated the U.S. EPA
to consider aggregate risks (from total exposure to a single agent via multiple sources and pathways) and cumulative risks (from exposure to two
or more distinct chemicals sharing a common mode of action) of pesticides. Aggregate risk assessment requires information on pesticide levels
in all relevant environmental media, bioavailability, route-specific absorption factors, and human exposure factors (air, food, and water consumption) specific for sex and age. The CDC’s recent biomonitoring activities have generated data on urinary organophosphate insecticide
metabolite levels for children and adults (Centers for Disease Control and
Prevention, 2001); this type of biomonitoring promises to improve exposure estimates used in risk assessments.
Risk Characterization
This phase of risk assessment describes the nature and magnitude of the
risk to human health related to a particular hazard and the attendant uncertainties. Outputs include (1) an evaluation of the nature and magnitude of potential adverse health effects based on synthesis of information
from hazard identification, dose–response assessment, and human exposure assessment, (2) quantitative estimates of exposures that cause mini-
52
CHILD HEALTH
AND THE
ENVIRONMENT
mal risk (for apparent nonthreshold toxicants) or exposures that likely
cause no excess risk (for toxicants with apparent thresholds), and (3) a
summary of the risk assessment process including reasons for selecting
key studies, critical effects and their relevance to humans, limitations of
available data, data gaps, areas of uncertainty, and level of scientific confidence in the risk assessment results.
Threshold Toxicants
The reference dose (RfD) or concentration (RfC) of a toxicant is the estimated amount that is assumed to be without an appreciable risk of noncancer adverse health outcomes over a lifetime of exposure (including the
noncancer toxic effects of carcinogens such as arsenic). Derivation of a
RfD or RfC includes
• Estimation of the no observed adverse effect level (NOAEL), the lowest observed adverse effect level (LOAEL), or the benchmark dose for
the adverse health effect most sensitive to the toxicant of interest (or
the most sensitive mammalian species)
• Division of the NOAEL, LOAEL, or benchmark dose by factors (often
referred to as uncertainty or safety factors) of up to 10 each to compensate for the uncertainties discussed below
The National Academy of Sciences has identified five uncertainty and
modifying factors for potential use in developmental risk assessments:
(1) extrapolation from animals to humans, (2) human chronic exposures
compared to subchronic exposures in animals, (3) variation in susceptibility within genetically diverse human populations, (4) incomplete toxicity databases, and (5) susceptibility of human developing systems (National Academy of Sciences, 2000). The last two criteria may be invoked
for pesticides used on foods under the FQPA, that is, an additional factor of up to 10 may be applied if children are thought to have increased
exposures or susceptibility or if supporting data are incomplete. The
OECD concluded that the product of uncertainty factors would normally
range from 100 to 500 for NOAELs for repeat dose toxicity and up to 5000
for severe or irreversible developmental effects that occur at exposure levels below those inducing well-defined parental toxicity.
Most RfDs and RfCs are based on studies of experimental animals or
adult humans (Table 3–1). Uncertainty factors are substantially larger for
assessments dependent on animal compared to human data (e.g., 10 for
methylmercury based on developmental neurotoxicity in humans versus
1400 for tetrachloroethylene based on hepatotoxicity and weight gain in
rats). The EPA has not developed an RfD for lead because of uncertainty
TABLE 3–1. Low-Dose Risk Estimates for Noncancer Health Effects from Chronic Exposure to Selected Toxicants
Substance
RfD or RfC a
Uncertainty
Factor b
Arsenic (inorganic)
3 104 mg/kg/day
3
Cadmium
Mercury (elemental)
5 104 mg/kg/day
3 104 mg/m3
10
30
Methylmercury
Bromodichloromethane
Tetrachloroethylene
PCBs (Arochlor-1016)
PCBs (Arochlor-1254)
1 104 mg/kg/day
0.02 mg/kg/day
0.01 mg/kg/day
7 105 mg/kg/day
2 105 mg/kg/day
10
1000
1000
100
300
Alachlor
Chlorpyrifos
Permethrin
0.01 mg/kg/day
0.003 mg/kg/day
0.05 mg/kg/day
Source: U.S. Environmental Protection Agency (2000b).
a RfD
mg/kg/day (diet) or mg/L (water); RfC mg/m3 (inhalation).
b NOAEL
or LOAEL divided by RfD or RfC.
100
10
100
Critical Effect
Oral (water)—hyperpigmentation and keratosis of skin, possible
vascular complications (humans)
Oral (water, food)—significant proteinuria (humans)
Inhalation—hand tremor, memory disturbances, slight autonomic
dysfunction (humans)
Oral (food)—developmental neurotoxicity (humans)
Oral (oral/gavage)—renal cytomegaly (mice)
Oral—hepatotoxicity in mice, weight gain in rats
Oral—reduced birth weight (monkeys)
Oral—inflamed Meibomian glands, distorted nails, decreased IgG
and IgM response to sheep red blood cells (monkeys)
Oral—hemosiderosis, hemolytic anemia (dogs)
Oral—decreased plasma cholinesterase activity after 9 days (humans)
Oral—increased liver weights (rats)
54
CHILD HEALTH
AND THE
ENVIRONMENT
about a threshold for lead toxicity; evidence indicates effects on certain
blood enzymes and auditory and cognitive functions at very low blood
lead levels (see also Chapter 4, Lead).
Although the EPA has set a maximum contaminant level for dioxin
in public drinking water of 3 108 mg/L, it recently concluded that
dioxin is a human carcinogen and that current population exposure levels may increase the lifetime excess probability of cancer by up to 103 to
102 (U.S. Environmental Protection Agency, 2000a). Because this risk is
much higher than that usually considered acceptable for general population exposure, the EPA did not recommend an RfD for dioxin (under its
traditional approach, the EPA would have had to assign an RfD that is
two to three orders of magnitude lower than current background exposure levels). A review panel has recommended that the EPA augment its
RfDs and RfCs for chronic exposures with those for acute, short-term, and
longer-term exposures to better address children and other sensitive subpopulations.
The U.S. Agency for Toxic Substances and Disease Registry (ATSDR)
estimates virtually safe exposure limits for noncarcinogens by the process
described above but refers to them as minimal risk levels (MRLs). The
ATSDR uses MRLs as screening levels to identify hazardous waste site
contaminants and potential health effects that may be of concern. Minimal risk levels have been derived for acute (15 days), intermediate
(15–364 days), and chronic (1 year) exposure durations and for oral and
inhalation routes (see selected examples in Table 3–2). Comparison of
Tables 3.1 and 3.2 indicates that the EPA and the ATSDR occasionally produce different risk levels for the same toxicant; for example, the EPA
RfD for chronic oral exposure to cadmium is 5 104 mg/kg/day, while
the ATSDR MRL for the same exposure is 2 104 mg/kg/day. Such
differences may arise from reviews conducted at different times, that is,
with differing sets of evidence. The ATSDR has defined an MRL of
109 mg/kg/day for chronic oral dioxin exposure based on neurotoxicity in monkeys.
The use of uncertainty factors is illustrated by the derivation of the
EPA’s RfD for methylmercury that was based on:
• A benchmark dose (BMD)3 from two epidemiologic studies of neurologic abnormalities among prenatally exposed infants.
3 A BMD may be defined as the dose (or its 95% lower confidence limit) of a toxicant
that produces a specified increased incidence of a response, for example, 1% or 5% of
the maximum toxic response, based on data within the observed dose range (International Programme on Chemical Safety, 1999).
TABLE 3–2. Minimum Risk Levels for Selected Toxicants by Duration of Exposure
Duration and Route
MRL
Uncertainty
Factors
Acute oral
Chronic oral
Acute inhalation
Intermediate inhalation
Chronic oral
Acute or
intermediate oral
Chronic oral
Acute oral
Intermediate oral
Chronic oral
0.005 mg/kg/day
3 104 mg/kg/day
0.05 ppm
0.004 ppm
2 104 mg/kg/day
0.003 mg/kg/day
10
3
300
90
10
10
0.001 mg/kg/day
0.008 mg/kg/day
104 mg/kg/day
2 105 mg/kg/day
100
300
90
1000
Methylmercury
Chronic inhalation
Acute oral
Intermediate oral
Chronic oral
2 104 mg/m3
0.007 mg/kg/day
0.002 mg/kg/day
3 104 mg/kg/day
30
100
100
4
PCBs
Intermediate oral
3 105 mg/kg/day
300
Chronic oral
2 105 mg/kg/day
300
Acute oral
Intermediate oral
Chronic oral
2 107 mg/kg/day
2 108 mg/kg/day
109 mg/kg/day
21
30
90
Substance
Arsenic
Benzene
Cadmium
Chlorpyrifos
Hexachlorobenzene
Mercury (elemental)
Mercury (inorganic)
Dioxin
(2,3,7,8-TCDD)
Source: (Agency for Toxic Substances and Disease Registry (2002).
Health Effects
Gastrointestinal effects and facial edema (humans)
Hyperpigmentation and keratosis of skin (humans)
LOAEL for immunologic effects (mice)
LOAEL for neurologic effects (mice)
Renal effects (humans)
NOAEL for AChE inhibition in human adult males orally
exposed to chlorpyrifos
NOAEL for AChE inhibition in orally exposed rats
LOAEL for hyperactivity in offspring of prenatally exposed rats
Degenerative changes in ovaries of orally exposed monkeys
LOAEL for peribiliary lymphocytosis and fibrosis of the liver in
orally exposed adult male rats
Hand tremor in industrially exposed men
NOAEL for increased absolute and relative kidney weights in rats
NOAEL for increased absolute and relative kidney weights in rats
Neurodevelopmental effects in Seychelles Islands children exposed
to methylmercury from fish prenatally and postnatally
LOAEL for neurobehavioral effects in infant monkeys exposed
from birth to a PCB congener mixture similar to that in human
breast milk
LOAEL for immunologic effects in adult monkeys exposed to
Aroclor 1254
Reduced serum total hemolytic complement activity in mice
Decreased thymus weight in guinea pigs
Altered social interactions in monkeys exposed perinatally
56
CHILD HEALTH
AND THE
ENVIRONMENT
• A threefold uncertainty factor for variable susceptibility in human populations. The EPA used this instead of a tenfold factor because the BMD
was based on effects in a sensitive subpopulation: developing fetuses.
• A threefold uncertainty factor for missing data. There was no twogeneration animal study, and there were no epidemiologic data on exposure duration in relation to developmental effects and adult peripheral neuropathy.
Nonthreshold Toxicants
There is no consensus on a method for risk assessment of chemicals suspected to have nonthreshold critical effects, such as genotoxic carcinogens
and germ cell mutagens; existing approaches are based largely on characterization of dose–response relationships. For carcinogens, the EPA estimates unit risks and slope factors; see, for instance, those for arsenic,
cadmium, and bromodichloromethane (Table 3–3). Note that the RfD for
chronic oral exposure to bromodichloromethane is 20 g/kg/day based
on noncancer risks (Table 3–1). Based on a cancer slope factor of 6.2
105 per g/kg/day in male mice (Table 3–3), chronic exposure at the
RfD would cause an estimated lifetime excess cancer risk of 20 6.2
105 or about 1.2 103; this is two to three orders of magnitude higher
than the commonly used definitions of acceptable risk for the general population, but the RfD incorporates an uncertainty factor of 1000. See further discussion below under “Carcinogens.”
Aggregate and Cumulative Risks
Methods for cumulative risk assessment of chemicals with a common
mode of action include estimation of toxic equivalents or a combined margin of exposure index. These methods require toxicity data for multiple
chemicals tested in the same animal species, and they assume that effects
of each chemical are additive and have similar dose–response relationships. The toxic equivalents method has been widely applied to dioxins
and related organochlorines that bind to the AhR receptor (see Chapter
6, PCBs, Dioxins, and Related Compounds). Potential problems in assessing aggregate or cumulative pesticide risks are illustrated by consideration of anticholinesterase (AchE) inhibitors (Wilkinson et al., 2000). In
addition to organophosphorus and carbamate insecticides, many drugs
and food constituents inhibit AChE to some degree. Aggregate or cumulative exposure estimates for AchE inhibitors would require many datasets
that do not exist (not all AChE inhibitors have undergone toxicity testing, especially testing involving perinatal exposures) and numerous assumptions concerning dose–response relationships and additivity of effects. Even if modeling methods are developed, much research will be
TABLE 3–3. Low-Dose Risk Estimates for Cancer from Chronic Exposure to Selected Toxicants
Arsenic (inorganic)
Cancer
Unit Risk a
Slope Factor b
Skin (humans exposed orally to water)
5 105 per g/L (water)
1.5 103 per g/kg/day
103
g/m3
Cadmium
Lung (men occupationally exposed
by inhalation)
1.8
(air)
Bromodichloromethane
Kidney (orally exposed male mice)
1.8 106 per g/L (water)
per
na
6.2 105 per g/kg/day
Source: U.S. Environmental Protection Agency (2000b).
a The upper-bound excess lifetime cancer risk from continuous exposure to an agent at 1 g/L in water or 1 g/m3 in air; for example, if the unit risk 1.5 106 g/L, up
to 1.5 excess tumors are expected to develop per 1 million people exposed for a lifetime to the chemical at a level of 1 g/L in water or 1 g/m3 in air.
b Upper 95% confidence limit on excess cancer risk from lifetime exposure to an agent, usually expressed as a proportion of a population affected per mg/kg/day; generally
reserved for use in the low-dose region, that is, exposures corresponding to risks less than 1 in 100.
58
CHILD HEALTH
AND THE
ENVIRONMENT
needed to generate the data needed to test model assumptions and support aggregate and cumulative risk assessments.
Selected Risk Assessment Practices
This section addresses those aspects of selected risk assessments that relate to child health protection. Approaches to assigning level of evidence
for carcinogenicity are particularly well developed and are discussed in
some depth.
Carcinogens
Hazard Identification
This section describes identification of carcinogens by the World Health
Organization’s International Agency for Research on Cancer (IARC) and
variations adopted by the U.S. EPA. To evaluate a given substance, the
IARC convenes an expert panel to assess the quality of relevant toxicologic and epidemiologic studies using standardized criteria. After assessing and summarizing the quality and results of studies, IARC panels judge
the weight of evidence that the agent in question is carcinogenic for humans using these criteria:
• Strength of association—although moderate associations do not imply lack
of causality, stronger associations are more likely to indicate causality.
• Consistency—associations replicated using different study designs and
under different circumstances of exposure are more likely to represent
a causal relationship than findings from single studies.
• Dose–response relationship—a monotonic increase in the cancer risk
with increasing exposure is considered a strong indication of causality;
a decline in risk after cessation of exposure also suggests causality.
• Specificity—increased risk of a specific anatomic and/or histologic type
of cancer supports causality, especially if limited to one histologic type
within the same organ.
Based on the above considerations, the panel decides the level of evidence
of carcinogenicity for humans and experimental animals (Table 3–4). Finally, the panel assesses the weight of evidence to decide the likelihood
that an agent is a human carcinogen (Table 3–5). The IARC and the EPA
have independently assessed the carcinogenicity of many substances;
most differences in ratings relate to designation of probable human carcinogens. With few exceptions, the EPA requires sufficient evidence in an-
TABLE 3–4. Level of Evidence of Carcinogenicity in Humans and Animals
Level of Evidence
Epidemiologic Evidence
Animal Study Evidence
Sufficient
A positive relationship has been observed between
the exposure and cancer in studies in which chance,
bias, and confounding could be ruled out with
reasonable confidence
Increased incidence of cancer or an appropriate combination of
benign and malignant neoplasms in (1) two or more species of
animals or (2) in two or more independent studies in one species;
exceptionally, a single study in which malignancies occur with
high incidence or unusual site/type of tumor, or early age at onset
Limited
A positive association has been observed between exposure
and cancer for which a causal interpretation is considered
to be credible, but chance, bias, or confounding could
not be ruled out with reasonable confidence
Evidence is restricted to a single experiment, unresolved questions
regarding the design, conduct, or interpretation of the study, or
only benign neoplasms or lesions of uncertain neoplastic potential
or neoplasms that commonly occur spontaneously
Inadequate
Available studies are of insufficient quality, consistency, or
statistical power to decide the presence or absence of
a causal association, or no data on cancer in humans
are available
Available studies have major qualitative or quantitative limitations,
or no data on cancer in experimental animals are available
Evidence
suggesting
lack of
carcinogenicity
Several adequate studies covering the full range of known
human exposure levels consistently do not show an
increased risk of cancer at any exposure level; the studies
should be methodologically sound and yield a pooled risk
estimate near unity with a narrow confidence interval; no
individual study nor the pooled results should show any
sign of a dose–response relationship; this decision applies
only to the tumor sites and levels of exposure studied
Adequate studies involving at least two species are available that
show no evidence of carcinogenicity; this applies only to the
species, tumor sites, and levels of exposure studied
Source: International Agency for Research on Cancer (1999).
Note: The IARC and the EPA both consider other evidence including data on preneoplastic lesions, tumor pathology, genotoxicity, structure–activity relationships, metabolism, pharmacokinetics, physicochemical parameters, analogous biologic agents, and mechanisms of carcinogenesis.
TABLE 3–5. Categorization of Overall Weight of Evidence for Human Carcinogenicity, IARCa and EPAb
Human Carcinogenicity
IARC
EPA
Human carcinogen
1. Sufficient evidence in humans or (occasionally)
limited human and sufficient animal evidence
plus strong evidence that carcinogenesis is
mediated by a mechanism that also operates
in humans
A. This group is used only when there is sufficient evidence from epidemiologic
studies to support a causal association between exposure to the agents and
cancer
Probable human
carcinogen
2A. Limited human and sufficient animal
evidence, or inadequate human and sufficient
animal evidence plus strong evidence that
carcinogenesis is mediated by a mechanism
that also operates in humans, or (occasionally)
limited human evidence
B1. Usually reserved for limited evidence of
carcinogenicity from epidemiologic studies
B2. Sufficient evidence from animal studies and inadequate or no evidence from
epidemiologic studies
Possible human
carcinogen
2B. Limited human and animal evidence, or
inadequate human and sufficient animal
evidence, or inadequate human, limited
animal, and other supporting evidence
C. Limited evidence of carcinogenicity in animals and absent or inadequate
human data; includes a variety of evidence such as a single well-conducted
positive study, tumor responses of marginal statistical significance in studies
having inadequate design or reporting, benign but not malignant tumors
caused by a nonmutagen, and responses of marginal statistical significance
in a tissue known to have a high or variable background rate
Not classifiable as
to human
carcinogenicity
3. Inadequate human and inadequate or limited
animal evidence, or, inadequate human and
sufficient animal evidence plus strong evidence
that the mechanism of carcinogenicity in
experimental animals does not operate in
humans
D. Inadequate or no human and animal evidence
Evidence of noncarcinogenicity
for humans
4. Evidence suggesting lack of carcinogenicity in
humans and in animals, or (occasionally),
inadequate human evidence but evidence
suggesting lack of carcinogenicity in animals
consistently and strongly supported by
a broad range of other relevant data
E. No evidence of carcinogenicity in at least two adequate animal studies in
different species or in both human and animal adequate studies; should not be
interpreted as a definitive conclusion that the agent will not be a carcinogen
under any circumstances
a Source:
International Agency for Research on Cancer (1999).
b Source:
U.S. Environmental Protection Agency (1986).
Risk Assessment
61
imals or limited human evidence, whereas the IARC requires both. Despite the uncertainties of cross-species extrapolations, the IARC concluded
that “In the absence of adequate data on humans, it is biologically plausible and prudent to regard agents and mixtures for which there is sufficient evidence of carcinogenicity in experimental animals as if they presented a carcinogenic risk to humans” (International Agency for Research
on Cancer, 1999). Agents identified as known or probable human carcinogens are candidates for the next steps of quantitative risk assessments.
Dose–Response and Exposure Assessment
Under its 1986 carcinogen risk assessment guideline, the EPA adopted the
health protective position that, in the absence of knowledge of mode of
action, a linear dose–response relationship should be assumed at exposure levels between zero and the lowest levels observed in human and
animal studies. The guidelines indicated a preference for human dose–
response data or, in their absence, data from animal species that respond
most like humans.
The EPA recently adopted new carcinogen risk assessment guidelines
(revised in 1999) but noted that they may be further revised, depending
on evolving science or comments from peer reviewers, the public, or others; the guidelines include these specific provisions to better protect the
fetus, infants, and children (U.S. Environmental Protection Agency, 1999,
2001): (1) risk assessors should evaluate the relevance and applicability
to children of a postulated mode of action (e.g., based on age-related differences in uptake, metabolism, and excretion of an agent), (2) an RfD or
RfC approach could be considered when there is sufficient evidence of a
biologic threshold4; low-dose linearity, however, should be assumed for
children if there is no strong evidence that children and adults are comparable with regard to a postulated mode of action, (3) the dose should
be based on weight (not surface area)—this provides increased protection
in extrapolations from animals to humans, and (4) dose–response assessment should include evaluation of risks by tumor type and the relevance
of each tumor type to children and other sensitive subpopulations.
These provisions are justified by the known increased cancer risks of
children after perinatal exposure to ionizing radiation and prenatal DES
exposure. Supporting evidence comes from rodent studies showing that
combined perinatal and adult carcinogen exposure often produces a higher
tumor incidence and a shorter latent period than adult exposure alone.
4 The
EPA’s Science Advisory Board recommended inclusion of criteria to evaluate if
the evidence on mode of action is valid.
62
CHILD HEALTH
AND THE
ENVIRONMENT
Risk Characterization
Risk characterization should summarize information from hazard identification, dose–response assessment, exposure assessment, and risk modeling. This includes toxicokinetics, genotoxicity, structure–activity relationships of chemicals and properties of microbial agents (e.g., genomic
integration) relevant to carcinogenicity, effects in model systems, and numerical risk estimates. The last, based on assumptions of low-dose linearity and constant lifetime exposure, should include (1) unit risk—the
excess lifetime cancer risk per unit dose of carcinogen, (2) dose at a given
risk level—carcinogen dose that corresponds to a given risk, and (3) population risk—the excess lifetime cancer risk at a specified carcinogen dose
in an exposed population.
Important aspects of carcinogen risk characterization in the 1999 EPA
draft guidelines include the following:
• The assigned level of evidence should describe both the likelihood of
human carcinogenicity and the conditions under which cancer risks
would arise; for example, an agent could be rated as a likely carcinogen when inhaled but not when ingested.
• Increased weight should be given to an agent’s properties, including mode
of action and structure–activity relationships to known carcinogens.
These changes assume that increased knowledge of the mode of action at
cellular and subcellular levels, as well as toxicokinetic and metabolic processes, justifies inferences that go beyond available data on the relationships between exposures and risks of cancer in humans and animals. For
instance, it has been argued that chloroform is likely a threshold carcinogen; if this conclusion is accepted, the result could be less stringent
regulations for exposures related to occupation and drinking water.
Developmental Toxicants
Developmental toxicants are agents that cause preimplantation loss, early
fetal death (spontaneous abortion), late fetal death (stillbirth), impaired
growth, structural abnormalities (birth defects), and functional deficits
(e.g., mental retardation) (National Academy of Sciences, 2001). A wider
definition includes adverse effects detected at any point in the life span
of the organism caused by exposure of either parent (preconceptual), the
fetus (in utero), or the child until sexual maturity (International Programme on Chemical Safety, 2001; U.S. Environmental Protection Agency,
1991). About 3% of newborn infants have major developmental disorders
characterized as life-threatening, requiring major surgery, or presenting a
significant disability (National Academy of Sciences, 2000).
Risk Assessment
63
Epidemiologic studies have demonstrated associations and exposure–
risk relationships between high-level exposures to environmental contaminants (e.g., methylmercury, PCBs contaminated by dioxins and
furans, ionizing radiation) and adverse developmental effects. The proportions of adverse human developmental outcomes attributable to environmental hazards, however, are uncertain because of limited epidemiologic research in this field and the difficulty of demonstrating
developmental effects of perinatal exposure at background contaminant
levels. The fact that known human developmental toxicants generally
cause similar effects in at least one experimental animal species tested indicates the relevance of animal models to human health. Experimental
and epidemiologic studies have shown the importance of the timing of
exposure during pregnancy; for example, exposure to toxicants during
early gestation can cause birth defects.
The EPA guideline for developmental toxicity risk assessment provides for the use of available animal and human data and information
from pharmacokinetic and structure–activity studies. The level of evidence is assessed using categories virtually identical to those shown in
Table 3–6 for reproductive toxicity risk assessment (the difference being
that minimum evidence of no hazard in animals requires data from studies in at least two species showing no developmental toxicity at doses that
were minimally toxic to the adult). The assumption that a biologic
TABLE 3–6. EPA Categorization of Overall Weight of Evidence for Human
Reproductive Hazards
Level of Evidence
Criteria
Sufficient human
evidence of hazard
Convincing epidemiologic evidence; biologic plausibility
Sufficient animal
and/or limited
human evidence
of hazard
Positive findings in animals in at least one well-executed
study in one species (e.g., one that meets the EPA’s
guidelines for two-generation studies) and/or limited
epidemiologic evidence
Insufficient evidence
of hazard
Evidence ranges from nonexistent to animal or human
studies that have a limited study design or conduct (e.g.,
low statistical power, limited array of adverse outcomes,
inadequate dose selection or exposure information,
uncontrolled confounders)
Sufficient evidence
of no hazard
Minimum evidence of no hazard requires data on an adequate
array of outcomes from more than one study with two
species showing no adverse reproductive effects at doses
that were minimally toxic; may be supplemented by human
data showing no apparent hazard
Source: U.S. Environmental Protection Agency (1996).
64
CHILD HEALTH
AND THE
ENVIRONMENT
dose–response threshold generally exists for developmental toxicants may
not apply to all toxicants; for example, it may be appropriate to use a linear low-dose extrapolation for developmental toxicants that act through
a genotoxic mechanism (Kimmel, 2001). Also, the spectrum of developmental abnormalities may vary with the dose because fetal death at high
doses may preclude expression of effects seen at lower doses; this suggests that LOELSs from animal studies that do not include an adequate
range of doses may be misleading.
Neurotoxins
Developmental neurotoxicity encompasses any adverse effects on nervous
system structure or function caused by prenatal and early childhood exposure to a toxicant (Mileson and Ferenc, 2001). Potential developmental
neurotoxic outcomes include adverse effects on autonomic, sensory, motor, and cognitive functions, social behaviors, and biologic rhythms; these
effects may be accompanied by molecular and structural changes (CorySlechta et al., 2001). An estimated 3%–8% of liveborn infants have neurodevelopmental disabilities including dyslexia, ADHD, major cognitive
deficits, and autism (U.S. Department of Health and Human Services,
1999; Weiss and Landrigan, 2000).
The nervous system differs from other organs and tissues by having
the widest variety of specialized cell functions, a blood-brain barrier that
modulates access of chemicals, and a severely limited ability to regenerate. The prefrontal cortex in humans has executive functions, including
planning and control of impulses and emotions, and does not reach adult
volume until about age 10 years; extensive remodeling of this region occurs during adolescence, including a 30%–40% reduction in synaptic density. Known developmental neurotoxins in humans and experimental
animals include lead, methylmercury, ionizing radiation, and PCBs. Examples of nervous system developmental processes disrupted by one or
more of these toxicants include differentiation, neuronal proliferation,
neurotrophic signals, neuronal migration, synaptogenesis, gliogenesis and
myelination, and apoptosis (Rice and Barone, 2000).
There are four approaches to detection of neurobehavioral effects in
humans: clinical neurologic examination, self-report checklists, performance tests, and neuropsychologic tests. Performance and neuropsychologic test deficits may be detectable at exposure levels below the generally high levels required to cause clinically apparent signs. Performance
tests have been developed to assess attention/vigilance, mathematical
processing, verbal processing, spatial processing, memory, psychomotor
skills (e.g., reaction time, manual dexterity), and multitasking. Perfor-
Risk Assessment
65
mance tests and infants tests applied before age 1 year are markedly less
affected by socioenvironmental influences than are IQ and school achievement tests. Standardized tests of neuropsychologic development during
infancy, childhood, and adolescence include (Fiedler et al., 1996; Rice and
Barone, 2000)
• Neonatal Behavioral Assessment Scale (NBAS)—tests young infants for
neuromuscular and motor reflexes, muscle tone, activity level, reaction
to sensory stimuli, and autonomic function. It is sensitive to effects of
prenatal exposure to PCBs, opiates, and certain pharmaceuticals but is
not predictive of neurobehavioral function during later childhood, and
no norms exist.
• Bayley Scales of Infant Development—an apical test, that is, successful
responses require the integrity of multiple elements of cognitive and fine
motor function, attention to the task, and motivation to perform. The test
is sensitive to the effects of prenatal exposure to alcohol, lead, PCBs, and
methadone but is nonspecific (standardized norms are available).
• Fagan Test of Infant Intelligence (visual recognition memory test)—
assesses the ability of infants to remember visual images and is moderately predictive of IQ measured later in childhood; it is sensitive to
effects of prenatal exposure to PCBs, methylmercury, and alcohol.
• McCarthy Scales of Children’s Abilities, Wechsler Preschool and Primary Scale of Intelligence—Revised, Wechsler Intelligence Scale for
Children III, Stanford-Binet, and Kaufman Assessment Battery for
Children—these tests assess older children’s neurologic and cognitive
development including psychomotor, memory, verbal, and perceptual
functions. They are sensitive to the effects of perinatal exposure to lead,
alcohol, and PCBs (standardized norms exist)
Until very recently, neurotoxicity testing required by regulatory agencies
focused almost exclusively on obvious neurologic dysfunction in adult
animals associated with general neuropathology, with little or no attention to neurobehavioral effects and developmental neurotoxicity. The EPA
has obtained adequate developmental neurotoxicity testing data on only
a few pesticides and industrial chemicals. Even for chemicals tested for
neurotoxicity in adult animals, risk assessment is virtually limited to qualitative hazard identification and to the early stage of risk characterization
because of insufficient data to support quantitative risk evaluations (Landrigan et al., 1994).
Problems arise in extrapolating the results of developmental neurotoxicity tests from animals to humans. The human brain has much more
prenatal development than the rodent brain, important differences in the
66
CHILD HEALTH
AND THE
ENVIRONMENT
relative rates at which specific brain regions develop, a vastly larger and
functionally more complex prefrontal cortex, and a much larger proportion of cortex devoted to vision including color discrimination, spatial resolution, and higher-order processing and integrating functions. The rodent brain devotes a much larger proportion of the cortex to olfactory
functions.
Reproductive Toxicants
Reproductive toxicity includes adverse effects of substances on sexual
maturation at puberty, gamete production and transport, the female reproductive cycle, sexual behavior, fertility, gestation, parturition, lactation, and reproductive senescence (International Programme on Chemical Safety, 2001; National Academy of Sciences, 2001). The EPA guideline
for reproductive toxicity risk assessment specifies level of evidence categories to evaluate potential reproductive hazards (Table 3–6). Positive
findings from a single study of high quality would generally provide sufficient animal evidence of reproductive hazard. Confidence in positive
findings is increased if there is evidence of a dose–response relationship,
biologic plausibility, consistent results from multiple studies, explanatory
evidence for discordant results, appropriate exposures (route, level, duration, frequency), an adequate array of observed outcomes, and appropriate statistical power and analyses.
Although about 10% of human couples experience increased time to
pregnancy or infertility, there has been relatively little epidemiologic research on early life environmental exposures and adverse reproductive
effects. DES is the only known human reproductive toxicant that acts in
utero. The few known environmental reproductive toxicants in humans
were discovered from studies of occupationally exposed men and include
lead, the pesticides 1,2-dibromo-3-chloropropane, chlordecone, and ethylene dibromide, and certain solvents (see later chapters for more details).
Animal models may be insensitive indicators of hazards to human male
fertility; the average number of sperm per ejaculate in human men is only
about two to four times that associated with impaired fertility, while rodent sperm counts are up to 1000-fold those linked with maximum fertility. Nevertheless, existing evidence of hazards from perinatal exposures
comes mainly from animal studies.
Reproductive toxicity dose–response assessments have generally
used curve-fitting models that have biologic plausibility but do not incorporate the mode of action. The EPA guideline, however, calls for use
of biologic information to assess the shape of the dose–response relationship below the observable range. For instance, a dose threshold for
Risk Assessment
67
reproductive effects may be supported by evidence of host defense mechanisms such as the testicular Sertoli cell barrier. The RfDs or RfCs for reproductive toxicants are based on the NOAEL, LOAEL, or BMD for the
adverse reproductive effect most sensitive to the toxicant of interest (or
the most sensitive mammalian species) and application of uncertainty factors. The dose–response curves for the antiandrogenic effects from perinatal exposure of rodents to vinclozolin vary widely, with several showing antiandrogenic effects at doses well below the NOAEL observed in
Tier I toxicity testing and no apparent threshold; such findings support
alternatives to NOAELs such as a BMD.
Conclusion
The 1983 National Academy of Sciences report on risk assessment provided a useful framework for integrating toxicologic and epidemiologic
data into quantitative estimates of human health risks at low exposure
levels. Importantly, it recommended separating the scientific process of
risk assessment from the political evaluation of its socioeconomic implications. The National Academy of Sciences report noted the pervasive uncertainties in risk assessments and attributed them mainly to knowledge
gaps, a finding that remains true two decades later. There has been a tendency in recent years to include cost considerations in risk assessments
and to give increased weight to the mode of action of toxicants, particularly carcinogens, as opposed to the results of animal bioassays. The 1983
report also stated that protection of public health would be better served
if scientists improved their communication of risk assessment findings to
decision makers and the public rather than factoring political considerations (such as costs) into risk assessments.
The four categories of risk assessment described in this chapter illustrate several issues and challenges: inadequate developmental toxicity
and developmental neurotoxicity testing of commercial chemicals, uncertainties surrounding recommended exposure limits, and the important
role of epidemiologic studies in carcinogen and neurotoxicity risk assessments but their weaker role in reproductive and developmental toxicity risk assessments (because of a dearth of epidemiologic research in
the latter areas). These uncertainties argue for substantial strengthening
of epidemiologic and toxicologic research, including postmarketing studies to ensure early detection of unexpected adverse health effects of existing and new products.
The reader is referred to further information on risk assessment from
these national and international agencies:
68
CHILD HEALTH
AND THE
ENVIRONMENT
• International Programme on Chemical Safety—established in 1980 by
collaboration of the International Labour Office, the United Nations
Environment Programme, and the World Health Organization to evaluate environmental chemical hazards to human health: http://www.
who.int/pcs
• OECD. (2002). Located at http://www.oecd.org
• U.S. EPA: http://cfpub1.epa.gov/ncea/cfm/nceahome.cfm
• Agency for Toxic Substances and Disease Registry: http://www.atsdr.
cdc.gov
• U.S. National Academy Press publications: http://www.nap.edu/
References
Agency for Toxic Substances and Disease Registry. (2002). Minimal risk levels
(MRLs) for hazardous substances. Located at http://www.atsdr.cdc.gov/
mrls.html
Centers for Disease Control and Prevention. (2001). National report on human exposure to environmental chemicals. Atlanta: Centers for Disease Control and
Prevention.
Cory-Slechta DA, Crofton KM, Foran JA, Ross JF, Sheets LP, Weiss B, Mileson B.
(2001). Methods to identify and characterize developmental neurotoxicity for
human health risk assessment. 1. Behavioral effects. Environ Health Perspect
109(Suppl 1):79–91.
Elwood JM. (1998). Critical appraisal of epidemiological studies and clinical trials. New York: Oxford University Press.
Fiedler N, Feldman RG, Jacobson J, Rahill A, Wetherell A. (1996). The assessment
of neurobehavioral toxicity: SGOMSEC joint report. Environ Health Perspect
104(Suppl 2):179–91.
Hertz-Picciotto I. (1995). Epidemiology and quantitative risk assessment: a bridge
from science to policy. Am J Public Health 85:484–91.
Hill AB. (1965). The environment and disease: association or causation? Proc R
Soc Med 58:295–300.
International Agency for Research on Cancer. (1999). Preamble to the IARC monographs. Lyon: International Agency for Research on Cancer.
International Programme on Chemical Safety. (1999). Environmental health criteria 210: Principles for the assessment of risks to human health from exposure
to chemicals. Geneva: International Programme on Chemical Safety.
International Programme on Chemical Safety. (2001). Environmental health criteria No. 225. Principles for evaluating health risks to reproduction associated
with exposure to chemicals. Geneva: International Programme on Chemical
Safety.
Kaufman JS, Poole C. (2000). Looking back on “causal thinking in the health sciences.” Annu Rev Public Health 21:101–19.
Kimmel CA. (2001). 1999 Warkany lecture: improving the science for predicting
risks to children’s health. Teratology 63:202–9.
Risk Assessment
69
Krewski D, Crump KS, Farmer J, Gaylor DW, Howe R, Portier C, Salsburg D,
Sielken RL, Van Ryzin J. (1983). A comparison of statistical methods for low
dose extrapolation utilizing time-to-tumor data. Fundam Appl Toxicol 3:
140–60.
Landrigan PJ, Graham DG, Thomas RD. (1994). Environmental neurotoxic illness:
research for prevention. Environ Health Perspect 102(Suppl 2):117–20.
Mileson BE, Ferenc SA. (2001). Methods to identify and characterize developmental neurotoxicity for human health risk assessment: overview. Environ
Health Perspect 109(Suppl 1):77–8.
National Academy of Sciences. (1983). Risk assessment in the federal government:
managing the process. Washington, D.C.: National Academy Press.
National Academy of Sciences. (2000). Scientific frontiers in developmental toxicology and risk assessment. Washington, D.C.: National Academy Press.
National Academy of Sciences. (2001). Evaluating chemical and other agent exposures for reproductive and developmental toxicity. Washington, DC: National Academy Press.
Rice D, Barone S Jr. (2000). Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models. Environ Health Perspect 108(Suppl 3):511–33.
Rothman KJ, Greenland S. (1998). Modern epidemiology. Philadelphia: LippincottRaven.
The Presidential/Congressional Commission on Risk Assessment and Risk Management. (1997a). Framework for environmental health risk management, Volume I (stock number 055-000-00567-2). Washington, DC: Government Printing
Office.
The Presidential/Congressional Commission on Risk Assessment and Risk
Management. (1997b). Risk assessment and risk management in regulatory
decision-making, Volume II (stock number 055-000-00568-1). Washington, DC:
Government Printing Office.
U.S. Department of Health and Human Services. (1999). Mental health: a report
of the surgeon general. Rockville, MD: U.S. Department of Health and Human
Services.
U.S. Environmental Protection Agency. (1986). Guidelines for carcinogen risk assessment. Fed Reg 51:33992-4003.
U.S. Environmental Protection Agency. (1991). Guidelines for developmental
toxicity risk assessment. Fed Reg 56:63798-826.
U.S. Environmental Protection Agency. (1996). Guidelines for reproductive toxicity risk assessment. Fed Reg 61:56274-322.
U.S. Environmental Protection Agency. (1999). Guidelines for carcinogen risk assessment (NCEA-F-0644). Washington, DC: U.S. Environmental Protection
Agency.
U.S. Environmental Protection Agency. (2000a). Exposure and human health reassessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and related compounds. Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2000b). Integrated risk information system. Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2001). Notice of opportunity to provide
additional information and comment. Fed Reg 66:59593-4.
70
CHILD HEALTH
AND THE
ENVIRONMENT
U.S. General Accounting Office. (2001). Chemical risk assessment. Selected federal agencies’ procedures, assumptions, and policies (GAO-01-810). Washington, DC: U.S. General Accounting Office.
U.S. Surgeon General. (1964). Smoking and health. (Public Health Service Publication No. 1103). Washington, DC: Public Health Service, U.S. Department of
Health, Education, and Welfare.
Weiss B, Landrigan PJ. (2000). The developing brain and the environment: an introduction. Environ Health Perspect 108(Suppl 3):373–4.
Wilkinson CF, Christoph GR, Julien E, Kelley JM, Kronenberg J, McCarthy J, Reiss
R. (2000). Assessing the risks of exposures to multiple chemicals with a common mechanism of toxicity: how to cumulate? Regul Toxicol Pharmacol 31:
30–43.
4
Metals—Lead
The twentieth century saw greatly expanded use and environmental dispersion of lead, a dense metal valued since antiquity because of its low
melting point, pliability, and durability. Analysis of a Swiss peat bog
showed that annual average lead deposition increased 1600-fold between
5000 B.C. and the maximum in 1979. Recognition of adult lead poisoning
with abdominal colic can be traced to Hippocrates in about 370 B.C. (in
metal workers) and Baker in 1767 (who linked Devonshire colic to consumption of lead-contaminated cider) (Table 4–1). Childhood lead poisoning was recognized as a distinct entity in 1892 and neurotoxicity in
experimental animals by the 1920s. There has been substantial progress
in reducing childhood lead exposure, but many children remain at risk.
The objective of this chapter is to illustrate how failure to apply the
precautionary principle allowed inappropriate uses of lead, massive environmental contamination, and major adverse impacts on child health.
The first section focuses mainly on the susceptibility of the developing
human nervous system to adverse neurobehavioral effects from relatively
low-level lead exposure, as evidenced by epidemiologic and toxicologic
studies. The discussion includes environmental indices and biomarkers
of lead exposure and toxicokinetics. The risk management section addresses lead sources (air, water, food, soil/dust) and interventions for preventing childhood lead exposure.
71
CHILD HEALTH
72
AND THE
ENVIRONMENT
TABLE 4 –1. Selective Chronology of Lead (82Pb207)
370 B.C.
1767
1892
1909
1923
1943
1970
1975
1978
1979
1976–80
1986
1987
1991
1994
1991–4
1999
2000
Hippocrates attributed abdominal colic in a metal worker to lead exposure
Devonshire colic attributed to lead-contaminated cider (Sir George Baker,
England)
Childhood poisoning from lead-based paint reported in Australia
France, Belgium, and Austria ban white lead interior paints
First public sale of leaded gasoline (Dayton, Ohio)
Lead poisoning shown to have persistent neurotoxic effects in children
U.S. Clean Air Act establishes the EPA; “acceptable” blood lead level
reduced from 60 to 40 g/dL
CDC investigation shows that lead emissions in air are a health hazard for
children; first catalytic converters used to reduce motor vehicle emissions
EPA bans lead-based paint and lead-based paint products; blood lead action
level reduced from 40 to 30 g/dL
Deficits in psychologic and classroom performance linked to moderate lead
levels in deciduous teeth
NHANES II shows that mean blood lead declined by about 40% in all age,
sex, and racial groups during this period in step with declining use of
leaded gasoline
EPA reduced drinking water lead standard from 50 to 5 g/L and banned
use of lead in plumbing solder and fixtures
Elevated hearing threshold discovered at blood lead levels as low as 10 g/dL
Lead solder in food and beverage cans banned; action level for blood lead
reduced to 10 g/dL
Meta-analyses showed IQ deficits in children with blood lead levels of
10–20 g/dL; United Nations calls on governments to eliminate lead from
gasoline worldwide
NHANES III shows geometric mean blood lead level of 2.7 g/dL
(CI 2.5–3.0) and prevalence of levels of 10 g/dL of 4.4%, among
children aged 1–5 years
NHANES 1999 data show that geometric mean blood lead level of children
aged 1–5 years was 2.0 g/dL (CI 1.7–2.3)
Neurobehavioral effects in children at blood lead levels below 5 g/dL
Health Effects
Acute childhood lead poisoning is now rare in developed countries, but
widespread low-level exposure continues. Low-level or moderate lead exposure during early childhood causes persistent adverse neurobehavioral
effects including cognitive deficits. Relatively subtle neurobehavioral
deficits occur at blood lead levels below 5 g/dL (Lanphear et al., 2000)
but clinically obvious symptoms occur only at levels exceeding about
50 g/dL (Table 4–2). Encephalopathy with delirium, convulsions, paralysis, coma, and death may occur at blood lead levels above 80 g/dL,
Metals—Lead
73
TABLE 4 –2. Health Effects of Lead in Children by Blood Lead Level
Blood lead
(g/dL)
125
80
60
20
15
10
10
Health Effects
Acute encephalopathy, death
Encephalopathy, renal toxicity (aminoaciduria)
Colic
Anemia, peripheral neuropathy, reduced nerve conduction velocity
Increased zinc protoporphyrin, impaired vitamin D activation
Growth deficits
IQ and hearing deficits, inhibition of ALAD and pyrimidine-5-nucleotidase
Source: Agency for Toxic Substances and Disease Registry (1999).
indicating that the margin between blood lead levels sufficient to cause
obvious symptoms and those that can cause massive brain damage or
death is only about a factor of 2.
Toxic Mechanisms
Rapid development of the human brain before age 3 years involves a complex cascade of processes including cell division and migration, differentiation of neuroblasts to neurons with axons, dendrites, synapses, and
neurotransmitter systems, and programmed cell death (apoptosis). This
section describes some of the mechanisms through which lead can disrupt brain development.
Genetic Susceptibility
There is no clear evidence of genetic susceptibility to lead toxicity, but genetic polymorphisms may play a role. ␦-Aminolevulinic acid dehydratase
(ALAD), the vitamin D receptor gene (VDR), and the hemochromatosis
gene (HFE—encodes the hereditary hemochromatosis protein), respectively, control heme synthesis, gastrointestinal calcium absorption, and
iron transport. At low to moderate lead exposure levels, over 99% of lead
in blood is firmly bound to red cell proteins, mainly ALAD. Environmentally exposed ALAD-2 homozygotes and heterozygotes have higher
blood lead levels than ALAD-1 homozygotes (Shen et al., 2001), possibly
because of a higher affinity of the ALAD-2 protein for lead. But no firm
evidence exists for an association between ALAD genotype and susceptibility to lead toxicity at background exposure levels, precluding its use
as a risk marker for lead screening programs. Calcitriol, the activated form
of vitamin D, binds to nuclear vitamin D receptors in intestinal, kidney,
and bone cells and activates transcription of genes that encode calcium-
74
CHILD HEALTH
AND THE
ENVIRONMENT
binding proteins; these include calbindin-D, an intestinal cell transport
protein that promotes calcium and lead absorption. Among men occupationally exposed to lead, carriers of the B allele of VDR had higher blood
and bone lead levels, suggesting that they had higher lead absorption capacity than noncarriers; there appear to be no reports on this relationship
in children (Schwartz et al., 2000).
Normal HFE protein controls gastrointestinal iron absorption by
binding to the transferrin receptor and reducing its affinity for iron-loaded
transferrin. The net impact of HFE mutations is increased gastrointestinal iron absorption; lead absorption may also be increased, but direct evidence is lacking. The role of HFE polymorphisms in susceptibility to lead
toxicity remains hypothetical but, given a prevalence of 4% in NHANES
III, potentially important.
Lead–Protein Interactions
The mechanisms of lead neurotoxicity are incompletely understood, but
its ability to compete with iron, zinc, and calcium at binding sites on proteins and disrupt their function may be fundamental. Even at low levels,
lead binds almost irreversibly to zinc-dependent proteins including gene
transcription factors, membrane ion-transport enzymes, intracellular signaling enzymes, and ALAD. By inhibiting ALAD, lead reduces production of heme and heme-dependent proteins (hemoglobin, cytochromes,
myoglobin), causes accumulation of ␦-aminolevulinic acid (␦-ALA, a neurotoxin at high levels), and impairs oxygen transport and storage, mitochondrial energy production, and P450 detoxification systems. Tissuespecific low molecular weight proteins bind lead, a property that may
reduce lead neurotoxicity but may also facilitate lead uptake and carcinogenesis in kidney cells.
Synapse Formation and Function
In the absence of synapse formation, neurons fail to obtain adequate
amounts of neurotrophins, antiapoptotic proteins produced by target tissues; approximately half of the neurons produced during development normally do not form synapses and undergo apoptosis. Lead reduces the likelihood of synapse formation by disrupting sialylation or desialylation of
neuronal cell adhesion molecules (NCAMs), and functioning of NMDA receptors, protein kinase C, and calcium-dependent ion channels. The
NCAMs are cell surface glycoproteins that mediate adhesion between neurons and between neurons and muscle cells. Glycosylation of NCAM by
the enzyme sialyltransfererase (ST) soon after training periods appears to
be a key mechanism for synapse formation and memory in rodents (Davey
and Breen, 1998). Extremely low levels of lead induce ST activity in neu-
Metals—Lead
75
rons, causing inappropriate glycosylation of NCAM that may contribute to
subsequent learning deficits in adult rats. Lead also inhibits dendrite formation and branching essential steps in synapse formation.
Neuronal plasticity, the ability of synaptic efficacy to change in response to neural activity, is a likely biologic mechanism for learning. Synaptic efficacy may undergo long-term depression (LTD) induced by low
levels of synaptic activity or long-term potentiation (LTP) induced by the
opposite conditions. Excitatory amino acids (glutamic and aspartic acids)
and their NMDA receptors play major roles in activity-dependent synaptic plasticity and in the stabilization and elimination of synaptic connections during development. Lead interferes with the NMDA-type receptor and raises the threshold for LTP, a mechanism that may partially
explain learning deficits from lead exposure (Guilarte, 1997).
By competing with calcium, lead disrupts calcium-dependent ion
channels, intracellular signaling enzymes such as protein kinase C, and
NMDA receptors and alters synaptic transmission. At picomolar levels,
lead activates protein kinase C and diminishes the synaptic signal-to-noise
ratio by stimulating spontaneous and inhibiting evoked presynaptic
neurotransmitter release (Johnston and Goldstein, 1998). These low-dose
effects may disrupt brain endothelial cell differentiation, blood-brain barrier development, learning, and memory.
Lead Absorption
Bioavailability of ingested lead varies by chemical form and amount of
lead, diet (intake of calcium, iron, phosphate, vitamin D, and fat), age,
and pregnancy status. Adults absorb 10%–15% of lead ingested with
meals, but children and pregnant women can absorb up to 50%. Because
lead absorption at intestinal calcium binding sites is inhibited by dietary
calcium, children and pregnant women with low calcium intakes have increased lead absorption. Low calcium intake induces activation of vitamin D to 1,25-dihydroxyvitamin D in kidney and calcium-binding protein (calbindin-D) in intestinal cells, thus increasing absorption of calcium,
lead, and other trace metals; lead, in turn, inhibits renal vitamin D activation. Iron deficiency also appears to increase duodenal lead absorption.
Maternal bone lead appears to be mobilized during pregnancy and lactation, particularly among women who do not take calcium supplements
(Gulson et al., 1998).
Neurotoxicity
Neurotoxicity of lead in children was first described in Australia (Gibson,
1892). This nineteenth-century study was remarkable for the use of post-
76
CHILD HEALTH
AND THE
ENVIRONMENT
card questionnaires sent to parents and the use of chemical analysis to
measure lead concentrations in paint samples. Gibson concluded that the
affected children lived in houses with paint that had oxidized to an easily removed powder containing high lead concentrations, had ingested
sufficient lead to cause poisoning, and should be removed from their
homes. Turner (1897) identified four clinical categories of moderate to severe childhood lead poisoning: (1) symmetrical paralysis of extensor muscles of the hands and feet, (2) abdominal colic, (3) epilepsy, and (4) ocular neuritis with paralysis of eye muscles. Current knowledge indicates
that low to moderate childhood lead exposure can cause behavioral abnormalities and deficits in global intelligence, short-term memory, reading, spelling, hearing sensitivity, visuomotor performance, perception integration, and reaction time (Agency for Toxic Substances and Disease
Registry, 1999; U.S. Environmental Protection Agency, 2001a).
Cognitive Function
As recently as 1972, the National Academy of Sciences stated that emissions of lead into air had no known harmful effects. In the same year,
ironically, the (CDC) was investigating the health of persons living near
a large lead smelter in El Paso, Texas. House dust and blood lead levels
were strongly associated with proximity to the smelter. Among children
living within 1.6 km of the smelter, 69% had blood lead levels of at least
40 g/dL; this group had significant IQ deficits and reduced finger-wrist
tapping speed compared to less exposed children (Landrigan et al., 1975a,
1975b). Subsequent studies of children living near other lead smelters produced similar findings.
A cross-sectional study of young children in Boston produced the
first strong evidence of neurotoxicity from moderate lead exposure during childhood (Needleman et al., 1979). Cumulative exposure since birth
was assessed by measuring the lead concentration in deciduous teeth,
thus avoiding reliance on blood lead, an indicator of both recent and
longer-term exposure. Children with high dentin lead levels had a mean
full-scale IQ deficit of 4.5 points (compared to those with low dentin lead
levels), with lower scores on verbal, auditory, speech, and attention subscales. When reassessed at age 18 years, the high-exposure group had
greatly increased risks of reading disability and high school dropout
(Needleman et al., 1996).
Several longitudinal and numerous cross-sectional studies have confirmed that low and moderate lead exposure during childhood causes
adverse neurobehavioral outcomes. The evidence has been summarized
in meta-analyses (Needleman and Gatsonis, 1990; Pocock et al., 1994;
Schwartz, 1994) and literature reviews (Agency for Toxic Substances and
Metals—Lead
77
Disease Registry, 1999; Banks et al., 1997; Thacker et al., 1992; U.S. Environmental Protection Agency, 1998). Major conclusions from these reports
include the following:
• A doubling of blood lead from 10 to 20 g/dL, or of tooth lead from 5
to 10 g/g, appears to cause an average full-scale IQ deficit of 1–3 points.
• Full-scale IQ deficits among school-age children were more consistently
associated with blood lead levels at about age 2 years in cohort studies and tooth lead levels in cross-sectional studies, compared to current
blood lead levels in cross-sectional studies
A reduction in average full-scale IQ of 5 points would reduce the proportion of very gifted people (IQ 130) in a population by a factor of
about 2.5 and double the proportion of mentally retarded persons (IQ
70) (Weiss, 1990). The predicted impact of a 10 g/dL increment in blood
lead at age 2 years in the Boston longitudinal study of middle-class children was an average full-scale IQ deficit at age 10 years of about 6 points,
independent of potential confounders (Bellinger et al., 1992). In the Port
Pirie smelter town longitudinal study, IQ at age 7 years was most closely
associated with average blood lead levels from birth to any age between
15 months and 4 years; the predicted impact of an increase in average
blood lead at ages 1–4 years from 10 to 30 g/dL was a full-scale IQ deficit
of 4–5 points (Baghurst et al., 1992). Of note was the particularly strong
relationship between lead and WISC subscale deficits related to perceptual organization and synthesis, spatial visualization, nonverbal concept
formation, and visuomotor coordination. Further follow-up of the Port
Pirie cohort showed that (1) children in the highest tertile of blood lead
at ages 2–4 years had cognitive deficits that persisted to at least age 11–
13 years even though their blood lead levels had declined substantially,
(2) there was a dose–response relation between lead and full-scale IQ,
with no evidence of a threshold, and, (3) children with lifetime average
blood lead levels above 15 g/dL had an increased prevalence of problem behaviors (see, e.g., Burns et al., 1999).
The importance of lead exposure during early childhood and the
peaking of blood lead levels at about age 2–3 years likely explain the
greater consistency in cross-sectional studies of older children that used
tooth lead rather than current blood lead to assess exposure. The crosssectional study of almost 5000 children in NHANES III (1988–1994) is
noteworthy in that performance scores for arithmetic, reading, nonverbal
reasoning, and short-term/working memory were inversely related to
blood lead levels even in the range below 10 g/dL (Lanphear et al., 2000).
The belief that the neurotoxic effects of even high-level lead exposure were reversible was first challenged by a follow-up study of children
78
CHILD HEALTH
AND THE
ENVIRONMENT
previously treated for lead poisoning demonstrating persistent poor
school performance, impulsive behavior, and short attention span (Byers
and Lord, 1943). The authors hypothesized that persistent neurotoxicity
was related to prolonged release of lead from stores in bone or to irreversible damage to the developing nervous system. A recent review of
five longitudinal studies concluded that even low to moderate early childhood lead exposure is associated with neurobehavioral deficits that persist after exposure levels decline (Tong, 1998).
Other factors associated with poverty can cause IQ deficits: malnutrition, iron deficiency, certain infections, head injury, and inferior prenatal care. Several studies, however, were able to adjust for such potential
confounders and still showed a lead effect. The biologic plausibility of
neurotoxicity from low-level lead exposure is supported by experiments
in rats and nonhuman primates; studies of monkeys exposed only prenatally or postnatally showed learning and memory deficits at blood lead
levels as low as 10 g/dL and learning and memory deficits that persisted years after exposure cessation (Rice, 1992).
Sensorimotor Effects
Children with blood lead levels less than 20 g/dL exhibited visualevoked potential abnormalities (Altmann et al., 1998), and those with levels below 10 g/dL had elevated hearing thresholds (Osman et al., 1999).
Lead causes retinal degeneration with apoptotic loss of rods and bipolar
cells in both neonatal and adult rats at blood lead levels below 20 g/dL.
Moderate lead exposure during early childhood is associated with motor
deficits including reduced finger-wrist tapping speed, motor nerve conduction velocity (with a possible threshold at blood lead levels of 20–
30 g/dL), and fine motor deficits (Schwartz et al., 1988).
Behavioral Problems
Children with high tooth lead levels are more likely to be distractible, dependent, hyperactive, easily frustrated, and unable to follow directions
(Needleman et al., 1979). Prenatal and childhood lead exposure indices
have been linked to delinquent and antisocial behaviors during adolescence in longitudinal studies (Burns et al., 1999; Dietrich et al., 2001).
Other Effects
Developmental Effects
A review of epidemiologic studies concluded that prenatal lead exposure
likely increases the risk of preterm delivery and is inconsistently associated with reduced birth weight (Andrews et al., 1994). More recent stud-
Metals—Lead
79
ies have provided limited evidence that birth weight is inversely related
to maternal lead exposure (Gonzalez-Cossio et al., 1997; Irgens et al., 1998)
and possibly to paternal lead exposure (Lin et al., 1998). Stature and head
circumference during early childhood appear to be inversely related to
current blood lead levels in cross-sectional and longitudinal studies (see,
e.g., Ballew et al., 1999; Rothenberg et al., 1999). There is also some evidence that lead exposure during early childhood may increase the risk of
obesity during late adolescence (Kim et al., 1995).
Historic and modern studies of low/moderate and high-level paternal or maternal lead exposure have produced limited evidence of an association with spontaneous abortion (Hertz-Picciotto, 2000). Of note was
the strong exposure–risk relationship extending to maternal first trimester
blood lead levels below 15 g/dL in a recent nested case-control study
(Borja-Aburto et al., 1999). There is inadequate evidence that parental occupational lead exposure increases the risk of stillbirth. Limited evidence
indicates associations between parental occupational or environmental
lead exposure and oral cleft, neural tube, and cardiac defects (Aschengrau
et al., 1993; Irgens et al., 1998; Kristensen et al., 1993; Vinceti et al., 2001).
Blood, Renal, and Immunologic Effects
Hemoglobin levels are reduced in young children with moderate or high
lead exposure, independent of iron deficiency. Iron deficiency anemia facilitates gastrointestinal lead absorption; children from low-income families, therefore, are at increased risk of anemia from iron deficiency and
both increased lead exposure and absorption. A recent review of lead
nephrotoxicity concluded that low-level lead exposure is associated with
increased urinary excretion of low molecular weight proteins and lysosomal enzymes, but the relation between these sensitive signs of renal
tubular damage and the risk of renal disease is uncertain (LoghmanAdham, 1997). Children may be more susceptible to lead-induced urinary
protein excretion than adults (Fels et al., 1998). There is inadequate evidence to assess whether lead exposure impairs immune function.
Exposures
Internal Dose Indicators
Available biomarkers of internal dose include lead concentrations in blood
(including cord blood), urine, breast milk, hair, deciduous teeth, and other
bones. Blood, urinary, and hair lead levels reflect exposures over the past
several months, while bone (including teeth) lead concentrations indicate
CHILD HEALTH
80
AND THE
ENVIRONMENT
lifetime cumulative exposures in children. At low to moderate lead exposure levels, over 99% of lead in blood is firmly bound to red cell proteins, mainly ALAD; about 0.3%–0.7% is unbound and is the biologically
active fraction. The half-life of lead in whole blood is about 30 days.
Plasma lead is difficult to measure accurately because even slight hemolysis of red blood cells causes major artifactual increases. Whole blood
lead levels are easily and reliably measured and have been used to assess
lead exposure in most childhood blood lead screening programs and epidemiologic studies. Because lead cycles between blood and bone, blood
lead levels in older children and adults reflect both recent and past exposures. Blood lead in infants and toddlers, however, reflects recent exposure because body lead stores (especially in bone) are still low. Among
children in the general population, blood lead levels usually peak at age
2 years in parallel to exposure from soil and dust. Maternal blood lead
levels peak during the first and third trimesters; older mothers had steeper
increases in blood lead during the third trimester, consistent with mobilization from bone during late pregnancy (Hertz-Picciotto et al., 2000).
It was not until 1982 that the first population-based prevalence data
were available in the United States using blood samples collected during
NHANES II (Mahaffey et al., 1982). Among children aged 1–5 years, 98%
of black and 85% of white children had blood lead levels of at least
10 g/dL (Table 4–3). During the next decade, the percentage of children
with elevated blood lead levels decreased to 21% among blacks and 5.5%
among whites. Similar declines occurred in all subgroups defined by age,
sex, race/ethnicity, income level, and urban status. Prevalence rates of elevated blood lead levels were associated with age of housing among all
ethnic and income groups except the highest income category (Table 4–4).
The geometric mean blood lead level of children declined from 15.0 g/dL
in 1976–1980 to 2.7 g/dL in 1991–1994 and to 2.0 g/dL in 1999 (Centers for Disease Control and Prevention, 2000; Pirkle et al., 1998). The
TABLE 4 –3. Percentage of Children with Elevated Blood Lead
Levels (10 g/dL), United States, NHANES II and III
1976–80
1988–91
All persons aged 1–74 yr
77.8
4.3
All children aged 1–5 yr
88.2
8.9
All children aged 6–19 yr
71.7
2.6
Black children aged 1–5 yr
97.7
20.6
White children aged 1–5 yr
85.0
5.5
Source: Brody et al. (1994), Pirkle et al. (1994).
Metals—Lead
81
TABLE 4 – 4. Percentage of Children Aged 1–5 Years with Blood
Lead Levels 10 g/dL, United States, 1991–94
Year Housing Was Built
Pre-1946
1946–73
Post-1973
21.9
13.0
5.6
13.7
2.3
1.4
3.4
1.6
1.5
16.4
4.1
0.9
7.3
2.0
2.7
4.3
0.4
NA
11.5
5.8
5.8
3.1
0.8
2.5
Race/ethnicity
Black, non-Hispanic
Mexican American
White, non-Hispanic
Family income level
Low
Middle
High
Urban status
Pop. 1,000,000
Pop. 1,000,000
NA, not available.
Source: Pirkle et al., 1998.
prevalence of elevated blood lead levels has likely decreased substantially
among children in many other countries, but few have tracking systems
to demonstrate this.
Bone, because of its mass and high affinity for lead, contains high
proportions of the body lead burden in children (73%) and adults (94%)
(Agency for Toxic Substances and Disease Registry, 1999). The bone lead
concentration reflects long-term exposure and can be measured in deciduous teeth or by X-ray fluorescence of bones in vivo. The half-life of lead
in bone ranges from years to decades, depending on bone type, metabolic
state, and age; the half-life is shorter in children because of bone remodeling and high turnover rates during growth. Release of bone lead contributes to blood lead levels long after exposure ends.
Over the range 5–100 g/dL, lead progressively inhibits the red
cell enzyme ALAD, increasing plasma ␦-ALA levels, but such increases
can also be caused by iron deficiency at low blood lead levels. Lead also
blocks the action of the enzyme heme synthetase, leading to elevated
blood levels of erythrocyte protoporphyrin. Neither heme precursor is a
specific indicator of lead exposure. Because hair grows about 1 cm per
month and lead levels in segments of hair reflect exposures at the time
they were formed, segmental hair analysis provides information on lead
82
CHILD HEALTH
AND THE
ENVIRONMENT
exposures over time. Hair is subject, however, to external contamination
and detects only 57% of children with elevated blood lead levels, precluding its use as a screening method. Urinary lead levels correlate weakly
with total blood lead but strongly with plasma lead levels. Urinary lead
excretion during the 24 hours after challenge with a chelating agent reflects body lead burden.
Modeled Exposure Estimates
The EPA developed the Integrated Exposure Uptake Biokinetic (IEUBK)
model to estimate human lead uptake using data on lead concentrations
in multiple media (air, dust, soil, paint, food, and water) and estimated
average human consumption levels. This model assumes that lead from
all sources is absorbed via the lungs and gastrointestinal tract and enters
a blood plasma reservoir that equilibrates with lead in all tissues with
age-specific parameters for each transfer process. Biokinetic models are
subject to several uncertainties including the bioavailability of lead in various media and the prevalence and intensity of behavioral factors, particularly hand–mouth activity. Nevertheless, using environmental lead
data for four communities, modeled geometric mean blood lead estimates
for each community fell within 0.7 g/dL of directly measured blood lead
levels (Hogan et al., 1998).
Risk Management
Tetraethyl lead in gasoline and lead pigments in paint were the main
sources of lead exposure for children in the twentieth century. Other important sources included foods in lead-soldered cans and drinking water
contaminated by lead or lead alloy plumbing materials. During the midtwentieth century, the Lead Industries Association (LIA) campaigned successfully against attempts to reduce or prohibit the use of lead in paint
and plumbing materials and the use of lead arsenate on foods (Silbergeld,
1997).
Prevention
Childhood lead control policies that emphasize primary prevention of
lead exposure will ultimately reduce or eliminate the need for screening
at-risk children (Lanphear, 1998). Based on NHANES III (1991–1994), there
were about 900,000 children in the United States with blood lead levels
above the health-concern blood lead level (10 g/dL) set by the CDC
compared to about 14 million children in 1978 (Table 4–5). The U.S. strate-
Metals—Lead
83
TABLE 4 –5. Number of Children with Blood
Lead Levels of 10 g/dL, United States
Year
Number
1976–80a
14.2 million
1991–94b
0.9 million
a Age
6 months to 5 years. Source: Mahaffey et al. (1982).
b Age
1 to 5 years. Source: Pirkle et al. (1998).
gies for eliminating blood lead levels over 10 g/dl among young children by the year 2010 include (1) controlling lead paint hazards in lowincome housing, (2) expanding blood lead screening programs and follow-up services for at-risk children, (3) conducting intervention research,
and (4) tracking progress and refining strategies (President’s Task Force
on Environmental Health Risks and Safety Risks to Children, 2000).
The United States appears to be the first country that banned lead
from gasoline, paint, and plumbing. Denmark was the first European
Union country to ban most uses of lead other than lead-acid car batteries (Danish Environmental Protection Agency 2001). The estimated future
economic benefits from increased worker IQ and productivity attributable to reduced blood lead levels since 1976 among children are $110–$319
billion increased earnings over the lifetime of the U.S. cohort aged 2 years
in 2000 (in 2000 dollars and assuming a 3% discount rate), with similar
savings for each year’s birth cohort (Grosse et al., 2002).
Air
Sources. In 1925, leading medical scientists, members of the U.S. Public Health Service, and industry representatives discussed the use of
leaded gasoline. Alice Hamilton of Harvard Medical School noted that
lead is a cumulative poison that usually does not produce easily recognized symptoms. A Standard Oil representative stated that industry could
not respond to a remote probability of harm from a product that allegedly
improved fuel economy. The U.S. government allowed continued use of
tetraethyl lead after a committee appointed by the Surgeon General reported in early 1926 that lead from this source was not an acute hazard
to the community. Subsequent exposure of virtually the entire populations of developed countries to lead from this source could have been
avoided if the warnings of health authorities had been heeded (Needleman, 1997).
Annual lead emissions in the United States decreased from 221,000
tons in 1970 to less than 4000 tons in 1997 (Table 4–6); at peak use in the
mid-1970s, combustion of leaded gasoline contributed about 90% of lead
CHILD HEALTH
84
AND THE
ENVIRONMENT
TABLE 4 –6. Lead Emissions into Air (tons), United States,
1970–1998
1970
1980
1990
1998
On-road vehicles
171,961
60,501
421
19
Metal processing
24,224
3,026
2,170
2,098
Other
24,684
10,626
2,384
1,856
Total
220,869
74,153
4,975
3,973
Source: U.S. Environmental Protection Agency (2000a).
emissions. Such emissions contained water-soluble, inorganic lead in respirable particles from which lead was easily absorbed into blood. Evidence that leaded gasoline was the major determinant of lead exposure in the general population came from NHANES II (1976–1980), a
population-based health survey of the United States (Annest et al., 1983).
Average blood lead levels among persons age 6 months to 74 years declined by a remarkable 37% during the short period 1976–1980. Similar
decreases occurred in each race, sex, and age subgroup and closely paralleled the declining sales of leaded gasoline and a 40% decline in ambient air lead concentrations during 1975–1980. For young children, there
was a strong correlation (r 0.95, p 0.001) between blood lead levels
and total lead used in gasoline for each 6-month period (see Fig. 4–1). Indoor air and umbilical cord blood lead levels in Boston during the period
1979–1981 correlated strongly with monthly leaded gasoline sales (Rabinowitz et al., 1984). Similar changes in airborne lead and childhood blood
lead levels were observed in other countries after the introduction of leadfree gasoline.
Emissions from industrial sources, mainly lead smelters and battery
manufacturers, have decreased by only 6% since 1988 and are now the
major sources of airborne lead emissions. Upgrading of the El Paso lead
smelter dramatically reduced lead exposure indices, but even after the
smelter closed in 1985, soil and blood lead levels of children living close
to the smelter were still elevated a decade later. Cross-sectional and longitudinal studies of children living in proximity to other lead smelters
confirmed this child health threat.
Intervention. The approximately 40% decline in average blood lead
levels in the United States during the late 1970s was mainly due to the
1970 Clean Air Act. Under this act, the EPA persuaded car manufacturers to add catalytic converters to new vehicles in 1975 to reduce exhaust
emissions; lead-free gasoline was introduced in the same year (lead ren-
Metals—Lead
85
20
19
18
17
16
15
14
13
12
11
10
9
8
40
50
60
70
80
90 100 110
Lead used per 6-month period (000’s of tons)
FIGURE 4–1. Average blood lead levels by 6-month periods of white children aged
6 months through 5 years and total lead used in gasoline production per 6 months
during 1976–1980. The blood lead values used to compute these averages were
preadjusted by regression analysis to account for the effects of sex, income, degree of urbanization, region of the country, and season (Annest et al., 1983). (Copyright © 1983, Massachusetts Medical Society. All Rights reserved.)
ders catalytic converters ineffective). The EPA set national ambient air
quality standards (NAAQS) for lead in 1978 and banned the use of leaded
gasoline in highway vehicles in 1995. Emissions from industrial sources
caused all recent violations of the lead NAAQS.
In some developing countries weak environmental standards, rapidly increasing numbers of cars, use of leaded gasoline, and elevated blood
lead levels still occur. The United Nations Commission on Sustainable Development in 1994 called for worldwide elimination of lead from gasoline. This action caused the World Bank, the OECD, and the United Nations Economic Commission for Europe to encourage and assist nations
to phase out leaded gasoline. By 1997, 13 countries had eliminated leaded
86
CHILD HEALTH
AND THE
ENVIRONMENT
gasoline and 18 had initiated a phase-out. The member states of the United
Nations Economic Commission for Europe agreed in 1998 to phase out
the use of leaded gasoline by the beginning of 2005.
Soil, House Dust, and Paint
For many years, seasonal variations in childhood blood lead levels (highest in summer) were attributed to sun exposure and dermal activation of
vitamin D, causing increased gastrointestinal lead absorption. Current evidence indicates that children are more exposed during summer because
of increased house dust lead concentrations and longer outdoor play periods in areas with contaminated soil (Yiin et al., 2000). Hand–mouth behavior of young children, combined with play activities at floor or ground
level, greatly increases the chance of exposure to lead-contaminated dust
and soil (Lanphear and Roghmann, 1997).
Sources. In populated areas, soil lead levels tend to be highest in the
urban core of large cities; soil lead levels of several hundred micrograms
per gram are common in urban cores where high traffic volumes occurred
for decades during the leaded gasoline era. Lead concentrations above
10,000 g/g have been observed in soil samples below the drip line of
buildings with exterior lead-based paint, but atmospheric deposition is
the main source of lead in soil outside the drip line. Soil tracked into
homes is a major source of house dust; lead isotope ratios of house dust
are virtually identical to those of street dust and soil. In the early 1980s,
about 95% of house dust lead in newer housing and at least 50% in older
housing originated in leaded gasoline (Fergusson and Schroeder, 1985).
Vinyl miniblinds containing lead as a plasticizer are an unusual but important source of lead-contaminated dust and have never been subjected
to a product recall.
A pooled analysis of 12 cross-sectional studies indicated that the major source of lead exposure for young children with blood lead levels of
10–25 g/dL was house dust (Lanphear et al., 1998). Blood lead levels of
young children were most closely related to interior floor dust lead loading and handwipe lead level (Succop et al., 1998). Children who ingest
soil while playing outdoors between meals may absorb a considerable
fraction of the lead content because a much larger fraction of ingested
lead is absorbed when fasting.
Houses built during 1920–1950 contained tens of pounds of paint lead
(Table 4–7). Children can take up lead from paint by eating paint chips
or by ingesting house dust or soil contaminated by the breakdown of old
paint. Ingested lead-based paint chips were an important cause of substantial blood lead elevations, particularly among African American chil-
Metals—Lead
87
TABLE 4 –7. White Lead Used in House Paint by Decade
Year
Total
(000’s of tons)
Amount per Housing
Unit (lb.s)
1920–29
1,307
87
1930–39
737
42
1940–49
476
22
1950–59
196
7
1960–69
82
3
1970–79
29
1
Source: President’s Task Force on Environmental Health Risks and
Safety Risks to Children (2000).
dren living in deteriorating pre-1950 housing. Radiographic evidence of
recent paint chip ingestion was found in about 25% of children with
markedly elevated blood lead levels (55 g/dL) but rarely among those
with lower levels (McElvaine et al., 1992). The marked and rapid decline
of blood lead levels concurrent with decreased use of lead in gasoline during the late 1970s occurred despite the absence of any significant remediation of lead-based paint. If lead-based paint was the most important
source of lead exposure for children, blood lead levels would likely peak
in winter, when children tend to remain indoors. In sum, it appears that
leaded gasoline caused high average background blood lead levels in the
general population and that lead-based paint was mainly a problem in
low-income population subgroups in older deteriorating housing. Since
such housing was commonly found in the high-traffic core of large urban
areas, affected populations must have often had major exposures from
both sources.
Intervention. France, Belgium, and Austria banned the interior use of
lead paint in 1909, but the United States and many other countries did
not act until the 1970s or later; in 1978 the U.S. Consumer Product Safety
Commission ruled that paint used for residences, toys, furniture, and public areas must contain no more than 0.06% lead by weight. About 26 million homes in the United States have significant lead-based paint hazards,
including, deteriorated lead-based paint areas, floor or window sill dust
lead loadings, or lead levels in bare soil above specified thresholds (U.S.
Department of Housing and Urban Development, 2001).
An evaluation of 11 U.S. residential lead hazard control projects concluded that floor and window sill dust lead levels remained substantially
lower over a 3-year follow-up period and children’s geometric mean
CHILD HEALTH
88
AND THE
ENVIRONMENT
blood lead levels decreased from 11.0 g/dL at baseline to 8.2 g/dL
1 year postintervention (Galke et al., 2001). A systematic review of residential residential lead abatement randomized trials concluded that there
was no impact on mean blood lead levels, but there was a significantly
lower prevalence of blood lead levels of 15 g/dL or greater in the intervention groups (6% vs. 14%) (Haynes et al., 2002).
The Residential Lead-Based Paint Hazard Reduction Act requires sellers and landlords of most housing built before 1978 to disclose known
lead-based paint and related hazards to buyers or renters and provide
printed information on how to protect their families from lead in the
home. Given the need for much more intensive intervention, this act appears to provide little real protection for children whose parents cannot
afford safer housing. The EPA developed new standards for lead in paint,
dust, and soil in 2000 (Table 4–8). These standards govern properties
receiving federal assistance and activities by certified lead services
providers, and guide action by state and local health departments, property owners, contractors, lenders, and insurers. The U.S. Department of
Housing and Urban Development (HUD) estimated that interim control
of lead hazards in pre-1960 low-income housing will yield a net benefit
TABLE 4 – 8. Standards for Lead in Environmental Media
Medium
Standard
Agency
Ambient air
0.5 g/m3 annual average
1.5 g/m3 quarterly average
WHO a
EPAb
Drinking water
10 g/L
15 g/L
WHO c
EPAd
Floor dust
40 g/sq ft
EPAe
Interior window sill dust
250 g/sq ft
EPAe
Bare soil in play areas
400 g/g
EPAe
Bare soil in rest of yard
1200 g/g
EPAe
Food
Children aged 0–6 yr
Children aged 7 yr
Pregnant women
6 g/day f
15 g/day
25 g/day
FDA f
a World
b U.S.
Health Organization (2000a).
Environmental Protection Agency (2001b).
c World
Health Organization (2000b).
d U.S.
Environmental Protection Agency (2000b).
e U.S.
Environmental Protection Agency (2001a).
f U.S.
Food and Drug Administration (1993): provisional tolerable total intake level (level of daily intake
likely to pose no risk of adverse effects to almost all individuals in the general population).
Metals—Lead
89
of $8.9 billion (at a 3% discount rate) over the next 10 years (President’s
Task Force on Environmental Health Risks and Safety Risks to Children,
2000). In 1999, Rhode Island became the first state to pursue legal action
against the manufacturers of lead-based paints.
Water
Water can leach lead from materials used in water distribution systems,
homes, or storage containers including lead pipes and solder. Lead pipes
are still present in many older communities and can cause substantial contamination, especially when water is soft and acidic. Elevated lead intakes
of infants occurred in areas of Germany and the United Kingdom where
high-lead tap water was consumed as such or added to dehydrated infant formulas and cereals. Among a nationally representative sample of
German children, blood lead levels were related to gender (male), age,
lead level in drinking water, and outdoor dust fall (Seifert et al., 2000).
Tap water lead in Glasgow decreased during 1981–1993 but remained the
leading cause of elevated maternal blood lead levels; about 13% of infants
were exposed to formula made from tap water containing lead above the
European Community’s limit value (10 g/L) (Watt et al., 1996). Tap water
lead levels above the EPA standard (15 g/L) occurred in 5% of standing
samples in a recent U.S. survey.
Food
Historically, food has been an important source of lead exposure, for example, foods stored in lead-soldered cans or lead-glazed ceramic containers. During the 1970s, over 90% of food cans were lead soldered, and
the FDA asked the food processing industry to intervene voluntarily. Lead
levels in infant foods and juices fell by 85% during 1973–1978 with the
conversion to lead-free soldered cans of infant formula and glass jars for
infant juices. Average daily dietary lead intake of a child aged 2 years in
the United States declined from 30 g/day in 1982 to 5 g/day during
1986–1988 and to 1.3 g/day in 1994–1996 (the FDA banned the use of
lead-soldered cans in 1991). Reduced contamination of food crops by airborne lead and reduced use of lead-glazed cookware also helped reduce
the dietary lead intake (Centers for Disease Control and Prevention, 1991).
Exceptions include northern Quebec, where 26% of reproductive-age Inuit
women had blood lead levels of at least 10 g/dL, associated with consumption of waterfowl contaminated by lead shot (Dewailly et al., 2001).
Lead-glazed pottery continues to be an important source of lead exposure in several countries; ceramics from Mexico, China, Korea, Italy, and
Spain imported into the United States as recently as 1995 were found to
release large amounts of lead into food and drink.
90
CHILD HEALTH
AND THE
ENVIRONMENT
The Joint FAO/WHO Expert Committee on Food Additives reduced
the Provisional Tolerable Weekly Intake for lead from all sources from 50
to 25 g per kilogram body weight because of concern for children. During the 1980s, mean weekly dietary lead intake of infants and children
was about 3 g/kg in the United States, 15–20 g/kg in Sweden and
Canada, and 26–28 g/kg in Poland, Germany, and Hungary. Diets high
in calcium or calcium supplements have been linked to lower blood lead
levels in pregnant women. Maternal wine consumption and cigarette
smoking were associated with increased cord blood lead levels; wine and
tobacco, respectively, may contain lead from the pesticide lead arsenate
and the use of brass fittings in wine-dispensing equipment.
Screening
Screening can identify children with elevated blood lead levels and trigger appropriate interventions. Baltimore, the first place in the United
States to do so, offered free blood lead testing of children beginning in
1935, mandated lead paint removal from housing in 1941, and assigned
public health nurses to investigate cases and educate parents in the late
1940s. Under the 1971 Lead-Based Paint Poisoning Prevention Act, the
CDC initiated the first mass screening nationwide with, about 400,000–
500,000 children under age 6 years screened annually during the next several years. Follow-up of children with high blood lead levels showed that
over 90% had received medical care, 75% of homes had been investigated,
and 80% of residential lead paint hazards had been abated. In 1991 the
CDC reduced the action level for childhood blood lead to 10 g/dL
and recommended universal blood lead screening of children aged 12–
72 months unless it could be shown that their community did not have
a childhood lead poisoning problem. By 1994, only half of the states had
implemented the CDC guideline.
In 1997 the CDC recommended that states develop policies and actions for primary prevention of lead exposure, targeted screening programs, surveillance, and actions to manage children with elevated levels
(Tables 4–9 and 4–10) (Centers for Disease Control and Prevention, 1997).
The CDC specified that screening programs should be targeted at children up to age 6 years who have never been screened and have one or
more risk factors. Among young children with elevated blood lead levels
during NHANES III (1988–1994), only one-third had been previously
screened (Kaufmann et al., 2000).
An analysis of the 1997 CDC guidelines suggested that it is most costeffective to do universal screening in high-prevalence populations and
targeted screening in low-prevalence populations. Targeted screening con-
Metals—Lead
91
TABLE 4–9. Model Childhood Lead Poisoning Prevention Policies and Actions
Policy
Action
Primary prevention
Housing codes or statutes
Lead education plan
Certified lead-abatement workers
Medicaid policies requiring
anticipatory guidance
Plan to reduce exposures from
industry and drinking water
Evaluate/control residential lead-based
paint hazards
Public lead education
Professional education/training
Anticipatory guidance by child health-care
providers
Identify/control sources other than leadbased paint
Secondary prevention
Screening plan; Medicaid screening
policies; protocols for providers
Policies and protocols for care,
environmental management,
follow-up
Childhood blood lead screening
Follow-up care for children with elevated
levels
Monitoring (surveillance)
Policy requiring laboratories to
report blood lead test results
Certification/licensing procedures
for safety of lead-hazard reduction
activities; procedures for tracking
lead-safe housing
Monitor children’s blood lead levels
Monitor older/deteriorated housing,
hazard-reduction activities, lead-safe
housing
Source: Centers for Disease Control and Prevention (1997).
sistent with the CDC guidelines would detect 90% of the cases at twothirds the cost of universal screening. The CDC concluded that venous
blood samples are more cost-effective than finger prick samples under
either universal or targeted screening; finger prick samples yield higher
false-positive rates because of contamination from lead on skin. The
prevalence of elevated blood lead levels (10 g/dL) among young children of lead-exposed workers was 52%, considerably above the U.S. population prevalence, indicating a need to reduce take-home lead exposures
and to screen at-risk children for elevated blood lead levels (Roscoe et al.,
1999).
The U.S. Preventive Services Task Force recommended that blood
lead be measured at age 12 months for all children having identifiable
risk factors or living in communities with a high or unknown prevalence
of elevated blood lead levels (U.S. Preventive Services Task Force, 1996).
CHILD HEALTH
92
AND THE
ENVIRONMENT
TABLE 4 –10. Blood Lead Categories and Actions Recommended by the CDC
Level (g/dL)
Recommended Action
10
Reassess or rescreen in 1 year. No additional action necessary unless
exposure sources change
10–14
Provide family lead education and follow-up testing; refer for social
services if necessary
15–19
Provide family lead education and follow-up testing; refer for social
services if necessary. If blood lead levels in this range persist
(two venous blood lead levels in this range at least 3 months
apart) or increase, proceed according to actions for blood lead
levels of 20–44 g/dL
20–44
Provide case management, clinical management, environmental
investigation, and lead-hazard control
45–69
Within 48 hours, begin case management, clinical management,
environmental investigation, and lead-hazard control
70
Hospitalize child and begin medical treatment immediately.
Begin case management, clinical management, environmental
investigation, and lead-hazard control immediately
Source: Centers for Disease Control and Prevention (1997).
Although the Task Force recommended targeted screening, it noted that
no randomized screening trials had been done and that most intervention studies on asymptomatic children had evaluated the impact on blood
lead levels, not on health status. The health benefits of blood lead screening programs are limited by irreversible neurotoxic effects before screening and intervention, prolonged mobilization of lead from bone stores to
blood postintervention, and continued contamination of house dust.
Surveillance
The CDC funded childhood blood lead surveillance systems and developed software to track those with elevated levels. Such systems were used
to document the extent of the problem, justify preventive and screening
programs, and reveal risks where none were expected including traditional medicines and ceramic ware. Large-scale population-based health
examination surveys are also important for monitoring health outcome
and health hazard exposure trends. Without NHANES, we would not
have such clear evidence of the role of leaded gasoline as the major exposure source for the general population, including persons in all demographic subgroups, during the leaded gasoline era.
Metals—Lead
93
Conclusion
Proven Child Health Outcomes
• Moderate to high childhood exposure—central nervous system toxicity
(ranging from headaches and agitation to somnolence, vomiting, coma,
and convulsions), peripheral neuropathy, anemia, kidney damage.
• Low exposure—deficits on standardized tests of global intelligence, increased threshold for hearing.
Unresolved Issues and Knowledge Gaps
• Developmental effects—there is limited evidence that low-level prenatal lead exposure may cause fetal death, reduced birth weight,
preterm delivery, birth defects (cardiac, neural tube, oral clefts), and reduced stature among preadolescent children.
• Neurotoxicity—there is limited evidence that low-level prenatal and/or
early childhood lead exposure may cause learning and behavior problems during childhood and adolescence.
• Delayed effects—there is inadequate evidence to assess the influence
of low-level childhood lead exposure on the risk of kidney cancer, other
renal disease, hypertension, and neurologic disease during adulthood
• Knowledge gaps—the efficacy of blood lead screening programs in reducing adverse health effects has not been tested in a randomized trial.
Risk Management Issues
• Prevention
° An estimated 900,000 American children aged 1–5 years in
1991–1994 had blood lead values above the CDC action level (10
g/dL); recent research has shown cognitive and hearing deficits
below this level.
Interventions
to reduce soil, dust, or paint lead hazards have had lim°
ited success in reducing children’s blood lead levels, probably because of recontamination from persistent lead sources and mobilization of bone lead.
° Ingestion of soil, dust, food, and water contaminated from past and
present lead sources continues to expose young children to varying
degrees of lead contamination.
° Only the virtual elimination of lead exposure during gestation and
early childhood can protect children from adverse neurobehavioral
effects of lead.
94
CHILD HEALTH
AND THE
ENVIRONMENT
• Biomonitoring—the tracking of population blood lead levels in
NHANES has enabled evaluation of preventive interventions and identification of continuing needs.
References
Agency for Toxic Substances and Disease Registry. (1999). Toxicological profile for
lead. Atlanta: Agency for Toxic Substances and Disease Registry.
Altmann L, Sveinsson K, Kramer U, Weishoff-Houben M, Turfeld M, Winneke G,
Wiegand H. (1998). Visual functions in 6-year-old children in relation to lead
and mercury levels. Neurotoxicol Teratol 20:9–17.
Andrews KW, Savitz DA, Hertz-Picciotto I. (1994). Prenatal lead exposure in relation to gestational age and birth weight: a review of epidemiologic studies.
Am J Ind Med 26:13–32.
Annest JL, Pirkle JL, Makuc D, Neese JW, Bayse DD, Kovar MG. (1983). Chronological trend in blood lead levels between 1976 and 1980. N Engl J Med 308:
1373–7.
Aschengrau A, Zierler S, Cohen A. (1993). Quality of community drinking water
and the occurrence of late adverse pregnancy outcomes. Arch Environ Health
48:105–13.
Baghurst PA, McMichael AJ, Wigg NR, Vimpani GV, Robertson EF, Roberts RJ,
Tong SL. (1992). Environmental exposure to lead and children’s intelligence at
the age of seven years. The Port Pirie Cohort Study. N Engl J Med 327:1279–84.
Ballew C, Khan LK, Kaufmann R, Mokdad A, Miller DT , Gunter EW. (1999). Blood
lead concentration and children’s anthropometric dimensions in the Third
National Health and Nutrition Examination Survey (NHANES III), 1988–1994.
J Pediatr 134:623–30.
Banks EC, Ferretti LE, Shucard DW. (1997). Effects of low level lead exposure on
cognitive function in children: a review of behavioral, neuropsychological and
biological evidence. Neurotoxicology 18:237–81.
Bellinger DC, Stiles KM, Needleman HL. (1992). Low-level lead exposure, intelligence and academic achievement: a long-term follow-up study. Pediatrics
90:855–61.
Borja-Aburto VH, Hertz-Picciotto I, Rojas Lopez M, Farias P, Rios C, Blanco J.
(1999). Blood lead levels measured prospectively and risk of spontaneous abortion. Am J Epidemiol 150:590–7.
Brody DJ, Pirkle JL, Kramer RA, Flegal KM, Matte TD, Gunter EW, Paschal DC.
(1994). Blood lead levels in the U.S. population. Phase 1 of the Third National
Health and Nutrition Examination Survey (NHANES III, 1988 to 1991). JAMA
272:277–83.
Burns JM, Baghurst PA, Sawyer MG, McMichael AJ, Tong SL. (1999). Lifetime lowlevel exposure to environmental lead and children’s emotional and behavioral
development at ages 11–13 years. The Port Pirie Cohort Study. Am J Epidemiol
149:740–9.
Byers RK, Lord EE. (1943). Late effects of lead poisoning on mental development.
Am J Dis Child 66:471–94.
Centers for Disease Control and Prevention. (1991). Preventing lead poisoning in
young children: a statement by the Centers for Disease Control. Atlanta: Cen-
Metals—Lead
95
ters for Disease Control and Prevention, Public Health Service, U.S. Department of Health and Human Services.
Centers for Disease Control and Prevention. (1997). Screening young children for
lead poisoning: guidance for state and local public health officials. Atlanta:
Centers for Disease Control and Prevention.
Centers for Disease Control and Prevention. (2000). Blood lead levels in young
children—United States and selected states, 1996–1999. MMWR 49:1133–7.
Danish Environmental Protection Agency. (2001). Denmark’s lead ban a world
first. Located at http://www.mst.dk/news/02070000.htm.
Davey FD, Breen KC. (1998). Stimulation of sialyltransferase by subchronic lowlevel lead exposure in the developing nervous system. A potential mechanism
of teratogen action. Toxicol Appl Pharmacol 151:16–21.
Dewailly E, Ayotte P, Bruneau S, Lebel G, Levallois P, Weber JP. (2001). Exposure
of the Inuit population of Nunavik (Arctic Quebec) to lead and mercury. Arch
Environ Health 56:350–7.
Dietrich KN, Ris MD, Succop PA, Berger OG, Bornschein RL. (2001). Early exposure to lead and juvenile delinquency. Neurotoxicol Teratol 23:511–8.
Fels LM, Wunsch M, Baranowski J, Norska-Borowka I, Price RG, Taylor SA, Patel S, De Broe M, Elsevier MM, Lauwerys R, and others. (1998). Adverse effects of chronic low level lead exposure on kidney function—a risk group study
in children. Nephrol Dial Transplant 13:2248–56.
Fergusson JE, Schroeder RJ. (1985). Lead in house dust of Christchurch, New
Zealand: sampling, levels and sources. Sci Total Environ 46:61–72.
Galke W, Clark S, Wilson J, Jacobs D, Succop P, Dixon S, Bornschein B, McLaine
P, Chen M. (2001). Evaluation of the HUD lead hazard control grant program:
early overall findings. Environ Res 86:149–56.
Gibson JL. (1892). Notes on lead-poisoning as observed among children in Brisbane. Proc Intercolonial Med Congr Aust 3:76–83.
Gonzalez-Cossio T, Peterson KE, Sanin LH, Fishbein E, Palazuelos E, Aro A,
Hernandez-Avila M, Hu H. (1997). Decrease in birth weight in relation to maternal bone lead burden. Pediatrics 100:856–62.
Grosse SD, Matte TD, Schwartz J, Jackson RJ. (2002). Economic gains resulting
from the reduction in children’s exposure to lead in the United States. Environ Health Perspect 110:563–69.
Guilarte TR. (1997). Glutamatergic system and developmental lead neurotoxicity.
Neurotoxicology 18:665–72.
Gulson BL, Mahaffey KR, Jameson CW, Mizon KJ, Korsch MJ, Cameron MA, Eisman JA. (1998). Mobilization of lead from the skeleton during the postnatal
period is larger than during pregnancy. J Lab Clin Med 131:324–9.
Haynes E, Lanphear BP, Tohn E, Farr N, Rhoads GG. (2002). The effect of interior
lead hazard controls on children’s blood lead concentrations: a systematic evaluation. Environ Health Perspect 110:103–7.
Hertz-Picciotto I. (2000). The evidence that lead increases the risk for spontaneous
abortion. Am J Ind Med 38:300–9.
Hertz-Picciotto I, Schramm M, Watt-Morse M, Chantala K, Anderson J, Osterloh
J. (2000). Patterns and determinants of blood lead during pregnancy. Am J Epidemiol 152:829–37.
Hogan K, Marcus A, Smith R, White P. (1998). Integrated exposure uptake biokinetic model for lead in children: empirical comparisons with epidemiologic
data. Environ Health Perspect 106(Suppl 6):1557–67.
96
CHILD HEALTH
AND THE
ENVIRONMENT
Irgens A, Kruger K, Skorve AH, Irgens LM. (1998). Reproductive outcome in offspring of parents occupationally exposed to lead in Norway. Am J Ind Med
34:431–7.
Johnston MV, Goldstein GW. (1998). Selective vulnerability of the developing brain
to lead. Curr Opin Neurol 11:689–93.
Kaufmann RB, Clouse TL, Olson DR, Matte TD. (2000). Elevated blood lead levels and blood lead screening among U.S. children aged one to five years: 1988–
1994. Pediatrics 106:E79.
Kim R, Hu H, Rotnitzky A, Bellinger D, Needleman H. (1995). A longitudinal
study of chronic lead exposure and physical growth in Boston children. Environ Health Perspect 103:952–7.
Kristensen P, Irgens LM, Daltveit AK, Andersen A. (1993). Perinatal outcome
among children of men exposed to lead and organic solvents in the printing
industry. Am J Epidemiol 137:134–44.
Landrigan PJ, Gehlbach SH, Rosenblum BF, Shoults JM, Candelaria RM, Barthel
WF, Liddle JA, Smrek AL, Staehling NW, Sanders JF. (1975a). Epidemic lead
absorption near an ore smelter. The role of particulate lead. N Engl J Med 292:
123–9.
Landrigan PJ, Whitworth RH, Baloh RW, Staehling NW, Barthel WF, Rosenblum
BF. (1975b). Neuropsychological dysfunction in children with chronic low-level
lead absorption. Lancet 1:708–12.
Lanphear BP. (1998). The paradox of lead poisoning prevention. Science 281:
1617–8.
Lanphear BP, Dietrich K, Auinger P, Cox C. (2000). Cognitive deficits associated
with blood lead concentrations 10 microg/dL in U.S. children and adolescents. Public Health Rep 115:521–9.
Lanphear BP, Matte TD, Rogers J, Clickner RP, Dietz B, Bornschein RL, Succop P,
Mahaffey KR, Dixon S, Galke W, and others. (1998). The contribution of leadcontaminated house dust and residential soil to children’s blood lead levels.
A pooled analysis of 12 epidemiologic studies. Environ Res 79:51–68.
Lanphear BP, Roghmann KJ. (1997). Pathways of lead exposure in urban children.
Environ Res 74:67–73.
Lin S, Hwang SA, Marshall EG, Marion D. (1998). Does paternal occupational lead
exposure increase the risks of low birth weight or prematurity? Am J Epidemiol
148:173–81.
Loghman-Adham M. (1997). Renal effects of environmental and occupational lead
exposure. Environ Health Perspect 105:928–39.
Mahaffey KR, Annest JL, Roberts J, Murphy RS. (1982). National estimates of blood
lead levels: United States, 1976–1980: association with selected demographic
and socioeconomic factors. N Engl J Med 307:573–9.
McElvaine MD, DeUngria EG, Matte TD, Copley CG, Binder S. (1992). Prevalence
of radiographic evidence of paint chip ingestion among children with moderate to severe lead poisoning, St Louis, Missouri, 1989 through 1990. Pediatrics
89:740–2.
Needleman HL. (1997). Clamped in a straitjacket: the insertion of lead into gasoline. Environ Res 74:95–103.
Needleman HL, Gatsonis CA. (1990). Low-level lead exposure and the IQ of children. A meta-analysis of modern studies. JAMA 263:673–8.
Needleman HL, Gunnoe C, Leviton A, Reed R, Peresie H, Maher C, Barrett P.
(1979). Deficits in psychologic and classroom performance of children with elevated dentine lead levels. N Engl J Med 300:689–95.
Metals—Lead
97
Needleman HL, Riess JA, Tobin MJ, Biesecker GE, Greenhouse JB. (1996). Bone
lead levels and delinquent behavior. JAMA 275:363–9.
Osman K, Pawlas K, Schutz A, Gazdzik M, Sokal JA , Vahter M. (1999). Lead exposure and hearing effects in children in Katowice, Poland. Environ Res 80:1–8.
Pirkle JL, Brody DJ, Gunter EW, Kramer RA, Paschal DC, Flegal KM, Matte TD.
(1994). The decline in blood lead levels in the United States. The National
Health and Nutrition Examination Surveys (NHANES). JAMA 272:284–91.
Pirkle JL, Kaufmann RB, Brody DJ, Hickman T, Gunter EW, Paschal DC. (1998).
Exposure of the U.S. population to lead, 1991–1994. Environ Health Perspect
106:745–50.
Pocock SJ, Smith M, Baghurst P. (1994). Environmental lead and children’s intelligence: a systematic review of the epidemiological evidence. BMJ 309:1189–97.
President’s Task Force on Environmental Health Risks and Safety Risks to Children. (2000). Eliminating childhood lead poisoning: a federal strategy targeting lead paint hazards. Washington, DC: U.S. Environmental Protection
Agency.
Rabinowitz M, Needleman H, Burley M, Finch H, Rees J. (1984). Lead in umbilical blood, indoor air, tap water, and gasoline in Boston. Arch Environ Health
39:299–301.
Rice DC. (1992). Lead exposure during different developmental periods produces
different effects on FI performance in monkeys tested as juveniles and adults.
Neurotoxicology 13:757–70.
Roscoe RJ, Gittleman JL, Deddens JA, Petersen MR, Halperin WE. (1999). Blood
lead levels among children of lead-exposed workers: a meta-analysis. Am J
Ind Med 36:475–81.
Rothenberg SJ, Schnaas L, Perroni E, Hernandez RM, Martinez S, Hernandez C.
(1999). Pre- and postnatal lead effect on head circumference: a case for critical
periods. Neurotoxicol Teratol 21:1–11.
Schwartz BS, Lee BK, Lee GS, Stewart WF, Simon D, Kelsey K, Todd AC. (2000).
Associations of blood lead, dimercaptosuccinic acid–chelatable lead, and tibia
lead with polymorphisms in the vitamin D receptor and ␦-aminolevulinic acid
dehydratase genes. Environ Health Perspect 108:949–54.
Schwartz J. (1994). Low-level lead exposure and children’s IQ: a meta-analysis
and search for a threshold. Environ Res 65:42–55.
Schwartz J, Landrigan PJ, Feldman RG, Silbergeld EK, Baker EL Jr, von Lindern
IH. (1988). Threshold effect in lead-induced peripheral neuropathy. J Pediatr
112:12–7.
Seifert B, Becker K, Helm D, Krause C, Schulz C, Seiwert M. (2000). The German
environmental survey 1990/1992 (GerES II): reference concentrations of selected environmental pollutants in blood, urine, hair, house dust, drinking
water and indoor air. J Expo Anal Environ Epidemiol 10:552–65.
Shen XM, Wu SH, Yan CH, Zhao W, Ao LM, Zhang YW, He JM, Ying JM, Li RQ,
Wu SM, and others. (2001). Delta-aminolevulinate dehydratase polymorphism
and blood lead levels in Chinese children. Environ Res 85:185–90.
Silbergeld EK. (1997). Preventing lead poisoning in children. Annu Rev Public
Health 18:187–210.
Succop P, Bornschein R, Brown K, Tseng CY. (1998). An empirical comparison of
lead exposure pathway models. Environ Health Perspect 106(Suppl 6):1577–83.
Thacker SB, Hoffman DA, Smith J, Steinberg K, Zack M. (1992). Effect of low-level
body burdens of lead on the mental development of children: limitations of
meta-analysis in a review of longitudinal data. Arch Environ Health 47:336–46.
98
CHILD HEALTH
AND THE
ENVIRONMENT
Tong S. (1998). Lead exposure and cognitive development: persistence and a dynamic pattern. J Paediatr Child Health 34:114–8.
Turner AJ. (1897). Lead poisoning among Queensland children. Aust Med Gazette
16:475–9.
U.S. Department of Housing and Urban Development. (2001). National survey of
lead and allergens in housing. Final report. Volume I: Analysis of lead hazards. Revision 6.0. Washington, DC: U.S. Department of Housing and Urban
Development.
U.S. Environmental Protection Agency. (1998). Risk analysis to support standards
for lead in paint, dust, and soil (EPA 747-R-97-006). Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2000a). National air pollutant emission
trends, 1900–1998. Research Triangle Park, NC: U.S. Environmental Protection
Agency.
U.S. Environmental Protection Agency. (2000b). National primary drinking water
regulations for lead and copper. Final rule. Fed Reg 65:1950–2015.
U.S. Environmental Protection Agency. (2001a). Lead. Identification of dangerous
levels of lead. Final rule. Fed Reg 66:1206–40.
U.S. Environmental Protection Agency. (2001b). National ambient air quality standards (NAAQS). Washington, DC: U.S. Environmental Protection Agency.
U.S. Food and Drug Administration. (1993). Guidance document for lead in shellfish. Washington, DC: U.S. Food and Drug Administration.
U.S. Preventive Services Task Force. (1996). Guide to clinical preventive services.
Washington, DC: U.S. Department of Health and Human Services.
Vinceti M, Rovesti S, Bergomi M, Calzolari E, Candela S, Campagna A, Milan M,
Vivoli G. (2001). Risk of birth defects in a population exposed to environmental
lead pollution. Sci Total Environ 278:23–30.
Watt GC, Britton A, Gilmour WH, Moore MR, Murray GD, Robertson SJ, Womersley J. (1996). Is lead in tap water still a public health problem? An observational study in Glasgow. BMJ 313:979–81.
Weiss B. (1990). Risk assessment: the insidious nature of neurotoxicity and the aging brain. Neurotoxicology 11:305–13.
World Health Organization. (2000a). Guidelines for air quality. Geneva: World
Health Organization.
World Health Organization. (2000b). Guidelines for drinking water quality.
Geneva: World Health Organization.
Yiin LM, Rhoads GG, Lioy PJ. (2000). Seasonal influences on childhood lead exposure. Environ Health Perspect 108:177–82.
5
Metals—Mercury, Arsenic,
Cadmium, and Manganese
The previous chapter documents the child health threats posed by lead,
the most intensely studied heavy metal. This chapter describes the known
and potential health hazards of other metals and metalloids including
mercury, arsenic, cadmium, and manganese. Except for mercury, it is the
inorganic and organic derivatives of these elements that are potential child
health hazards. In common with PCBs and certain other stable organochlorine compounds, cadmium and methylmercury tend to persist in
environmental media and to bioaccumulate in certain foods eaten by humans. While lead, mercury, arsenic, and cadmium have no known essential role in human biology, inorganic manganese is an essential trace
element required for the normal function of several important enzymes.
Inhaled inorganic manganese, however, can cause neurotoxicity among
occupationally exposed adults. Although high-level exposures to mercury
(especially methylmercury) cause severe neurotoxicity among children
and adults, there has been little epidemiologic research on the potential
roles of dental amalgam (a widespread source of elemental mercury
exposure), arsenic, cadmium, and manganese in adverse child health
outcomes. This chapter summarizes current knowledge about these elements and points to the need for increased epidemiologic research and
biomonitoring.
99
100
CHILD HEALTH
AND THE
ENVIRONMENT
I. MERCURY
Mercury exists in the natural environment as methylmercury, mercuric
sulfide (cinnabar ore), and mercuric chloride. Microbial biotransformation of inorganic mercury creates virtually all of the methylmercury found
in environmental media. Synthetic organic mercurials have been used as
antimicrobial preservatives in vaccines, other medicines, paints, and seed
grain. Elemental mercury is a dense, shiny, silver-white metal that is liquid at room temperature and has a relatively high vapor pressure. Uses
of elemental mercury have included mercury cathodes for electrolysis of
sodium chloride to produce chlorine gas and caustic soda, extraction of
gold from ore, dental amalgam for repairing carious teeth, thermometers,
barometers, mercury vapor lamps, electrical switches, and religious remedies and rituals in Latin America and Asia. Environmental inorganic mercury is a minor source of mercury exposure but inorganic mercury products have been used as disinfectants in diaper washes and as analgesics
in teething powders for infants. The major issues explored in Part I of this
chapter are the uncertainties about potential health risks of low-level exposure to methylmercury from dietary sources and elemental mercury
from dental amalgam and other sources.
Methylmercury
Health Effects
Severe neurotoxicity of organic mercury was evident as early as 1866,
when exposure in a chemistry laboratory killed two persons. Grave or fatal neurotoxic effects also occurred among syphilitics treated with diethylmercury (1887) and among workers engaged in organic mercury pesticide production during the early twentieth century. After acute adult
methylmercury exposure, a latent period of several weeks or even a few
months passes before symptoms appear. Methylmercury is extremely neurotoxic in the human fetus and the developing infant.
Molecular Mechanisms
About 95% of ingested methylmercury is absorbed and readily crosses
placental and blood-brain barriers. After crossing the blood–brain and
placental barriers, methylmercury enters tissues, where it is demethylated
and oxidized to divalent mercury that readily reacts with sulfhydryl
groups of proteins and thiols (e.g., tubulin, glutathione). Possible mech-
Metals—Mercury, Arsenic, Cadmium, and Manganese
101
anisms for toxicity of methylmercury and divalent mercury include
(Agency for Toxic Substances and Disease Registry, 1999b)
• Oxidative stress with generation of free radicals that attack protein
and DNA
• Disruption of microtubule formation, impairing cell motility and control of chromosome movement during cell division
• Increased permeability of the blood–brain barrier
• Disruption of DNA replication and protein synthesis
• Interference with proteins involved in neuronal calcium metabolism
In experimental animals, prenatal low-level methylmercury exposure inhibits neuronal cell division and migration, key processes in the developing brain, causing widespread brain damage. Neonatal exposures cause
focal cell loss, primarily in the cerebellum and occipital cortex.
Neurotoxicity: Epidemic Poisonings
Minamata. In 1953, a strange polio-like disease struck inhabitants of
Minamata, Japan, most victims being coastal villagers who regularly ate
fish from the adjacent bay. Onset of the epidemic coincided with the
startup of acetaldehyde production at a coastal factory later shown to
have used mercuric oxide as a catalyst. Investigators observed that stray
cats developed neurotoxicity after eating local shellfish. A heat-stable
compound present in shellfish and factory effluents caused neurotoxicity
in experimental animals. The local government did not intervene at this
stage, as requested by public health authorities, on the grounds that the
causative agent was not identified with certainty. Researchers finally identified the neurotoxin as methylmercury in 1963, but it was not until 1968
that Japanese authorities officially recognized it as the causal agent and
intervened.
Investigators identified over 2000 victims, including about 64 prenatally exposed infants (Harada, 1995). Affected infants generally appeared
normal at birth but later developed signs and symptoms of severe neurotoxicity: mental retardation, abnormal reflexes, ataxia, dysarthria, involuntary movements, and cerebral palsy. None crawled, stood, or talked
before age 3 years, and many could not walk at age 7 years. Some infants
exhibited severe neurotoxic effects, while their mothers had mild or no
symptoms. Autopsies of infants who died showed greatest brain damage
among those exposed during the third trimester. Because they did not
suspect mercury initially, investigators did not collect blood or hair samples but many families followed the Japanese custom of preserving a dried
102
CHILD HEALTH
AND THE
ENVIRONMENT
section of umbilical cord. Children born during 1950–1965, the peak period of acetaldehyde production, had the highest umbilical cord mercury
levels. Fish mercury levels (10–30 g/g) and average fish consumption
(300 g/day) at Minamata were much higher than current levels in United
States.
Iraq. Methylmercury and other synthetic organic mercurials were
used for several decades during the early twentieth century to protect
seeds from fungal damage and improve crop yields. Unwitting use of
methylmercury-treated seed grain for food caused several recognized epidemics of severe neurotoxicity in Iraq during 1955–1972; the largest
epidemic (1971–1972) involved over 6000 cases with several hundred
deaths. Similar outbreaks occurred in other countries (Pakistan, Guatemala,
Ghana), and these disastrous experiences finally led to worldwide bans
of alkylmercurials for seed treatment.
As in Minamata, some prenatally exposed infants had substantial
neurologic deficits even though their mothers reported no symptoms or
only mild, transitory paresthesias (Amin-Zaki et al., 1979). Signs and
symptoms among 32 prenatally exposed infants included microcephaly,
irritability, exaggerated reactions to stimuli, and abnormal reflexes.
Among eight infants with severe cerebral palsy, six were blind and two
had minimal sight; among their mothers, peak hair mercury levels occurred during the third trimester (average, 444 g/g). Among infants with
milder signs, the lowest peak maternal hair mercury level during pregnancy was 32 g/g. Up to age 4 months, blood methylmercury among
infants exceeded maternal levels, consistent with continued exposure
through breast milk. Follow-up of severe cases to age 5 years showed persistent neurologic abnormalities and delayed developmental milestones,
such as, inability at age 2 years to walk two steps without support or respond to simple verbal communication.
Among 49 Iraqi children aged 2–16 years with high postnatal exposures, about half had severe effects including ataxia, dysarthria, visual
deficits (blurred vision, constricted fields, blindness), hearing deficits,
glove and stocking numbness and paresthesias, involuntary movements,
and incontinence (Amin-zaki et al., 1978). The severity of neurologic abnormalities was associated with estimated blood mercury concentrations
near the end of the exposure period (using a blood mercury half-life of
56 days and extrapolating back in time from current blood levels). The
degree of recovery over a 2-year follow-up period was inversely related
to the initial severity of signs and symptoms; all children had persistent
hyperreflexia, even those with initially mild poisoning. About a third of
the initially blind children had recovered partial sight, and about a third
Metals—Mercury, Arsenic, Cadmium, and Manganese
103
of the severely poisoned children remained physically and mentally incapacitated (Amin-zaki et al., 1978). A WHO expert group reviewed
dose–response data from Iraq and estimated risks of fetal neurotoxicity
of 5% and 30%, respectively, at maternal hair mercury levels of 10–20 g/g
and 70 g/g (World Health Organization, 1990).
Neurotoxicity: Environmental Exposures
Studies of several fish-eating populations exposed to methylmercury at
levels considerably below those in Minamata and Iraq have not shown
consistent evidence of neurotoxic effects (Myers and Davidson, 1998). Results from the two largest longitudinal studies, the Faroe Islands and
Seychelles Islands birth cohort studies, are shown in Table 5–1. Faroe Islands residents eat diets rich in fish and marine mammals (pilot whales)
containing relatively high concentrations of methylmercury, PCBs, and
potentially protective antioxidants (selenium and vitamin E); the median
maternal hair mercury level during pregnancy was 4.5 g/g, much
lower than at Minamata (41 g/g) but higher than in the United States
(1 g/g). Cord blood methylmercury levels were inversely related to
scores on a standardized neurologic examination at age 2 weeks, independent of PCBs (Steuerwald et al., 2000). Breast-feeding was associated
with higher infant hair mercury level at age 12 months and early developmental milestone attainment (sitting, creeping, and rising). Assessment
at age 7 years, however, indicated that cord-blood mercury level was associated with deficits in language, attention, and visuospatial memory, independent of cord blood PCB level (Grandjean et al., 1999).
The Seychelles Islands study population is remote from polluting industry, consumes large amounts of fish but not whales, and has a low
prevalence of tobacco and alcohol use among women. The median maternal prenatal hair mercury level during pregnancy was 5.9 g/g, slightly
higher than that of the Faroese women. Prenatal or postnatal mercury exposure indices were inconsistently related to developmental milestones
or neurobehavioral scores up to age 5 years.
High-level prenatal methylmercury exposure causes similar behavioral and pathologic effects in young animals and humans, that is, mental retardation, cerebellar ataxia, primitive reflexes, dysarthria, and seizures. Relatively low prenatal exposures cause visual memory deficits,
abnormal social behavior, and reduced growth at puberty in nonhuman
primates, while low neonatal exposures produce visual spatial contrast
sensitivity deficits. Monkeys exposed from birth to adulthood to low
doses of methylmercury displayed visual recognition memory deficits
during infancy, slower retrieval of treats, impaired fingertip vibration
sense in middle age, and slight visual field deficits as adults (Rice and
TABLE 5–1. Major Birth Cohort Studies of Mercury and Neurobehavioral Effects
Main Author, Population
Population
Exposures and Associations
Grandjean et al. (1992)
1023 mother–infant pairs; exposed to methylmercury,
PCBs, and other contaminants mainly from eating
pilot whales; marine fish minor source of
exposure
Cord blood mercury—median, 24.2 g/L; 75th percentile, 40 g/L;
maternal hair mercury—median, 4.5 g/g; 13% exceeded 10 g/g
Grandjean et al. (1995)
583 infants followed to age 12 months; recorded age
first sat without support, crawled, and stood
without support
Early milestone development associated with breast-feeding and
increased infant hair mercury level at age 12 months but not
with maternal hair (at delivery) or cord blood mercury level
Grandjean et al.
(1997, 1999)
917 children tested at age 7 years; clinical
examination and neurophysiologic
and neuropsychologic tests
Inverse associations between cord blood and maternal hair
mercury levels and scores on tests of language, attention,
memory, and visuospatial and motor functions that persisted at
maternal hair mercury levels below 10 g/g; cord blood mercury
most closely associated with language, attention, and memory
deficits; concurrent child hair and blood mercury levels were
less predictive but were inversely associated with visuospatial
memory
Grandjean et al. (1998)
112 children whose maternal hair mercury level was
10–20 g/g and 112 matched children whose
maternal hair mercury level was ⬍3 g/g,
age 7 years
High-exposure group had small deficits in motor function
(especially fingertapping), language, and memory
Murata et al. (1999)
Reanalysis of data on brainstem auditory evoked
potentials
Maternal hair and cord blood but not child’s concurrent hair
mercury level associated with brainstem auditory evoked
potential abnormalities
Faroe Islands
Budtz-Jorgensen et al.
(2000)
Estimated BMDs of cord blood mercury
for deficits in attention, language,
and verbal memory scores at age 7 years
95% confidence lower limit of estimated BMDs for cord blood
mercury was about 5 g/L (equivalent to a maternal hair
mercury level of about 1 g/g)
Grandjean et al. (2001)
435 children age 7 years; measured PCB levels
in umbilical cord tissue
Association between cord tissue PCBs and deficits on 2 of 17 neuropsychologic outcomes; possible interaction between PCBs and
mercury in highest mercury tertile
Myers et al. (1995)
779 mother–infant pairs, methylmercury from
marine fish; infants tested at age 6 months
(visual recognition memory and developmental
screening tests)
Maternal hair total mercury during pregnancy—median, 5.9 g/g
(range, 0.5–27 g/g); no associations with neurodevelopmental
scores at age 6.5 months
Davidson et al. (1995);
Myers et al. (1997)
738 infants assessed at 19 months, 736 retested at
29 months (infant development and behavior)
No association between maternal hair mercury during pregnancy
and psychomotor or mental development scores or age at first
walking or talking
Axtell et al. (1998);
Davidson et al. (1998);
Myers et al. (2000);
Palumbo et al. (2000)
711 children aged 5.5 years
No consistent associations between prenatal or postnatal mercury
exposure indices and reduced neuropsychologic scores (including
overall indices, subscales, and recombined subscales); positive
association between postnatal mercury exposure and memory
subscale scores
Crump et al. (2000)
Estimated maternal hair mercury BMDs for
neurologic tests, neuropsychologic tests, and
developmental milestone data at four follow-up
examinations (age 6, 19, 29, and 66 months)
The average lower 95% confidence limit BMD for maternal hair
mercury was about 25 g/g (range, 19–30 g/g)
Seychelles Islands
106
CHILD HEALTH
AND THE
ENVIRONMENT
Hayward, 1999). Animal studies have not yet replicated the usual pattern
of human methylmercury exposure, that is, generally intermittent and related to fish consumption.
Other Effects
All three forms of mercury cumulate to higher levels in kidney than any
other tissue and can cause toxicity ranging from increased urinary protein levels (indicative of renal tubular damage) to renal failure with
nephrosis and necrosis of proximal tubules. For instance, infants dermally
exposed to diapers washed with soap containing phenylmercury have increased urinary excretion of ␥-glutamyl transpeptidase. After exposure to
organic mercury, children are more susceptible than adults to skin changes
(acrodynia or “pink disease”) including rash followed by peeling skin
on the palms of the hands and soles of the feet, itching, and joint pain.
Acrodynia was more common in the past, when mercury-containing laxatives, worming medications, teething powders, and diaper rinses were
widely used. There is limited animal and inadequate human evidence that
methylmercury is carcinogenic, and the EPA concluded that it is unlikely
to be a human carcinogen at exposure levels generally encountered from
environmental sources.
Exposures
Given the high toxicity of mercury, it is surprising that only Germany and
the United States appear to have collected nationally representative data
on mercury levels in children and reproductive-age women (Table 5–2)
(Centers for Disease Control and Prevention 2001b, 2001c; Seifert et al.,
2000).
Exposure Biomarkers
Blood mercury, about 95% of which is bound to red blood cells, has a halflife of about 50 days and thus reflects recent exposure. The half-life of
methylmercury in the blood of lactating women is about half that in nonlactating women due to excretion in breast milk. The cord blood mercury
concentration is about 20%–30% higher than that of maternal blood and
reflects fetal exposure during late gestation, the period of greatest susceptibility to neurotoxicity. Maternal blood and hair but not breast milk
mercury levels are associated with methylmercury exposure from fish
consumption. About 90% of methylmercury is excreted in bile/feces and
the remainder in urine and breast milk. Incomplete development of biliary transport systems contributes to a longer half-life of methylmercury
in infants compared to adults.
Metals—Mercury, Arsenic, Cadmium, and Manganese
107
TABLE 5–2. Norms and Health-Based Limits for Selected Metals in
Human Specimens
Sample
Norm or Limit
Mercury
Blood (age 1–5 years)
Blood (age 6–14 years)
Blood (women age 16–49 years)
Cord blood
Hair (age 1–5 years)
Hair (women age 16–49 years)
Maternal hair
Urine (age 6–14 years)
1.4 g/L (CI 0.7–4.8), 90th percentile a
1.1 g/L, 90th percentile b
6.2 g/L (CI 4.7–7.9), 90th percentile a
5 g/L (BMDL) c
0.4 g/g (CI 0.3–1.8), 90th percentile d
1.4 g/g (CI 0.9–1.7), 90th percentile d
1 g/g (BMDL) c
25 g/g (BMDL) e
10–20 g/g (BMD) f
12 g/g (BMDL) g
1.9 g/g creatinine, 90th percentile b
Arsenic
Urine (age 6–14 years)
14.1 g/g creatinine, 90th percentile b
Cadmium
Blood (age 1–19 years)
Blood (age 6–14 years)
Urine (age 6–14 years)
0.4 g/L (CI 0.3–1.0), 90th percentile a
0.3 g/L, 90th percentile b
0.15 g/g creatinine, 90th percentile b
a Centers
b Seifert
for Disease Control and Prevention (2001a).
et al. (2000).
c Budtz-Jorgensen
d Centers
e Crump
f World
et al. (2000).
for Disease Control and Prevention (2001c).
et al. (2000).
Health Organization (1990).
g National
Academy of Sciences (2000).
BMDL benchmark dose limit (lower 95% confidence limit on BMD).
BMD benchmark dose.
Hair grows at the rate of about 1 cm per month, and the mercury
content in a given segment of hair is about 250-fold that in blood at the
time the segment was formed. Maternal hair mercury levels correlate
strongly with those in fetal brain, cord blood, and newborn hair; thus segmental hair mercury analysis is valuable for retrospective mercury exposure estimation. Published studies vary as to whether they measured total mercury or methylmercury, but over 80% of total mercury in hair from
fish-eating populations is methylmercury. Women exposed during the
major poisoning incidents had hair mercury levels of up to 700 g/g in
Minamata (median, 41 g/g) and over 400 g/g in Iraq. Mothers in longitudinal studies of fish-eating populations all had hair mercury levels
below 40 g/g.
108
CHILD HEALTH
AND THE
ENVIRONMENT
United States and German biomonitoring surveys showed that mercury levels were generally low; in the United States, about 10% of women
had hair mercury levels within one-tenth of potentially hazardous levels,
indicating a relatively narrow margin of safety. Average hair mercury levels in local studies in the United States have usually been under 1 g/g,
that is, about the level expected for exposure at the EPA reference dose
for methylmercury. A Canadian methylmercury screening program tested
almost 40,000 aboriginal persons during 1972–1992 (Wheatley and Paradis, 1998). Inuit communities dependent on diets high in fish and sea
mammals had the highest average cord blood and adult mercury levels;
over 30% of reproductive-age Inuit women had hair methylmercury levels over 10 g/g. Blood mercury levels varied substantially by season,
corresponding to high consumption of fish and seafood in the early fall
and early winter.
Risk Management
It was not until the late 1960s and 1970s that investigators discovered the
ability of aquatic microbes to methylate mercury and the bioaccumulation
of methylmercury from concentrations in water to 1 million-fold higher
levels in predators such as tuna and marine mammals atop the aquatic
food chain. By then, however, vast amounts of mercury from chlor-alkali,
pulp and paper, mining, and other industries had been discharged into
aquatic environments worldwide. The largest current users of mercury are
chlor-alkali plants (production of chlorine and caustic soda) and electrical/electronic industries (electric lighting, wiring, switches, batteries).
Air
Analyses of mercury in peat and lake sediments in remote parts of North
America indicate that mercury emissions into air have increased about
fivefold since the beginning of the industrial period. The major sources
of air emissions are coal-fired utility/industrial boilers (50%), municipal
waste combustors (20%), and medical waste incinerators (10%). Future
mercury emission levels will be heavily influenced by increasing use of
coal to meet energy needs, especially since emissions from municipal and
medical waste combustion declined 50%–75% during the 1990s in the
United States.
Airborne mercury emissions from natural and anthropogenic sources
disperse in the environment by long-range atmospheric transportation.
Elemental mercury vapor tends to remain airborne, while inorganic mercury is rapidly cleared to soil and water compartments, deposition being
enhanced by precipitation. Modeling indicates that the highest deposition
Metals—Mercury, Arsenic, Cadmium, and Manganese
109
rates of airborne mercury in the United States occur in the southern Great
Lakes region, the Ohio River valley, the northeastern states, and other
scattered areas. Because atmospheric mercury deposition accounts for
much of the mercury in fish in the northeastern United States, even modest increases in atmospheric mercury loading could further elevate levels
in fish. Air mercury levels over the Atlantic Ocean increased until about
1990 and have continued to increase in northern Canada and Alaska due
to long-range transport of increasing global emissions.
Phenylmercuric acetate was used as a fungicide/bactericide to prolong the shelf life of interior latex paint up to 1990 in the United States
and was the source for two reported cases of childhood mercury poisoning (acrodynia). At that time, the EPA permitted interior latex paint to
contain up to 300 ppm mercury but did not require a label warning about
the presence and concentration of mercury; paint used in the home of one
patient actually contained about 950 ppm mercury, or three times the EPA
limit. After application, phenylmercuric acetate apparently breaks down
and releases elemental mercury. Air mercury levels were greatly elevated
during application of latex paint and decreased rapidly thereafter but remained above background levels for at least several years, reflecting continued mercury release. Subsequent investigations showed that exposed
children had higher urinary mercury levels than older persons. Even in
homes where paint contained less than 200 ppm mercury, air mercury
levels were up to 1.5 g/m3 (median, 0.3 g/m3), with some homes
exceeding the ATSDR acceptable indoor concentration for continuous exposure of 0.5 g/m3 (Beusterien et al., 1991). By 1991, all registrations for
mercury compounds in paints had been canceled by the EPA or voluntarily withdrawn by manufacturers. This occurrence shows the importance of regulatory measures to ensure that children are not exposed to
toxicants in the indoor environment arising from the use of household
products.
Reduced mercury emissions can be achieved through manufacturing
controls (product substitution, process modification, and materials separation), coal cleaning, and flue gas treatment technologies. Specific examples of manufacturing controls include replacement of mercury cathodes in chlor-alkali plants, reduced use of mercury in household batteries
and fluorescent lights, and removal of mercury-containing materials (e.g.,
batteries, fluorescent lights, thermostats) from wastes prior to incineration. Conventional cleaning methods reduce the coal mercury content by
about half, and control devices on utility and industrial boilers can remove over 90% of mercury emissions.
Under the Clean Air Act, the EPA has set rules for municipal and
medical waste incineration with the goal of reducing mercury emissions
110
CHILD HEALTH
AND THE
ENVIRONMENT
to 10% of 1995 levels. The EPA has also proposed mercury emission standards for hazardous waste incinerators and is evaluating reductions for
industrial boilers, chlor-alkali plants, and Portland cement kilns. Other
Clean Air Act regulations, including those for fine particulate matter
(PM2.5), will contribute to reduced mercury emissions. In 1998, the EPA
required coal-fired power plants to monitor smokestack mercury emissions. Mercury consumption in the United States decreased by 75% between 1988 and 1996 due to federal bans on its use in latex paints, pesticides, and batteries, state regulation of emissions and products, and
state-mandated recycling programs.
Water
The Castner Kellner method for the electrolytic production of sodium hydroxide and chlorine from brine was introduced in the late 1800s. Each
plant required about 100 tons of elemental mercury for cathodes to start
production and intermittent supplements to replace losses in cooling
water. As shown through investigation at Minamata and other sites, effluents from such plants contaminated the aquatic environment and
food chain with methylmercury. Mercury levels in aquatic environments
remote from polluting industries and other forms of development correlate strongly with air levels. Hydroelectric dams create new or enlarged
aquatic environments where mercury leaches from rocks and soil and enters the aquatic food chain. Average mercury concentrations in pike in
northern Quebec increased fivefold after construction of the James Bay
hydroelectric dams, thereby exposing aboriginal and other populations
dependent on local fish to increased methylmercury levels. Massive
amounts of liquid elemental mercury have been used to extract gold, silver, copper, and tin from ores. Between 1550 and 1880, an estimated
200,000 metric tons of elemental mercury were used in South American
gold mining. This practice persists in the Amazon basin, exposing workers to elemental mercury and local populations to methylmercurycontaminated fish.
Food
Predatory fish, shellfish, and sea mammals comprise the main methylmercury exposure sources for the general population. Dietary exposures
before and during pregnancy are important because the half-life of
methylmercury in maternal tissues is 1–9 months. In fish-eating populations, children are exposed prenatally, during breast-feeding (methylmercury in breast milk), and by eating fish from an early age. All fish contain some mercury, mostly as methylmercury, and large carnivorous
freshwater and marine fish and fish-eating mammals can have levels
Metals—Mercury, Arsenic, Cadmium, and Manganese
111
above 1 g/g, the current FDA action level; average levels of most commercially important marine fish are 0.1 g/g or lower. About 95% of
methylmercury in ingested fish is absorbed. The EPA estimated that 7%
of reproductive-age women and about 20% of all fish-eating children aged
3–6 years exceed the EPA methylmercury RfD (0.1 g/kg/day) (Table 5–3)
(U.S. Environmental Protection Agency, 1997).
The Minamata epidemic remains the only known occurrence of severe methylmercury poisoning due to fish consumption. The hazards of
methylmercury in fish must be balanced against the nutritional benefits
of fish, especially in indigenous populations dependent on fish as a major dietary component. Fish are excellent sources of selenium and omega3 fatty acids, substances important in brain growth that may offset to some
degree the neurotoxicity of low-level methylmercury contamination.
In general, the FDA and other regulatory agencies issue advisories
for mercury in fish rather than promulgating limits. A Health Canada advisory states that (1) reproductive-age women and children should limit
their consumption of swordfish, shark, and fresh or frozen tuna to one
TABLE 5–3. Guidelines and Standards for Chronic Exposure to Mercury
Exposure
Guideline or Limit
Agency
0.2
0.3 g/m3
ATSDRb
EPAc
Air (inorganic)
1 g/m3 (annual average)
WHO d
Drinking water
1 g/L (total mercury)
2 g/L (inorganic mercury)
WHO e
EPAf
Food (methylmercury)
0.1g/kg/day
0.3 g/kg/day
1.0 g/g (fish, edible portion)
Women who are pregnant or may become
pregnant or are breast-feeding and children
should not eat shark, swordfish, king
mackerel, or tilefish
EPAc
ATSDRb
FDAg
FDAh
Chronic oral intake
(inorganic)
0.3 g/kg/day
EPAc
Air
(elemental)a
a Inhalation,
b Agency
c U.S.
g/m3
chronic exposure.
for Toxic Substances and Disease Registry (2002).
Environmental Protection Agency (2001b).
d World
Health Organization (2000a).
e World
Health Organization (2000b).
f U.S.
Environmental Protection Agency (2001a).
g U.S.
Food and Drug Administration (2000).
h U.S.
Food and Drug Administration (2001).
112
CHILD HEALTH
AND THE
ENVIRONMENT
meal monthly, (2) mercury concentrations in these fish (0.5–1.5 g/g) are
above the Canadian guideline of 0.5 g/g, but the nutritional value of
fish justifies occasional consumption, and (3) mercury levels in canned
tuna are usually well below 0.5 g/g. The FDA analyses of canned
tuna indicated that the average mercury level was 0.17 g/g (range,
0.1–0.75 g/g). The FDA recently advised pregnant women not to eat
shark, swordfish, king mackerel, and tilefish because of their high
methylmercury levels. Surveys of the Cree native Indian population in
northern Quebec indicate that the prevalence of hair mercury levels
greater than 15 g/g declined from 14% to 3% during 1988–1994, a change
attributed to education on avoidance of high-mercury fish species
(Dumont et al., 1998). Mercury levels in meat, liver, and kidney from
Swedish pigs declined during the 1980s, possibly due to reduced use of
fish in pig feed.
Other Sources
Despite their extreme neurotoxicity, childhood exposure to organic mercury
compounds continues. An estimated 7000–10,000 infants in Argentina were
exposed to phenylmercury compounds used as disinfectants during commercial diaper washing; health outcomes included acrodynia and elevated
urinary ␥-glutamyl transferase levels (Gotelli et al., 1985). Until recently, sodium ethylmercurithiosalicylate (thimerosal) was used to prevent microbial contamination of vaccines and other biologics; three professional societies and the U.S. Public Health Service recommended rapid introduction
of thimerosal-free vaccines (Centers for Disease Control and Prevention,
2000). As of late 2001, all routine pediatric vaccines were being produced
in thimerosal-free or thimerosal-reduced (95% reduction) formulations.
General Considerations
Given the extraordinary sensitivity of the developing fetus to methylmercury, reproductive-age women have the highest priority for preventive actions including
• Policies and actions to further reduce food-related exposure while preserving access to beneficial and traditional diets
• Surveillance to evaluate progress and identify residual problems
• Continued research on neurobehavioral and other potential health effects of low-level perinatal methylmercury exposure, including delayed
effects in adulthood
Major uncertainties exist regarding a safe level for ingested methylmercury. The available data from Iraq and Minamata were inadequate for robust estimates of a NOAEL and an RfD. The EPA, however, set an RfD
for methylmercury of 0.1 g/kg/day in 1995 (Table 5–3) based on de-
Metals—Mercury, Arsenic, Cadmium, and Manganese
113
velopmental delays and neurologic abnormalities among Iraqi children,
incorporating an uncertainty factor of 10. The EPA plans to review its RfD
for methylmercury using new data from the Faroe Islands, the Seychelles
Islands, and the U.S. Great Lakes region.
The ATSDR based its MRL for methylmercury on data from the
Seychelles Islands study, in which no adverse effects were evident; on the
assumption that the average exposure among study participants (mean
maternal hair mercury was 15.3 g/g) was a NOAEL, the ATSDR set the
MRL at 0.3 g/kg/day. A benchmark analysis of the Seychelles Islands
study assessed neurobehavioral and developmental milestone data at four
follow-up examinations; the average lower 95% confidence limit on estimated BMDs for maternal hair mercury was about 25 g/g (range, 19–
30 g/g) (Crump et al., 2000). Among the epidemiologic studies, only the
Faroe Islands study has shown adverse effects (reduced performance on
language, attention, and memory tests at age 7 years) at maternal hair
mercury levels less than 10 g/g (Grandjean et al., 1997). The National
Academy of Sciences recently concluded that the RfD for methylmercury
should be based on the BMD lower limit for abnormal scores on the Boston
Naming Test in the Faroe Islands study (equivalent to a mercury level of
12 g/g in maternal hair); because the Academy also recommended an
uncertainty factor of at least 10, maternal hair mercury concentrations
should be no higher than 1.2 g/g (National Academy of Sciences, 2000).
Elemental and Inorganic Mercury
Health Effects
Dental amalgam is the major source of elemental mercury and (after oxidation in tissues) inorganic mercury exposure in the general population.
Elemental mercury is highly lipophilic, and about 70%–80% of inhaled
material is absorbed and disseminated in blood; it readily crosses the
blood-brain and placental barriers and enters tissues, where it is rapidly
oxidized to divalent mercury. The longest half-life of mercury in tissues
occurs in brain, while kidneys accumulate mercury to levels about ten
times those in other tissues (Agency for Toxic Substances and Disease
Registry, 1999b).
Most evidence of the health effects of elemental mercury comes from
adult studies. Acute exposure to very high levels of elemental mercury
vapor can cause pulmonary edema, respiratory distress, and death; the
nervous system and the kidneys are the most sensitive targets at lower
exposure levels. Neurotoxic effects in children accidentally exposed to
moderately high levels include dizziness, insomnia, peripheral neuropa-
114
CHILD HEALTH
AND THE
ENVIRONMENT
thy (numbness and tingling in the hands and feet), tremors, and irritability. Although major debates about the safety of dental amalgam have
occurred since its introduction in the 1800s, the potential child health effects of prenatal and childhood exposure to this source remain virtually
unresearched and unknown.
Mercurous chloride, also known as calomel, was used to treat many
ailments beginning in the eighteenth century and in teething powders
during the 1940s, causing many childhood poisonings. Even in recent
years, there have been mercury poisoning cases caused by Mexican products containing calomel for treatment of acne and other skin conditions
(Centers for Disease Control and Prevention, 1996a). Renal toxicity of divalent mercury appears to be related mainly to actions on enzymes and
transport proteins that take up mercury in proximal tubular cells. A small
fraction of persons, possibly due to genetic susceptibility, produce anti–
glomerular basement membrane and anti-DNA antibodies in response to
inorganic mercury exposure, and some develop autoimmune glomerulonephritis. Inorganic mercury triggers an autoimmune syndrome in susceptible rodents including autoreactive T cells, high IgE, anticollagen
types I and II IgGs, glomerulonephropathy, and gastrointestinal necrotizing vasculitis.
Exposures
The estimated percentage of total mercury uptake attributable to dental
amalgam is about 30% for children and 50% for adults, making it the single largest source for the general population (Richardson, 1995). Amalgam dental fillings continually release elemental mercury vapor that is
inhaled and readily absorbed through the lungs. The average number of
carious permanent teeth per person among children aged 2–10 years decreased from 2.3 in NHANES I to 1.4 in NHANES III; corresponding values for persons aged 6–18 years were, respectively, 4.4 and 1.9 (Brown et
al., 2000). Release of mercury vapor from amalgam increases about fivefold when chewing, brushing, or consuming hot beverages; mouth breathing, common in children, increases exposure.
The number of dental amalgam fillings in children and adults is associated with blood, breast milk, and urinary mercury levels. Inhaled elemental mercury is transported in blood to tissues and oxidized to inorganic mercury, which is transported mainly in plasma and excreted in
urine; plasma and urinary mercury levels appear to be the best indices of
mercury uptake from dental amalgam fillings. Placental mercury levels
are strongly correlated with maternal blood levels and with the number
of amalgam fillings (Ask et al., 2002).
Estimated infant mercury exposure from breast-feeding was up to
Metals—Mercury, Arsenic, Cadmium, and Manganese
115
0.3 g/kg/day (approximately half as inorganic mercury and half as
methylmercury) (Oskarsson et al., 1996). The number of amalgam fillings
estimated to cause exposure at the tolerable daily intake level for mercury vapor (0.014 g/kg/day) was one for toddlers and children, three
for teenagers, and four for adults (Richardson, 1995). A survey of urban
homes showed that the average mercury concentration in house dust samples (1.7 g/g) was 30 times that in soil samples, suggesting the existence
of important indoor sources (Rasmussen et al., 2001). Drinking water is a
minor source of mercury exposure.
Risk Management
Child health issues related to elemental mercury include its widespread
use in dental amalgam and the difficulty of recognizing sporadic cases
caused by exposures such as accidental indoor spillage or religious practices. Although Egyptians used a mercury amalgam in dentistry over 1500
years ago, widespread use has occurred only during the past 150 years.
Dental amalgam causes continuous exposure to mercury; on precautionary grounds alone, this argues for not using amalgam among children
and reproductive-age women. Removal of amalgam causes an exponential decline of blood, plasma, and urinary mercury levels, with values at
2 months being about 60% of preremoval levels. The importance of dental amalgam should decrease as prevalence rates of dental caries decline
(fluoridation and fluoride supplements) and as the use of substitute materials increases.
Potential indoor sources of elemental mercury exposure include broken switches, thermostats, and thermometers and religious practices. Mercury can easily be obtained for religious or cultural purposes including
sprinkling on the floor of a home or car, burning in a candle, and mixing
with perfume. Many users are not aware of its toxicity and may be exposed to indoor air mercury levels far above occupational exposure limits. If mercury is spilled on floors, it tends to remain in cracks and carpets and emit dense mercury vapor; sporadic childhood poisoning cases
from such sources continue to be reported (Centers for Disease Control
and Prevention, 1996b).
Conclusions
Proven Child Health Outcomes
• Moderate to high prenatal exposure to methylmercury can cause severe neurotoxic effects ranging from abnormal reflexes, irritability,
116
•
•
•
•
CHILD HEALTH
AND THE
ENVIRONMENT
delayed milestones, and visual disturbances to cerebral palsy, microcephaly, blindness, and major cognitive deficits.
Postnatal exposure to high methylmercury levels can cause ataxia,
dysarthria, visual deficits (including blindness), hearing deficits, peripheral neuropathy, and involuntary movements.
Neurotoxic effects in children accidentally exposed to moderately high
levels of elemental mercury include dizziness, insomnia, peripheral
neuropathy, tremors, and irritability.
Inorganic mercury (mercurous chloride) used in teething powders during the early twentieth century caused many childhood poisonings
(acrodynia) characterized by irritability, stomatitis, insomnia, and
erythema of the hands, feet, and other areas.
All three forms of mercury can cause increased urinary protein excretion indicative of renal tubular damage.
Unresolved Issues and Knowledge Gaps
• Neurotoxic effects of low-level prenatal or childhood exposure to
methylmercury from diet. The two major birth cohort studies of prenatal maternal exposure to fish or marine mammals contaminated by
methylmercury at levels well below those in Minamata, but higher than
those in the general population, have produced conflicting results:
° The Faroes Islands study showed that cord blood mercury levels were
associated with significant deficits in language, attention, and visuospatial memory at age 7 years, independent of cord blood PCB levels.
° The Seychelles Islands study showed no consistent associations between prenatal or postnatal mercury exposure indices and developmental milestones or neurobehavioral scores up to age 5 years.
• The health effects of exposure to elemental mercury from dental amalgam during fetal development and childhood, if any, remain unknown.
• Knowledge gaps exist concerning neurobehavioral and other potential
health effects of low-level perinatal methylmercury exposure.
Risk Management Issues
• Prevention—policies and actions are required to further reduce mercury
emissions and education to reduce consumption of highly contaminated
aquatic foods.
• Biomonitoring—biomonitoring of children and reproductive-age women
is needed to measure progress in reducing perinatal mercury exposure
(only the United States and Germany have implemented national tracking of mercury levels in children).
Metals—Mercury, Arsenic, Cadmium, and Manganese
117
II. OTHER METALS AND METALLOIDS
In contrast to the relatively well-known effects of mercury on children,
most epidemiologic research on arsenic, cadmium, and manganese has
involved occupationally and environmentally exposed adults. Arsenic
and cadmium can affect a vast array of biochemical and nutritional processes by binding to sulfhydryl groups, generating free radicals, inhibiting antioxidative enzymes, and depleting intracellular glutathione. Countering such effects are metallothioneins (MTs), small cysteine-rich proteins
that bind metals and scavenge free radicals. Tissues vary in their MT content and their susceptibility to metals; for example, testes have very low
levels of MT, and cadmium exposure triggers Leydig cell death and decreased testosterone production. Cadmium and arsenic compounds inhibit DNA repair systems and are well-known carcinogens in experimental animals and humans.
Arsenic
Arsenic, a metalloid with both metallic and nonmetallic physicochemical
properties, occurs naturally in rocks, soil, water, air, plants, and animals
and has four valency states (0, 3, 3, and 5). Inorganic arsenates (pentavalent arsenic) dominate in aerobic surface waters and arsenites (trivalent arsenic) in anaerobic groundwaters (U.S. Environmental Protection
Agency, 2000). Although organic arsenic has been considered to be relatively nontoxic, recent evidence suggests otherwise.
Health Effects
Molecular Mechanisms
About 50%–70% of absorbed inorganic arsenate is rapidly reduced in vivo
to arsenite; the latter reacts readily with tissue components and inhibits
many enzymes. In humans, arsenite is methylated to metabolites
(monomethylarsonic acid, dimethylarsinic acid, and trimethylarsine oxide) that are excreted more rapidly. Arsenic may contribute to carcinogenicity in humans by inhibiting DNA repair, causing DNA methylation
and oxidative stress, and inhibiting transcription of the hTERT gene
(which encodes the reverse transcriptase subunit of human telomerase),
causing chromosome end-to-end fusions and other abnormalities. Reduction of pentavalent to trivalent arsenic in vivo may be viewed as an
activation pathway, as the trivalent form is more reactive with sulfhydryl
groups of tissue components. Although methylation of inorganic arsenic
118
CHILD HEALTH
AND THE
ENVIRONMENT
is still commonly viewed as a detoxification pathway, methylated and dimethylated trivalent arsenicals are more potent cytotoxins, genotoxins,
and enzyme inhibitors (Thomas et al., 2001). Dimethylarsinic acid (DMA)
is genotoxic and a complete carcinogen in rodents.
Developmental Effects
Inorganic arsenic crosses the human placenta, but there has been little research on adverse developmental outcomes. Ecologic and case-control
studies have shown elevated risks of spontaneous abortion, birth defects,
and/or stillbirths in areas with elevated drinking water or airborne arsenic levels (Ahmad et al., 2001; Aschengrau et al., 1989; Hopenhayn-Rich
et al., 2000; Ihrig et al., 1998; Zierler et al., 1988). Prenatal exposure to
high-dose inorganic arsenic caused neural tube birth defects, growth retardation, and fetal death in hamsters, mice, rats, and rabbits. The National Research Council and the ATSDR concluded that there is insufficient evidence to judge if inorganic arsenic can affect reproduction or
development in humans (Agency for Toxic Substances and Disease Registry, 2000a; National Academy of Sciences, 1994).
Cancer and Other Chronic Diseases
Several major agencies concluded that arsenic compounds cause skin, lung,
bladder, and kidney cancers in humans (International Agency for Research
on Cancer, 1980; National Academy of Sciences, 1999; U.S. Environmental
Protection Agency, 1994). These conclusions were based mainly on epidemiologic studies of adults occupationally exposed to airborne arsenic
and populations exposed to drinking water containing high arsenic levels. The National Academy of Sciences recently concluded that chronic exposure to drinking water with arsenic levels less than 50 g/L is associated with increased risks of bladder and lung cancers (National Academy
of Sciences, 2001). An estimated 4–7 excess lifetime cancer cases per 10,000
persons would occur at chronic exposure levels as low as 3 g/L.
There have been very few epidemiologic studies of the potential role
of arsenic in childhood cancer. Exposure to airborne arsenic and other
metals (lead, cadmium) from smelters was associated with a doubling of
the childhood cancer risk in an ecologic study (Wulff et al., 1996). Children in northern Chile exposed to drinking water containing 750–800
g/L arsenic developed skin pigmentation and keratoses (Smith et al.,
2000). Lymphocytes from children and women exposed to inorganic arsenic in drinking water at levels of about 200 g/L displayed an increased
frequency of micronuclei and trisomy compared to less exposed controls
(Dulout et al., 1996). Epidemiologic studies of adults exposed to relatively
high levels of inorganic arsenic from drinking water or arsenical drugs
Metals—Mercury, Arsenic, Cadmium, and Manganese
119
have shown increased risks of several types of cancer, skin hyperpigmentation/keratoses in areas not exposed to the sun, peripheral neuropathy, cardiovascular disease, anemia, and diabetes.
Exposures
Ingested trivalent and pentavalent soluble inorganic arsenic compounds
are readily absorbed and transported in blood and are excreted mainly in
urine. The National Academy of Sciences noted that urinary total inorganic arsenic measurement avoids interference from organic arsenic in
seafood and better reflects recent and ongoing exposure than blood, hair,
or nail levels. Among persons exposed to arsenic in drinking water, children had higher urinary arsenic levels than adults, reflecting their higher
daily intake of water per unit body weight. The German Environmental
Survey showed that the 90th percentile urinary arsenic level among children was 14.1 g/g creatinine (Table 5–2). Geometric mean urinary levels among children who consumed fish less than or more than once
weekly, respectively, were 5.9 and 10.5 g/L (Seifert et al., 2000). The halflife of arsenic in blood is only 1 hour and is strongly influenced by recent
exposure; hair and nails are subject to contamination. Breast milk arsenic
levels are quite low even in areas with high levels in drinking water.
Risk Management
Food, drinking water, and soil are the main potential arsenic exposure
sources for children. Children living near point sources of arsenic are at
risk of increased exposure, particularly from soil and drinking water. An
estimated 13% of the general U.S. population exceed the EPA RfD for inorganic arsenic (0.3 g/kg/day).
Air
Combustion of fossil fuels and wastes, mining, smelting, pulp and paper
production, glass manufacturing, and cement manufacturing are the main
industrial sources of airborne arsenic. Although air is a minor exposure
source for the general population, young children in zinc and copper
smelter communities may have elevated blood and urinary arsenic levels.
Water
Millions of persons in Bangladesh are exposed to high drinking water arsenic levels from deep wells installed to reduce gastrointestinal infections
related to contaminated surface water; more localized problems exist in
other regions including India, China, Taiwan, Chile, and Argentina. The
CHILD HEALTH
120
AND THE
ENVIRONMENT
EPA estimated that 5.4% of groundwater and only 0.7% of surface water
sources in the United States have average arsenic levels above 10 g/L.
High groundwater arsenic levels occur in some western states with sulfide mineral deposits high in arsenic. The National Academy of Sciences
assessed epidemiologic data on drinking water arsenic levels and adult
cancer and concluded that cancer risks are significant even at the former
EPA arsenic drinking water standard (50 g/L). In 2001 the EPA reduced
its arsenic drinking water standard from 50 to 10 g/L, to be effective
3–5 years after approval (Table 5–4).
Food
Marine fish (e.g., tuna) and shellfish have the highest mean arsenic levels in Canadian and American total diet studies. Current evidence indicates that organic arsenic compounds in fish and shellfish (arsenobetaine
and arsenocholine) are excreted rapidly in urine and do not pose a significant risk to humans.
Other Sources
Arsenic, cadmium, and mercury were the metals in indoor and outdoor
dust samples in Louisiana that most frequently exceeded EPA risk-based
concentrations (Lemus et al., 1996). The last agricultural application involving inorganic arsenic pesticides was voluntarily canceled in 1993, but
young children continue to be at risk of unintentional arsenic poisoning
from arsenic-based rodenticides. The EPA recently began a review of the
potential for child exposure to chromated copper arsenate (CCA), widely
used in pressure-treated wood. Leaching from CCA-treated wood used
in playground structures and decks can raise arsenic and chromium levTABLE 5–4. Guidelines and Standards for Chronic Exposure to Inorganic Arsenic
Exposure
Guideline or Limit
103
per
g/m3
WHOa
Air
1.5 ⫻
Drinking water
10 g/L
10 g/L
WHOb
EPAc
Total oral exposure
0.3 g/kg/day
0.3 g/kg/day
ATSDRd
EPAe
a World
(unit risk)
Agency
Health Organization (2000a); because arsenic is a human carcinogen, the WHO specifies a unit
risk (excess lifetime cancer risk of inorganic arsenic).
b World
c U.S.
Health Organization (2000b).
Environmental Protection Agency (2001a).
d Agency
e U.S.
for Toxic Substances and Disease Registry (2002).
Environmental Protection Agency (2001b).
Metals—Mercury, Arsenic, Cadmium, and Manganese
121
els in underlying sand or soil, including both trivalent and the more toxic
hexavalent chromium. There have been no measurements of arsenic levels in children exposed to CCA-treated wood but adult workers handling
arsenic-treated wood have elevated urinary arsenic levels (Jensen et al.,
1991).
Cadmium
Elemental cadmium is a relatively rare, soft, extremely toxic metal used
in many products including batteries, pigments, metal coatings, plastics,
and metal alloys. Cadmium is ubiquitous in natural environments, usually as inorganic salts. Adults exposed to high levels of inorganic cadmium may develop kidney dysfunction, lung diseases, disturbed calcium
metabolism, and osteomalacia.
Health Effects
Endemic itai-itai (“ouch-ouch”) disease occurred in the downstream basin
of the Jinzu River in Japan beginning about 1912, the name coming from
the cries of victims suffering from severe bone pain. Although chronic
cadmium poisoning was suspected as a possible cause during the late
1950s, it was not until 1968 that the Japanese government completed research and concluded that the source was rice grown in water contaminated with cadmium from a mine. Health outcomes included irreversible,
progressive kidney damage, calcium loss, and osteomalacia (Iwata et al.,
1993). The delayed onset and progression of kidney damage reflect the
cumulation and persistence of cadmium in tissues. There appear to have
been no epidemiologic studies of the potential health effects of prenatal
and childhood cadmium exposure.
Molecular Mechanisms
Cadmium enters cells through calcium channels, cumulates intracellularly, and interferes with the uptake and functions of several essential metals (including calcium, zinc, selenium, chromium, and iron). At noncytotoxic doses, cadmium interferes with p53 and other transcription factors,
enhancing the genotoxicity of direct mutagens. Tumor suppressor p53 is
a zinc-dependent DNA transcription factor that controls DNA repair, survival, proliferation, and differentiation in cells with DNA damage; inactivation of this gene by mutation is a key event in many human cancers.
Cadmium also interferes with chromosome spindle formation and appears to cause chromosome aberrations among occupationally exposed
122
CHILD HEALTH
AND THE
ENVIRONMENT
men. Most cadmium in vivo is bound to MT, but this low molecular
weight complex can enter plasma, be excreted into the renal glomerular
filtrate, and be reabsorbed in renal tubular cells that split the complex and
are exposed to the toxic effects of free cadmium. In liver, cadmium binds
to mitochondrial protein sulfhydryl groups and triggers oxidative stress
and liver cell injury. Although inhaled cadmium is taken up by olfactory
axonal projections, it appears that it is not transported into other parts of
the brain.
Reproductive and Developmental Effects
Human evidence is insufficient to assess the potential role of cadmium in
prenatal and childhood development. Animals experienced fetal growth
deficits and skeletal malformations after prenatal exposure and testicular
atrophy from high postnatal exposure.
Kidney
Urinary excretion of small proteins (␣1-microglobulin, 2-microglobulin,
retinol-binding protein, and N-acetyl--D-glucosaminidase) is a sensitive
indicator of renal tubular cell injury caused by cadmium and certain other
nephrotoxins. A cross-sectional study showed a borderline association between urinary cadmium and small proteins in children (Noonan et al.,
2002).
Cancer
The IARC classifies cadmium as a known human carcinogen, while the
EPA deems it a probable human carcinogen (International Agency for Research on Cancer, 1994; U.S. Environmental Protection Agency, 1994). Men
occupationally exposed to inhaled cadmium dust and fumes had elevated
lung cancer risks. In experimental animals, inhaled cadmium caused lung
cancer, while ingested cadmium caused leukemia, testicular tumors, and
proliferative prostatic lesions; there appear to have been no epidemiologic
studies of the potential role of cadmium in childhood cancer.
Other Effects
The few epidemiologic studies of cadmium and cognitive function in children have yielded inconclusive findings because of inadequate exposure
assessment and lack of control for potential confounders. Prenatal exposure of rodents to relatively low cadmium levels caused adverse neurobehavioral effects (Agency for Toxic Substances and Disease Registry,
1999a). Other potential effects, based on very limited epidemiologic evi-
Metals—Mercury, Arsenic, Cadmium, and Manganese
123
dence, include reduced serum free thyroxine (T4) levels and suppressed
immediate hypersensitivity and IgG levels.
Exposures
Blood and urine cadmium levels in humans are both good indicators of
dietary cadmium intake, the main exposure source among nonsmokers
in the general population. Iron deficiency, relatively prevalent among reproductive-age women, increases gastrointestinal cadmium absorption.
Blood and hair cadmium levels in children are related to residential proximity to zinc and copper smelters. Cord blood cadmium levels are associated with maternal blood levels but are about tenfold lower; pregnant
smokers have higher blood and placental cadmium levels than nonsmokers. At low exposure levels, blood cadmium reflects exposure during the past 2–3 months, while urine cadmium indicates the body burden, particularly that in kidney. Hair is not an ideal indicator, as it is prone
to external contamination by airborne cadmium.
The German Environmental Survey and NHANES III appear to provide the only nationally representative cadmium exposure data (Table
5–2) (Centers for Disease Control and Prevention 2001b; Paschal et al.,
2000; Seifert et al., 2000). The 90th percentile blood cadmium levels for
children in the German Environmental Survey and NHANES III (1999),
respectively, were 0.3 and 0.4 g/L. During 1988–1994, about 0.2%–0.5%
of U.S. children aged 6–19 years had urine cadmium concentrations above
5 g/g creatinine, the current WHO health-based exposure limit (Paschal
et al., 2000). Results from NHANES III, however, indicate that the likelihood of renal tubular damage (indicated by urine microalbumin levels
above 30 g/mL) increased by 1% for each 10% increase in urinary cadmium above the median 0.23 g/g creatinine (Paschal et al., 2000).
Cadmium levels in diet, blood, and urine in Japan and in cadaver
kidneys in Sweden decreased by about 40% over the past 10–20 years,
possibly due to changed dietary habits or reduced food contamination. A
considerable fraction of the adult cadmium body burden may arise from
exposures during childhood, when gastrointestinal absorption rates are
higher.
Risk Management
Sources
The main industrial sources of airborne cadmium include metal production, waste incineration, battery production, fossil fuel combustion, and
124
CHILD HEALTH
AND THE
ENVIRONMENT
cement production. Airborne cadmium, mainly in respirable particulate
matter, ranges from 5 ng/m3 in rural areas to 15 ng/m3 in urban areas,
60 ng/m3 in industrial areas, and 300 ng/m3 near metal smelters. Drinking water usually contains less than 1 g/L but can reach 10 g/L in
areas influenced by anthropogenic activities. Agricultural soil cadmium
levels increase with use of phosphate or sewage sludge fertilizers; food
crops readily take up soil cadmium.
Food is the main source of cadmium intake by children and nonsmoking adults, with amounts averaging 10–50 g/day but up to tenfold
higher in polluted areas. Compared to breast-fed children, dietary cadmium intake may be up to 12 times higher in infants fed soy-based or
cereal-based formula; the estimated weekly intake at age 6 months is
3.1 g/kg, below the WHO provisional tolerable weekly intake based on
kidney effects in adults (7 g/kg) (Eklund and Oskarsson, 1999). Cadmium intake may be higher if the water used to make the formula also
contains cadmium. Only 5%–10% of ingested cadmium is absorbed, but
this increases if diets are deficient in zinc, iron, or calcium. The main dietary sources of cadmium in the U.S. adult population are liver, potatoes,
spinach, iceberg lettuce, and pasta; this is consistent with evidence that
leafy vegetables and grain crops readily take up cadmium from soil contaminated by the use of phosphate or sewage sludge fertilizers. Cereals
and shellfish are the main sources of cadmium exposure among nonsmoking Swedes.
Uptake of cadmium by rice varies with soil cadmium levels; concentrations in rice varied from less than 10 ng/g (geometric mean) in Australia, Finland, Spain, and the United States to 30 ng/g or more in China,
Japan, Italy, and Colombia. Most food crops other than rice contain sufficient zinc to inhibit absorption of ingested cadmium. Surveys of nonsmoking women aged 20–50 years in Asia showed that most cadmium
was from the diet, with rice alone accounting for about 40%. Northern indigenous populations who regularly consume organ meat, particularly
the liver and kidney of caribou, moose, and seal liver, may have cadmium
intakes above the EPA RfD for food (Table 5–5). Drinking water usually
contains very low cadmium levels (1 g/L), with important exceptions
such as water contaminated by leaching from waste disposal sites or certain plumbing materials (Agency for Toxic Substances and Disease Registry, 1999a).
Intervention
The EPA RfDs for cadmium are based on the observation that a renal cortex cadmium concentration of 200 g/g is the highest level not associated with significant proteinuria in adults. Given the toxicity of cadmium
Metals—Mercury, Arsenic, Cadmium, and Manganese
125
TABLE 5–5. Guidelines and Standards for Chronic Exposure
to Cadmium
Exposure
Guideline or Limit
Agency
Air
5 ng/m3 (average annual)
WHO a
Drinking water
3 g/L
5 g/L
WHO b
EPAc
Oral (food)
1 g/kg/day
EPAd
Total oral
0.2 g/kg/day
ATSDRe
a World
Health Organization (2000a).
b World
Health Organization (2000b).
c U.S.
Environmental Protection Agency (2001a).
d U.S.
Environmental Protection Agency (2001b).
e Agency
for Toxic Substances and Disease Registry (2002).
in adults and its accumulation in tissues, especially kidney, it would be
prudent to minimize exposures during pregnancy and childhood.
Manganese
Manganese is a transition element closely related to iron, an essential nutrient, and an integral part of several enzymes. Although manganese has
11 possible valence states, the main forms found in mammalian tissues
are di-, tri-, and tetravalent cations. Metallic manganese is employed primarily in steel production, while inorganic manganese compounds are
components of dry-cell batteries, matches, fireworks, glazes, varnishes,
ceramics, and nutritional supplements. Synthetic organic manganese compounds include the fuel additive methylcyclopentadienyl manganese tricarbonyl (MMT) and the fungicides maneb and mancozeb. Given the neurotoxicity of inhaled manganese among occupationally exposed adults
and the susceptibility of the developing nervous system to neurotoxins,
some jurisdictions have banned MMT use on precautionary grounds.
Health Effects
Ingested manganese is relatively nontoxic because of low absorption rates
and homeostatic mechanisms. Inhaled manganese can cause adverse developmental and respiratory effects and neurotoxicity similar to Parkinson’s disease in occupationally exposed persons and experimental animals.
126
CHILD HEALTH
AND THE
ENVIRONMENT
Molecular Mechanisms
The brain has a high energy requirement, making it susceptible to defects
in mitochondrial function, a feature common to Parkinson’s disease,
Huntington’s disease, and Friedreich’s ataxia. Manganese disrupts mitochondrial functions by (1) competing with iron, notably in certain irondependent mitochondrial enzymes, (2) catalyzing dopamine oxidation to
a reactive quinone that generates reactive oxygen species that damage mitochondrial DNA, and (3) inhibiting aconitase, an enzyme essential for
mitochondrial energy production. Trivalent manganese appears to be a
much more potent generator of reactive oxygen species than divalent
manganese. Manganese accumulates in astrocytes and is only slowly eliminated from the brain; its persistence in mitochondria may explain progressive loss of function after exposure ends (Aschner et al., 1999).
Rodent models indicate that inhaled manganese may enter the brain
via projections of olfactory neurons. Gastrointestinally absorbed manganese is transported in blood bound to plasma proteins including transferrin; transferrin bound manganese enters the brain by binding to
transferrin-receptors in cerebral capillaries and migrates to basal ganglia
via axonal transport. Because they compete for the same receptors, iron
deficiency promotes brain uptake of manganese; given the high prevalence of iron deficiency globally, this could be an important determinant
of manganese toxicity.
Neurotoxicity
Children on parenteral nutrition containing relatively high manganese
levels have developed signs of neurotoxicity including tremors and
seizures. There have been no adequate studies of children exposed to environmental manganese. A case report, however, noted marked verbal
and visual memory deficits but normal cognitive function in a child aged
10 years exposed for 5 years to drinking water high in manganese
(1.2 mg/L; compare this to the EPA MCL of 0.05 mg/L) (Woolf et al., 2002).
Compared to unexposed children, those exposed to manganese in drinking water contaminated by sewage irrigation had increased hair manganese levels and reduced scores on short-term memory, manual dexterity, and visuoperceptual speed tests (He et al., 1994). An epidemiologic
study in Greece revealed an association between drinking water manganese levels and neurotoxicity scores among adults (Kondakis et al.,
1989). Endemic motor neuron disease and other chronic neurologic disorders among certain island populations in the western Pacific have been
linked to diets low in calcium and iron and high in manganese, about half
the cases developing during early childhood (Cawte et al., 1989).
Metals—Mercury, Arsenic, Cadmium, and Manganese
127
Neurotoxic effects among men occupationally exposed to inhaled
manganese for many years include tremor, reduced response speed, possible memory and intellectual deficits, and mood changes (Agency for
Toxic Substances and Disease Registry, 2000b; Mergler and Baldwin, 1997).
Among Quebec adults exposed to ambient airborne manganese, those with
higher blood manganese levels had reduced ability to perform regular,
rapid, and precise pointing movements, lower maximum rotation speeds
in rapid alternating movements, and increased tremor (Beuter et al., 1999).
Manganese is a cumulative neurotoxin in experimental animals, causing adverse effects at relatively low levels when inhaled. Neurotoxic effects of manganese are linked to its accumulation in basal ganglia, especially the globus pallidus. Although the animal evidence of neurotoxicity
is extensive, a NOAEL has not been established. Among rats exposed to
ingested manganese, neonates but not adults had an increased acoustic
startle response.
Other Effects
Studies of animals perinatally exposed to manganese by various routes
have shown adverse effects including transient ataxia, decreased hypothalamic dopamine levels, reduced birth weight, skeletal abnormalities
(club foot), and reduced size of testes and seminal vesicles (Agency for
Toxic Substances and Disease Registry, 2000b). No studies have assessed
the potential carcinogenicity of inhaled manganese in humans or animals;
one animal study of ingested manganese revealed an increased incidence
of pancreatic tumors, but several other studies were negative. The EPA
designated manganese as not classifiable for carcinogenicity in humans.
Exposures
The EPA estimated that if MMT were used in all unleaded gasoline, about
5%–10% of people would be exposed to airborne manganese levels exceeding 0.1 g/m3, a potential inhalation RfC. Canada appears to be the
only country that has adopted MMT as the major replacement for
tetraethyl lead. Tailpipe emissions of manganese from combustion of fuel
containing MMT occur mainly as fine particulate (PM2.5). Mean ambient
manganese levels in Canadian cities (12 ng/m3) are higher than those in
California (3 ng/m3), where MMT use has been restricted (Wallace and
Slonecker, 1997). Airborne PM2.5 manganese levels generally declined during recent years in the United States but not in Canada. Blood and urine
manganese levels, respectively, appear to reflect the body burden and recent exposures (Agency for Toxic Substances and Disease Registry, 2000b).
CHILD HEALTH
128
AND THE
ENVIRONMENT
Risk Management
Sources
Food is the main source of manganese for the general population, with
air and water contributing about 1% of the daily intake. Combustion of
MMT produces manganese oxides. Average airborne manganese levels in
high and low traffic density areas of Montreal, respectively, were 24 ng/m3
and 15 ng/m3 (Loranger and Zayed, 1997). Manganese levels over time
at both sites were significantly correlated with changes in traffic density.
About 40% of inhaled manganese is absorbed.
Intervention
The EPA set the RfC for airborne manganese at 0.4 g/m3 in 1990 and reduced it to 0.05 g/m3 in 1993 based on neurobehavioral deficits in workers exposed to airborne manganese (Table 5–6). Unresolved risk assessment issues included the need to extrapolate from subchronic to chronic
exposures, lack of reproductive and developmental toxicity data, and unknown differences in the toxicity of different forms of manganese. The
EPA RfD for ingested manganese is 140 g/kg/day for adults; an RfD for
children has not been developed.
In 1976, MMT was introduced into the U.S. fuel supply, to raise octane ratings and reduce engine knock, but its use was restricted to
leaded gasoline in 1977. The Ethyl Corporation subsequently applied
to the EPA several times for permission to use MMT in unleaded gasoline. The EPA initially denied these applications on the grounds of concern about the impact on exhaust hydrocarbon emissions; later, the EPA
based its denial on public health concerns related to inhaled particu-
TABLE 5–6. Guidelines and Standards for Chronic Exposure to Manganese
Exposure
Guideline or Limit
Agency
Air
0.15 g/m3 (annual average)
0.05 g/m3
0.04 g/m3
WHO a
EPAb
ATSDRc
Drinking water
500 g/L
50 g/L
WHO d
EPAa
Oral RfD
140 g/kg/day
EPAb
a World
b U.S.
Health Organization (2000a).
Environmental Protection Agency (2001b).
c Agency
d World
for Toxic Substances and Disease Registry (2002).
Health Organization (2000b).
Metals—Mercury, Arsenic, Cadmium, and Manganese
129
late manganese emissions (Davis, 1999). The Ethyl Corporations’s subsequent challenge succeeded because the Clean Air Act allows the EPA
to ban a fuel additive only if it interferes with any emission control device or system.
Canada legislated a ban on importation of MMT in 1997. A consortium of 21 automotive manufacturers supported this decision on the
grounds that MMT interferes with antipollution devices and can harm
human health by increasing toxic contaminant emissions. The ban was
overturned in 1998 after a challenge by the Ethyl Corporation under
the North American Free Trade Act, again giving precedence to economic concerns over public health and the precautionary principle. The
Council on Scientific Affairs of the American Medical Association concluded that it would be prudent to call for more research and testing
before MMT is introduced widely into the U.S. gasoline supply
(Lyznicki et al., 1999).
Conclusions
Proven Health Outcomes
• Arsenic
° Large doses of inorganic arsenic can cause severe acute toxicity and
death.
° Chronic occupational exposure, mainly by inhalation, can cause lung
cancer in adults.
° Chronic environmental exposure, mainly by ingestion of contaminated drinking water, can cause skin, lung, bladder, and kidney cancers, peripheral vascular disease, and skin pigmentation and keratoses in adults.
• Cadmium
° Chronic high-level exposure to ingested cadmium can cause severe
kidney disease and osteomalacia in adults.
° Men occupationally exposed to inhaled cadmium dust and fumes
have an elevated lung cancer risk.
Unresolved Issues and Knowledge Gaps
• Arsenic—there is inadequate evidence to assess the role of chronic environmental exposure in developmental effects (fetal deaths, birth defects, low birth weight), childhood cancer, and delayed effects (hypertension, diabetes).
130
CHILD HEALTH
AND THE
ENVIRONMENT
• Cadmium—there is inadequate evidence to assess the role of environmental cadmium in childhood renal disease; cadmium can cause reduced fetal growth, skeletal malformations, leukemia, and testicular
cancer in animals, but its role in these conditions in humans is unknown.
• Manganese—although inhaled manganese can cause neurotoxic and
other adverse health effects in occupationally exposed men and experimental animals, there has been almost no research on its potential
effects on human prenatal development and child health.
• Knowledge gaps—the potential role of arsenic, cadmium, and manganese in fetal and child health is still to be investigated.
Risk Management Issues
• Prevention
° Arsenic—drinking water standards for arsenic should be reduced to
10 g/L as soon as possible; the potential for childhood exposure to
arsenic from CCA-treated wood requires early assessment.
° Cadmium—use of cadmium-contaminated sewage sludge on land for
food crops should be avoided.
° Manganese—use of the fuel additive MMT in countries such as
Canada requires assessment for its potential neurotoxic effects on the
human fetus and the developing child.
• Biomonitoring—Germany and the United States appear to be the only
countries with population-based biomonitoring systems for measuring
exposures to metals and other environmental toxicants among children.
References
Agency for Toxic Substances and Disease Registry. (1999a). Toxicological profile
for cadmium. Atlanta: Agency for Toxic Substances and Disease Registry.
Agency for Toxic Substances and Disease Registry. (1999b). Toxicological profile
for mercury (update). Atlanta: Agency for Toxic Substances and Disease
Registry.
Agency for Toxic Substances and Disease Registry. (2000a). Toxicological profile
for arsenic. Atlanta: Agency for Toxic Substances and Disease Registry.
Agency for Toxic Substances and Disease Registry. (2000b). Toxicological profile
for manganese. Atlanta: Agency for Toxic Substances and Disease Registry.
Agency for Toxic Substances and Disease Registry. (2002). Minimal risk levels
(MRLs) for hazardous substances. Located at http://www.atsdr.cdc.gov/
mrls.html.
Ahmad SA, Sayed MH, Barua S, Khan MH, Faruquee MH, Jalil A, Hadi SA,
Talukder HK. (2001). Arsenic in drinking water and pregnancy outcomes. Environ Health Perspect 109:629–31.
Metals—Mercury, Arsenic, Cadmium, and Manganese
131
Amin-Zaki L, Majeed MA, Clarkson TW, Greenwood MR. (1978). Methylmercury
poisoning in Iraqi children: clinical observations over two years. BMJ 1:613–6.
Amin-Zaki L, Majeed MA, Elhassani SB, Clarkson TW, Greenwood MR, Doherty
RA. (1979). Prenatal methylmercury poisoning. Clinical observations over five
years. Am J Dis Child 133:172–7.
Aschengrau A, Zierler S, Cohen A. (1989). Quality of community drinking water
and the occurrence of spontaneous abortion. Arch Environ Health 44:283–90.
Aschner M, Vrana KE, Zheng W. (1999). Manganese uptake and distribution in
the central nervous system (CNS). Neurotoxicology 20:173–80.
Ask K, Akesson A, Berglund M, Vahter M. (2002). Inorganic mercury and methylmercury in placentas of Swedish women. Environ Health Perspect 110:523–6.
Axtell CD, Myers GJ, Davidson PW, Choi AL, Cernichiari E, Sloane-Reeves J, Cox
C, Shamlaye C, Clarkson TW. (1998). Semiparametric modeling of age at
achieving developmental milestones after prenatal exposure to methylmercury
in the Seychelles child development study. Environ Health Perspect 106:
559–63.
Beusterien KM, Etzel RA, Agocs MM, Egeland GM, Socie EM, Rouse MA,
Mortensen BK. (1991). Indoor air mercury concentrations following application of interior latex paint. Arch Environ Contam Toxicol 21:62–4.
Beuter A, Edwards R, deGeoffroy A, Mergler D, Hundnell K. (1999). Quantification of neuromotor function for detection of the effects of manganese. Neurotoxicology 20:355–66.
Brown LJ, Wall TP, Lazar V. (2000). Trends in total caries experience: permanent
and primary teeth. J Am Dent Assoc 131:223–31.
Budtz-Jorgensen E, Grandjean P, Keiding N, White RF, Weihe P. (2000). Benchmark
dose calculations of methylmercury-associated neurobehavioural deficits. Toxicol Lett 112–13:193–9.
Cawte J, Kilburn C, Florence M. (1989). Motor neurone disease of the western Pacific: do the foci extend to Australia? Neurotoxicology 10:263–70.
Centers for Disease Control and Prevention. (1996a). Update: mercury poisoning
associated with beauty cream—Arizona, California, New Mexico, and Texas,
1996. MMWR 45:633–5.
Centers for Disease Control and Prevention. (1996b). Mercury exposure among
residents of a building formerly used for industrial purposes—New Jersey, 1995.
MMWR 45:422–4.
Centers for Disease Control and Prevention. (2000). Summary of the joint statement on thimerosal in vaccines. American Academy of Family Physicians,
American Academy of Pediatrics, Advisory Committee on Immunization Practices, Public Health Service . MMWR 49:622, 631.
Centers for Disease Control and Prevention. (2001a). National report on human
exposure to environmental chemicals. Located at http://www.cdc.gov/nceh/
dls/report/PDF/CompleteReport.pdf.
Centers for Disease Control and Prevention. (2001b). National report on human
exposure to environmental chemicals. Selection of chemicals for the report. Located at http://www.cdc.gov/nceh/dls/report/totalreport/substance.htm
Centers for Disease Control and Prevention. (2001c). Blood and hair mercury levels in young children and women of childbearing age—United States, 1999.
MMWR 50:140–3.
Crump KS, Van Landingham C, Shamlaye C, Cox C, Davidson PW, Myers GJ,
Clarkson TW. (2000). Benchmark concentrations for methylmercury obtained
132
CHILD HEALTH
AND THE
ENVIRONMENT
from the Seychelles Child Development Study. Environ Health Perspect 108:
257–63.
Davidson PW, Myers GJ, Cox C, Axtell C, Shamlaye C, Sloane-Reeves J, Cernichiari
E, Needham L, Choi A, Wang Y, and others. (1998). Effects of prenatal and
postnatal methylmercury exposure from fish consumption on neurodevelopment: outcomes at 66 months of age in the Seychelles Child Development
Study. JAMA 280:701–7.
Davidson PW, Myers GJ, Cox C, Shamlaye CF, Marsh DO, Tanner MA, Berlin M,
SloaneReeves J, Cernichiari E, Choisy O, and others. (1995). Longitudinal neurodevelopmental study of Seychellois children following in utero exposure to
methylmercury from maternal fish ingestion: outcomes at 19 and 29 months.
Neurotoxicology 16:677–88.
Davis JM. (1999). Inhalation health risks of manganese: an EPA perspective. Neurotoxicology 20:511–8.
Dulout FN, Grillo CA, Seoane AI, Maderna CR, Nilsson R, Vahter M, Darroudi F,
Natarajan AT. (1996). Chromosomal aberrations in peripheral blood lymphocytes from Native Andean women and children from northwestern Argentina
exposed to arsenic in drinking water. Mutat Res 370:151–8.
Dumont C, Girard M, Bellavance F, Noel F. (1998). Mercury levels in the Cree population of James Bay, Quebec, from 1988 to 1993/94. CMAJ 158:1439–45.
Eklund G, Oskarsson A. (1999). Exposure of cadmium from infant formulas and
weaning foods. Food Addit Contam 16:509–19.
Gotelli CA, Astolfi E, Cox C, Cernichiari E, Clarkson TW. (1985). Early biochemical effects of an organic mercury fungicide on infants: “dose makes the poison.” Science 227:638–40.
Grandjean P, Budtz-Jorgensen E, White RF, Jorgensen PJ, Weihe P, Debes F, Keiding N. (1999). Methylmercury exposure biomarkers as indicators of neurotoxicity in children aged 7 years. Am J Epidemiol 150:301–5.
Grandjean P, Weihe P, Burse VW, Needham LL, Storr-Hansen E, Heinzow B, Debes
F, Murata K, Simonsen H, Ellefsen P, and others. (2001). Neurobehavioral
deficits associated with PCB in 7-year-old children prenatally exposed to seafood neurotoxicants. Neurotoxicol Teratol 23:305–17.
Grandjean P, Weihe P, Jorgensen PJ, Clarkson T, Cernichiari E, Videro T. (1992).
Impact of maternal seafood diet on fetal exposure to mercury, selenium, and
lead. Arch Environ Health 47:185–95.
Grandjean P, Weihe P, White RF. (1995). Milestone development in infants exposed
to methylmercury from human milk. Neurotoxicology 16:27–33.
Grandjean P, Weihe P, White RF, Debes F. (1998). Cognitive performance of children prenatally exposed to “safe” levels of methylmercury. Environ Res 77:
165–72.
Grandjean P, Weihe P, White RF, Debes F, Araki S, Yokoyama K, Murata K,
Sorensen N, Dahl R, Jorgensen PJ. (1997). Cognitive deficit in 7-year-old children with prenatal exposure to methylmercury. Neurotoxicol Teratol 19:417–28.
Harada M. (1995). Minamata disease: methylmercury poisoning in Japan caused
by environmental pollution. Crit Rev Toxicol 25:1–24.
He P, Liu DH, Zhang GQ. (1994). Effects of high-level-manganese sewage irrigation
on children’s neurobehavior. Chung Hua Yu Fang I Hsueh Tsa Chih 28:216–8.
Hopenhayn-Rich C, Browning SR, Hertz-Picciotto I, Ferreccio C, Peralta C, Gibb
H. (2000). Chronic arsenic exposure and risk of infant mortality in two areas
of Chile. Environ Health Perspect 108:667–73.
Metals—Mercury, Arsenic, Cadmium, and Manganese
133
Ihrig MM, Shalat SL, Baynes C. (1998). A hospital-based case-control study of stillbirths and environmental exposure to arsenic using an atmospheric dispersion
model linked to a geographical information system. Epidemiology 9:290–4.
International Agency for Research on Cancer. (1980). IARC monographs on the
evaluation of carcinogenic risks to humans. Vol. 23. Arsenic and arsenic compounds. Lyon, France: International Agency for Research on Cancer.
International Agency for Research on Cancer. (1994). Beryllium, cadmium, mercury, and exposures in the glass manufacturing industry. Lyon, France: International Agency for Research on Cancer.
Iwata K, Saito H, Moriyama M, Nakano A. (1993). Renal tubular function after reduction of environmental cadmium exposure: a ten-year follow-up. Arch Environ Health 48:157–63.
Jensen GE, Christensen JM, Poulsen OM. (1991). Occupational and environmental exposure to arsenic—increased urinary arsenic level in children. Sci Total
Environ 107:169–77.
Kondakis XG, Makris N, Leotsinidis M, Prinou M, Papapetropoulos T. (1989). Possible health effects of high manganese concentration in drinking water. Arch
Environ Health 44:175–8.
Lemus R, Abdelghani AA, Akers TG, Horner WE. (1996). Health risks from exposure to metals in household dusts. Rev Environ Health 11:179–89.
Loranger S, Zayed J. (1997). Environmental contamination and human exposure
assessment to manganese in the St. Lawrence River ecozone (Quebec, Canada)
using an environmental fate/exposure model: GEOTOX. SAR QSAR Environ
Res 6:105–19.
Lyznicki JM, Karlan MS, Khan MK. (1999). Manganese in gasoline. Council on
Scientific Affairs, American Medical Association. J Occup Environ Med 41:
140–3.
Mergler D, Baldwin M. (1997). Early manifestations of manganese neurotoxicity
in humans: an update. Environ Res 73:92–100.
Murata K, Weihe P, Araki S, Budtz-Jorgensen E, Grandjean P. (1999). Evoked potentials in Faroese children prenatally exposed to methylmercury. Neurotoxicol Teratol 21:471–2.
Myers GJ, Davidson PW. (1998). Prenatal methylmercury exposure and children:
neurologic, developmental, and behavioral research. Environ Health Perspect
106(Suppl 3):841–7.
Myers GJ, Davidson PW, Palumbo D, Shamlaye C, Cox C, Cernichiari E, Clarkson TW. (2000). Secondary analysis from the Seychelles Child Development
Study: the child behavior checklist. Environ Res 84:12–9.
Myers GJ, Davidson PW, Shamlaye CF, Axtell CD, Cernichiari E, Choisy O,
Choi A, Cox C, Clarkson TW. (1997). Effects of prenatal methylmercury exposure from a high fish diet on developmental milestones in the Seychelles Child
Development Study. Neurotoxicology 18:819–29.
Myers GJ, Marsh DO, Davidson PW, Cox C, Shamlaye CF, Tanner M, Choi A,
Cernichiari E, Choisy O, Clarkson TW. (1995). Main neurodevelopmental study
of Seychellois children following in utero exposure to methylmercury from a
maternal fish diet: outcome at six months. Neurotoxicology 16:653–64.
National Academy of Sciences. (1994). Science and judgment in risk assessment.
Washington, DC: National Academy Press.
National Academy of Sciences. (1999). Arsenic in drinking water. Washington, DC:
National Academy Press.
134
CHILD HEALTH
AND THE
ENVIRONMENT
National Academy of Sciences. (2000). Toxicological effects of methylmercury.
Washington, DC: National Academy Press.
National Academy of Sciences. (2001). Arsenic in drinking water. 2001 update.
Washington, DC: National Academy Press.
Noonan CW, Sarasua SM, Campagna D, Kathman SJ, Lybarger JA, Mueller PW.
(2002). Effects of exposure to low levels of environmental cadmium on renal
biomarkers. Environ Health Perspect 110:151–5.
Oskarsson A, Schultz A, Skerfving S, Hallen IP, Ohlin B, Lagerkvist BJ. (1996). Total and inorganic mercury in breast milk in relation to fish consumption and
amalgam in lactating women. Arch Environ Health 51:234–41.
Palumbo DR, Cox C, Davidson PW, Myers GJ, Choi A, Shamlaye C, Sloane-Reeves
J, Cernichiari E, Clarkson TW. (2000). Association between prenatal exposure
to methylmercury and cognitive functioning in Seychellois children: a reanalysis of the McCarthy Scales of Children’s Ability from the main cohort
study. Environ Res 84:81–8.
Paschal DC, Burt V, Caudill SP, Gunter EW, Pirkle JL, Sampson EJ, Miller DT, Jackson RJ. (2000). Exposure of the U.S. population aged 6 years and older to cadmium: 1988–1994. Arch Environ Contam Toxicol 38:377–83.
Rasmussen PE, Subramanian KS, Jessiman BJ. (2001). A multi-element profile of
housedust in relation to exterior dust and soils in the city of Ottawa, Canada.
Sci Total Environ 267:125–40.
Rice DC, Hayward S. (1999). Comparison of visual function at adulthood and during aging in monkeys exposed to lead or methylmercury. Neurotoxicology
20:767–84.
Richardson GM. (1995). Assessment of mercury exposure and risks from dental
amalgam. Ottawa: Health Canada.
Seifert B, Becker K, Helm D, Krause C, Schulz C, Seiwert M. (2000). The German
environmental survey 1990/1992 (GerES II): reference concentrations of selected environmental pollutants in blood, urine, hair, house dust, drinking
water and indoor air. J Expo Anal Environ Epidemiol 10:552–65.
Smith AH, Arroyo AP, Mazumder DN, Kosnett MJ, Hernandez AL, Beeris M,
Smith MM, Moore LE. (2000). Arsenic-induced skin lesions among Atacameno
people in northern Chile despite good nutrition and centuries of exposure. Environ Health Perspect 108:617–20.
Steuerwald U, Weihe P, Jorgensen PJ, Bjerve K, Brock J, Heinzow B, BudtzJorgensen E, Grandjean P. (2000). Maternal seafood diet, methylmercury exposure, and neonatal neurologic function. J Pediatr 136:599–605.
Thomas DJ, Styblo M, Lin S. (2001). The cellular metabolism and systemic toxicity
of arsenic. Toxicol Appl Pharmacol 176:127–44.
U.S. Environmental Protection Agency. (1994). Cadmium. Located at http://www.
epa.gov/iris/subst/0141.htm.
U.S. Environmental Protection Agency. (1997). Mercury study report to Congress.
Research Triangle Park, NC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2000). 40 CFR Parts 141 and 142. National
drinking water regulations; arsenic and clarifications to compliance and new
source contaminants monitoring; proposed rule. Fed Reg 65:38888–983.
U.S. Environmental Protection Agency. (2001a). Current drinking water standards.
Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2001b). IRIS substance list. Located at
http://www.epa.gov/iris/subst/index.html
Metals—Mercury, Arsenic, Cadmium, and Manganese
135
U.S. Food and Drug Administration. (2000). Action levels for poisonous or
deleterious substances in human food and animal feed. Located at http://
www.cfsan.fda.gov/⬃lrd/fdaact.html.
U.S. Food and Drug Administration. (2001). An important message for pregnant
women and women of childbearing age who may become pregnant about the
risks of mercury in fish. Located at http://vm.cfsan.fda.gov/⬃dms/admehg.
html
Wallace L, Slonecker T. (1997). Ambient air concentrations of fine (PM2.5) manganese in U.S. national parks and in California and Canadian cities: the possible impact of adding MMT to unleaded gasoline. J Air Waste Manag Assoc
47:642–52.
Wheatley B, Paradis S. (1998). Northern exposure: further analysis of the results
of the Canadian aboriginal methylmercury program. Int J Circumpolar Health
57(Suppl 1):586–90.
Woolf A, Wright R, Amarasiriwardena C, Bellinger D. (2002). A child with chronic
manganese exposure from drinking water. Environ Health Perspect 110:613–16.
World Health Organization. (1990). Environmental health criteria 101. Methylmercury. Geneva: International Programme on Chemical Safety.
World Health Organization. (2000a). Guidelines for air quality. Geneva: World
Health Organization.
World Health Organization. (2000b). Guidelines for drinking water quality.
Geneva: World Health Organization.
Wulff M, Hogberg U, Sandstrom A. (1996). Cancer incidence for children born in
a smelting community. Acta Oncol 35:179–83.
Zierler S, Theodore M, Cohen A, Rothman KJ. (1988). Chemical quality of maternal drinking water and congenital heart disease. Int J Epidemiol 17:589–94.
6
PCBs, Dioxins, and
Related Compounds
Polyhalogenated aromatic hydrocarbons (PHAHs) comprise a large group
of semi-volatile chemicals that are stable at high temperatures, highly soluble in lipids, and resistant to biodegradation. Unfortunately, these properties enable PHAHs to disperse and persist in the environment, to bioaccumulate in terrestrial and aquatic food chains, and to cause unforeseen
adverse health effects among wildlife and humans. All of us probably
have detectable PHAHs in our bodies, the concentrations of PCBs generally being much higher than those of other PHAHs.
Chlorinated, brominated, and mixed halogenated PHAHs have similar structures and toxicity but widely variable potencies; major subgroups
include biphenyls, dibenzo--dioxins, and dibenzofurans (Table 6–1).
Monsanto, the sole manufacturer of PCBs in the United States, produced
about 700,000 tons during the period 1929–1979, annual output peaking
in 1970 at 43,000 tons. Given their high heat capacity and stability, PCBs
were ideal for uses in heat-resistant solvents, sealants, and lubricants, and
as dielectric fluids in electrical transformers, fluorescent light ballasts, and
other electrical equipment. The most intensely studied PHAH is 2,3,7,8tetrachloro--dibenzodioxin (TCDD), one of the most potent known toxicants (Schiestl et al., 1997).
This chapter describes the health threats to children of the widespread
use and dispersion of these highly toxic chemicals. The PCBs illustrate a
136
PCBs, Dioxins, and Related Compounds
137
TABLE 6–1. PHAHs: Types and Acronyms
Polychlorinated
Polybrominated
Dibenzo--dioxins
PCDDs
PBDDs
Biphenyls
PCBs
PBBs
Dibenzofurans
PCDFs
PBDFs
Diphenylethers
PCDEs
PBDEs
Naphthalenes
PCNs
PBNs
recurrent theme: the banning of a toxicant after recognition of severe human health impacts but continued use of chemically similar substances
with ill-defined human health risks. The key role of the aryl hydrocarbon
receptor in the toxicity of coplanar dioxin-like PHAHs and current knowledge of adverse health effects in humans and experimental animals are
discussed. The identified exposure sources and pathways point to the interventions needed to further reduce environmental contamination and
human exposures. See also Chapter 7 (Pesticides) and Chapter 8 (Hormonally Active Agents) for further discussion of the potential effects of
PHAHs and related compounds on reproductive system development
and function.
Health Effects
Several events led to the discovery of the toxicity of PCBs, dioxins, and
related compounds:
• A major outbreak of chloracne (a severe acne-like skin eruption) occurred at a trichlorophenol plant in West Virginia after a reactor
explosion in 1949. Only later was it realized that combustion of trichlorophenol can generate large amounts of TCDD and dioxin-like
compounds, now known to cause chloracne.
• The deaths of millions of chickens from chick edema in the United
States during the late 1950s and 1960s were linked to poultry feeds contaminated by fat-soluble chlorinated aromatic compounds, including
polychlorinated dibenzo- -dioxins (PCDDs); the contamination was
traced to pentachlorophenol used in hide-stripping operations followed
by removal of fat and alkaline hydrolysis to produce fatty acids for animal feeds. Pentachlorophenol is soluble in fat and yields PCDDs during alkaline hydrolysis.
CHILD HEALTH
138
AND THE
ENVIRONMENT
• Investigation of unexpected gas chromatographic peaks during analyses of tissues from fish and marine mammals revealed the bioaccumulation of PCBs and DDT/DDE in the aquatic food chain.
• Outbreaks of an apparently new illness occurred in Japan (1968) and
Taiwan (1979) characterized by severe acne-like skin eruptions and
other signs and symptoms in children and adults and by low birth
weight, pigmentation, developmental delays, and cognitive deficits in
prenatally exposed infants. Both outbreaks were traced to leakage of
PCBs and other PHAHs from heat-exchange coils.
Epidemiologic studies have focused on populations exposed to PHAHs
through diet (consumption of highly contaminated cooking oils or moderately tainted fish), occupation, and industrial accidents. In general,
PHAH exposures of these populations involved mixtures, precluding
definitive assignment of toxic effects to specific congeners. Some health
effects attributed to PCBs, for example, may have been caused by other
PHAHs.
Molecular Mechanisms
The three major PHAH families each have two aromatic rings but differ
in how the rings are joined (Fig. 6–1). The major families comprise over
400 members (congeners) that vary by the number and placement of chlorine or bromine atoms. Among PCBs, those with zero or one chlorine in
the ortho position (positions 2, 2, 6, or 6 on biphenyl (see Fig. 6–1) are
TABLE 6–2. Equitoxic Levels of Selected PHAHs in the
Diet of Rhesus Macaque Monkeys
Compound
Level (ng/g)
Aroclor 1242
100,000
Arcola 1248
25,000
Firemaster FF-1
25,000
3,4,3,4-Tetrachlorobiphenyl
1,000
3,4,5,3,4,5-Hexachlorodibenzofuran
1,000
2,3,7,8-Tetrachlorodibenzofuran
2,3,4,6,7,8-Hexachlorodibenzo--dioxin
50
5
2,3,6,7-Tetrachlorodibenzo--dioxin
0.5–2
2,3,7,8-Tetrachlorodibenzo--dioxin
1.0
Source: McNulty (1985).
Equitoxicity based on red, swollen eyelids and elevated fingernails after 1 month and severe morbidity or death after 2 months.
PCBs, Dioxins, and Related Compounds
139
3
Biphenyl
2
3'
4
4'
5
3,5,3',5' Tetrachlorobiphenyl
6
6'
5'
Cl
Cl
Cl
Cl
1
2
Dibenzo--dioxin
2'
10
O
9
8
7
3
4
O
5
6
O
Cl
2,3,7,8-Tetrachlorodibenzo--dioxin
Cl
Cl
O
Cl
2,3,7,8-Tetrachlorodibenzofuran
Cl
Cl
Cl
O
Cl
FIGURE 6–1. Structure of biphenyls, dioxins, and selected chlorinated congeners.
often called coplanar because they have a flat configuration. Less than 10%
of PHAH congeners, those with chlorine and/or bromine in at least the
2, 3, 7, and 8 positions, have dioxin-like toxicity; certain halogenated benzenes and naphthalenes have similar toxicity (U.S. Environmental Protection Agency, 2000b). Congeners with zero, one, or two or more chlorine atoms in the ortho position, respectively, have high, intermediate, or
low AhR affinities and dioxin-like toxicities. The toxicity of PHAHs varies
widely, with PCDDs and polychlorinated dibenzofurans (PCDFs) generally being much more potent than PCBs (Table 6–2). Among the 209 PCB
congeners, the 13 coplanar members with four or more chlorines have
dioxin-like toxicity. Some evidence, however, suggests that noncoplanar
PCB congeners are more neurotoxic than coplanar compounds and may
act through AhR-independent mechanisms.
AhR-Mediated Toxicity
TCDD and dioxin-like compounds appear to have three main biochemical effects: enzyme induction and modulation of multiple growth factors
140
CHILD HEALTH
AND THE
ENVIRONMENT
and hormones (Birnbaum, 1995). The reproductive, developmental, immunologic, and carcinogenic effects of dioxin-like PHAHs occur at extremely low body-burden levels in experimental animals and are mediated by AhR, a ligand-dependent transcription factor. The AhR occurs in
human tissues and functions similarly to that in rodents but generally has
lower affinity for PHAHs, possibly explaining the higher susceptibility of
rodents to PHAH toxicity. When activated by TCDD or other ligands, AhR
dimerizes with AhR nuclear translocator (Arnt) and binds with high affinity to enhancer elements in the upstream region of several oncogenes and
genes that encode growth factors, hormones, hormone receptors, cytochrome P450 enzymes (CYP1A1, CYP1A2, CYP1B1), and electrophile response element (EPRE) genes (Safe, 2001). Upregulation of these genes
can enhance or mitigate the effects of toxicants and hormones; for instance, CYPlA1 and CYPlA2 increase the metabolism of endogenous and
exogenous substances to reactive oxygenated metabolites and cause oxidative stress, a major signal for triggering apoptosis. The most potent
known inducer of the CYP1A1 gene is TCDD.
The ability of dioxin-like PHAHs to cause wasting, immunosuppression, birth defects, chloracne, and cancer in experimental animals
and/or humans may arise from persistent AhR activation due to their
strong binding affinity and long biologic half-life. Knockout mice missing the AhR gene are not susceptible to TCDD and are resistant to
benzo[a]pyrene-induced tumors, demonstrating the key role of AhR in
carcinogenesis. The duration of health effects depends on the half-lives of
congeners in body fat; for example, recovery is rapid from certain PHAH
congeners with short half-lives but protracted for persistent congeners.
Toxic Equivalency
A PHAH congener is dioxin-like if it binds to AhR, causes dioxin-like effects, and bioaccumulates. Toxic equivalency factors (TEFs) for PHAHs
are the ratios of the toxicity of specific congeners relative to that of TCDD.
There is no universally accepted set of TEFs, and congener-specific TEFs
may vary among species; TEFs developed by the U.S. EPA and the WHO
are widely used. The toxic equivalent (TEQ) of each congener in a mixture is the product of its TEF and its molar concentration; the sum of TEQs
in a mixture is the amount of TCDD estimated to equal the toxicity of the
mixture. Although use of TEQs entails uncertainties, they address the
need to assess health risks of the mixed PHAH exposures among humans.
For instance, the affinity of TCDD for AhR is about 10,000 times that of
hexachlorobenzene, but the latter occurs at much higher concentrations,
comprising up to 60% of total TEQ in breast milk in some countries.
PCBs, Dioxins, and Related Compounds
141
AhR-Independent Mechanisms
Ortho-substituted (noncoplanar) PCB congeners appear to cause neurotoxic effects through AhR-independent mechanisms in experimental
animals. Reduced dopamine levels in the caudate nucleus, putamen, and
hypothalamus of adult monkeys exposed to commercial PCB mixtures
are associated with brain tissue PCB residues composed almost entirely
of di-ortho-substituted congeners. Noncoplanar PCB congeners are the
most potent in reducing dopamine levels and disrupting calcium transport in human pheochromocytoma cells in vitro. Finally, the ability of ortho-substituted PCBs to bind with high affinity to the thyroid hormone
transport protein transthyretin may contribute to the neurotoxicity of prenatal PCB exposure.
Reproduction
See Chapter 8 (Hormonally Active Agents) for discussion of the potentially adverse effects of dioxins, PCBs, and related compounds on reproductive health.
Development
In 1968, a 3-year-old girl with severe acne-like skin lesions visited a dermatology clinic in Japan and became the first recognized case of a new
disease called Yusho (oil disease). About 1800 persons were eventually
recognized to have Yusho, characterized in older children and adults by
chloracne, eye discharge, hyperpigmentation, fatigue, nausea, elevated
serum triglyceride levels, and reduced sensory nerve conduction velocity (Urabe and Asahi, 1985). Prenatally exposed infants experienced increased perinatal mortality, intrauterine growth retardation, small head
circumference, eye discharges, hyperpigmentation (skin, nails, buccal cavity), gingival hyperplasia, precocious dentition, and abnormal skull calcification (Yamashita and Hayashi, 1985). Lengthy investigations identified the cause to be cooking oil contaminated by PCBs leaking through a
heating coil during rice oil production. Subsequent studies revealed other
highly toxic contaminants, including PCDFs, polychlorinated triphenyls,
and polychlorinated quaterphenyls in the contaminated oil and blood and
adipose tissue samples from exposed persons.
In 1979, acne-like skin lesions occurred among the students and staff
of a school for blind children in Taiwan; over 2000 Yucheng (oil disease)
cases were eventually recognized. Some children developed severe, scarring chloracne (Fig. 6–2). Follow-up at ages of up to 6 years showed that
142
CHILD HEALTH
AND THE
ENVIRONMENT
FIGURE 6–2. Chloracne on the face of a Yusho case (Urabe and Asahi, 1985).
prenatally exposed Yucheng children had delayed developmental milestones, abnormal behaviors, lower stature and weight, gingival hypertrophy, tooth chipping, acne or acne scars, hyperpigmentation, lymphadenopathy, hirsutism, hypertelorism (eyes abnormally far apart), and short
metacarpal bones (Rogan et al., 1988). Despite the close clinical resemblance to Yusho and epidemiologic evidence showing an association with
PCBs, Dioxins, and Related Compounds
143
a specific brand of rice oil, several months passed before Taiwanese authorities formally recognized the cause and prohibited further distribution of the oil. PCDFs, PCBs, and PCDDs, respectively, contributed 53%,
35%, and 12% of the dioxin-TEQ in the contaminated cooking oil.
The Yusho and Yucheng incidents showed that high prenatal exposures to PCBs and other dioxin-like PHAHs can cause low birth weight
and structural and functional abnormalities. A review concluded that
there is inconsistent evidence in populations exposed to relatively lowlevel PCBs (from background sources such as fish) of inverse associations
between maternal PCB levels and birth weight or gestation length (Longnecker et al., 1997); recent European studies, however, have shown fairly
consistent associations between maternal PCB or dioxin-TEQ exposure indices and low birth weight (Patandin et al., 1998; Rylander et al., 1998;
Vartiainen et al., 1998). Prenatal PCB exposures were associated with
growth deficits during early childhood in three cohort studies (Blanck et
al., 2002; Gladen et al., 2000; Jacobson et al., 1990). Follow-up of daughters of women prenatally exposed to polybrominated biphenyls (PBBs)
showed a strong inverse association between maternal serum PCBs, but
not PBBs, and weight adjusted for height at ages 5–24 years (average difference of 11 pounds for maternal serum PCB levels above the median
versus below the median) (Blanck et al., 2002). Stature at age 10 years was
not associated with blood PCB levels measured at age 8 years (Karmaus
et al., 2002).
An explosion at a trichlorophenol plant in Seveso, Italy, in 1976 exposed the surrounding population to relatively pure TCDD. The ratio of
male:female births was reduced among the highest exposure group at
Seveso (Mocarelli et al., 2000) but not among Yusho infants (Yoshimura
et al., 2001). Among the 26 births in the most contaminated part of Seveso,
2 had minor structural defects and none had major malformations. An
early case-control study showed an increased risk of some birth defects
related to likely paternal exposure to Agent Orange in Vietnam, but later
studies found no consistent associations between paternal serum TCDD
levels and risks of fetal death, birth defects, preterm birth, intrauterine
growth retardation, infant death, or altered sex ratio.
Developmental effects of TCDD in experimental animals include fetal death, cleft palate, hydronephrosis (caused by ureteral epithelial hyperplasia), hypomineralization of teeth, ovarian atrophy, cleft phallus (female), reduced size of testes and male accessory sex glands (especially
the ventral prostate), delayed testicular descent, reduced anogenital distance (males), reduced growth, thymic atrophy, and immunosuppression
(Agency for Toxic Substances and Disease Registry, 1998) (see also Chapter 8, Hormonally Active Agents).
144
CHILD HEALTH
AND THE
ENVIRONMENT
Neurotoxicity
Epidemiologic Studies
Although some Yusho infants had reduced head circumference, there appear to have been no follow-up studies to assess neurobehavioral function. Yucheng children had global IQ deficits of about 5 points at ages 4–7
years; those born up to 6 years after their mothers’ exposure were just as
affected as those born within 2 years of exposure (Chen et al., 1992). These
findings are consistent with the long half-lives of many PHAH congeners
in vivo; serum PCB and PCDF levels in Yucheng women 14 years after
exposure were still 10–100 times higher than those of controls. A birth cohort study of children whose mothers were exposed to PCBs from eating
contaminated Lake Michigan fish showed that prenatal PCB exposure was
associated with weak reflexes and an increased startle response at birth,
visual recognition memory deficits during infancy, verbal and numerical
memory deficits at age 4 years, and full-scale and verbal IQ deficits at age
11 years (Jacobson and Jacobson, 1996). Children with the highest prenatal PCB exposure had an average full-scale IQ deficit of 6 points, independent of potential confounders, and were two to three times more likely
to score poorly on verbal comprehension, freedom from distractibility,
and written word comprehension. Benchmark dose analyses of the Lake
Michigan cohort were conducted for four PCB-related cognitive outcomes:
full-scale IQ, word comprehension, and average reaction time at age 11
years and the McCarthy Memory Scale at age 4 years (Jacobson et al.,
2002); the estimated benchmark dose for a 5% increased incidence of fullscale IQ deficits was 0.36–0.63 g/g (average lipid-adjusted PCB level in
cord and maternal serum and breast milk).
A systematic review of seven cohort studies of prenatal PCB exposure and neurologic development noted associations with abnormal reflexes among newborn infants, reduced motor skills among infants, and
cognitive deficits at about age 4 years; importantly, lactational PCB exposure was not clearly associated with any adverse neurologic effect
(Ribas-Fito et al., 2001). A recent German study, however, reported independent effects of prenatal and postnatal PCB exposures on cognitive
deficits at age 42 months (Walkowiak et al., 2001). Although divergent
from most other studies, this finding is consistent with evidence that postnatal exposure of monkeys to a relatively low dose of PCBs can cause
learning deficits (Rice, 1997).
Cognitive deficits observed in Michigan but not in North Carolina
might be attributable to differences in exposure levels. For example, a later
study showed that persons who averaged one meal of Great Lakes fish
per week had mean serum dioxin-TEQ and PCB-TEQ levels, respectively,
about twofold and tenfold those of nonconsumers (Anderson et al., 1998).
PCBs, Dioxins, and Related Compounds
145
Other possible differences include “spiking” of serum PCB levels among
fish eaters (serum PCB levels increase two- to fivefold after a PCBcontaminated fish meal and require a week to return to baseline levels)
and the presence of toxic contaminants other than PCBs in Lake Michigan fish.
Toxicologic Studies
Perinatal exposure of rodents and nonhuman primates to PCBs causes
neurotoxic effects involving motor activity (increased open field activity
during early life, hypoactivity as adults), cognitive function (decreased
active avoidance learning, increased errors in radial arm maze, decreased
performance on spatial and nonspatial discrimination reversal tasks), and
altered dopamine levels in several brain regions (Agency for Toxic Substances and Disease Registry, 2000). Prenatal PCB exposure also causes
markedly low serum thyroxine (T4) levels in experimental animals; perinatal hypothyroidism is a known cause of cognitive deficits in animals
and humans and may contribute to the neurotoxicity of PCBs. Although
neurotoxicity of PCBs may relate mainly to noncoplanar congeners, coplanar PCBs may also be important; for instance, rats perinatally exposed to
3,3,4,4,5-pentachlorobiphenyl developed low-frequency hearing deficits
at doses as low as 1 g/kg/day (Crofton and Rice, 1999). There has been
little research on neurotoxicity of TCDD in animals.
Cancer
Dioxins
The IARC concluded that TCDD is a human carcinogen based on limited
human evidence from occupationally exposed persons, strong evidence
of carcinogenicity in experimental animals, and a shared mechanism
(AhR) in animals and humans. The IARC concluded that the human carcinogenicity of other dioxins and furans could not be determined. The
EPA independently concluded that TCDD is a human carcinogen, other
dioxins are probable human carcinogens, and general population exposure levels to TCDD may increase the lifetime (up to age 70 years) absolute cancer risk by as much as 0.1–1.0% (U.S. Environmental Protection
Agency, 2000b).
Few studies of TCDD and childhood cancer have been conducted.
People exposed as children to TCDD at Seveso had statistically insignificant excesses of ovarian cancer, Hodgkin’s disease, myeloid leukemia,
and thyroid cancer (Pesatori et al., 1993). Given the limited exposure assessment, the small number of highly exposed children, and the usual
long latency of human cancers, longer follow-up is needed to assess potential cancer risks. Children living near municipal or hospital incinera-
146
CHILD HEALTH
AND THE
ENVIRONMENT
tors, potential sources of TCDD exposure, had a twofold increased risk
of leukemia and other cancers (Knox, 2000). The EPA concluded that existing epidemiologic data are insufficient to assess the risk of childhood
cancer from TCDD exposure. However, TCDD is a potent animal carcinogen, causing cancer (usually at multiple sites) in all species tested at
doses as low as 1 ng/kg/day. Rats exposed prenatally to a single low dose
of TCDD develop more mammary gland terminal end buds, the breast
structures most susceptible to carcinogenesis; such animals are more susceptible as adults to chemically induced breast cancer.
PCBs and Related Compounds
The IARC concluded that several PCB mixtures are reasonably anticipated
to be human carcinogens based on sufficient evidence in animal studies
and inadequate human evidence (International Agency for Research on
Cancer, 1987). Cohort studies of occupationally exposed adults showed
slightly increased risks of skin melanoma and liver, biliary tract, gallbladder, colorectal, and hematopoietic cancers. In experimental animals,
PCBs caused liver cancer and hyperplasia of the bile duct, gallbladder,
and urinary tract, with the higher chlorinated PCBs being most potent.
The EPA estimated that lifetime ingestion of PCBs at a rate of 1 g/kg/day
would cause an extra 77 cancer cases per 10,000 persons.
Population-based epidemiologic studies have shown associations between PCBs and adult cancers, including non-Hodgkin’s lymphoma
(Rothman et al., 1997) and breast cancer. Although a review concluded
that the epidemiologic evidence concerning PCBs and breast cancer was
mixed and inconclusive, two subsequent well-conducted studies showed
associations with breast adipose tissue levels of specific PCB congeners
(Aronson et al., 2000; Holford et al., 2000). Breast adipose tissue PCB levels reflect cumulative exposure in the target organ of interest and likely
provide a better indicator of the breast cancer risk. Studies that assessed
total blood PCB levels may be misleading because breast tissue levels of
specific PCB congeners are only weakly correlated with blood levels and
specific PCB congeners vary widely with respect to estrogenic activity,
affinity for AhR, cytochrome P450 response, and half-lives in vivo. Apart
from a small case-control study of childhood leukemia in Germany showing no associations with PCB, DDE, hexachlorocyclohexane, or dieldrin
levels in bone marrow fat (Scheele et al., 1992), the role of PCBs in childhood cancer remains virtually unexplored.
Immune System
Yucheng children reexamined at ages 8–16 years were much more likely
than matched controls to have experienced recent middle ear infections
PCBs, Dioxins, and Related Compounds
147
but had normal immunologic markers (Chao et al., 1997; Yu et al., 1998).
Among Dutch children, current plasma PCB levels were associated with
recurrent middle ear infections, chicken pox, and allergies but not with
antibody levels, leukocyte counts, or T-cell markers (Weisglas-Kuperus
et al., 2000). Among Inuit infants, the risk of otitis media before age 1 year
was associated with prenatal exposure to DDE, hexachlorobenzene, and
dieldrin but not PCBs (Dewailly et al., 2000).
The EPA and the ATSDR assessed the immunologic effects of dioxins, PCBs, and related PHAHs in humans and experimental animals and
concluded that (Agency for Toxic Substances and Disease Registry, 1998,
2000; U.S. Environmental Protection Agency, 2000a)
• No consistent exposure-related immunologic effects have been observed in adult humans exposed to PCDDs at levels several orders of
magnitude above background levels.
• Yucheng infants and infants whose mothers ate fish or marine mammals contaminated by PCBs and other organochlorine compounds
had increased infections (middle ear infections, bronchitis) during early
infancy.
• The immune system is one of the most sensitive targets for dioxins in
experimental animals. Perinatal exposure to low TCDD doses causes
thymic gland atrophy, and much lower doses disrupt specific immune
receptor functions, with a reduced antibody response and decreased
host resistance to infections.
• TCDD and related PHAHs should be considered nonspecific immunosuppressants, at least until better data are available.
Endocrine System
See Chapter 8 (Hormonally Active Agents) for discussion of potential effects of dioxins and related compounds on the endocrine system.
Chloracne
Chloracne is the earliest clinically recognizable and most consistent health
effect of exposure to dioxin-like PHAHs in several species, including humans and monkeys. Based on TCDD levels in soil and a limited number
of blood samples, chloracne was the only dose-related health effect among
the local population at Seveso; 88% of the 187 cases diagnosed by an expert panel were children. Chloracne developed in children over a period
of 3 days to 2 weeks after the explosion; lipid-adjusted serum TCDD levels were highest among those most severely affected and were several orders of magnitude higher than adipose tissue levels among U.S. children
CHILD HEALTH
148
AND THE
ENVIRONMENT
TABLE 6–3. TCDD Levels among Seveso Children
Aged 3–14 Years with Chloracne and U.S. Populations
pg/g Fat
Grade 4 (severe chloracne)
828–7,420a
Grade 3
U.S. children
U.S. general population
a Serum.
2b
5.4c
Mocarelli et al. (1991).
bAdipose
c Whole
12,100–56,000a
tissue (Orban et al., 1994).
blood (Needham et al., 1996).
(Table 6–3). Children but not adults with serum TCDD levels below 10,000
pg/g fat developed chloracne, suggesting that children were more susceptible. The Seveso incident caused the greatest documented human
TCDD exposures, the highest blood TCDD levels occurring among six
children with grade 4 chloracne (12,100–56,000 ng/L) (Pocchiari et al.,
1979). All children with chloracne soon after exposure were clear of lesions by 6 years after the incident.
Based on levels in blood samples and contaminated cooking oil, persons (mainly children) who developed chloracne in the Yusho and
Yucheng incidents had average exposures to PCDFs of 4–6 g/kg, measured as 2,3,4,7,8-pentachlorodibenzofuran equivalents (PEQ) (Ryan et al.,
1990). These exposures are toxicologically equivalent to TCDD levels
known to cause chloracne in monkeys and are about 200 times higher
than current body burden levels in North American populations. It is possible that TCDD causes chloracne by interfering with the regulatory role
of retinol (vitamin A) and retinoic acid in skin epithelial cell proliferation
and differentiation.
Other Effects
Some of the children exposed to PCBs and other PHAHs during the Yusho
and Yucheng incidents had natal teeth, discoloration of teeth during infancy, and tooth chipping during later childhood. Combined prenatal and
postnatal PCB exposure has been linked to enamel defects in permanent
teeth. Lactational exposure to dioxin-TEQs in Finland was associated with
the frequency and severity of mineralization defects in permanent first
molars at relatively low exposure levels; such changes may provide a lifelong visible indicator of exposure to TCDD and dioxin-like compounds
PCBs, Dioxins, and Related Compounds
149
(Alaluusua et al., 1999). After a single postnatal dose of TCDD, young
adult male rats experienced defective dentin and enamel formation, possibly due to interference with vitamin A by TCDD.
Exposures
Internal Dose Indices
The body PHAH burden is stored mainly in adipose tissue and the liver.
Until the late 1980s, the gold standard for TCDD body burden assessment
required a biopsy of adipose tissue, a highly invasive procedure. After the
discovery that lipid-based TCDD concentrations are similar in adipose
tissue, liver, and serum, and with the the development of sensitive analytic methods, blood analysis became the most common method for estimating human exposure to TCDD and other PHAHS. But the correlation
between serum and adipose tissue levels of individual PCB congeners
varies substantially; also, the relative amounts of coplanar PCBs in adipose tissue vary more among individuals than do PCDD and PCDF levels. The persistence of individual PCB congeners in blood varies widely;
after cessation of chronic exposure, the half-lives of nine PCB congeners
in monkeys ranged from a few months to about 8 years.
Internal Dose Estimates
After 13 years of follow-up, serum PCB and PCDF levels among Yucheng
women were still one to two orders of magnitude higher than those in
controls and were inversely associated with the total duration of breastfeeding, consistent with the importance of lactation as a PHAH excretory
mechanism. Dioxin-TEQ levels in placental samples from Yucheng
women were up to 1000 times those observed in American women. Compared to unexposed persons, frequent consumers of Great Lakes fish had
twofold higher mean serum TEQs for dioxins and furans and tenfold
higher TEQs for coplanar PCBs (Anderson et al., 1998). Monitoring data
for PHAH levels in the general population are virtually nonexistent; a
small German study of neonates showed relatively low cord blood levels
of individual PCB congeners and hexachlorobenzene, consistent with a
substantial decline during 1994–1998 (Lackmann, 2002). The CDC plans
to include dioxins, furans, and PCBs in NHANES in future years (Centers for Disease Control and Prevention 2001).
Breast milk PHAH levels reflect the body burden of reproductive-age
women and the potential for prenatal and lactational infant exposure. Al-
CHILD HEALTH
150
AND THE
ENVIRONMENT
though limited by small sample sizes and methodologic differences, existing data suggest that breast milk dioxin-TEQ levels declined by about
50% in several countries during the period 1970–1998 (LaKind et al., 2001).
To the extent that there has been a real decrease, this would be consistent
with reduced PHAH emissions.
Breast milk PCB levels in Canada have decreased since 1982 (Craan
and Haines, 1998). However, PCB levels in breast milk from Inuit women
in the Canadian Arctic were up to ten times those in southern Canada,
reflecting high consumption of contaminated traditional foods (Dewailly
et al., 1993). In Sweden, total breast milk dioxin-TEQ decreased by about
two-thirds during recent decades, but polybrominated diphenyl ethers
(PBDEs) increased sharply (see Fig. 6–3) (Noren and Meironyte, 2000). If
the latter trend continues, PBDEs will displace PCBs and DDT as the leading breast milk organochlorines during the next 15–30 years.
Risk Management
The PHAHs from manufacturing or combustion processes and waste disposal disperse widely through evaporation, long-range airborne transport, and deposition into soil and water, persisting in and cycling between
4500
120
4000
100
3000
80
2500
60
2000
1500
40
PBDE (pg/g fat)
TEQ (pg/g fat)
3500
1000
20
500
0
0
1972
1980
1988
1991
1994
1997
year
TEQ
PBDE
FIGURE 6–3. Breast milk contaminant levels, Sweden, 1972–1997 (chart is based on
data in Noren and Meironyte, 2000). PBDE polybrominated diphenyl ethers;
TEQ TCDD toxic equivalents.
PCBs, Dioxins, and Related Compounds
151
environmental compartments and biota. Widely used during the midtwentieth century, PCBs are now tightly controlled chemicals but substantial amounts remain in use, mainly in electrical equipment manufactured before the 1979 ban. Large amounts of PCBs from electrical
equipment removed from service and other sources remain in storage and
waste disposal sites, subject to fires that generate extremely toxic PCDFs
and PCDDs. The EPA estimates that up to 200 chemical processes, including production of chlorinated solvents, paints, printing inks, plastics,
and detergent bars, may inadvertently generate PCBs.
In general, PCDDs and PCDFs arise unintentionally through four
main processes: (1) combustion of wastes and fossil fuels, (2) metal smelting, refining, and processing (including the reprocessing of scrap metal
from cars and other products containing plastics), (3) production, use, and
disposal of chlorinated organic compounds, and (4) production of chlorine-bleached wood pulp. Other sources include photolysis of highly
chlorinated phenols and reservoir sources, that is, contaminated soils, sediments, biota, and water.
Airborne concentrations of PCDDs, PCDFs, and PCBs in Europe and
the United States were low before the 1930s, increased until about 1970,
and then declined. Total PCBs in Great Lakes sediments and biota decreased during the 1970s and 1980s, but this trend slowed during the
1990s. Industrial production of chlorinated organic chemicals and chlorine use by pulp and paper mills increased during 1930–1970, followed
by a period of increasing pollution abatement activities. The latter included elimination of most open burning, particulate controls on combustors, the phase-out of leaded gas, bans on PCBs, 2,4,5-trichlorophenoxyacetic acid (2,4,5-T), and hexachlorophene and restrictions on the
use of pentachlorophenol. Emissions of PCDD and PCDF in the United
States declined about 80% during 1987–1995, mainly due to reduced municipal and medical waste incinerator emissions.
Food
Sources
The vast majority of dioxin-TEQ intake is from dietary sources, especially
meat and dairy products (Table 6–4). The primary pathways for PCDDs
in the human diet are air-plant-animal and air-water-fish, with bioaccumulation factors generally higher for congeners with a higher chlorine
content. Dairy products from cows grazed on pasture contaminated by
the application of sewage sludge (as fertilizer) can raise the daily dietary
intake of PCDDs and PCDFs by up to 40%. Among pregnant Dutch
women, dairy products and oils of animal origin contributed, respectively,
CHILD HEALTH
152
AND THE
ENVIRONMENT
TABLE 6–4. Estimated Average Daily Intake
of Dioxin-TEQ by the U.S. Population
Source/Pathway
Intake (pg/day)
Air
0.4
Water
0.004
Soil (ingestion)
0.02
Food sources (total)
Fruits and vegetables
Dairy products
Beef
Fish
Eggs
34.4
1.2
8.0
18.0
6.7
0.5
Total intake
34.8
Source: Travis (1991).
about 50% and 25% of dietary intake of PCDDs and coplanar PCBs. In
the United Kingdom, the estimated daily dietary intake of PCDD/PCDF
and PCB TEQs each declined by about 70% during 1982–1992. Similarly,
estimated adult daily PHAH intakes in the United States peaked about
1970 and then declined.
Breast milk is the major exposure source during infancy, far exceeding prenatal exposures. Infants absorb 90%–100% of tetra and higher chlorinated PCB congeners from breast milk. A U.S. national survey showed
that average total dioxin-TEQ intakes in breast-fed infants (42 pg/kg/day)
were 10–20 times those in adolescents or adults (Schecter et al., 2001); in
all age groups, average daily intakes exceeded the ATSDR MRL for chronic
exposure (1 pg/kg/day) (Table 6–5). Infant blood PCB levels increase with
the duration of breast-feeding and may be up to fivefold higher than those
of formula-fed children. The total daily TEQ intake among breast-fed
infants in Holland is about 110–120 pg/kg, substantially higher than
the tolerable daily intake for the general population set by the WHO
(1–4 pg/kg). Breast milk PHAH levels decline by over 50% after prolonged breast-feeding, reflecting the importance of lactation as an excretory mechanism.
Children consume more food per unit body weight than adults and
substantially more for specific foods such as milk. Analysis of hamburgers, pizza, deep-fried chicken, and ice cream showed that daily total
dioxin-TEQ exposure (per unit body weight) from consumption of these
foods would be almost fourfold higher for children than adults (Schecter
and Li, 1997). Dairy products, meat, and processed foods, respectively,
contributed 43%–50%, 14%–19%, and 15%–23% of PCB/dioxin-TEQ in-
PCBs, Dioxins, and Related Compounds
153
TABLE 6–5. Standards and Guidelines for TCDD or Dioxin TEQs
Medium
Standard
Agency
Air
Emission limits
EPAa
Drinking water
3 ⫻ 108 mg/L
EPAb
5 pg/g fat
3 pg/g fat
Francec
Belgiumd
1 ng/g
0.05 and 1 ng/g
0.05 ng/g
ATSDR
ATSDR
ATSDR
1 pg/kg/day
1–4 pg/kg/day
ATSDR f
WHO g
Foods of animal origin
Milk and dairy products
Pork and derived products
Soil e
Action level
Evaluation level
Screening level
Total daily intake
MRL (chronic exposure)
TCDD and dioxin-like compounds
a The
EPA has regulations for release of dioxin-like compounds from air sources and from pulp and paper mills into water bodies.
b U.S.
Environmental Protection Agency (2001).
c France
(World Health Organization, 2001).
d Belgium
(World Health Organization, 2001).
e Dioxin-TEQ
f Agency
g Van
(De Rosa et al., 1997).
for Toxic Substances and Disease Registry (2002).
Leeuwen et al. (2000).
take among toddlers (Patandin et al., 1999). In the United States, average
dioxin-TEQ levels were highest in farm-grown freshwater fish fillets
(1.7 pg/g), intermediate in ocean fish, beef, chicken, pork, sandwich meat,
eggs, cheese, ice cream, and human milk (0.33–0.51 pg/g), and lowest in
a simulated vegan diet (0.09 pg/g) (Schecter et al., 2001). Intake of dietary
dioxins in some European countries has declined by about 50% since the
late 1980s.
Intervention
Dietary intervention during pregnancy would make little difference in fetal exposure; effective intervention will require lifelong reduced PHAH
dietary intake, especially from foods of animal origin. In its three risk assessments of PHAHs since 1987, the WHO each time has concluded that
there are known benefits of breast-feeding, that evidence of harm from
PHAHs is insufficient to limit breast-feeding or to eliminate specific foods
from the diet, that reducing the release of PHAHs to the environment is
the best way to minimize human exposures, and that it should continue
to endorse breast-feeding (Brouwer et al., 1998).
CHILD HEALTH
154
AND THE
ENVIRONMENT
Since the 1970s, fish advisories have been used to reduce consumption of contaminated sport fish, especially by women. Although almost
10% of Great Lakes States residents consume regional sport fish, less than
half of exposed women are aware of the advisories. During 1986–1989,
Mohawk women at Akwasasne in New York had breast milk PCB levels
almost double those of nonaboriginal women and associated with fish
consumption; after issuance of fish advisories and decreased local fish
consumption, breast milk PCB levels declined to general population levels by 1990–1992. See Table 6–6 for reference doses and standards for PCBs.
Air and Water
Sources
The EPA estimates that PCDD/PCDF releases totaled about 3.3 kg TEQ
in the United States during 1995 compared to about 14 kg in 1987. Reduced municipal and medical waste incinerator releases accounted for
most of the improvement, but these sources plus burning of refuse in barrels still comprise about 70% of quantifiable emissions. Outdoor air has
generally been a minor direct source of human TCDD exposure, with important exceptions including the Seveso incident. Low levels of PCBs are
widely found in surface waters but rarely occur at high levels in groundwater because of their low solubility; exceptions include groundwater
contaminated by leachates from hazardous waste sites. Chlorine bleached
pulp and paper production has been a major source of dioxins in water.
Intervention
The Clean Air Act and the Clean Water Act require the EPA to set emissions limits and ambient water quality criteria for TCDD (Table 6–5). The
TABLE 6–6. Standards and Guidelines for PCBs
Medium
Standard
Drinking water
0.5 g/La
Foodb,c
Eggs
Milk
Poultry
Fish (edible portion)
Red meat
0.3 g/g fat
1.5 g/g fat
3 g/g fat
2 g/g
3 g/g fat
a U.S.
Environmental Protection Agency (2001).
b U.S.
Food and Drug Administration (2000b).
c U.S.
Food and Drug Administration (2000a).
PCBs, Dioxins, and Related Compounds
155
1998 pulp and paper effluent guidelines should eliminate 96% of TCDD
discharges from this industry, once the largest industrial source of TCDD
in water. The EPA maximum contaminant level for TCDD in public drinking water is 3 ⫻ 108 mg/L. Virtual elimination of PHAH air emissions
is essential to reduce population exposures to these toxicants over the
long term, including prenatal and lactational exposures at critical stages
of growth and development. The May 2001 Stockholm Convention legally
binds the United States, the European countries, and several other countries to (1) regulate 16 persistent organochlorine compounds, ban production and use of 8 organochlorine pesticides, and ban production and
limit uses of PCBs, (2) require the use of the best available technology to
limit air emissions from major stationary sources of PCDDs, PCDFs, and
certain other toxicants, (3) limit PCDD/PCDF emissions from waste incinerators, (4) reduce total national air emissions below the levels for a
reference year, and (5) manage stockpiles of waste persistent organochlorine pollutants (POPs) in an environmentally sound manner.
Other Sources
Urban levels of the higher chlorinated PCDD congeners generally exceed
those in rural areas and are associated with local combustion sources including incinerators. Sludges contaminated with PHAH from pulp and
paper mills and sewage treatment plants have been widely used as agricultural soil enhancers. Although food crops take up only small amounts
of PHAHs from soil, continuation of this practice could cause higher uptakes in the future as PHAHs accumulate in soil.
There has been no systematic survey of dioxins in commercial products. Pentachlorophenol (PCP) and the herbicide 2,4,5-T were banned because of concern about their content of dioxins and other PHAHs. The EPA
reported that 14% of 2,4-D samples contained 1,2,3,7,8-pentachlorodibenzo-dioxin, a congener that is as potent as TCDD. Manufacturers have cooperated with the EPA to prevent the manufacturing of new products significantly contaminated with TCDD. The FDA and the Consumer Product
Safety Commission concluded that dioxins in paper and paper products including coffee filters, food packaging, tampons, surgical dressings, and diapers do not pose a significant health risk. Pentachlorophenol is ubiquitous
in the environment because of past uses in wood preservation.
The PHAHs are chemicals of concern in many of the Superfund hazardous waste sites, the best known being Love Canal (New York) and
Times Beach (Missouri). In 1999, the EPA proposed regulations to limit
dioxins in cement kiln dust and sludges from pulp and paper and sewage
treatment plants used as soil additives. The recent EPA TCDD assessment
concluded that the human cancer risk related to current background lev-
156
CHILD HEALTH
AND THE
ENVIRONMENT
els of dioxins is 0.1 to 1.0% (U.S. Environmental Protection Agency, 2000b).
The EPA did not recommend an RfD because, under its traditional approach, it would have been two to three orders of magnitude lower than
current background exposure levels. The WHO used noncarcinogenic animal toxicity data and human body burden data to set a tolerable daily
intake of dioxin-TEQ at 1–4 pg/kg (World Health Organization, 2000).
Conclusions
Proven Health Effects
• Prenatal exposure to relatively high amounts of PCBs (contaminated
by other PHAHs) can cause intrauterine growth retardation, eye discharges, increased pigmentation, developmental delays, and cognitive
deficits.
• Relatively high exposure to TCDD can cause chloracne, with children
possibly being more susceptible than adults.
• The IARC and the EPA concluded that TCDD can cause cancer in humans; the excess lifetime cancer risk of current population TCDD exposure levels may be as high as 1.0%.
Unresolved Issues and Knowledge Gaps
• Developmental effects—there is limited evidence that low-level preconceptual and gestational exposure to PCBs and related PHAHs may
cause fetal deaths, intrauterine growth retardation, and preterm delivery and that high-level prenatal exposure may cause developmental abnormalities (hypertelorism, natal teeth).
• Neurotoxicity
° There is limited epidemiologic evidence that preconceptual and prenatal maternal exposure to PCBs and related PHAHs from dietary
and/or other background environmental sources can cause abnormal
reflexes among newborn infants, reduced motor skills among infants,
and cognitive deficits in children
° There is inadequate evidence to assess the possible independent roles
of lactational PCB exposure, coplanar versus noncoplanar PCB congeners, and prenatal PHAH-induced thyroid suppression in causing
neurotoxicity
• Cancer
° Occupational PCB exposure is a probable cause of cancer in adults.
° There is limited evidence that cumulative exposure to PCBs and related PHAHs may cause adult cancers including breast cancer and
PCBs, Dioxins, and Related Compounds
157
non-Hodgkin’s lymphoma; the relative importance of PHAH exposure during childhood and adulthood is unknown.
° The potential role of PHAHs in childhood cancers is unknown.
• Other toxicity
° There is inadequate evidence to assess the role of perinatal PHAH
exposure in the development of the endocrine, reproductive, and immune systems; TCDD and related PHAHs should be considered nonspecific immunosuppressants at least until better data are available.
° There is limited evidence that perinatal PHAH exposure may affect
preadolescent growth (reduced stature).
Risk Management Issues
• Prevention
° Although PCB production ceased in 1979 and emissions of other
PHAHs have generally decreased, PHAHs persist throughout the
global environment and bioaccumulate in aquatic and terrestrial food
chains.
° Foods, particularly breast milk, dairy products, fish, meat, and poultry, continue to be the major sources of PHAH uptake for the general
population.
° The benefits of breast-feeding appear to outweigh the potential neurobehavioral effects of lactational PHAH exposure.
° It is necessary to reduce PHAH emissions and human exposures
through various strategies including
Replacement of chlorine and chlorine-containing products with safer
alternatives
Reduced incineration of chlorine-containing products and emission
controls to further reduce PHAH emissions from incinerators
Continued modification of industrial processes to reduce inadvertent production of PHAH contaminants
Reduced use of PHAH-contaminated sludge on croplands
• Biomonitoring
° Breast milk dioxin-TEQ levels have decreased by about 50% in several countries during recent decades, with the notable exception of
dioxin-like PCDEs; relatively high levels persist in populations dependent on contaminated fish and sea mammals.
References
Agency for Toxic Substances and Disease Registry. (1998). Toxicological profile for
chlorinated dibenzo- -dioxins (update). Atlanta: Agency for Toxic Substances
and Disease Registry.
158
CHILD HEALTH
AND THE
ENVIRONMENT
Agency for Toxic Substances and Disease Registry. (2000). Toxicological profile for
polychlorinated biphenyls (update). Atlanta: Agency for Toxic Substances and
Disease Registry.
Agency for Toxic Substances and Disease Registry. (2002). Minimal risk levels
(MRLs) for hazardous substances. Located at http://www.atsdr.cdc.gov/
mrls.html.
Alaluusua S, Lukinmaa PL, Torppa J, Tuomisto J, Vartiainen T. (1999). Developing teeth as biomarker of dioxin exposure. Lancet 353:206.
Anderson HA, Falk C, Hanrahan L, Olson J, Burse VW, Needham L, Paschal D,
Patterson D Jr, Hill RH Jr. (1998). Profiles of Great Lakes critical pollutants: a
sentinel analysis of human blood and urine. The Great Lakes Consortium. Environ Health Perspect 106:279–89.
Aronson KJ, Miller AB, Woolcott CG, Sterns EE, McCready DR, Lickley LA, Fish
EB, Hiraki GY, Holloway C, Ross T, and others. (2000). Breast adipose tissue
concentrations of polychlorinated biphenyls and other organochlorines and
breast cancer risk. Cancer Epidemiol Biomarkers Prev 9:55–63.
Birnbaum LS. (1995). Developmental effects of dioxins. Environ Health Perspect
103(Suppl 7):89–94.
Blanck HM, Marcus M, Rubin C, Tolbert PE, Hertzberg VS, Henderson AK, Zhang
RH. (2002). Growth in girls exposed in utero and postnatally to polybrominated biphenyls and polychlorinated biphenyls. Epidemiology 13:205–10.
Brouwer A, Ahlborg UG, van Leeuwen FX, Feeley MM. (1998). Report of the WHO
working group on the assessment of health risks for human infants from exposure to PCDDs, PCDFs and PCBs. Chemosphere 37:1627–43.
Centers for Disease Control and Prevention. (2001). National report on human exposure to environmental chemicals. Selection of chemicals for the report. Located at http://www.cdc.gov/nceh/dls/report/totalreport/substance.htm.
Chao WY, Hsu CC, Guo YL. (1997). Middle-ear disease in children exposed prenatally to polychlorinated biphenyls and polychlorinated dibenzofurans. Arch
Environ Health 52:257–62.
Chen YC, Guo YL, Hsu CC, Rogan WJ. (1992). Cognitive development of YuCheng (“oil disease”) children prenatally exposed to heat-degraded PCBs.
JAMA 268:3213–8.
Craan AG, Haines DA. (1998). Twenty-five years of surveillance for contaminants
in human breast milk. Arch Environ Contam Toxicol 35:702–10.
Crofton KM, Rice DC. (1999). Low-frequency hearing loss following perinatal exposure to 3,3,4,4,5-pentachlorobiphenyl (PCB 126) in rats. Neurotoxicol Teratol 21:299–301.
De Rosa CT, Brown D, Dhara R, Garrett W, Hansen H , Holler J, Jones D, JordanIzaguirre D, O’Connor R, Pohl H, and others. (1997). Dioxin and dioxin-like
compounds in soil, Part I: ATSDR interim policy guideline. Agency for Toxic
Substances and Disease Registry. Toxicol Ind Health 13:759–68.
Dewailly E, Ayotte P, Bruneau S, Gingras S, Belles-Isles M, Roy R. (2000). Susceptibility to infections and immune status in Inuit infants exposed to
organochlorines. Environ Health Perspect 108:205–11.
Dewailly E, Ayotte P, Bruneau S, Laliberte C, Muir DC, Norstrom RJ. (1993). Inuit
exposure to organochlorines through the aquatic food chain in Arctic Quebec.
Environ Health Perspect 101:618–20.
Gladen BC, Ragan NB, Rogan WJ. (2000). Pubertal growth and development
PCBs, Dioxins, and Related Compounds
159
and prenatal and lactational exposure to polychlorinated biphenyls and dichlorodiphenyl dichloroethene. J Pediatr 136:490–6.
Holford TR, Zheng T, Mayne ST, Zahm SH, Tessari JD, Boyle P. (2000). Joint effects of nine polychlorinated biphenyl (PCB) congeners on breast cancer risk.
Int J Epidemiol 29:975–82.
International Agency for Research on Cancer. (1987). Overall evaluations of carcinogenicity: an updating of IARC Monographs Volumes 1 to 42 (Supplement
7). Lyon, France: International Agency for Research on Cancer.
Jacobson JL, Jacobson SW. (1996). Intellectual impairment in children exposed to
polychlorinated biphenyls in utero. N Engl J Med 335:783–9.
Jacobson JL, Jacobson SW, Humphrey HE. (1990). Effects of exposure to PCBs and
related compounds on growth and activity in children. Neurotoxicol Teratol
12:319–26.
Jacobson JL, Janisse J, Banerjee M, Jester J, Jacobson SW, Ager JW. (2002). A benchmark dose analysis of prenatal exposure to polychlorinated biphenyls. Environ Health Perspect 110:393–8.
Karmaus W, Asakevich S, Indurkhya A, Witten J, Kruse H. (2002). Childhood
growth and exposure to dichlorodiphenyl dichloroethene and polychlorinated
biphenyls. J Pediatr 140:33–9.
Knox E. (2000). Childhood cancers, birthplaces, incinerators and landfill sites. Int
J Epidemiol 29:391–7.
Lackmann GM. (2002). Polychlorinated biphenyls and hexachlorobenzene in fullterm neonates. Reference values updated. Biol Neonate 81:82–5.
LaKind JS, Berlin CM, Naiman DQ. (2001). Infant exposure to chemicals in breast
milk in the United States: what we need to learn from a breast milk monitoring program. Environ Health Perspect 109:75–88.
Longnecker MP, Rogan WJ, Lucier G. (1997). The human health effects of DDT
(dichlorodiphenyltrichloroethane) and PCBs (polychlorinated biphenyls) and an
overview of organochlorines in public health. Annu Rev Public Health 18:211–44.
McNulty WP. (1985). Toxicity and fetotoxicity of TCDD, TCDF and PCB isomers
in rhesus macaques (Macaca mulatta). Environ Health Perspect 60:77–88.
Mocarelli P, Needham LL, Marocchi A, Patterson DG Jr, Brambilla P, Gerthoux PM,
Meazza L, Carreri V. (1991). Serum concentrations of 2,3,7,8-tetrachlorodibenzop-dioxin and test results from seleced residents of Seveso, Italy. J Toxicol Environ Health 32:357–66.
Mocarelli P, Gerthoux PM, Ferrari E, Patterson DG Jr, Kieszak SM, Brambilla P,
Vincoli N, Signorini S, Tramacere P, Carreri V, and others. (2000). Paternal concentrations of dioxin and sex ratio of offspring. Lancet 3555:1858–63.
Needham LL, Patterson DG Jr, Burse VW, Paschal DC, Turner WE, Hill RH Jr.
(1996). Reference range data for assessing exposure to selected environmental
toxicants. Toxicol Ind Health 12:507–13.
Noren K, Meironyte D. (2000). Certain organochlorine and organobromine contaminants in Swedish human milk in perspective of past 20–30 years. Chemosphere 40:1111–23.
Orban JE, Stanley JS, Schwemberger JG, Remmers JC. (1994). Dioxins and dibenzofurans in adipose tissue of the general U.S. population and selected subpopulations. Am J Public Health 84:439–45.
Patandin S, Dagnelie PC, Mulder PG, Op de Coul E, van der Veen JE, WeisglasKuperus N, Sauer PJ. (1999). Dietary exposure to polychlorinated biphenyls
160
CHILD HEALTH
AND THE
ENVIRONMENT
and dioxins from infancy until adulthood: a comparison between breastfeeding, toddler, and long-term exposure. Environ Health Perspect 107:45–51.
Patandin S, Koopman-Esseboom C, de Ridder MA, Weisglas-Kuperus N, Sauer
PJ. (1998). Effects of environmental exposure to polychlorinated biphenyls and
dioxins on birth size and growth in Dutch children. Pediatr Res 44:538–45.
Pesatori AC, Consonni D, Tironi A, Zocchetti C, Fini A, Bertazzi PA. (1993). Cancer in a young population in a dioxin-contaminated area. Int J Epidemiol 22:
1010–3.
Pocchiari F, Silano V, Zampieri A. (1979). Human health effects from accidental
release of tetrachlorodibenzo-p-dioxin (TCDD) at Seveso, Italy. Ann NY Acad
Sci 320:311–20.
Ribas-Fito N, Sala M, Kogevinas M, Sunyer J. (2001). Polychlorinated biphenyls
(PCBs) and neurological development in children: a systematic review. J Epidemiol Commun Health 55:537–46.
Rice DC. (1997). Effect of postnatal exposure to a PCB mixture in monkeys on multiple fixed interval-fixed ratio performance. Neurotoxicol Teratol 19:429–34.
Rogan WJ, Gladen BC, Hung KL, Koong SL, Shih LY, Taylor JS, Wu YC, Yang D,
Ragan NB, Hsu CC. (1988). Congenital poisoning by polychlorinated biphenyls
and their contaminants in Taiwan. Science 241:334–6.
Rothman N, Cantor KP, Blair A, Bush D, Brock JW, Helzlsouer K, Zahm SH, Needham LL, Pearson GR, Hoover RN, and others. (1997). A nested case-control
study of non-Hodgkin lymphoma and serum organochlorine residues. Lancet
350:240–4.
Ryan JJ, Gasiewicz TA, Brown JF Jr. (1990). Human body burden of polychlorinated dibenzofurans associated with toxicity based on the Yusho and Yucheng
incidents. Fundam Appl Toxicol 15:722–31.
Rylander L, Stromberg U, Dyremark E, Ostman C, Nilsson-Ehle P, Hagmar L.
(1998). Polychlorinated biphenyls in blood plasma among Swedish female fish
consumers in relation to low birth weight. Am J Epidemiol 147:493–502.
Safe S. (2001). Molecular biology of the Ah receptor and its role in carcinogenesis. Toxicol Lett 120:1–7.
Schecter A, Cramer P, Boggess K, Stanley J, Papke O, Olson J, Silver A, Schmitz
M. (2001). Intake of dioxins and related compounds from food in the U.S. population. J Toxicol Environ Health A 63:1–18.
Schecter A, Li L. (1997). Dioxins, dibenzofurans, dioxin-like PCBs, and DDE in
U.S. fast food, 1995. Chemosphere 34:1449–57.
Scheele J, Teufel M, Niessen KH. (1992). Chlorinated hydrocarbons in the bone
marrow of children: studies on their association with leukaemia. Eur J Pediatr 151:802–5.
Schiestl RH, Aubrecht J, Yap WY, Kandikonda S, Sidhom S. (1997). Polychlorinated biphenyls and 2,3,7,8-tetrachlorodibenzopdioxin induce intrachromosomal recombination in vitro and in vivo. Cancer Res 57:4378–83.
Travis CC, Hattemer-Frey HA. (1991). Human exposure to dioxin. Sci Total Environ 104:97–127.
U.S. Environmental Protection Agency. (2000a). Exposure and human health reassessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and related compounds, part III: integrated summary and risk characterization for 2,3,7,8tetrachloro-p-dioxin (TCDD) and related compounds. Washington, DC: U.S.
Environmental Protection Agency.
PCBs, Dioxins, and Related Compounds
161
U.S. Environmental Protection Agency. (2000b). Exposure and human health reassessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and related compounds. Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2001). Current drinking water standards.
Washington, D.C: U.S. Environmental Protection Agency.
U.S. Food and Drug Administration. (2000a). Action levels for poisonous or deleterious substances in human food and animal feed. Washington, DC: U.S. Food
and Drug Administration.
U.S. Food and Drug Administration. (2000b). Food and drugs chapter 1: Unavoidable contaminants in food for human consumption and food-packaging
material. Washington, DC: U.S. Government Printing Office.
Urabe H, Asahi M. (1985). Past and current dermatological status of Yusho patients. Environ Health Perspect 59:11–5.
van Leeuwen FX, Feeley M, Schrenk D, Larsen JC, Farland W, Younes M. (2000).
Dioxins: WHO’s tolerable daily intake (TDI) revisited. Chemosphere 40:1095–
101.
Vartiainen T, Jaakkola JJ, Saarikoski S, Tuomisto J. (1998). Birth weight and sex of
children and the correlation to the body burden of PCDDs/PCDFs and PCBs
of the mother. Environ Health Perspect 106:61–6.
Walkowiak J, Wiener JA, Fastabend A, Heinzow B, Kramer U, Schmidt E, Steingruber HJ, Wundram S, Winneke G. (2001). Environmental exposure to polychlorinated biphenyls and quality of the home environment: effects on psychodevelopment in early childhood. Lancet 358:1602–7.
Weisglas-Kuperus N, Patandin S, Berbers GA, Sas TC, Mulder PG, Sauer PJ,
Hooijkaas H. (2000). Immunologic effects of background exposure to polychlorinated biphenyls and dioxins in Dutch preschool children. Environ Health
Perspect 108:1203–7.
World Health Organization. (2000). Consultation on assessment of the health risk
of dioxins; re-evaluation of the tolerable daily intake (TDI): executive summary. Food Addit Contam 17:223–40.
World Health Organization. (2001). Joint FAO/WHO food standards programme.
Codex committee on food additives and contaminants. Thirty-third Session,
The Hague, The Netherlands, 12–16 March 2001. Position paper on dioxins and
dioxin-like PCBs. Located at http://www.who.int/fsf/Chemicalcontaminants/
fa01_29e.pdf.
Yamashita F, Hayashi M. (1985). Fetal PCB syndrome: clinical features, intrauterine growth retardation and possible alteration in calcium metabolism. Environ Health Perspect 59:41–5.
Yoshimura T, Kaneko S, Hayabuchi H. (2001). Sex ratio in offspring of those affected by dioxin and dioxin-like compounds: the Yusho, Seveso, and Yucheng
incidents. Occup Environ Med 58:540–1.
Yu ML, Hsin JW, Hsu CC, Chan WC, Guo YL. (1998). The immunologic evaluation of the Yucheng children. Chemosphere 37:1855–65.
7
Pesticides
Conventional pesticides comprise a diverse group of substances intended
to destroy, repel, or control organisms identified as pests. Some are broadspectrum biocides and others are relatively selective, targeting specific organisms such as insects, fungi, or plants. By 1999, the EPA had registered
almost 900 active ingredients in pesticides, including 350 approved for
use on food crops. The discovery during the 1960s of toxic effects in
wildlife linked to bioaccumulation of organochlorines in food chains contributed to public concern and decisions of the U.S. government to form
the Environmental Protection Agency (EPA) in 1970 and to ban the use
of DDT on food crops in 1972. Pesticides in the Diets of Infants and Children
was a landmark report documenting the vulnerability of children to pesticides and the potential for disrupting growth and development processes (National Academy of Sciences, 1993). This report noted that tests
conducted by manufacturers usually involved sexually mature animals
and did not assess neurobehavioral, immunologic, and endocrine effects
of prenatal and early-life pesticide exposures.
Milestones in the history of pesticide development include:
• Use of elemental sulfur and oil mixtures as insecticides by the Greeks
and Romans
162
Pesticides
163
• Fungicidal and insecticidal applications of inorganic sulfur and arsenic
compounds during the nineteenth century
• Development of synthetic organic pesticides during the 1930s and the
early post–World War II era; for example, DDT was registered for use
on 334 crops by 1961
• Introduction during the 1960s and 1970s of organophosphate (OP) insecticides, highly toxic but less persistent compounds that generally do
not bioaccumulate
• Use during the 1980s and 1990s of natural plant pyrethrins and synthetic pyrethroids, nonpersistent insecticides effective at low doses with
relatively low toxicity
• Introduction during the 1990s of genetically modified plants, pesticideresistant crops, and relatively nontoxic biopesticides (microbials, plant
pesticides, pheromones)
This chapter explores pesticides with respect to acute poisonings, developmental toxicity, developmental neurotoxicity, carcinogenicity, and exposure issues. Other topics include the new measures under the 1996 Food
Quality Protection Act (FQPA) that address the unique exposure patterns
and susceptibility of children.
Health Effects
Major classes and examples of active ingredients in conventional pesticides are shown in Table 7–1. Herbicides comprise about half of the 5 billion pounds of active ingredients used globally each year. Almost 80% of
these active ingredients are used in agriculture and forestry, the remainder being used by industry and homeowners. Pesticide formulations incorporate so-called inert substances, that is, ingredients not claimed to be
pesticidally active (e.g., surfactants, dyes, suspending agents, preservatives, and emulsifiers). The EPA has registered about 1600 inert ingredients including 56 substances for which it has significant concern about
carcinogenicity, reproductive and developmental toxicity, and neurotoxicity (U.S. Environmental Protection Agency, 1999a, 2001b).
Susceptibility
The developing fetus and child are at risk of adverse effects from pesticides because of their unique characteristics and their potential for higher
exposures than adults.
TABLE 7–1. Classes and Examples of Conventional Pesticides
Class
Antimicrobials—substances used
to destroy or suppress the growth
of harmful bacteria, viruses,
fungi, algae, protozoa
Subclasses
Examples
Fungicides
Vinclozolin, procymidone, iprodione, ziram, ferbam, thiram, maneb, mancozeb, zineb,
benomyl, captan, methylmercury, pentachlorophenol, hexachlorobenzene
Algicides
Nonselective biocides
Copper sulfate, lithium hypochlorite, pentachlorophenol, various herbicides
Methyl bromide, phosphine gas
Biopesticides—pesticides derived
from natural materials including
animals, plants, bacteria, and
certain minerals
Microbials
Plant pesticides
Pheromones
Bacillus thuringiensis (used as an insecticide)
Pyrethrins (insecticides)
11-Hexadecenal, many other compounds
Herbicides—agents used to
destroy unwanted plants
Selective
Chlorophenoxy herbicides—2,4-D, 2,4,5-T; other—atrazine,
cyanazine, alachlor, metolachlor, acetochlor
Paraquat, diquat, glyphosate
Nonselective
Insecticides—any substance used
to destroy insects
Rodenticides—pesticides specially
designed to kill rodents
Other
Chlorpyrifos, malathion
Aldicarb, aminocarb, carbaryl, carbofuran
DDT, hexachlorocyclohexanes, aldrin, dieldrin, endrin, endosulfan, heptachlor,
chlordane, toxaphene, chlordecone
Pyrethroids—allethrin, permethrin
Anticoagulants
Other
Hydroxycoumarins, indandiones
Vitamin D3, bromethalin, zinc phosphide, strychnine
OPs
N-methyl carbamates
Organochlorines
Pesticides
165
Immature Detoxification Systems
Liver metabolism of xenobiotics increases dramatically during early infancy with some enzymes not maturing until age 5 years or later. Several
OP insecticides, for instance, are activated by cytochrome P450 systems
to oxons, and the latter are inactivated by plasma and liver paraoxonases,
enzymes that do not reach adult levels until at least age 6 months.
Developmental Molecular Targets
Normal development of the nervous system requires a complex array
of precisely timed events including neuronal proliferation, migration,
differentiation, synapse formation, gliogenesis, myelination, and programmed cell death (apoptosis) that appear to depend in part on earlylife neurotransmitter systems. Transient bursts of anticholinesterase
(AChE) expression coincide with periods of axonal outgrowth in neonatal primate brains, and some selective AChE inhibitors suppress neurite
outgrowth in neuroblastoma cells in vitro. Anticholinesterase exists as intracellular monomers and dimers before synapse formation and as cell
surface tetramers afterward, paralleling a functional change from generating and stabilizing synapses to enabling neurotransmission. In humans,
nicotinic cholinergic receptors are maximally expressed in the brainstem
at mid-gestation and may mediate adverse effects of maternal smoking
on brain development (Eskenazi and Castorina, 1999).
Prenatal exposure of experimental animals to nicotine or chlorpyrifos
inhibits DNA synthesis, triggers apoptosis, and causes reduction of neuron populations in brain regions enriched in cholinergic innervation. Repeated exposure to OPs during gestation or the early postnatal period
causes persistent reduced brain AChE activity and deficits in learning, locomotion, and balance. In neonatally exposed rats, chloropyrifos also
causes reduced expression of the adenyl cyclase cascade signaling system
that serves global functions in the coordination of cell differentiation during development. Development of serotonergic and noradrenergic neurons depends on GABA. Prenatal exposure of rats to dieldrin reduces the
expression of developing GABA receptor subunits.
Exposure
Young children may have higher pesticide exposures than adults because
they have (1) a higher intake of air, water, and food per unit body weight
per day, (2) a larger skin surface area per unit body weight (dermal absorption is an important exposure route for many pesticides), (3) markedly
higher intake per unit body weight per day of certain foods (e.g., apple
juice), and (4) behaviors that favor pesticide uptake, such as hand–mouth
166
CHILD HEALTH
AND THE
ENVIRONMENT
behavior and frequent contact with potentially contaminated surfaces
(floors, carpets, lawns).
Poisonings
There appear to be no accurate estimates of the global burden of pesticide poisoning among children. The Toxic Exposure Surveillance System
(TESS) of the American Association of Poison Control Centers covers 96%
of the U.S. population. During 2000, TESS received reports of 54,544 childhood pesticide exposures attributed to insecticides (41%), rodenticides
(36%), insect repellents (16%), and herbicides (6%); about 25% of all
exposures were treated in a health care facility, mostly on a precautionary basis (Litovitz et al., 2001). The main causes of moderate or severe
poisonings were OPs (alone or in combination), pyrethins, piperonyl
butoxide/pyrethin mixtures, herbicides, and carbamate insecticides.
Because they are used in and around homes, bait and tracking rodenticides are one of the leading causes of poisoning among young children; fortunately, most cases are mild and do not require medical intervention. Severe neurotoxicity, including coma, seizures, and death, has
occurred after ingestion or excessive dermal exposure to N,N-diethyl-mtoluamide (DEET) (Briassoulis et al., 2001). Some type II pyrethroids have
acute oral toxicities comparable to those of OPs (salivation, hyperexcitability, choreoathetosis, seizures, sympathetic activation). In contrast to
adults, the major clinical signs in children with severe carbamate and OP
poisoning involve central nervous system (CNS) depression and hypotonia rather than the cholinergic symptoms seen in adults.
During the late 1950s, about 4000 persons in Turkey (mainly children)
developed porphyria cutanea tarda. This condition includes fragile skin,
blisters, sores, and small cysts on sun-exposed skin areas caused by increased liver porphyrin production and photosensitization; at least several
hundred infants died, apparently from exposure through breast milk (Gocmen et al., 1989). Investigation linked the illness to consumption of foods
prepared from wheat seed grain treated with the fungicide hexachlorobenzene (HCB). Many breast-fed children of exposed mothers died from
pembe yara (“pink sore”), a condition including weakness, convulsions,
and skin lesions from photosensitization. The most frequent clinical symptoms observed during a follow-up study of victims 20 years later were severe scarring from skin lesions, hyperpigmentation, arthritis, small hands,
hypertrichosis, short stature, weakness, paresthesias, and cogwheeling. Release of methyl isocyanate from an insecticide production plant in Bhopal,
India, during 1984 was the worst pesticide-related poisoning incident to
date, exposing at least 200,000 persons and causing over 3000 deaths.
Pesticides
167
Development
Among the relatively few epidemiologic studies of pesticides and fetal
development reported to date, most have had limited statistical power
and crude exposure assessment.
Fetal Death
A review of pre-1998 epidemiologic studies concluded that parental pesticide exposure may increase the risk of early and late fetal deaths; although fetal deaths could not be attributed to specific pesticides, there
were associations with broad pesticide categories such as organochlorines
and carbamates (Arbuckle and Sever, 1998). Fetal deaths before 12 weeks
were associated with both paternal and maternal preconceptual pesticide
exposure (especially to herbicides and thiocarbamates), while later fetal
deaths were associated with postconception exposure (Arbuckle et al.,
1999, 2001). There is also limited evidence of associations between early
fetal deaths and serum HCB or DDE levels (Jarrell et al., 1998; Korrick et
al., 2001) and between late fetal deaths and maternal residential proximity to agricultural use of restricted pesticides (Bell et al., 2001b).
Fetal Growth and Gestational Length
The few epidemiologic studies of pesticide exposure and fetal growth and
gestational length have produced mixed results. There were exposure–risk
relationships between third trimester maternal serum DDE levels and intrauterine growth retardation and preterm birth in a recent study that
took advantage of stored serum samples collected during 1959–1966,
when DDT exposures were much higher than today (Longnecker et al.,
2001). Two analytic studies of paternal occupational exposure to herbicides showed no relation to intrauterine growth retardation (IUGR) and
inconsistent evidence of a link to preterm delivery (Michalek et al., 1998;
Savitz et al., 1997). Maternal but not paternal occupational exposure to
PCP was associated with IUGR (Dimich-Ward et al., 1996; Karmaus and
Wolf, 1995). There was no relation between birth weight and community
exposure to malathion or parental occupation as farmers (Kristensen et
al., 1997; Thomas et al., 1992). In one of the few studies to examine postnatal growth, stature at intervals up to age 8 years was inversely associated with blood DDE in girls but not boys (Karmaus et al., 2002).
Birth Defects
Two reviews of epidemiologic studies of pesticide exposure and birth defects concluded that the evidence is suggestive but inconclusive (Garcia,
1998; Nurminen, 1995). In these studies, the main indicator of pesticide
168
CHILD HEALTH
AND THE
ENVIRONMENT
exposure was parental occupational exposure in agriculture and the most
frequent associations were with limb reduction defects and orofacial clefts.
Epidemiologic studies published since these reviews have shown:
• Orofacial clefts—associated with paternal occupational pesticide exposure (Shaw et al., 1999).
• Limb reduction defects—associated with maternal occupational and
residential pesticide exposure (Engel et al., 2000; Shaw et al., 1999).
• Genital defects—cryptorchidism was associated with adipose tissue heptachlor levels (Hosie et al., 2000) and maternal but not paternal occupational pesticide exposure (Weidner et al., 1998); neither cryptorchidism
nor hypospadias was associated with maternal serum DDT/DDE levels
(Longnecker et al., 2002).
• Neural tube defects—associated with residential pesticide use and maternal residence within 0.25 mile of agricultural crops (Shaw et al., 1999)
and paternal occupational exposure to phenoxy herbicides (National
Academy of Sciences, 2001).
• Cardiac defects—associated with maternal first trimester pesticide exposure, with some evidence of higher risks with herbicide and rodenticide exposure (Loffredo et al., 2001).
• Late fetal deaths from birth defects—associated with maternal residential proximity to agricultural pesticide use during gestational weeks
3–8 (especially pyrethroid and organochlorine pesticides) (Bell et al.,
2001a, 2001c).
Reproductive Toxicity
Men occupationally exposed during production of the organochlorine insecticide chlordecone and the nematocide 1,2-dibromo-3-chloropropane
(DBCP) developed greatly reduced sperm counts and infertility; DBCP
was also linked to Y chromosome nondisjunction in the sperm of exposed
men. Animal studies showed that DBCP caused testicular atrophy at the
lowest dose tested and several types of cancer in rodents. In addition,
DBCP covalent binds to DNA and causes single-strand DNA breaks in
round spermatids at low doses. Chlordecone binds to the estrogen receptor and has estrogenic activity in vivo, but the mechanism by which
it causes oligospermia remains uncertain. Occupational exposure during
application of pesticides has been associated with reduced semen quality
in men and reduced fertility in women. See also Chapter 8 (Hormonally
Active Agents) for a discussion of hormonally active pesticides.
Neurotoxicity
Compared to studies of lead, mercury, and PCBs, few epidemiologic studies have assessed the developmental neurotoxicity of pesticides. Also,
Pesticides
169
most pesticides and other commercial chemicals have not been tested in
animals for developmental neurotoxicity. Hypothesized effects of perinatal pesticide exposure in humans include social and emotional development deficits, autism, cerebral palsy, and mental retardation (Goldman
and Koduru, 2000). Most insecticides are designed to target a variety of
neuroreceptors and ion channels causing effects such as hyperexcitation
and paralysis in target species, but they are also neurotoxic in human and
animal bystanders. In one of the few epidemiologic studies of neurobehavioral development and perinatal pesticide exposure, there was no association between cord blood DDE levels and scores on the Fagan Test of
Infant Intelligence at ages 6 or 12 months (Darvill et al., 2000). Occupational exposure of older children (and adults) to OPs, carbamates, and
some other pesticides may cause prolonged suppression of AChE, visual
disturbances, and peripheral neuropathy.
Cancer
Children may be exposed to pesticides used in and around homes and
schools, residues in food and water, airborne drift from agricultural use,
and carry-home residues by occupationally exposed parents. Recent reviews of childhood cancers and pesticide exposure concluded that (1) reported associations were modest but stronger when exposure was assessed
in more detail, (2) most childhood cancers associated with pesticide exposure are the same types linked to pesticide exposure in adults, but the
associations among children tend to be stronger, and (3) methodologic limitations preclude strong conclusions (Daniels et al., 1997; Zahm and Ward,
1998). Pesticide exposures linked to specific childhood cancers included
• Brain cancer—in-home, pet, and garden use of pesticides and parental
occupational pesticide exposure (including exposure–risk relationships)
• Leukemia—parental occupational pesticide exposure and in-home and
garden use of pesticides during pregnancy or childhood (exposure–risk
relationships in two studies that assessed children’s direct exposure to
pesticides)
• Non-Hodgkin’s lymphoma, Wilms’ tumor, Ewing’s sarcoma—postnatal
exposure indices
Recent studies have increased the evidence of associations between residential or parental pesticide exposure indices and childhood brain cancer, acute lymphatic leukemia, non-Hodgkin’s lymphoma, and kidney
cancer (mainly Wilms’ tumor) (see, e.g., Cordier et al., 2001; Daniels et al.,
2001; Feychting et al., 2001; Krajinovic et al., 2002). Insecticide exposure
during year 2 of infancy appeared to confer a higher risk of childhood
170
CHILD HEALTH
AND THE
ENVIRONMENT
leukemia than later exposures (Ma et al., 2002). Children with certain polymorphisms of phase I and phase II enzymes involved in the metabolism
of xenobiotics had substantially increased risks of acute lymphoblastic
leukemia, and there was evidence of positive interactions between
CYP1A1 polymorphisms and pesticide exposure (Infante-Rivard et al.,
1999; Krajinovic et al., 2002; Sinnett et al., 2000).
Because of the relative rarity of childhood cancer, most etiologic studies of childhood cancer have used a case-control design and may be subject to recall bias. Some have argued that many natural biocides in foods
are rodent carcinogens at high doses and that carcinogenic risks of synthetic pesticides at levels found in foods are likely minimal. The limited
epidemiologic evidence of a role for pesticides in childhood cancer, however, is consistent with the following evidence:
• Epidemiologic studies of adult cancers
° Although one reviewer concluded that evidence of associations between herbicide exposure and soft tissue sarcomas, non-Hodgkin’s
lymphoma, leukemia, and ovarian cancer was limited (Dich et al.,
1997), the National Academy of Sciences concluded that there was
sufficient evidence of an association between herbicide and/or
TCDD exposure and soft tissue sarcomas, non-Hodgkin’s lymphoma, and Hodgkin’s disease (National Academy of Sciences,
2001).
There
were interactions between serum anti-Epstein-Barr virus anti°
body and , -DDE, HCB, or chlordane levels as risk factors for hairy
cell leukemia (Nordstrom et al., 2000).
• Genotoxicity—cytogenetic studies of fumigant and herbicide applicators have shown sister chromatid breaks and exchanges in peripheral
lymphocytes at bands that contain oncogenes and genes involved
in tumor suppression and apoptosis including bands 14q32 and
18q21, the most common sites of chromosomal rearrangements in nonHodgkin’s lymphoma.
• Carcinogenicity risk assessments—the EPA concluded that pesticide active ingredients include 5 known, 71 probable, and 82 possible human
carcinogens, many of which are still registered for use on foods.
Immunotoxicity
In one of the few epidemiologic studies of immunologic abnormalities
and pesticide exposure among children, combined blood DDE plus PCB
or DDE plus HCB levels were associated with otitis media; blood DDE
Pesticides
171
was also associated with a history of asthma and elevated IgE levels (Karmaus et al., 2001).
Exposures
Potential indicators of pesticide exposures among children and reproductive-age adults include environmental levels (air, water, soil, house dust,
food) and internal dose. Epidemiologic studies and environmental surveys
revealed that the pesticides most frequently used in homes or yards include chlorpyrifos, diazinon, DEET, dichlorvos, malathion, piperonyl butoxide, pyrethrins, 2-methyl-4-chlorophenoxyacetic acid (MCPA), 2,4-D,
cygon, propoxur, chlordane, carbaryl, and heptachlor. In agricultural areas,
children may be exposed to pesticides through various media including
substantially higher house dust concentrations of OPs compared to nonagricultural families (Lu et al., 2000). Among California children living in
agricultural areas with intense pesticide use, the highest-ranked pesticides
(weighted by amount used and carcinogenic potential in experimental animals) were propargite, methyl bromide, trifluralin, simazine, molinate,
and metam sodium (Gunier et al., 2001).
Environmental pesticide monitoring and human activity survey data
enable exposure estimates, but actual uptake can be measured only
through biomonitoring of human tissues and fluids (Fenske et al., 2000).
Breast milk, adipose tissue, and blood samples contain measurable
amounts of many fat-soluble, persistent contaminants including organochlorine pesticides, while urine analyses may detect evidence of recent exposure to nonpersistent pesticides and some organochlorines. Some biochemical effects of pesticides reflect the internal dose (e.g., cholinesterase
depression reflects recent exposure to AChE insecticides), while others are
nonspecific (e.g., chromosome aberrations).
Blood and Adipose Tissue
Dichlorodiphenyldichloroethylene, the major long-lived metabolite of
DDT, was detected in 99% of blood serum and adipose tissue samples from
U.S. national surveys in the late 1970s (median maternal serum DDE level
was 25 g/L during the early 1960s), but average concentrations have decreased severalfold since then. Over 90% of adipose tissue samples contained multiple organochlorine pesticide metabolites indicative of past and
continuing exposure to these persistent, bioaccumulative compounds.
Plasma butyrylcholinesterase (BChE) and red blood cell AChE levels
are the most sensitive indicators of OP and carbamate insecticide uptake;
172
CHILD HEALTH
AND THE
ENVIRONMENT
for example, red blood cell AChE is 12–14 times more sensitive to chlorpyrifos than brain or retinal AChE (Chen et al., 1999). The plasma BChE
level is normally depressed during the first two trimesters of pregnancy
among younger women but is still strongly correlated with self-reported
pesticide exposure status. Hemoglobin adducts of pesticides or their
metabolites are detectable after exposure to several urea and carbamate
insecticides. The frequency of chromosome damage, as indicated by
micronuclei in peripheral lymphocytes, has been associated with occupational exposure to pesticides, especially among persons with certain
phase II enzyme polymorphisms.
Urine
About 75% of registered OPs are metabolized in vivo to dimethyl and
other dialkyl phosphate metabolites excreted in urine. Population-based
surveys have shown high prevalence rates among children of urinary
metabolites of several pesticides including chlorpyrifos (80%–90%), PCP
(100%), -dichlorobenzene (a widely used toilet deodorant and moth
repellent) (96%), lindane (54%), and 2,4-D (20%). During recent decades,
the prevalence of the main urinary chlorpyrifos metabolite increased
substantially, while that of PCP decreased markedly. Organophosphate metabolites were detected in 95%–100% of meconium samples
from neonates in New York City, indicating fetal exposure (Whyatt and
Barr, 2001). Concentrations of chlorpyrifos and other OPs in house dust
and of their metabolites in children’s urine samples were associated
with parental occupational exposure and residential proximity to
treated farmland (Fenske et al., 2000, 2002). Based on urinary metabolite levels, children living near orchards are more likely than unexposed
children to exceed EPA acute and chronic RfDs for azinphosmethyl and
phosmet.
Among the U.S. population aged 6–59 years, the 90th percentile urinary dimethyl phosphate level was 10.1 g/g (Centers for Disease Control and Prevention, 2001). As in children, the most frequently detected
urinary pesticide metabolite among adults during NHANES III was
2,5-dichlorophenol (98%), indicative of ubiquitous exposure to dichlorobenzene (Table 7–2). Compared to NHANES II (1976–1980), the
prevalence of detectable chlorpyrifos (metabolite) in urine increased fivefold (from 6% to 31%) during NHANES III (1988–1994), while that of PCP
decreased by about half (from 72% to 39%). Biomonitoring of urinary
metabolites and red blood cell cholinesterase levels in agricultural workers has been used to assess pesticide exposure.
Pesticides
173
TABLE 7–2. Prevalence of Pesticide Urinary Metabolites, NHANES III (1988–1994)
Metabolite
Parent Pesticide
Prevalence
(%)
95th Percentile
Concentration
(g/g creatinine)
2,5-Dichlorophenol
-Dichlorobenzene
98
670
1-Naphthol
Naphthalene, carbaryl
86
36
3,5,6-Trichloro-2-pyridinol
Chlorpyrifos
82
8
2-Naphthol
Naphthalene
81
18
Pentachlorophenol
Pentachlorophenol
64
5
2,4-Dichlorophenol
2,4-D, dichloroprop, other
64
45
Source: Hill et al. (1995).
Breast Milk
Organochlorine levels in breast milk have generally declined dramatically
since the 1960s; for instance, DDE levels decreased by over 90% in Canada
between 1967 and the mid-1990s (Craan and Haines, 1998; Noren and
Meironyte, 2000). Breast milk organochlorine levels vary by geographic
region and maternal characteristics; DDE levels in breast milk decrease
with lifetime months of lactation because of cumulative depletion of fat
stores. Hexachlorobenzene, a fungicide, comprises 10%–60% of the total
dioxin TEQ in human milk samples in most countries. Lactational intake
of HCB, chlordane, dieldrin, and heptachlor epoxide may exceed WHO
acceptable daily intakes.
Risk Management
Over 4.6 billion pounds of pesticide active ingredient chemicals were sold
in the United States in 1997, including about 1 billion pounds of conventional pesticides (Table 7–3). Leading pesticides sold for domestic use
included -dichlorobenzene (used as a moth repellent and deodorizer),
2,4-D and glyphosate (herbicides), DEET (insect repellant), and chlorpyrifos (an insecticide). Total annual sales of pesticides for home and garden use in the United States decreased by 14% during 1979–1995, but sales
of herbicides for home and garden use increased by 42%. By 1990, production of -dichlorobenzene in the United States reached 70,000 tons, indicative of the widespread cosmetic uses of this known animal carcinogen.
Over 90% of current herbicide use occurs on just four crops—corn,
soybeans, cotton, and wheat. By amount of active ingredient, atrazine and
CHILD HEALTH
174
AND THE
ENVIRONMENT
TABLE 7–3. Pesticide Sales, United States, 1997
Pesticide Category
Millions of Lb
%
Herbicides
Insecticides
Fungicides
Other
568
128
81
198
12
3
2
4
Subtotal (conventional pesticides)
975
21
Sulfur, miscellaneous
Wood preservatives
Specialty biocides
Chlorine/hypochlorites
256
665
272
2459
6
15
6
54
Total
4627
100
Source: U.S. Environmental Protection Agency (1999a).
2,4-D, respectively, are the most widely used herbicides in the agricultural
and nonagricultural sectors. By 1990, relatively nonpersistent OPs and
carbamates had replaced most food uses of organochlorine pesticides in
the United States but not in many other countries; OPs now comprise
about half of the insecticides used in United States. The intensity of insecticide use on corn, soybeans, cotton, and wheat decreased after 1970
with the introduction of synthetic pyrethroids applied in much smaller
quantities per acre. Fungicide use remained high because of increased
acreage of fruits and vegetables.
Foods
Sources
Foods frequently consumed by children include apples, other fruits, and
vegetables that often contain pesticide residues in the parts per billion range.
Even the few countries that have conducted regular food consumption and
food pesticide residue surveys to assess population exposure to pesticides
from foods have not focused on exposures of infants and children. Canada,
for instance, has not conducted a national nutrition survey including children and pregnant women since 1970–1972. Food consumption data by single year of age among children less than 5 years old are needed to estimate
their pesticide exposures accurately (National Academy of Sciences, 1993).
The most intense pesticide applications include potatoes (43 lb of active ingredient per acre), apples and other vegetables (23 lb/acre), citrus
fruit (10 lb/acre), corn (2.9 lb/acre), soybeans (1.2 lb/acre), and wheat
Pesticides
175
(0.4 lb/acre) (National Academy of Sciences, 2000). The 10 most frequently
detected pesticide or pesticide metabolite residues in U.S. baby foods and
some of their characteristics are shown in Table 7–4; three of the five leading pesticides were moderate to strong neurotoxins (endosulfan, chlorpyrifos, carbaryl). Apples were the most important food source of ingested
chlorpyrifos, followed by tomatoes and grapes.
Despite heavy use of herbicides early in the growing season, food
residues of insecticides and fungicides are generally higher because these
are applied directly to food closer to or even after its harvest. Even when
average daily pesticide ingestion levels can be estimated, these are insufficient for assessing the potential toxicity of peak exposures. Incidents of
food-related OP poisoning during the 1980s and early 1990s were linked
TABLE 7–4. Ten Most Frequent Pesticide Residues Found in Selected Baby Foods
in 1999
Pesticide or
Metabolite
% of
Foodsa
Endosulfan
26
Organochlorine insecticide; potent neurotoxin, fetal
resorption and skeletal defects at high maternal
doses in rats, mutagenic, unknown carcinogenicity
Iprodione
21
Dicarboximide fungicide; toxicity studies mostly
negative
Chlorpyrifos
14
OP insecticide; moderately strong neurotoxin on acute
exposure
Carbaryl
13
Carbamate insecticide; moderate to potent neurotoxin
on acute exposure, possible teratogen at high doses,
weak mutagen but can react with nitrite to
produce N-nitrosocarbaryl, a potent mutagen
Permethrin
13
Pyrethroid insecticide; low mammalian toxicity
Chlorpyrifosmethyl
12
OP insecticide; equivocal evidence of delayed neurotoxicity, inadequate developmental toxicity databasec
Malathion
12
OP insecticide; neurotoxic, mutagenic
Thiabendazole
12
Benzimidazole fungicide; toxicity seen only at high
doses
Dimethoate
6
OP insecticide; potent neurotoxin, fetal deaths,
reduced fetal growth rate, teratogenic, mutagenic,
possible carcinogen
Ethylene thiourea
4
Metabolite of ethylene bisthiocarbamate fungicides,
possible human carcinogend
a U.S.
Food and Drug Administration (2000).
b EXTOXNET
cU.S.
Toxicityb
(Extension Toxicology Network), (2000).
Environmental Protection Agency (2000).
dIARC
(1987).
176
CHILD HEALTH
AND THE
ENVIRONMENT
to aldicarb, a carbamate insecticide used on fruits, nuts, potatoes, and
other vegetables. Aldicarb is not removable by peeling or washing because it is a systemic agent, that is, it is taken up by roots into the plant
itself. Although aldicarb levels in blended samples of banana were within
the legal limit, individual bananas had up to ten times this limit. The manufacturer agreed to stop selling aldicarb for use on bananas since they are
commonly eaten by children at levels up to fivefold those of adults (on a
body weight basis).
The FDA tests about 40 food samples per day for a limited number
of pesticides, a rate that precludes careful monitoring of the residues of
the many registered pesticides and the many suppliers of food products.
The FDA program has detected pesticide residues in 56% of fruit, 41% of
grain, 32% of fish/shellfish, and 31% of vegetable products. Levels above
FDA maximum residue levels (MRLs) occurred on 5% or more of domestic
strawberries, spinach, red beets, head lettuce, and other leaf and stem vegetables. The most frequently detected pesticides or metabolites in animal
fat and hen eggs were PCP (35%) and DDE (21%). Whole milk was the
main source of DDE for persons in the upper decile of estimated daily intake (MacIntosh et al., 1996).
Intervention
Pesticide active ingredients for use on foods must each receive a registration and a tolerance from the EPA. Tolerances, the most important
means by which the EPA limits pesticide residues in foods, are the legal
limits of pesticides allowed in raw or processed foods and are the highest levels likely to occur with normal agricultural pesticide uses. Until the
1993 report Pesticides in the Diets of Infants and Children was published, tolerances had generally been based on good agricultural practice, not risk
to human health; this report strongly recommended that tolerances be set
to safeguard infant and child health (National Academy of Sciences, 1993).
The FQPA requires an additional tenfold uncertainty factor in risk assessments for pesticide residue tolerances if there is evidence of special
sensitivities of infants and children or if data on toxicity and exposures
are incomplete. The FQPA and the 1996 Safe Drinking Water Act both require assessment of chemicals for hormonal activity. In setting tolerances,
the EPA now considers (1) RfDs based on animal toxicity tests including
developmental toxicity, developmental neurotoxicity, and two-generation
reproductive tests and (2) data on pesticide residues in foods and
age-specific food consumption patterns of children and other population
subgroups.
Based on evidence that young children may ingest methyl parathion,
one of the most toxic OPs, at levels up to eight to nine times the RfD, the
Pesticides
177
EPA in 1999 accepted voluntary cancellation of its use on fruit and vegetable crops representing 90% of the dietary risk to children. The EPA also
accepted voluntary measures in 1999 to reduce use of the OP azinphosmethyl because of an unacceptable dietary risk to young children and
agricultural workers. Ten years after Sweden initiated pesticide riskreduction programs during the mid-1980s, the weight of pesticide active
ingredients sold annually decreased by about two-thirds. Pesticide
residues exceeded Swedish MRLs in 3%–5% of imported fruits and vegetables compared to 0.5%–1% of domestically grown products. Food
preparation such as washing, peeling, and cooking appears to reduce the
prevalence of detectable pesticide residues in foods by about half, results
varying by food and pesticide; for example, endosulfan residues were almost completely removed from apples and pears by peeling but persisted
on spinach after washing.
Indoor and Home Environment
Methods to assess pesticide exposures of young children include measurements of pesticide concentrations in indoor air, carpet dust, outdoor
soil, and handwipes. Children’s exposures can be estimated using timelocation-activity diaries, videotaped activity studies, probability distributions of measured surface residues and exposure factors, and pharmacokinetic rate constants. Personal air sampling has shown that indoor
airborne pesticide exposure may exceed dietary doses for pesticides used
mainly in the home. Exposure to airborne insecticides tends to be higher
in warmer climates because of increased indoor use (e.g., for termite control) and higher evaporation of semivolatile compounds.
Urban Homes
Six million urban children living in poverty in the United States are at risk
of exposure to pesticides used extensively in schools, homes, and day-care
centers (Landrigan et al., 1999). Pesticide products are used indoors in over
90% of U.S. households, the main types being insecticide bombs, broadcast applications, crack and crevice treatments, no-pest strips, pet shampoos, and flea collars. Pesticide use was reported in over 70% of households with pregnant women or infants less than 6 months of age. The most
frequently detected pesticides in dust and surface wipe samples from U.S.
homes were chlorpyrifos, atrazine, malathion, chlordane, DDT/DDE,
methoxychlor, propoxur, carbaryl, permethrin, o-phenylphenol, PCP, and
2,4-D. Tracking of pesticide-contaminated soil and dust into homes by pets
and people is a major source of pesticide residues in house dust.
178
CHILD HEALTH
AND THE
ENVIRONMENT
Young children may be exposed to pesticides through inhalation, dermal absorption, or ingestion of pesticides in rugs, furniture, stuffed toys,
other absorbent surfaces, dust, or soil. Exposure is enhanced by behaviors common among young children including hand–mouth and object–mouth behaviors and playing barefoot indoors and outdoors. After
application to residential lawns, 2,4-D can be detected in indoor air and
on surfaces throughout homes (Nishioka et al., 2001; U.S. Environmental
Protection Agency, 1999b). The main sources of indoor 2,4-D deposition
were track-in by pets and people and settling of resuspended floor dust
on tables and sills; the estimated exposure of young children to 2,4-D was
1–40 g/kg/day (the EPA chronic oral RfD for 2,4-D is 10 g/kg/day).
On occasion, insecticides are applied indoors in entire communities, particularly in endemic malaria regions, where large amounts of DDT continue to be used; in 1993, over 1 million kg of DDT was used to spray
house walls in the Western Hemisphere alone. In 2000 the EPA banned
some indoor uses of chlorpyrifos including broadcast spraying and direct
uses on pets; crack and crevice treatment is still allowed.
Agricultural Area Homes
In agricultural areas, children may be more exposed to pesticides because
of higher environmental levels in their indoor and outdoor environments
and maternal breast milk. In rural California, ten pesticides were detected
in house dust and hand-wipe samples from young children; the most exposed toddlers could ingest diazinon doses above the EPA chronic RfD.
The highest house dust OP levels in Washington State occurred in homes
with at least one farm worker and in those close to apple or pear orchards.
Schools
There have been few studies of pesticide contamination in school environments, but school kitchens, cafeterias, classrooms, offices, athletic fields,
and playgrounds are often treated with a variety of pesticides. Incidents
involving illness of students and staff in some U.S. schools have been attributed to unwitting exposure to excessive levels of pesticides including
resmethrin, chlorpyrifos, diazinon, and chlordane. In its first statewide survey of pesticide use in public schools, New York State found that 87% of
schools used pesticides and usually took few precautions. The Attorney
General of New York recommended that schools should adopt and communicate written pest management policies and practices and select the
least toxic pesticides when their use is deemed necessary. Other recommendations included notification of school staff, students, and parents, use
of warning signs around treated areas, application only by certified personnel, maintenance of detailed records on pesticide use, and avoidance
Pesticides
179
of pesticides containing known or probable carcinogens (Attorney General of New York, 1996). The EPA noted a health concern for use of chlorpyrifos in crack and crevice treatment in schools, day-care centers, or other
rooms that children may occupy for extended periods of time.
Other Sources
Automatic insecticide dispensers are registered by the EPA for use in restaurants, schools, supermarkets, hospitals, day-care centers, and other facilities to control indoor flying insects in food service or work areas (Centers for Disease Control and Prevention, 2000). These automatically
dispense a fine mist of pyrethrin or resmethrin (a pyrethroid) along with
other active and inert ingredients at frequent intervals (e.g., every 15 minutes). Data for the period 1993–1996 from the TESS, based on poisoning
reports from 85% of U.S. poison control centers, showed 54 cases of illness
(age 3 to 73 years) associated with automatic insecticide dispensers. The
CDC recommended that automatic insecticide dispensers be installed according to manufacturers’ labeling instructions, that warning stickers be
placed on the dispensers, that installation near air ducts be avoided, and
that timers be set to dispense insecticide during nonbusiness hours. There
appears to be no ongoing monitoring of pesticide exposure in commercial,
institutional, day-care, and other indoor areas accessible to children.
Water
Drinking water is a minor source of pesticide exposure for children or
adults in the general population but may be important for subgroups such
as farm families. At least 130 pesticides and metabolites have been detected
in groundwater, particularly aldicarb, atrazine, DDE, dieldrin, and soil fumigants including ethylene dibromide, dichloropropane, and DBCP. The
median total herbicide concentration in Iowa municipal well water increased about tenfold with inclusion of their breakdown products, some
of which have toxicity similar to that of their parent compounds. Drinking water guidelines and standards for individual pesticides vary by factors of up to 10 between the WHO and the EPA (Table 7–5). In contrast,
the European Union Drinking Water Directive sets maximum limits of
0.1 g/L for any individual pesticide, 0.03 g/L for certain specific pesticides, and 0.5 g/L for total pesticides (European Union, 1998).
General Issues
Pesticide Toxicity Testing
Child health protection requires societal targets for reduction of pesticide
use, monitoring of pesticide exposures, and sophisticated premarket de-
CHILD HEALTH
180
AND THE
ENVIRONMENT
TABLE 7–5. Drinking Water Guidelines and Standards
for Selected Pesticides
Pesticide
WHOa
Alachlor
20 g/L
2 g/L
Atrazine
2 g/L
3 g/L
Aldrin/dieldrin
EPAb
30 ng/L
—
DDT
2 g/L
—
2,4-D
30 g/L
70 g/L
9 g/L
1 g/L
Pentachlorophenol
a World
b U.S.
Health Organization (1998).
Environmental Protection Agency (2001a).
velopmental toxicity testing (Landrigan et al., 1999). Governments must
balance public and scientific concerns about pesticide safety against the
benefits of pesticides, especially increased, affordable, and high-quality
food supplies. In developed countries, the main control of pesticides is
the registration process; this generally includes evaluation of standardized information provided by the manufacturer on efficacy, toxicity, environmental fate and dispersion, and potential ecologic effects. The herbicide 2,4-D was introduced during the 1940s and has been one of the
most widely used pesticides for several decades, including applications
in areas where children may be exposed. The IARC has designated
chlorophenoxy herbicides as possible human carcinogens, but both the
IARC and the EPA concluded that animal carcinogenicity tests of 2,4-D
were inadequate for evaluation.
The report Pesticides in the Diets of Infants and Children recommended
that pesticide evaluations include developmental neurotoxicity data; at
present, however, only adult neurotoxicity test data are available for most
of the 900 pesticide active ingredients in use. The EPA developmental neurotoxicity testing guideline has been criticized for failing to require studies that expose developing animals during all vulnerable periods, assess
delayed effects, and use standard neurobehavioral and neuropathology
test methods (Claudio et al., 2000). The EPA recently requested developmental neurotoxicity data from manufacturers on about 140 pesticides
considered to be neurotoxic.
Inert ingredients are considered trade secrets, and the EPA does not
require disclosure except for highly toxic compounds. Unless an inert ingredient is determined to be highly toxic, it does not have to be identi-
Pesticides
181
fied by name or percentage on the label, but the total percentage of such
ingredients must be declared. In what appears to be a precedent, the New
York City mosquito control program for prevention of West Nile virus infections included both active and inert ingredients in its environmental
impact assessment (New York City Department of Health, 2001).
Child-Oriented Interventions
The FQPA required the EPA to evaluate pesticide safety in light of aggregate exposures from nondietary and dietary routes. The insecticide chlorpyrifos illustrates some of the key pesticide risk management issues relevant to child health protection, including cumulative risks from multiple
agents sharing a common mechanism and aggregate exposure from multiple environmental media. After the banning of heptachlor and chlordane,
chlorpyrifos rapidly became one of the most widely used OPs in the United
States. By 2000, about 800 chlorpyrifos-containing products were registered
in the United States for use on food crops, lawns, playgrounds, parks, and
pets (flea collars) and in homes, schools, day-care centers, and many other
settings. Broadcast application of chlorpyrifos indoors appears to cause accumulation in polyurethane foam in pillows, toys, bedding, and mattresses, potentially causing uptake of over 200 g/kg bw/day among
young children (40% from dermal and 60% from oral contact), far higher
than the EPA RfD for chronic exposure of 3 g/kg bw/day from all sources
(Gurunathan et al., 1998).
In a reevaluation of chlorpyrifos, concern was raised by new evidence
including the following:
• Neonatal rats were more sensitive than adults to inhibition of cholinesterase in frontal cortex, plasma, and red blood cells after acute chlorpyrifos exposure.
• Lack of a NOAEL for developmental neurotoxicity in rats based on
structural changes in the developing brains of offspring.
The EPA decided to implement a phased partial ban of chlorpyrifos that
will eliminate its uses in and around homes and nonresidential settings
(mainly for termite control) and on tomatoes and restrict its use on apples. This action should reduce total chlorpyrifos use by up to 50% when
fully implemented in 2004.
Persistent Organochlorine Pesticides
Extensive use of DDT, mainly to control malaria in tropical countries, remains an issue more than 30 years after it was banned in Canada and
182
CHILD HEALTH
AND THE
ENVIRONMENT
other countries. Transmission of malaria is enhanced by changes in land
use that provide suitable habitats for malaria-transmitting mosquitoes
close to human populations, such as, road building, mining, logging, agricultural, and irrigation projects in the Amazon Basin and Southeast Asia.
The WHO malaria vector control efforts now favor pyrethroid-treated protective nets for beds, nonpersistent insecticides with minimal impact on
nontarget organisms, and limitation of house spraying to specific highrisk and epidemic-prone areas. Although DDT use has decreased sharply
in countries such as Mexico (from 25,000 tons in 1970 to about 300 tons
in 1998), illegal use of DDT and other persistent organochlorine pesticides
continues in parts of Asia. The United Nations Economic Commission for
Europe (UN-ECE) has a protocol that legally binds Canada, the United
States, Europe, Russia, and several other countries to ban production and
use of eight organochlorine pesticides (aldrin, chlordane, dieldrin, endrin,
hexabromobiphenyl, chlordecone, mirex, and toxaphene). The protocol
will also limit the production and/or use of DDT, heptachlor, lindane,
and HCB.
Exposure to Pesticides Sharing a Common Mechanism
The National Academy of Sciences evaluated the risk of exposure to multiple pesticides with a common toxic mechanism by assessing five OPs
(National Academy of Sciences, 1993). Although the Academy had to rely
on scanty data, it concluded that some children have sufficient aggregate
OP exposures to produce symptoms of acute poisoning.
Pesticide Benefits
A National Academy of Sciences expert committee recently reviewed the
future role of pesticides in agriculture (National Academy of Sciences,
2000). Net benefits of pesticides on crop yields depend on crop type,
weather conditions, types of pests controlled, and other factors. Fumigants, fungicides, and other pesticides appear to be effective in reducing
loss and spoilage during storage, distribution, and marketing of crops and
foodstuffs. Herbicides allow corn, wheat, and other cereal grains to be
planted earlier in the growing season at high densities that raise productivity but preclude machine weed control. Several fungicides, soil fumigants, and insecticides control fungal and insect diseases that are major
problems in potato production; up to half of vegetable and fruit produce
could be lost to insects and fungi during storage and transportation without the use of pesticides. The expert committee concluded that chemical
pesticides would continue to be important because reduced-risk chemical products are being registered and competitive alternatives are not
available.
Pesticides
183
Conclusions
Proven Child Health Outcomes
• Poisonings—50,000 reports annually of U.S. children exposed to pesticide products in and around homes, with about 25% requiring health
care services and about 3% causing moderate or severe symptoms
• Reduced sperm counts and infertility—among men occupationally exposed during production of chlordecone and DBCP
Unresolved Issues and Knowledge Gaps
• Developmental effects—limited evidence that periconceptual parental
pesticide exposure may cause early and late fetal deaths, preterm delivery, IUGR, and birth defects (particularly orofacial and limb reduction defects)
• Neurotoxicity
° Inadequate evidence to assess the role of perinatal pesticide exposure
in brain development
° Limited evidence that occupational exposure of older children (and
adults) to OPs or carbamates can cause prolonged suppression of
AChE, visual disturbances, and peripheral neuropathy
• Childhood cancer—limited evidence that parental exposures before or
during pregnancy and childhood exposures may cause childhood cancers (brain cancer, leukemia, non-Hodgkin’s lymphoma, Wilms’ tumor,
Ewing’s sarcoma)
• Knowledge gaps—longitudinal studies needed beginning in the first
trimester to assess the risk of subtle and delayed health effects of pesticides and other contaminants with known or suspected developmental toxicity
Risk Management Issues
• Prevention
° Toxicity testing—need comprehensive testing of pesticide active ingredients and inert components for developmental toxicity
° Exposure reduction—need to minimize occupational and domestic
exposures of reproductive-age persons, pregnant women, and young
children to pesticides, especially those known or suspected to be developmental toxins or carcinogens
° Efficacy testing—need rigorous efficacy testing of pesticides to justify the
explicit trade-off between benefits and harm in their risk management
184
CHILD HEALTH
AND THE
ENVIRONMENT
• Biomonitoring—need to measure internal doses of pesticides periodically in representative samples of children and reproductive-age
men and women to monitor progress in exposure reduction and to
identify high-exposure groups
References
Arbuckle TE, Lin Z, Mery LS. (2001). An exploratory analysis of the effect of pesticide exposure on the risk of spontaneous abortion in an Ontario farm population. Environ Health Perspect 109:851–7.
Arbuckle TE, Savitz DA, Mery LS, Curtis KM. (1999). Exposure to phenoxy herbicides and the risk of spontaneous abortion. Epidemiology 10:752–60.
Arbuckle TE, Sever LE. (1998). Pesticide exposures and fetal death: a review of
the epidemiologic literature. Crit Rev Toxicol 28:229–70.
Attorney General of New York. (1996). Pesticides in schools: reducing the risks.
Located at http://www.oag.state.ny.us/environment/schools96.html.
Bell EM, Hertz-Picciotto I, Beaumont JJ. (2001a). A case-control study of pesticides
and fetal death due to congenital anomalies. Epidemiology 12:148–56.
Bell EM, Hertz-Picciotto I, Beaumont JJ. (2001b). Case-cohort analysis of agricultural pesticide applications near maternal residence and selected causes of fetal death. Am J Epidemiol 154:702–10.
Bell EM, Hertz-Picciotto I, Beaumont JJ. (2001c). Pesticides and fetal death due to
congenital anomalies: implications of an erratum. Epidemiology 12:595–6.
Briassoulis G, Narlioglou M, Hatzis T. (2001). Toxic encephalopathy associated
with use of DEET insect repellents: a case analysis of its toxicity in children.
Hum Exp Toxicol 20:8–14.
Centers for Disease Control and Prevention. (2000). Illnesses associated with use
of automatic insecticide dispenser units—selected states and United States,
1986–1999. MMWR 49:492–5.
Centers for Disease Control and Prevention. (2001). National report on human exposure to environmental chemicals: dimethylphosphate. Atlanta: Centers for
Disease Control and Prevention.
Chen WL, Sheets JJ, Nolan RJ, Mattsson JL. (1999). Human red blood cell acetylcholinesterase inhibition as the appropriate and conservative surrogate endpoint for establishing chlorpyrifos reference dose. Regul Toxicol Pharmacol
29:15–22.
Claudio L, Kwa WC, Russell AL, Wallinga D. (2000). Testing methods for developmental neurotoxicity of environmental chemicals. Toxicol Appl Pharmacol
164:1–14.
Cordier S, Mandereau L, Preston-Martin S, Little J, Lubin F, Mueller B, Holly E,
Filippini G, Peris-Bonet R, McCredie M, and others. (2001). Parental occupations and childhood brain tumors: results of an international case-control
study. Cancer Causes Control 12:865–74.
Craan AG, Haines DA. (1998). Twenty-five years of surveillance for contaminants
in human breast milk. Arch Environ Contam Toxicol 35:702–10.
Daniels JL, Olshan AF, Savitz DA. (1997). Pesticides and childhood cancers. Environ Health Perspect 105:1068–77.
Daniels JL, Olshan AF, Teschke K, Hertz-Picciotto I, Savitz DA, Blatt J, Bondy ML,
Pesticides
185
Neglia JP, Pollock BH, Cohn SL, and others. (2001). Residential pesticide exposure and neuroblastoma. Epidemiology 12:20–7.
Darvill T, Lonky E, Reihman J, Stewart P, Pagano J. (2000). Prenatal exposure to
PCBs and infant performance on the Fagan Test of Infant Intelligence. Neurotoxicology 21:1029–38.
Dich J, Zahm SH, Hanberg A, Adami HO. (1997). Pesticides and cancer. Cancer
Causes Control 8:420–43.
Dimich-Ward H, Hertzman C, Teschke K, Hershler R, Marion SA, Ostry A, Kelly
S. (1996). Reproductive effects of paternal exposure to chlorophenate wood
preservatives in the sawmill industry. Scand J Work Environ Health 22:267–73.
Engel LS, O’Meara ES, Schwartz SM. (2000). Maternal occupation in agriculture
and risk of limb defects in Washington State, 1980–1993. Scand J Work Environ Health 26:193–8.
Eskenazi B, Castorina R. (1999). Association of prenatal maternal or postnatal child
environmental tobacco smoke exposure and neurodevelopmental and behavioral problems in children. Environ Health Perspect 107:991–1000.
European Union. (1998). Council directive 98/83/EC on the quality of water
intented for human consumption. Located at http://europa.eu.int/comm/
environment/water/water-drink/index_en.html.
EXTOXNET (Extension Toxicology Network). (2000). Pesticide active ingredient
profiles. Located at http://pmep.cce.cornell.edu/profiles/extoxnet/index.html.
Fenske RA, Kedan G, Lu C, Fisker-Andersen JA, Curl CL. (2002). Assessment of
organophosphorus pesticide exposures in the diets of preschool children in
Washington State. J Expo Anal Environ Epidemiol 12:21–8.
Fenske RA, Kissel JC, Lu C, Kalman DA, Simcox NJ, Allen EH, Keifer MC. (2000).
Biologically based pesticide dose estimates for children in an agricultural community. Environ Health Perspect 108:515–20.
Feychting M, Plato N, Nise G, Ahlbom A. (2001). Paternal occupational exposures
and childhood cancer. Environ Health Perspect 109:193–6.
Garcia AM. (1998). Occupational exposure to pesticides and congenital malformations: a review of mechanisms, methods, and results. Am J Ind Med 33:
232–40.
Gocmen A, Peters HA, Cripps DJ, Bryan GT, Morris CR. (1989). Hexachlorobenzene episode in Turkey. Biomed Environ Sci 2:36–43.
Goldman LR, Koduru S. (2000). Chemicals in the environment and developmental toxicity to children: a public health and policy perspective. Environ Health
Perspect 108(Suppl 3):443–8.
Gunier RB, Harnly ME, Reynolds P, Hertz A, Von Behren J. (2001). Agricultural
pesticide use in California: pesticide prioritization, use densities, and population distributions for a childhood cancer study. Environ Health Perspect 109:
1071–8.
Gurunathan S, Robson M, Freeman N, Buckley B, Roy A, Meyer R, Bukowski J,
Lioy PJ. (1998). Accumulation of chlorpyrifos on residential surfaces and toys
accessible to children. Environ Health Perspect 106:9–16.
Hill RH Jr, Head SL, Baker S, Gregg M, Shealy DB, Bailey SL, Williams CC, Sampson EJ, Needham LL. (1995). Pesticide residues in urine of adults living in the
United States: reference range concentrations. Environ Res 71:99–108.
Hosie S, Loff S, Witt K, Niessen K, Waag KL. (2000). Is there a correlation between
organochlorine compounds and undescended testes? Eur J Pediatr Surg 10:304–9.
Infante-Rivard C, Labuda D, Krajinovic M, Sinnett D. (1999). Risk of childhood leu-
186
CHILD HEALTH
AND THE
ENVIRONMENT
kemia associated with exposure to pesticides and with gene polymorphisms.
Epidemiology 10:481–7.
International Agency for Research on Cancer. (1987). Overall evaluations of carcinogenicity: an updating of IARC Monographs Volumes 1 to 42 (Supplement
7). Lyon, France: International Agency for Research on Cancer.
Jarrell J, Gocmen A, Foster W, Brant R, Chan S, Sevcik M. (1998). Evaluation of
reproductive outcomes in women inadvertently exposed to hexachlorobenzene
in southeastern Turkey in the 1950s. Reprod Toxicol 12:469–76.
Karmaus W, Asakevich S, Indurkhya A, Witten J, Kruse H. (2002). Childhood
growth and exposure to dichlorodiphenyl dichloroethene and polychlorinated
biphenyls. J Pediatr 140:33–9.
Karmaus W, Kuehr J, Kruse H. (2001). Infections and atopic disorders in childhood and organochlorine exposure. Arch Environ Health 56:485–92.
Karmaus W, Wolf N. (1995). Reduced birthweight and length in the offspring of
females exposed to PCDFs, PCP, and lindane. Environ Health Perspect 103:
1120–5.
Korrick SA, Chen C, Damokosh AI, Ni J, Liu X, Cho SI, Altshul L, Ryan L, Xu X.
(2001). Association of DDT with spontaneous abortion: a case-control study.
Ann Epidemiol 11:491–6.
Krajinovic M, Sinnett H, Richer C, Labuda D, Sinnett D. (2002). Role of NQO1,
MPO and CYP2E1 genetic polymorphisms in the susceptibility to childhood
acute lymphoblastic leukemia. Int J Cancer 97:230–6.
Kristensen P, Irgens LM, Andersen A, Bye AS, Sundheim L. (1997). Gestational
age, birth weight, and perinatal death among births to Norwegian farmers,
1967–1991. Am J Epidemiol 146:329–38.
Landrigan PJ, Claudio L, Markowitz SB, Berkowitz GS, Brenner BL, Romero H,
Wetmur JG, Matte TD, Gore AC, Godbold JH, and others. (1999). Pesticides
and inner-city children: exposures, risks, and prevention. Environ Health Perspect 107:431–7.
Litovitz TL, Klein-Schwartz W, White S, Cobaugh DJ, Youniss J, Omslaer JC, Drab
A, Benson BE. (2001). 2000 Annual report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med
19:337–95.
Loffredo CA, Silbergeld EK, Ferencz C, Zhang J. (2001). Association of transposition of the great arteries in infants with maternal exposures to herbicides and
rodenticides. Am J Epidemiol 153:529–36.
Longnecker MP, Klebanoff MA, Brock JW, Zhou H, Gray KA, Needham LL, Wilcox
AJ. (2002). Maternal serum level of 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene and risk of cryptorchidism, hypospadias, and polythelia among male offspring. Am J Epidemiol 155:313–22.
Longnecker MP, Klebanoff MA, Zhou H, Brock JW. (2001). Association between
maternal serum concentration of the DDT metabolite DDE and preterm and
small-for-gestational-age babies at birth. Lancet 358:110–4.
Lu C, Fenske RA, Simcox NJ, Kalman D. (2000). Pesticide exposure of children in
an agricultural community: evidence of household proximity to farmland and
take home exposure pathways. Environ Res 84:290–302.
Ma X, Buffler PA, Gunier RB, Dahl G, Smith MT, Reinier K, Reynolds P. (2002).
Critical windows of exposure to household pesticides and risk of childhood
leukemia. Environ Health Perspect 110:955–60.
Pesticides
187
MacIntosh DL, Spengler JD, Ozkaynak H, Tsai L, Ryan PB. (1996). Dietary exposures to selected metals and pesticides. Environ Health Perspect 104:202–9.
Michalek JE, Rahe AJ, Boyle CA. (1998). Paternal dioxin, preterm birth, intrauterine growth retardation, and infant death. Epidemiology 9:161–7.
National Academy of Sciences. (1993). Pesticides in the diets of infants and children. Washington, DC: National Academy Press.
National Academy of Sciences. (2000). The future role of pesticides in U.S. agriculture. Washington, DC: National Academy Press.
National Academy of Sciences. (2001). Veterans and agent orange. Update 2000.
Washington, DC: National Academy Press.
New York City Department of Health. (2001). Adult mosquito control programs.
Final environmental impact statement. New York: New York City Department
of Health.
Nishioka MG, Lewis RG, Brinkman MC, Burkholder HM, Hines CE, Menkedick
JR. (2001). Distribution of 2,4-D in air and on surfaces inside residences after
lawn applications: comparing exposure estimates from various media for
young children. Environ Health Perspect 109:1185–91.
Nordstrom M, Hardell L, Lindstrom G, Wingfors H, Hardell K, Linde A. (2000).
Concentrations of organochlorines related to titers to Epstein-Barr virus early
antigen IgG as risk factors for hairy cell leukemia. Environ Health Perspect
108:441–5.
Noren K, Meironyte D. (2000). Certain organochlorine and organobromine contaminants in Swedish human milk in perspective of past 20–30 years. Chemosphere 40:1111–23.
Nurminen T. (1995). Maternal pesticide exposure and pregnancy outcome. J Occup Environ Med 37:935–40.
Savitz DA, Arbuckle T, Kaczor D, Curtis KM. (1997). Male pesticide exposure and
pregnancy outcome. Am J Epidemiol 146:1025–36.
Shaw GM, Wasserman CR, O’Malley CD, Nelson V, Jackson RJ. (1999). Maternal
pesticide exposure from multiple sources and selected congenital anomalies.
Epidemiology 10:60–6.
Sinnett D, Krajinovic M, Labuda D. (2000). Genetic susceptibility to childhood
acute lymphoblastic leukemia. Leuk Lymphoma 38:447–62.
Thomas DC, Petitti DB, Goldhaber M, Swan SH, Rappaport EB, Hertz-Picciotto I.
(1992). Reproductive outcomes in relation to malathion spraying in the San
Francisco Bay Area, 1981–1982. Epidemiology 3:32–9.
U.S. Environmental Protection Agency. (1999a). Pesticide industry sales and usage: 1996 and 1997 market estimates (EPA733-R-99-01). Washington, DC: U.S.
Environmental Protection Agency.
U.S. Environmental Protection Agency. (1999b). Transport of lawn-applied 2,4-D
from turf to home: assessing the relative importance of transport mechanisms
and exposure pathways (EPA 600-R-99-040). Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2000). Chlorpyrifos methyl: toxicology
section of the RED chapter. Located at http://www.epa.gov/pesticides/op/
chlorpyrifosmethyl/rev_toxicology.pdf.
U.S. Environmental Protection Agency. (2001a). Current drinking water standards.
Located at http://www.epa.gov/safewater/mcl.html.
U.S. Environmental Protection Agency. (2001b). Lists of other (inert) pesticide
188
CHILD HEALTH
AND THE
ENVIRONMENT
ingredients. Washington, DC: Located at http://www.epa.gov/opprd001/
inerts/lists.html.
U.S. Food and Drug Administration. (2000). Pesticide program: residue monitoring 1999. Washington DC: U.S. Food and Drug Administration.
Weidner IS, Moller H, Jensen TK, Skakkebaek NE. (1998). Cryptorchidism and hypospadias in sons of gardeners and farmers. Environ Health Perspect 106:
793–6.
Whyatt RM, Barr DB. (2001). Measurement of organophosphate metabolites in
postpartum meconium as a potential biomarker of prenatal exposure: a validation study. Environ Health Perspect 109:417–20.
World Health Organization. (1998). Guidelines for drinking water quality, 2nd edition, Vol. 2, Health criteria and other supporting information, 1996 (p 940–949)
and Addendum to Vol. 2, 1998 (p 281–283). Located at http://www.who.int/
water_sanitation_health/GDWQ/index.html.
Zahm SH, Ward MH. (1998). Pesticides and childhood cancer. Environ Health Perspect 106(Suppl 3):893–908.
8
Hormonally Active Agents
In multicellular organisms, cell signaling through the nervous, immune,
and endocrine systems is essential for coordinating metabolic functions
among all cells. Signals range from nutrient and metabolite levels in extracellular fluids to specialized systems including chemical messengers,
storage and transportation mechanisms, and receptors. Specialized endocrine cells produce chemical messengers (hormones) that interact with
receptors in local and distant target cells. There are intimate links between
the endocrine and nervous systems such as the hypothalamic–pituitary
gland complex in which specialized neurons produce hormones.
Depending on the hormone, target cells may be confined to a specific organ or may be more widespread. By binding to receptors and influencing intracellular signaling systems, hormones control many processes at the molecular level during gestation and postnatal life. Major
endocrine glands include the pituitary gland, thyroid gland, parathyroid
glands, pancreatic islets, adrenal glands, testes, and ovaries. When activated by a hormone, receptors on or in target cells trigger a cascade of
intracellular reactions. An agonist is a hormone or chemical that binds to
a receptor and induces specific biochemical effects; antagonists compete
for binding to a receptor but do not induce the biochemical effects associated with the receptor.
189
190
CHILD HEALTH
AND THE
ENVIRONMENT
Endocrine disruptors have been defined as exogenous agents that interfere with the synthesis, secretion, transport, binding, action, or elimination of natural hormones in the body and may cause adverse effects
at the level of an organism, progeny, and populations. As recommended
by the National Academy of Sciences, the term hormonally active agents
(HAAs) is used here to describe substances with hormone-like activity,
regardless of the mechanism (National Academy of Sciences, 1999). Although there has been concern during recent years about environmental
HAAs, especially in relation to apparent estrogenic effects in wildlife, the
ability of DDT/DDE to inhibit testicular and secondary sexual development in roosters was reported several decades earlier (Burlington and
Lindeman, 1950). Experience with the potent synthetic estrogen, diethylstilbestrol (DES), showed that prenatal exposure of women to an HAA,
albeit a therapeutic product and not an environmental contaminant, could
cause early and delayed adverse developmental, reproductive, and carcinogenic effects in offspring; use of other drugs including estrogens, androgens, and progestins during pregnancy has also been linked to abnormal reproductive development. Wildlife and experimental animal
studies have shown important adverse effects of HAAs at high doses on
reproductive system development, sexual behavior, fertility, and immune
function (Vos et al., 2000).
Several studies indicate that incidence rates of birth defects of the penis (hypospadias) and testes (cryptorchidism) and testicular cancer in humans have increased during recent decades, while sperm quality has declined. The validity of these trends and the possible role of environmental
HAAs as potential causes remain uncertain and controversial. On the one
hand, the increased testicular cancer incidence rates, particularly among
young men, are supported by high-quality population-based incidence
data from several countries (see, e.g., Power et al., 2001; Weir et al., 1999).
These data indicate that testicular cancer incidence rates increased twofold
or more during the past three to four decades, especially among more recent birth cohorts. On the other hand, cryptorchidism birth prevalence
rates from the late 1960s to the late 1990s have varied considerably among
countries and irregularly over time within countries (Paulozzi, 1999).
Prevalence rates of cryptorchidism are influenced by diagnostic efficiency
and age at examination; in about 70% of affected neonates, testes spontaneously descend by age 3 months. Increased clinical surveillance and
intervention, especially soon after birth, tend to detect many cases that
would otherwise have resolved spontaneously.
Birth prevalence rates of hypospadias, a birth defect in which the urethra opens on the ventral surface of the penis, also vary substantially
among countries (e.g., 0.26 per 1000 male births in Mexico and 2.11 in
Hormonally Active Agents
191
Hungary); even in countries with high reported rates, there was a
30%–40% underascertainment of cases later treated surgically (Kallen et
al., 1986). Hypospadias rates approximately doubled in the United States
during 1968–1993; rates of severe hypospadias and the ratio of severe to
mild cases both increased over time, findings suggestive of a true increase
(Paulozzi et al., 1997). Increases have been reported by surveillance systems in Alberta (Canada), Norway, and Israel but not in several other developed countries. In Finland, where all children are examined at birth
by a pediatrician, the cumulative prevalence of hypospadias to age 8 years
remained constant during 1970–1986 (Aho et al., 2000). Variable case detection and reporting over time and place may explain the inconsistent
hypospadias incidence trends.
Available evidence suggests that average sperm concentrations (number of sperm per unit volume) and semen volumes both declined substantially during 1938–1991 (Carlsen et al., 1992). These trends occurred
in several but not all countries for which data were available, and most
of the apparent decrease occurred before 1970. A recent reanalysis concluded that changes in counting methods and population characteristics
such as abstinence time did not explain the observed trends (Swan and
Elkin, 1999). A survey of semen quality in four European cities, using standardized measurement methods, showed that the range of average sperm
concentrations across centers was 98–132 million/ml (Jorgensen et al.,
2001). There appear to have been no population-based studies in which
semen quality and exposures to nonoccupational environmental HAAs
were measured at the individual level.
There is continuing controversy about the reality of declining sperm
production and potential environmental links. A review of available evidence concluded that estrogenic environmental HAAs play only a minor
role in the trends noted above and that further research is needed on other
endocrine mechanisms including thyroid and androgen-dependent processes (Foster, 1998). The National Research Council also concluded that
there is insufficient toxicologic and epidemiologic evidence to attribute
changes in hypospadias, cryptorchidism, testicular cancer, and sperm
counts to environmental HAAs (National Academy of Sciences, 1999). The
potential for exposure to environmental HAAs and evidence of the sensitivity of the developing reproductive tract from studies in animals and
humans, as well as the many knowledge gaps, support continuing concern about this issue.
This chapter can do little more than introduce the reader to the complexities of the endocrine system and the known or potential impacts of
environmental HAAs on animal and human health. The objective is to
describe current knowledge and uncertainties in this field, particularly in
192
CHILD HEALTH
AND THE
ENVIRONMENT
relation to the limited human evidence and the more definitive animal
evidence of developmental, reproductive, and carcinogenic effects. The
focus is on areas where much of the animal research has been done, that
is, the effects of environmental HAAs on the normal functioning of endogenous estrogens, androgens, and thyroid hormones. The ability of anthropogenic environmental HAAs to act through both hormonal and nonhormonal mechanisms is described. The chapter closes with a discussion
of the need for improved risk management including bioassays to detect
HAAs, biomonitoring, and research on major knowledge gaps.
Normal Endocrine Function
The concentrations of hormones at receptor sites in target cells relate to
rates of synthesis and secretion, delivery, degradation, and elimination.
The strongest controls are feedback loops, usually negative feedback, at
the level of hormone synthesis and secretion. The two main categories of
hormone receptors are cell surface and nuclear receptors:
• Cell surface receptors—protein and peptide hormones usually bind to
cell surface receptors, activating intracellular second messengers that
modulate the activity of enzymes and other intracellular targets; second messengers and examples of hormones that activate them include
cyclic adenosine monophosphate (AMP) (LH, FSH, TSH1), protein kinase (insulin, growth hormone, prolactin, several growth factors), and
calcium and/or phosphoinositides (gonadotropin-releasing hormone,
TRH1).
• Nuclear receptors—steroids and thyroid hormones bind to nuclear receptors that are ligand-dependent transcription factors; the hormone–
nuclear receptor complexes bind to promoter regions of hormoneresponsive genes and stimulate or inhibit their transcription
Through either type of receptor, the effect of small hormone concentrations is greatly amplified to produce biologic effects.
Gonadal Development
The Y chromosome gene SRY and other regulator genes control development of the male gonadal ridge into a testis with seminiferous cords con1 LH luteinizing hormone, FSH follicle-stimulating hormone, TSH thyroidstimulating hormone, TRH thyrotropin-releasing hormone.
Hormonally Active Agents
193
taining Sertoli and primordial germ cells and interstitial tissue containing
Leydig cells. During early pregnancy, SRY directs embryonic bipotential
gonadal ridge cells to produce Mullerian inhibitory hormone (MIH) and
testosterone (T) and to downregulate expression of CYP19, a gene that encodes the aromatase enzyme that converts T to estradiol-17 (E2). Mullerian inhibitory hormone causes regression of the Mullerian ducts that
form the fallopian tubes, uterus, and upper vagina in females. Testosterone
and MIH induce the differentiation of the mesonephric (Wolffian) ducts
into the epididymis, seminal vesicles, and prostate. During early infancy,
testicular volume increases rapidly, reflecting Sertoli cell division and
growth in the length of the seminiferous cords; thyroid hormones regulate
the duration of infant Sertoli cell proliferation, affecting adult Sertoli cell
number and sperm-producing capacity.
At puberty, pulsatile release of hypothalamic gonadotropin-releasing
hormone stimulates similar pulsatile release of LH and FSH from the anterior pituitary gland. Luteinizing hormone binds to receptors on target
cells (testicular Leydig cells in boys and, in girls, ovarian theca interna,
theca lutein, and granulosa lutein cells) and stimulates T production in
boys and sustained follicle growth, E2 production, inhibin A production,
granulosa cell luteinization, ovulation, luteal formation, and progesterone production in girls (Table 8–1). Follicle-stimulating hormone binds to
receptors on ovarian granulosa cells and testicular Sertoli cells, stimulating proliferation and differentiation of granulosa cells (including inhibin
B secretion and CYP19 induction) in girls, and Sertoli cell inhibin B secretion and spermatogenesis in boys.
Testosterone stimulates Sertoli cell differentiation, spermatogenesis,
growth in seminiferous tubule diameter, and further testicular growth and
has anabolic effects, that is, it stimulates growth of nonreproductive tissues including muscle, kidney, liver, and salivary gland. Sertoli cells produce several hormones of the transforming growth factor (TGF-) family, including MIH, inhibin, and activin. Production of MIH occurs in the
fetus and child until about age 8–10 years; at puberty, Sertoli cells develop
androgen receptors, and increased T levels suppress MIH production. The
peptide hormones inhibin A and B, produced by Sertoli and ovarian granulosa cells, provide negative feedback signals to suppress pituitary FSH
synthesis.
In the early female fetus, primordial germ cells migrate to the bipotential gonadal ridge and are incorporated in the developing ovary as
oogonia that divide by mitosis initially and then enter meiosis. It appears
that female fetal reproductive tract development requires signals to maintain the Mullerian duct, repress the Wolffian duct, suppress Leydig cell
development, and maintain meiotic oocytes. Meiosis is halted during
TABLE 8–1. Functions of Selected Endogenous Hormones
Hormone, Main Site of Production
Function
LH; anterior pituitary gland
Binds to LH receptors on testicular Leydig cells and ovarian theca interna, theca lutein, and
granulosa lutein cells; stimulates T production in males and stimulates sustained follicle growth,
E2 production, inhibin A production, granulosa cell luteinization, ovulation, luteal formation, and
progesterone production in females
FSH; anterior pituitary gland
Binds to FSH receptors on testicular Sertoli cells and ovarian granulosa cells; stimulates proliferation
and differentiation of granulosa cells (including inhibin A production and CYP19 induction) and
Sertoli cell inhibin B production and spermatogenesis
Human chorionic gonadotropin (hCG)—fetal
form is produced in fetal liver and kidney
Fetus—little maternal hCG crosses the placenta; fetal hCG may control fetal androgen synthesis, go
nadal steroid production, and brain growth and differentiation
Estradiol (E2); ovarian theca interna, theca
lutein, and granulosa lutein cells (and small
amounts in testicular Leydig cells)
Binds to estrogen receptors in target tissues (ovary, breast, uterus, cervix, liver, bone, other) and
stimulates transcription of estrogen-responsive genes; exerts positive feedback on frequency and
amplitude of hypothalamic gonadotropin-releasing hormone and LH during the late follicular
phase of the menstrual cycle but negative feedback at other times
T; testicular Leydig cells
Male fetus—fetal Leydig cells produce T that stabilizes Wolffian ducts and stimulates growth and
development of the epididymis, vasa deferentia, and seminal vesicles
Adolescent boys and men—Leydig cells produce T that stimulates spermatogenesis; T is
converted to DHT in prostate, liver, kidney, skin, and muscle; adolescent girls and women—
T serves as precursor for E2 synthesis; exerts negative feedback on frequency and amplitude of
hypothalamic gonadotropin-releasing hormone and LH; stimulates growth of nonreproductive
tissues including muscle, kidney, liver, salivary gland
Dihydrotestosterone (DHT); fetal external
genitalia
Male fetus—differentiation of penis, scrotum, prostate, Cowper glands
Inhibin; Sertoli and granulosa cells
Exerts negative feedback on FSH secretion
TSH; anterior pituitary gland
Binds to TSH receptor on thyroid follicular epithelial cells; stimulates all aspects of thyroid hormone
production except storage, that is, production of thyroglobulin, endocytosis of thyroglobulin,
proteolytic cleavage of thyroglobulin to release T4, deiodination of T4 to T3, and secretion of T4
and T3 into blood
Thyroid hormones—triiodothyronine (T3),
thyroxine (T4); thyroid gland
Bind to receptors (TR␣, TR, and their isoforms); TR2 mediates negative feedback of T4 on
pituitary TSH secretion; other TR isoforms are ubiquitously expressed and mediate the effects of
T3 and T4 needed for normal function of every organ system
Source: Darlington and Dallman (2001).
196
CHILD HEALTH
AND THE
ENVIRONMENT
prophase of the first meiotic division of oogonia, yielding primordial follicles, that is, primary oocytes enclosed in single layers of follicular cells.
The primordial follicles in each fetal ovary decrease during later life until relatively few remain at menopause. Follicle-stimulating hormone
binds to ovarian granulosa cell FSH receptors and stimulates proliferation and differentiation; the follicle(s) most responsive to FSH express the
LH receptor to which LH binds and stimulates sustained follicle growth
and E2 production. Pituitary LH surges cause further granulosa cell differentiation (luteinization) and ovulation.
Hormone Synthesis
Steroid synthesis in the gonads and adrenals is subject to acute and chronic
regulation, the latter occurring at the level of gene transcription. The
steroidogenic acute regulatory protein (StAR) controls the rate-limiting
step in steroid hormone synthesis; by activating a regulated channel, it
allows translocation of cholesterol from the outer to the inner mitochondrial membrane and rapid increases in steroid synthesis. Expression of
StAR in gonads is controlled by LH and FSH. Most steroidogenic enzymes
belong to the cytochrome P450 family encoded by CYP genes, several of
which are controlled by LH and FSH (see Fig. 8–1):
• LH—controls several genes including
° CYP11A1—encodes the enzyme P450scc (cholesterol-20,22-desmolase), which converts cholesterol to pregnenolone
° CYP17—encodes P450c17, which has 17␣-hydroxylase and 17,20lyase activities and acts at key branch points in steroidogenesis
° 3-Hydroxysteroid dehydrogenase—converts pregnenolone to progesterone in ovarian thecal cells
° 17-Hydroxysteroid dehydrogenase—converts 4-androstenedione to
T and E2 to estrone and vice versa
• Follicle-stimulating hormone increases expression of CYP19 in ovarian
granulosa cells; this gene encodes P450arom, an enzyme with aromatase activity that converts T to E2 and 4-androstenedione to estrone
• 5␣-Reductase—this enzyme converts T to DHT, which has about 2.5
times more potency than T as an androgen receptor agonist; occurs in
Leydig cells, prostate, and skin; expression in skin is influenced by genetic factors, transforming growth factor, insulin-like growth factor, and
circulating androgens
Thyroid hormones control growth, development, differentiation, and metabolism in vertebrates. Hypothalamic neurons produce TRH, which, in
Hormonally Active Agents
197
Cholesterol
Controlled by LH
P450scc/CYP11A1 (controlled by LH)
3βHSD
Progesterone
Pregnenolone
3βHSD
17α-Hydroxyprogesterone
P450c17/CYP17 (controlled by LH)
17α-Hydroxypregnenolone
P450c17/CYP17 (controlled by LH)
Dehydroepiandrosterone
P450c17/CYP17
(controlled by LH)
3β-HSD
4Androstendione
P450arom/CYP19
(controlled by FSH)
17β-HSD
(controlled by LH)
Testosterone
Estradiol-17β
17β-HSD
5α-Reductase
Dihydrotestosterone
Estrone
FIGURE 8–1. Androgen and estrogen biosynthesis, showing enzymes/genes that
control each step.
turn, stimulates cells in the anterior pituitary gland to secrete TSH; the
latter binds to thyroid epithelial cell receptors, activates a G-protein-linked
signaling system, and stimulates all aspects of thyroid hormone synthesis (except storage): production, endocytosis, and proteolytic cleavage of
thyroglobulin, deiodination of T4 to T3, and secretion of T4 and T3 into
blood. At high circulating levels, T3 and T4 inhibit secretion of TRH by
hypothalamic neurons; this inhibition is reversed when T3 and T4 levels
198
CHILD HEALTH
AND THE
ENVIRONMENT
are low. Triiodothyronine and T4 bind to ubiquitously expressed receptors (TR␣, TR, and their isoforms); TR2 mediates negative feedback of
T4 on pituitary TSH secretion. The various TR isoforms mediate the essential roles of T3 and T4 in the normal function of every organ system.
Hormone Receptors
Several small, lipid-soluble hormones (T, E2, T3, T4) and signaling molecules (vitamin D3, retinoic acid, retinoids, and certain fatty acids) bind to
nuclear receptors coded by genes of the nuclear receptor superfamily. The
best-known effects of sex steroid hormones are delayed because of the
time required for their interaction with nuclear receptors in target tissues,
activation of gene transcription, and impacts on phenotype arising from
cell division and cell differentiation.
Estrogen Receptors
There are two known estrogen receptor (ER) subtypes in rodents and humans, ER␣ and ER, having tissue-specific patterns of expression and reacting differently to the same ligand. A wide variety of chemical ligands
can bind to ERs with variable affinities and half-lives and with effects
ranging from full agonist to full antagonist. E2 binds to either receptor
with similar affinities, and the resulting complexes interact with short
DNA promoter sequences known as estrogen-response elements (EREs) and
induce expression of several genes (Kuiper et al., 1998). In contrast, the
affinities of the phytoestrogens genistein and coumestrol for ER substantially exceed those for ER␣.
The relative expression of ER␣ and ER appears to be a key determinant of cell responses to ER ligands; for instance, antiestrogens such as
the drug tamoxifen show some agonistic activity with ER␣ but not with
ER. Human ER decreases cell sensitivity to E2 and modulates ER␣ transcriptional activity. Estrogen receptor  is expressed in human Sertoli,
Leydig, and peritubular myoid cells, variably in germ cells, and in epithelial and stromal cells throughout other parts of the male reproductive
tract; ER␣ is expressed in seminal vesicles. The presence of ERs in fetal
and adult male germ and Sertoli cells suggests that estrogen influences
testicular development and spermatogenesis.
During embryogenesis, ERs are also expressed in many nonreproductive tissues including the fetal rodent brain, suggesting a role for estrogen in early brain development. In human brain, ER␣ appears to dominate in the hypothalamus and amygdala, while ER is prominent in the
hippocampal formation, entorhinal cortex, and thalamus. In rodents, T induces CYP19 expression in cortical and hypothalamic neurons during fe-
Hormonally Active Agents
199
tal development, enabling such neurons to convert T to E2. Despite low
ER levels, bone responds strongly to estrogen, possibly through an ERrelated receptor (ERR␣).
Androgen Receptors
A member of the nuclear receptor gene superfamily on the X chromosome
encodes the androgen receptor (AR). The ARs are expressed ubiquitously
in genital and nongenital tissues in both sexes including ovarian granulosa cells, skin, eccrine glands, sebocytes, hair follicle root sheath, liver,
prostate, oral mucosa, breast, and muscle. By mediating the biologic actions of T and DHT, AR controls the growth, differentiation, and function
of the male urogenital tract. In addition, AR modulates transcription
of androgen-responsive genes (e.g., sex-limited protein, probasin, and
prostate-specific antigen). Also, AR interacts with other signaling pathways. Gene polymorphisms and mutations of AR may cause up to 10%
of human male infertility.
Progesterone Receptors
Progesterone, produced in the ovary during the latter half of the menstrual cycle and early pregnancy and by the placenta during later gestation in humans, targets two types of progesterone receptors, PR-A and
PR-B, encoded by one gene with two distinct estrogen-regulated promoter
sites. Although the distinct functions of the PR subtypes are unknown,
knockout mice missing the PR gene display anovulation, uterine hyperplasia, and lack of mammary gland development. Progesterone signaling
through PRs is essential for lobulo-alveolar breast development during
pregnancy but not for ductal growth at puberty. E2 increases responsiveness to progesterone by inducing PRs in neural and uterine tissues. Progesterone also plays important roles in brain function; for example, it inhibits the neuronal nicotinic acetylcholine receptor and stimulates glial
cell myelin production.
Thyroid Hormone Receptors
Thyroid hormone receptors (TRs) and T4-activating enzymes are expressed in the human fetal brain during early gestation, consistent with
the important role of T4 in perinatal brain development. When taken up
by tissues, T4 may undergo outer-ring monodeiodination to active T3 or
inner-ring monodeiodination to inactive T3 (reverse T3 or rT3). Nuclear
receptors TR␣ and TR each have two isoforms and mediate most thyroid hormone actions. When activated by T3, TR promotes transcription
of several genes, facilitated by coactivators that acetylate histones, loosening the chromatin structure and facilitating access of TR and other key
200
CHILD HEALTH
AND THE
ENVIRONMENT
transcription factors. Independent of TR, T3 and T4 have important effects on calcium, sodium, and glucose transport across cell membranes,
enzyme activity, mitochondrial respiration, and actin polymerization.
Acute effects of T3 on mitochondrial respiration may be mediated by
T3 receptors on the inner mitochondrial membrane; T3 also stimulates
transcription of mitochondrial genes through both TR-mediated and
mitochondrial-mediated mechanisms including p43 (Wrutniak-Cabello et
al., 2001). Functions of thyroid hormone and its receptors during development include:
• Development of brain and other systems—severe mental deficits and
multiple birth defects accompany fetal hypothyroidism.
• Auditory development—TR2 is an essential transcription factor for
auditory development
• Thyroid hormone-dependent differentiation of Sertoli cells.
• Astrocyte structure and function—astrocytes comprise nearly 40% of
cells in the human brain and are distinguished by their content of glial
fibrillary acidic protein (GFAP), a component of their cytoskeleton essential for myelination, cell adhesion, and signaling pathways; thyroid
hormone modulates expression of GFAP.
Hormone Transport and Inactivation
Protein-bound hormones provide a reservoir of hormone that can be released quickly as plasma levels of free hormone fall. Transport proteins
also protect hormones from peripheral metabolism, especially by liver enzymes, thus increasing their half-lives. E2 and T bind to sex hormone binding globulin (SHBG) and albumin; thyroid hormones bind to thyroxinebinding globulin, transthyretin, and albumin. The purpose of SHBG is to
regulate the concentration of free androgens and estrogens in plasma;
about 98% of endogenous E2 is bound to plasma proteins, leaving about
2% free to enter target cells. Complexes of SHBG bind to SHBG receptors
on sex steroid responsive cell membranes, activating an adenyl cyclase/
cyclic AMP second messenger cascade that may modulate the effects of
sex steroids; for instance, SHBG inhibits E2-induced proliferation of human breast cancer cells in vitro.
Cytochrome P450 enzymes in liver and kidney hydroxylate sex
steroids to more hydrophilic forms, thus aiding urinary and biliary excretion. In breast tissue, various progestins stimulate the activity of sulfotransferase, an enzyme that converts E2 to inactive estrogen sulfates.
Luteinizing hormone stimulates sulfotransferase activity and inhibits sulfatase enzymes in testicular Leydig cells, thereby inactivating E2. Human
Hormonally Active Agents
201
sulfotransferase also sulfates T3, T4, and reverse T3, contributing to thyroid hormone processing and iodide recycling.
Mechanisms of Environmental
Hormonally Active Agents
Bioassays for Hormonally Active Agents
Hormonal and other effects of selected HAAs are shown in Table 8–2 and
discussed below. No internationally recognized methods exist for quantifying the hormonal activity of pure or mixed substances. Of 17 chemicals
tested in eight different short-term bioassays, only 3 had above-background
ER-binding affinities (bisphenol A, nonylphenol, ,-DDT), with bisphenol A having the highest estrogenic activity in all the bioassays (Andersen
et al., 1999). Based on results of three in vitro screening tests and one in
vivo bioassay of estrogenic activity, the National Academy of Sciences concluded that (1) several chemicals were positive in all four assays (e.g.,
,-DDT, methoxychlor, bisphenol A), while others reacted in only one or
two assays (e.g., toxaphene, dieldrin, diphenylphthalate), and (2) a few were
positive in all three in vitro tests but negative in vivo (e.g., chlordecone, ␣endosulfan, butylbenzyl phthalate) (National Academy of Sciences, 1999).
A systematic examination of selected HAAs in mammalian species
by peer-review panels concluded that definite low-dose effects included
(Melnick et al., 2002)
• Estrogenic—E2, DES, genistein, methoxychlor, nonylphenol
• Immunologic—genistein, methoxychlor, nonylphenol
HAAs with possible low-dose effects included bisphenol (estrogenic) and
vinclozolin (antiandrogenic).
Estrogen Modulators
Estrogen Receptor Agonists
The synthetic estrogen drug DES was used in the 1950s and 1960s for
treatment of threatened miscarriage, causing an unusual type of vaginal
cancer among exposed daughters about 20 years later; it also causes cancer in experimental animals. It is discussed here to show the potential human health effects of high-dose prenatal exposure to a potent HAA. Diethylstilbestrol readily crosses the placenta and differs from maternal E2
in that it is not inactivated by placental enzymes and has weak affinity
TABLE 8–2. Mechanisms of Selected Environmental HAAsa
Compound
Mechanisms of
Hormonal Activity
Genotoxicity
Developmental and Reproductive Toxicity
of Perinatal Exposures
Carcinogenicityb
,⬘-DDE (main
metabolite of the
organochlorine
insecticide, DDT)
AR antagonist
⫹ in vitro
Animals—males had retained nipples,
reduced anogenital distance, delayed
puberty, reduced sperm production,
reduced testicular and accessory sex
organ weight
Humans—DDT is a
probable carcinogen
Animals—DDT causes
liver and thyroid tumors
,-DDT (DDT
isomer; 15–21%
of technical grade
DDT)
ER agonist
⫹ in vitro
Animals—females had increased
uterine weight, premature vaginal
opening
See above
Chlordecone
(Kepone)
ER agonist
⫹ in vitro
Humans—occupationally exposed
men developed reduced sperm
concentration and infertility
Animals—females: neonatal exposure
caused earlier vaginal opening and
persistent vaginal estrus in adulthood
Animals—liver
tumors
Vinclozolin
(fungicide)
AR antagonist—vinclozolin
⫹ in vitro
binds fairly weakly to AR;
its metabolites, M1 and M2,
are stronger antiandrogens
Animals—males had reduced anogenital
distance, hypospadias, agenesis of sex
accessory tissues, retained nipples and
areolas, cryptorchidism, epididymal
agenesis
Not available
TCDD
PCBs
Benzo[a]pyrene
(example of
a PAH)
Aryl hydrocarbon receptor
(AhR) agonist; decreased
AR and ER expression;
disruption of cell differentiation; interference by
TCDD-AhR-Arnt (aromatic
hydrocarbon receptor
nuclear translocator)
complex with binding of
ER to enhancer elements of
estrogen-responsive genes
⫹ in vivo; genotoxicity
Humans—men: increased LH and
appears to be an indirect
FSH levels, decreased T levels
effect; for example,
induction of CYP1A1
Animals—fetal death, low birth weight,
by TCDD may activate
cleft palate, hydronephrosis of the
other endogenous and
kidneys (caused by hyperplasia
exogenous substances
of the epithelium of the ureters),
hypomineralization of teeth, thymic
to reactive metabolites
atrophy, and immunosuppression;
capable of damaging DNA
males had hypospadias, cryptorchidism,
delayed puberty, feminized sexual
behavior, reduced size of testes and
male accessory sex glands (especially
ventral prostate), reduced anogenital
distance; females had ovarian atrophy,
cleft phallus
Humans—known
carcinogen
Coplanar PCBs—see TCDD
above; hydroxylated PCBs
compete with T4 for
binding to transthyretin
and inhibit E2 metabolism
⫹ in vitro and in vivo
(Schiestl et al., 1997);
commercial PCB mixtures
not potent genotoxins
Humans—IUGR, perinatal mortality,
small head circumference (after
high-level prenatal exposure)
Animals—early and late fetal deaths,
reduced fetal growth rate,
hydronephrosis, cleft palate
Humans—probable
carcinogen
AhR agonist—see TCDD
above
⫹ in vitro and in vivo
Animals—early and late fetal deaths,
birth defects, reduced growth rates
prenatally and postnatally, reduced
fertility of offspring
Humans—probable
carcinogen; animals—
skin, lung, forestomach
tumors
Animals—tumors at
multiple sites
Animals—liver, bile duct,
thyroid tumors
(continued)
TABLE 8–2. Mechanisms of Selected Environmental HAAsa (continued)
Mechanisms of
Hormonal Activity
Compound
Developmental and Reproductive Toxicity
of Perinatal Exposures
Genotoxicity
Diesel exhaust
(contains PAHs,
other toxins)
Bisphenol A
(plastic monomer)
AhR agonist—see PAHs
⫹ in vitro and in vivo
hER␣ and hER agonist
⫹ in vitro
Di(2-ethylhexyl)
phthalate (DEHP)
Main testicular target
appears to be Sertoli cells
that proliferate during the
neonatal period in humans
Chromosome aberrations
in vitro
Animals—increased anogenital distances
(both sexes), reduced ovarian
primary follicles and Sertoli cells
Animals—males had increased anogenital
distance, increased prostate size,
decreased epididymal weight,
persistent increased prostate AR levels;
females had increased weight and
earlier onset of puberty
Animals—decreased fertility, degeneration
of seminiferous tubules, fetal death,
birth defects; females had normal
reproductive development, consistent
with DEHP as an antiandrogen
aToxicological
Profiles, Agency for Toxic Substances and Disease Registry, unless otherwise indicated.
bInternational
Agency for Research on Cancer.
Carcinogenicityb
Humans—probable
carcinogen
Animals—lung tumors
Animals—equivocal
carcinogen, with males
being more
susceptible; leukemia,
testicular tumors
Humans—probable
carcinogen; animals—
liver tumors
Hormonally Active Agents
205
for ␣-fetoprotein and SHBG, allowing free DES to access ERs in fetal tissues. In the presence of ␣-fetoprotein or SHBG, a much greater proportion of DES remains free and able to bind to human ER (hER), compared
to E2. Diethylstilbestrol has about two-fold higher affinity than E2 for
hERs (hER␣ and hER) in vitro (Table 8–3), modulates ER-dependent
DNA transcription, causes reduced and mutated transcripts of DNA polymerase- (a DNA repair gene), stimulates mammary gland proliferation,
and causes persistently altered expression of several HOX genes involved
in patterning of the fetal mouse reproductive tract (Block et al., 2000).
TABLE 8–3. Relative Activities of Sex Steroids, DES, Suspected Environmental
Hormonally Active Agents, and Phytoestrogens for Human ERs (hER␣ and hER)
Relative Binding
Affinity a
Compound
Relative
Transactivation
Activityb
hER␣
hER
hER␣
hER
100
100
100
100
Progesterone
0.01
0.01
NA
NA
Testosterone
0.01
0.01
NA
NA
Estradiol-17 (E2)
DES
236
221
117
69
,-DDT
0.01
0.02
54
10
,-DDE
0.01
0.01
NA
NA
2.4
4.7
77
62
2,4,6-Trichloro-4-biphenylolc
2,3,4,5-tetrachloro-4-biphenylolc
3.4
7.2
68
41
-Octylphenol
0.02
0.07
61
57
-Nonylphenol
0.05
0.09
62
34
Bisphenol A
0.01
0.01
50
41
Methoxychlor
0.01
0.01
9
2
Endosulfan
0.01
0.01
6
1
Chlordecone
0.06
0.1
27
1
Genistein
4
87
198
182
Coumestrol
20
140
102
98
Zearalenone
7
5
91
27
0.1
0.5
97
80
Daidzein
a Affinity
for hER relative to estradiol-17.
b Ratio
of luciferase reporter gene induction values of each compound relative to estradiol-17 at concentrations of 1 M each.
c Hydroxylated
PCB congeners.
NA not available in this source (Kuiper et al., 1998).
CHILD HEALTH
206
AND THE
ENVIRONMENT
Although phytoestrogens are not environmental contaminants, some
are hormonally active and are ubiquitous in plant foods commonly eaten
by children, including soy-based infant formula, fruits, and vegetables.
About 200 naturally occurring phytoestrogens have been identified, the
main categories being lignans and isoflavones. Phytoestrogens occur at
particularly high levels in flaxseeds and soybeans. Soy-based infant formulas contain isoflavones (mainly glycosides of genistein and daidzein)
at levels far higher than those in human breast milk. Phytoestrogens vary
widely in their affinity for hER␣ and hER; genistein, coumestrol, and
zearalenone have the highest relative binding affinities for both hER and
ER␣, but their rank order varies for the two subtypes (Table 8–4). Genistein has much higher affinity for ER than for ER␣, suggesting that it may
influence tissues high in ER (e.g., human vascular smooth muscle, prostate epithelial cells, testicular Leydig and Sertoli cells). Hydroxylation by
P450 enzymes increases the ability of some phytoestrogens to bind to and
activate ERs. Although diets high in fruits and vegetables appear to reduce the risk of cardiovascular disease, cancer, and other chronic diseases,
there is no convincing evidence that these benefits are attributable to phytoestrogens per se.
The relative hER binding affinities and ER-dependent gene expression activities of selected sex steroids, suspected environmental HAAs,
and phytoestrogens are shown in Tables 8–3 and 8–4. Important features
include the following: (1) potencies span at least four orders of magnitude, (2) ER binding affinity is weakly correlated with the ability to activate ER-dependent gene expression, (3) ,-DDT, certain hydroxylated
PCB metabolites, and several phytoestrogens are relatively potent inducers of ER-dependent gene expression, and (4) the test substances behave
similarly with ER␣ and ER. Other ER agonists include tris(4-chlorophenyl)methane (TCPM—a recently identified global contaminant of
TABLE 8–4. Relative Binding Affinity of Estradiol-17
and Selected Phytoestrogens for hER␣ and hER
Hormone or Phytoestrogen
hER␣
hER
Estradiol-17 (E2)
100
100
Genistein
1.6
100
Coumestrol
12
34
Zearalenone
22
75
Daidzein
0.2
1.8
Note: Relative affinity, compared to E2 set at 100, based on concentration of competitor needed to displace half of ER-bound E2
(Nikov et al., 2000).
Hormonally Active Agents
207
unknown origin but structurally related to DDT), brominated bisphenol
A compounds, and metabolites of methoxychlor, polybrominated
diphenylethers, and TCPM. Reported findings of extreme synergism between pairs of weakly estrogenic organochlorine pesticides in a human
ER-dependent in vitro assay were not confirmed in subsequent studies,
and the original paper was withdrawn (McLachlan, 1997).
Other Mechanisms
The complexities of environmental HAAs are illustrated by considering
some of their known or suspected mechanisms:
• Mixed effects—some methoxychlor metabolites are agonists of ER and
antagonists of ER and AR.
• Altered expression of ER—prenatal exposure of swine to TCDD caused
increased testicular ER␣ levels.
• Aryl hydrocarbon receptor agonists—TCDD and dioxin-like compounds bind to AhR and trigger a cascade of biochemical responses in
rodents including:
° Upregulation of CYP1A1 and CYP1B1—this may increase hydroxylation and inactivation of E2 and hydroxylation and activation of
xenobiotics (e.g., some hydroxylated PAH metabolites are ER agonists in vitro).
Tissue-specific
downregulation of ER protein and ER mRNA in re°
productive tissues and liver.
• Unknown mechanisms—triazine herbicides are estrogenic in rodents
in vivo; while the mechanism is unknown, it may involve modulation
of hypothalamic control of LH secretion and/or induction of aromatase
(the enzyme that catalyzes conversion of androgens to estrogens).
Androgen Modulators
Androgen Receptor Agonists/Antagonists
Certain isomers of DDT and DDE and some PAHs such as dibenzo[a,h]
anthracene are weak agonists of human AR (hAR) in vitro. ,-DDE (the
major DDT metabolite found in human tissues), vinclozolin and procymidone (fungicides), linuron (a herbicide), fenitrothion (an insecticide),
and some PAHs inhibit human and/or animal AR-dependent transcriptional activity. Vinclozolin binds fairly weakly to AR, but two of its
metabolites (M1, M2) are stronger antiandrogens.
Other Mechanisms
• Sertoli cell toxicants—aphthalate metabolite (mono-(2-ethylhexyl)
phthalate) inhibits neonatal rat Sertoli cell division at submicromolar
208
CHILD HEALTH
AND THE
ENVIRONMENT
concentrations in vitro; the fungicide benomyl and its main metabolite
(carbendazim) interfere with microtubules and intermediate filaments
of Sertoli and germ cells, causing germ cell death and abnormal spermatid development.
• Steroid synthesis and catabolism
° The phthalates DEHP and di-n-butyl phthalate (DBP) reduce fetal and
neonatal T production, possibly by inhibiting expression of StAR and
genes encoding steroidogenic enzymes.
° TCDD reduces testicular levels of CYP11A, thereby reducing T production; it may also increase inactivation of T in liver and testicles.
° Tributyltin inhibits aromatase (reducing conversion of androgens to
estrogens) and reduces the conjugation and excretion of androgens.
° Altered AR expression—prenatal exposure of animals to TCDD or
DES causes decreased AR levels in testes, prostate, and accessory sex
glands.
Thyroid Modulators
Chemicals known to interfere with thyroid hormone function inhibit the
synthesis, transport, or catabolism of thyroid hormones. TCDD induces
the enzyme UDP-glucuronosyltransferase-1, triggering increased glucuronidation and excretion of T4, increased TSH levels, and thyroid hyperplasia. Hydroxylated metabolites of PCBs, dioxins, and furans bind
strongly to human transthyretin, the only thyroid hormone-binding protein synthesized in the brain, with affinities up to several times those of
T4; by binding to transthyretin in brain, some hydroxylated PCBs, dioxins, and furans may alter brain free T4 levels and interfere with brain development and function. Neonatal cord plasma free T4 levels are inversely
associated with cord plasma hydroxylated PCB levels but not with individual or total PCBs (Sandau et al., 2002).
Health Effects
Multiple male reproductive system abnormalities may comprise a testicular dysgenesis syndrome (Boisen et al., 2001). The syndrome of male genital tract birth defects, testicular cancer, poor semen quality, and subfertility may all involve disruption by HAAs of embryonic programming
and fetal gonadal development in genetically susceptible individuals.
There is considerable animal evidence to support this hypothesis and
some evidence from the few directly relevant epidemiologic studies.
Hormonally Active Agents
209
Developmental Effects
Animal studies indicate that prenatal exposure to environmental HAAs can
cause abnormal development of the male and female reproductive tracts
through mechanisms including germ cell toxicity and steroid hormone receptor modulation. Developmental exposure of animals to germ cell toxicants generally reduces fertility without causing malformations of external
genitalia and accessory glands. In contrast, prenatal exposure to antiandrogenic agents causes male reproductive tract malformations (hypospadias,
cryptorchidism, vaginal pouches, agenesis of the ventral prostate, and nipple retention) while generally having less effect on the testes. There is a
marked dearth of epidemiologic studies on the potential role of environmental HAAs in reproductive tract developmental abnormalities in humans.
Reproductive Tract Birth Defects
Final testicular descent in humans appears to require a T surge just before birth, suggesting that antiandrogens could cause cryptorchidism.
Only a few studies of cryptorchidism have assessed environmental exposures, providing limited evidence of links to prenatal parental pesticide exposure (Hosie et al., 2000; Kristensen et al., 1997; Weidner et al.,
1998), increased free E2 and decreased T levels in maternal blood during
early gestation (Bernstein et al., 1988; Key et al., 1996), maternal smoking
(Akre et al., 1999; McBride et al., 1991), and maternal use of exogenous
estrogens during pregnancy (Depue, 1984). A small case-control study of
cryptorchidism in Germany showed associations with heptachlor and
hexachlorobenzene levels in fat samples (Hosie et al., 2000). In another of
the few epidemiologic studies to measure internal doses of environmental HAAs, there were weak, nonsignificant associations between elevated
maternal serum DDE levels and cryptorchidism and hypospadias, as well
as a borderline association with accessory nipples among male infants
(OR 1.9, CI 0.9–4.0) (Longnecker et al., 2002). This study took advantage of stored, frozen maternal serum samples collected during the Collaborative Perinatal Project (1959–1966), a time when human serum
DDT/DDE levels were much higher than they are currently.
Hypospadias has been associated with low birth weight (Akre et al.,
1999; Weidner et al., 1999), proximity of residence to landfill sites (Dolk
et al., 1998), and prenatal DES exposure (Beral and Colwell, 1981). In the
first human evidence of a transgenerational effect of DES, the sons of
women exposed to DES in utero (i.e., the grandsons of women who took
DES during pregnancy) had a substantially increased risk of hypospadias
(RR 21, CI 6.5–70) (Klip et al., 2002). The daughters of women who consumed DES during pregnancy had increased risks of vaginal adenosis,
cervical ectropion, transverse cervical and vaginal ridges, and hypoplastic
210
CHILD HEALTH
AND THE
ENVIRONMENT
uterus (Swan, 2000). In addition, DES produced similar abnormalities in
prenatally exposed male and female rodents and monkeys.
Feminization of Males
Although maternal serum DDE has been linked to accessory nipples in
male infants (see above), most evidence linking HAA exposures to developmental feminization comes from animal studies. Developmental abnormalities among male rodents prenatally exposed to pesticidal AR antagonists (vinclozolin, procymidone, linuron, and DDT/DDE) include
feminization (reduced anogenital distance, retained nipples/areolas), reproductive tract birth defects (agenesis of sex accessory tissues, cryptorchidism, hypospadias), and sexual dysfunction (inability after puberty
to mate with sexually receptive females) (Gray et al., 2001). The timing of
exposure is critical; the fetal male rat is most susceptible on gestational
days 14 through 19.
The antiandrogenic effects of phthalates appear to relate to Sertoli
cell toxicity and inhibition of fetal T synthesis. Expert review panels concluded that prenatal exposure of experimental animals to high doses of
the plasticizers butylbenzyl phthalate (BBP), DEHP, DBP, diisononyl phthalate (DINP), or di-n-hexyl phthalate (DnHP) can cause feminization of
males, reproductive tract birth defects, and various abnormalities of testes
(reduced Sertoli cells, seminiferous tubular atrophy, reduced weight) and
accessory glands (reduced prostate weight) (Agency for Toxic Substances
and Disease Registry, 2000a). The sensitivity of testes to phthalates appears to be greatest prenatally, intermediate among juvenile animals, and
lowest among adults. As noted above, bisphenol A and nonylphenols have
estrogenic activity in in vitro systems. Low-dose prenatal exposure of male
mice to bisphenol A caused feminization, increased prostate size, and persistent increased prostate AR levels. There is mixed evidence of estrogenic
effects in male and female offspring after perinatal exposure to high-dose
nonylphenol.
Male rats exposed prenatally to single low doses of TCDD had reduced ventral prostate weight and reduced anogenital distance. Prenatal exposure of swine to TCDD caused male reproductive tract abnormalities including cryptorchidism and reduced testicular germ cell
populations. See Chapter 6 (PCBs, Dioxins, and Related Compounds)
for further discussion of developmental effects of TCDD and dioxin-like
chemicals.
Other Developmental Abnormalities
Prenatal exposure of rodents to DEHP at high doses caused eye, neural
tube, skeletal, and cardiac birth defects, but it is not known if these involved hormonal mechanisms; DEHP is a structural isomer of the anti-
Hormonally Active Agents
211
convulsant valproic acid that is known to cause similar birth defects in
humans. Female mice prenatally exposed to low-dose bisphenol A were
heavier and had earlier onset of puberty than controls. Female rats exposed prenatally to single low doses of TCDD had genital clefting with
hypospadias and vaginal threads, effects that can also be produced by
prenatal exposure to DES or E2. Rats prenatally exposed to diesel exhaust
or filtered diesel exhaust had increased anogenital distance (both sexes)
and reduced ovarian primary follicles and Sertoli cells, possibly from inhibition of CYP19 activity and synthesis, causing reduced conversion of
T to E2 and accumulation of T (Watanabe and Kurita, 2001).
Reproductive Effects
The average age at menarche may have decreased during recent decades
in the United States and other countries (see, e.g., de Muinck Keizer and
Mul, 2001). This apparent trend may reflect improved nutrition and other
factors; the possible role of environmental HAAs remains undefined. Premature breast development among young Puerto Rican girls was associated with serum levels of phthalates (especially DEHP) and their metabolites (Colon et al., 2000); this potentially important finding needs
verification. Breast-fed daughters of women with high serum polybrominated biphenyl (PBB) levels had earlier menarche than daughters of
women with low serum PBB levels or formula-fed girls (Blanck et al.,
2000). Daughters of women who consumed DES during pregnancy had
increased risks of menstrual irregularity, infertility, ectopic pregnancy,
spontaneous abortion, and preterm delivery (Goldberg and Falcone,
1999). Epidemiologic studies that measured internal doses of environmental contaminants and sperm quality have shown inverse associations
between PCBs and DDE in serum or semen specimens and sperm concentration, motility, and morphology (Bush et al., 1986; Hauser et al.,
2002). The EPA reviewed 63 pesticides and identified 8 that have effects
on reproduction (see Chapter 7 for further discussion of reproductive effects of pesticides).
Endocrine System
There is good evidence in experimental animals and limited evidence in
humans and wildlife that PCBs and other dioxin-like compounds inhibit
thyroid function. Polychlorinated biphenyl or total TCDD toxic equivalent (TEQ) levels in breast milk, maternal blood, or cord blood were generally associated with reduced plasma thyroid hormone and elevated TSH
212
CHILD HEALTH
AND THE
ENVIRONMENT
levels in mothers during late pregnancy, in neonates, and in older children (see, e.g., Osius et al., 1999). Because untreated congenital hypothyroidism causes severe cognitive deficits and because the severity of maternal hypothyroidism during pregnancy has been linked to cognitive
deficits at age 8 years, modestly reduced thyroid hormone levels in the
mother and/or the fetus may have adverse neurobehavioral effects on the
developing child. Several brominated fire retardants (tetrabromobisphenol A, pentabromophenol, and hydroxylated metabolites of polybrominated diphenyl ethers) have high affinity for human transthyretin in vitro.
Occupational and environmental TCDD exposures have been linked to
diabetes in adults, but no studies of children have been reported. In experimental animals, TCDD causes profound reductions in glucose transport into tissues (Kern et al., 2002).
Immune System
TCDD and many other organochlorine chemicals are immunotoxic in animals, but there have been few studies of their potential effects in humans.
Evidence of the potential for HAAs to interfere with immune system development and function includes the following:
• Genes regulated by nuclear thyroid hormone receptors appear to determine the B-cell pool size in mice; hypothyroid mice have reduced
numbers of pro-B, pre-B, and B cells in bone marrow, an effect reversible
by treatment with T4.
• E2 reduces T cell–dependent immune function but enhances B-cell antibody production; androgens suppress both T-cell and B-cell immune
responses.
• Prenatal DES exposure in rodents causes persistent deficits in T-cell, Bcell, and natural killer cell function, as well as increased susceptibility
to transplanted and carcinogen-induced cancers.
• A systematic examination of low-dose effects of selected HAAs in mammalian species by peer-review panels concluded that exposure (prenatal to puberty) to genistein, methoxychlor, and nonylphenol has definite low-dose immunologic effects including (Melnick et al., 2002)
° Genistein—increased splenic T-lymphocyte proliferation after antiCD3 stimulation
° Methoxychlor—decreased percentage of CD4/CD8 thymocytes
and decreased antibody plaque-forming cell response in males
° Nonylphenol—increased relative thymus weight and splenic
T-lymphocyte proliferation after anti-CD3 stimulation
Hormonally Active Agents
213
Cancer
Discovery of clear cell adenocarcinoma of the vagina and cervix among
teenage girls and young adult women exposed in utero to DES provided
the first clear evidence of adverse effects from prenatal exposure to a synthetic estrogen in humans (Herbst et al., 1971). Exposed women had increased risks of clear cell cervicovaginal cancer and breast cancer; exposed
men had a marginally increased risk of testicular cancer (Strohsnitter et
al., 2001). The mechanism by which DES causes vaginal cancer is not
known but may involve activation of ER␣, genotoxic effects, or both. Diethylstilbestrol and its metabolites bind to DNA and tubulin, increasing
the rate of aneuploidy and nondisjunction in dividing cells.
Incidence rates of testicular cancer, the most common cancer among
young men in most developed countries, are increasing in many countries; for example, Canadian men born during 1959–1968 are twice
as likely to develop testicular cancer as those born during 1904–1913
(Liu et al., 1999). Risk factors for testicular cancer include persistent
cryptorchidism, early birth order, low birth weight, prenatal exposure to
exogenous estrogen, and delayed puberty (protective). Correction of
cryptorchidism early in life does not reduce the risk of testicular cancer,
and men with unilateral cryptorchidism have an increased risk of contralateral testicular cancer, suggesting that these conditions have a common cause. Given the strong animal evidence that environmental AR antagonists can cause cryptorchidism, it is possible that some human
testicular cancers may be attributable to such agents.
Thyroid cancer incidence rates, especially papillary thyroid cancers,
increased markedly in the United States and Canada during the late twentieth century (see, e.g., Liu et al., 2001). Birth cohort analysis of the
Connecticut data indicate that the increase occurred among persons born
during 1915–1945, a period when radiation was used to treat benign conditions of the head and neck region of children (see Chapter 9, Radiation).
Pooled analyses of thyroid cancer studies in several countries showed associations with current oral contraceptive use (especially with papillary
thyroid carcinomas), a history of exposure to fertility drugs, and lactation
suppression treatment but not with postmenopausal hormone replacement
treatment or fish consumption (Bosetti et al., 2001; La Vecchia et al., 1999).
No epidemiologic studies have been done on prenatal exposures to
environmental HAAs and cancer risks later in life. Epidemiologic studies of hormone-sensitive cancers (breast, ovarian, endometrial, prostate,
testicular) and blood or tissue concentrations of HAAs during adulthood
(e.g., PCB, DDT, and DDE) showed inconsistent evidence of associations.
CHILD HEALTH
214
AND THE
ENVIRONMENT
Exposures
Data from NHANES III on urinary phthalate metabolite levels in representative samples of the U.S. population are the only population-based
phthalate exposure measurements for any country (Table 8–5). Metabolites of diethylphthalate (DEP) and DBP occurred at highest concentrations; this result was surprising, as these phthalates are produced in much
lower quantities than DEHP and DINP. Urinary DBP and DEHP metabolite levels were higher among low-income groups (Koo et al., 2002). Reproductive-age women had average urinary DBP metabolite levels 50%
higher than those of other subgroups (Blount et al., 2000). Diethylphthalate and DBP are used extensively in products with volatile components
such as perfumes, nail polishes, and hair sprays. The CDC has initiated
additional studies to identify the sources of exposure to DEP and DBP.
The highest lipid-adjusted PCB concentrations occur in placenta, where
they can exceed 5 g/g fat or almost three times those in breast milk.
Organochlorines including ␣-hexachlorocyclohexane, ,-DDE, and PCBs
were detected in one-third of human amniotic fluid samples in Los Angeles (Foster et al., 2000). The median concentration of ,-DDE in human
serum, on a molar basis, is about 100-fold greater than that of E2 in premenopausal women and about the same level as T in men. Tris(4-chlorophenyl)methane (TCPM) and TCPM-OH were detected at levels of about
1–20 ng/g lipid in all human adipose samples tested in Japan. See also
Chapters 6 (PCBs, Dioxins, and Related Compounds) and 7 (Pesticides) for
further details on population exposure levels of potential HAAs.
Infants raised on soy-based formula ingest about 5–8 mg/kg/day of
total isoflavones and have average plasma isoflavone levels of about
TABLE 8–5. Urinary Phthalate Metabolites, United States
50th Percentile
Urinary Concentration
(g/g creatinine)a
50th Percentile
Urinary Concentration
(g/g creatinine)b
Mono-ethyl (DEP)
280
134
Mono-butyl (DBP)
33
22
Mono-benzyl (BBP)
20
14
3
3
Urinary Phthalate Metabolite
(parent phthalate)
Mono-2-ethylhexyl (DEHP)
Mono-isononyl (DINP)
LOD
LOD
a U.S.
population aged 20–60 years, NHANES III, 1988–1994 (Blount et al., 2000).
b U.S.
population aged 6 years and older, 1999 (Centers for Disease Control and Prevention, 2001).
LOD limit of detection.
Hormonally Active Agents
215
1 mg/L, that is, more than three orders of magnitude higher than E2 concentrations in early life (Setchell et al., 1998). Average plasma levels of
isoflavones among infants fed breast or cow’s milk are about 5 g/L.
Isoflavone exposures among formula-fed infants are six- to elevenfold
higher (on a body weight basis) than those that have hormonal effects on
adults consuming soy foods, that is, such exposures may be sufficient to
produce biologic effects in infants (Setchell et al., 1997).
Risk Management
The 1996 FQPA and the Safe Drinking Water Act require screening tests
for HAAs in food and source drinking water. An EPA advisory committee recommended evaluation of pesticides, commercial chemicals, and environmental contaminants for effects on estrogen, androgen, and thyroid
hormone function. The EPA must prioritize more than 87,000 chemicals
including 75,500 in the Toxic Substances Control Act Inventory, 900 active
pesticide ingredients, 2500 pesticide inerts, and 8000 cosmetic ingredients,
food additives, and nutritional supplements for endocrine disruption
screening and testing. The EPA has also committed to test six mixtures
representative of contaminants in human breast milk, phytoestrogens in
soy-based infant products, chemicals commonly found in hazardous
waste sites, pesticide/fertilizer mixtures, drinking water disinfection byproducts, and gasoline (U.S. Environmental Protection Agency, 1998).
The EPA adopted a tiered approach that will prioritize chemicals for
evaluation based on existing information on exposure and hormonal activity (U.S. Environmental Protection Agency, 2000a). Evaluation will require the validation of existing in vitro and animal tests and the development of new methods. The EPA decided to use a two-tiered approach
and is collaborating with the Office of Economic Cooperation and Development and the National Toxicology Program to standardize and validate Tier 1 and 2 bioassays of effects on male and female reproductive
systems and behaviors and thyroid function.
The European Commission in 1999 decided to address HAAs by prioritizing substances for testing when appropriate test methods become
available, to identify HAAs covered by existing European Community
legislation, to expand epidemiologic studies of cause–effect relations, and
to consider policies for vulnerable groups including children. The Commission identified 564 HAAs with various levels of evidence, among
which 146 were high production and/or highly persistent chemicals (BKH
Consulting Engineers and TNO Nutrition and Food Research, 2000). Information on production and use of selected HAAs is shown in Table 8–6.
TABLE 8–6. Production and Use of Selected HAAs
Chemical
Uses (Past and/or Present)a
Production
Major Source of Human Exposurea
PCBs
Dielectrics, flame retardants,
hydraulic fluids, electrical
equipment, pigments, others
Banned in USA in 1977;
annual production in USA
42,500 tons (1970), 2 million
tons produced globally since
the 1930sb
Food (especially fatty foods), breast milk
DDT
Insecticide
Banned in the USA in 1973; peak
annual production was 90,000
tons (USA, 1962); estimated
cumulative global production
about 2 million tonsc
Food (especially fatty foods), breast milk
Bisphenol A
Monomer for production of
polycarbonate polymers,
epoxy resins, dyes, flame
retardants, dental sealants
2.6 million tons, 1999
(USA, South America,
Europe, Asia)d
Food (especially fatty foods)
Nonylphenol (accounts for
about 85% of alkylphenol
production)
Surfactants—nonylphenol ethoxylates
used in surfactants for industrial
and institutional formulations,
household liquid detergents; used
in antioxidants such as
tris(nonylphenol)phosphite
120,000 tons (USA, 2000)e
Water, food
Diethylphthalate (DEP)—
monoester metabolite had
highest urinary level in a
CDC study
Toothbrushes, automobile parts, tools,
toys, food packaging, cosmetics,
insecticides, aspirin
USA—production 13,000
tons (1988) f
Food in plastic packaging, contaminated
fish and shellfish, contaminated water
near waste sites and landfills, certain
consumer products
Di-n-butyl phthalate (DBP)—
monoester metabolite had
second highest urinary level
in a CDC study
Plastics, paints, glue, hair spray, and
other chemical products
USA—production 13,000
tons (1988)g
Fish, shellfish, cosmetic use
Di(2-ethylhexyl) phthalate
(DEHP)
Polyvinyl chloride (PVC) plastics
may contain up to 40% DEHP—
toys, vinyl upholstery, adhesives,
coatings; DEHP also used in inks,
pesticides, cosmetics, and vacuum
pump oil
USA—one U.S. company
produced 143,000 tons of
DEHP and ocytlphthalates
in 1998h
Medical products packaged in plastic
(e.g., blood products, dialysis fluids),
foods packaged in plastics (especially
fatty foods like milk products, fish/
seafood, oils), groundwater near waste
sites, workplace or indoor air where
DEHP is released from plastic
materials, coatings, flooring
aAgency
for Toxic Substances and Disease Registry (2001a).
bNational
cAgency
Academy of Sciences (1999).
for Toxic Substances and Disease Registry (2000b).
dBisphenol
eChemical
A global industry group (2002).
Market Reporter (2001).
fAgency
for Toxic Substances and Disease Registry (1995).
gAgency
for Toxic Substances and Disease Registry (2001b).
hAgency
for Toxic Substances and Disease Registry (2000c).
218
CHILD HEALTH
AND THE
ENVIRONMENT
Water
Nonylphenol ethoxylates, a subgroup of nonionic surfactants known as
alkylphenol ethoxylates, are high-volume chemicals (over 300,000 tons per
year produced globally). These chemicals have been used for over 40 years
as cleaners, degreasers, detergents, and emulsifiers in domestic and industrial products (textiles, pulp and paper, paints, resins and protective
coatings, pesticides, and cosmetics). Microbial biodegradation of the parent compounds during sewage treatment releases estrogenic alkylphenols, primarily nonylphenol and octylphenol.
Relatively high amounts of nonylphenol ethoxylates and their degradation products (particularly nonylphenol, nonylphenol ethoxylate, and
nonylphenol diethoxylate) occur in municipal wastewater treatment plant
effluents, often at levels exceeding 1 mg/L. Nonylphenol occurs at high
concentrations in some sewage sludges applied to agricultural lands and
is relatively persistent in groundwater. Although nonylphenol ethoxylate
has been detected in human urine, there are no population-based data on
the prevalence of exposure. Other HAAs potentially present in water include the natural hormones E2 and estrone and the synthetic hormone
ethynyl estradiol-17␣ (from municipal wastewater discharges) and organotin compounds (may leach from PVC pipes).
Food
Foods are potential sources of exposure to certain HAAs including phytoestrogens, phthalates, dioxins, PCBs, certain pesticides, and organotin
compounds. The synthetic estrogen zeranol, used as a growth promoter
in beef production, has potency similar to that of DES and E2 in inducing estrogen-dependent gene expression in human MCF7 cells in vitro.
Although zeranol is not an environmental contaminant, its potency shows
the need for integrated risk management of HAAs.
Phytoestrogens
Soy protein is relatively cheap, and about 60% of processed foods contains soy derivatives. Although diets high in phytoestrogens are associated with reduced risks of certain cancers among adults, it is not known
whether exposure during infancy and early childhood has any adverse
effects.
Phthalates
Global phthalate production appears to be about 5–6 million tons per year.
Phthalates are used to make flexible polyvinyl chloride products with
smaller amounts being used as components in consumer products includ-
Hormonally Active Agents
219
ing soaps, lotions, perfumes, insect repellants, and other products applied
to skin. Phthalates can migrate from polyvinyl chloride food packaging in
contact with foods; such migration is increased by heat and fat content.
Foods wrapped in plastic film contained di(2-ethylhexyl) adipate (DEHA)
at levels up to 310 mg/kg (cheese). Di(2-ethylhexyl) phthalate from plastic
tubing used in commercial milking equipment can leach into milk. Dairy
products from the United Kingdom contained average total phthalate levels of 0.06–0.32 mg/kg in milk, 1.8–19 mg/kg in cream, 4.8–56 mg/kg in
butter, and 2.4–114 mg/kg in various types of cheese; similar levels occurred in Norway and Spain. After Denmark banned the use of DEHPplasticized milk tubing, average DEHP levels in whole milk fell to less than
50 g/L within 6 months.
Average adult human daily phthalate intakes, mainly from foods, are
generally far lower than the estimated NOAELs in rodents. Infants and
toddlers, however, may be exposed to relatively high DEHP levels from
mouthing plastic toys and objects (National Institute of Environmental
Health Sciences, 2000). During the 1980s, toy manufacturers voluntarily
phased out the use of DEHP in soft toys for children; DINP became the
predominant phthalate used to soften the polyvinyl chloride used in some
children’s products until its use was voluntarily restricted in 1998 because
of evidence that it causes liver cancer in rodents. Although the U.S. Consumer Produce Safety Commission (CPSC) concluded that few if any children are at risk from DINP because of low absorbed doses, it asked industry to remove phthalates from soft rattles and teethers, and most
manufacturers agreed to comply by early 1999 (U.S. Consumer Product
Safety Commission, 1998). The CPSC also asked industry to find a substitute for phthalates in other products intended for children aged 3 years
or less that are likely to be mouthed or chewed. Canada and the European Union (EU) have acted to stop using DEHP and DINP and to limit
the use of certain other phthalates (e.g., di-n-octyl phthalate) in toys intended for mouthing (nipples, teethers, pacifiers, and rattles) but not in
larger toys designed for older children.
Persons receiving intensive treatment including parenteral fluids
(e.g., premature infants or very ill children) may have high DEHP exposures (because of leaching from plastic medical devices). United States
and Canadian expert panels have identified serious health concerns about
DEHP (Health Canada, 2002; National Institute of Environmental Health
Sciences, 2000). The Canadian panel concluded that DEHP exposure from
medical devices in infants, toddlers, critically ill children, and pregnant
and lacatating women may cause reproductive and developmental risks
and recommended that alternatives to DEHP in medical devices be
assured. Whereas most previous research and regulatory attention to
220
CHILD HEALTH
AND THE
ENVIRONMENT
phthalates has focused on DEHP and DINP, the results of the CDC survey of urinary phthalate levels noted above indicate the need to control
exposures to DEP, DBP, and other phthalates (Blount et al., 2000).
Other Sources
Bisphenol A is a monomer precursor used in production of polycarbonate plastic (used in baby bottles, household appliances, food and drink
containers, and many other products) and epoxy resins. Total bisphenol
A production in the United States, Western Europe, and Japan was about
2 million tons in 1999, with annual growth of about 7% during the period
1996–1999. Cans for preserved foods are often lined with organosol
(polyvinylchloride type) or epoxy lacquers. Bisphenol A diglycidyl ether
(BADGE) is used in organosol lacquers, and it and related derivatives can
migrate into canned foods, especially fatty foods, at levels of up to several milligrams per kilogram, exceeding the limit of 1 mg/kg proposed
by the European Commission Scientific Committee for Food. Bisphenol
A has been found in the liquid phase of preserved vegetables in lacquercoated cans at levels capable of inducing proliferation of MCF-7 human
breast cancer cells in vitro. Tris(nonylphenyl) phosphite is used in some
plastic food wrap products and can release nonylphenol into foods.
Other Products
Pesticides
Many pesticides are applied after harvesting to fruits and vegetables to
extend their lives and preserve their quality during storage, transport,
and marketing. Persistence and distribution of pesticide residues in the
edible portions of produce have been reported for foods, including fruits
and vegetables, commonly eaten by children and pregnant women. Vinclozolin is registered as a fungicide for use on fruits and vegetables in the
United States and Europe; its metabolites include two potent antiandrogens (see above) and 3,5-dichloroaniline, a genotoxin that is structurally
similar to a known animal carcinogen (parachloroaniline). The EPA was
concerned about ground and surface water 3,5-dichloroaniline levels and
about the fact that this metabolite is also formed from the related dicarboximide fungicides, iprodione and procymidone. Under the FQPA, the
EPA has assessed the cumulative hazard of these fungicides, which have
a common antiandrogenic toxic mechanism. In what appears to be the
first pesticide regulatory action arising in part from concern about hormonal activity, phased restrictions on some food crop and ornamental
uses of the fungicide vinclozolin in the United States were implemented
because of concerns about its antiandrogenic and carcinogenic effects (U.S.
Hormonally Active Agents
221
Environmental Protection Agency, 2000b). After 2004, predicted annual
use of this pesticide on canola, nondomestic wine grapes, and turf will
be reduced by about half from the current 141,000 pounds.
Others
Brominated flame-retardants, including polybrominated diphenyl ethers
(PBDEs), pentabromophenol, and tetrabromobisphenol A, are used in
large quantities in electronic equipment, plastics, fabrics, and building
materials. Tributyltin bioaccumulates in crabs, oysters, and salmon, particularly those grown in captivity and exposed to organotin used as an
antifouling agent. In Japan and the United Kingdom, organotin levels in
marine products decreased after the introduction of legal controls during
1987-1990. Several sunscreens had estrogen agonist activity in bioassays
(Schlumpf et al., 2001).
Diethylstilbestrol was used as a pregnancy medication in North
America and several European countries from the 1940s to the 1970s in
the mistaken belief that it would prevent miscarriage. It was also used to
prevent lactation in women who wished to bottle-feed, as a postcoital contraceptive, and as a growth promoter for livestock. In the United States
alone, at least 3–5 million pregnant women were given DES and related
synthetic estrogens, which are now known to be carcinogens and teratogens. Although DES is not an environmental HAA, the experience with
it illustrates the potential for harm in using therapeutic HAAs without
proven efficacy and the value of modern regulatory practices.
General Considerations
Although there is strong evidence of the adverse developmental and reproductive effects of environmental HAAs at high doses in animals, few
studies have been done in humans and the results to date are largely inconclusive. The dearth of human evidence should not be interpreted as
evidence of no risk. Rather, the large-scale epidemiologic studies with
good exposure and outcome assessment needed to detect the potential
risks of multiple low-level environmental HAA exposures in humans have
not been done. The planned U.S. longitudinal study of children is the type
of study needed to fill this important knowledge gap (National Institute
of Child Health and Human Development, 2001).
In assessing animal or epidemiologic studies of environmental HAAs
and developmental and reproductive outcomes, it is important to consider that (1) these processes can be disrupted not only by hormonal mechanisms but also by toxic effects of the same agents, (2) results from highdose testing of single agents in animal systems must be interpreted with
222
CHILD HEALTH
AND THE
ENVIRONMENT
caution because human exposures are usually much lower and there are
many other uncertainties in extrapolating from animals to humans, and
(3) humans are exposed to multiple low-level HAAs and other toxicants
during critical prenatal and childhood development periods.
Conclusions
Proven Health Outcomes
• Although DES is not an environmental HAA, past experience has
shown that prenatal exposure to this potent synthetic estrogen causes
reproductive tract abnormalities (hypospadias, hypoplastic testicles,
epididymal cysts) and vaginal cancer in offspring.
Unresolved Issues and Knowledge Gaps
• Developmental effects—despite considerable animal evidence, the potential roles of prenatal parental exposure to environmental HAAs in
human fetal deaths, growth deficits, and birth defects are unknown;
for example, only two epidemiologic studies of cryptorchidism have
measured internal doses of environmental HAAs, and none have measured biomarkers of susceptibility to HAAs.
• Reproductive system development and function
° There is limited, inconclusive evidence that the average age at menarche has decreased in several countries during recent decades; the validity of these trends and the potential roles of environmental HAAs
and other factors remain undefined.
° Limited evidence suggests that the ability of the pesticide chlordecone to cause severely reduced sperm concentrations among occupationally exposed men may involve a hormonal mechanism, that is,
ER agonist activity.
° Some pesticides have antiandrogenic activity in animals and in in
vitro bioassays, but their role in human subfertility is unknown.
° Internal HAA doses have rarely been measured in studies of human
sperm quality.
• Cancer
° Testicular cancer—incidence rates have increased substantially in
several countries during recent decades; definitive epidemiologic
studies of the possible role of HAAs have not been done.
° Thyroid cancer—incidence rates have increased in several countries
during recent decades, especially among young women; the potential role of environmental HAAs remains unknown.
Hormonally Active Agents
223
• Knowledge gaps
° Longitudinal studies of children beginning in the first trimester of
pregnancy (or earlier) are needed to assess the risk of subtle and delayed health effects of HAAs and other contaminants with known or
suspected developmental toxicity.
° It is necessary to identify the exposure sources of the phthalates DEP
and DBP in reproductive-age women and to assess potential developmental and child health effects (see “Biomonitoring” below).
Risk Management Issues
• Prevention
° Toxicity testing—ongoing efforts to implement tiered toxicity testing
for hormonal activity on a prioritized set of chemicals to which
reproductive-age women and children are most exposed should be
encouraged.
° Exposure reduction—it is necessary to minimize exposures of reproductive-age women, pregnant women, and young children to environmental HAAs identified in screening tests.
• Biomonitoring—NHANES III detected relatively high urinary levels of
metabolites of the phthalates DEP and DBP in a representative sample
of the U.S. population, with reproductive-age women having average
DBP metabolite levels about 50% higher than those of other age/sex
groups; it is necessary to initiate biomonitoring for HAAs in other countries, with a focus on reproductive-age women and children.
• Disease tracking—the limited, inconclusive evidence of increasing incidence rates of hypospadias and cryptorchidism and increasing prevalence of reduced sperm quality shows the need for tracking systems to
enable population-based monitoring of these conditions.
References
Agency for Toxic Substances and Disease Registry. (1995). Toxicological profile for
diethyl phthalate. Atlanta: Agency for Toxic Substances and Disease Registry.
Agency for Toxic Substances and Disease Registry. (2000a). Final phthalate expert
panel reports (Center for the Evaluation of Risks to Human Reproduction).
Located at http://cerhr.niehs.nih.gov/news/index.html
Agency for Toxic Substances and Disease Registry. (2000b). Toxicological profile
for DDT, DDE, DDD. Atlanta: Agency for Toxic Substances and Disease
Registry.
Agency for Toxic Substances and Disease Registry. (2000c). Toxicological profile
for di(2-ethylhexyl)phthalate. Atlanta: Agency for Toxic Substances and Disease Registry.
224
CHILD HEALTH
AND THE
ENVIRONMENT
Agency for Toxic Substances and Disease Registry. (2001a). ToxFAQS. Located at
http://atsdr1.atsdr.cdc.gov/toxfaq.html
Agency for Toxic Substances and Disease Registry. (2001b). Toxicological profile for
di-n-butyl phthalate. Atlanta: Agency for Toxic Substances and Disease Registry.
Aho M, Koivisto AM, Tammela TL, Auvinen A. (2000). Is the incidence of hypospadias increasing? Analysis of Finnish hospital discharge data 1970–1994.
Environ Health Perspect 108:463–5.
Akre O, Lipworth L, Cnattingius S, Sparen P, Ekbom A. (1999). Risk factor patterns for cryptorchidism and hypospadias. Epidemiology 10:364–9.
Andersen HR, Andersson AM, Arnold SF, Autrup H, Barfoed M, Beresford NA,
Bjerregaard P, Christiansen LB, Gissel B, Hummel R, and others. (1999). Comparison of short-term estrogenicity tests for identification of hormonedisrupting chemicals. Environ Health Perspect 107(Suppl 1):89–108.
Beral V, Colwell L. (1981). Randomised trial of high doses of stilboestrol and ethisterone therapy in pregnancy: long-term follow-up of the children. J Epidemiol
Commun Health 35:155–60.
Bernstein L, Pike MC, Depue RH, Ross RK, Moore JW, Henderson BE. (1988). Maternal hormone levels in early gestation of cryptorchid males: a case-control
study. Br J Cancer 58:379–81.
Bisphenol A global industry group. (2002). Bisphenol A home page. Located at
http://www.bisphenola.org/
BKH Consulting Engineers and TNO Nutrition and Food Research (for the
European Commission DG ENV). (2000). Towards the establishment of a priority list of substances for further evaluation of their role in endocrine
disruption—preparation of a candidate list of substances as a basis for priority setting. (M0355008/1786Q/10/11/00). Located at http://europa.eu.int/
comm/environment/docum/bkh_main.pdf.
Blanck HM, Marcus M, Tolbert PE, Rubin C, Henderson AK, Hertzberg VS, Zhang
RH, Cameron L. (2000). Age at menarche and tanner stage in girls exposed in
utero and postnatally to polybrominated biphenyl. Epidemiology 11:641–7.
Block K, Kardana A, Igarashi P, Taylor HS. (2000). In utero diethylstilbestrol (DES)
exposure alters Hox gene expression in the developing mullerian system.
FASEB J 14:1101–8.
Blount BC, Silva MJ, Caudill SP, Needham LL, Pirkle JL, Sampson EJ, Lucier GW,
Jackson RJ, Brock JW. (2000). Levels of seven urinary phthalate metabolites in
a human reference population. Environ Health Perspect 108:972–82.
Boisen KA, Main KM, Rajpert-De Meyts E, Skakkebaek NE. (2001). Are male reproductive disorders a common entity? The testicular dysgenesis syndrome.
Ann NY Acad Sci 948:90–9.
Bosetti C, Kolonel L, Negri E, Ron E, Franceschi S, Dal Maso L, Galanti MR, Mark
SD, Preston-Martin S, McTiernan A, and others. (2001). A pooled analysis of
case-control studies of thyroid cancer. VI. Fish and shellfish consumption. Cancer Causes Control 12:375–82.
Burlington H, Lindeman VF. (1950). Effect of DDT on testes and secondary sex
chacters of white leghorn cockerels. Proc Soc Exp Biol Med 74:48–51.
Bush B, Bennett AH, Snow JT. (1986). Polychlorobiphenyl congeners, p,p-DDE,
and sperm function in humans. Arch Environ Contam Toxicol 15:333–41.
Carlsen E, Giwercman A, Keiding N, Skakkebaek NE. (1992). Evidence for decreasing quality of semen during past 50 years. BMJ 305:609–13.
Centers for Disease Control and Prevention. (2001). National report on human ex-
Hormonally Active Agents
225
posure to environmental chemicals. Atlanta: Centers for Disease Control and
Prevention.
Chemical Market Reporter. (2001). Chemical Market Reporter. Located at
http://www.aperc.org/nonylphenol.pdf
Colon I, Caro D, Bourdony CJ, Rosario O. (2000). Identification of phthalate esters in the serum of young Puerto Rican girls with premature breast development. Environ Health Perspect 108:895–900.
Darlington DN, Dallman MF. (2001). Feedback control in endocrine systems. Principles and practice of endocrinology and metabolism. New York: Lippincott
Williams & Wilkins.
de Muinck Keizer SM, Mul D. (2001). Trends in pubertal development in Europe.
Hum Reprod Update 7:287–91.
Depue RH. (1984). Maternal and gestational factors affecting the risk of cryptorchidism and inguinal hernia. Int J Epidemiol 13:311–8.
Dolk H, Vrijheid M, Armstrong B, Abramsky L, Bianchi F, Garne E, Nelen V, Robert
E, Scott JE, Stone D, and others. (1998). Risk of congenital anomalies near
hazardous-waste landfill sites in Europe: the EUROHAZCON study. Lancet
352:423–7.
Foster W, Chan S, Platt L, Hughes C. (2000). Detection of endocrine disrupting
chemicals in samples of second trimester human amniotic fluid. J Clin Endocrinol Metab 85:2954–7.
Foster WG. (1998). Endocrine disruptors and development of the reproductive
system in the fetus and children: is there cause for concern? Can J Public Health
89(Suppl 1):S37–41, S52, S41–6.
Goldberg JM, Falcone T. (1999). Effect of diethylstilbestrol on reproductive function. Fertil Steril 72:1–7.
Gray LE, Ostby J, Furr J, Wolf CJ, Lambright C, Parks L, Veeramachaneni DN,
Wilson V, Price M, Hotchkiss A, and others. (2001). Effects of environmental
antiandrogens on reproductive development in experimental animals. Hum
Reprod Update 7:248–64.
Hauser R, Altshul L, Chen Z, Ryan L, Overstreet J, Schiff I, Christiani DC. (2002).
Environmental organochlorines and semen quality: results of a pilot study. Environ Health Perspect 110:229–33.
Health Canada. 2002. Health Canada Expert advisory panel on DEHP in medical
devices. Located at http://www.hc-sc.gc.ca/hpb-dgps/therapeut/zfiles/english/
advcomm/eap/dehp/eap-dehp-final-report-2002-jan-11_e.pdf.
Herbst AL, Ulfelder H, Poskanzer DC. (1971). Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy with tumor appearance in young
women. N Engl J Med 284:878–81.
Hosie S, Loff S, Witt K, Niessen K, Waag KL. (2000). Is there a correlation between
organochlorine compounds and undescended testes? Eur J Pediatr Surg 10:
304–9.
Jorgensen N, Andersen AG, Eustache F, Irvine DS, Suominen J, Petersen JH, Andersen AN, Auger J, Cawood EH, Horte A, and others. (2001). Regional differences in semen quality in Europe. Hum Reprod 16:1012–9.
Kallen B, Bertollini R, Castilla E, Czeizel A, Knudsen LB, Martinez-Frias ML, Mastroiacovo P, Mutchinick O. (1986). A joint international study on the epidemiology of hypospadias. Acta Paediatr Scand Suppl 324:1–52.
Kern PA, Dicker-Brown A, Said ST, Kennedy R, Fonseca VA. (2002). The stimulation of tumor necrosis factor and inhibition of glucose transport and lipopro-
226
CHILD HEALTH
AND THE
ENVIRONMENT
tein lipase in adipose cells by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Metabolism
51:65–8.
Key TJ, Bull D, Ansell P, Brett AR, Clark GM, Moore JW, Chilvers CE, Pike MC.
(1996). A case-control study of cryptorchidism and maternal hormone concentrations in early pregnancy. Br J Cancer 73:698–701.
Klip H, Verloop J, van Gool JD, Koster ME, Burger CW, van Leeuwen FE. (2002).
Hypospadias in sons of women exposed to diethylstilbestrol in utero: a cohort
study. Lancet 359:1102–7.
Koo JW, Parham F, Kohn MC, Masten SA, Brock JW, Needham LL, Portier CJ.
(2002). The association between biomarker-based exposure estimates for phthalates and demographic factors in a human reference population. Environ
Health Perspect 110:405–10.
Kristensen P, Irgens LM, Andersen A, Bye AS, Sundheim L. (1997). Birth defects
among offspring of Norwegian farmers, 1967–1991. Epidemiology 8:537–44.
Kuiper GG, Lemmen JG, Carlsson B, Corton JC, Safe SH, van der Saag PT, van
der Burg B, Gustafsson JA. (1998). Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology 139:4252–63.
La Vecchia C, Ron E, Franceschi S, Dal Maso L, Mark SD, Chatenoud L, Braga C,
Preston-Martin S, McTiernan A, Kolonel L, and others. (1999). A pooled analysis of case-control studies of thyroid cancer. III. Oral contraceptives, menopausal replacement therapy and other female hormones. Cancer Causes Control 10:157–66.
Liu S, Semenciw R, Ugnat AM, Mao Y. (2001). Increasing thyroid cancer incidence
in Canada, 1970–1996: time trends and age-period-cohort effects. Br J Cancer
85:1335–9.
Liu S, Wen SW, Mao Y, Mery L, Rouleau J. (1999). Birth cohort effects underlying
the increasing testicular cancer incidence in Canada. Can J Public Health
90:176–80.
Longnecker MP, Klebanoff MA, Brock JW, Zhou H, Gray KA, Needham LL, Wilcox
AJ. (2002). Maternal serum level of 1,1-dichloro-2,2-bis(p-chlorophenyl)
ethylene and risk of cryptorchidism, hypospadias, and polythelia among male
offspring. Am J Epidemiol 155:313–22.
McBride ML, Van den Steen N, Lamb CW, Gallagher RP. (1991). Maternal and gestational factors in cryptorchidism. Int J Epidemiol 20:964–70.
McLachlan JA. (1997). Synergistic effect of environmental estrogens: report withdrawn. Science 277:462–3.
Melnick R, Lucier G, Wolfe M, Hall R, Stancel G, Prins G, Gallo M, Reuhl K, Ho
SM, Brown T, and others. (2002). Summary of the National Toxicology Program’s report of the endocrine disruptors low-dose peer review. Environ
Health Perspect 110:427–31.
National Academy of Sciences. (1999). Hormonally active agents in the environment. Washington, DC: National Academy Press.
National Institute of Child Health and Human Development. (2002). The National
Children’s Study. Located at http://nationalchildrensstudy.gov/about/
042002.cfm.
National Institute of Environmental Health Sciences. (2000). Expert panel review
of phthalates. Research Triangle Park. Located at http://cerhr.niehs.nih.gov.
Nikov GN, Hopkins NE, Boue S, Alworth WL. (2000). Interactions of dietary estrogens with human estrogen receptors and the effect on estrogen receptor–
Hormonally Active Agents
227
estrogen response element complex formation. Environ Health Perspect 108:
867–72.
Osius N, Karmaus W, Kruse H, Witten J. (1999). Exposure to polychlorinated
biphenyls and levels of thyroid hormones in children. Environ Health Perspect
107:843–9.
Paulozzi LJ. (1999). International trends in rates of hypospadias and cryptorchidism. Environ Health Perspect 107:297–302.
Paulozzi LJ, Erickson JD, Jackson RJ. (1997). Hypospadias trends in two U.S. surveillance systems. Pediatrics 100:831–4.
Power DA, Brown RS, Brock CS, Payne HA, Majeed A, Babb P. (2001). Trends in
testicular carcinoma in England and Wales, 1971–99. BJU Int 87:361–5.
Sandau CD, Ayotte P, Dewailly E, Duffe J, Norstrom RJ. (2002). Pentachlorophenol and hydroxylated polychlorinated biphenyl metabolites in umbilical cord
plasma of neonates from coastal populations in Quebec. Environ Health Perspect 110:411–7.
Schiestl RH, Aubrecht J, Yap WY, Kandikonda S, Sidhom S. (1997). Polychlorinated biphenyls and 2,3,7,8-tetrachlorodibenzopdioxin induce intrachromosomal recombination in vitro and in vivo. Cancer Res 57:4378–83.
Schlumpf M, Cotton B, Conscience M, Haller V, Steinmann B, Lichtensteiger W.
(2001). In vitro and in vivo estrogenicity of UV screens. Environ Health Perspect 109:239–44.
Setchell KD, Zimmer-Nechemias L, Cai J, Heubi JE. (1997). Exposure of infants to
phyto-oestrogens from soy-based infant formula. Lancet 350:23–7.
Setchell KD, Zimmer-Nechemias L, Cai J, Heubi JE. (1998). Isoflavone content of
infant formulas and the metabolic fate of these phytoestrogens in early life.
Am J Clin Nutr 68:1453S–61S.
Strohsnitter WC, Noller KL, Hoover RN, Robboy SJ, Palmer JR, Titus-Ernstoff L,
Kaufman RH, Adam E, Herbst AL, Hatch EE. (2001). Cancer risk in men exposed in utero to diethylstilbestrol. J Natl Cancer Inst 93:545–51.
Swan SH. (2000). Intrauterine exposure to diethylstilbestrol: long-term effects in
humans. APMIS 108:793–804.
Swan SH, Elkin EP. (1999). Declining semen quality: can the past inform the present? Bioessays 21:614–21.
U.S. Consumer Product Safety Commission. (1998). The risk of chronic toxicity
associated with exposure to diisononyl phthalate (DINP) in children’s products. Bethesda, MD: U.S. Consumer Product Safety Commission.
U.S. Environmental Protection Agency. (1998). Endocrine disruptor screening and
testing advisory committee (EDSTAC) final report. Located at http://www.epa.
gov/oscpmont/oscpendo/history/finalrpt.htm
U.S. Environmental Protection Agency. (2000a). Endocrine disruptor screening program report to Congress. Located at http://www.epa.gov/scipoly/oscpendo/
reporttocongress0800.pdf.
U.S. Environmental Protection Agency. (2000b). Reregistration eligibility decision.
Vinclozolin (EPA 738-R-00-023). Washington, DC: U.S. Environmental Protection Agency.
Vos JG, Dybing E, Greim HA, Ladefoged O, Lambre C, Tarazona JV, Brandt I,
Vethaak AD. (2000). Health effects of endocrine-disrupting chemicals on wildlife, with special reference to the European situation. Crit Rev Toxicol 30:
71–133.
228
CHILD HEALTH
AND THE
ENVIRONMENT
Watanabe N, Kurita M. (2001). The masculinization of the fetus during pregnancy
due to inhalation of diesel exhaust. Environ Health Perspect 109:111–9.
Weidner IS, Moller H, Jensen TK, Skakkebaek NE. (1998). Cryptorchidism and
hypospadias in sons of gardeners and farmers. Environ Health Perspect 106:
793–6.
Weidner IS, Moller H, Jensen TK, Skakkebaek NE. (1999). Risk factors for cryptorchidism and hypospadias. J Urol 161:1606–9.
Weir HK, Marrett LD, Moravan V. (1999). Trends in the incidence of testicular
germ cell cancer in Ontario by histologic subgroup, 1964–1996. CMAJ 160:
201–5.
Wrutniak-Cabello C, Casas F, Cabello G. (2001). Thyroid hormone action in mitochondria. J Mol Endocrinol 26:67–77.
9
Radiation
The electromagnetic spectrum ranges from ionizing radiation through ultraviolet light, visible light, infrared light, microwaves, and radio waves
(Figure 9–1). Ionizing radiation comprises very short wavelength/high
frequency electromagnetic waves and particles with sufficient energy to
remove electrons from atoms or molecules, thereby creating ions. At the
other extreme, transmission and use of electricity that cycles 50 or 60 times
per second (i.e., 50 or 60 Hz) produces distinct electric and magnetic waves
in the extremely long wavelength/low frequency range of the electromagnetic spectrum.
The known and potential health effects of electromagnetic radiation
vary markedly by wavelength/frequency. Prenatal and early childhood
high-level ionizing radiation exposure can cause severe neurotoxicity and
delayed effects including cancer; such effects arise from the ability of ionizing radiation to damage DNA. Intense childhood sun exposure, especially among genetically susceptible subgroups (e.g., fair-skinned Caucasians), is a major cause of malignant melanoma of skin; although UV
light is less energetic than ionizing radiation, it causes photochemical and
oxidative DNA damage. There is some evidence (albeit inconsistent) of
an association between childhood exposure to relatively high-level powerfrequency magnetic fields and childhood acute lymphoblastic leukemia.
229
CHILD HEALTH
230
Gamma rays
AND THE
ENVIRONMENT
< IE-10 cm
X-rays
Ultraviolet
Visible
Infrared
Microwave
Radio
IE-10 IE-09 IE-08 IE-07 IE-06 IE-05 IE-04 0.001 0.01
0.1
> 10cm
1
10
Wavelength (cm)
FIGURE 9–1. Electromagnetic spectrum.
The potential biologic mechanism remains unknown; power-frequency
radiation at intensities found in the general population appears to be
nongenotoxic. Despite much concern about the safety of wireless phones,
there is no convincing evidence that radiofrequency radiation exposure
at levels found among humans is genotoxic or carcinogenic. This chapter
explores these and other potential health effects of prenatal or childhood
radiation exposure.
I. IONIZING RADIATION
All persons are continually exposed to ionizing radiation from natural
sources including cosmic rays from the sun and outer space and natural
radionuclides in the earth’s crust, air, water, and foods. Important anthropogenic sources include medical X-rays, radionuclide emissions from
nuclear reactor accidents, and nuclear weapons detonations. Radionuclide
emissions during a nuclear reactor incident in 1979 at Three Mile Island
in Pennsylvania and a much larger release in 1986 at Chernobyl in the Soviet Union raised renewed public concern about radiation and health.
Large increases in thyroid cancer among children living close to Chernobyl within a few years of exposure point to the vulnerability of children to radiation, whether through high exposures or because of age-
Radiation
231
related susceptibility. The biological effects of ionizing radiation (BEIR) V
report concluded that the excess risk of cancer from radiation exposure
during childhood is about twice that of adults but subject to uncertainties because of the limited follow-up of exposed children (National Academy of Sciences, 1990).
Radiation from radionuclides comprise ␣- and -particles and highenergy photons (␥-rays). ␣-Particles are high-energy positively charged
helium nuclei (emitted by heavier radionuclides such as uranium) that
cause extensive damage over short distances in tissues. -Particles are
electrons (emitted by radionuclides such as tritium) that penetrate tissues
further than ␣-particles but cause less damage per unit distance. Health
risks from ␣- or -particles are related mainly to inhaled or ingested
radionuclides, for example, lung cancer from inhalation of radon, an ␣emitter. ␥-Rays are high-energy photons emitted by radionuclides; lacking charge and mass, they penetrate the body easily, causing tissue damage. X-rays are also photons but generally have lower energy than ␥-rays
and have a different source, that is, the bombardment of matter with
charged particles causing X-ray emissions from extranuclear processes.
Globally, average population exposure from medical uses of radiation is
about half that from natural sources (United Nations Scientific Committee on the Effects of Atomic Radiation, 2000).
The objective of Part I of this chapter is to define the known and potential impacts of ionizing radiation on child health including genotoxicity, adverse pregnancy outcomes, neurotoxicity, and cancer. The discussion addresses the major categories of anthropogenic radiation exposure,
particularly radioactive fallout from atmospheric nuclear weapons tests,
preconceptual and prenatal parental exposure to diagnostic X-rays or occupational sources, and postnatal exposure of children to X-rays, radon,
nuclear detonations, and nuclear power plant emissions. Radiation exposure levels, sources, and risk management issues are explored. The
reader is referred to other sources for more detailed information on ionizing radiation hazards (Agency for Toxic Substances and Disease Registry, 1999; National Academy of Sciences, 1990, 1999; United Nations Scientific Committee on the Effects of Atomic Radiation, 2000).
Health Effects
The health effects of ionizing radiation depend on the type, dose, and duration of radiation exposure, time since exposure (or first exposure), tissue, age, sex, and other personal characteristics. Knowledge of radiationrelated adult cancer risks comes mainly from epidemiologic studies of
232
CHILD HEALTH
AND THE
ENVIRONMENT
Japanese atomic bomb survivors and persons exposed to relatively high
doses of medical X-rays during diagnosis or treatment of diseases (tuberculosis, ankylosing spondylitis, cervical cancer, tinea capitis, and alleged thymus enlargement). Radiation exposure units include
• Gy (gray)—1 Gy equals 1 joule of energy deposited in 1 kg of a material; equivalent to 100 rads
• Sv (sievert)—the biologically effective dose, estimated by multiplying
the absorbed dose in grays by a quality factor Q specific for the type
of incident radiation; equivalent to 100 rem
• Bq (becquerel)—quantity of a radionuclide that will have one transformation in 1 second
Genotoxic Effects
By depositing energy directly in DNA or by creating free radicals, ionizing radiation can cause DNA base damage, single- and double-strand
breaks, and DNA–protein cross-links. Nuclear DNA repair enzymes
usually restore normal DNA structure and function, but unsuccessful, incomplete, or inaccurate repair can cause mutations, chromosomal aberrations (deletions, inversions, translocations), and cell death. Doublestrand DNA breaks are prone to inaccurate repair (e.g., nonhomologous
DNA end joining by DNA repair enzymes), causing mutations and chromosomal aberrations characteristic of leukemia and other cancers. Potential mechanisms for radiation-induced cancer include activation of
oncogenes (e.g., ras and c-myc) and inactivation of tumor-suppressor
genes such as p53 and Rb (retinoblastoma gene). In cells with DNA damage, p53 normally blocks cell division and may trigger apoptosis; inactivated p53 allows cells with DNA damage to divide, increasing the risk of
malignant transformation. Rapidly dividing fetal cells appear to be more
prone to radiation-induced damage than more slowly dividing differentiated cells (Agency for Toxic Substances and Disease Registry, 1999).
Estimated genetic effects of exposure to low-level ionizing radiation
are shown in Table 9–1. Radiation-induced heritable mutations, that is,
germ cell mutations, have not been clearly demonstrated in humans. The
BEIR V report concluded that the acute radiation dose required to double the spontaneous mutation rate in humans is likely to be at least that
observed in mice (1 Sv) but noted considerable uncertainties; doubling
dose estimates, for instance, do not include diseases related to multiple
genes, the likely largest category of genetically related diseases (National
Academy of Sciences, 1990).
Elevated indoor air radon levels in homes or schools have been linked
to DNA damage, chromosomal abnormalities, and micronuclei in pe-
Radiation
233
TABLE 9–1. Estimated Genetic Effects of 1 rem (10 mSv) per Generationa
Additional Cases per 106
Live-Born Infants per Rem
per Generation
Background
Incidence per 106
Live-Born Infants
First
Generation
Equilibrium
2,500
7,500
5–20
1–15
25
75
X-linked disorder
400
1
5
Recessive disorder
2,500
1
Very slow
increase
600
5
Very little
increase
3,800
1
1
20,000–30,000
10
10–100
600,000
300,000
NA
NA
NA
NA
Genetic Effect
Autosomal dominant disorder—
clinically severe
clinically mild
Chromosomal—unbalanced
translocations
Chromosomal—trisomies
Birth defects
Disorders of complex etiology—
heart disease
cancer
Source: National Academy of Sciences (1990).
a Risks
are for an average population exposure of 10 mSv per generation with background gene mutation incidence rates and an assumed doubling dose for chronic exposure of 1 Sv.
NA not available.
ripheral lymphocytes from exposed children. Compared to their parents
and siblings born before the Chernobyl disaster, children conceived afterward by parents occupationally exposed in the cleanup had a sevenfold increase in new bands on multisite DNA fingerprinting, a finding
suggestive of germ-cell mutations. Adults exposed as children or young
adults to relatively high-dose X-rays had higher proportions of lymphocytes with stable chromosome aberrations.
Radiation-induced somatic cell mutations may also cause genetic instability. Several inherited conditions conferring increased cancer risks involve defects of genes encoding DNA repair enzymes needed to correct
damage caused by ionizing radiation and other mutagens1 and clastogens.2 The genes BRCA1, BRCA2, and ATM, for instance, encode proteins
1Chemical
or physical agents that produce heritable changes in nucleotide sequence
(by causing base-pair substitutions or small additions or deletions of one or more base
pairs in genetic material).
2Agents that can cause breaks in chromosomes that result in the gain, loss, or rearrangement of chromosomal segments or cause sister chromatid exchanges during
DNA replication.
234
CHILD HEALTH
AND THE
ENVIRONMENT
involved in homologously directed repair of double-strand DNA breaks;
mutations in these genes are causally linked to an increased risk of breast
and other cancers.
Developmental Effects
Japanese children prenatally exposed to atomic bomb radiation had increased risks of reduced head circumference, mental retardation, longlasting complex chromosomal aberrations, and growth deficits at ages
9–19 years. Findings from the limited number of epidemiologic studies
in other populations include
• Stillbirths—associations with self-reported exposure of either parent to
ionizing radiation (Parker et al., 1999; Savitz et al., 1989; Zhang et al.,
1992)
• Birth defects—associations between preconceptual parental or first
trimester maternal exposure to occupational radiation or diagnostic
X-rays and the risk of neural tube defects (Kallen et al., 1998; Parker
et al., 1999; Sever et al., 1988; Zhang et al., 1992) and cardiac defects
(Correa-Villasenor et al., 1993; Loffredo et al., 2001)
• Birth weight—IUGR associated with maternal periconceptual occupational radiation or paternal preconceptual gonadal exposure to diagnostic X-rays (Shea and Little, 1997; Zhang et al., 1992)
Animal studies indicate that preconceptual exposure of gametes to ionizing radiation (or chemical mutagens) causes pre- and peri-implantation
deaths and growth retardation without birth defects; exposure of the embryo during early organogenesis causes fetal deaths, reduced body
weight, and birth defects (Rutledge, 1997).
Cancer
Ionizing radiation is a known cause of cancer in humans and can produce
cancer in virtually every tissue in animals and likely in humans (Agency
for Toxic Substances and Disease Registry, 1999; International Agency for
Research on Cancer, 2000). The risk of radiation-induced cancer depends
on several factors including the total dose, dose rate, age at exposure, sex,
time since first exposure, and tissue. Readily inducible malignancies include leukemia and thyroid, breast, stomach, and colon cancers. Prenatal
exposure to diagnostic X-rays, especially during the last trimester of pregnancy, is associated with an increased risk of childhood cancer (Doll and
Wakeford, 1997). Several generalizations about radiation carcinogenesis
can be made:
Radiation
235
• A single exposure is sufficient to elevate the cancer incidence many
years later.
• Although body organs vary in sensitivity to radiation, the risks of most
cancers increase after exposure.
• Age at exposure influences the subsequent cancer risk.
° Short-term radiation exposure during childhood appears to double
the lifetime excess risk of fatal cancer compared to the same exposure during adulthood.
° Children exposed to ionizing radiation have an increased risk of leukemia and thyroid cancer that may develop as soon as 2–4 years after exposure.
° Girls exposed to X-rays during or before adolescence have higher excess breast cancer risks as adults than those exposed at older ages.
• At the same dose, high linear energy transfer (LET) radiation (e.g.,
␣-particles) is far more cytotoxic and oncogenic than low-LET radiation (e.g., X-rays) and can increase cell transformation rates at doses
that have little effect on cell survival.
Atomic Bomb Survivors
Only two cases of cancer before age 15 years occurred among 1829 Japanese children prenatally exposed to atomic bomb radiation. Postnatally exposed Japanese children had increased risks of adult cancers including
leukemia and breast cancer. The estimated lifetime risk of solid tumors
per sievert for those exposed at age 10 years was 1.0–1.8 times that of persons exposed at age 30 years (Pierce et al., 1996).
Radioactive Fallout
Atmospheric nuclear detonations and nuclear reactor accidents have released iodine, cesium, strontium, and other radionuclides over wide areas.
Findings from investigations of children exposed to radioactive fallout
from nuclear testing in Nevada and the Marshall Islands and the Chernobyl nuclear reactor disaster include:
• Thyroid cancer—fairly consistent evidence of increased risks of benign
and malignant thyroid tumors, particularly among persons exposed as
children, with latent periods ranging from a few years to several
decades
• Leukemia—limited evidence of increased risks, especially among children and persons exposed before age 20 years
The Chernobyl accident released large amounts of radionuclides including 131I, 137Cs, and 134Cs in a plume that moved mainly north into Belarus,
west into Ukraine, and to a lesser degree throughout the Northern Hemi-
236
CHILD HEALTH
AND THE
ENVIRONMENT
sphere. In the immediate surroundings there were 143 adult cases of acute
radiation syndrome including 34 deaths. About 5%–8% of children in Belarus aged less than 15 years received estimated thyroid doses of at least
2 Gy. Childhood thyroid cancer incidence rates increased from 0.2 to 4.0
per 100,000 persons per year during 1987–1995. An intensive thyroid cancer screening survey of children within 150 km of the Chernobyl power
plant and born before, during, or after the year of the accident showed a
markedly increased risk among those born during the 3 years before the
accident and exposed as young children (Shibata et al., 2001). Although
thyroid screening likely contributed to the increased incidence rates, a
case-control study showed a strong association between childhood thyroid cancer and estimated thyroid radiation dose from ground deposition
of 131I (Astakhova et al., 1998). Factors possibly contributing to the
markedly increased risk of childhood thyroid cancer include relatively
high thyroid doses from energy-rich, shorter-lived radioiodides and iodide deficiency in the region of the Chernobyl reactor.
The International Program on the Health Effects of the Chernobyl Accident, established by the WHO in 1991, concluded that childhood thyroid cancer increased significantly in Belarus and Ukraine beginning 3
years after the accident but there was no increase in leukemia or mental
retardation among children exposed prenatally. It may be expected that
radiation-induced thyroid cancer and adenomas will continue among
those exposed as children well into their adult lives. Elevated childhood
leukemia rates among children living near nuclear facilities have been reported, but other large studies have shown no association, justifying the
conclusion that there is no convincing evidence that radiation released
during normal operations of nuclear plants increases childhood cancer
rates in surrounding populations.
Prenatal Exposure to Ionizing Radiation
Although medical X-rays are not part of the environment as defined here,
they are a major source of radiation exposure during childhood, and studies of exposed children have provided insights into potential risks from
environmental sources of ionizing radiation. The Oxford Survey of Childhood Cancers produced the first substantive evidence that prenatal exposure to diagnostic abdominal X-rays is associated with an increased
risk of childhood cancer, especially leukemia (Stewart et al., 1958). The
estimated radiation dose from prenatal X-rays (mainly to detect pelvic
disproportion) was 0.46 cGy/film; 9% of pregnant women were exposed
to five or more films, that is, to at least 2.3 cGy. A similar association was
observed in a large U.S. study (MacMahon, 1962). These findings were in
contrast to the apparent absence of excess childhood cancer among pre-
Radiation
237
natally exposed Japanese atomic bomb survivors; there were, however,
only about 1800 prenatally exposed Japanese children, substantially limiting the statistical power of the study. (Note: the childhood leukemia incidence rate in the United States is about 45 cases per million children per
year; assuming the same risk in unexposed Japanese children, about two
cases would be expected over a 30-year follow-up period).
A recent review of epidemiologic studies of childhood leukemia and
prenatal maternal exposure to diagnostic X-rays, especially in the third
trimester, concluded that there is strong evidence of an increased risk
(pooled RR 1.38, 1.31–1.47) (Doll and Wakeford, 1997); other reviewers
reached a similar but less definitive conclusion (Boice and Miller, 1999).
Negative results from recent studies of maternal X-ray exposure and childhood leukemia could reflect a true null association or differences from
earlier studies, such as, reduced radiation doses of diagnostic X-rays
(Meinert et al., 1999; Naumburg et al., 2001; Shu et al., 2002). Prenatal maternal X-ray exposure in a German study was associated with childhood
lymphomas (Meinert et al., 1999). There is inadequate evidence of an association between childhood leukemia and preconceptual paternal exposure to X-rays or occupation-related radiation. The UNSCEAR estimated
that the absolute excess risk of childhood cancer from prenatal maternal
exposure to X-rays is about 5% per sievert (United Nations Scientific
Committee on the Effects of Atomic Radiation, 2000).
Postnatal Exposure of Children to Therapeutic and Diagnostic Radiation
A pooled analysis of five cohort and two case-controls studies of thyroid
cancer indicated a strong association with radiation exposure during
childhood, a linear exposure–risk relationship extending to doses as low
as 0.10 Gy, and higher absolute and relative excess risks for females compared to males (Ron et al., 1995). The UNSCEAR concluded that children
are more susceptible to radiation-induced thyroid cancer than older persons; for example, the risk for children aged 0–5 years at exposure was
fivefold that for those exposed at ages 10–14 years (United Nations Scientific Committee on the Effects of Atomic Radiation, 2000).
Children given radiotherapy for ringworm of the scalp or skin hemangiomas had substantially increased brain and thyroid cancer risks and
strong exposure–risk relationships. In Connecticut, cohorts born during
1915–1945, the period when use of X-rays to treat benign conditions of
the head and neck was common, had increased thyroid cancer risks
(Zheng et al., 1996). Among three recent studies of childhood leukemia
and postnatal diagnostic X-ray exposure, one showed an overall association (including interactions between X-ray exposure and polymorphisms
in several DNA repair genes), another showed a link only to pre-B acute
238
CHILD HEALTH
AND THE
ENVIRONMENT
lymphatic leukemia, and the other showed no relationship (InfanteRivard et al., 2000; Meinert et al., 1999; Shu et al., 2002).
Postnatal Exposure of Children to Environmental Radiation
Radon, a proven human carcinogen, is the most ubiquitous source of radiation exposure for the general population; residential radon may cause
6600–24,000 lung cancer deaths annually among adults in the United
States, the risk being about tenfold higher for persons first exposed at age
15 compared to age 50. Polonium, a radon decay product, behaves like
calcium in vivo and is taken up by bone, potentially causing significant
radiation exposure of bone marrow cells of the fetus and child, but there
is inadequate evidence of an association between radon exposure and
childhood leukemia. There was a moderately strong association between
CNS tumors and measured indoor radon levels (Kaletsch et al., 1999). The
BEIR VI report concluded that the limited evidence linking general population radon exposure to nonlung cancers does not warrant consideration in setting radon standards and guidelines (National Academy of Sciences, 1999). Other sources of residential radiation exposure have been
linked to childhood leukemia (␥-radiation from uranium-containing concrete) and bone sarcomas (drinking water radium levels) (Axelson et al.,
2002; Finkelstein and Kreiger, 1996).
Neurotoxicity
Cases of reduced head circumference and mental retardation after prenatal exposure to therapeutic ionizing radiation were first reported in the
1920s. Children of mothers who received whole-body radiation from
atomic bombs at Hiroshima and Nagasaki had an increased risk of microcephaly, mental retardation, and lower school performance. The risk of
mental retardation was greatest for exposures during gestational weeks
8–15, the peak time of neuron proliferation and neuroblast migration to
the cerebral cortex, processes that are disrupted by ionizing radiation. The
average IQ deficit for exposure during early gestation was 21–29 points
per gray (National Academy of Sciences, 1990). Children exposed to cranial radiation for treatment of tinea capitis had increased risks of IQ
deficits, lower scores on high school aptitude tests, poor school performance, hospitalization for mental disorder (including a dose–response relationship), and a borderline increase in mental retardation. Among survivors of childhood cancer exposed to therapeutic whole-brain radiation,
substantial cognitive deficits occurred mainly among those aged less than
3 years at exposure.
Radiation
239
Exposures
The global average exposure to natural sources of radiation is about
2.4 mSv/year (Table 9–2). Other sources include the following (United
Nations Scientific Committee on the Effects of Atomic Radiation, 2000):
• Atmospheric nuclear testing—global average exposure peaked at about
150 Sv/year in 1963, declining to about 5 Sv/year in 2000.
• Medical uses of radiation—the average annual dose in developed countries is about 1.2 mSv, but the range of individual exposures is very
wide.
• Occupationally exposed persons—the average annual dose among
monitored workers is about 0.6 mSv from anthropogenic sources and
1.8 mSv from enhanced natural sources (e.g., air crew, mining, mineral
processing).
Nuclear Test and Reactor Emissions
Among the nuclear tests conducted in the Marshall Islands during
1946–1958, radioactive fallout was greatest after the large 1954 Bravo thermonuclear test, causing average absorbed thyroid doses of about 1 Gy in
TABLE 9–2. Average Annual Radiation Exposure (mSv)
of General Population
Worldwide Average
Annual Dose (Range)
Source
Cosmic rays
Terrestrial ␥-rays
Inhalation (mainly radon)
Ingestion
0.4
0.5
1.2
0.3
Total natural sources
2.4 (1–10)
Medical uses of radiation
Atmospheric nuclear testing
Chernobyl accident
Nuclear power production
(0.3–1.0)
(0.3–0.6)
(0.2–10)
(0.2–0.8)
0.4 (0.2–1.2)a
0.005
0.002
0.0002
Total anthropogenic
1.2
Source: United Nations Scientific Committee on the Effects of
Atomic Radiation (2000).
a Varies
by level of health-care resources per capita.
240
CHILD HEALTH
AND THE
ENVIRONMENT
the most contaminated islands. Exposed children had a high prevalence
of screen-detected thyroid nodules; iodine deficiency may have been a
contributory factor. Average cumulated thyroid 131I doses among persons
most exposed to the Nevada atmospheric nuclear bomb tests during
1951–1958 were 12–16 cGy; a few persons, including young children, may
have received thyroid doses of over 1 Gy. Early summer tests caused the
greatest doses because of fallout when cows were on pasture and fresh
local vegetables were used. The Chernobyl nuclear reactor accident released large amounts of radioisotopes, including several iodine and cesium radioisotopes. The estimated average thyroid 131I dose among controls in a case-control study of childhood thyroid cancer in Belarus was
20 cGy.
Radon
Radioactive radon gas (222Rn) and its decay products occur naturally in
many homes and cause average doses of about 1.2 mSv per year, that is,
about half of the average background radiation dose to the U.S. population. Average outdoor radon concentrations are about 10 Bq/m3, whereas
about 5%–10% of American homes have radon levels above 148 Bq/m3
(the current remedial action level). Surveys of schools have detected occasional high radon levels. Average radon levels in drinking water from
ground sources are about 20 Bq/L, much higher than those of surface
sources.
Risk Management
Sources
Radon is released by the decay of naturally occurring thorium and uranium, elements present in rock and soil in highly variable concentrations.
Radon decays into polonium isotopes (218Po and 214Po) that are ␣-particle
emitters. Depending on the amount of radon in the underlying soil, building construction, and air exchange rates, indoor radon concentrations can
reach relatively high levels. Radon can enter homes through exposed soil
(e.g., crawl spaces), cracks in concrete floors and walls, floor drains,
groundwater (e.g., sumps, showering), and building materials.
Atmospheric testing of nuclear weapons during 1945–1980 was the
main anthropogenic environmental radiation source globally (United Nations Scientific Committee on the Effects of Atomic Radiation, 2000). These
tests caused widespread dispersion of radionuclides such as 90Sr, 137Cs,
Radiation
241
and 131I in the atmosphere; those with long half-lives (90Sr and 137Cs) remain widespread in ecosystems. 90Sr mimics calcium and accumulates in
bone, 131I accumulates in thyroid, and 137Cs emulates potassium and disperses throughout the body. Food is normally a minor source of radionuclides but radioactive iodine can be widely dispersed in air, deposit on
pastureland, and enter the human food chain via fresh cow milk. Dairy
products such as cheese are less hazardous, as they take time to prepare
and the half-life of iodine radionuclides is quite short; for example, the
half-life of 131I is 8 days. Food crops and meat from wildlife and domestic animals may be contaminated by 137Cs.
Intervention
The International Commission for Radiological Protection and national
agencies such as the U.S. EPA have recommended radiation exposure limits for people and environmental media (Table 9–3). Cessation of atmospheric nuclear bomb detonations has reduced an important source of airborne radioactivity. Compliance with regulations on the design, operation,
and maintenance of nuclear power plants must be monitored and enforced to protect the public from disasters such as that at Chernobyl. When
large atmospheric releases of 131I and other iodine radionuclides occur,
TABLE 9–3. Recommended Exposure Limits to Ionizing
Radiation for the General Population
Exposure
Recommended Limit
Total exposure a
Woman’s abdomen
General public (whole body,
5 year average)
2 mSv/year
1 mSv/year
Radon in home indoor air b
Remedial action level
148 Bq/m3
Drinking water (MCLs)c
␣-Particles
-Particles
226Ra and 228Ra combinedd
0.555 Bq/L
0.04 mSv/year
0.185 Bq/L
a Source:
International Commission on Radiological Protection (1991).
b Source:
U.S. Environmental Protection Agency (1993).
c Maximum
dRadium
contaminant levels.
isotopes.
Source: U.S. Environmental Protection Agency (2001).
242
CHILD HEALTH
AND THE
ENVIRONMENT
ingestion of iodide tablets reduces thyroid uptake of inhaled or ingested
radioiodides but only if taken within hours after exposure (Zanzonico and
Becker, 2000).
Control of radon exposure in homes, schools, and other buildings frequented by children also require monitoring, regulatory, and educational
programs. Monitoring is important because radon levels vary enormously,
with some buildings having very high levels. Only 43% of new singlefamily detached homes built in areas with high radon potential during
1998 incorporated radon-reducing features (U.S. Environmental Protection Agency 2002). The EPA advice to homeowners concerning radon includes the following: (1) conduct a short-term test (2–90 days) for radon
levels, (2) if the result is 148 Bq/m3 or higher, conduct a short- or longterm confirmatory test or, if the level is 370 Bq/m3 or higher, conduct a
second short-term test immediately, and (3) take remedial action if a longterm test result or the average of two short-term test results is 148 Bq/m3
or higher. The EPA has proposed a standard for radon in public drinking
water supplies that would limit radon to either 11.1 Bq/L or to 148 Bq/L
with a requirement to develop indoor air radon programs.
Inappropriate radiotherapy for benign conditions subsided when the
U.S. National Academy of Sciences and the U.K. Medical Research Council in 1956 recognized the health risks of ionizing radiation. There appear
to be no population-based data on measured exposures of pregnant
women or children to medical or dental diagnostic X-rays; although these
are not environmental hazards as defined in this book, they are important sources of radiation exposure.
Conclusions
Proven Health Outcomes
• Microcephaly and mental retardation (high-dose first trimester exposure)
• Cancer
° Low-dose prenatal exposure—childhood leukemia
° High-dose childhood exposure—leukemia and thyroid, brain, and
breast cancers in adulthood and childhood thyroid cancer
Unresolved Issues and Knowledge Gaps
• Developmental effects of parental preconceptual or first trimester
exposure
Radiation
243
° Inadequate evidence of an association between paternal preconcep-
tual occupational exposure and childhood leukemia
° Limited evidence of associations with birth defects (neural tube,
cardiac)
° Limited evidence of associations with IUGR
• Cancer related to low-dose radiation exposure during childhood
° Limited evidence of associations between diagnostic X-ray exposure
during childhood and childhood leukemia
° Inadequate evidence of associations between environmental radiation exposure (radon in indoor air, radium in drinking water, other)
and childhood cancers (leukemia and brain cancers, bone sarcomas)
• Knowledge development—need epidemiologic studies with improved
exposure assessment and statistical power to assess childhood and delayed health effects of low-level, early-life exposures
Risk Management Issues
• Prevention—need to minimize exposure to occupational, medical, and
environmental sources of ionizing radiation among children and reproductive-age persons, especially pregnant women
• Biomonitoring—need to measure ionizing radiation exposure among
reproductive-age men and women and children in homes, schools,
workplaces, and medical/dental applications (no population-based
data available except for certain occupational exposures)
II. POWER FREQUENCY ELECTRIC
AND MAGNETIC FIELDS AND
RADIOFREQUENCY RADIATION
A 1979 report of an association between childhood cancer and residential
electric wiring configurations in Denver, Colorado, raised concern about
the potential health effects of electromagnetic fields (EMFs) and triggered
substantial research (Wertheimer and Leeper, 1979). Electric field strength
(volts/meter) depends on the voltage independent of current size. Magnetic fields are produced by moving electric charges and their strength
(amps/meter) reflects the amount of current passing through a conductor, independent of voltage; most epidemiologic studies have measured
magnetic flux density.3 When electrical equipment is connected to a power
3 Magnetic
flux density units are gauss (G) or tesla (T); 1 T 10 mG.
244
CHILD HEALTH
AND THE
ENVIRONMENT
supply, there is a 50 or 60 Hz electrical field along the power cord, and
when the equipment is turned on, there is also a 50 or 60 Hz magnetic
field. The strengths of electric and magnetic fields both decrease rapidly
with distance from the source; common building materials and even
shrubs shield against electric but not magnetic fields. Power-frequency
EMFs have wavelengths of about 5,000 km, far greater than UV-light or
ionizing radiation (Figure 9–1).
Radiofrequency (RF) radiation spans the frequency range 3 kHz to
300 GHz, that is, above EMFs and below visible light, with wavelengths
of less than 1 cm to tens or hundreds of meters. Having much lower energy than shorter-wavelength electromagnetic radiation such as X-rays or
␥-radiation, RF radiation has insufficient energy to ionize biologically important atoms but it can cause tissue heating. Microwave (MW) radiation
comprises the subset of RF radiation at 300 MHz to 300 GHz (Figure
9–1). Exposure to RF radiation is ubiquitous and is increasing with the
growing use of wireless phones and other communication systems. Existing safety codes for RF radiation are designed to protect the public and
occupationally exposed persons from potential heat-induced adverse
health effects. At distances more than several wavelengths from an RF
emitter, RF radiation levels can be accurately estimated as power density
(e.g., W/m2).
The objective of Part II of this chapter is to describe the evidence relating environmental power-frequency magnetic fields and RF radiation
to potential impacts on child health including biologic effects, reproductive outcomes, and childhood cancer. The discussion addresses parental
and childhood exposure sources including proximity to high-voltage
power lines, residential wire configuration, and use of electric devices,
wireless phones, and radio/TV communication systems. Risk management issues are explored in light of the widespread exposure to potential
hazards that are still poorly understood.
Health Effects
Epidemiologists face many obstacles in assessing potential health effects
of EMF and RF radiation exposures. The relative rarity of serious health
outcomes such as birth defects and childhood cancer, the low prevalence
of high exposures to power-frequency magnetic fields, and the choice and
accurate measurement of biologically relevant exposure indicators are
among the most important challenges. Research to date has barely begun
to address the numerous potentially relevant exposure parameters, susceptible population subgroups, and interactions with other factors.
Radiation
245
Molecular Mechanisms and Biologic Effects
Power-Frequency Electromagnetic Fields
Power-frequency magnetic fields normally transmit negligible energy to
tissues. The two main mechanisms by which biologic systems may interact with power-frequency EMFs are (International Radiation Protection Association, 1993) (1) induced electric currents, polarization of bound charge
(formation of electric dipoles), and reorientation of existing electric dipoles
in tissues and (2) induced electric fields and currents in tissues.
Electric and magnetic fields are independent entities for significant
sources of human exposure. Unlike electric fields, magnetic fields can easily penetrate buildings and tissues. Power-frequency magnetic fields induce tissue currents of about 1 A/m2, three orders of magnitude lower
than tissue currents associated with electrical activity in neurons and
seven or eight orders of magnitude higher than those induced by external electric fields. The earth’s static magnetic field does not induce tissue
currents. The strength of both electric and magnetic fields is inversely proportional to a power function of the distance from the source that depends on the geometry of the source. Expert panel reviews of the potential biologic effects of power-frequency magnetic fields have concluded
that (International Commission on Non-Ionizing Radiation Protection,
1998; National Academy of Sciences, 1997; National Institute of Environmental Health Sciences, 1998)
• Magnetic flux densities commonly experienced in residences (0.01–
1 T) have no reproducible in vitro biologic effects.
• Exposures above 50 T have fairly consistent effects on intracellular
calcium levels, signal transduction, and gene expression in experimental systems; exposures above 100 T increase cell proliferation, disrupt signal transduction pathways, and inhibit differentiation, and exposures above 500 T affect bone healing.
• Direct genotoxicity has not been consistently demonstrated at any magnetic flux density, but very strong magnetic fields may enhance mutation rates after initiation with ionizing radiation.
• There is no convincing evidence of adverse developmental effects or
carcinogenicity in animals at relatively high magnetic flux densities,
with the possible exception of enhanced growth of chemically induced
mammary cancers in rodents exposed to magnetic flux densities in the
range 0.01–30 mT.
The hormone melatonin appears to reduce the incidence and growth rate
of chemically induced breast cancers in experimental animals. It has been
hypothesized that power-frequency magnetic fields may reduce mela-
246
CHILD HEALTH
AND THE
ENVIRONMENT
tonin levels in humans and increase the breast cancer risk; a recent review, however, concluded that exposure to light at night profoundly suppressed blood melatonin levels, but magnetic fields had little or no effect
(Graham et al., 2001).
Radiofrequency Electromagnetic Radiation
Absorption of RF and tissue heating are highest at frequencies near the
body’s resonant frequencies, for example, about 35 MHz for an ungrounded adult and 700 MHz for an infant’s head. Animal studies have
provided some evidence of genotoxicity and carcinogenicity at very high
RF radiation levels; the weight of evidence from toxicologic studies, however, suggests that RF radiation is not directly mutagenic and that reported chromosomal abnormalities in experimental models are likely
caused by tissue warming (Brusick et al., 1998). Reviews of the literature
on the biologic effects of RF radiation at the frequencies used by wireless
phones have concluded that (Expert panel on the potential health risks of
RF fields from wireless telecommunication devices, 1999; International
Commission on Non-Ionizing Radiation Protection, 1998)
• RF radiation penetrates tissues up to 1 cm, causing tissue heating.
• Ornithine decarboxylase activity and calcium efflux from neurons increase when the amplitude of RF radiation is modulated by EMFs; these
effects are not known to cause adverse health effects and require further research.
• Reported effects of RF radiation on cell proliferation, genotoxicity, and
increased permeability of the blood–brain barrier have been inconsistent.
• There is little evidence that RF radiation is carcinogenic in experimental animals.
Developmental Effects
Power-Frequency Magnetic Fields
A meta-analysis of nine case-control and cohort studies of pregnancy outcome and maternal exposure to video display monitors during pregnancy
yielded pooled odds ratios of unity for spontaneous abortion and birth
defects (Parazzini et al., 1993). A recent review of epidemiologic studies
concluded that there were few associations and no consistent relationships between occupational or residential power-frequency magnetic field
exposure and the risk of spontaneous abortion, birth defects, low birth
weight, or preterm birth (Shaw, 2001). Nevertheless, two California cohort studies (one used a nested case-control design) demonstrated a doubling of the spontaneous abortion risk at dosimeter-measured personal
Radiation
247
magnetic flux densities greater than 1.6 T (Li et al., 2002) or 3.5 T (Lee
et al., 2002). Toxicologic evidence indicates that power-frequency magnetic fields are inconsistently teratogenic in chick embryos and generally
have no effect on fetal growth, birth defects, fetal loss, and neurobehavioral development in other animal species.
Radiofrequency Radiation
The few epidemiologic studies of RF radiation and adverse pregnancy
outcomes have shown no convincing relationships, that is, no strong and
precise associations with quantitative exposure assessment, exposure–risk
relationships, and control of confounders (Royal Society of Canada, 1999).
Only a few RF radiation frequencies have been tested in experimental systems, and most have assessed acute high-dose exposure rather than the
chronic low-level exposure typical of human populations; the most consistent high-dose effect was reduced fetal growth. Evidence of teratogenic
effects is inconsistent and is seen only at RF radiation doses capable of
causing substantial tissue heating.
Cancer
Childhood brain cancer and leukemia incidence rates have generally been
stable or have increased slightly during recent decades. In the United
States, most of the increase in childhood brain cancer and leukemia incidence rates occurred during the mid-1980s, suggestive of an artifact such
as improved diagnosis or cancer registration (Linet et al., 1999). Despite
extensive research into potential causes, including power-frequency magnetic fields, there are no known modifiable risk factors for childhood cancer with high attributable risks.
Power-Frequency Magnetic Fields
The first study of childhood cancer and high-current power line configurations in Denver, Colorado, showed a significant association and an
exposure–risk relationship (Wertheimer and Leeper, 1979). An independent study in Denver confirmed this association and showed that the
Wertheimer-Leeper wire code and a simplified version of it were both correlated with measured bedroom magnetic flux densities (Kaune and
Savitz, 1994; Savitz and Kaune, 1993).
Leukemia. Leukemia has been associated with exposure to magnetic
flux densities at home (childhood leukemia) and at work (especially adult
chronic lymphocytic leukemia). Because childhood cancer is rare, almost
all investigators have used case-control designs, estimating past expo-
248
CHILD HEALTH
AND THE
ENVIRONMENT
sures through parent-reported information, residential wire configuration, distance from electric power lines, and residential and personal exposure monitoring over very limited time periods (Savitz, 1995). Among
the few studies focusing on electric fields, one reported a moderately
strong exposure–risk relationship between measured nighttime bedroom
electric fields and childhood leukemia (Coghill et al., 1996), but several
larger studies, including two with personal electric field measurements,
were negative (Green et al., 1999; London et al., 1991; McBride et al., 1999).
Key conclusions from several meta-analyses and expert panel reviews
of childhood cancer and power-frequency magnetic flux density include
(Ahlbom et al., 2000, 2001; International Agency for Research on Cancer,
2002; National Institute of Environmental Health Sciences, 1999) (see also
Table 9–4):
• Childhood leukemia
° Pooled odds ratios of 1.3 to 1.5 for residences within 50 m of power
transmission and distribution lines
° Pooled odds ratios of 1.3 to 1.7 for estimated high historic magnetic
fields based on wire configuration; inconsistent odds ratios for present measured magnetic flux densities
° An apparent concentration of risk among children with timeweighted average magnetic flux densities of 0.4 T or greater; only
1.4% of cases and 0.6% of controls had such exposure (Ahlbom et al.,
2000)
° Inadequate evidence of an association with electric fields
• Other cancers—there were no consistent relationships between residential electric or magnetic fields and brain or other childhood cancers.
• Animal evidence—animal evidence of power-frequency magnetic field
carcinogenicity was judged as inadequate; there were no data to assess
the carcinogenicity of static magnetic fields or static or power-frequency
electric fields or to support a specific exposure metric.
Of the individual studies of power-frequency magnetic fields and childhood leukemia, only two studies used personal exposure monitoring, one
showing a positive association that was stronger among children aged
less than 6 years (Green et al., 1999) and the other showing no relationship (McBride et al., 1999). Other reported associations with childhood
leukemia include paternal but not maternal occupational exposure to
power-frequency magnetic fields, residential power-frequency magnetic
flux density at night, and time spent watching television or playing games
connected to television sets (television sets produce very weak powerfrequency magnetic fields but do produce stronger EMFs at certain
frequencies).
TABLE 9–4. Meta-Analyses of Childhood Cancer and Power-Frequency Magnetic Fields
Reference
Studies Included
Pooled Estimates
Washburn et al. (1994)
13 studies of childhood cancer and residential
proximity to electric power distribution
equipment and risk of childhood leukemia,
lymphoma, and CNS tumors
Pooled odds ratios based on residential distance from
transmission and distribution wires included leukemia
(OR ⫽ 1.5, CI 1.1–2.0), lymphoma (OR ⫽ 1.6, CI 0.9–2.8),
and CNS tumors (OR ⫽ 1.9, CI 1.3–2.7); no association
between pooled odds ratios and any of 15 indicators of
epidemiologic quality
Miller et al. (1995)
7 case-control studies of childhood leukemia and
wire code, distance to power distribution
equipment, spot magnetic flux density
measurements, and calculated indices based on
distance and historic power load data
Pooled odds ratios by exposure index were wire codes
(OR ⫽ 1.6, CI 1.3–2.0), distance (OR ⫽ 2.1, CI 1.2–3.7),
spot measures (OR ⫽ 1.1, CI 0.7–1.7), and calculated
index (OR ⫽ 2.5, CI 1.0–6.0)
Meinert and Michaelis (1996)
13 studies of childhood cancer and wire code,
distance to power distribution equipment, spot
magnetic flux density measurements; note: each
OR based on pooling of 2 to 6 studies
Two-level wire code: all cancers combined (OR ⫽ 1.4,
CI 0.9–2.0), leukemia (OR ⫽ 1.7, CI 1.1–2.5), CNS tumors
(OR ⫽ 1.5, CI 0.7–3.3), and lymphomas (OR ⫽ 1.3,
(CI 0.5–3.4)
Distance odds ratios for leukemia for distances of ⬍100 m,
⬍50 m, and ⬍25 m, respectively, were 1.1 (0.8–1.6),
1.3 (0.9–1.9), and 1.9 (1.0–3.5)
Measured magnetic flux densities: odds ratios for leukemia
for exposures ⬎0.1 T, ⬎0.2 T, and ⬎0.3 T, respectively,
were 1.6 (0.9–2.7), 1.9 (1.1–3.3), and 1.3 (0.3–5.8)
National Academy of
Sciences (1997)
Pooled analyses of 11 case-control studies of
childhood leukemia in relation to wire codes
and distance
Pooled odds ratios for high exposure: wire code
(OR ⫽ 1.5, CI 1.1–2.1), distance (OR ⫽ 1.4, CI 1.1–1.8)
(continued)
TABLE 9–4. Meta-Analyses of Childhood Cancer and Power-Frequency Magnetic Fields (continued)
Reference
Studies Included
Pooled Estimates
Wartenberg (2001)
19 studies of childhood leukemia and magnetic
flux density and distance to power
distribution equipment
Pooled odds ratios for high exposure: calculated and
measured magnetic flux density (OR ⫽ 1.3, CI 1.1–1.7),
distance to power distribution equipment (OR ⫽ 1.2,
CI 1.0–1.6)
Angelillo and Villari (1999)
Pooled analysis of 1 cohort and 14 case-control
studies of childhood leukemia and wire codes,
proximity to power distribution equipment,
and spot and 24-hour measures of magnetic flux
density
Pooled odds ratio for high exposure: wire code (OR ⫽ 1.5,
CI 1.1–2.0), 24-hour measured magnetic flux densities
(OR ⫽ 1.6, CI 1.1–2.2)
Ahlbom et al. (2000)
Pooled analysis of 9 epidemiologic studies of
childhood leukemia and measured or estimated
magnetic flux density and wire code
Pooled odds ratio for estimated magnetic flux densities
of 0.4 T or greater (versus less than 0.1 T) was 2.0
(CI 1.3–3.1); adjustment for potential confounders
made little difference; among North American subjects,
the odds ratio for the highest wire code category was
1.2 (CI 0.8–1.9)
Greenland et al. (2000)
Pooled analysis of 15 epidemiologic studies of
childhood leukemia and magnetic flux density
and wire codes
Pooled odds ratios: magnetic flux density above 0.3 T,
relative to ⱕ0.2 T, OR ⫽ 1.7 (CI 1.2–2.4); very high
current wire configuration, relative to lowest wire code,
ORs varied from 0.7 to 3.0 across studies and authors
decided not to include summary estimate because
of heterogeneity
Radiation
251
Personal exposures depend on activity patterns and power-frequency
magnetic flux densities that vary substantially over time and place; shortterm exposure measures are likely poor proxies for lifetime dose or frequency of exposure to peak levels. Given the modest strength of the association in positive studies, ignorance about the etiology of childhood
leukemia, the complexity of power-frequency magnetic field exposures,
and the potential modifying effects of other environmental and socioeconomic factors, continuing uncertainty about the relation between magnetic fields and childhood leukemia seems inevitable (Savitz and Poole,
2001). The National Institute of Environmental Health Sciences (NIEHS)
expert panel concluded that power-frequency magnetic fields should be
considered a possible human carcinogen, noting that positive associations
with childhood leukemia are consistent with the limited evidence of increased chronic lymphocytic leukemia risk among occupationally exposed
adults.
Brain tumors. Meta-analyses of studies of childhood brain tumors and
power-frequency magnetic fields yielded pooled odds ratios of 1.9 (CI
1.3–2.7) for proximity to high-voltage wires (Washburn et al., 1994) and
1.5 (CI 0.7–3.3) for high wire code configuration (Meinert and Michaelis,
1996). More recent reviews, however, concluded that there was no consistent evidence of associations with exposure of parents or children to
power-frequency magnetic fields based on wire code, distance, measured
or estimated magnetic flux density, or electrical appliance use (Ahlbom
et al., 2001; Kheifets et al., 1999).
Radiofrequency Radiation
Epidemiologic studies of cancer and RF radiation from proximity to television/radio transmitters or occupational exposures have produced inadequate evidence to conclude that RF radiation is a likely cause of human cancer (Elwood, 1999). Among four studies published since this
review, three showed no associations between RF radiation related to cell
phone use or occupation and the risk of adult CNS tumors or leukemia.
One study revealed an association between brain gliomas, but not meningiomas or salivary gland cancers, and exposure to analog but not digital
wireless phones (Auvinen et al., 2002). Wireless phone use was also linked
to a three- to fourfold increased risk of adult uveal melanoma (Stang et
al., 2001). A large case-control study of childhood neuroblastoma showed
a borderline association with maternal occupational RF radiation exposure (De Roos et al., 2001). The question of a role for RF radiation in childhood cancer will remain open until further large analytical studies with
adequate exposure assessment are conducted.
252
CHILD HEALTH
AND THE
ENVIRONMENT
Exposures
Power-Frequency Magnetic Field Measurement
The exposure metrics that best indicate biologically relevant exposure to
power-frequency magnetic fields or RF radiation are unknown. Epidemiologic studies of power-frequency magnetic fields and child health have
used (1) indirect indices including job-exposure matrices, self-reported use
of electrical appliances, proximity to power lines, and residential electric
wire configuration coding, (2) spot and time-weighted measures of residential power-frequency magnetic flux densities, and (3) personal magnetic flux density exposures measured by wearing personal dosimeters for
up to several days.
The Wertheimer-Leeper wire coding system and a simplified derivative were developed as low-cost, nonintrusive methods to assess longterm residential exposure to power line magnetic fields. Wire codes are
thought to be less likely than short-term measurements to misclassify exposure because they depend on relatively constant characteristics, while
measured power-frequency magnetic flux density varies substantially
over time. Power lines strongly influence background magnetic fields over
long time periods, while currents in residential electrical grounding systems cause the largest variations over shorter time periods. In homes
where the electrical system is grounded to plumbing, appliances can transmit imperceptible contact currents to a person in contact with the appliance and produce stronger electric fields in target tissues compared to
magnetic or electric fields (Kavet et al., 2000). The geometry of a source
influences the rate of decline of magnetic field strength with distance; for
example, magnetic field strength declines more rapidly with distance from
a point source such as an electric motor than it does with distance from
an extended source like a power line. Children receive geometric mean
extremely low frequency (ELF) and very low frequency (VLF) magnetic
flux densities of about 0.002–0.02 T while watching TV or playing video
games, very close to background levels.
Radiofrequency Radiation Measurement
Most RF radiation in the environment comes from commercial radio and
television broadcasting and telecommunications facilities. There have
been very few studies of population RF exposure levels. The average background RF power density in large U.S. cities is about 50 W/m2, with
about 1% of persons having exposures exceeding 10 mW/m2. Within
homes, exposures from appliances (MW ovens, video display units, television sets) contribute a few tens of microwatts per square meter, much
Radiation
253
TABLE 9–5. Reference Levels for the General Public for Electromagnetic Fields
and Radiofrequency Radiation
Exposure
Guideline or Standard
Power-frequency (50–60 Hz)
electromagnetic fields a
Electric fields
Power-frequency magnetic fields
Radiofrequency radiation
100 kHz to 10 GHz (e.g., wireless phones,
radio, television, microwave ovens),
specific absorption rates a
a International
b Department
5 kV/m (50 Hz)
4.2 kV/m (60 Hz)
0.1 T
0.08 W/kg (whole body average)
2 W/kg (localized, head and trunk)
4 W/kg (localized, limbs)
Non-essential or prolonged calls by
children age 16 years should
be discouragedb
Commission on Non-Ionizing Radiation Protection (1998).
of Health (2000).
lower than international reference levels for ambient RF power density
for such frequencies (Table 9–5).
Wireless phones are RF transmitters with maximum power of 0.2–
0.6 W and expose the head at very close range, with peak energy specific
absorption rates in the head of 0.12-2.8 W/kg, near the International Commission on Non-Ionizing Radiation Protection (ICNIRP) specific absorption rate limit for hand-held wireless phones (2.0 W/kg partial body) (Table
9.5). Wireless telecommunication base stations operate at power levels far
lower than those of radio and TV broadcast antennas; ground-level RF
power densities near wireless base stations are on the order of 10 mW/m2
but can be much higher if multiple antennas are clustered at a site (Royal
Society of Canada, 1999). At the frequency of wireless phone base stations
(1850–1990 MHz), the FCC limits for RF power density for the general public is 10 W/m2 (Federal Communications Commission, 1999).
Risk Management
Sources
Although wire code categories are correlated with measured magnetic
flux densities, they explain only 10%–20% of the variance in spot, 24-hour
stationary, and personal dosimetry measurements. The main determinants of personal magnetic flux densities measured by dosimetry appear
to be electric appliances, grounding to metallic water lines, and currents
254
CHILD HEALTH
AND THE
ENVIRONMENT
in nearby power lines (mainly low-voltage distribution lines). Powerfrequency magnetic flux density from household electrical appliances
(hair dryers, vacuum cleaners, MW ovens, electric blankets, televisions,
air conditioners, and computers) decreases rapidly with distance. Young
children spend about 70% of their time at home, over half of which is in
their bedrooms; magnetic flux densities measured by personal dosimetry
of children are strongly correlated with 24-hour average bedroom levels.
Median 24-hour magnetic flux densities above 0.2 T occurred in only
1.4% of children’s bedrooms in Germany (Schuz et al., 2000).
Radiofrequency radiation produced by anthropogenic sources, by ascending frequencies, includes AM and FM radio, VHF and ultrahigh frequency (UHF) television, two-way radios, base stations for transmitting
wireless phone and MW communication signals, MW ovens, radar, and
satellite communications. Other RF radiation sources include magnetic
resonance imaging systems, video display monitors, and antitheft and security devices. There are now more wireless than fixed-line phone users
globally and the former will grow to about 1.6 billion by 2005, driven by
deployment of the lower-cost wireless systems in developing countries
(Repacholi, 2001). Wireless phones, used increasingly by children and adolescents, continually transmit RF signals to base stations when turned on
(Royal Society of Canada, 1999).
Intervention
No national body in the United States has recommended a precautionary
policy to reduce power-frequency EMF exposure. The NIEHS concluded
that there is insufficient evidence of adverse health effects upon which to
base a regulatory limit, but it would be prudent to limit children’s exposure by requiring a minimum distance between power lines and homes,
schools, and day-care centers. The Australian government recommended
that new electric power transmission lines be routed away from schools
and that power line loads be phased to reduce nearby magnetic fields.
The ICNIRP guidelines for general population exposure to EMF in the
power-frequency range are 10 V/m for electric fields and 0.1 T for magnetic flux density (International Commission on Non-Ionizing Radiation
Protection, 1998).
The ICNIRP reviewed health issues related to the use of hand-held
radiotelephones and base transmitters and concluded that (1) there is no
substantive evidence that adverse health effects, including cancer, are
caused by exposure to RF radiation levels at or below the ICNIRP specific absorption rate limits, and (2) neither the epidemiologic nor the toxicologic studies reported to date provide a basis for health hazard assessments of RF radiation exposure or for setting quantitative restrictions
Radiation
255
on human exposure. The ICNIRP guidelines for limiting EMF and RF radiation exposure, based on established health effects, include (International Commission on Non-Ionizing Radiation Protection, 1998)
• 1 Hz to 10 MHz—restrictions on current density (mA/m2) to prevent
effects on nervous system function
• 100 kHz to 10 GHz—restrictions on specific absorption rates (W/kg) to
prevent whole-body heat stress and excessive localized tissue heating
• 10–300 GHz—restrictions on power density (W/m2) to prevent excessive heating in tissue at or near the body surface
The ICNIRP guidelines address neither the potential for increased susceptibility during childhood nor the possibility of health outcomes other
than tissue heating and acute effects on nervous system function. This approach, however, is consistent with the recent position of other agencies.
Existing RF radiation exposure standards for the 100 kHz to 10 GHz range
are all based on biologic data and a safety factor but vary because of different interpretations of the biologic data, magnitude of the safety factor,
consideration of averaging times, and dependence on frequency (Erdreich
and Klauenberg, 2001).
The Royal Society of Canada recently reviewed the evidence of health
effects from RF radiation and concluded that the Canadian Safety Code
6 limits for specific energy absorption rates for RF and MW frequencies
100 kHz to 10 GHz generally protect the general public from adverse
health effects related to tissue warming (Royal Society of Canada, 1999).
The expert panel noted that biologic effects may occur at nonthermal exposure levels but concluded that there is insufficient evidence to assess
whether these may cause adverse health effects. The U.K. Department of
Health recommended that nonessential and prolonged calls on wireless
phones by children below age 16 years be discouraged; it adopted this
precautionary approach because of evidence that changes in brain activity occur at RF radiation levels below current guidelines, the significant
gaps in scientific knowledge, and the possibility that the developing
brains of children and teenagers may be susceptible to unknown health
risks of RF radiation (Department of Health UK, 2000).
Conclusions
Proven Health Effects
• Tissue heating—at very high RF exposure levels not encountered in
normal daily living
256
CHILD HEALTH
AND THE
ENVIRONMENT
• Cell proliferation—increased at very high power-frequency magnetic
field levels not encountered in daily life
Unresolved Issues and Knowledge Gaps
• Childhood cancer—limited evidence that exposure to relatively high
power-frequency magnetic fields may cause childhood leukemia and
inadequate evidence for an association with brain tumors; inadequate
evidence to assess role of RF radiation
• Spontaneous abortion—inadequate evidence of an association with maternal magnetic field exposure
• Knowledge gaps—need epidemiologic studies with improved exposure assessment and statistical power to detect potential health effects
of low-level exposures to power-frequency magnetic fields and RF
radiation
Risk Management Issues
• Exposure reduction—prudent to minimize exposure to environmental
sources of power-frequency magnetic fields and RF radiation, particularly among pregnant women and children
• Biomonitoring—need to monitor power-frequency magnetic fields and
RF radiation exposure among children in homes and other settings to
measure progress in reducing exposure levels and to identify high-risk
groups for targeted actions including research and intervention
III. SUNLIGHT
Solar ultraviolet (UV) radiation has been subdivided into UVA (320–
400 nm), UVB (290–320 nm), and UVC (200–290 nm). Over 90% of total
ground-level UV radiation is UVA because UVC and about 90% of UVB
are absorbed by ozone, water vapor, oxygen, and carbon dioxide as sunlight passes through the earth’s atmosphere. Ground-level UVA radiation
is most intense in the early morning and afternoon, passes through glass,
penetrates skin to the dermis, and causes modest tanning and wrinkling
(Ferrini et al., 1998). Ultraviolet B radiation is most intense at midday and
causes sunburn and tanning but cannot penetrate glass. Concern about
sun exposure and child health comes from evidence of rapidly increasing
incidence rates of malignant melanoma (3%–-7% per year from the mid1960s to the mid-1980s in many Caucasian populations), discovery of the
Radiation
257
importance of intense sunlight exposures during childhood and youth as
causes of melanoma, and evidence of thinning of the earth’s stratospheric
ozone layer. About 80%–90% of melanomas and nonmelanomas are
caused by exposure to sunlight (International Agency for Research on
Cancer, 2001).
The objective of Part III of this chapter is to define the importance of
sun exposure during childhood and adolescence on the risk of benign
nevi (moles) and skin cancer, with a focus on malignant melanoma. The
genotoxic effects of UV light, the role of intermittent intense sun exposure and host factors in skin cancer, and risk management strategies are
explored.
Health Effects
Small amounts of UV radiation are beneficial for dermal activation of vitamin D, but intense exposure to sunlight during childhood can cause
sunburn, common and atypical nevi (moles), and freckling and can increase the risk of delayed effects including melanoma and basal cell skin
cancers, other skin changes, and possibly cataracts.
Molecular Mechanisms
Ultraviolet A, UVB, and UVC all induce DNA damage including chromosomal aberrations, sister chromatid exchange, and mutations in human cells in vitro and can induce or enhance cell and viral gene expression. Ultraviolet B, a direct genotoxin, stimulates photochemical reactions
in DNA that produce cyclobutane pyrimidine dimers and pyrimidinepyrimidone (6-4) photoproducts. In human volunteers, UVA exposure
caused increased urinary excretion of 8-oxo-2-deoxyguanosine and cyclobutane thymine dimers, which peaked 3–4 days after exposure; the
former metabolite indicates oxidative DNA damage, probably from UVAinduced free radicals, and the latter metabolite indicates thymidine dimer
formation, possibly due to interaction of UVA with endogenous sensitizers (Cooke et al., 2001). The importance of DNA repair systems in reducing the genotoxicity of UV radiation is illustrated by the genetic disease
xeroderma pigmentosum; affected persons have greatly increased risks of
melanoma and other skin cancers because of defects in either of two DNA
repair systems.
Inactivation of the p53 tumor suppressor gene is a key molecular
change in skin cancer and many other types of cancer. Molecular studies
of human melanoma cells indicate that p53 activation in response to UV
258
CHILD HEALTH
AND THE
ENVIRONMENT
damage is diminished and the regulation of its target genes is disrupted.
About a third of p53 mutations in human sun-related skin cancers occur
at trinucleotide sequences containing the rare base 5-methylcytosine; experimental studies show that simulated sunlight preferentially causes mutations involving dimers at dipyrimidine sequences with 5-methylcytosine.
Precancerous Skin Lesions
Although benign, nevi are important because at least 20%–30% of melanomas appear to arise in preexisting common or dysplastic nevi. Epidemiologic studies of nevi have shown the importance of a low latitude
of residence, recreational sun exposure, and frequent sunburns during
childhood and youth. The prevalence of large numbers of nevi per person is associated with frequent intense sun exposure and sunburn. Atypical (or dysplastic) nevi are generally bigger than a pencil eraser and have
irregular shapes and pigmentation; these lesions appear to be most closely
associated with intense sun exposure and sunburns during childhood.
Other risk factors for nevi include fair skin and hair color, freckling, and
difficult tanning. Density (number per unit area of skin) of nevi less than
5 mm in diameter tends to be highest on frequently exposed skin (face,
back, shoulders, and dorsal surfaces of arms), while density of larger nevi
is higher on intermittently exposed areas, especially the trunk.
Skin Cancer
The IARC concluded that (1) solar radiation is a known cause of cancer
in humans (melanoma and nonmelanoma skin cancers), (2) there is inadequate evidence for carcinogenicity due to fluorescent lighting in humans, (3) sunlamps, sunbeds, and UVA, UVB, and UVC are each probable human carcinogens, and (4) solar radiation, broad-spectrum UV light,
UVA, UVB, and UVC are each known causes of cancer in animals (International Agency for Research on Cancer, 1992).
Malignant Melanoma
Because of its rarity during childhood, most epidemiologic studies of melanoma have involved adults. One of the few studies of childhood melanoma showed associations with a family history of melanoma, inability
to tan, and indirect markers of sun exposure (number of nevi, facial freckling) (Whiteman et al., 1997). Important findings from systematic reviews
of epidemiologic studies of melanoma and sun exposure include (Elwood
and Jopson, 1997)
Radiation
259
• Strong associations with a history of intermittent intense sun exposure
and sunburns during childhood, adolescence, and adulthood and with
skin changes related to childhood sun exposure (freckling, nevi)
• A propensity to occur on intermittently exposed body sites (trunk and
limbs)
• Associations with evidence of genetic susceptibility including inability
to tan and lightly colored hair and eyes
Several case-control studies of melanoma published subsequent to these
reviews have confirmed the strong consistent associations between adult
melanoma and intensity of recreational sun exposure and frequency of
sunburns during childhood and youth (see, e.g., Pfahlberg et al., 2001).
The number of skin nevi is a strong predictor of the adult melanoma risk.
Ultraviolet B greatly increases the risk of melanoma in experimental animals.
Overall melanoma incidence and mortality rates increased substantially from the mid-1950s to the late 1980s in several countries, with evidence of age-specific rates stabilizing or even decreasing among post–
World War II birth cohorts in Sweden, Connecticut, Canada, New Zealand,
and Australia but not among similar birth cohorts in France, Italy, and
Czechoslovakia (see, e.g., Marrett et al., 2001). Observed mortality rate
decreases may have been partially caused by earlier diagnosis and improved survival.
Nonmelanoma Skin Cancers
The main known risk factor for basal cell carcinoma of skin is intermittent recreational sun exposure during childhood and adolescence among
persons who tend to burn rather than tan. The other main type of nonmelanoma skin cancer, squamous cell carcinoma, has generally been associated with lifelong sun exposure, but site-specific lesions have been
linked to site-specific sunburn frequency during childhood. In Finland,
one of the few countries having population-based incidence data for nonmelanoma skin cancer, basal cell and other nonmelanoma skin cancer incidence rates increased from about 1960 to 1995.
Other Health Effects
Although there is strong evidence that chronic exposure to sunlight and
occupational exposure to UV light are risk factors for cortical cataracts of
the eye, there appear to have been no epidemiologic studies of the role
of sun exposure during childhood and adolescence. Within the UV spec-
260
CHILD HEALTH
AND THE
ENVIRONMENT
trum, UVB at 297 nm is most efficient in causing human lens epithelial
cell death in vitro and is much more potent than UVA in causing posterior cortical cataracts in mice. The potential impact of sun exposure on
childhood immune function remains largely undefined. Exposure of human skin to UVA or UVB reduces T-cell-mediated contact hypersensitivity. Animal studies showed that UV-mediated suppression of contact hypersensitivity is reduced by broad-spectrum sunscreens, especially those
containing high levels of a UVA filter (Fourtanier et al., 2000). Animals
exposed to UVB have reduced immune responses including increased susceptibility to infection and tolerance of highly antigenic UV-induced skin
tumors.
Exposure
Incident UV light intensity at ground level depends on the solar zenith
angle, stratospheric ozone, atmospheric pollutants, weather, and altitude.
In a network of ground-level solar UV monitoring stations in Australia,
the main factors influencing daily total ground-level UV intensity at a
given site were cloud cover and stratospheric ozone levels (Roy et al.,
1998). Chlorofluorocarbons and bromine compounds are the main anthropogenic contaminants that reduce stratospheric ozone levels; a decrease
in stratospheric ozone of 50% at 60 degrees north (e.g., Scandinavia)
would give such regions the effective UV dose of California or Mediterranean countries.
Surveys in several countries have shown a continuing high prevalence of sunburns among children, including infants as young as 6 months
of age. Skin exposure varies by ground-level UV intensity, the presence
of reflective surfaces (e.g., water, snow), body region, behavior, and clothing. Personal UV dosimetry of children in Japan showed that average
daily exposures measured during different seasons were not directly correlated with outdoor UV intensity but were related to sporadic outdoor
activities, suggesting the dominant role of behavioral factors.
Risk Management
Prevention of adult skin cancer must start in childhood and adolescence,
when people receive a large fraction of their lifetime UV exposure. Downward trends of melanoma incidence rates among post–World War II birth
cohorts in several countries (noted above) suggest that preventive programs may have already had a favorable impact. Recent surveys of pre-
Radiation
261
ventive sun behaviors indicate that a third to a half of children used sunscreen and protective clothing while outdoors on the preceding weekend,
with a tendency for favorable behaviors to decrease with age among
adolescents.
There are uncertainties concerning the most important wavelengths
of UV and near-UV visible radiation that may contribute to the risk of
melanoma; the genotoxic portion of solar radiation may even extend to
the near-visible light range. Although sunscreens are highly effective in
blocking UVB and preventing skin erythema, not all products, especially
older ones, adequately block UVA. Given that UVA comprises over 90%
of the energy in ground-level UV radiation and is known to be genotoxic
in human cells, the public should be advised to use broad-spectrum sunscreens that block both UVA and UVB. It appears, however, that no mandated regulation exists to specify testing of sunscreens for UVA efficacy.
Testing showed that 6 of 11 products claiming to provide UVA protection
had actual effectiveness levels of 6%–52% in the UVA range. There have
been no randomized trials with long-term follow-up to assess sunscreen
efficacy during childhood in the prevention of skin cancers, but available
studies indicate the following:
• Nevi—a survey of Israeli children showed an increased risk of nevi
among sunscreen users, but a randomized trial of sunscreen use among
young children in Canada showed a significant reduction in the number of nevi in the intervention arm within 3 years
• Melanoma—there was a twofold increased risk of melanoma among
Swedes who used sunscreens regularly; the increase was mainly on the
trunk and was related to increased frequency of sunbathing.
• Nonmelanoma skin cancers and actinic keratoses—randomized trials
have shown the efficacy of sunscreens in reducing sunburn, actinic keratoses, and squamous cell but not basal cell skin cancers in adults.
Sunscreens reduce epidermal damage and the frequency of cyclobutane
pyrimidine dimers, pyrimidine-pyrimidone (6-4) photoproducts, and
photooxidative lesions in human skin in vitro models. Despite evidence
that older sunscreens prevented nonmelanoma but not melanoma skin
cancers in UVR-exposed animals, new broad-spectrum sunscreens with
higher sun protection factors (SPFs) may prove to be effective in reducing the risk of melanoma as well as other skin cancers. In any event, there
is a need to improve the appropriate use of existing sunscreens, including the amount applied and the frequency of application. Controlled trials are also needed to assess the potential value of population screening
for early detection of melanomas.
262
CHILD HEALTH
AND THE
ENVIRONMENT
The IARC concluded that protection of skin from solar damage ideally involves several actions and has recommended public health guidelines including the following (International Agency for Research on Cancer, 2001):
• Adopt multiple measures, that is, wear tightly woven protective clothing that adequately covers the arms, trunk, and legs and a hat that
shades the whole head, seek shade, avoid outdoor activities during
peak midday sun intensity, and use sunscreens.
• Educate the public that sunscreen use should not be the first or the only
choice for skin cancer prevention, as a means to prolong solar exposure, or as a substitute for clothing on usually unexposed sites such as
the trunk and buttocks (and labeling of sunscreen products to inform
consumers of this recommendation).
• Recognize that adequate solar protection during childhood is more important than at any other time in life and promote adoption of the above
recommendations by parents and school managers.
• Residents of areas with high sun intensity who work outdoors or enjoy regular outdoor recreation should use sunscreens with high SPFs
(15) daily on usually exposed skin.
• Stringent evaluation of sunscreen safety should be performed using the
same regulatory safety requirements as for pharmaceuticals, with a particular focus on potential long-term effects; data on the safety evaluation of sunscreens must be in the public domain so that they are available for independent scientific evaluation.
• Once the optimal method for specifying protection against broadspectrum UVA has been agreed upon, a labeling method should be introduced that is internationally congruent and understandable to the
public.
• Advertising for sunscreens should promote a global sun protection
strategy, avoid promoting sunscreen use for intentional exposure to the
sun, and avoid using messages likely to provide a false sense of security among users.
• Health promotion interventions should be designed to increase the appropriate and effective use of sunscreens by the general public and by
subgroups at risk of skin cancer because of their phenotype or a predisposition to intentional solar exposure.
Whereas sunbeds probably cause skin cancer, public education should
also address this issue on a precautionary basis. Finally, a broad strategy
Radiation
263
should include measures to avoid further reductions in the earth’s stratospheric ozone layer. With a 10% loss of stratospheric ozone, skin cancer
rates may rise by 30% for basal cell carcinoma and by 50% for squamous
cell carcinoma, assuming no adaptive changes in human behavior (Jones,
1992).
Conclusions
Proven Health Outcomes of Intense Sun Exposure
During Childhood
• Nevi (moles)
• Skin cancer (malignant melanoma, basal cell carcinoma) during
adulthood
Unresolved Issues and Knowledge Gaps
• Cancer—limited evidence that intermittent intense sun exposure during childhood may increase the risk of squamous cell carcinoma
during adulthood
• Cataracts—inadequate evidence to assess the role of sun exposure during childhood and the risk of cortical cataracts during adulthood
• Immune function—inadequate evidence to assess adverse effects of sun
exposure on children’s immune system
• Knowledge development—need epidemiologic studies to detect potential delayed health effects of UV exposure (immunologic, cataracts)
and sunscreen use (protective and/or adverse effects)
Risk Management Issues
• Prevention—need to reduce exposure of children to intense sun exposure through several actions including adequate clothing, use of a hat
that shades the whole head, seeking shade, avoiding outdoor activities
during peak midday sun intensity, and use of broad-spectrum sunscreens
• Monitoring—need to measure ground-level and personal UV exposure
and sun-protective behaviors of children and their guardians, that is,
to measure progress in exposure reduction and to identify high-risk
groups for targeted interventions
264
CHILD HEALTH
AND THE
ENVIRONMENT
References
Agency for Toxic Substances and Disease Registry. (1999). Toxicological profile for
ionizing radiation. Atlanta: Agency for Toxic Substances and Disease Registry.
Ahlbom IC, Cardis E, Green A, Linet M, Savitz D, Swerdlow A. (2001). Review of
the epidemiologic literature on EMF and Health. Environ Health Perspect 109
(Suppl 6):911–33.
Ahlbom A, Day N, Feychting M, Roman E, Skinner J, Dockerty J, Linet M, McBride
M, Michaelis J, Olsen JH and others. (2000). A pooled analysis of magnetic
fields and childhood leukaemia. Br J Cancer 83:692–8.
Angelillo IF, Villari P. (1999). Residential exposure to electromagnetic fields and
childhood leukaemia: a meta-analysis. Bull WHO 77:906–15.
Astakhova LN, Anspaugh LR, Beebe GW, Bouville A, Drozdovitch VV, Garber V,
Gavrilin YI, Khrouch VT, Kuvshinnikov AV, Kuzmenkov YN, Minenko VP,
Moschik KV, and others. (1998). Chernobyl-related thyroid cancer in children
of Belarus: a case-control study. Radiat Res 150:349–56.
Auvinen A, Hietanen M, Luukkonen R, Koskela RS. (2002). Brain tumors and salivary gland cancers among cellular telephone users. Epidemiology 13:356–9.
Axelson O, Fredrikson M, Akerblom G, Hardell L. (2002). Leukemia in childhood
and adolescence and exposure to ionizing radiation in homes built from uranium-containing alum shale concrete. Epidemiology 13:146–50.
Boice JD, Miller RW. (1999). Childhood and adult cancer after intrauterine exposure to ionizing radiation. Teratology 59:227–33.
Brusick D, Albertini R, McRee D, Peterson D, Williams G, Hanawalt P, Preston J.
(1998). Genotoxicity of radiofrequency radiation. DNA/Genetox Expert Panel.
Environ Mol Mutagen 32:1–16.
Coghill RW, Steward J, Philips A. (1996). Extra low frequency electric and magnetic fields in the bedplace of children diagnosed with leukaemia: a casecontrol study. Eur J Cancer Prev 5:153–8.
Cooke MS, Evans MD, Burd RM, Patel K, Barnard A, Lunec J, Hutchinson PE.
(2001). Induction and excretion of ultraviolet-induced 8-oxo-2-deoxyguanosine and thymine dimers in vivo: implications for PUVA. J Invest Dermatol
116:281–5.
Correa-Villasenor A, Ferencz C, Loffredo C, Magee C. (1993). Paternal exposures
and cardiovascular malformations. The Baltimore-Washington Infant Study
Group. J Expo Anal Environ Epidemiol 3(Suppl 1):173–85.
De Roos AJ, Teschke K, Savitz DA, Poole C, Grufferman S, Pollock BH, Olshan
AF. (2001). Parental occupational exposures to electromagnetic fields and radiation and the incidence of neuroblastoma in offspring. Epidemiology 12:
508–17.
Department of Health UK. (2000). Mobile phones and health. Located at www.
doh.gov.uk/mobile.htm.
Doll R, Wakeford R. (1997). Risk of childhood cancer from fetal irradiation. Br J
Radiol 70:130–9.
Elwood JM. (1999). A critical review of epidemiologic studies of radiofrequency
exposure and human cancers. Environ Health Perspect 107(Suppl 1):155–68.
Elwood JM, Jopson J. (1997). Melanoma and sun exposure: an overview of published studies. Int J Cancer 73:198–203.
Radiation
265
Erdreich LS, Klauenberg BJ. (2001). Radio frequency radiation exposure standards:
considerations for harmonization. Health Phys 80:430–9.
Expert panel on the potential health risks of radiofrequency fields from wireless
telecommunication devices. (1999). A review of the potential health risks of radiofrequency fields from wireless telecommunication devices. Ottawa: The
Royal Society of Canada/La Société royale du Canada.
Federal Communications Commission. (1999). Questions and answers about biological effects and potential hazards of radiofrequency electromagnetic fields. Located at http://www.fcc.gov/Bureaus/Engineering_Technology/Documents/
bulletins/oet56/oet56e4.pdf.
Ferrini RL, Perlman M, Hill L. (1998). American College of Preventive Medicine
practice policy statement: skin protection from ultraviolet light exposure. The
American College of Preventive Medicine. Am J Prev Med 14:83–6.
Finkelstein MM, Kreiger N. (1996). Radium in drinking water and risk of bone
cancer in Ontario youths: a second study and combined analysis. Occup Environ Med 53:305–11.
Fourtanier A, Gueniche A, Compan D, Walker SL, Young AR. (2000). Improved
protection against solar-simulated radiation-induced immunosuppression by
a sunscreen with enhanced ultraviolet A protection. J Invest Dermatol 114:
620–7.
Graham C, Cook MR, Gerkovich MM, Sastre A. (2001). Examination of the melatonin hypothesis in women exposed at night to EMF or bright light. Environ
Health Perspect 109:501–7.
Green LM, Miller AB, Agnew DA, Greenberg ML, Li J, Villeneuve PJ, Tibshirani
R. (1999). Childhood leukemia and personal monitoring of residential exposures to electric and magnetic fields in Ontario, Canada. Cancer Causes Control 10:233–43.
Greenland S, Sheppard AR, Kaune WT, Poole C, Kelsh MA. (2000). A pooled analysis of magnetic fields, wire codes, and childhood leukemia. Childhood Leukemia-EMF Study Group. Epidemiology 11:624–34.
Infante-Rivard C, Mathonnet G, Sinnett D. (2000). Risk of childhood leukemia associated with diagnostic irradiation and polymorphisms in DNA repair genes.
Environ Health Perspect 108:495–8.
International Agency for Research on Cancer. (1992). IARC monographs on the
evaluation of carcinogenic risks to humans. Vol. 55. Solar and ultraviolet radiation. Lyon, France: International Agency for Research on Cancer.
International Agency for Research on Cancer. (2000). IARC monographs on the
evaluation of carcinogenic risks to humans. Vol. 75. Ionizing radiation, Part 1:
X- and gamma-radiation, and neutrons. Lyon, France: International Agency
for Research on Cancer.
International Agency for Research on Cancer. (2001). Cancer-preventive effect of
sunscreens. Lyon, France: International Agency for Research on Cancer.
International Agency for Research on Cancer. (2002). IARC monographs on the
evaluation of carcinogenic risks to humans. Vol. 80. Non-ionizing radiation,
Part 1: Static and extremely low frequency electric and magnetic fields. Lyon,
France: International Agency for Research on Cancer.
International Commission on Radiological Protection. (1991). 1990 recommendations of the International Commission on Radiological Protection, ICRP Publication 60. Oxford: Pergamon Press.
266
CHILD HEALTH
AND THE
ENVIRONMENT
International Commission on Non-Ionizing Radiation Protection. (1998). Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz). Health Phys 74:494–522.
Jones RR. (1992). Ozone depletion and its effects on human populations. Br J Dermatol 127(Suppl 41):2–6.
Kaletsch U, Kaatsch P, Meinert R, Schuz J, Czarwinski R, Michaelis J. (1999). Childhood cancer and residential radon exposure—results of a population-based
case-control study in Lower Saxony (Germany). Radiat Environ Biophys 38:
211–5.
Kallen B, Karlsson P, Lundell M, Wallgren A, Holm LE. (1998). Outcome of reproduction in women irradiated for skin hemangioma in infancy. Radiat Res
149:202–8.
Kaune WT, Savitz DA. (1994). Simplification of the Wertheimer-Leeper wire code.
Bioelectromagnetics 15:275–82.
Kavet R, Zaffanella LE, Daigle JP, Ebi KL. (2000). The possible role of contact current in cancer risk associated with residential magnetic fields. Bioelectromagnetics 21:538–53.
Kheifets LI, Sussman SS, Preston-Martin S. (1999). Childhood brain tumors and residential electromagnetic fields (EMF). Rev Environ Contam Toxicol 159:111–29.
Lee GM, Neutra RR, Hristova L, Yost M, Hiatt RA. (2002). A nested case-control
study of residential and personal magnetic field measures and miscarriages.
Epidemiology 13:21–31.
Li DK, Odouli R, Wi S, Janevic T, Golditch I, Bracken TD, Senior R, Rankin R, Iriye
R. (2002). A population-based prospective cohort study of personal exposure
to magnetic fields during pregnancy and the risk of miscarriage. Epidemiology 13:9–20.
Linet MS, Ries LA, Smith MA, Tarone RE, Devesa SS. (1999). Cancer surveillance
series: recent trends in childhood cancer incidence and mortality in the United
States. J Natl Cancer Inst 91:1051–8.
Loffredo CA, Hirata J, Wilson PD, Ferencz C, Lurie IW. (2001). Atrioventricular
septal defects: possible etiologic differences between complete and partial defects. Teratology 63:87–93.
London SJ, Thomas DC, Bowman JD, Sobel E, Cheng TC, Peters JM. (1991). Exposure to residential electric and magnetic fields and risk of childhood leukemia. Am J Epidemiol 134:923–37.
MacMahon B. (1962). Prenatal X-ray exposure and childhood cancer. J Natl Cancer Inst 28:1173–91.
Marrett LD, Nguyen HL, Armstrong BK. (2001). Trends in the incidence of cutaneous malignant melanoma in New South Wales, 1983–1996. Int J Cancer 92:
457–62.
McBride ML, Gallagher RP, Theriault G, Armstrong BG, Tamaro S, Spinelli JJ,
Deadman JE, Fincham S, Robson D, Choi W. (1999). Power-frequency electric
and magnetic fields and risk of childhood leukemia in Canada. Am J Epidemiol
149:831–42.
Meinert R, Kaletsch U, Kaatsch P, Schuz J, Michaelis J. (1999). Associations between childhood cancer and ionizing radiation: results of a population-based
case-control study in Germany. Cancer Epidemiol Biomarkers Prev 8:793–9.
Meinert R, Michaelis J. (1996). Meta-analyses of studies on the association between
electromagnetic fields and childhood cancer. Radiat Environ Biophys 35:11–8.
Radiation
267
Miller MA, Murphy JR, Miller TI, Ruttenber AJ. (1995). Variation in cancer risk
estimates for exposure to powerline frequency electromagnetic fields: a metaanalysis comparing EMF measurement methods. Risk Anal 15:281–7.
National Academy of Sciences. (1990). Biological effects of ionizing radiation
(BEIR) V report: Health effects of exposure to low levels of ionizing radiation.
Washington, DC: National Academy Press.
National Academy of Sciences. (1997). Possible health effects of exposure to residential electric and magnetic fields. Washington, DC: National Academy Press.
National Academy of Sciences. (1999). Biological effects of ionizing radiation
(BEIR) VI report: The health effects of exposure to Indoor radon. Washington,
DC: National Academy Press.
National Institute of Environmental Health Sciences. (1998). Assessment of health
effects from exposure to power-line frequency electric and magnetic fields.
Working group report. (NIH Publication No. 98–3981). Research Triangle Park,
NC: National Institute of Environmental Health Sciences.
National Institute of Environmental Health Sciences. (1999). Report on health effects from exposure to power-line frequency electric and magnetic fields. (NIH
Publication No. 99-4493). Research Triangle Park, NC: National Institute of Environmental Health Sciences.
Naumburg E, Bellocco R, Cnattingius S, Hall P, Boice JD Jr, Ekbom A. (2001). Intrauterine exposure to diagnostic X rays and risk of childhood leukemia subtypes. Radiat Res 156:718–23.
Parazzini F, Luchini L, La Vecchia C, Crosignani PG. (1993). Video display terminal use during pregnancy and reproductive outcome—a meta-analysis. J Epidemiol Commun Health 47:265–8.
Parker L, Pearce MS, Dickinson HO, Aitkin M, Craft AW. (1999). Stillbirths among
offspring of male radiation workers at Sellafield nuclear reprocessing plant.
Lancet 354:1407–14.
Pfahlberg A, Kolmel KF, Gefeller for the Febim Study Group O. (2001). Timing of
excessive ultraviolet radiation and melanoma: epidemiology does not support
the existence of a critical period of high susceptibility to solar ultraviolet
radiation-induced melanoma. Br J Dermatol 144:471–5.
Pierce DA, Shimizu Y, Preston DL, Vaeth M, Mabuchi K. (1996). Studies of the
mortality of atomic bomb survivors. Report 12, part I. Cancer: 1950–1990. Radiat Res 146:1–27.
Repacholi MH. (2001). Health risks from the use of mobile phones. Toxicol Lett
120:323–31.
Ron E, Lubin JH, Shore RE, Mabuchi K, Modan B, Pottern LM, Schneider AB,
Tucker MA, Boice JD. (1995). Thyroid cancer after exposure to external radiation: a pooled analysis of seven studies. Radiat Res 141:259–77.
Roy CR, Gies HP, Lugg DJ, Toomey S, Tomlinson DW. (1998). The measurement
of solar ultraviolet radiation. Mutat Res 422:7–14.
Royal Society of Canada. (1999). A review of the potential health risks of radiofrequency fields from wireless telecommunication devices. Ottawa: Royal
Society of Canada.
Rutledge JC. (1997). Developmental toxicity induced during early stages of mammalian embryogenesis. Mutat Res 396:113–27.
Savitz DA. (1995). Exposure assessment strategies in epidemiological studies of
health effects of electric and magnetic fields. Sci Total Environ 168:143–53.
268
CHILD HEALTH
AND THE
ENVIRONMENT
Savitz DA, Kaune WT. (1993). Childhood cancer in relation to a modified residential wire code. Environ Health Perspect 101:76–80.
Savitz DA, Poole C. (2001). Do studies of wire code and childhood leukemia point
towards or away from magnetic fields as the causal agent? Bioelectromagnetics Suppl 5:S69–85.
Savitz DA, Whelan EA, Kleckner RC. (1989). Self-reported exposure to pesticides
and radiation related to pregnancy outcome—results from National Natality
and Fetal Mortality Surveys. Public Health Rep 104:473–7.
Schuz J, Grigat JP, Stormer B, Rippin G, Brinkmann K, Michaelis J. (2000). Extremely low frequency magnetic fields in residences in Germany. Distribution
of measurements, comparison of two methods for assessing exposure, and predictors for the occurrence of magnetic fields above background level. Radiat
Environ Biophys 39:233–40.
Sever LE, Gilbert ES, Hessol NA, McIntyre JM. (1988). A case-control study of congenital malformations and occupational exposure to low-level ionizing radiation. Am J Epidemiol 127:226–42.
Shaw GM. (2001). Adverse human reproductive outcomes and electromagnetic
fields: a brief summary of the epidemiologic literature. Bioelectromagnetics
Suppl 5:S5–18.
Shea KM, Little RE. (1997). Is there an association between preconception paternal X-ray exposure and birth outcome? The ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. Am J Epidemiol 145:546–51.
Shibata Y, Yamashita S, Masyakin VB, Panasyuk GD, Nagataki S. (2001). 15 years
after Chernobyl: new evidence of thyroid cancer. Lancet 358:1965–6.
Shu XO, Potter JD, Linet MS, Severson RK, Han D, Kersey JH, Neglia JP, Trigg
ME, Robison LL. (2002). Diagnostic X-rays and ultrasound exposure and risk
of childhood acute lymphoblastic leukemia by immunophenotype. Cancer
Epidemiol Biomarkers Prev 11:177–85.
Stang A, Anastassiou G, Ahrens W, Bromen K, Bornfeld N, Jockel KH. (2001). The
possible role of radiofrequency radiation in the development of uveal melanoma. Epidemiology 12:7–12.
Stewart A, Webb J, Hewitt D. (1958). A survey of childhood malignancies. Br Med
J 1:1495–1508.
United Nations Scientific Committee on the Effects of Atomic Radiation. (2000).
UNSCEAR 2000 report. Sources and effects of ionizing radiation. UNSCEAR
2000 report to the General Assembly, with scientific annexes. Vienna: United
Nations Scientific Committee on the Effects of Atomic Radiation.
U.S. Environmental Protection Agency. (1993). Protocols for radon and radon
decay product measurements in homes (EPA 402-R-93-003). Located at
http://www.epa.gov/RadonPubs/homprot1.html.
U.S. Environmental Protection Agency. (2001). Current drinking water standards.
Located at http://www.epa.gov/safewater/mcl.html.
U.S. Environmental Protection Agency. (2002). Radon resistant new construction.
Located at http://www.epa.gov/iaq/radon/index.html.
Wartenberg D. (2001). Residential EMF exposure and childhood leukemia: metaanalysis and population attributable risk. Bioelectromagnetics S5:S86–S104.
Washburn EP, Orza MJ, Berlin JA, Nicholson WJ, Todd AC, Frumkin H, Chalmers
TC. (1994). Residential proximity to electricity transmission and distribution
equipment and risk of childhood leukemia, childhood lymphoma, and child-
Radiation
269
hood nervous system tumors: systematic review, evaluation, and meta-analysis.
Cancer Causes Control 5:299–309.
Wertheimer N, Leeper E. (1979). Electrical wiring configurations and childhood
cancer. Am J Epidemiol 109:273–84.
Whiteman DC, Valery P, McWhirter W, Green AC. (1997). Risk factors for childhood melanoma in Queensland, Australia. Int J Cancer 70:26–31.
Zanzonico PB, Becker DV. (2000). Effects of time of administration and dietary iodine levels on potassium iodide (KI) blockade of thyroid irradiation by 131I
from radioactive fallout. Health Phys 78:660–7.
Zhang J, Cai WW, Lee DJ. (1992). Occupational hazards and pregnancy outcomes.
Am J Ind Med 21:397–408.
Zheng T, Holford TR, Chen Y, Ma JZ, Flannery J, Liu W, Russi M, Boyle P. (1996).
Time trend and age-period-cohort effect on incidence of thyroid cancer in Connecticut, 1935–1992. Int J Cancer 67:504–9.
10
Indoor Air
Young children in developed countries spend over 90% of their time in
homes, day-care centers, schools, motor vehicles, and other indoor environments. The EPA’s Science Advisory Board over the past several years
has consistently ranked indoor air pollution among the top five risks to
public health in the United States. Sources of indoor air contaminants include occupant behaviors (especially smoking), water, building materials,
consumer products, pets, insects, fungi, inadequately ventilated cooking
and heating devices, and influx of outdoor air pollutants. Children inhale
relatively high amounts of air per unit body weight per day and play at
floor level, where concentrations of airborne contaminants may be relatively high. Airborne contaminants can disrupt lung growth and function
during early childhood, causing persistent functional deficits and respiratory diseases.
The major categories of indoor air contaminants are gases/vapors
and particulate matter, ranging from small, respirable particles to large,
transiently suspended particles including house dust (Table 10–1). Smoking is the main source of airborne fine particulate matter (PM2.5) in homes
with smokers; other sources include other indoor combustion processes
and infiltration of outdoor air. Larger particles can originate from indoor
materials (e.g., textiles, clothing), biologic sources (pets, insects, fungi),
270
TABLE 10–1. Types of Indoor Air Contaminants
Category
Examples
Sources
Gases and vapors
Carbon monoxide, nitrogen dioxide, formaldehyde,
radon, volatile organic chemicals, pesticides
Influx of outdoor air, fuel combustion (cooking, heating), building
materials, geologic formations (radon), solvents, cleaning agents,
chlorinated water
Particulate matter
Toxicants—ETS, particulate matter from other
combustion processes, asbestos
Biologics—animal dander, fungal spores, bacteria,
viruses, pollens, arthropod antigens
Tobacco smoking, fuel combustion (cooking, heating), influx of
outdoor air, building materials, furnishings
Pets, molds (closely linked to dampness), insects, humans
Dusta
Pesticides, heavy metals
Interior use of pesticides, influx of outdoor air, tracking in of
contaminated soil/dust
a
Included because airborne contaminants often deposit and accumulate in house dust and carpets, on furnishings, and on other surfaces.
272
CHILD HEALTH
AND THE
ENVIRONMENT
and soil tracked indoors from external sources. Economically disadvantaged children generally have higher exposures to indoor air hazards because of a higher prevalence of household smokers and microenvironmental conditions such as dampness, inadequate ventilation, and building
deterioration that favor proliferation of molds, house-dust mites, and
cockroaches.
The purpose of this chapter is to describe the known and suspected
health impacts of indoor air contaminants, with a major focus on environmental tobacco smoke (ETS), biologic agents, and volatile organic
chemicals (VOCs). Topics include known and suspected links between indoor air pollutants and fetal development, sudden infant death syndrome
(SIDS), respiratory diseases, and cancer. Available data on exposure levels for ETS and major VOCs document high-risk groups. The section on
risk management notes the important role of children’s parents and caregivers to provide a safe indoor environment but also points to the need
for regulatory measures. See Chapters 4, 5, 7, 9, 11, and 12 for discussion
of other potential indoor air contaminants including heavy metals, pesticides, radon, ambient air pollutants, and volatile disinfection by-products.
Asthma
This section describes the general characteristics of asthma, the most common chronic disease among children in developed countries, to facilitate
subsequent discussion of various airborne agents in this important childhood condition. Asthma has two main characteristics: chronic inflammation of the airways and recurrent episodes of wheezing and coughing.
The etiology of asthma appears to involve the interaction of multiple
genes, environmental allergens, and chemical toxicants. Genes possibly
important in the onset of this disease include those for cytokine/
chemokine and IgE regulation, glutathione-S-transferase, and 5-lipoxygenase (see, e.g., Ober and Moffatt, 2000). Asthma prevalence rates among
children have increased substantially during recent decades; in Canada,
for instance, the prevalence of physician-diagnosed childhood asthma increased from 2.5% to 11.2% during 1978–1995 (Millar and Hill, 1998). Although the presence and degree of airway inflammation determine the
severity of asthma, the causes of airway inflammation remain uncertain
(National Academy of Sciences, 2000).
The main types of childhood airway obstruction involve episodic
wheezing (1) after bronchiolitis during infancy, often associated with maternal smoking and low birth weight, or (2) after sensitization to allergens, the dominant type among children aged 5–18 years. About 20% of
Indoor Air
273
children diagnosed with asthma before age 5 years may be asymptomatic
by age 10 years, but many still have significant airway hyperreactivity.
Diagnosis is facilitated by the use of standard criteria such as the CDC
case definition for asthma.1 Symptom questionnaires appear to best
balance sensitivity and specificity for asthma prevalence studies, while
highly specific diagnostic criteria are better suited for etiologic studies
(Pekkanen and Pearce, 1999).
Asthma risk factors include those that cause the development of
asthma (i.e., incident asthma) and those that exacerbate established
asthma. Epidemiologists have studied prevalent asthma much more frequently than incident disease. The U.S. National Academy of Sciences
(NAS) reviewed indoor air contaminants and assigned levels of evidence
for their roles as causes of asthma development or as triggers of episodes
in known asthmatics (Table 10–2). The NAS concluded that (1) house-dust
mite antigen is the only known cause of asthma development, (2) ETS is
associated with asthma development but causality remains uncertain, and
(3) cat, cockroach, and house-dust mite antigens and ETS trigger episodes
in known asthmatics.
Analysis of NHANES III showed that the risk of physician-diagnosed
asthma in young children was approximately doubled for those with a
family history of allergic disease, ETS exposure, home use of a gas stove
or oven for heat, or the presence of a dog in the household; the population attributable risk for children with one or more risk factors was 39%
(Lanphear et al., 2001). Recent evidence indicates that early childhood infections do not protect against asthma, arguing against the hypothesis that
a reduced risk of such infections may have contributed to increased
asthma prevalence rates (McKeever et al., 2002). Nonenvironmental risk
factors (e.g., obesity) may also contribute to asthma (von Mutius et al.,
2001). Asthma is discussed further below and in Chapter 11 (Outdoor Air);
readers may consult other sources for more detailed information on
asthma (National Academy of Sciences, 2000; Pearce et al., 1998).
1 (1) The presence of wheezing lasting for 2 or more consecutive days that responds to
bronchodilators, or a chronic cough that persists for 3–6 weeks in the absence of allergic rhinitis or sinusitis, or nocturnal awakening with dyspnea, cough, and/or wheezing in the absence of other medical conditions known to cause these symptoms and
(2) a 12% increment in FEV1 and/or FVC (See Lung Function later in Chapter for explanation of these terms) after inhaling a short-acting bronchodilator, or a 20% decrement in FEV1 after a challenge by histamine, methacholine, exercise, or cold air, or a
20% diurnal variation in peak expiratory flow over 1 to 2 weeks. A confirmed case is
a person who has had any of the clinical symptoms at least three times during the past
year and had at least one of the laboratory criteria.
TABLE 10–2. Summary of Indoor Environmental Risk Factors for Asthma
Level of Evidence
Factors That Cause Development of Asthma
Factors That Precipitate Episodes in Known Asthmatics
Sufficient evidence of a
causal relationship
House-dust mite antigens
Cat, cockroach, and house-dust mite antigens in specifically
sensitized persons, ETS (preschool-age children)
Sufficient evidence of an
association
ETS (preschool-age children)
Dog, rodent, and fungal antigens in specifically sensitized persons,
rhinovirus, nitrogen dioxide from indoor sources
Limited or suggestive
evidence of an association
Cockroach antigen (preschool-age children),
respiratory syncytial virus
Domestic birds, C. pneumoniae, M. pneumoniae, respiratory syncytial
virus, formaldehyde, fragrances, ETS (school-age children)
Inadequate or insufficient
evidence to determine
whether or not an
association exists
Cat, cow, horse, dog, domestic bird, fungal, and
rodent antigens, endotoxins, molds, rhinovirus
(infants), Chlamydia pneumoniae, C. trachomatis,
Mycoplasma pneumoniae, houseplants, indoor
pollens, nitrogen dioxide (from indoor sources),
pesticides, plasticiziers, VOCs, formaldehyde,
fragrances, ETS (school-age children)
Cow and horse antigens, down pillows, C. trachomatis, endotoxins,
houseplants, indoor pollens, insects other than cockroaches or
house-dust mites, pesticides, plasticizers, VOCs
Limited or suggestive
evidence of no association
Rhinovirus (school-age children)
No agents
Source: National Academy of Sciences (2000).
Indoor Air
275
Environmental Tobacco Smoke
Health Effects
Environmental tobacco smoke is a common indoor air contaminant
and an important preventable cause of childhood illnesses and deaths
(Table 10–3). The EPA concluded that ETS
• Causes lung function deficits, respiratory tract irritation (cough, sputum
production, wheezing), lower respiratory tract infections, middle ear infections, and increased frequency and severity of asthmatic episodes
• Likely causes 8000–26,000 new asthma cases annually and a substantial fraction of SIDS deaths
Reviews since the 1992 EPA report was published support these conclusions and indicate that ETS likely increases the risk of childhood meningococcal infections (see, e.g., California Environmental Protection Agency,
1997). The estimated annual direct and indirect costs of the health effects
of parental smoking in the United States were $13 billion in 1993 dollars.
TABLE 10–3. Child Health Impacts of Parental Smoking
in the United States
Category
Attributable Events (per year)
Low birth weight
(2500 g)
46,000 cases
2,800 perinatal deaths
SIDS
2,000 deaths
Bronchiolitis
(age 2 yr)
22,000 hospitalizations
1,100 deaths
Middle ear infections
(age 15 yr)
3.4 million outpatient visits
Tympanostomies
(age 15 yr)
110,000
Asthma
(age 18 yr)
1.8 million outpatient visits
28,000 hospitalizations
14 deaths
Burnsa
10,000 outpatient visits
590 hospitalizations
250 deaths
Source: Aligne and Stoddard (1997).
a Fires
initiated by smoking materials.
CHILD HEALTH
276
AND THE
ENVIRONMENT
Mechanisms of Toxicity
Particle deposition. Inhaled particles may be deposited in the (1)
mouth, nose, and larynx, (2) tracheobronchial airways, and (3) terminal
bronchioles and alveoli of the lungs. The first two regions have high flow
rates, and large and some very small particles are deposited there by impaction; fine particles are deposited mainly by diffusion and sedimentation in regions with low airflow rates, that is, the terminal bronchioles
and alveoli. Deposited particles may dissolve in body fluids, may be
cleared by the mucociliary system, or may be transported to lymph nodes.
Genotoxicity. Tobacco smoke contains over 40 known carcinogens.
Environmental tobacco smoke originates mainly from sidestream smoke;
the lower temperatures of idling cigarettes cause less complete combustion and higher concentrations of most carcinogens in sidestream smoke
compared to mainstream smoke (Table 10–4). In the absence of smoking,
indoor air particulate matter has little detectable mutagenic activity; after smoking, however, genotoxic activity (mutagenicity and sister chromatid exchange) increases markedly. Levels of blood cotinine (a nicotine
metabolite) and ETS-related adducts (4-aminobiphenyl-hemoglobin and
PAH-albumin) in young children are associated with ETS exposure. Cord
blood lymphocytes from infants prenatally exposed to maternal smoking
had hypoxanthine-guanine phosphoribosyltransferase (HPRT) gene dele-
TABLE 10– 4. Ratios of Sidestream to Mainstream
Concentrations of Known and Probable Human
Carcinogens in Tobacco Smoke
SS/MS Ratio
Vapor phase
Benzene
Hydrazine
Formaldehyde
N-nitrosodimethylamine
5–10
3
0.1–50
20–100
Particulate phase
Total particulate phase
Benzo(a)pyrene
2-naphthylamine
4-aminodiphenyl
Cadmium
Nickel
Polonium-210
Source: Wigle et al. (1987).
1.3–1.9
2.5–3.5
30
31
7.2
13–30
1–4
Indoor Air
277
tions similar to those associated with childhood acute lymphatic leukemia (e.g., the t(4;11) translocation) (Finette et al., 1998).
Developmental Effects
Maternal ETS exposure has been linked to an increased risk of spontaneous abortion, particularly late events (see, e.g., Windham et al., 1992).
Two recent reviews concluded that maternal ETS exposure is associated
with average birth weight reductions of 25–90 g (adjusted for gestation
length) and a pooled odds ratio for IUGR of 1.2 (CI 1.1–1.3) (Misra and
Nguyen, 1999; Windham et al., 1999). Several studies published since these
reviews showed links between maternal ETS exposure and preterm birth
(see, e.g., Dejmek et al., 2002). Only a few of the studies reporting links
between prenatal ETS exposure among nonsmoking women and adverse
pregnancy outcomes (spontaneous abortions, low birth weight, preterm
delivery) have used biomarkers of exposure (e.g., cotinine) during early
or late pregnancy. Polymorphisms in CYP1A1 and GSTT1 have been
linked to marked birth weight deficits among women who smoked during pregnancy (Wang et al., 2002); the potential role of such polymorphisms in the risk of low birth weight among ETS-exposed nonsmoking
women is not known. Height at age 5 years was not related to prenatal
ETS exposure status among children of nonsmoking mothers (Eskenazi
and Bergmann, 1995).
Lung Function
Spirometry. Depending on host characteristics and exposure intensity
and duration, air pollutants may cause acute and reversible or chronic
and persistent reductions in lung volumes and flow rates measured by
spirometric tests including FEV1 (forced expiratory volume 1, the volume
in liters of air that is forcefully exhaled in 1 second), FVC (forced vital capacity, the maximum volume of air that can be exhaled after full inspiration), FEV1/FVC (ratio of FEV1 to FVC, expressed as a percentage), MMEF
(also referred to as FEF25-75, the maximum midexpiratory flow rate in milliliters per second during the middle half of a FVC test), FEF75 (forced expiratory flow rate in milliliters per second at 75% of FVC), and PEF (peak
expiratory flow rate, the peak flow rate in milliliters per second during
expiration).
Deficits of 20% or more in FEV1 and MMEF, respectively, indicate
large or small airway obstruction. Measured flow rates may be compared
to population norms for children of the same age, sex, and height. Assessment may also include measurement of FEV1 before and after inhalation of a -agonist bronchodilator; an increase of 0.2 L (or 15% of baseline FEV1) after the use of a bronchodilator indicates the presence of
reversible airflow obstruction characteristic of asthma.
CHILD HEALTH
278
AND THE
ENVIRONMENT
Lung function deficits. Disruption of the signaling systems that control normal lung morphogenesis and growth during critical prenatal and
childhood time periods can cause irreversible structural and/or functional
deficits (Pinkerton and Joad, 2000). Lung volumes and flow rates grow
until about age 25 years, plateau for several years, and then gradually decline. Although small lung function deficits generally do not cause symptoms or clinically obvious disability, they provide objective measures of
preclinical respiratory tract damage. A meta-analysis of 22 surveys of
school-age children exposed to parental smoking concluded that ETS
exposure is associated with lung function deficits among asthmatic and
healthy children, with evidence of exposure-risk relations (Table 10–5)
(Cook et al., 1998). Longitudinal studies and investigations using biomarkers of ETS exposure have confirmed links between childhood ETS
exposure and lung function deficits (see, e.g., Li et al., 2000; Mannino et
al., 2001). A meta-analysis of 10 epidemiologic studies showed a significant association between maternal smoking and bronchial hyperreactivity (pooled OR 1.3, CI 1.1–1.5) (Cook and Strachan, 1998). The biologic
plausibility of these findings is shown by similar findings in experimental animal studies (Witschi et al., 1997).
Asthma
A meta-analysis of six longitudinal studies concluded that incident asthma
development, especially before age 7 years, and asthma severity were associated with maternal smoking (Strachan and Cook, 1998b). It appears
that prenatal maternal smoking and childhood ETS exposure are independently related to the risk of childhood asthma (Infante-Rivard et al.,
1999; McGready et al., 2001). The NAS concluded that ETS: (1) causes exacerbations of asthma in preschool-age children, (2) is associated with
new-onset asthma in young children, and (3) is a possible cause of asthma
exacerbations in older children (National Academy of Sciences, 2000). The
TABLE 10–5. Meta-Analysis of
Parental Smoking and Children’s
Lung Function
Parameter
Average Deficit a
FEV1
1.4% (1.0–1.9)
FEF25-75
5.0% (3.3–6.6)
FEF75
4.3% (3.1–5.5)
Source: Cook et al. (1998).
a Reduction
for children exposed to parental
smoking compared to unexposed children.
Indoor Air
279
Academy stated that there is inadequate evidence to determine if ETS is
associated with new-onset asthma in school-age children.
Other Respiratory Diseases
In a pooled analysis of lower respiratory illness before age 3 years, all but
one of 38 epidemiologic studies showed significant associations with
smoking by either parent (pooled OR 1.6, CI 1.4–1.7), and most studies
showed exposure–risk relationships (Strachan and Cook, 1997). An independent meta-analysis indicated that ETS exposure during early childhood approximately doubled the risk of lower respiratory infections requiring hospitalization (Li et al., 1999). Environmental tobacco smoke
exposure alone (from fathers and/or other household members) is a risk
factor for respiratory illness among infants whose mothers never smoked.
The frequency and duration of otitis media among young children are
also strongly associated with ETS exposure; a meta-analysis of 42 epidemiologic studies yielded pooled odds ratios of 1.5 (CI 1.1–2.0) for recurrent middle ear infection and 1.4 (CI 1.2–1.6) for middle ear effusion
among children exposed to parental smoking (Strachan and Cook, 1998a).
Cancer
Based mainly on studies of lung cancer among adults exposed to ETS for
many years, the EPA determined that there is sufficient evidence that ETS
causes cancer in humans (U.S. Environmental Protection Agency, 1992).
Some 50 epidemiologic investigations of childhood cancer have shown
that maternal smoking during pregnancy and postnatal ETS exposure are
weakly associated with brain tumors, leukemia, and lymphomas (Sasco
and Vainio, 1999). A meta-analysis showed moderately strong associations
between paternal but not maternal smoking and non-Hodgkin’s lymphoma, acute lymphatic leukemia, and brain tumors (Boffetta et al., 2000).
The association with paternal smoking does not necessarily indicate a link
to direct ETS exposure; the mechanism could involve paternal germ cell
mutations. There is inadequate evidence to assess reported links between
ETS exposure during childhood and breast cancer and nasopharyngeal
cancer in adults. Although paternal smoking was associated with childhood cancers, no tobacco smoke carcinogen has been tested in animals
with preconceptual exposure of male parents (Anderson et al., 2000).
Sudden Infant Death Syndrome
Sudden infant death syndrome is a diagnosis of exclusion based on the
sudden unexpected death of an infant usually aged 2–5 months with no
adequate cause of death identified at autopsy; SIDS likely includes deaths
that have heterogeneous etiologies. A meta-analysis of 39 studies of SIDS
yielded pooled, adjusted odds ratios of 2.1 (CI 1.8–2.4) for prenatal ma-
280
CHILD HEALTH
AND THE
ENVIRONMENT
ternal smoking and 1.9 (CI 1.6–2.4) for postnatal maternal smoking (Anderson and Cook, 1997). Evidence includes the following: (1) consistent,
strong, dose-related associations with maternal smoking during pregnancy were found, (2) infants of women who smoked prenatally and postnatally had the greatest risk of SIDS, and (3) maternal and household
smoking intensity were independently associated with SIDS. Evidence on
SIDS and prenatal maternal smoking meets these criteria for causality:
strength of association, exposure–risk relationship, consistency across
studies and study designs, biologic plausibility, and appropriate time
relationship.
Exposures
Self-reported information, indoor air monitoring (air nicotine and fine particulate levels), and biomonitoring are useful for estimating ETS exposure. In homes and public buildings where smoking takes place, ETS is
the main source of indoor fine particulate matter, with 24-hour average
PM2.5 levels increasing by about 1 g/m3 per cigarette smoked per day.
During recent years, about 30% of nonsmoking Canadian children have
experienced ETS exposure, primarily at home. Nicotine is almost specific
to tobacco, and its major proximate metabolite, cotinine, is a very specific
and sensitive biomarker of recent ETS exposures, having a half-life of
7–40 hours in urine, saliva, and blood. Cord serum cotinine, the most useful biomarker of fetal exposure during late pregnancy, distinguishes between mothers who smoked, who were passively exposed to ETS, and
who had neither exposure. Maternal smoking is the major determinant
of urinary and salivary cotinine and hair nicotine levels among young
children. Hair nicotine level reflects adsorption of ETS onto hair but is
rarely used in epidemiologic studies. During the period 1990–1999, the
median plasma cotinine level among nonsmokers in the United States decreased fourfold (from 0.20 to 0.05 g/L) (Centers for Disease Control
and Prevention, 2001). Despite this progress, more than half of American
youth are still exposed and have higher plasma cotinine levels than older
persons. A known human carcinogen in ETS, 4-aminobiphenyl, forms
adducts with hemoglobin, with a half-life of about 4 months.
Most epidemiologic studies of particulate matter have used measures
of particle size and concentration, not toxicity. Although there are no calibration standards for measurement of the suspended particle mass, existing methods identify relative exposure levels and time trends. Indoor
PM0.1 and PM2.5-10 concentrations vary substantially with brief peaks related to indoor activities. Peak indoor air and personal PM exposure levels are associated with ETS, influx of outdoor air, cooking, vacuuming,
Indoor Air
281
movement (resuspended particles), wood smoke, and motor vehicle
travel. Freestanding kerosene heaters add about 40 g/m3 of PM2.5 to
background residential levels. Ambient PM2.5 from motor vehicles, fuel
combustion, incineration, and industries can permeate homes, school
buses, and other indoor environments frequented by children, resulting
in indoor/outdoor PM2.5 ratios of 0.4 to 0.8, depending on particle size
and density, the air exchange rate, and the surface-to-volume ratio of the
indoor environment.
Asbestos fibers have been detected in air and dust in homes, schools,
and other public buildings built approximately during the period 1920–
1977 [the Consumer Product Safey Commission (CPSC) and the EPA
started to restrict asbestos use during the 1970s]. Parents occupationally exposed to asbestos can carry fibers home on their clothing. Although the cancer risks associated with such exposures during childhood have not been assessed systematically, there have been sporadic
reports of childhood mesothelioma linked to parental occupational exposure, presumably caused by asbestos contamination of residential indoor air.
Risk Management
Children are more likely than adults to develop health effects from ETS
exposure, and the home is their most important exposure source. Although
measures have variably been adopted to control ETS exposure in public
places, transportation, schools, day-care centers, and workplaces, the protection of children from exposure at home has received relatively little attention. Only California meets the nation’s Healthy People 2010 objective
of eliminating exposure to ETS by either banning indoor smoking or limiting it to separately ventilated areas. Community programs and clinical
interventions can reduce children’s exposures to ETS at home; for instance,
ETS exposure of adolescents living with adult smokers was reduced in
homes with parent-imposed smoking restrictions (Biener et al., 1997).
Biologic Agents
Health Effects
Immunologic Sensitization and Inflammation
In sensitized persons, IgE is the main mediator of allergic responses to inhaled allergens, and serum IgE levels are associated with the development and severity of asthma. Allergen–IgE complexes bind to immune
282
CHILD HEALTH
AND THE
ENVIRONMENT
system cells, triggering release of proinflammatory agents that cause early
(airway constriction, mucus production, and airway edema) and delayed
effects (prolonged bronchoconstriction) (National Academy of Sciences,
2000). Allergen priming of Th cells begins prenatally because of transplacental transport of very small amounts of antigens. By age 5 years, most
children convert to the normal adult Th1 cytokine pattern and respond to
allergen exposure with moderate IgG and IgA production and no symptoms. If the dominant Th cells produce the Th2 cytokine pattern, especially
interleukin-4 (IL-4) and IL-5, antigen exposure will trigger IgE production, eosinophilia, and allergic symptoms.
Asthma
This section focuses on biologic agents in the home environment that are
potential causes of asthma development or triggers of episodes in known
asthmatics. Rhinovirus infection appears to trigger episodes among known
asthmatics, and there is limited evidence for associations between respiratory syncytial virus, Chlamydia pneumoniae, and Mycoplasma pneumoniae
infections and exacerbation of childhood asthma (Table 10–2). There is limited evidence for respiratory syncytial virus and inadequate evidence for
other infectious agents to assess their role in asthma development.
Animal and insect aeroallergens. Important findings of the National
Institute of Medicine and other studies include the following (see also
Table 10–2):
Pets
• Sensitization—cat allergen Fel d 1 can sensitize children at very low
levels (nanograms per cubic meter) and trigger severe asthma attacks
in sensitized persons.
• Asthma development—the strongest risk factor for recent-onset,
physician-diagnosed asthma among young children in NHANES III
was allergy to a pet (Lanphear et al., 2001); a longitudinal cohort study
of incident asthma in southern California also showed an association
with pets in the home (McConnell et al., 2002).
• Asthma episodes—there is sufficient evidence that cat and dog allergens can cause episodes in sensitized asthmatics.
House-dust mites
• Sensitization—African American and Mexican American children in
NHANES III had relatively high house-dust mite sensitization rates.
• Asthma development—there is sufficient evidence that house-dust mite
antigens can cause asthma development.
Indoor Air
283
• Asthma episodes—there is sufficient evidence that house-dust mite
antigens can trigger episodes in sensitized asthmatics.
Cockroaches
• Sensitization—African American and Mexican American children in
NHANES III had relatively high cockroach antigen sensitization rates;
there was a dose–response relationship between cockroach allergen exposure and sensitization.
• Asthma development—there is suggestive evidence that cockroach
antigens can cause asthma development in preschool children.
• Asthma episodes—there is sufficient evidence that cockroach antigens
can trigger episodes in sensitized asthmatics.
Fungi
Measurement of fungal exposure is complicated by large numbers of fungi
and fungal products. Many fungal products are potential allergens, and
some have been characterized; the amounts and profiles produced depend on environmental factors and the type of substrate. Fungal mycotoxins comprise some 400 entities including aflatoxins, trichothecenes, fumonisins, and ochratoxin, with toxic activities including cytotoxicity,
immunotoxicity, carcinogenicity, estrogenic activity, and inflammation.
Few epidemiologic studies have directly measured indoor air fungal allergen levels, resorting instead to indirect indicators such as the presence
of visible mold or dampness.
Approximately 6%–10% of the general population and 15%–50% of
persons with allergic conditions are sensitized to fungal allergens. Dampness and other indicators of fungal allergens have been associated with
asthma in several studies. A review of epidemiologic studies of indoor
dampness and respiratory symptoms, asthma, and allergy concluded that
dampness increases the risk of cough, wheezing, and asthma by up to
twofold; the mechanism is unknown but may include house-dust mites
or molds (Bornehag et al., 2001). The IOM concluded that there was sufficient evidence of an association between molds and triggering of asthma
episodes but inadequate evidence of an association between molds and
development of asthma (Table 10–2). See also Chapter 11 (Outdoor Air)
for discussion of ambient molds and other aeroallergens in childhood respiratory disease.
Endotoxin. Endotoxin is a component of the outer membrane of gramnegative bacteria that is a potent toxin, activating human airway macrophages to release proinflammatory cytokines and other substances at levels as low as 1 ng/mL in vitro. Exposure to elevated endotoxin levels in
house dust has been linked to airway inflammation when inhaled by adult
284
CHILD HEALTH
AND THE
ENVIRONMENT
volunteers, wheezing episodes before age 1 year, and asthma severity
among adults. At present there is inadequate evidence to assess its role
in asthma development or triggering of asthma episodes (Table 10–2).
Multiple exposures. Among asthmatic children, the likelihood of sensitization to house-dust mite or cockroach antigens is associated with their
concentration in house dust. Deficits in FEV1 among asthmatic children
in NHANES II were greater among those sensitized to any indoor allergen than for those reactive to any outdoor allergen tested (Schwartz and
Weiss, 1995). A recent review concluded that existing evidence does not
prove a causal relationship between allergen exposure level and asthma
development, noting that no published studies had linked allergen exposure during infancy to the risk of asthma after age 6 years in a random
population sample (Pearce et al., 2000).
Other Diseases
An outbreak of idiopathic pulmonary hemosiderosis among 37 infants in
Cleveland included 12 deaths; investigation showed strong associations
with residence in households with major water damage, increased air
levels of Stachybotrys chartarum fungal spores, and combined exposure to
S. chartarum and ETS (Dearborn et al., 1999). Over 100 additional similar
cases have since been reported in the United States. It is thought that
trichothecene mycotoxins produced by S. chartarum inhibit protein synthesis in the rapidly growing infant lung, causing capillary fragility and
hemorrhage.
Exposures
Comparison of results from health studies of indoor aeroallergens has
been constrained by inconsistent use of the sampling sites, measurement
devices, and sampling schedules needed to measure children’s exposures
during activities in different household areas; most studies have measured
aeroallergen levels in reservoir dust, but air levels may better reflect respiratory tract exposures. But airborne particle levels from resuspended
house dust vary substantially throughout the day and across seasons,
making adequately integrated exposure measures difficult and expensive.
The correlation between cat, dust mite, and cockroach antigen concentrations in floor dust and serum allergen-specific IgE levels appears to
vary by type of dust sampler (Mansour et al., 2001). The first National
Survey of Lead and Allergens in Housing in United States (1998–1999)
measured cockroach allergen Bla g 1, the dust mite allergens Der f 1 and
Indoor Air
285
Der p 1, the cat allergen Fel d 1, the dog allergen Can f 1, the rodent
allergens Rat n 1 and mouse urinary protein, allergens of the fungus
Alternaria alternata, and endotoxin in dust samples from a bed, the bedroom floor, a sofa or chair, the living room floor, the kitchen floor, and the
basement floor (Vojta et al., 2002) (results to be published soon).
Animal and Insect Aeroallergens
Pets. About 30% of households in the United States keep cats or dogs
as pets. The main cat allergen, Fel d I, occurs in settled dust and in fine
airborne particles (PM2.5) that remain airborne for prolonged times, are
very adherent, spread throughout the house, and are ubiquitous even in
households without cats. The main dog allergens, Can f I and Can f II, are
also widely distributed in fine airborne particles and in settled dust. Most
homes with dogs or cats and substantial proportions of pet-free homes
have high levels of their specific antigens in dust and detectable levels in
airborne fine particulate matter; curtains, desks, and chairs in school classrooms act as reservoirs for cat and dog allergens apparently transported
to school on clothing.
House-dust mites. The house-dust mite (Dermatophagoides pteronyssinus) is a tiny insect that scavenges materials including human skin scales,
pollen grains, insect scales, house dust, and plant fibers. Antigens from
mite feces and dead body parts (Der p I and II) occur mainly as large particles (10 m) that accumulate in settled dust and remain airborne for
short periods if disturbed; antigen concentrations in house dust are highest in geographic regions with persistent high humidity during several
months annually. Within homes, concentrations (micrograms per gram of
dust) are highest in mattresses, carpets, quilts, and sofas. Floor and mattress dust antigen levels vary by factors including type of ventilation
(higher in the absence of forced air), season, type and age of home, age
of mattress, and number of occupants. Young children spend a great deal
of time on or near the floor, where allergens are concentrated in dust, and
can be sensitized to house-dust mite antigens at very low levels.
Cockroaches. Blattella germanica and Periplaneta americana are the two
most common cockroaches in the United States; high levels of their antigens (Bla g 1, 2, and 4) are linked to warm, humid climates, lower socioeconomic status, living in apartments, and availability of water and
food (particularly in kitchens). Cockroach antigens are associated with
large particles that accumulate in settled dust and remain airborne for
short periods if disturbed. There is a strong association between poverty
and high house dust cockroach antigen levels and skin sensitization.
286
CHILD HEALTH
AND THE
ENVIRONMENT
Fungi
There are at least 1 million species of fungi including 200 to which humans are commonly exposed. The most prevalent mold genera in homes
are Alternaria, Cladosporium, and Penicillium. Fungal spores are ubiquitous
in the outdoor environment and indoor dust, and can readily grow indoors under conditions of persistent high humidity and in homes with a
dirt floor or crawl space type of basement. Compared to white children,
African American children are twice as likely to be sensitized to A. alternata (OR 2.1, CI 1.5–2.8) (Stevenson et al., 2001). Indoor storage of organic wastes (for composting) for a week or longer is associated with a
five- to eightfold increase in levels of fungal extracellular polysaccharides
and -glucans on living room and kitchen floors. Other factors contributing to mold growth include carpeted floors, dampness, past flooding, indoor storage of firewood, and unvented dryers.
Fungal exposure has often been assessed by crude indicators such as
the presence of visible mold or dampness. Quantitative indices include
spore counts, culturable fungi, ergosterol (a fungal cell membrane sterol),
(1-3)-glucans, extracellular polysaccharides, fungal volatile organic compounds (3-methylfuran, 1-octene-3-ol, geosmin), mycotoxins, specific
DNA sequences, and fungal-specific lipids (Dillon et al., 1999). (1-3)Glucans are potent inflammatory agents that occur in cell walls of fungi,
yeasts, some bacteria, and many plants. Concentrations of (1-3)-glucans
per unit area in settled house dust are associated with carpets, the presence of a dog, more than four persons, infrequent cleaning, and the presence of visible mold. Settled dust concentrations of (1-3)-glucans are
highly correlated with those of endotoxin, house-dust mite and cat allergens, and mold spores.
Endotoxin
High endotoxin levels in settled dust are associated with the presence of
a dog or cat in the home, indoor storage of organic wastes (for composting), and farm environments where livestock and poultry are kept. Mattress dust endotoxin levels may better reflect exposure than floor dust or
air levels (Park et al., 2000).
Risk Management
Animal and Insect Aeroallergens
Pets. Removal of pets from the home is the most effective way to reduce cat or dog allergens but is often avoided; keeping pets out of bedrooms and regular cleaning reduce exposure levels somewhat. After removal of cats from homes, Fel d I levels fall but may remain elevated for
several months unless carpets and upholstery are removed and mattresses
Indoor Air
287
and pillows are encased. Intervention using a high-efficiency particulate
air cleaner (HEPA filter) over a 3-month period significantly reduced airborne but not settled dust cat allergen levels, and there was no significant improvement in cat-induced asthma or rhinitis symptom scores, sleep
disturbance, medication use, or PEFs (Wood et al., 1998). Removal of dogs
from homes appears to reduce dog allergen levels, but there is inadequate
evidence to assess the impact of dog or cat removal or related interventions on symptoms of sensitized asthmatics (National Academy of Sciences, 2000).
House-dust mites. A review of 23 intervention trials that used chemical and physical methods to control house-dust mite antigen levels in
homes of asthmatics concluded that there was negligible impact on symptom scores, medication use, and lung function (Gotzsche et al., 1998).
Nevertheless, some randomized intervention trials have succeeded in reducing bedroom mite allergen levels, bronchial hyperreactivity, and
bronchodilator use and in improving symptoms and FEV1 levels (Carswell et al., 1996; Shapiro et al., 1999). Possible reasons for lack of efficacy
include failure to reduce antigen levels adequately and sensitization to
other allergens persistent in the home environment (atopic persons are
usually sensitized to multiple indoor allergens). Aeroallergen levels can
also be reduced to some degree by eliminating carpets and upholstered
furniture (especially from bedrooms), vacuuming with a central vacuum
or one with a HEPA filter, damp mopping floors, washing bedding weekly
in hot water (at least 55°C), increasing ventilation, and keeping relative
humidity below 50% and temperatures below 25°C.
Cockroaches. Insecticides are often used for cockroach control, especially in inner-city housing and apartments. This measure, however, introduces children to the risk of pesticide toxicity. Other measures include
general cleanliness (especially removal of food debris), control of dampness, and caulking of crevices. A randomized intervention trial that recruited asthmatic children sensitized to cockroach antigen showed only
a minor, transitory benefit of cockroach extermination and house cleaning education on house-dust antigen levels (Gergen et al., 1999). There is
sufficient evidence that intensive cockroach exposure interventions can
cause transiently reduced allergen levels, but there is inadequate evidence
to assess its efficacy in improving symptoms or lung function in sensitized asthmatics (National Academy of Sciences, 2000).
Fungi and Endotoxins
Mold control depends mainly on maintenance of low humidity levels and
avoidance of water damage. There are no recognized guidelines or stan-
288
CHILD HEALTH
AND THE
ENVIRONMENT
dards for concentrations of fungi or fungal products in indoor air or house
dust (Dillon et al., 1999). A review of the recent epidemiologic literature
concluded that the use of cross-sectional study designs and inadequate
validation of exposure measures precluded a quantitative risk assessment
and an air standard for fungi (Verhoeff and Burge, 1997). Finland, apparently the only country that has done so, declared indoor visible mold
growth to be a health hazard and legislated public health measures to
control indoor dampness and molds (Husman, 1999). Some countries have
developed guidelines for relative humidity and ventilation rates. There is
suggestive evidence that interventions can reduce fungal allergen levels
indoors but inadequate evidence to determine if such interventions improve asthma symptoms or lung function in sensitized asthmatics; there
is inadequate evidence to determine if interventions reduce endotoxin levels in homes (National Academy of Sciences, 2000).
Mixed Exposures
The European Environment and Health Committee recommended that
European Union member states take actions on asthma including (1) development of guidelines for house dust, humidity, molds, cockroaches,
and pets, (2) promotion of improved ventilation, (3) communication of
the dangers of ETS and smoking during pregnancy, and (4) strict enforcement of smoking prohibitions in areas frequented by children (European Centre for Environment and Health, 1999). There is sufficient evidence that intervention (ventilation, mattress and pillow encasement,
regular hot washing of bedding, etc.) can reduce multiple house-dust allergen levels and cause an improvement in asthma symptoms but inadequate evidence to determine if mitigation can reduce asthma development (National Academy of Sciences, 2000).
Volatile Organic Chemicals and Gases
This section briefly addresses indoor volatile organic chemicals (VOCs)
and gases (carbon monoxide and nitrogen dioxide) as potential threats to
child health. See also Chapter 11 (Outdoor Air) for discussion of outdoor
sources of VOCs and other toxic gases and their potential health effects.
Health Effects
Volatile Organic Chemicals
Among the over 300 chemicals that have been detected in indoor air, about
30 have been linked to asthma in occupational or animal studies; among
Indoor Air
289
these, children may be exposed to several in their homes, such as, acrolein
(ETS), formaldehyde (ETS, synthetic materials), and chlorine (chlorinated
water, household bleach). There is limited evidence that exposure to
mixed indoor air contaminants may be a factor in childhood asthma and
other allergic diseases in children. Limited evidence links formaldehyde
to wheezing and other respiratory symptoms, but there is inadequate evidence to determine if it can cause asthma development (National Academy of Sciences, 2000). The ability of relatively low-level formaldehyde
to cause eye irritation among youth and young adults was shown in a
controlled exposure study (Yang et al., 2001). Experimental animal evidence indicates that inhaled formaldehyde exposure increases sensitization to aeroallergens.
Home use of organic solvents for model building and artwork, and
maternal prenatal exposure to extensive indoor painting, were associated
with a two- to fourfold increased risk of childhood leukemia (Freedman
et al., 2001). Benzene is a frequent contaminant of indoor air and a known
cause of aplastic anemia and acute myeloid leukemia in persons with prolonged high occupational exposures. Sequential metabolism of benzene
in liver and bone marrow generates semiquinone radicals that react covalently with tubulin, chromosomal proteins, and topoisomerase II and
induce oxidative DNA damage. When attacked by free radicals, topoisomerase II can break DNA strands and cause the 11q23 chromosomal band
abnormalities commonly seen in childhood leukemia. At high concentrations in air, formaldehyde is a known animal (nasal sinus cancer) and a
probable human carcinogen, but the potential role of it and other VOCs
in childhood cancer is unknown.
Carbon Monoxide and Nitrogen Dioxide
Carbon monoxide (CO) is a colorless, odorless gas produced by incomplete combustion of carbonaceous fuels and materials. The most serious
health effect in children is acute, high-dose CO poisoning that can cause
symptoms ranging from headache and dizziness to nausea, vomiting, loss
of consciousness, and death. Inhaled CO is rapidly absorbed into the
bloodstream, where it binds to hemoglobin with an affinity about 240
times that of oxygen. Carbon monoxide also binds to cytochrome aa3,
blocking mitochondrial energy production, and to myoglobin, interfering
with oxygen storage in skeletal and heart muscle. Children are at increased risk of CO poisoning because of their relatively high metabolic
rates; the fetus is also susceptible because fetal hemoglobin has a higher
affinity for CO than adult hemoglobin. Because of their high oxygen
needs, the central nervous and cardiovascular systems are most adversely
affected by CO poisoning. The American Association of Poison Control
290
CHILD HEALTH
AND THE
ENVIRONMENT
Centers Toxic Exposure Surveillance System (TESS) received 1787 reports
of moderate, severe, or fatal CO poisonings among children aged 0–5
years during 2000 (Litovitz et al., 2001). Potential health effects of lowlevel ambient CO levels are discussed in Chapter 11 (Outdoor Air).
There is limited but inconsistent evidence of associations between indoor nitrogen dioxide (NO2) exposure and respiratory disease in children.
The best evidence comes from studies that had reasonable sample sizes
and direct measures of NO2 exposure, such as, the Harvard Six Cities
Study (Neas et al., 1991) and a large case-control study of incident asthma
in Montreal (Infante-Rivard, 1993). Acute high-level NO2 exposure in
hockey arenas employing propane-powered ice resurfacers has caused
high rates of acute respiratory symptoms (in 57% of adolescent players
in one outbreak).
Exposures
Volatile Organic Chemicals
Exposure to VOCs can be estimated from their levels in indoor, ambient,
and personal air and water, and VOC uptake can be assessed from urine
and alveolar breath samples. Personal VOC exposures are dominated by
indoor sources, with ambient outdoor exposures contributing only about
2%–25% of personal exposures for most toxic and carcinogenic VOCs
(Wallace, 1993). Detailed monitoring of adult volunteers showed that
most of 25 commonly performed activities increase personal exposure to
one or more target VOCs, often by a factor of 10. Activities causing high
VOC exposures included use of deodorizers (-dichlorobenzene), washing clothes and dishes (chloroform), visiting a dry cleaner (1,1,1trichloroethane, tetrachloroethylene), smoking (benzene, styrene), and
painting and paint stripping (n-decane, n-undecane). Specific findings
from studies of VOC exposures include the following:
• All VOCs—based on measured personal exposures and estimated
upper-bound lifetime cancer risks, the most important indoor VOCs
were benzene, vinylidene chloride, chloroform, and -dichlorobenzene;
airborne sources accounted for 80%–100% of the exposure to each of
these VOCs (Wallace, 1991).
• Benzene
° Children—personal air benzene levels of urban children are higher
than those of suburban children; levels of urinary trans-muconic acid
(a benzene metabolite) among inner-city children are associated with
time spent playing near the street (note: motor vehicles emit benzene).
Indoor Air
291
° Adults—the main sources are active and passive smoking, auto ex-
haust, and driving or riding in automobiles; built-in garages are a
major source for nonsmokers
Indoor air formaldehyde levels may be elevated in homes, especially mobile homes, containing large amounts of particleboard and other building
materials that release formaldehyde and have low air exchange rates. An
indoor air quality survey of homes in southern Louisiana showed indoor
air formaldehyde levels of up to 6.6 mg/m3, with 60% of homes exceeding the ASHRAE guideline (123 g/m3) (Lemus et al., 1998). See Chapter 12 (Water) for discussion of exposure to chloroform and other VOCs
while showering or swimming in indoor chlorinated pools.
Carbon Monoxide and Nitrogen Dioxide
Carbon monoxide binds tightly to hemoglobin, forming carboxyhemoglobin with a half-life of 2–6 hours. Carboxyhemoglobin blood levels reflect
the concentration and duration of CO exposure, exercise, and other factors and can be estimated by direct blood analysis or by measuring CO in
exhaled breath. Ambient NO2 monitoring does not adequately reflect children’s personal exposures because they spend over 90% of their time indoors, where NO2 levels reflect sources such as gas stoves and smokers.
Homes with inadequately ventilated gas stoves generally have markedly
higher NO2 levels than those with electric stoves, particularly in winter.
Risk Management
Volatile Organic Chemicals
Indoor sources of VOCs include built-in garages, smoking, building materials and furnishings, chlorinated water, dry-cleaned clothes, cleaning
solutions, solvents, paint, and other consumer products (Table 10–6). Potential control measures for indoor air contaminants include ventilation,
source removal or substitution, source modification, air cleaning, and behavioral change (Moeller, 1997). Parents and other child-care providers
can reduce children’s exposures to certain VOCs by, for example, not using moth balls or bathroom deodorizers containing -dichlorobenzene,
avoiding dry-cleaning (or at least storing dry-cleaned materials outdoors
for 24 hours), and not storing volatile solvents, gasoline, or gasolinecontaining machines in attached garages.
Major indoor sources of formaldehyde include products containing
formaldehyde-based resins such as finishes, plywood, paneling, fiberboard,
particleboard, permanent press fabric, draperies, and urea formaldehyde
foam. Given the ubiquitous and largely unavoidable exposure to such prod-
TABLE 10–6. VOCs: Sources and Levels in Indoor Air, Exhaled Breath, and Blood
VOC
Main Indoor Sources
Indoor
air levels
(g/m3)a
Exhaled
breath levels
(g/m3)b
Blood
levels
(ng/L) c
Formaldehyde
Building materials, furniture, cabinets, acid-cured floor finishes
50
NA
NA
Chloroform
Chlorinated water (drinking, showering/bathing, washing dishes or clothes, indoor
swimming pools)
1
ND–2.3
NA
Carbon tetrachloride
Past uses—aerosol cans, cleaning fluids, degreasing agents, fire extinguishers, spot removers
⬍5
ND–0.7
NA
Benzene
Gasoline or motor vehicle in built-in garage, smoking
5
0.9–12
460
Ethylbenzene
Gasoline or motor vehicle in built-in garage, smoking
1,1,1-Trichloroethane
Past uses—glues, paints, aerosol sprays
Trichloroethylene
Adhesives, paint removers, and spot removers
Tetrachloroethane
Past uses—paints
Tetrachloroethylene
Dry-cleaned fabrics, ingredient in some consumer products
5
1.8–6.8
NA
-Dichlorobenzene
Mothballs, toilet deodorizers
1
ND–1.3
2050
Styrene
Smoking
2
ND–0.8
190
Toluene
Gasoline or motor vehicle in built-in garage, smoking, paints, paint thinners, fingernail
polish, lacquers, adhesives
20
NA
1460
Xylenes (o,m,p)
Gasoline or motor vehicle in built-in garage, smoking, cleaning agent, thinner for paint,
paints, varnishes
15
0.6–8.6
1040
a U.S.
Environmental Protection Agency (1998).
b Exhaled
c 95th
breath median concentrations (Wallace et al., 1996).
percentiles (estimated from graph) (Needham et al., 1995).
ND ⫽ not detectable.
5
0.2–2.9
225
NA
0.1–6.6
340
5
ND–0.9
NA
NA
NA
580
Indoor Air
293
ucts at present, improved ventilation is one of the few potential interventions in the short term; even this measure is complicated by the associated
energy cost. Longer term, there is the potential for introduction of safer
products.
The European Environment and Health Committee recommended
that European Union member states take actions on VOCs and other indoor air toxicants by (1) establishment of pollution-free schools by limiting access of vehicles, especially diesel-powered vehicles, and by restricting the siting of pollution-emitting sources around schools, (2) elimination
of wall-to-wall carpets and irritant chemical cleaning products in schools,
(3) establishment of guidelines for the quality of the home environment to
minimize risk factors including gas stoves, and (4) promotion of improved
home ventilation (European Centre for Environment and Health, 1999).
Carbon Monoxide and Nitrogen Dioxide
Potential sources of CO in indoor air include inadequately ventilated nonelectric space heaters and stoves, leaking furnaces or chimneys, backdrafting from furnaces, gas water heaters, wood stoves, and fireplaces,
running cars in attached closed garages, indoor use of barbeques, and
smoking. The main indoor sources of NO2 are inadequately ventilated gas
appliances for cooking and heating. Preventive measures for both CO and
NO2 include proper installation, maintenance, and use of fuel-burning
appliances, adequate ventilation, avoidance of unventilated space heaters
(or at least avoiding prolonged use), and use of CO detectors.
Conclusion
Proven Child Health Outcomes
• Environmental tobacco smoke
° Prenatal maternal smoking increases the risk of SIDS
° Childhood ETS exposure increases the risk of bronchiolitis, pneumonia, and middle ear infections, the number and severity of asthma
episodes in preschool-age children, bronchial hyperreactivity, lung
function deficits, and reduced lung function growth rates during
childhood.
• Biologic agents
° House-dust mite antigen exposure can cause new-onset (incident)
asthma.
° Cat, cockroach, and house-dust mite antigen exposure can precipitate episodes in sensitized asthmatics.
294
CHILD HEALTH
AND THE
ENVIRONMENT
• Volatile chemicals and gases
° Carbon monoxide from combustion sources can cumulate indoors to
levels capable of causing acute toxicity and death.
° Acute exposure to formaldehyde or high-level NO2, respectively, can
cause acute eye irritation or respiratory symptoms.
Unresolved Issues and Knowledge Gaps
• Environmental tobacco smoke
° There is limited evidence that prenatal maternal ETS exposure can
cause spontaneous abortions, birth weight deficits, and preterm birth.
° There is limited evidence that postnatal ETS exposure can cause SIDS,
independent of prenatal maternal smoking.
° There is limited evidence that prenatal and postnatal maternal smoking can cause new-onset, persistent asthma in preschool-age children,
independent of other risk factors.
° There is limited evidence that ETS causes asthma exacerbations in
school-age children but inadequate evidence to determine if ETS can
cause new-onset asthma in this age group.
° There is limited evidence that paternal smoking is associated
with childhood brain tumors, acute lymphatic leukemia, and nonHodgkin’s lymphoma.
° There is inadequate evidence that childhood ETS exposure is linked
to adult breast and nasopharyngeal cancers.
• Biologic agents
° Dog antigens, molds, and rhinovirus are associated with episodes in
known asthmatics but causality is uncertain.
° There is limited evidence that cockroach antigen can cause newonset asthma in preschool-age children.
° There is limited evidence that dog, rodent, and fungal antigens can
exacerbate asthma among preschool-age children.
° There is limited evidence that the fungus Stachybotrys chartarum can
cause idiopathic pulmonary hemosiderosis in infants, especially those
also exposed to ETS.
• Volatile chemicals and gases
° There is limited evidence that personal air or indoor NO2 levels are
associated with lung function deficits, respiratory symptoms, and
early-onset asthma.
° Children have low-level exposures to several VOCs in indoor air that
are known human or animal carcinogens, but their potential role in
childhood or adult cancer remains undefined (occupational exposure
to benzene, however, can cause adult leukemia).
Indoor Air
295
Risk Management
• Prevention
° It appears that no country has comprehensive programs to address
indoor air health hazards and to evaluate progress in reducing children’s exposures.
° Although paternal smoking is linked to childhood cancer, no tobacco
smoke carcinogen has been tested for intergenerational carcinogenicity in animals.
• Monitoring
° Biomonitoring—periodic population-based measurement of indicators of internal dose of ETS and selected VOCs is needed.
° Indoor environment surveys—periodic population-based measurement of the prevalence of key indoor air hazards (e.g., aeroallergens,
dampness/molds, VOCs, CO, NO2) is needed.
See Chapter 11 (Outdoor Air) for further discussion of CO, NO2, and certain VOCs and other chapters for other indoor pollutants (e.g., radon, pesticides, metals, volatile drinking water disinfection by-products).
References
Aligne CA, Stoddard JJ. (1997). Tobacco and children. An economic evaluation of
the medical effects of parental smoking . Arch Pediatr Adolesc Med 151:648–53.
Anderson HR, Cook DG. (1997). Passive smoking and sudden infant death syndrome: review of the epidemiological evidence. Thorax 52:1003–9.
Anderson LM, Diwan BA, Fear NT, Roman E. (2000). Critical windows of exposure for children’s health: cancer in human epidemiological studies and neoplasms in experimental animal models. Environ Health Perspect 108(Suppl
3):573–94.
Biener L, Cullen D, Di ZX, Hammond SK. (1997). Household smoking restrictions
and adolescents’ exposure to environmental tobacco smoke. Prev Med 26:
358–63.
Boffetta P, Tredaniel J, Greco A. (2000). Risk of childhood cancer and adult lung
cancer after childhood exposure to passive smoke: a meta-analysis. Environ
Health Perspect 108:73–82.
Bornehag CG, Blomquist G, Gyntelberg F, Jarvholm B, Malmberg P, Nordvall L,
Nielsen A, Pershagen G, Sundell J. (2001). Dampness in buildings and health.
Nordic interdisciplinary review of the scientific evidence on associations between exposure to “dampness” in buildings and health effects (NORDDAMP).
Indoor Air 11:72–86.
California Environmental Protection Agency. (1997). Health effects of exposure to
environmental tobacco smoke. Sacramento: Office of Environmental Health
Hazard Assessment.
296
CHILD HEALTH
AND THE
ENVIRONMENT
Carswell F, Birmingham K, Oliver J, Crewes A, Weeks J. (1996). The respiratory
effects of reduction of mite allergen in the bedrooms of asthmatic children—
a double-blind controlled trial. Clin Exp Allergy 26:386–96.
Centers for Disease Control and Prevention. (2001). National report on human exposure to environmental chemicals. Cotinine. Atlanta: Centers for Disease Control and Prevention.
Cook DG, Strachan DP. (1998). Parental smoking, bronchial reactivity and peak
flow variability in children. Thorax 53:295–301.
Cook DG, Strachan DP, Carey IM. (1998). Parental smoking and spirometric indices in children. Thorax 53:884–93.
Dearborn DG, Yike I, Sorenson WG, Miller MJ, Etzel RA. (1999). Overview of investigations into pulmonary hemorrhage among infants in Cleveland, Ohio.
Environ Health Perspect 107:495–9.
Dejmek J, Solansky I, Podrazilova K, Sram RJ. (2002). The exposure of nonsmoking and smoking mothers to environmental tobacco smoke during different
gestational phases and fetal growth. Environ Health Perspect 110:601–6.
Dillon HK, Miller JD, Sorenson WG, Douwes J, Jacobs RR. (1999). Review of methods applicable to the assessment of mold exposure to children. Environ Health
Perspect 107(Suppl 3):473–80.
Eskenazi B, Bergmann JJ. (1995). Passive and active maternal smoking during
pregnancy, as measured by serum cotinine, and postnatal smoke exposure. I.
Effects on physical growth at age 5 years. Am J Epidemiol 142:S10–8.
European Centre for Environment and Health WHO. (1999). Children’s health and
the environment (EUR/ICP/EHCO 02 02 05/16). Rome: European Centre for
Environment and Health.
Finette BA, O’Neill JP, Vacek PM, Albertini RJ. (1998). Gene mutations with characteristic deletions in cord blood T lymphocytes associated with passive maternal exposure to tobacco smoke. Nat Med 4:1144–51.
Freedman DM, Stewart P, Kleinerman RA, Wacholder S, Hatch EE, Tarone RE, Robison LL, Linet MS. (2001). Household solvent exposures and childhood acute
lymphoblastic leukemia. Am J Public Health 91:564–7.
Gergen PJ, Mortimer KM, Eggleston PA, Rosenstreich D, Mitchell H, Ownby D,
Kattan M, Baker D, Wright EC, Slavin R, and others. (1999). Results of the national cooperative inner-city asthma study (NCICAS) environmental intervention to reduce cockroach allergen exposure in inner-city homes. J Allergy
Clin Immunol 103:501–6.
Gotzsche PC, Hammarquist C, Burr M. (1998). House dust mite control measures
in the management of asthma: meta-analysis. BMJ 317:1105–10.
Husman TM. (1999). The Health Protection Act, national guidelines for indoor air
quality and development of the national indoor air programs in Finland. Environ Health Perspect 107(Suppl 3):515–7.
Infante-Rivard C. (1993). Childhood asthma and indoor environmental risk factors. Am J Epidemiol 137:834–44.
Infante-Rivard C, Gautrin D, Malo JL, Suissa S. (1999). Maternal smoking and
childhood asthma. Am J Epidemiol 150:528–31.
Lanphear BP, Aligne CA, Auinger P, Weitzman M, Byrd RS. (2001). Residential exposures associated with asthma in U.S. children. Pediatrics 107:505–11.
Lemus R, Abdelghani AA, Akers TG, Horner WE. (1998). Potential health risks
from exposure to indoor formaldehyde. Rev Environ Health 13:91–8.
Indoor Air
297
Li JS, Peat JK, Xuan W, Berry G. (1999). Meta-analysis on the association between
environmental tobacco smoke (ETS) exposure and the prevalence of lower respiratory tract infection in early childhood. Pediatr Pulmonol 27:5–13.
Li YF, Gilliland FD, Berhane K, McConnell R, Gauderman WJ, Rappaport EB, Peters JM. (2000). Effects of in utero and environmental tobacco smoke exposure
on lung function in boys and girls with and without asthma. Am J Respir Crit
Care Med 162:2097–2104.
Litovitz TL, Klein-Schwartz W, White S, Cobaugh DJ, Youniss J, Omslaer JC, Drab
A, Benson BE. (2001). 2000 Annual report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med
19:337–95.
Mannino DM, Moorman JE, Kingsley B, Rose D, Repace J. (2001). Health effects
related to environmental tobacco smoke exposure in children in the United
States: data from the Third National Health and Nutrition Examination Survey. Arch Pediatr Adolesc Med 155:36–41.
Mansour M, Lanphear BP, Hornung R, Khoury J, Bernstein DI, Menrath W, Decolongon J. (2001). A side-by-side comparison of sampling methods for settled,
indoor allergens. Environ Res 87:37–46.
McConnell R, Berhane K, Gilliland F, Islam T, Gauderman WJ, London SJ, Avol
E, Rappaport EB, Margolis HG, Peters JM. (2002). Indoor risk factors for asthma
in a prospective study of adolescents. Epidemiology 13:288–95.
McGready R, Hamilton KA, Simpson JA, Cho T, Luxemburger C, Edwards R,
Looareesuwan S, White NJ, Nosten F, Lindsay SW. (2001). Safety of the insect
repellent N,N-diethyl-M-toluamide (DEET) in pregnancy. Am J Trop Med Hyg
65:285–9.
McKeever TM, Lewis SA, Smith C, Collins J, Heatlie H, Frischer M, Hubbard R.
(2002). Early exposure to infections and antibiotics and the incidence of allergic disease: a birth cohort study with the West Midlands General Practice Research Database. J Allergy Clin Immunol 109:43–50.
Millar WJ, Hill GB. (1998). Childhood asthma. Health Rep 10:9–21.
Misra DP, Nguyen RH. (1999). Environmental tobacco smoke and low birth
weight: a hazard in the workplace? Environ Health Perspect 107:879–904.
Moeller DW. (1997). Environmental health. Cambridge, MA: Harvard University
Press.
National Academy of Sciences. (2000). Clearing the air. Asthma and indoor air exposures. Washington, DC: National Academy Press.
Neas LM, Dockery DW, Ware JH, Spengler JD, Speizer FE, Ferris BG Jr. (1991).
Association of indoor nitrogen dioxide with respiratory symptoms and pulmonary function in children. Am J Epidemiol 134:204–19.
Needham LL, Hill RH, Ashley DL, Pirkle JL, Sampson EJ. (1995). The priority
toxicant reference range study: interim report. Environ Health Perspect 103
(Suppl 3):89–94.
Ober C, Moffatt MF. (2000). Contributing factors to the pathobiology. The genetics of asthma. Clin Chest Med 21:245–61.
Park JH, Spiegelman DL, Burge HA, Gold DR, Chew GL, Milton DK. (2000). Longitudinal study of dust and airborne endotoxin in the home. Environ Health
Perspect 108:1023–8.
Pearce N, Beasley R, Burgess C, Crane J. (1998). Asthma epidemiology. principles
and methods. New York: Oxford University Press.
298
CHILD HEALTH
AND THE
ENVIRONMENT
Pearce N, Douwes J, Beasley R. (2000). Is allergen exposure the major primary
cause of asthma? Thorax 55:424–31.
Pekkanen J, Pearce N. (1999). Defining asthma in epidemiological studies. Eur
Respir J 14:951–7.
Pinkerton KE, Joad JP. (2000). The mammalian respiratory system and critical windows of exposure for children’s health. Environ Health Perspect 108(Suppl 3):
457–62.
Sasco AJ, Vainio H. (1999). From in utero and childhood exposure to parental
smoking to childhood cancer: a possible link and the need for action. Hum
Exp Toxicol 18:192–201.
Schwartz J, Weiss ST. (1995). Relationship of skin test reactivity to decrements in
pulmonary function in children with asthma or frequent wheezing. Am J
Respir Crit Care Med 152:2176–80.
Shapiro GG, Wighton TG, Chinn T, Zuckrman J, Eliassen AH, Picciano JF, PlattsMills TA. (1999). House dust mite avoidance for children with asthma in homes
of low-income families. J Allergy Clin Immunol 103:1069–74.
Stevenson LA, Gergen PJ, Hoover DR, Rosenstreich D, Mannino DM, Matte TD.
(2001). Sociodemographic correlates of indoor allergen sensitivity among
United States children. J Allergy Clin Immunol 108:747–52.
Strachan DP, Cook DG. (1997). Parental smoking and lower respiratory illness in
infancy and early childhood. Thorax 52:905–14.
Strachan DP, Cook DG. (1998a). Parental smoking, middle ear disease and adenotonsillectomy in children. Thorax 53:50–6.
Strachan DP, Cook DG. (1998b). Parental smoking and childhood asthma: longitudinal and case-control studies. Thorax 53:204–12.
U.S. Environmental Protection Agency. (1992). Respiratory health effects of passive smoking: lung cancer and other disorders (EPA/600/9-90/006F). Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (1998). Inside IAQ (EPA/600/N-98/002).
Research Triangle Park, NC: U.S. Environmental Protection Agency.
Verhoeff AP, Burge HA. (1997). Health risk assessment of fungi in home environments. Ann Allergy Asthma Immunol 78:544–54.
Vojta PJ, Friedman W, Marker DA, Clickner R, Rogers JW, Viet SM, Muilenberg
ML, Thorne PS, Arbes SJ Jr, Zeldin DC. (2002). First national survey of lead
and allergens in housing: survey design and methods for the allergen and endotoxin components. Environ Health Perspect 110:527–32.
von Mutius E, Schwartz J, Neas LM, Dockery D, Weiss ST. (2001). Relation of body
mass index to asthma and atopy in children: the National Health and Nutrition Examination Study III. Thorax 56:835–8.
Wallace L. (1993). A decade of studies of human exposure: what have we learned?
Risk Anal 13:135–9.
Wallace L, Buckley T, Pellizzari E, Gordon S. (1996). Breath measurements as
volatile organic compound biomarkers. Environ Health Perspect 104(Suppl 5):
861–9.
Wallace LA. (1991). Comparison of risks from outdoor and indoor exposure to
toxic chemicals. Environ Health Perspect 95:7–13.
Wang X, Zuckerman B, Pearson C, Kaufman G, Chen C, Wang G, Niu T, Wise PH,
Bauchner H, Xu X. (2002). Maternal cigarette smoking, metabolic gene polymorphism, and infant birth weight. JAMA 287:195–202.
Indoor Air
299
Wigle DT, Collishaw NE, Kirkbride J. (1987). Exposure of involuntary smokers to
toxic components of tobacco smoke. Can J Public Health 78:151–4.
Windham GC, Eaton A, Hopkins B. (1999). Evidence for an association between
environmental tobacco smoke exposure and birthweight: a meta-analysis and
new data. Paediatr Perinat Epidemiol 13:35–57.
Windham GC, Swan SH, Fenster L. (1992). Parental cigarette smoking and the risk
of spontaneous abortion. Am J Epidemiol 135:1394–1403.
Witschi H, Joad JP, Pinkerton KE. (1997). The toxicology of environmental tobacco
smoke. Annu Rev Pharmacol Toxicol 37:29–52.
Wood RA, Johnson EF, Van Natta ML, Chen PH, Eggleston PA. (1998). A placebocontrolled trial of a HEPA air cleaner in the treatment of cat allergy. Am J Respir
Crit Care Med 158:115–20.
Yang X, Zhang YP, Chen D, Chen WG, Wang R. (2001). Eye irritation caused by
formaldehyde as an indoor air pollution—a controlled human exposure experiment. Biomed Environ Sci 14:229–36.
11
Outdoor Air
About 1.5 billion people in developed countries live in urban areas and
are chronically exposed to outdoor air pollution, mainly from vehicular
and industrial emissions. The WHO forecasts that urban populations will
grow fastest in economically disadvantaged countries, comprise 62% of
the world’s population by the year 2025, and live in increasingly dense
centers of anthropogenic emissions. Over 150 million tons of air pollutants are emitted annually in United States alone. Closely related to population growth and urbanization are the uncertain potential impacts of
global climatic change on local, regional, and global air quality.
A severe air pollution episode in a river valley in Belgium during
1930 caused several thousand acute respiratory illnesses and about 60
deaths. A similar air pollution episode was associated with several thousand respiratory illness cases and 19 deaths in Donora, Pennsylvania, in
1948 when industrial emissions were trapped in a river valley by a temperature inversion. The severe London smog episode of December 1952
caused about 4000 excess deaths over a 5-day period during which visibility was reduced to as little as 1–5 m. Although most of the excess deaths
were from cardiorespiratory diseases among the elderly, death rates doubled among young children. The smog was caused by a persistent temperature inversion combined with extensive use of coal by industry and
300
Outdoor Air
301
for home heating; estimated particulate matter (PM) and sulfur dioxide
(SO2) levels were greatly elevated. A similar episode in London during
1956 lasted for only 18 hours but was associated with about 1000 excess
deaths. These incidents showed that severe air pollution can cause substantial excess mortality and led to the U.K. 1956 Clean Air Act.
The EPA has defined two major air pollutant categories: (1) criteria
air pollutants—CO, NO2, SO2, PM, lead, and ozone are clearly recognized
human health threats and are regulated under national ambient air quality standards (NAAQS) and (2) hazardous air pollutants (HAPs)—this
group comprises 189 substances reasonably expected to cause serious
health effects and under review for establishment of emission standards.
Motor vehicles produce most of the CO, NO2, and VOCs in ambient urban air. Fossil fuels combusted by industry (especially coal) and motor
vehicles (especially diesel fuel) are the major sources of ambient air PM2.5
and SO2. Photochemical smog was first recognized in California during
the 1940s and 1950s when greatly increased motor vehicle emissions and
secondary pollutants from photochemical reactions were trapped in the
Los Angeles basin; about 75% of southern urban Californians reported
eye irritation during peak smog periods in 1960.
Although concentrations of individual toxic chemicals in ambient air
are highly correlated, substantial evidence links each of the criteria air
pollutants to respiratory and other health effects in children. Children
may be more exposed than adults to outdoor air pollutants because they
breathe more air per unit body weight at rest, spend more time outdoors,
and have higher activity levels. Asthmatic children are generally more
susceptible than healthy children to respiratory impacts of outdoor air
pollution.
The objective of this chapter is to describe the major known and probable effects of outdoor air pollution, with a focus on children living in
modern urban environments. The discussion includes the relationships
between major categories of outdoor air toxics and potential developmental and respiratory effects. The chapter closes with a discussion of
major sources of outdoor air toxics and interventions aimed at reducing
their health effects. Chapters 4 and 5 (Metals), Chapter 7 (Pesticides),
Chapter 9 (Radiation), and Chapter 10 (Indoor Air) address other indoor
and outdoor air contaminants.
Health Effects
Many studies have shown consistent associations between daily ambient
PM levels and daily cardiovascular, respiratory, and total deaths involv-
302
CHILD HEALTH
AND THE
ENVIRONMENT
ing mainly older persons; associations have generally been stronger after
applying lag periods of up to a few days. The estimated average increases
in daily total, cardiovascular, and respiratory disease mortality, respectively, are about 1%, 1.4%, and 3.4% per 10 g/m3 increment in PM10 levels (Dockery and Pope, 1994). Elevations of PM10 have also been accompanied by increased hospitalizations and emergency room and physician
visits of children for asthma and other respiratory diseases.
Mechanisms of Toxicity
Particulate Matter
Particulate matter comprises solid, liquid, or mixed particles suspended
in air with variable size, composition, and origins (Table 11–1). Because
the health effects of PM are strongly related to particle size, PM size categories have been defined for regulatory purposes. Primary PM is emitted directly into the atmosphere and occurs in ultrafine, fine, and coarse
size ranges. Older studies measured total suspended particles (TSP), a
fraction including particles up to 40 m in diameter. Whereas fine
(PM1.0–2.5) particles can remain in the atmosphere for days to weeks and
may travel through the atmosphere hundreds to thousands of kilometers
from sources, coarse particles (PM2.5–10) usually settle to earth within minutes to hours and within tens of kilometers from sources. Coarse particles often originate from the earth’s crust and usually contain oxides of
iron, calcium, silicon, and aluminum.
Secondary PM is formed in the atmosphere through chemical and
physical transformations of ultrafine particles (PM1.0) and gases that coalesce to form PM2.5; the latter contains sulfates, nitrates, ammonium ion,
elemental carbon, PAHs, other toxic organic carbon compounds, and metals. Particle size is influenced by humidity; uptake of water by airborne
particles increases their size and contributes to visible haze during summer smog episodes.
Toxicity of inhaled particles depends on their physical and chemical
properties, particularly the size, solubility, and content of toxic substances
including organic chemicals and metals. Health concerns about PM have
increasingly focused on fine and ultrafine particles because they are deposited in peripheral airways and alveoli and contain higher concentrations of toxic chemicals than larger particles. Fine particles are highly respirable and have very large surface areas, making excellent carriers for
adsorbed inorganic and organic toxics, particularly PAHs, nitro-PAHs,
and oxidized PAH derivatives. Coarse particles are deposited mainly in
the upper airways or larger bronchi and larger particles (ⱖ10 m) in the
nasopharynx.
TABLE 11–1. Criteria Air Pollutants
Contaminant
Sources
Characteristics
PM
1. Coarse particles (PM2.5–10)
2.5 m ⱕ diameter ⬍ 10 m
Mechanical processes, such as crushing or grinding activities, construction, farming,
and mining activities, paved and unpaved roadways; fuel combustion; biologic
sources—grass, tree, and other plant pollens, mold spores
2. Fine particles (PM1.0–2.5)
1.0 m ⱕ diameter ⬍ 2.5 m
Direct emissions from motor vehicles, industries, wood burning, construction, tilled
fields, unpaved roads, stone crushing; secondary formation in atmosphere from gases
emitted by motor vehicles and industry (SO2, NOx, ammonia, and VOCs)
3. Ultrafine particles
(PM1.0)
NOx
Diameter ⬍ 1.0 m
Condensation of hot vapors formed during high-temperature combustion, such as
in motor vehicle catalytic converters
Fuel combustion at high temperature, such as in internal combustion engines,
electric utilities
Odorless, highly reactive gases;
NO2 contributes to reddishbrown color of urban smog
SO2
Colorless gas with a
choking odor
Combustion of fossil fuels such as diesel fuel, high-sulfur gasoline, coal
CO
Colorless, odorless gas
Incomplete combustion, such as in internal combustion engines, industries,
wood burning
Ozone
Pale blue gas with strong odor;
potent oxidizing agent and
respiratory tract irritant
Photochemical reactions of NOx and VOCs mainly from motor vehicle emissions
in the presence of sunlight under summer conditions
Leada
Metal
Leaded gasoline, lead smelters
a See
Chapter 4 for further discussion of lead.
304
CHILD HEALTH
AND THE
ENVIRONMENT
Deposited PM is cleared from the larger ciliated airways by the mucociliary ladder to the pharynx and swallowed or expectorated; impaired
clearance at high exposure levels appears to be greater among children
than adults. Fine particles deposited in peripheral airways and alveoli are
phagocytosed by lung macrophages that migrate via the mucociliary ladder or enter the lymphatic system and migrate to regional lymph nodes.
In human volunteers, inhaled ultrafine carbon particles labeled with technetium-99 were rapidly absorbed into the bloodstream and distributed to
the liver and other tissues, showing the potential for inhaled particles to
cause toxic effects in extrapulmonary tissues (Nemmar et al., 2002). Some
particles are retained in the lungs; a comparison of autopsy lung tissue
from never-smoking women in Mexico City and Vancouver, Canada,
showed that the Mexican women had sevenfold higher geometric mean
particles per gram of lung tissue, a finding consistent with the higher PM
levels in Mexico City (Brauer et al., 2001).
Diesel particles have a mass median aerodynamic diameter of 0.2 m
and comprise a carbonaceous core (about 80% of the particle mass) and
organic chemicals (about 20% of the mass). Based on human lung models, about 10% of inhaled diesel exhaust particles are deposited in alveoli; the modeled particle mass deposited per unit alveolar surface area
per minute peaked at age 2 years at levels twice those of adults. Diesel
exhaust PM promotes release of cytokines, chemokines, immunoglobulins, and oxidants that can cause respiratory tract inflammation (Pandya
et al., 2002). In adult volunteers exposed to diesel exhaust, there were increased neutrophils and myeloperoxidase in sputum and increased expression of inflammatory response genes (IL-5, IL-8, and growth-regulated oncogene-␣) in bronchial tissue and bronchial wash cells. Diesel
exhaust PM contains several known carcinogens including polycyclic
mononitroarenes; exposure is widespread, as shown by the presence of
mononitroarene-hemoglobin adducts in blood samples from most persons, with levels being higher among urban compared to rural residents.
Among toxic chemicals in diesel exhaust, 3-nitrobenzanthrone is a particularly powerful mutagen.
Gases
Ozone reacts with polyunsaturated fatty acids and with sulfhydryl,
amino, and other tissue compounds and generates free radicals that cause
further oxidative tissue damage. Ozone exposure of healthy and atopic
children is associated with nasal lavage inflammatory reaction biomarkers (increased leukocyte counts, eosinophilic cationic protein, and myeloperoxidase activity) and urinary eosinophil protein X levels (a marker of
Outdoor Air
305
eosinophil activation). Controlled exposure of human volunteers to ozone,
NO2, or diesel exhaust causes an acute inflammatory response in airways
(Krishna et al., 1998). Acute effects in animals exposed to ozone at nearambient levels or to NO2 at levels far above ambient concentrations
include lipid peroxidation, inflammation and increased permeability of
airways, bronchial hyperreactivity to inhaled aeroallergens and bronchoconstrictive agents, lung function deficits, and reduced mucociliary
clearance. Monkeys exposed to high-level ozone for 90 days develop a
dose-related bronchiolar inflammatory response with thicker walls and
increased nonciliated bronchiolar epithelial cells, interstitial smooth muscle cells, macrophages, mast cells, and neutrophils; monkeys exposed for
a year develop persistent pulmonary fibrosis.
Acute exposure to high SO2 levels in controlled exposure studies also
causes an inflammatory response indicated by neutrophilia in bronchoalveolar lavage fluid. Ozone appears to be much potent than NO2 in triggering airway inflammation in healthy persons. But NO2 is a precursor of
photochemically produced stronger oxidants including ozone, nitric oxide
(NO), peroxyacetyl nitrate (PAN), and peroxypropionyl nitrate. Exposure
of human nasal or bronchial epithelial cells in vitro to NO2, ozone, and
diesel exhaust PM causes release of proinflammatory mediators; the latter
increase eosinophil chemotaxis and adherence to endothelial cells, possibly offering a mechanism for pollution-induced airway inflammation.
Carbon monoxide binds with high affinity to the iron present in
heme-proteins, that is, hemoglobin, myoglobin, and cytochromes. At low
CO levels, toxicity is mainly caused by tissue hypoxia due to conversion
of hemoglobin to carboxyhemoglobin, a relatively stable complex incapable of transporting oxygen to tissues. Secondary toxic mechanisms include binding of CO to myoglobin in heart and skeletal muscle and to cytochromes in various tissues, thereby interfering with oxygen storage and
mitochondrial energy generation.
Volatile Organic Carbons
Emissions by motor vehicles of incompletely combusted hydrocarbons
and evaporative fuel losses are the major sources of ambient urban VOCs.
There are over 600 VOCs having three main types of toxicity: (1) participation in atmospheric photochemical reactions that increase ground-level
ozone levels, (2) genotoxicity, and (3) carcinogenicity. In the absence of
anthropogenic air pollution, ozone levels remain relatively low because
ozone reacts with NO to produce NO2 and oxygen, reactions that equilibrate in the absence of VOCs. By reacting with NO, VOCs cause a net increase in ozone levels. Photochemically driven oxidation of VOCs pro-
306
CHILD HEALTH
AND THE
ENVIRONMENT
duces highly reactive hydrocarbon radicals that react with other smog
components to form PAN and aldehydes, reactive chemicals that contribute to eye and respiratory tract irritation.
Volatile organic chemicals include benzene, a known human carcinogen that causes acute myeloid leukemia after prolonged occupational
exposure at relatively high levels. It is not known if lifelong low-level exposure to benzene in ambient air increases the risk of cancer (see also
Chapter 10, Indoor Air, for further discussion of benzene).
Developmental Effects
Birth Weight
Several epidemiologic studies have demonstrated fairly consistent associations between low birth weight, preterm birth, IUGR, and forms of ambient air pollution including CO, SO2, and PM during pregnancy, particularly during the first trimester (Bobak, 2000; Bobak and Leon, 1999; Ha
et al., 2001; Maisonet et al., 2001; Ritz and Yu, 1999; Ritz et al., 2000; Wang
et al., 1997). There is some indication that IUGR is associated with ambient air PAHs, independent of PM10 levels (Dejmek et al., 2000). Sensitivity of the fetus during organogenesis to mutagens such as PAHs may
relate in part to their ability to inhibit trophoblast invasion of the endometrium and impair placental function. Ambient CO levels routinely exceeded 300 ppm in the Los Angeles basin during the early 1970s, levels
that could produce maternal blood carboxyhemoglobin levels equivalent
to those due to smoking 20 cigarettes per day (6%–10% of total hemoglobin), thereby reducing the fetal oxygen supply. Ambient CO levels also
increase fetal blood carboxyhemoglobin levels, further reducing the oxygen supply to rapidly growing fetal tissues. Animal models have shown
that low-level CO (75–150 ppm) caused reduced birth weight and postnatal weight gain in rats.
Stillbirths and Early Childhood Deaths
There is some epidemiologic evidence, mainly from time series ecologic
studies, of associations between multiple ambient air pollutants and stillbirths (Pereira et al., 1998), PM and SIDS (Woodruff et al., 1997), and multiple pollutants and early childhood respiratory deaths (Conceicao et al.,
2001).
Birth Defects
The first large, population-based study of outdoor air pollution and
birth defects was a record-based case-control study of orofacial clefts and
cardiac defects in the Los Angeles region (Ritz et al., 2002). There were
Outdoor Air
307
exposure–risk relationships between average ambient CO levels near the
mother’s residence during the second month of gestation and the risk of
cardiac ventricular septal defects (OR 3.0, CI 1.4–6.1, for the fourth versus the first CO quartiles) and aortic/pulmonary artery valve defects. No
clear relationships between these birth defects and other air pollutants
were apparent.
Respiratory Function and Diseases
Respiratory health effects from exposure to ambient air pollution during
childhood range from subtle transient lung function deficits to coughing,
asthma episodes, and irreversible lung function deficits. Supporting evidence for health effects of individual major air pollutants comes from controlled exposure, epidemiologic, and animal studies. Controlled exposure
studies have generally involved acute exposures of volunteers in test
chambers or through mouthpieces to air/contaminant mixtures over periods ranging from less than an hour to a few hours, small sample sizes
(often fewer than 20 subjects), intermittent exercise (to increase ventilation rates and contaminant doses), and spirometry to measure changes in
lung function indicators (see Chapter 10, Indoor Air, for a description of
spirometry tests).
Observational epidemiologic studies of ambient air quality enable assessment of the mixed, variable, and chronic exposures common in the
general population over periods of days to years, for example, major air
pollution episodes lasting a few days, intermittent exposure to summertime smog episodes, or average air quality during pregnancy or childhood. Commonly used variations of general epidemiologic study designs
in this field include (1) panel studies involving small cohorts (often fewer
than 100 subjects) that are intensively studied for a limited time (weeks
to months), for example, through daily symptom diaries, daily selfadministered peak expiratory flow (PEF) tests, frequent ambient air monitoring, and, sometimes, personal air monitoring, (2) ecologic time-trend
analytic studies of daily respiratory health events (e.g., emergency room
visits or physician contacts) and daily ambient air quality data, and
(3) hybrid studies—many cohort and cross-sectional studies have collected certain exposure and health outcome data for individual subjects
and regional ambient air quality data. Only a few epidemiologic studies
have used personal air sampling (and these were small, short-term studies) or repeated standardized respiratory function tests (instead of selfadministered PEF tests).
Lung function deficits after brief exposure to ambient air pollutants
or intense air pollution episodes are generally reversible within hours or
308
CHILD HEALTH
AND THE
ENVIRONMENT
days but may cause clinically significant effects in sensitive groups including children (especially asthmatics) and the elderly. Although small
lung function deficits are not clinically obvious, some children experience
relatively large effects and chronic exposure to intermittently high air pollution levels is associated with lung function growth deficits in preadolescent children, that is, reduced growth rates of lung capacity. Epidemiologic studies have not shown clear evidence of a threshold for the inverse
relationships between ozone or PM and lung function deficits in children.
The majority of studies have focused on the relation between ambient air
pollution and exacerbation of existing asthma rather than incident disease. Bronchial hyperreactivity is a major feature of asthma and appears
to predispose children to larger decrements of lung function during air
pollution episodes. See Chapter 10 (Indoor Air) for further discussion of
childhood asthma.
Particulate Matter
Ambient air PM, acid, and NO2 levels are highly correlated, precluding
conclusive attribution of health effects to any one pollutant. Reviews by
the California Environmental Protection Agency and the EPA concluded
that PM10 is associated with exacerbation of asthma and increased respiratory illness in children, with no evidence of a threshold (California Environmental Protection Agency, 2000; U.S. Environmental Protection
Agency, 2001a). Specific findings from these reviews and individual studies include
• Acute exposure—elevated daily PM levels appear to cause PEF deficits,
cough, asthma exacerbation, and acute respiratory illness in healthy
and/or asthmatic children.
• Chronic exposure—average PM (or acid aerosol) levels are associated
with lung function deficits, lung function growth deficits, respiratory
symptoms, acute respiratory illness, and school absenteeism in healthy
and/or asthmatic children.
Ozone
Ozone is a powerful oxidant and a pulmonary irritant that causes average FEV1 deficits of 5%–10% after controlled exposure (up to 3 hours) of
adolescents or young adults to levels as low as 80–120 ppb while exercising intermittently (see, e.g., McDonnell et al., 1999). After such exposure, some sensitive subjects develop much higher FEV1 deficits. Chronic
ambient ozone exposure appears to cause persistent lung function deficits,
but the exposure–risk relationship between average daily exposures and
Outdoor Air
309
chronic effects remains poorly defined (California Environmental Protection Agency, 2000). Associations between ozone and respiratory effects
are stronger during the summer because of the higher ambient ozone levels (Burnett et al., 2001; Kopp et al., 2000). Asthmatics are more susceptible than nonasthmatics to lung function deficits, airway inflammation,
and bronchial hyperreactivity after ozone exposure.
Personal air ozone levels are generally much lower than ambient
levels except during outdoor activities (Suh et al., 2000); thus, ambient
levels tend to misclassify individual ozone exposures, a factor that may
explain some inconsistencies in observational epidemiologic studies.
Ozone appears to produce stronger effects among active children in field
settings than in controlled exposure studies, possibly because of (1) longer
exposures, (2) potentiation by other ambient air pollutants, (3) persistence
of effects from exposures on previous days, and (4) persistence of a transient response associated with the daily peak of exposure. Findings concerning respiratory effects of ozone include the following:
• Acute exposure
° Daily ozone levels while exercising outdoors during summer are associated with lung function deficits among healthy and asthmatic
children, with exposure–risk relationships extending to levels below
the current ozone 1-hour NAAQS (120 ppb); a pooled analysis of six
summer camp panel studies of nonasthmatic children yielded a slope
of 0.50 ml FEV1/ppb ozone (p .0001) based on the average ozone
level during the hour before lung function measurements (Kinney et
al., 1996).
° Asthma exacerbations severe enough to require medical care and
school absenteeism are linked to ozone, especially after prolonged
episodes (see, e.g., Burnett et al., 2001; Gilliland et al., 2001).
• Chronic exposure
° Recent studies produced conflicting evidence of lung function growth
deficits among preadolescent children living in areas subject to high
summertime ozone levels (Frischer et al., 1999; Gauderman et al.,
2000).
° Average lifetime ozone exposure appeared to cause persistent lung
function deficits among nonsmoking, nonasthmatic college freshmen
(Kunzli et al., 1997).
° Asthma—incident asthma was strongly associated with involvement
in three or more sports in high-ozone areas in southern California
(McConnell et al., 2002); asthma symptom scores were much more
closely associated with personal air than ambient ozone levels
(Delfino et al., 1996).
310
CHILD HEALTH
AND THE
ENVIRONMENT
Sulfur Dioxide and Related Pollutants
Sulfur dioxide and sulfuric acid are strong irritants that appear to cause
transient lung function deficits, reduced bronchial mucociliary clearance,
respiratory symptoms, asthma exacerbation, and respiratory illness (California Environmental Protection Agency, 2000). Sulfur dioxide is very soluble in water and tends to be absorbed in the upper airways by subjects
at rest, but some of it reaches peripheral lung airways at higher ventilation rates. In air, SO2 is converted to sulfuric acid and sulfate, substances
associated with the fine PM fraction; the health effects attributed to fine
PM may, therefore, be partially caused by the sulfate component. Health
effects attributable to SO2 and derivatives include the following (see “Particulate Matter” above for effects of chronic exposure to acid aerosols):
• Lung function deficits—adolescent asthmatics are more sensitive than
nonasthmatics to lung function deficits after exposure to SO2 or sulfuric acid aerosol during controlled studies or intense air pollution episodes; the exposure–risk relationship between controlled SO2 concentration and FEV1 deficits appears to be continuous, with no evidence
of a threshold (World Health Organization, 2000).
• Respiratory symptoms—many asthmatics and some nonasthmatics
with allergies develop respiratory symptoms after brief exposure to SO2
in controlled exposure studies; such effects are found consistently in
exercising asthmatics at SO2 levels of 400 ppb or greater (California Environmental Protection Agency, 2000).
Nitrogen Dioxide
Nitrogen dioxide is a much weaker oxidant than ozone and appears to
be much less potent in causing airway inflammation in healthy persons.
A meta-analysis of controlled NO2 exposure studies of adults indicated a
trend for airflow rates to decrease among healthy persons at levels above
1 ppm and among asthmatics at lower levels. Exposure to NO2 at levels
as low as 260 ppb for 30 minutes increased the bronchial responsiveness
of asthmatics to subsequent challenge with common aeroallergens (California Environmental Protection Agency, 2000). Relatively high personal
air NO2 levels were associated with incident asthma in Montreal (InfanteRivard, 1993) and asthma exacerbation (Linaker et al., 2000). Prevalent
and recent-onset asthma among children in Austria were associated with
ambient NO2 levels in communities where children had lived for at least
2 years (Studnicka et al., 1997). Firm conclusions are not possible, however, because ambient NO2 is correlated with PM and other emissions
from high-temperature combustion processes (California Environmental
Protection Agency, 2000).
Outdoor Air
311
Mixed Pollutants
There is limited evidence of lung function deficits after controlled exposure of asthmatic children to mixed air pollutants, possibly because of the
small sample sizes and short duration of exposure and observation. For
instance, controlled exposure of young adult asthmatics to NO2 and ozone
or to NO2 and SO2 increased bronchoconstriction after challenge with
house-dust mite antigen. Several cross-sectional and longitudinal studies
showed associations between prolonged exposure to multiple ambient air
pollutants and lung function deficits and lung function growth deficits
(see, e.g., Gauderman et al., 2000; Schwartz, 1989). Asthma exacerabation
and respiratory symptoms have been associated with long-term exposure
to multiple air pollutants and traffic density indices (see, e.g., Boezen et
al., 1999; McConnell et al., 1999). Daily asthma events severe enough to
require medical care have been associated with exposure to two or more
major ambient air pollutants (see, e.g., Hajat et al., 1999).
Aeroallergens
Aeroallergens are usually proteins carried on inhalable particles, the major outdoor sources being plant pollens and mold spores. Approximately
80% of asthmatic children are skin-prick positive to one or more aeroallergens (see Chapter 10, Indoor Air, for discussion of indoor aeroallergens). Increased daily ambient air spore counts for specific fungi or total
spore counts have been linked to respiratory symptoms, asthma exacerbation, and daily asthma deaths. Other findings include
• Acute exposure—daily ambient air fungal spore counts and soybean
allergen concentrations have been associated with asthma exacerbation
severe enough to require medical care and asthma deaths among children and young adults (see, e.g., Dales et al., 2000; Downs et al., 2001).
• Chronic exposure—skin prick sensitization to Alternaria has been linked
to incident and prevalent asthma among children and young adults
(Gergen and Turkeltaub, 1992; Halonen et al., 1997).
Cancer
Diesel exhaust is a major source of fine and ultrafine particles in urban
ambient air and has been identified as a probable human carcinogen by
the IARC, the U.S. National Toxicology Program, and the EPA (International Agency for Research on Cancer, 1989; National Institute for Environmental Health Sciences, 2001; U.S. Environmental Protection Agency,
2000b). A cohort study of 1.2 million U.S. adults showed that each PM2.5
increment of 10 g/m3 was associated with an 8% increased risk of lung
312
CHILD HEALTH
AND THE
ENVIRONMENT
cancer mortality, independent of smoking, occupation, and other risk factors (Pope et al., 2002). The role of childhood ambient air exposures in
adult lung cancer is unknown.
The few epidemiologic studies of childhood cancer in relation to
ambient air pollution were relatively small and used relatively crude exposure measures such as neighborhood traffic density or parent’s occupation. There is limited evidence of associations between childhood leukemia and prenatal parental occupational exposure to benzene or gasoline
or childhood exposure to high-density traffic (Fig. 11–1). A large Danish
case-control study showed no association between maternal occupational
benzene exposure or residential NO2 exposure and childhood leukemia
or CNS cancers but did show associations with childhood Hodgkin’s disease (Raaschou-Nielsen et al., 2001). Two ecologic studies showed generally negative results but used very crude exposure indicators.
Exposures
Under the Clean Air Act, the EPA was required to set NAAQS, to ensure
that NAAQS are met through source control, and to monitor the effectiveness of the program. The act also required each state to establish a
network of air monitoring stations for criteria air pollutants (CO, NO2,
ozone, PM10, PM2.5, SO2, lead). The State and Local Monitoring Network
(SLAMS) is a network of about 4000 monitoring stations sited and designed to meet the needs of state and local air pollution control agencies
to support air program activities. The National Air Monitoring Network
(NAMS) is a subset of about 1000 stations in SLAMS sited and designed
to monitor the highest contaminant levels in areas of greatest population
density. A separate network of stations monitors ozone precursors (about
60 VOCs and carbonyl) in 24 large population areas with extremely high
ozone levels with two to five sites per area, depending on the population.
Ideally, air-monitoring stations would provide accurate indices relevant
to the occurrence of health effects in populations exposed to air pollutants. Obstacles to achieving this goal include the limited number of stations, sampling infrequency (e.g., 24-hour averages obscure short-term
peaks), small-area variations in air contaminants (e.g., CO levels may vary
substantially over short distances, depending on traffic density and other
factors), and the dependence of personal exposure levels on behavioral
factors.
The concentrations of pollutants in ambient air depend on the rate
of emissions and the efficiency of dispersion; daily variations in pollutant levels depend more on meteorological conditions than on changes in
odds ratio and 95% confidence limits
100
10
1
0.1
Denmark prenatal
Denmark child
UK - traffic
UK gasoline
stations
Sweden
UK -parent
occup exp
Denver 5,000/d
Denver 10,000/d
Shanghai
FIGURE 11–1. Epidemiologic studies of childhood leukemia and indicators of exposure to benzene and
other motor vehicle emissions in outdoor air. Shanghai (Shu et al., 1988), Denver traffic density (Savitz
and Feingold, 1989), U.K. parent occupational exposure (McKinley et al., 1991), Sweden NO2 levels
(Feychting et al., 1998), U.K. proximity to gasoline stations and main roads (Harrison et al., 1999), Denmark prenatal maternal and childhood residential exposure (Raaschou-Nielsen et al., 2001).
CHILD HEALTH
314
AND THE
ENVIRONMENT
emission rates. Wind speed is the major factor influencing efficiency of dispersion; other factors depend on local conditions and include the effective
trapping of atmospheric pollutants in valleys or basins by wind or atmospheric thermal inversions. Trends in ambient air levels of major air pollutants in the United States are shown in Table 11–2. Annual average ambient air levels decreased by 36% for CO and SO2, 18% for PM10, and 10%
for NOx (NO and NO2) but by only 4% for ozone during the 1990s.
Particulate Matter
Early epidemiologic studies measured (TSP), that is particles up to about
40 m in diameter. The TSP values include nonrespirable particles and
are subject to measurement artifacts; since the late 1970s, most studies
have measured PM10, PM2.5, and, occasionally, PM1.0. Polycyclic aromatic
hydrocarbons, other toxic organic carbon compounds, and sulfate occur
mainly in PM2.5. Ambient air PM2.5 concentrations tend to be relatively
uniform across large cities in eastern United States during the summer
months, unlike PM10 levels, which show more spatial variation. The
dependence of PM2.5 levels on traffic volume is reflected in peak hourly
levels during morning and evening rush hours and peak 24-hour averages during midweek, particularly at roadway monitoring sites. Both
PM2.5 and PM10 particles penetrate buildings, with indoor to outdoor concentration ratios approaching unity at high air exchange rates. Levels of
PM vary by microenvironment; for example, geometric mean PM10 levels in subways and buses may be two to four times those in ambient outdoor air.
TABLE 11–2. Percent Change in Annual Average Emission and
Ambient Air Pollutant Levels, United States, 1991–2000
Percent Change
Pollutant
Emissions
Ambient Air Levels
CO
5
41
NOx
3
11
Ozone (1-hr)
na
10
VOCs
16
na
PM10
6a
19
SO2
24
37
Source: U.S. Environmental Protection Agency (2001b).
a Direct
PM10 emissions.
Outdoor Air
315
Ozone
Elevated ambient ozone levels arise mainly from photochemical reactions
of air pollutants under summer conditions. Indoor ozone levels are usually much lower than ambient levels because ozone is highly reactive and
the indoor environment is rich in substances that react with ozone. Children are at risk of ozone exposure because they are often outdoors and
physically active during summer, when ozone levels are highest; for example, children in summer camps may inhale ozone doses close to those
that produce lung function deficits and respiratory symptoms in adults
during controlled exposure studies. Annual average ozone levels in 24
communities in Canada and the United States varied from 16 to 35 ppb;
1-hour ozone levels exceeded the Canadian guideline (100 ppb) in 18 communities and the U.S. NAAQS (120 ppb) in 10 communities (Spengler et
al., 1996). Monitoring of children in southern California during the summer ozone season showed that average personal summer ozone exposures were three to four times winter levels.
Sulfur Dioxide and Strong Acidity
Sulfurous and sulfuric acids, partially neutralized ammonium bisulfate
salts, and fine PM are formed in atmospheric reactions between SO2 and
other fossil fuel emissions. Ambient air concentrations of these pollutants
are highly correlated; for example, sulfates may comprise 20% or more
of the PM2.5 mass in regions such as the northeastern United States subject to air pollution from combustion of high-sulfur fuels, especially coalburning electric power utility and diesel engine emissions. Total airborne
acidic species are higher in summer because of atmospheric conversion of
SO2 and NO2, respectively, to sulfuric acid (H2SO4) and nitric acid (HNO3)
through photochemical processes. The main vapor phase strong acids in
the southwestern United States are HNO3 and hydrochloric acid (HCl).
Annual average particle strong acidity levels in 24 Canadian and U.S. communities ranged from below detection limits to about 50 nM/m3, but single-day levels often exceeded 150 nM/m3 during the summer in the “sulfate belt” comprising major portions of Tennessee, Kentucky, Ohio,
Virginia, Pennsylvania, New York, and West Virginia (Spengler et al., 1996).
Nitrogen Dioxide
The annual average ambient NO2 level in U.S. cities in 1998 was 18 ppb
(U.S. Environmental Protection Agency, 2000d). Urban outdoor NOx levels display diurnal variation related to morning and late afternoon rush
hour traffic but little seasonal variation. In cities with high traffic density,
CHILD HEALTH
316
AND THE
ENVIRONMENT
1-hour outdoor NO2 levels reach about 100 ppb and occasionally reach
400 ppb during urban air pollution episodes.
Volatile Organic Chemicals
Modeled estimates of long-term outdoor air levels of 148 VOCs in the
United States indicated an average of 14 VOCs in each census tract for
which concentrations exceeded health-based benchmarks for cancer or
noncancer outcomes (Woodruff et al., 1998). Formaldehyde, benzene, and
1,3-butadiene levels exceeded cancer benchmark levels in over 90% of all
census tracts (Table 11–3). About 10% of census tracts had one or more
VOCs at levels conferring a cancer risk of at least 104.
Personal Air
Individual exposure to air pollutants varies with ambient and indoor air
pollutant concentrations in the microenvironments that a person moves
through during daily activities, the time spent and activities undertaken
in each microenvironment, and ventilation rates. Personal air PM2.5 levels among nonsmoking children and adults living in nonsmoking homes
were correlated with ambient PM2.5 but not with personal gaseous pollutant levels (Sarnat et al., 2001). Personal exposures to reactive gaseous
pollutants are driven more by time spent outdoors than by ambient air
levels, suggesting that they should not be confounders of ambient PM2.5
and health effects. A person may be exposed to more ozone during an
TABLE 11–3. Hazardous Air Pollutants with Ambient Concentrations Above Cancer
Benchmark Levels
HAP
Average Level
(g/m3)
Cancer Benchmark
(g/m3)
Ratio
% of
Census Tracts
Above Level
Formaldehyde
0.25
0.077
3.2
94
Benzene
0.48
0.12
4.0
92
Ethylene dichloride
0.061
0.038
1.6
21
Chloroform
0.083
0.043
1.9
8
Carbon tetrachloride
0.88
0.067
13
3
Ethylene dibromide
0.0077
0.0045
1.7
1
Bis(2-ethylhexyl) phthalate
1.6
0.25
6.4
1
Methyl chloride
1.2
0.56
2.2
1
Source: Woodruff et al. (1998).
Outdoor Air
317
hour of vigorous exercise outdoors than while spending several hours indoors with low average physical activity.
Findings from monitoring of children’s personal air contaminant levels include:
• Particulate matter—personal air PM2.5 levels among children without
ETS exposure were strongly correlated with but higher than ambient
levels.
• Ozone—in southern California, average personal air ozone exposure
levels were about 22 ppb during summer months and 6 ppb during
winter months.
• Nitrogen dioxide—personal air exposures among children tend to be
more closely related to indoor sources (gas appliances, smokers) than
to ambient NO2 levels.
• Volatile organic chemicals—personal air benzene levels were associated
with front-door ambient levels, riding in cars, moped driving, and refueling of cars (Raaschou-Nielsen et al., 1997).
Biomarkers
There are relatively few biomarkers of ambient air pollution uptake.
Hemoglobin adducts of 1-nitropyrene and 2-nitrofluorene and urinary
1-hydroxypyrene are non-specific biomarkers of exposure to diesel PM
(other sources include mainstream tobacco smoke and cooked animal
fats). Urinary 1-hydroxypyrene was detected in the majority of Harlem
seventh-grade students, including nonsmokers (Northridge et al., 1999).
Urinary levels of trans-muconic acid (a metabolite of benzene) among
inner-city children were associated with time spent playing near the street
but not with ETS exposure or urinary cotinine (Weaver et al., 1996). Carbon monoxide binds tightly to hemoglobin, forming carboxyhemoglobin
(half-life of 2–6 hours); carboxyhemoglobin blood levels reflect the concentration and duration of CO exposure, exercise, and other factors and
can be estimated by direct blood analysis or by measuring CO in exhaled
breath.
Risk Management
Sources
The major sources of ambient air pollutants in the United States are motor vehicles (53% of NOx, 79% of CO, 44% of VOCs) and electricity generation (67% of SO2) (Table 11–4). Emissions from these sources are ex-
CHILD HEALTH
318
AND THE
ENVIRONMENT
TABLE 11–4. Major Sources of Ambient Air Pollutant Emissions, United States, 1998
(Percent by Source)
PM2.5a
Source
SO2
NO2
CO
VOCs
Electricity generation
2.0
67.3
25.8
0.5
0.3
Motor vehicles
7.3
7.2
53.3
78.6
43.5
Industries
Solvent use
Biogenic sources
11.7
25.3
19.9
9.6
12.6
0.1
0.1
0.1
0.1
29.5
9.5
0.1
0.1
0.1
0.1
Other
69.4b
0.3
1.8
11.4c
14.1c
Total (percent)
100.0
100.0
100.0
100.0
100.0
Total (thousands of tons)
8,379
19,647
24,454
89,454
17,917
Source: U.S. Environmental Protection Agency (2000c).
a Directly
emitted particles; secondary fine particles also formed from SO2, NOx, NHs, and VOCs in the
atmosphere.
b Includes
agriculture, forestry, fugitive dust (roads, construction), and other combustion.
c Includes
domestic gasoline-powered devices, recreational marine vessels, aircraft, and other sources.
tremely complex, with hundreds of identified compounds, including PM,
gases, VOCs, and metals. See Chapters 4 and 5 (Metals) for further discussion of selected metals.
Particulate Matter
Exhaust emissions from gasoline and diesel engines of similar power contain similar types of toxic chemicals, but diesel engines are more important sources of NO2, PM, and certain PAH carcinogens (e.g., nitroarenes)
than gasoline engines with catalytic converters. Diesel engines produce
69% of the PM10 emissions from motor vehicles in the United States (U.S.
Environmental Protection Agency, 2000c). Catalytic converters of the type
used with gasoline vehicles substantially reduce total PM and VOC emissions but increase emissions of ultrafine particles. In many rural temperate and northern regions, wood burning for heat and cooking and forest
fires, respectively, are major sources of ambient PM during winter and
summer months.
Ozone
Ground-level ozone has two main origins: downward transport of ozone
from the stratosphere and photochemical reactions near the earth’s surface driven by anthropogenic emissions. The former contributes about
5–15 ppb to background terrestrial ozone levels. The major reactions driving the formation of ground-level ozone are illustrated in Figure 11–2; the
Outdoor Air
319
Motor vehicles emit NO2 and VOCs
NO2 + solar energy
O + O2
NO + O (ground state oxygen)
O3 (ozone)
NO + O3
NO2 + O2
O3 + solar energy
Oc + H2O
2 OH* (hydroxyl radicals)
CO + OH*
CO2 + HO2* (hydroperoxy radical)
HO2* + NO
NO2 + OH*
O3 + NO2
O2 + NO3* (nitrate radical – forms mainly at night)
2 NO3* + H2O
HNO2 + hv
R + OH*
R* + O2
O2 + Oc (charged oxygen)
2HNO2 + O3
OH* + NO
R* + H2O (R = a volatile organic carbon)
ROO* (peroxy radical)
ROO* + NO
NO2 + OH*
SO2 + 2 OH*
RO + NO2
HNO3 (nitric acid – forms inorganic vapor acid)
H2SO4
FIGURE 11–2. Formation of ground-level ozone and related photochemical oxidative reactions.
net result is to produce peak ground-level NO2 levels by midmorning and
maximum ozone levels by midday or early afternoon. Ozone from urban
sources is transported over considerable distances to rural downwind locations and is relatively persistent there because of lower NO levels (NO
reacts with and depletes ozone).
Sulfur Dioxide
Coal contains about 4% sulfur by weight and was the first fossil fuel to
be extensively exploited during the nineteenth-century Industrial Revolution. Initially, coal was used primarily in industrial boilers to create
steam to power machinery. The 1952 London smog incident was caused
by the combination of a prolonged inversion and widespread use of lowgrade coal for home heating; at its peak, smog extended for 30 km around
London and reduced visibility to 1–5 m. During recent decades, coal has
been widely used in electricity-generating utilities and releases more SO2
than either oil or gas. A 1000 MW plant burns about 700 tons of coal per
320
CHILD HEALTH
AND THE
ENVIRONMENT
hour and emits about a half million tons of SO2 annually; similar amounts
are generated by utilities burning high-sulfur oil. Other sources of SO2 include pulp and paper mills, oil refineries, and smelters.
Levels of SO2 vary by season, with generally higher levels during
winter months. Total airborne acidic species, however, are higher in summer months because of atmospheric conversion of SO2 and NO2, respectively, to H2SO4 and HNO3 through photochemical processes (Figure 11.2).
The oxidation of SO2 by molecular oxygen occurs very slowly in clean,
dry air but much faster in air containing particulates and moisture. Sulfuric acid reacts with atmospheric ammonia (from animal, human, and
other sources), forming ammonium sulfate and bisulfate salts that accumulate in the PM2.5 fraction.
Nitrogen Oxides
Nitric oxide and NO2 are formed at high combustion temperatures from
nitrogen and oxygen in air, electric utilities and motor vehicles being the
two main sources. Diesel engines produce 50% of the NO2 emissions from
motor vehicles. Emissions of NOx increased 17% in the United States between 1970 and 1999, mainly from heavy-duty diesel vehicles and coalfired electric generation plants. Natural NOx sources include stratospheric
intrusion, oceans, lightning, soil, and wildfires. In the United States, natural and anthropogenic NOx emissions, respectively, are about 2.2 and
21.4 Tg annually.
Carbon Monoxide
Motor vehicle interiors have the highest average CO levels (10–29 mg/m3
or 9–25 ppm) of all microenvironments; commuting exposures are highly
variable, with some commuters breathing CO in excess of 40 mg/m3
(35 ppm). Important sources of CO exposure for children include living
in homes with elevated levels (related to cigarette smoking, inadequately
ventilated nonelectric cooking and heating appliances, vehicle start-up
and idling in attached garages), commuting in cars or buses, and engaging in physical activity (e.g., playing, running, bicycling) adjacent to hightraffic roadways.
Volatile Organic Carbons
Leading sources of VOC emissions are vegetation, motor vehicles, consumer and commercial solvent use, open burning and forest fires, dry
cleaning, and pulp and paper production. The VOCs emitted by vegetation are mainly monoterpenes and isoprene, which are widely dispersed,
while anthropogenic VOCs are concentrated in population centers. Older,
poorly maintained vehicles have far higher VOC emissions than newer
models. Benzene and MTBE were added to gasoline at low levels in the
Outdoor Air
321
late 1970s during the phase-out of tetraethyl lead to enhance fuel combustion. Children may be exposed to MTBE in air (from incomplete fuel
combustion and other emissions into ambient air) or water (groundwater
contaminated by leaking underground storage tanks or surface water
contaminated by recreational water crafts). See also Chapter 5 (Metals—
Mercury, Arsenic, Cadmium, and Manganese) for discussion of manganese exposure from emissions of motor vehicles using gasoline containing MTBE. Monitoring of ambient air and personal exposures to toxic
VOCs in the United States showed that the major sources were motor vehicle exhaust, gasoline vapors, or ETS for personal exposures and motor
vehicle exhaust or gasoline vapors for ambient air levels (Anderson et al.,
2001; Wallace, 1996).
Over the period 1900–1998, PM10 emissions increased until about 1950
and then decreased by about 50% by 1998, while emissions of other pollutants generally increased until about 1970 (Fig. 11–3); total emissions of
the six principal air pollutants decreased 31% between 1970 and 1999,
mainly due to reduced SO2 and VOC emissions. During 1990–1999, emissions of PM10, PM2.5, SO2, CO, and VOCs decreased by 9%–17% but NOx
emissions increased slightly (2%) (Table 11–2). The observed emission
trends are likely related to several factors including:
• Coal consumption by electric utilities and their PM10 and SO2 emissions
approximately doubled every decade from 1940 to 1970; subsequent regulation led to reduced SO2 and PM10 emissions by electric utilities.
Emissions (millions of tons)
35000
30000
25000
20000
15000
10000
5000
0
1900
1920
1940
1960
1980
1998
Year
NOX
VOC
SO2
PM10
FIGURE 11–3. Trends in national emissions of major air pollutants, United States,
1900–1998. PM10 represents directly emitted PM from sources other than fugitive
dust, agriculture, or forestry (U.S. Environmental Protection Agency, 2000c).
322
CHILD HEALTH
AND THE
ENVIRONMENT
• Fossil fuel combustion increased during periods of strong economic
growth and decreased during recessions; for example, emissions of SO2
and VOCs declined during the 1930s and after the 1974 energy crisis
(Figure 11–3).
• Emission reductions per motor vehicle (related to the introduction of
catalytic converters in 1975, improved fuels, and new engine technologies) have been offset by greatly increased numbers of motor vehicles and greater fuel consumption.
• Ground-level ozone levels have remained fairly constant since the 1970s
in parallel with levels of their precursors, that is, VOCs and NOx.
• Decreased use of wood for residential heating and cooking has contributed to the long-term decline in PM10.
Limited data on emissions of hazardous air pollutants by the EPA indicate decreases of 23% in total HAPs, 66% in tetrachloroethylene, and 39%
in benzene during the mid-1990s. Ambient air benzene levels in Canadian urban areas decreased by about 25% during 1990–1997.
Intervention
Air Quality Guidelines and Standards
Air quality standards legally define clean air by specifying maximum concentrations and durations of pollutants that pose a threat to public health.
The WHO developed air quality guidelines for Europe in 1987 to help
countries develop their own national air quality standards and broadened
the guidelines in 2000 to address air quality issues in developing countries (World Health Organization, 2000). The WHO air quality guidelines,
the EPA NAAQS, and the California ambient air quality standards (AAQS)
are shown in Table 11–5. The main differences are the higher values in the
United States for annual NO2 (100 vs. 40 g/m3) and 24-hour SO2 (365
vs. 125 g/m3), but the WHO values are proposed guidelines and the
EPA NAAQS are legally binding standards. The number of persons living in U.S. counties where 1-hour or 8-hour average ambient levels of any
criteria air pollutant exceeded NAAQS in 1999 were, respectively, 62 million and 125 million.
The EPA is required by the Clean Air Act to review NAAQSs every
5 years to identify any revisions needed to adequately protect public
health with an adequate margin of safety. The EPA NAAQS review process is very lengthy and can substantially delay needed revisions. For instance, the EPA introduced revised ozone and PM standards in 1997, but
they have not been implemented because of a legal challenge initiated
by the American Trucking Association. The EPA planned to (1) phase out
Outdoor Air
323
TABLE 11–5. Guidelines for Major Air Pollutants
Pollutant
Type of
Average
Ozone
1-hour
8-hour d
Units
EPA
NAAQSa
California
AAQSb
WHO
Guidelinec
g/m3
(ppb)
g/m3
(ppb)
235
(120)
157
(80)
180
(90)
NA
NA
120
(61)
PM10
24-hour
Annual
g/m3
g/m3
150
50
50
30e
NA
NA
PM2.5
24-hour d
Annual d
g/m3
g/m3
65
15
NA
NA
NA
NA
SO4
24-hour
g/m3
NA
25
NA
1-hour
g/m3
NA
1300
(500)
365
(140)
80
(30)
655
(250)
NA
NA
(ppb)
g/m3
(ppb)
g/m3
(ppb)
g/m3
(ppb)
105
(40)
NA
125
(48)
50
(19)
mg/m3
(ppm)
mg/m3
(ppm)
40
(35)
10
(9)
23
(20)
10
(9)
30
(26)
10
(9)
g/m3
(ppb)
g/m3
(ppb)
NA
470
(250)
NA
200
(106)
40
(21)
SO2
3-hour
24-hour
Annual
CO
1-hour
8-hour
NO2
1-hour
Annual
a Source:
100
(53)
NA
U.S. Environmental Protection Agency (2001c).
b Source:
California Environmental Protection Agency (2000).
c Source:
World Health Organization (2000).
dA
1999 federal court ruling blocked implementation of these standards, proposed by the EPA in 1997;
the EPA has appealed this decision to the U.S. Supreme Court (U.S. Environmental Protection Agency,
2001c).
e Geometric
mean.
the 1-hour primary ozone standard with an 8-hour standard of 80 ppb
based on the 3-year average of the annual fourth highest daily maximum
8-hour ozone concentrations and (2) add new PM2.5 standards for annual
(15 g/m3) and 24-hour (65 g/m3) averages and retain existing PM10
standards (Table 11–5). As of early 2002, the EPA had still not implemented
these proposals. About 121 million persons in the United States lived in
counties that exceeded the NAAQS for 8-hour ozone in 2000 (Table 11–6).
CHILD HEALTH
324
AND THE
ENVIRONMENT
TABLE 11–6. Population in Counties that
Exceeded NAAQS, United States, 2000
Pollutant
Number of People
(millions)
CO
9.7
NO2
0
Ozone a
8-hour
1-hour
81.5
34.7
PM10
8.3
PM2.5
75.0
SO2
Any NAAQS
0
121.4
Source: U.S. Environmental Protection Agency
(2001b).
aA
1999 federal court ruling blocked implementation of these standards, proposed by the EPA in
1997; the EPA has appealed this decision to the
U.S. Supreme Court (U.S. Environmental Protection Agency, 2001c).
Under the Children’s Environmental Health Protection Act, California has reviewed state AAQSs to assess whether they adequately protect
infants and children (California Environmental Protection Agency, 2000).
The California Environmental Protection Agency conducted critical reviews of potential health effects of eight major air pollutants and concluded that health effects may occur in infants, children, and other potentially susceptible subgroups at or near the maximum levels specified
by California AAQSs. The Agency concluded that PM10, ozone, and NO2
posed the greatest public health threats at current AAQS levels and should
be included in the first tier of pollutants for review and potential changes
in air quality standards; the review of PM10 included sulfates because they
occur in this component. Pollutants in the second tier included lead, CO,
hydrogen sulfide, and SO2. The review of tier one pollutants concluded
that
• PM10—has the potential to produce health effects in infants and children including reduced birth weight, premature birth, asthma exacerbation, acute respiratory infections, and death; most Californians are
exposed to levels above the state standard during parts of the year.
• Ozone—has adverse effects on lung function and exacerbates asthma
and other respiratory illnesses; many Californians are exposed to ozone
levels above the state standard during the day in summer months.
Outdoor Air
325
• Nitrogen dioxide—may enhance the response of asthmatics to aeroallergens at NO2 levels close to the standard; NO2 levels in California are
occasionally close to the state standard.
• Carbon monoxide and SO2—SO2 levels are very low relative to the standard throughout the state, and ambient CO levels are only weakly related to personal air levels; existing AAQSs are reasonably protective
of public health, including that of infants and children.
Motor Vehicles
Motor vehicle emissions can be reduced by three main means: engine
modifications, exhaust system technologies, and fuel reformulation. Introduction of catalytic converters and lead-free gasoline beginning in the
United States in 1976 (in response to regulations under the 1970 Clean Air
Act) substantially reduced emissions of VOCs per car, but total motor vehicle VOC emissions have declined only slowly during recent years because of the greatly increased number of vehicles. During smog episodes,
some European cities limit car use to alternate days based on odd/even
vehicle registration numbers. The Ontario (Canada) Medical Association
has called for measures to reduce health impacts of air pollution on children and adults including California-level emission standards for light
and heavy-duty vehicles, expanded vehicle inspection and maintenance
programs, and tougher standards for sulfur levels in gasoline.
Fuels
The 1990 amendments to the U.S. Clean Air Act required the use of oxygenated gasoline in areas that did not meet the federal ambient air standard for CO, a problem most frequent during winter months. To meet the
requirement for 2.7% oxygen content in gasoline, gasoline producers may
add 15% MTBE, 7.5% ethanol, or other oxygenates. Oxygenates reduce
total VOC emissions, but alcohol increases emissions of formaldehyde.
Use of oxygenated gasoline appears to reduce CO levels when ambient
temperatures are above 50°F but has little benefit at lower temperatures.
Quantification of the public health benefits of fuel oxygenates is problematic because exposure–response relationships for ozone and respiratory health are not well understood for the population subgroups most
likely to be affected.
Leaking underground gasoline storage tanks have produced high
concentrations of benzene, MTBE, and other VOCs in groundwater. The
first publicized incident (1996) involved MTBE contamination of two well
fields supplying 50% of the drinking water to Santa Monica, California.
Air emissions of MTBE from motor vehicle exhaust and evaporation from
gasoline partition into water and enter the hydrologic cycle, for example,
via storm drain runoff. Given the low biodegradability of MTBE and its
326
CHILD HEALTH
AND THE
ENVIRONMENT
low affinity for soil particles, it can remain free in water (especially
groundwater) for extended times and quickly migrate long distances. California has served notice of its intent to phase out MTBE by the end of
2002, and an expert group advised the EPA to remove the requirement
for oxygenates from the Clean Air Act because they provide less benefit
in modern motor vehicles. In Canada, federal regulations in 1997 limited
the gasoline benzene content to an annual average of 1% and to 1.5% at
any one time.
Industry
In response to the London smog episode, the U.K. government introduced
its first Clean Air Act in 1956 to control domestic sources of air pollution.
The introduction of cleaner coals and the increased use of North Sea gas
helped reduce SO2 emissions; also, some electric power stations were relocated from urban to more rural areas. The U.K. Clean Air Act of 1968
required the use of tall chimneys for industries burning fossil fuels, but
increased stack height has increased residence time in air and promoted
conversion of atmospheric SO2 to sulfate compounds including acids. After the U.S. EPA was formed in 1970, it was immediately engaged in meeting requirements under the Clean Air Act to establish national air quality and emission standards. This led to the NAAQS for six criteria
pollutants (CO, NO2, PM, lead, SO2, and ozone) that are proven human
health threats. In 1995, the EPA finalized a rule aimed at reducing NOx
emissions from coal-fired power plants by over 400,000 tons per year during 1996–1999 and by over 2 million tons per year beginning in 2000.
The EPA has developed objectives to reduce potential health risks of
HAPs including cancer, birth defects, and reproductive effects related to
emissions from stationary and mobile sources (U.S. Environmental Protection Agency, 2000e). To achieve these objectives, the EPA plans to
(1) develop standards to address sources of 33 HAPs that present the
greatest threats to public health, (2) ensure that sources accounting for
90% of area-source emissions are subject to standards, with specific priorities to reduce benzene and formaldehyde emissions from motor vehicles, (3) conduct air toxics assessments to identify areas of concern and
priorities, (4) track progress through emission inventories and monitoring networks (there is no national ambient air quality monitoring network
to measure ambient HAP levels), (5) initiate national and local initiatives
focusing on multimedia and cumulative risks within urban areas, and
(6) educate and engage state, local, and other stakeholders.
Public Education
The EPA has developed an air quality index (AQI) to communicate information on potential health risks of local air levels of the five criteria
TABLE 11–7. Air Quality Index
Pollutant Concentration at Upper Limit of AQI Category
AQI
Air Quality Level
0–50
Color
Good
Green
Ozone
(8-hr)
(ppb)
Ozone
(1-hr)
(ppb)
PM2.5
(24-hr)
(g/m3)
PM10
(24-hr)
(g/m3)
CO
(8-hr)
(ppm)
SO2
(8-hr)
(ppb)
NO2
(24-hr)
(ppm)
0–64
—
0–15.4
0–54
0–4.4
0–34
—
51–100
Moderate
Yellow
65–84
—
15.5–40.4
55–154
4.5–9.4
35–144
—
101–150
Unhealthy for sensitive
groups
Orange
85–104
125–164
40.5–65.4
155–254
9.5–12.4
145–224
—
151–200
Unhealthy
Red
105–124
165–204
65.5–150.4
255–354
12.5–15.4
225–304
—
201–300
Very unhealthy
Purple
125–374
205–504
150.5–350.4
355–524
15.5–40.4
305–804
0.65–1.64
301–500
Hazardous
Maroon
NAa
ⱖ505
ⱖ350.5
ⱖ525
ⱖ40.5
ⱖ805
ⱖ1.65
Source: U.S. Environmental Protection Agency (1999, 2000a).
a Not
available (1-hr ozone levels are used to identify hazardous ozone levels).
328
CHILD HEALTH
AND THE
ENVIRONMENT
air pollutants other than lead. The AQI is designed with a range of 0–500
for each criterion air pollutant and six color-coded health concern categories (Table 11–7). The AQI for a given day is the highest of the five individual pollutant AQIs. For large metropolitan areas (350,000 people),
state and local agencies are required to report AQIs to the public daily
and, if two or more individual pollutant AQIs exceed 100, specify susceptible population groups such as children and people with asthma or
heart disease. Cautionary statements to the public vary somewhat, depending on the pollution profile.
Conclusions
Proven health effects
• Acute exposure effects
° Airway inflammation—controlled exposure of human volunteers to
ozone, NO2, or diesel exhaust causes an acute inflammatory response
in airways
° Transient lung function deficits
After controlled exposure to individual air pollutants (ozone, SO2),
especially among asthmatics
After same-day exposure to summer elevated ozone levels while engaged in outdoor activities or after moderately severe PM air pollution episodes among healthy and asthmatic children
Bronchial
hyperreactivity—after brief controlled exposure to high
°
levels of individual air pollutants (ozone, NO2)
° Exacerbation of asthma—after controlled exposure to SO2 and after
same-day exposure to increased ambient air pollution (ozone, PM10)
° Acute respiratory disease—after ambient air pollution increases (PM10)
Unresolved Issues and Knowledge Gaps
• Acute exposure effects
° Lung function deficits—limited evidence from controlled exposure to
high NO2 levels
° Bronchial hyperreactivity—limited evidence from controlled exposure
studies that moderately elevated ozone and high NO2 or mixed pollutant exposure levels increase bronchial reactivity to aeroallergens
° Asthma exacerbation—limited evidence of an association with ambient air fungal spore counts and soybean allergen
° Acute childhood respiratory disease—limited evidence of an association between daily respiratory disease events serious enough to re-
Outdoor Air
329
quire medical care or to cause death and daily concentrations of individual and multiple outdoor air pollutants (SO2, ozone, PM10)
• Prenatal maternal exposure effects
° Preterm delivery or intrauterine growth retardation—limited evidence of associations with first trimester exposure (CO, SO2, PM)
Stillbirths—inadequate
evidence of an association with exposure to
°
multiple outdoor air pollutants
° Birth defects—suggestive evidence of an association between orofacial clefts and cardiac defects and first trimester CO exposure in one
study
• Chronic exposure effects
° Lung function deficits and respiratory symptoms—limited evidence
of associations with residence in regions with intermittent exposure
to elevated ozone, PM10, particle strong acidity, multiple pollutants,
or residence in areas of high traffic density (in both healthy and asthmatic children)
° Lung function growth deficits—limited evidence of an association
with residence for at least a few years in areas with high ambient pollution levels (PM10, ozone, multiple pollutants)
° Incident asthma—limited evidence of associations with outdoor activity in high-ozone regions and personal NO2 exposure levels; inadequate evidence of association with the fungus Alternaria.
° Childhood cancer—inadequate evidence of an association between
leukemia or other childhood cancers and traffic density or ambient
benzene levels
• Knowledge gaps—need epidemiologic studies with improved exposure assessment, statistical power, and biomarkers of susceptibility to
assess the relative importance of ambient air pollutants and other risk
factors for incident asthma, other respiratory diseases, and other potential health effects including developmental effects and childhood
cancer
Risk Management Issues
• Prevention—need to prevent exposure of children and pregnant
women to ambient air pollutant levels in excess of existing standards;
interventions should target major sources including motor vehicles and
industry
• Monitoring—need to monitor personal air exposure levels of children
to ambient air pollutants, prevalence of lung function deficits and lung
function growth deficits, and incidence of asthma and other respiratory diseases
330
CHILD HEALTH
AND THE
ENVIRONMENT
References
Anderson MJ, Miller SL, Milford JB. (2001). Source apportionment of exposure to
toxic volatile organic compounds using positive matrix factorization. J Expo
Anal Environ Epidemiol 11:295–307.
Bobak M. (2000). Outdoor air pollution, low birth weight, and prematurity. Environ Health Perspect 108:173–6.
Bobak M, Leon DA. (1999). Pregnancy outcomes and outdoor air pollution: an
ecological study in districts of the Czech Republic 1986–8. Occup Environ Med
56:539–43.
Boezen HM, van der Zee SC, Postma DS, Vonk JM, Gerritsen J, Hoek G, Brunekreef
B, Rijcken B, Schouten JP. (1999). Effects of ambient air pollution on upper and
lower respiratory symptoms and peak expiratory flow in children. Lancet
353:874–8.
Brauer M, Avila-Casado C, Fortoul TI, Vedal S, Stevens B, Churg A. (2001). Air
pollution and retained particles in the lung. Environ Health Perspect 109:
1039–43.
Burnett RT, Smith-Doiron M, Stieb D, Raizenne ME, Brook JR, Dales RE, Leech
JA, Cakmak S, Krewski D. (2001). Association between ozone and hospitalization for acute respiratory diseases in children less than 2 years of age. Am
J Epidemiol 153:444–52.
California Environmental Protection Agency. (2000). Adequacy of California ambient air quality standards: Children’s Environmental Health Protection Act.
Sacramento: Office of Environmental Health Hazard Assessment.
Conceicao GM, Miraglia SG, Kishi HS, Saldiva PH, Singer JM. (2001). Air pollution and child mortality: a time-series study in São Paulo, Brazil. Environ
Health Perspect 109(Suppl 3):347–50.
Dales RE, Cakmak S, Burnett RT, Judek S, Coates F, Brook JR. (2000). Influence of
ambient fungal spores on emergency visits for asthma to a regional children’s
hospital. Am J Respir Crit Care Med 162:2087–90.
Dejmek J, Solansky I, Benes I, Lenicek J, Sram RJ. (2000). The impact of polycyclic
aromatic hydrocarbons and fine particles on pregnancy outcome. Environ
Health Perspect 108:1159–64.
Delfino RJ, Coate BD, Zeiger RS, Seltzer JM, Street DH, Koutrakis P. (1996). Daily
asthma severity in relation to personal ozone exposure and outdoor fungal
spores. Am J Respir Crit Care Med 154:633–41.
Dockery DW, Pope CA III. (1994). Acute respiratory effects of particulate air pollution. Annu Rev Public Health 15:107–32.
Downs SH, Mitakakis TZ, Marks GB, Car NG, Belousova EG, Leuppi JD, Xuan
W, Downie SR, Tobias A, Peat JK. (2001). Clinical importance of Alternaria exposure in children. Am J Respir Crit Care Med 164:455–9.
Feychting M, Svensson D, Ahlbom A. (1998). Exposure to motor vehicle exhaust
and childhood cancer. Scand J Work Environ Health 24:8–11.
Frischer T, Studnicka M, Gartner C, Tauber E, Horak F, Veiter A, Spengler J, Kuhr
J, Urbanek R. (1999). Lung function growth and ambient ozone: a three-year
population study in school children. Am J Respir Crit Care Med 160:390–6.
Gauderman WJ, McConnell R, Gilliland F, London S, Thomas D, Avol E, Vora H,
Berhane K, Rappaport EB, Lurmann F, Margolis HG, Peters J. (2000). Association between air pollution and lung function growth in southern California
children. Am J Respir Crit Care Med 162:1383–90.
Outdoor Air
331
Gergen PJ, Turkeltaub PC. (1992). The association of individual allergen reactivity with respiratory disease in a national sample: data from the second National Health and Nutrition Examination Survey, 1976–80 (NHANES II). J Allergy Clin Immunol 90:579–88.
Gilliland FD, Berhane K, Rappaport EB, Thomas DC, Avol E, Gauderman WJ, London SJ, Margolis HG, McConnell R, Islam KT, Peters JM. (2001). The effects of
ambient air pollution on school absenteeism due to respiratory illnesses. Epidemiology 12:43–54.
Ha EH, Hong YC, Lee BE, Woo BH, Schwartz J, Christiani da VC. (2001). Is air
pollution a risk factor for low birth weight in Seoul? Epidemiology 12:643–8.
Hajat S, Haines A, Goubet SA, Atkinson RW, Anderson HR. (1999). Association
of air pollution with daily GP consultations for asthma and other lower respiratory conditions in London. Thorax 54:597–605.
Halonen M, Stern DA, Wright AL, Taussig LM, Martinez FD. (1997). Alternaria as
a major allergen for asthma in children raised in a desert environment. Am J
Respir Crit Care Med 155:1356–61.
Harrison RM, Leung PL, Somervaille L, Smith R, Gilman E. (1999). Analysis of
incidence of childhood cancer in the West Midlands of the United Kingdom
in relation to proximity to main roads and petrol stations. Occup Environ Med
56:774–80.
Infante-Rivard C. (1993). Childhood asthma and indoor environmental risk factors. Am J Epidemiol 137:834–44.
International Agency for Research on Cancer. (1989). IARC monographs on the
evaluation of carcinogenic risks to humans. Vol. 46. Diesel and gasoline engine exhausts and some nitroarenes. Lyon, France: International Agency for
Research on Cancer.
Kinney PL, Thurston GD, Raizenne M. (1996). The effects of ambient ozone on
lung function in children: a reanalysis of six summer camp studies. Environ
Health Perspect 104:170–4.
Kopp MV, Bohnet W, Frischer T, Ulmer C, Studnicka M, Ihorst G, Gardner C,
Forster J, Urbanek R, Kuehr J. (2000). Effects of ambient ozone on lung function in children over a two-summer period. Eur Respir J 16:893–900.
Krishna MT, Chauhan AJ, Frew AJ, Holgate ST. (1998). Toxicological mechanisms
underlying oxidant pollutant-induced airway injury. Rev Environ Health 13:
59–71.
Kunzli N, Lurmann F, Segal M, Ngo L, Balmes J, Tager IB. (1997). Association between lifetime ambient ozone exposure and pulmonary function in college
freshmen—results of a pilot study. Environ Res 72:8–23.
Linaker CH, Coggon D, Holgate ST, Clough J, Josephs L, Chauhan AJ, Inskip HM.
(2000). Personal exposure to nitrogen dioxide and risk of airflow obstruction
in asthmatic children with upper respiratory infection. Thorax 55:930–3.
Maisonet M, Bush TJ, Correa A, Jaakkola JJ. (2001). Relation between ambient air
pollution and low birth weight in the northeastern United States. Environ
Health Perspect 109(Suppl 3):351–6.
McConnell R, Berhane K, Gilliland F, London SJ, Islam T, Gauderman WJ, Avol
E, Margolis HG, Peters JM. (2002). Asthma in exercising children exposed to
ozone: a cohort study. Lancet 359:386–91.
McConnell R, Berhane K, Gilliland F, London SJ, Vora H, Avol E, Gauderman WJ,
Margolis HG, Lurmann F, Thomas DC, Peters JM. (1999). Air pollution and
bronchitic symptoms in Southern California children with asthma. Environ
Health Perspect 107:757–60.
332
CHILD HEALTH
AND THE
ENVIRONMENT
McDonnell WF, Stewart PW, Smith MV, Pan WK, Pan J. (1999). Ozone-induced
respiratory symptoms: exposure–response models and association with lung
function. Eur Respir J 14:845–53.
McKinney PA, Alexander FE, Cartwright RA, Parker L. (1991). Parental occupations of children with leukaemia in west Cumbria, north Humberside, and
Gateshead. BMJ 302:681–7.
National Institute of Environmental Health Sciences. (2001). Ninth report on carcinogens. Research Triangle Park, NC: National Toxicology Program.
Nemmar A, Hoet PH, Vanquickenborne B, Dinsdale D, Thomeer M, Hoylaerts
MF, Vanbilloen H, Mortelmans L, Nemery B. (2002). Passage of inhaled particles into the blood circulation in humans. Circulation 105:411–4.
Northridge ME, Yankura J, Kinney PL, Santella RM, Shepard P, Riojas Y, Aggarwal M, Strickland P. (1999). Diesel exhaust exposure among adolescents in
Harlem: a community-driven study. Am J Public Health 89:998–1002.
Pandya RJ, Solomon G, Kinner A, Balmes JR. (2002). Diesel exhaust and asthma:
hypotheses and molecular mechanisms of action. Environ Health Perspect 110
(Suppl 1):103–12.
Pereira LA, Loomis D, Conceicao GM, Braga AL, Arcas RM, Kishi HS, Singer JM,
Bohm GM, Saldiva PH. (1998). Association between air pollution and intrauterine mortality in São Paulo, Brazil. Environ Health Perspect 106:325–9.
Pope CA 3rd, Burnett RT, Thun MJ, Calle EE, Krewski D, Ito K, Thurston GD.
(2002). Lung cancer, cardiopulmonary mortality, and long-term exposure to
fine particulate air pollution. JAMA 287:1132–41.
Raaschou-Nielsen O, Hertel O, Thomsen BL, Olsen JH. (2001). Air pollution from
traffic at the residence of children with cancer. Am J Epidemiol 153:433–43.
Raaschou-Nielsen O, Lohse C, Thomsen BL, Skov H, Olsen JH. (1997). Ambient
air levels and the exposure of children to benzene, toluene, and xylenes in
Denmark. Environ Res 75:149–59.
Ritz B, Yu F. (1999). The effect of ambient carbon monoxide on low birth weight
among children born in southern California between 1989 and 1993. Environ
Health Perspect 107:17–25.
Ritz B, Yu F, Chapa G, Fruin S. (2000). Effect of air pollution on preterm birth
among children born in Southern California between 1989 and 1993. Epidemiology 11:502–11.
Ritz B, Yu F, Fruin S, Chapa G, Shaw GM, Harris JA. (2002). Ambient air pollution and risk of birth defects in Southern California. Am J Epidemiol 155:17–25.
Sarnat JA, Schwartz J, Catalano PJ, Suh HH. (2001). Gaseous pollutants in particulate matter epidemiology: confounders or surrogates? Environ Health Perspect 109:1053–61.
Savitz DZ, Feingold L. (1989). Association of childhood cancer with residential
traffic density. Scand J Work Environ Health 15:360–3.
Schwartz J. (1989). Lung function and chronic exposure to air pollution: a crosssectional analysis of NHANES II. Environ Res 50:309–21.
Shu XO, Gao YT, Brinton LA, Linet MS, Tu JT, Zheng W, Fraumeni JF. (1988). A
population-based case-control study of childhood leukemia in Shanghai. Cancer 62:635–44.
Spengler JD, Koutrakis P, Dockery DW, Raizenne M, Speizer FE. (1996). Health
effects of acid aerosols on North American children: air pollution exposures.
Environ Health Perspect 104:492–9.
Outdoor Air
333
Studnicka M, Hackl E, Pischinger J, Fangmeyer C, Haschke N, Kuhr J, Urbanek
R, Neumann M, Frischer T. (1997). Traffic-related NO2 and the prevalence of
asthma and respiratory symptoms in seven year olds. Eur Respir J 10:2275–8.
Suh HH, Bahadori T, Vallarino J, Spengler JD. (2000). Criteria air pollutants and
toxic air pollutants. Environ Health Perspect 108(Suppl 4):625–33.
U.S. Environmental Protection Agency. (1999). Air quality index reporting; final
rule (40 CFR Part 58). Fed Reg 64:42530–49.
U.S. Environmental Protection Agency. (2000a). Air quality index. A guide to air
quality and your health (EPA-454/R-00-005). Research Triangle Park, NC: U.S.
Environmental Protection Agency.
U.S. Environmental Protection Agency. (2000b). Health assessment document for
diesel exhaust. Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2001b). Latest findings on national air
quality: 2000 status and trends. Research Triangle Park, NC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2000c). National air pollutant emission
trends, 1900–1998. Research Triangle Park, NC: U.S. Environmental Protection
Agency.
U.S. Environmental Protection Agency. (2000d). National air quality and emissions trends report, (1998). Research Triangle Park, NC: U.S. Environmental
Protection Agency.
U.S. Environmental Protection Agency. (2000e). National air toxics program: the
integrated urban strategy report to Congress. Research Triangle Park, NC: U.S.
Environmental Protection Agency.
U.S. Environmental Protection Agency. (2001a). Air quality criteria for particulate
matter (third external review draft). Located at http://cfpub.epa.gov/ncea/
cfm/partmatt.cfm.
U.S. Environmental Protection Agency. (2001c). National ambient air quality standards (NAAQS). Located at http://www.epa.gov/airs/criteria.html.
Wallace L. (1996). Environmental exposure to benzene: an update. Environ Health
Perspect 104(Suppl 6):1129–36.
Wang X, Ding H, Ryan L, Xu X. (1997). Association between air pollution and low
birth weight: a community-based study. Environ Health Perspect 105:514–20.
Weaver VM, Davoli CT, Heller PJ, Fitzwilliam A, Peters HL, Sunyer J, Murphy
SE, Goldstein GW, Groopman JD. (1996). Benzene exposure, assessed by urinary trans,trans-muconic acid, in urban children with elevated blood lead levels. Environ Health Perspect 104:318–23.
Woodruff TJ, Axelrad DA, Caldwell J, Morello-Frosch R, Rosenbaum A. (1998).
Public health implications of 1990 air toxics concentrations across the United
States. Environ Health Perspect 106:245–51.
Woodruff TJ, Grillo J, Schoendorf KC. (1997). The relationship between selected
causes of postneonatal infant mortality and particulate air pollution in the
United States. Environ Health Perspect 105:608–12.
World Health Organization. (2000). Guidelines for air quality. Geneva: World
Health Organization.
12
Water
I. CHEMICAL CONTAMINANTS
The introduction of chlorinated drinking water during the early twentieth century virtually eliminated cholera, typhoid fever, and other waterborne infectious diseases in the regions served. Chlorine, a potent oxidizing agent, not only kills microbial agents but also reacts with natural
organic material in raw water to produce hundreds of disinfection byproducts (DBPs). These DBPs include trihalomethanes (THMs), haloacetic
acids, haloacetonitriles, haloketones, halophenols, halogenated furanones,
and other halogenated hydrocarbons. There is limited epidemiologic evidence of associations between DBPs and adverse reproductive outcomes
and certain cancers. Drinking water is also subject to contamination by
natural sources (e.g., nitrate from agricultural operations), waste disposal
(including leachates from hazardous waste disposal sites), and leakage
from solvent storage containers. Trichloroethylene, for instance, is the
most frequently reported organic contaminant in groundwater in the
United States and is a probable human carcinogen; it has been used extensively as a degreaser in metal and automotive industries since the 1920s
and was widely used in dry cleaning for several decades.
Part I of this chapter considers chemical contaminants from disinfection and hazardous waste disposal, and Part II focuses on certain child-
334
Water
335
hood infectious diseases for which waterborne transmission is an important route of infection. The objective of Part I is to describe potential threats
to child health from widespread drinking water chemical contaminants,
with a major focus on risks of adverse developmental effects and cancer.
The discussion includes recent developments in the epidemiology and
risk management of DBPs. See other chapters for contaminants that may
occur in water including lead and arsenic (Chapters 4 and 5, Metals), pesticides (Chapter 7, Pesticides), and radionuclides (Chapter 9, Radiation).
Health Effects
Genotoxicity
Formation of THMs (chloroform, dichlorobromomethane, dibromochloromethane, and bromoform) during chlorination of water was reported in
1974. Although only a fraction of chlorinated DBPs, THM concentrations
have been widely used by regulators and researchers as a proxy for total
DBPs. Organic extracts of chlorinated drinking water were later shown
to be mutagenic in the Ames Salmonella assay, while extracts of raw water
had little or no activity; most of the mutagenic activity in chlorinated
water was associated with nonvolatile DBPs. The DBP 3-chloro-4(dichloromethyl)-5-hydroxy-2[5H]-furanone (MX) and related chlorohydroxyfurnanones are potent mutagens; MX alone comprised about
30%–60% of the total mutagenic activity of chlorinated drinking water in
Finland and Massachusetts (Wright et al., 2002).
Ozonation of raw water has been shown to produce considerably
lower mutagenic activity than chlorination, and high ozone doses virtually eliminate mutagenicity. Other findings from studies of DBP mutagenicity include the following: (1) ingested chloroform and bromoform
both cause chromosome abnormalities, and inhaled chloroform causes increased micronuclei in rats, (2) MX and three related chlorohydroxyfurnanones at low doses produce DNA strand breaks and complex chromatid
rearrangements in human and rodent cells in vitro, (3) chlorinated and
brominated acetonitriles produce DNA strand breaks in human cells in
vitro, and (4) trichloroacetic acid, dichloroacetic acid, and chloral hydrate
are all weakly mutagenic in mouse lymphoma cells in vitro.
Under the Comprehensive Environmental Response, Compensation,
and Liability Act (CERCLA), the Agency for Toxic Substances and Disease Registry (ATSDR) and the EPA are responsible for identifying contaminants and groups of contaminants at hazardous waste sites that pose
the greatest public health risks in the United States. The ATSDR is re-
CHILD HEALTH
336
AND THE
ENVIRONMENT
quired to periodically prepare a list (known as the CERCLA list) of such
substances in priority order based on frequency, known toxicity, and potential for human exposure. Most of the highest-ranked chemicals on the
2001 CERCLA list are genotoxic and probable or known human carcinogens (Table 12–1).
TABLE 12–1. Highest-Priority Contaminants at Hazardous Waste Sites in the
United States, in Rank Order by Risk to Public Healtha
Genotoxicityb
Contaminant
Carcinogenicity
in humansc
Arsenic (inorganic)
(highest ranked)
Clastogenic in vitro and in vivo;
mutagenic in vitro
Known
Lead (inorganic)
Clastogenic in vitro and in vivo
Possible
Mercury—inorganic
Clastogenic in vitro and in vivo; mixed
evidence of point mutations in vitro;
inhibits mitotic spindle formation
Not classifiable
Mercury-elemental
Possibly clastogenic in occupationally
exposed persons
Not classifiable
Vinyl chloride
Metabolites are mutagenic and clastogenic
in vitro and in vivo
Known
PCBs
Most studies showed no mutagenicity;
induces Phase I cytochrome enzymes
that can activate procarcinogens
and generate reactive oxygen species
Probable
Benzene
Metabolites appear to be clastogenic
Known
Cadmium
Mixed evidence of mutagenicity and
clastogenicity; interferes with mitotic
spindle in vitro and in vivo
Known
Benzo[a]pyrene,
Mutagenic and clastogenic in vitro
benzo[a]fluoranthene,
other PAHs
Probable
Chloroform
Not strongly mutagenic
Possible
,-DDT
Mixed evidence of mutagenicity and
clastogenicity
Possible
Trichloroethylene
EPA-not available; mixed evidence of
mutagenicity in vitro noted in IARC
review
Probable
Dieldrin
Clastogenic in vitro; mixed evidence of
mutagenicity
Not classifiable
Chromium
(hexavalent)
Mutagenic and clastogenic; causes oxidative
DNA and chromosomal damage
Known
a Agency
b U.S.
for Toxic Substances and Disease Registry (2001).
Environmental Protection Agency (2001b).
c International
Agency for Research on Cancer (2002).
Water
337
Developmental Outcomes
Given the widespread exposure to DBPs and the evidence of reproductive toxicity in experimental animals, there have been surprisingly few
epidemiologic studies incorporating strong statistical power and exposure assessment. Until recently, evidence of links between DBPs and adverse reproductive outcomes in humans came from a few epidemiologic
studies involving diverse exposure assessments and health outcomes, precluding a meaningful synthesis (Reif et al., 1996). Subsequent epidemiologic studies have strengthened the case that DBPs may be reproductive
toxicants in humans; recent reviews concluded that
• The weight of evidence from toxicologic and epidemiologic studies of
DBPs was suggestive of associations between DBP exposure and IUGR
and urinary tract birth defects, but the limited exposure data precluded
definitive conclusions (Graves et al., 2001).
• Epidemiologic studies provide moderate evidence of associations between THMs and IUGR, neural tube defects, and spontaneous abortions; evidence of a role for other VOCs in drinking water is weaker
because there have been fewer studies (Bove et al., 2002).
Fetal Deaths
Epidemiologic studies have yielded limited evidence of an association between spontaneous abortion and exposure to THMs during early pregnancy (Savitz et al., 1995; Waller et al., 1998). Important findings included
a moderately strong association with a specific THM, bromodichloromethane (BDCM), independent of several potential confounders (Waller
et al., 1998). The latter finding is consistent with evidence that BDCM
causes fetal death in rodents. Drinking water may contain non-DBP VOCs
such as tetrachloroethylene and trichloroethylene; there is fairly consistent evidence of associations between spontaneous abortion and occupational exposure to tetrachloroethylene, trichloroethylene, toluene, xylene,
and chloroform.
There is also limited evidence of an association between stillbirths
and THMs, particularly BDCM levels (Fig. 12–1). The subgroup of stillbirths caused by asphyxia was strongly associated with total THMs (OR
4.6, CI 1.9–11) in what appears to be the first epidemiologic study to assess this category of stillbirths in relation to environmental hazards (King
et al., 2000). Asphyxia-related stillbirths include those related to placental vascular abnormalities (e.g., placental abruption). Trichloroacetic acid
and other DBPs deplete folic acid in experimental animals (possibly
through a free radical mechanism that induces vitamin B12 and folate deficiencies); folate deficiency increases the plasma homocysteine level, a
CHILD HEALTH
odds ratio and 95% confidence interval
338
AND THE
ENVIRONMENT
100
10
1
0.1
Boston
NS, all
SBs,
THMs
NS, all
SBs,
BDCM
NS,
NS,
asphyxia, asphyxia,
THMs
BDCM
FIGURE 12–1. Associations between DBPs and stillbirths (SBs). Boston (Aschengrau et al., 1993), Nova Scotia (NS) (Dodds et al., 1999; King et al., 2000).
risk factor for neural tube defects and placental abruption (Dow and
Green, 2000).
Birth Defects
The main findings from epidemiologic studies of DBPs and birth defects
include
• Neural tube defects—fairly consistent associations with DBPs, the highest risks being those among women in the highest BDCM category and
women in the highest tertile of THM levels who did not take multivitamin or folate supplements during the 3 months before pregnancy
(Fig. 12–2)
• Urinary tract defects—limited evidence of an association with a chlorinated surface water source (Magnus et al., 1999)
• Oral clefts—inconsistent associations with DBPs
• Cardiac defects—generally no association with DBPs
There is inadequate evidence to assess the role of drinking water nitrate
in neural tube defects; epidemiologic studies have shown associations
with nitrate in well water but not in municipal water or diet (Arbuckle
et al., 1988; Croen et al., 2001; Dorsch et al., 1984). Given that dietary nitrate intake is usually much greater than that from water, it is possible
that nitrate is a marker for another toxicant(s) in water (especially groundwater). Animal studies have shown reproductive toxicity but not birth defects at high exposure levels to nitrate or nitrite.
Water
339
odds ratio and 95% confidence interval
10
1
0.1
NJ
Norway
NJ
NJ vits
NJ no
vits
Swe
ClO2
Swe
NaClO
NS
CHCl3
NS
BDCM
FIGURE 12–2. Associations between DBPs in drinking water and the risk of neural tube birth defects (Sweden restricted to spina bifida). Nova Scota (Dodds and
King, 2001), Swe (Sweden) (Kallen and Robert, 2000), NJ (New Jersey: no vits,
took vits, all subjects) (Klotz and Pyrch, 1999), Norway (Magnus et al., 1999), New
Jersey (Bove et al., 1995).
Exposure to non-THM VOCs from contaminated wells or proximity
to hazardous waste disposal sites has been linked to cardiac and neural
tube defects (Croen et al., 1997; Dolk et al., 1998; Goldberg et al., 1990).
Recent reviews of epidemiologic studies of pregnancy outcome and water
contaminants or proximity to hazardous waste disposal sites concluded
that (Bove et al., 2002; Vrijheid, 2000).
• There was limited and inconsistent evidence of associations between
trichloroethylene or tetrachloroethylene in drinking water and cardiac,
neural tube, and cleft defects.
• Some studies of individual and multiple hazardous waste disposal sites
have shown associations with adverse health effects including low birth
weight and birth defects; the potential for exposure to hazardous wastes
through contaminated groundwater is real, but epidemiologic studies
have generally lacked direct exposure measures, adequate statistical
power, and control of potential confounders.
In summary, there is insufficient evidence to infer or reject a causal relationship between specific drinking water contaminants and birth defects.
CHILD HEALTH
340
AND THE
ENVIRONMENT
In general, however, there is suggestive evidence of possible associations
between neural tube and urinary tract defects and THMs. Importantly, an
association between neural tube defects and THMs was restricted to
women who did not take daily multivitamins or folate supplements during the 3 months before pregnancy (Klotz and Pyrch, 1999). Although
based on small numbers, this finding is consistent with evidence that
DBPs create folate deficiency, thereby increasing the plasma homocysteine
level, a risk factor for neural tube defects (Dow and Green, 2000; Eskes,
2000).
Low Birth Weight
The most consistent findings in epidemiologic studies of DBPs and birth
weight and gestation length were associations between THMs and IUGR
(Fig. 12–3). Findings include evidence of exposure-risk relationships between IUGR and chloroform and BDCM (Kramer et al., 1992) and an association between third trimester THM exposure and IUGR (Gallagher et
al., 1998). These findings are consistent with the observation of reduced
birth weight among offspring of experimental animals that ingested chloroform at high doses during pregnancy (Thompson et al., 1974). Exposure to tetrachloroethylene-contaminated water was associated with re-
odds ratio and 95% confidence interval
100
10
1
0.1
Iowa
Iowa
CHCl3 BCDM
NJ
Italy
ClO2
Italy
NaClO
Den
NS
Tai
Swe
ClO2
Swe
NaClO
Nor
FIGURE 12–3. Associations between DBPs and the risk of IUGR (or small head circumference—Italy and Sweden); Nor (Norway) (Jaakkola et al., 2001), Sweden
(Kallen and Robert, 2000), Tai (Taiwan) (Yang et al., 2000), Nova Scotia (Dodds et
al., 1999), Den (Denver) (Gallagher et al., 1998), Italy (Kanitz et al., 1996), New
Jersey (Bove et al., 1995), Iowa (Kramer et al., 1992).
Water
341
duced birth weight and a twofold increased risk of IUGR among mothers aged 35 years or older but not among younger women (Sonnenfeld
et al., 2001). There were significant exposure–risk relations between
groundwater nitrate levels and IUGR and preterm birth in an agricultural
region dependent on groundwater potentially contaminated by agricultural activities (Bukowski et al., 2001).
Cancer
Evidence from the few epidemiologic studies of childhood cancer and
drinking water is inadequate to assess the potential etiologic roles of DBPs
and other contaminants. There were weak or no associations between
childhood cancer or childhood leukemia and DBP exposure in two studies (Infante-Rivard et al., 2001; Kallen and Robert, 2000). Further analysis
of the leukemia study showed strong but imprecise associations with
total THM exposure among children with polymorphisms involving a
CYP2E1 variant with increased transcriptional activity or a GSTT1 null
genotype, that is, variants that may increase generation or reduce detoxification of reactive THM metabolites (Infante-Rivard et al., 2002). Childhood leukemia was linked to drinking water containing trichloroethylene
in women but not men in an ecologic study (Cohn et al., 1994). The THMs
cause liver, kidney, and large intestinal tumors in rodents; BDCM and bromoform each caused substantially increased risks of large intestinal tumors. Rats of both sexes exposed to MX in drinking water had doserelated increases in thyroid and liver bile duct tumors, including increases
at the lowest dose tested; the highest MX doses caused adrenal, lung,
breast, and pancreatic tumors, lymphomas, and leukemias.
Investigation of a cluster of childhood leukemia in Woburn, Massachusetts, showed increased risk in an area served by water that may have
been contaminated by trichloroethylene and other toxicants from a hazardous waste disposal site (Cutler et al., 1986); this incident led to legal
actions popularized in the book and the motion picture A Civil Action.
Childhood brain tumors were linked to maternal well water consumption in western Washington state (Mueller et al., 2001). Until studies with
adequate statistical power and exposure assessment are conducted, assessment of the role of drinking water contaminants in childhood cancer
will be problematic.
Epidemiologic studies have shown fairly consistent associations between THMs and adult bladder cancer and less consistent links to other
adult cancers. Evidence from studies of persons occupationally exposed
to trichloroethylene indicates increased risks of kidney cancer, liver cancer, non-Hodgkin’s lymphoma, Hodgkin’s disease, multiple myeloma,
342
CHILD HEALTH
AND THE
ENVIRONMENT
and cervical cancer (Wartenberg et al., 2000). Although most subjects had
exposures to other solvents, it appears that trichloroethylene likely contributes to human cancers caused by solvent exposures. Biologic plausibility comes from evidence that trichloroethylene causes dose-related increased risks of liver and kidney cancers in experimental animals, with
no evidence of a threshold within the observable dose range; the
trichloroethylene metabolites trichloroacetic acid and dichloroacetic acid
are also rodent carcinogens (liver cancer).
Other Health Effects
Infantile methemoglobinemia has been attributed to high water nitrate
levels, particularly from farm wells. Compared to older persons, infants
under age 6 months have lower levels of NADH-cytochrome b5 reductase, an enzyme that converts methemoglobin to hemoglobin. Recent
evidence suggests that infantile methemoglobinemia may be caused by
gastrointestinal infection; NO is produced in gastrointestinal and other
tissues in response to infection and inflammation and can oxidize hemoglobin to methemoglobin; thus water nitrate levels may be a proxy for
waterborne enteric pathogens. There is inadequate evidence of an association between nitrate/nitrite in drinking water and the risk of childhood type I diabetes from ecologic studies; the only analytical study
showed associations with prenatal maternal dietary nitrite intake and
postnatal child dietary nitrite intake but not with drinking water nitrate
or nitrite.
Exposures
Chlorination Disinfection By-Products
Measurement of DBP exposure is complicated by variation in DBP concentrations in relation to
• Place—DBP formation increases with time after chlorination, and levels vary among sites including water treatment facilities, distribution
systems, and homes.
• Time—DBP levels are generally higher during warmer months because
of increased precursor levels in raw water (natural organic carbon from
plant and other sources) and increased DBP formation rates.
• Raw water characteristics—DBP formation after chlorination varies by
total organic carbon content, pH, ammonium and bromide ion concentrations, and temperature.
Water
343
• Water treatment practices—filtration (if done before chlorination, especially for surface water, this substantially reduces the concentration
of total organic carbon available to react with chlorine), choice of disinfection agent (e.g., chloramination produces lower DBP levels compared to chlorine or hypochlorite), and use of carbon filters to remove
DBPs (either at the treatment facility or the point of use).
Epidemiologic studies of DBPs have usually collected self-reported information on residential history and water consumption habits but not
on bathing/showering habits. The extent and quality of current and historic water quality data vary by jurisdiction, as do the frequency and number of specific DBP measurements. For monitoring DBPs in drinking
water, the EPA requires individual and total THMs, dichloroacetic acid,
trichloroacetic acid, total haloacetic acids, and bromate (U.S. Environmental Protection Agency, 1998). The infrequency of DBP measurements
is a problem for epidemiologic studies of developmental outcomes; for
example, the critical exposure period for birth defects is only a few weeks
early in the first trimester. The limited ability of epidemiologic studies to
accurately assess exposure to DBPs has likely caused some of the inconsistencies among studies (Arbuckle et al., 2002).
Exposure studies of human volunteers have shown that THM concentrations in water and exhaled alveolar breath or blood, collected after
swimming in chlorinated pools or after showering, are strongly correlated. A 1-hour swim in water with a chloroform level of 160 g/L produced an average chloroform uptake of about 65 g/kg body weight (75%
dermal, 25% inhalation); this is about thirtyfold higher than the dose from
ingesting 2 L of water with a THM level of 80 g/L (the current EPA maximum contaminant level [MCL]). Blood bromoform levels spiked sharply
after showering in a region with high brominated THM levels (Lynberg
et al., 2001). Such spikes could be etiologically significant if they occur
during critical exposure periods for adverse developmental outcomes. Although THM levels in exhaled breath and blood and urinary haloacetic
acid levels are promising biomarkers of recent exposure, technical and
cost problems must be overcome before such biomarkers are widely used
in epidemiologic studies (Arbuckle et al., 2002).
Most monitoring of DBP levels in chlorinated drinking water has
been limited to total THMs, a fraction that contains only a small fraction
of the mutagenic activity of chlorinated water. Only very limited or no
data on MX or total mutagenicity levels in chlorinated drinking water are
available in most countries. A tap water survey in Massachusetts surface
water systems showed that the average MX concentration was 27.5 ng/L
(range, 5–88 ng/L) and that MX comprised about half of the mutagenicity (Wright et al., 2002).
CHILD HEALTH
344
AND THE
ENVIRONMENT
Other Toxics
The most frequently detected non-THM VOCs in Canadian and American drinking water supplies have been trichloroethylene, tetrachloroethylene, 1,2-dichloroethane, dichloromethane, benzene, toluene, ethylbenzene, xylenes, vinyl chloride, and 1,1,1-trichloroethane. For all regulated
VOCs, 6% of groundwater and 8% of surface water supplies had one or
more VOCs at levels exceeding the MCL (Table 12–2). Groundwater systems serving populations of over 10,000 experience relatively frequent
VOC exceedances because of the shallow, unconfined groundwater supplies needed to sustain large flow rates but vulnerable to contamination.
Solvents persist in groundwater because of relatively low biologic and
chemical reactivity, low temperatures, absence of light, lack of contact
with the atmosphere, and relatively low microbial concentrations. Rural
groundwater supplies are vulnerable to nitrate contamination from animal and human feces, other organic waste, and chemical fertilizers; high
drinking water nitrate levels are associated with shallow or dug wells and
large farms.
Three to 4 million children in the United States live within 1.6 km of
a major hazardous waste disposal site and are at risk of exposure to VOCs
and other contaminants. Exposure pathways of VOCs for children include
ingestion of contaminated soil and drinking water and inhalation of VOCs
(e.g., volatilization during showers or from groundwater into basements).
Contaminated groundwater was an exposure pathway at 91% of the 1300
sites on the National Priority List targeted for remediation because of public health threats and for several of the 20 priority contaminants on the
TABLE 12–2. Relative Frequency of Regulated VOC Occurrences in Exceedance of
MCLs in Surface Water and Ground Water Supplies, United States
VOC
% of Surface
Water Supplies
% of GroundWater Supplies
4.7
2.3
Methylene chloride
Tetrachloroethylene
1.7
1.8
Trichloroethylene
1.2
1.5
Benzene
0.3
0.4
Vinyl chloride
0.3
0.2
8.2
6.1
All regulated
VOCsa
Source: U.S. Environmental Protection Agency (1999).
Note: limited to those solvents for which at least 1% of water systems had MCL exceedances.
a There
are 21 regulated VOCs including those shown above.
Water
345
2001 CERCLA list, including arsenic, vinyl chloride, benzene, cadmium,
chloroform, trichloroethylene, chromium VI, and hexachlorobutadiene
(Table 12–1). The average drinking water trichloroethylene level in the
United States during 1998 was 3.0 g/L, with average daily intakes of
about 11–33 g by inhalation and 2–20 g by ingestion; higher exposures
may occur in homes with private wells located near waste disposal sites.
Volatile organic chemicals are readily measured in exhaled breath and
generally reflect recent exposures; exhaled breath tetrachloroethylene reflects exposures over somewhat longer periods because its half-life in vivo
is much longer than that of most VOCs.
Risk Management
The provision of safe drinking water to a large population is a massive
enterprise. There are about 170,000 public water systems in the United
States alone, among which about 54,000 community water systems serve
the majority of the population (U.S. Environmental Protection Agency,
2001c). The EPA regulates public water systems (publicly or privately
owned water systems that serve at least 25 people or 15 service connections for at least 60 days annually) and has promulgated drinking water
standards that address several hazard categories including microorganisms, disinfectant chemicals and DBPs, inorganic chemicals, organic
chemicals, and radionuclides. This chapter addresses the first four categories to some degree; the reader should consult Chapter 9 for radionuclides, Chapter 7 for pesticides, and Chapters 4 and 5 for metals.
Disinfection By-Products
Trihalomethanes and other DBPs in drinking water originate mainly from
the reaction of chlorine with humic and fulvic acids, natural organic substances arising from the decomposition of lignins and other phenolic compounds in vegetation. Levels of natural organics are generally higher in
surface waters because of runoff from land, and thus chlorinated drinking
surface water usually has higher average DBP levels than water from
ground sources (Fig. 12–4). Most exposure to THMs comes from water and
beverages. Trihalomethane levels in hot water are higher than those in cold
water, tending to increase exposures through inhalation and
bathing/showering. Chlorinated swimming pools are an important source
of exposure, especially for children who swim frequently. Brominated
DBPs tend to occur at higher levels in water from coastal areas and in sys-
CHILD HEALTH
346
AND THE
ENVIRONMENT
Trihalomethanes
80
Percent
60
40
20
0
0-19
20-39
40-59
THM G
60-79
80+
THM S
Haloacetic acids
100
Percent
80
60
40
20
0
0-19
20-39
40-59
Halo G
60-79
80+
Halo S
FIGURE 12–4. Distribution of total THM and haloacetic acid levels ( g/L) in U.S.
ground and surface water supplies, 1998 (U.S. Environmental Protection Agency,
2000). G ground water, S surface water, Halo haloacetic acids.
tems that use ozone. Reduction of DBP levels in drinking water, while preserving adequate disinfection, can be achieved through means such as
• Using raw water supplies low in natural organic material; this generally means groundwater or surface waters not prone to runoff with
high organic content
Water
347
• Protecting surface waters from contamination with natural organic material, for example, controlled disposal of animal manure and agricultural fertilizer runoff
• Filtration of raw water before disinfection, thereby reducing both the
disinfectant dose needed and the microbial risks
• Using disinfection methods that reduce DBP formation (e.g., chloramination, ozonation, UV light)
• Removing DBPs from disinfected water with activated carbon filters
(at the source or the point of use).
The main problem with many of these options relates to cost, particularly
in older water systems that do not include filtration and serve relatively
small populations. Although chlorination remains the dominant method
in North America, ozonation is commonly used in several European countries; in such systems, some chlorinating agent is generally added to water
before it enters the distribution system to maintain disinfection. Also,
ozonation of water containing bromide increases production of bromate,
a known animal carcinogen (kidney and thyroid tumors), and brominated
DBPs. Granulated activated charcoal filters remove almost all THMs,
other DBPs, and VOCs such as trichloroethylene; the filters must be replaced frequently, as their efficiency drops sharply when they become saturated. Drinking water standards for DBPs and other chemical contaminants are presented in Table 12–3. There is no drinking water standard
for MX at present, but the proposed virtual safe dose for MX genotoxicity is 5 ng/kg/day and the proposed tolerable daily intake for nongenotoxic effects is 40 ng/kg/day (Hirose et al., 1999).
Other Chemical Contaminants
Approximately 85% of all U.S. drinking water systems have at least one
potential source of contamination within 2 miles of their water intake or
well, but only about 10% of water suppliers have implemented protective measures ranging from land use ordinances to public education. Hazardous waste disposal and leakage from storage tanks are major causes
of groundwater contamination and potential threats to the health of persons dependent on well water; for example, disposal of trichloroethylene
and leakage from storage tanks explain its frequent occurrence in groundwater samples. The ATSDR was established in 1980 under CERCLA (also
known as the Superfund Act) to identify and remediate the estimated
40,000 uncontrolled hazardous waste sites in the United States. Trichoro-
CHILD HEALTH
348
AND THE
ENVIRONMENT
TABLE 12–3. Drinking Water Guidelines and Standards for Microorganisms and
Chemical Contaminants
Contaminant
EPA
Standard
TTa
5.0%b
Disinfectants and DBPs
Bromate
Chloramine (as chlorine)
(maximum residual
disinfectant level) (MRDL)
Chlorine (MRDL)
Chlorine dioxide (MRDL)
Chlorite
Haloacetic acids
Total THMs
EPA
Standard
Selected other organic chemicalsc
Microorganisms
Cryptosporidium, Giardia
lamblia, heterotrophic plate
count, Legionella, turbidity,
enteric viruses
Total coliforms
Contaminant
10 g/L
4 mg/L
4 mg/L
0.8 mg/L
1.0 mg/L
60 g/L
80 g/L
Benzene
1,2-Dichloroethane
Dichloromethane
Ethylbenzene
Tetrachloroethylene
Toluene
1,1,1-Trichloroethane
Trichloroethylene
Vinyl chloride
Xylenes (total)
5 g/L
5 g/L
5 g/L
0.7 mg/L
5 g/L
1 mg/L
0.2 mg/L
5 g/L
2 g/L
10 mg/L
Inorganic chemicals
Nitrate
Nitrite
10 mg/L
1 mg/L
Source: U.S. Environmental Protection Agency (2001a).
a TT
treatment technique, that is, a required process intended to reduce the level of a contaminant in
drinking water; for further information on application of this standard see U.S. Environmental Protection Agency (2001a).
b No
more than 5% of samples can be total coliform-positive in a month (for systems that collect fewer
than 40 routine samples per month, no more than 1 sample can be total coliform-positive); every sample that has total coliforms must be analyzed for fecal coliforms, and no fecal coliforms or E. coli may
be present.
c Frequently
detected non-THM VOCs in Canadian and U.S. water supplies.
See other chapters for standards for metals, pesticides, and radionuclides in drinking water.
ethylene, lead, tetrachloroethylene, benzene, and chromium were the
most frequently detected substances in water near the 1300 National Priorities List sites.
To protect source waters from chemical contamination, the EPA has
planned a strategy to (1) ensure strong and useful source water assessments for all public water supplies, (2) target relevant state and federal
programs to address source water contamination prevention priorities,
(3) increase awareness, education, and involvement by private industry,
government, and the public, (4) foster local control and capacity, and
(5) document and report on risks identified and progress made on reducing those risks. Even if this strategy is implemented, there will be ma-
Water
349
jor challenges including the need to protect persons dependent on private
wells and small water systems (U.S. Environmental Protection Agency,
2001d).
Conclusions
Unresolved Issues and Knowledge Gaps
• Disinfection by-products
° Developmental effects—there is limited evidence that THMs (or related DBPs) may cause spontaneous abortions, stillbirths, IUGR, and
certain birth defects (neural tube, urinary tract); spontaneous abortions and stillbirths may be more closely related to bromodichloromethane than total THMs.
° Cancer
Although THMs were associated with childhood leukemia among
children with certain polymorphisms, there is inadequate evidence
for the role of DBPs in childhood cancer.
There is limited evidence that DBPs may cause certain adult cancers
(especially bladder cancer).
• Non-THM VOCs and nitrate/nitrite
° Developmental effects
There is inadequate evidence to assess the role of drinking water
nitrate/nitrite in neural tube defects.
There is inadequate evidence to assess the role of trichloroethylene
and related VOCs in birth defects (cardiac, neural tube) and childhood leukemia.
° Drinking water (especially farm well water) containing high nitrate
levels can cause infantile methemoglobinemia; it is not clear if the
causal agent is nitrate per se or waterborne infectious agents.
Risk Management
• Prevention
° Protection of source waters, prefiltration of raw water, and use of activated charcoal (at the source or the point of use) may be necessary
to simultaneously achieve adequate disinfection and minimal DBP
levels in chlorinated surface tap water.
° Control of hazardous waste disposal is needed to prevent contamination of groundwater by VOCs and other hazardous chemicals.
350
CHILD HEALTH
AND THE
ENVIRONMENT
• Monitoring—periodic measurement of DBP and other chemical contaminant levels is needed to evaluate progress in meeting existing
drinking water standards and to identify unmet needs.
II. WATERBORNE INFECTIONS
An adequate supply of safe drinking water is an elusive goal in many
parts of the world, particularly among economically disadvantaged
groups. Each year, about 11 million children die worldwide, primarily
from pneumonia, diarrhea, measles, malaria, and malnutrition. Diarrhea
causes about 2 million childhood deaths annually in developing countries, mostly because of fecally contaminated water used for drinking and
washing, and is an important cause of malnutrition (World Health Organization, 1998). Major causes of childhood diarrhea in developing countries include rotavirus, Shigella dysenteriae, Campylobacter jejuni, Cryptosporidium parvum, Vibrio cholerae, and various Salmonella serotypes.
Children in developing countries also have high illness and death rates
because of water-related parasitic infections, such as, Schistosoma infection from bathing or wading in contaminated water (Pruss et al., 2002).
The continuing threat of waterborne infections to public health even in
economically privileged countries was dramatically illustrated by the
massive 1993 outbreak of C. parvum in Milwaukee that affected over
400,000 people. Unprotected water sources and absent or inadequate
drinking water disinfection processes contribute to continuing outbreaks
of Giardia lamblia and Shigella sonnei and sporadic cases of E. coli 0157:H7
and other agents.
Waterborne pathogens originating in human or animal feces primarily affect the gastrointestinal system, with outcomes that vary from unapparent infections to life-threatening bloody diarrhea, dehydration, and
liver or kidney failure. Toddlers and young children are exposed to
enteric pathogens by ingesting contaminated water and through hand–
mouth behavior while playing with other children or in fecally contaminated areas; older children have lower risks because of acquired immunity and less hand–mouth behavior. Foods exposed to contaminated water
can also cause gastrointestinal infections; for example, formula-fed infants
are at much greater risk than breast-fed infants in regions with microbiologically contaminated drinking water. This part of the chapter briefly
focuses mainly on the major microbial threats to children in developed
countries; see other sources for further information (Detels et al., 1997).
Water
351
Bacteria
Globally, about 50% of the fatal cases of childhood diarrhea episodes annually are caused by enteric pathogens including Shigella, V. cholerae,
Campylobacter, Salmonella, and E. coli (Centers for Disease Control and Prevention, 2001a) (Table 12–4).
Shigella Species
Shigella, the most important global cause of epidemic dysentery (diarrhea
containing blood), is spread by contaminated water and food and personto-person contact. S. sonnei is the most common serotype in developed
countries while S. boydii, S. dysenteriae, and S. flexneri cause most cases in
developing countries. At-risk populations include low-income groups dependent on water supplies that are inadequately disinfected, such as, indigenous groups subject to household crowding, lack of piped water, and
inadequate sewage disposal. About 14,000 confirmed cases of shigellosis
are reported annually in the United States, and the true number may be
twentyfold higher (Centers for Disease Control and Prevention 2000).
Young children are at greatest risk, especially those in child-care centers.
After infection with S. flexneri, persons with the genetic marker HLA-B27
may develop Reiter’s syndrome, characterized by joint pain, eye irritation, painful urination, and, occasionally, chronic arthritis.
Escherichia coli
Enterotoxigenic E. coli is a common cause of diarrhea among travelers and
young children in developing countries. The first evidence that E. coli
could cause serious illness in otherwise healthy people came from an investigation in Oregon and Michigan showing a strong association between
consumption of undercooked hamburgers from the same fast-food restaurant chain and bloody diarrhea caused by E. coli O157:H7. Escherichia
coli O157:H7 produces Shiga-like toxins that damage renal tubular epithelial and endothelial cells; it usually causes mild diarrhea but can cause
life-threatening hemorrhagic colitis and hemolytic uremic syndrome and
is an important cause of acute renal failure in children. Animals, especially cattle, are the main reservoirs of E. coli O157:H7, and the principal
source of human infection is undercooked contaminated meat; other
sources include unpasteurized milk and contaminated water. Outbreaks
of bloody diarrhea caused by E. coli O157:H7 were first linked to contaminated water supplies in Missouri and Japan; the latter incident in-
TABLE 12–4. Potential Waterborne Infection Threats for Childrena
Infectious Agent
Occurrence
Reported Annual
Cases, USA, 1999b
All Ages
(Age ⬍ 15 Years)
Shigella—S. dysenteriae,
S. flexneri, S. boydii,
S. sonnei
Worldwide; most cases and deaths
occur among children aged
⬍10 years
Vibrio cholerae
Sporadic cases in developed
countries, endemic areas, global
pandemic (Asia, Africa, and Latin
America) during the past 40 years
Salmonella typhi
Sporadic cases in developed
countries, about 16 million cases
and 600,000 deaths globally
346
(121)
Enterohemorrhagic
Escherichia coli O157:H7
Children are at greatest risk; acute
renal failure occurs in 2–7% of
cases
Enterotoxigenic
Escherichia coli
Travelers’ diarrhea, young children
in developing countries
Helicobacter pylori
Widespread globally; prevalence
increases with age; likely the most
common chronic infection in
humans; causes gastritis and
stomach cancer
17,521
(9,656)
Reservoir
Mode of Transmission
Humans
Direct or indirect fecal-oral from
case or carrier, water, milk
Humans, association
with copepods and
other zooplankton
Ingestion of fecal-contaminated
water, contaminated foods,
raw or undercooked seafoods
naturally contaminated from
polluted waters
Humans
Fecal-contaminated water or food
4,513
(1,806)
Cattle
Meat, other foods, drinking water,
swimming in contaminated
water, person-to-person
79,000
(est.)
Humans, animals
Fecal-contaminated food or water
Humans
Suspected oral-oral or fecal-oral
or contaminated water
6
(2)
NA
Cryptosporidium parvum
Worldwide but prevalence higher
in developing countries; outbreaks
in day-care centers; waterborne
outbreak in Milwaukee in 1993
affected about 400,000 people
2,361
(821)
Entamoeba histolytica
Ubiquitous organism; infection rare
before age 5 years
Giardia intestinalis
(G. lamblia)
Humans, animals
Fecal-oral including ingestion of
contaminated drinking or
recreational water, person-toperson, food
NA
Humans
Fecal-oral including ingestion of
contaminated water and food,
person-to-person
Worldwide but prevalence higher in
developing countries; risk higher
for children than for adults
NA
Humans, animals
Fecal-oral including ingestion of
contaminated drinking or
recreational water or food,
person-to-person
Toxoplasma gondii
Worldwide; prevalence of infection
quite high even in developed
countries
NA
Cats, other felines
Transplacental if mother first
infected ⬍6 months before
conception or during gestation;
ingestion of contaminated meat,
soil/sand, or waterc
Cyclospora cayetanensis
Worldwide; outbreaks during
recent years in the United States
and Canada
56
(10)
Unknown
Fecal-oral including ingestion of
contaminated food or water
Rotavirus
Most common cause of severe
diarrhea among children aged
⬍5 years worldwide; all children
infected by age 3–4 years
55,000
(est.)
Humans
Fecal-oral including ingestion of
contaminated water or food
and contact with contaminated
surfaces
(continued)
TABLE 12–4. Potential Waterborne Infection Threats for Childrena (continued)
Infectious Agent
Occurrence
Reported Annual
Cases, USA, 1999b
All Ages
(Age ⬍ 15 Years)
Hepatitis A
Worldwide; children at risk, e.g., in
day-care centers
17,047
(4,519)
Adenoviruses
Worldwide; adenoviruses 40 and 41
cause gastroenteritis in children
Norwalk and related
caliciviruses
Sporadic and epidemic cases
aCenters
for Disease Control and Prevention (2000, 2001b).
bCenters
for Disease Control and Prevention (2001d).
cMullens
(1996).
NA, not available.
Reservoir
Mode of Transmission
Humans, other
primates
Fecal-oral including person-toperson and ingestion of
contaminated food or water
NA
Humans
Fecal-oral including person-toperson and ingestion of
contaminated drinking or
recreational water
181,000
(est.)
Humans
Ingestion of contaminated oysters
or water
Water
355
volved hemorrhagic colitis and hemolytic uremic syndrome among children in a day-care school. Other water-related risk factors include swimming and consumption of foods such as lettuce exposed to contaminated
irrigation water.
Helicobacter pylori
About half of the world’s population has been infected with H. pylori, but
infection rates vary widely by geographic region and socioeconomic status. Prevalence of infection often approaches 75% by age 10 years in developing countries compared to less than 10% even among adults in developed countries. Helicobacter pylori likely contributes to the syndrome
of diarrhea, malnutrition, and growth failure of children in developing
countries by suppressing the stomach’s acid barrier and is an important
cause of gastritis, peptic ulcers, and stomach cancer during adulthood. A
longitudinal study of Japanese persons with gastrointestinal ulcers and
related conditions showed that stomach cancer developed in 3% of those
infected with H. pylori and in none of the uninfected persons (Uemura
et al., 2001).
Children in the developing world acquire H. pylori soon after birth,
and although the mode and route of transmission remain uncertain, it appears that infection may be acquired mainly from other family members
or close contacts. Evidence supports both fecal-oral and oral-oral (through
vomitus or possibly saliva) pathways. Risk factors for infection include
drinking water from nonmunicipal water sources, inadequate sanitation,
low social class, and high-density living conditions. Helicobacter pylori has
been detected in water, soil, flies, cow feces, and most surface and shallow groundwater samples in a U.S. survey. There is little relation between
the occurrence of total coliforms or E. coli in water and the presence of
H. pylori, suggesting that routine screening of water supplies for traditional
indicator organisms may fail to protect consumers from this organism.
Other Bacteria
Since the early 1800s, cholera has spread periodically to other parts of the
world in pandemic waves from endemic areas in Southeast Asia. The most
recent pandemic began in 1961 in Indonesia and spread rapidly to other
countries in Asia, Europe, Africa, and Latin America. Peru, Guatemala,
and other Latin American countries reported almost 1 million cases to the
Pan American Health Organization during 1991–1993. Cholera is spread
mainly through contaminated drinking water and food including consumption of raw or undercooked shellfish. Algal blooms may have sup-
356
CHILD HEALTH
AND THE
ENVIRONMENT
ported the growth of V. cholerae in marine and fresh waters in Peru and
contamination of fish, mollusks, and crustacea. Typhoid fever, a lifethreatening illness, affects over 10 million persons annually in the developing world, with case-fatality rates being highest among infants and the
elderly.
Cyanobacterial Toxins
Cyanobacteria (blue-green algae) produce several toxins that have adverse
effects in experimental animals ranging from birth defects to neurotoxicity and liver cancer. Proliferation of cyanobacteria in a newly flooded
reservoir area above the Itaparica dam in Brazil was the apparent cause
of about 2000 gastroenteritis cases, including 88 deaths, mostly among
children. Cyanobacteria levels in drinking water consumed by pregnant
women were inconsistently related to low birth weight, prematurity, and
birth defects in an ecologic study.
Protozoa
Protozoal infections pose threats even in developed countries because
these organisms are shed from infected hosts in feces as oocysts with thick
walls that enable them to resist chlorination. Disinfection of water contaminated by protozoa generally requires efficient filtration that is not always available.
Cryptosporidium parvum
There are no vaccines or effective treatments for C. parvum, a microbe responsible for up to 20% of childhood diarrhea in developing countries.
Most cases are not life-threatening, but children with immune deficiency
may develop serious, potentially fatal infections (e.g., children undergoing chemotherapy for cancer or organ transplantation). About 400,000
people developed acute watery diarrhea due to C. parvum in Milwaukee
in early 1993 because one of the city’s water treatment plant filtration systems failed to remove Cryptosporidium oocysts. Investigation showed that
emergency room visits and hospital admissions for gastroenteritis before
the outbreak were strongly associated with turbidity, with lag periods of
about 8 days among children.
There have been at least 20 smaller cryptosporidiosis waterborne outbreaks in the United Kingdom and North America since the Milwaukee
epidemic. Surveillance of C. parvum infections in seven states during
Water
357
1997–1998 revealed over 1000 laboratory-confirmed cases including a seasonal peak in late summer among children aged 0–14 years (Dietz et al.,
2000). Outbreaks have occurred after swimming in contaminated pools,
suggesting that the summer peak may relate in part to recreational exposure to contaminated water. Children are at risk of infection in swimming pools because C. parvum
• Can survive the usual chlorine levels for several days
• Is not removed by conventional pool water filters
• Has a low infective dose
A single fecal accident can contaminate an entire pool to such a degree
that a few mouthfuls of water can cause infection (Centers for Disease
Control and Prevention, 2001c). Cryptosporidium parvum oocysts in human
and animal feces can enter surface waters directly or through wastewater,
leaky septic tanks, and runoff and have been detected in shellfish intended
for human consumption. Calves appear to be a major reservoir, shedding
oocysts for up to 2 weeks; a mixture of human and bovine C. parvum genotypes were detected in about half of raw surface water samples from several areas in the United States (Xiao et al., 2001).
Giardia intestinalis
Giardia intestinalis (also known as G. lamblia) is a ubiquitous enteric parasite affecting humans and a range of wild and domestic mammals,
particularly dogs and dairy cattle. It is a globally important cause of waterborne diarrheal disease and the most common protozoan infection of
the human small intestine, causing diarrhea particularly in children, for
example, in day-care centers. Dormant Giardia cysts persist for months in
cold fresh water, and ingestion of ten cysts can initiate infection.
Toxoplasma gondii
The largest recorded human toxoplasmosis outbreak, and the only one
caused by a contaminated municipal water supply, occurred in Victoria,
Canada, during 1994–1995. Ultimately, 112 serologically confirmed cases,
including 37 pregnant women and 12 infants with T. gondii retinitis, were
identified. There was no association with domestic cats, but mapping of
cases and case-control studies showed that the incidence rate was highest
for persons living or working in an area served by one of the city’s water
distribution systems. Domestic cats, cougars, and deer mice living in
the watershed of a surface water reservoir showed serologic evidence of
358
CHILD HEALTH
AND THE
ENVIRONMENT
T. gondii infection, and cougars were found to shed T. gondii oocysts. The
municipal water system used unfiltered, chloraminated surface water and
was the likely source of this community-wide outbreak of toxoplasmosis.
Cyclospora cayetanensis
Surveillance and investigations of cyclosporiasis in Guatemala and Peru
have shown that prevalence rates were highest for young children and
that cases were associated with drinking untreated water. Investigation
of 96 cyclosporiasis outbreaks in the United States and Canada during
1996–1997 (totaling over 2000 cases) showed strong associations with consumption of fresh Guatemalan raspberries. The mode of contamination
of the raspberries remains unknown, but C. cayetanensis oocysts have been
detected in wastewater, a commodity used for irrigation in some regions;
also, preharvest pesticide applications used on raspberries may have been
diluted with contaminated water.
Viruses
Fecal-contaminated drinking water is an important source of the more
than 100 different viruses that can infect the gastrointestinal tract.
Rotavirus
Rotavirus mainly affects infants aged 6–24 months and is the most common cause of infantile watery diarrhea worldwide, with potentially lethal
dehydration occurring in almost 1% of affected infants. In developing
countries, rotavirus infections account for about 6% of all diarrheal episodes and 20% of diarrhea-associated deaths of young children. In the
United States, virtually all children have been infected by age 4 years. Reviews of childhood diarrhea in Australia, the United States, and Canada
showed that rotavirus was the leading identified pathogen (up to half of
all children aged 12–23 months). The virus particle is stable and can persist in the environment, including potable and recreational waters, and can
be concentrated by shellfish; waterborne outbreaks have been reported.
Enteric Adenoviruses
Although much less common than rotaviruses, enteric adenoviruses (especially Ad40 and Ad41) are important causes of infantile watery diarrhea globally, accounting for 3%–12% of cases. Water is an important route
of enteric adenovirus infection.
Water
359
Hepatitis Viruses
Hepatitis A and E viruses transmitted via contaminated water, often
through food vehicles, cause environment-related hepatitis. Hepatitis A
virus infects all age groups, causing an estimated 1.4 million cases globally, while the less well-studied hepatitis E virus is found mainly in the
developing world, often causing severe childhood infections. Assessment
of hepatitis A control programs has relied mainly on cross-sectional surveys of local populations. Prevalence rates of anti-hepatitis A antibody
among children aged 6–10 years in six Latin American countries varied
from 30% in Chile to 70%–80% in Mexico and the Dominican Republic
(Tanaka, 2000). Until about 1980, up to 90% of children aged 10–15 years
in Southeast Asia were anti-hepatitis A antibody-positive; by the early
1990s, the prevalence had declined substantially in several regions, for example, from 31% in 1987 to 13% in 1996 among children aged 10–19 years
in Bangkok. Hepatitis E infection often occurs in relation to the use of surface water for drinking, cooking, personal washing, and human excreta
disposal. Many large outbreaks of hepatitis E have occurred after heavy
rains and flooding that caused fecal contamination of surface waters and
shallow wells.
Other Viruses
Norwalk and Norwalk-like viruses and related caliciviruses frequently
cause gastroenteritis epidemics in developing countries and occasionally
in economically advantaged countries. In a Finnish community, there were
1500–3000 cases in 1994 caused by a contaminated well, the main infectious agent being a calicivirus. Molecular diagnostic methods have shown
that caliciviruses are the most common cause of acute gastroenteritis outbreaks in the United States and may be a common cause of sporadic cases
among children and adults. Astrovirus is a significant cause of acute diarrhea among children, with transmission thought to involve fecal-oral
spread from water or food sources. About 3% of hospitalized pediatric
gastroenteritis cases in Australia were attributed to astrovirus, and most
children had serologic evidence of previous infection by age 6 years.
Childhood diarrhea sometimes involves coinfection with astrovirus, rotavirus, and caliciviruses.
Risk Management
The main source of human pathogens in drinking or recreational waters
is feces of human or animal origin. Even in developed countries, problems arise when massive microbial contamination overwhelms disinfec-
360
CHILD HEALTH
AND THE
ENVIRONMENT
tion processes. Rainfall or flooding enables the spread of E. coli O157:H7
and other enteric pathogens in animal feces into runoff and leachate
from contaminated soils. For example, over 1300 cases of gastroenteritis,
including infections with E. coli O157:H7 and Campylobacter species, occurred in an Ontario town in 2000 due to heavy rain, farm runoff, and
contamination of at least one well (Health Canada, 2000).
Use of water free of human pathogens in food preparation, ingestion,
and personal hygiene is essential for personal and public health. Safe disposal of human feces, protection of water supplies from microbial contamination, adequate water disinfection, and personal hygiene continue
to be challenges in low-income populations globally. Although watershed
management is recognized as an important preventive strategy, water
supplies that can be safely used with no disinfection are relatively rare;
even in relatively pristine environments, surface water may be contaminated with animal excreta containing human pathogens. High population
density and poverty combine to favor microbial contamination of water
and inadequate or nonexistent water treatment technologies; simple measures such as boiling water for ingestion are problematic in many impoverished regions. Even where available, conventional sewage and
water treatment methods are relatively inefficient in the removal and inactivation of most enteric viruses and several protozoa.
Water treatment options aimed specifically at disinfection range from
boiling water to complex water treatment systems in large urban centers;
the reader is referred to other sources for more detailed information on
this topic (Craun et al., 2001). The effectiveness of water treatment facilities is especially important for surface or near-surface waters subject to
microbial contamination; for large populations, this generally requires relatively expensive technology that combines filtration and disinfection
processes. Proposed regulations to control waterborne C. parvum will require billions of dollars to improve water-treatment facilities (particularly
filtration) in the United States. It is important to note, however, that methods to reduce microbial contamination are similar to and compatible with
those aimed at reducing DBP levels.
Waterborne disease surveillance is an important component of communicable disease control programs. In the absence of timely surveillance,
pharmacists were the first to notice the Milwaukee outbreak after their
antidiarrheal remedies were sold out. After the outbreak, analysis of cryptosporidiosis cases in Milwaukee during 1993 revealed a smaller outbreak
about 3 weeks before the major epidemic. Effective surveillance might
have detected this early warning and triggered actions to prevent the major epidemic. Drinking water standards for microorganisms are presented
in Table 12–3.
Water
361
Conclusions
Proven Health Outcomes
• Waterborne infectious agents are the major cause of childhood diarrheal diseases globally, including an estimated 2 million annual childhood deaths mainly in developing countries
• Escherichia coli 0157:H7, a major cause of acute renal failure in children,
can be transmitted through contaminated drinking water
Unresolved Issues and Knowledge Gaps
• Water appears to be a source of H. pylori infection during childhood;
this organism causes chronic gastrointestinal infection and an increased
risk of peptic ulcers and stomach cancer during adulthood. Routine
screening of water supplies for E. coli or total coliforms may not protect consumers from this organism.
• There is inadequate evidence to assess the potential role of waterborne
infections in infantile methemoglobinemia.
Risk Management Issues
• Prevention—filtration is needed to reduce the hazards of both chlorineresistant microbes (e.g., C. parvum) and DBPs.
• Monitoring—frequent measurements of microbial concentrations in
drinking water are needed for early detection of breakdowns in disinfection practices and to evaluate progress in meeting drinking water
standards.
References
Agency for Toxic Substances and Disease Registry. (2001). 2001 CERCLA priority
list of hazardous substances. Located at http://www.atsdr.cdc.gov/clist.html
Arbuckle TE, Hrudey SE, Krasner SW, Nuckols JR, Richardson SD, Singer P, Mendola P, Dodds L, Weisel C, Ashley DL, Froese KL, Pegram RA, and others.
(2002). Assessing exposure in epidemiologic studies to disinfection by-products in drinking water: report from an international workshop. Environ Health
Perspect 110(Suppl 1):53–60.
Arbuckle TE, Sherman GJ, Corey PN, Walters D, Lo B. (1988). Water nitrates and
CNS birth defects: a population-based case-control study. Arch Environ Health
43:162–7.
Aschengrau A, Zierler S, Cohen A. (1993). Quality of community drinking water
and the occurrence of late adverse pregnancy outcomes. Arch Environ Health
48:105–13.
362
CHILD HEALTH
AND THE
ENVIRONMENT
Bove FJ, Fulcomer MC, Klotz JB, Esmart J, Dufficy EM, Savrin JE. (1995). Public
drinking water contamination and birth outcomes. Am J Epidemiol 141:850–62.
Bove F, Shim Y, Zeitz P. (2002). Drinking water contaminants and adverse pregnancy outcomes: a review. Environ Health Perspect 110(Suppl 1):61–74.
Bukowski J, Somers G, Bryanton J. (2001). Agricultural contamination of groundwater as a possible risk factor for growth restriction or prematurity. J Occup
Environ Med 43:377–83.
Centers for Disease Control and Prevention. (2000). Division of Bacterial and
Mycotic Diseases. Disease information. Located at http://www.cdc.gov/
ncidod/dbmd/diseaseinfo/default.htm
Centers for Disease Control and Prevention. (2001a). Bacterial waterborne diseases. Atlanta: Centers for Disease Control and Prevention.
Centers for Disease Control and Prevention. (2001b). Parasitic disease information. Located at http://www.cdc.gov/ncidod/dpd/parasites/listing.htm
Centers for Disease Control and Prevention. (2001c). Protracted outbreaks of cryptosporidiosis associated with swimming pool use—Ohio and Nebraska, 2000.
MMWR 50:406–10.
Centers for Disease Control and Prevention. (2001d). Summary of notifiable diseases, United States, 1999. MMWR 48 (No. 53):1–18.
Cohn P, Klotz J, Bove F, Berkowitz M, Fagliano J. (1994). Drinking water contamination and the incidence of leukemia and non-Hodgkin’s lymphoma. Environ Health Perspect 102:556–61.
Craun GF, Hauchman FS, Robinson DE. (2001). Microbial pathogens and disinfection by-products in drinking water: health effects and management of risks.
Washington, DC: International Life Science Institute.
Croen LA, Shaw GM, Sanbonmatsu L, Selvin S, Buffler PA. (1997). Maternal residential proximity to hazardous waste sites and risk for selected congenital
malformations. Epidemiology 8:347–54.
Croen LA, Todoroff K, Shaw GM. (2001). Maternal exposure to nitrate from drinking water and diet and risk for neural tube defects. Am J Epidemiol 153:325–31.
Cutler JJ, Parker GS, Rosen S, Prenney B, Healey R , Caldwell GG. (1986). Childhood leukemia in Woburn, Massachusetts. Public Health Rep 101:201–5.
Detels R, Holland W, McEwen J, Omenn GS. (1997). Oxford textbook of public
health. New York: Oxford University Press.
Dietz V, Vugia D, Nelson R, Wicklund J, Nadle J , McCombs KG, Reddy S. (2000).
Active, multisite, laboratory-based surveillance for Cryptosporidium parvum.
Am J Trop Med Hyg 62:368–72.
Dodds L, King W, Woolcott C, Pole J. (1999). Trihalomethanes in public water supplies and adverse birth outcomes. Epidemiology 10:233–7.
Dodds L, King WD. (2001). Relation between trihalomethane compounds and
birth defects. Occup Environ Med 58:443–6.
Dolk H, Vrijheid M, Armstrong B, Abramsky L, Bianchi F, Garne E, Nelen V, Robert
E, Scott JE, Stone D, Tenconi R. (1998). Risk of congenital anomalies near
hazardous-waste landfill sites in Europe: the EUROHAZCON study. Lancet
352:423–7.
Dorsch MM, Scragg RK, McMichael AJ, Baghurst PA, Dyer KF. (1984). Congenital malformations and maternal drinking water supply in rural South Australia: a case-control study. Am J Epidemiol 119:473–86.
Water
363
Dow JL, Green T. (2000). Trichloroethylene induced vitamin B(12) and folate deficiency leads to increased formic acid excretion in the rat. Toxicology 146:
123–36.
Eskes TK. (2000). Homocysteine and human reproduction. Clin Exp Obstet Gynecol 27:157–67.
Gallagher MD, Nuckols JR, Stallones L, Savitz DA. (1998). Exposure to trihalomethanes and adverse pregnancy outcomes. Epidemiology 9:484–9.
Goldberg SJ, Lebowitz MD, Graver EJ, Hicks S. (1990). An association of human
congenital cardiac malformations and drinking water contaminants. J Am Coll
Cardiol 16:155–64.
Graves CG, Matanoski GM, Tardiff RG. (2001). Weight of evidence for an association between adverse reproductive and developmental effects and exposure to
disinfection by-products: a critical review. Regul Toxicol Pharmacol 34:103–24.
Health Canada. (2000). Waterborne outbreak of gastroenteritis associated with a
contaminated municipal water supply, Walkerton, Ontario, May–June 2000.
Can Commun Dis Rep 26:170–3.
Hirose A, Nishikawa A, Kinae N, Hasegawa R. (1999). 3-Chloro-4-(dichloromethyl)5-hydroxy-2(5H)-furanone (MX): toxicological properties and risk assessment
in drinking water. Rev Environ Health 14:103–20.
Infante-Rivard C, Amre D, Sinnett D. (2002). GSTT1 and CYP2E1 polymorphisms
and trihalomethanes in drinking water: effect on childhood leukemia. Environ Health Perspect 110:591–3.
Infante-Rivard C, Olson E, Jacques L, Ayotte P. (2001). Drinking water contaminants and childhood leukemia. Epidemiology 12:13–9.
International Agency for Research on Cancer. (1995). IARC monographs on the
evaluation of carcinogenic risks to humans. Vol. 63. Dry cleaning, some chlorinated solvents and other industrial chemicals. Lyon, France: International
Agency for Research on Cancer.
International Agency for Research on Cancer. (2002). Search IARC Agents. Located at http://monographs.iarc.fr/htdig/search.html.
Jaakkola JJ, Magnus P, Skrondal A, Hwang BF, Becher G, Dybing E. (2001). Foetal
growth and duration of gestation relative to water chlorination. Occup Environ Med 58:437–42.
Kallen BA, Robert E. (2000). Drinking water chlorination and delivery outcome—
a registry-based study in Sweden. Reprod Toxicol 14:303–9.
Kanitz S, Franco Y, Patrone V, Caltabellotta M, Raffo E, Riggi C, Timitilli D, Ravera G. (1996). Association between drinking water disinfection and somatic parameters at birth. Environ Health Perspect 104:516–20.
King WD, Dodds L, Allen AC. (2000). Relation between stillbirth and specific chlorination by-products in public water supplies. Environ Health Perspect 108:
883–6.
Klotz JB, Pyrch LA. (1999). Neural tube defects and drinking water disinfection
by-products. Epidemiology 10:383–90.
Kramer MD, Lynch CF, Isacson P, Hanson JW. (1992). The association of waterborne chloroform with intrauterine growth retardation. Epidemiology 3:
407–13.
Lynberg M, Nuckols JR, Langlois P, Ashley D, Singer P, Mendola P, Wilkes C,
Krapfl H, Miles E, Speight V, Lin B, Small L, and others. (2001). Assessing
364
CHILD HEALTH
AND THE
ENVIRONMENT
exposure to disinfection by-products in women of reproductive age living in
Corpus Christi, Texas, and Cobb County, Georgia: descriptive results and
methods. Environ Health Perspect 109:597–604.
Magnus P, Jaakkola JJ, Skrondal A, Alexander J, Becher G, Krogh T, Dybing E.
(1999). Water chlorination and birth defects. Epidemiology 10:513–7.
Mueller BA, Newton K, Holly EA, Preston-Martin S. (2001). Residential water
source and the risk of childhood brain tumors. Environ Health Perspect 109:
551–6.
Mullens A. (1996). “I think we have a problem in Victoria”: MDs respond quickly
to toxoplasmosis outbreak in BC. CMAJ 154:1721–4.
Pruss A, Kay D, Fewtrell L, Bartram J. (2002). Estimating the burden of disease
from water, sanitation, and hygiene at a global level. Environ Health Perspect
110:537–42.
Reif JS, Hatch MC, Bracken M, Holmes LB, Schwetz BA, Singer PC. (1996). Reproductive and developmental effects of disinfection by-products in drinking
water. Environ Health Perspect 104:1056–61.
Savitz DA, Andrews KW, Pastore LM. (1995). Drinking water and pregnancy outcome in central North Carolina: source, amount, and trihalomethane levels.
Environ Health Perspect 103:592–6.
Sonnenfeld N, Hertz-Picciotto I, Kaye WE. (2001). Tetrachloroethylene in drinking water and birth outcomes at the U.S. Marine Corps Base at Camp Lejeune,
North Carolina. Am J Epidemiol 154:902–8.
Tanaka J. (2000). Hepatitis A shifting epidemiology in Latin America. Vaccine
18(Suppl 1):S57–60.
Thompson DJ, Warner SD, Robinson VB. (1974). Teratology studies on orally administered chloroform in the rat and rabbit. Toxicol Appl Pharmacol 29:348–57.
U.S. Environmental Protection Agency. (1998). National primary drinking water
regulations: disinfectants and disinfection byproducts; final rule. Fed Reg 63:
69389–476.
U.S. Environmental Protection Agency. (1999). A review of contaminant occurrence in public water systems (EPA 816-R-99-006). Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency. (2000). National treated water levels of total trihalomethanes for July–September 1998. Located at http://www.epa.gov/
enviro/html/icr/national/report/tthm/5.html
U.S. Environmental Protection Agency. (2001a). Current drinking water standards.
Located at http://www.epa.gov/safewater/mcl.html
U.S. Environmental Protection Agency. (2001b). IRIS substance list. Located at
http://www.epa.gov/iris/subst/index.html
U.S. Environmental Protection Agency. (2001c). Public drinking water systems:
facts and figures. Located at http://www.epa.gov/safewater/pws/factoids.
html
U.S. Environmental Protection Agency. (2001d). National source water contamination prevention strategy. Seventh draft for discussion. Washington, DC: U.S.
Environmental Protection Agency.
Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S, Yamakido M,
Taniyama K, Sasaki N, Schlemper RJ. (2001). Helicobacter pylori infection and
the development of gastric cancer. N Engl J Med 345:784–9.
Water
365
Vrijheid M. (2000). Health effects of residence near hazardous waste landfill sites:
a review of epidemiologic literature. Environ Health Perspect 108(Suppl 1):
101–12.
Waller K, Swan SH, DeLorenze G, Hopkins B. (1998). Trihalomethanes in drinking water and spontaneous abortion. Epidemiology 9:134–40.
Wartenberg D, Reyner D, Scott CS. (2000). Trichloroethylene and cancer: epidemiologic evidence. Environ Health Perspect 108(Suppl 2):161–76.
World Health Organization. (1998). Fact sheet No 178. Reducing mortality from
major killers of children. Geneva: World Health Organization.
Wright JM, Schwartz J, Vartiainen T, Maki-Paakkanen J, Altshul L, Harrington JJ,
Dockery DW. (2002). 3-Chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone
(MX) and mutagenic activity in Massachusetts drinking water. Environ Health
Perspect 110:157–64.
Xiao L, Singh A, Limor J, Graczyk TK, Gradus S, Lal A. (2001). Molecular characterization of Cryptosporidium oocysts in samples of raw surface water and
wastewater. Appl Environ Microbiol 67:1097–101.
Yang CY, Cheng BH, Tsai SS, Wu TN, Lin MC, Lin KC. (2000). Association between chlorination of drinking water and adverse pregnancy outcome in
Taiwan. Environ Health Perspect 108:765–8.
13
Conclusion
Environmental Threats to Child Health
This section describes the burden of child health conditions and summarizes their known and suspected environmental risk factors as discussed
in previous chapters. The purpose is to provide an overview of progress
to date in identifying environmental threats to the fetus and child and
to highlight areas where research and monitoring on environmental exposures and child health outcomes are needed. The summary tables (Tables 13–1 to 13–5) have important limitations:
• Critical evidence for several of the known causal factors came from
studies of children exposed to contaminant levels much higher than
those experienced today in the general population; there is considerable uncertainty about health risks at much lower exposure levels.
• With very few exceptions, existing knowledge does not permit estimation of attributable risks; thus, inclusion of risk factors for a given
adverse health outcome does not mean that these factors cause all or
even most of the cases.
• The relationships are limited to those for which exposure occurs prenatally or during childhood.
366
Conclusion
367
• The tables do not identify critical exposure pathways or timing (e.g.,
ingested vs. inhaled, prenatal vs. postnatal); for information on these
and other details, see the subject-area chapters.
• The tables do not address nonenvironmental risk factors such as genetic traits, infections (other than waterborne infections), maternal prenatal alcohol and drug use, and parental or child diet; for information
on these issues, see other sources.
Adverse Developmental Outcomes
Behavioral (tobacco, alcohol), pharmaceutical, and microbial factors are
important risk factors for some adverse developmental outcomes but the
causes of these conditions remain poorly defined. Based on sheer numbers, fetal deaths, low birth weight, and birth defects are major child
health burdens (Table 13–1). Although about 1 million recognized fetal
deaths (excluding therapeutic abortions) occur annually in the United
States, about the same number of fetal deaths occur very soon after conception but are not clinically recognized. There are few proven environmental causes of fetal death in humans, at least at widely prevalent exposure levels; there is limited evidence implicating prenatal parental
(usually maternal) exposure to lead, PCBs/dioxin-like compounds, pesticides, ETS, ionizing radiation, and THMs and suggestive but inadequate
evidence for ambient air pollutants. Suspected environmental causes of
IUGR or preterm birth include prenatal maternal exposure to lead, ETS
(independent of prenatal maternal smoking), ambient air pollution, and
THMs.
Birth defects comprise many distinct conditions with few known
causes. First trimester exposure to high-dose ionizing radiation caused
microcephaly (and mental retardation) among infants of Japanese atomic
bomb survivors. Suspected environmental causes of birth defects include
lead, pesticides, low-dose ionizing radiation, THMs, and methylmercury.
There is inadequate evidence for possible links between birth defects and
arsenic, TCDD, HAAs, power-frequency magnetic fields, certain drinking
water contaminants (nitrate/nitrite, non-THM solvents), and ambient air
pollution. Despite public and scientific concern about reports of declining sperm counts in young adult men and premature breast development
in young girls and potential links to hormonally active environmental
contaminants, no adequately designed epidemiologic studies have assessed such relationships. There is limited evidence that prenatal or early
childhood exposure to lead or PCBs/dioxin-like compounds can cause
growth deficits during childhood.
TABLE 13–1. Role of Prenatal Exposure to Environmental Agents in Adverse Developmental Outcomes
Outcome
Number of
events per year
(United States)
Fetal deaths (spontaneous
abortions and stillbirths)
983,000b
Low birth weight (2500 g)
301,000c
Preterm delivery
467,000c
Total birth defectsd
120,000e
Neural tube defects
2,500
Microcephaly f
2,500
Environmental risk factors and level of epidemiologic evidencea
Sufficient
PCBs/dioxin-like
compounds
Limited
Inadequate
Lead, PCBs/dioxin-like
compounds, pesticides, prenatal
maternal ETS exposure, parental
ionizing radiation exposure,
THMs (or related DBPs)
Arsenic, HAAs, maternal exposure
to power-frequency magnetic
fields, multiple outdoor air
pollutants
Lead, pesticides, prenatal maternal
ETS exposure, outdoor air
pollution (CO, SO2, PM), THMs
(or related DBPs)
Arsenic, cadmium, pesticides, HAAs,
ionizing radiation exposure,
power frequency magnetic
fields, nitrate in drinking
water, non-THM VOCs in
drinking water
Lead, pesticides, prenatal
maternal ETS exposure
HAAs
Arsenic, TCDD, HAAs, powerfrequency magnetic fields
Lead, herbicides, preconceptual
parental or first trimester
maternal ionizing radiation
exposure, THMs (or related
DBPs)
Ionizing radiation
(prenatal atomic
bomb exposure)
Methylmercury
Nitrate/nitrite in drinking water,
non-THM VOCs in drinking
water, lead
Cardiac defects
13,000
Lead, pesticides, preconceptual
parental or first trimester
maternal ionizing radiation
exposure
Urinary tract defectsg
3,600
THMs (or related DBPs)
Hypospadias (second
and third degree)
1,000
Cryptorchidism
(undescended testicles)
3,900
Oral clefts
6,700
Lead, pesticides
Limb reduction defects
2,200
Pesticides
Skeletal defects
NA
Reduced stature during
childhood
NA
Ambient air pollution (CO), THMs,
non-THM VOCs in drinking
water, lead
Pesticides, HAAs
Ambient air pollution (CO), THMs,
non-THM solvents in drinking
water
Cadmium
Lead, PCBs/dioxin-like
compounds
ETS environmental tobacco smoke; TCDD 2,3,7,8-tetrachlorodibenzo--dioxin; THMs trihalomethanes.
a Sufficient evidence—based on peer-reviewed reports of expert groups or authoritative reviews that concluded that a causal relationship existed; limited evidence—includes relationships for which several epidemiologic studies, including at least one case-control or cohort study, showed fairly consistent associations and evidence of exposure–risk relationships after control for potential confounders; inadequate evidence—relationships for which epidemiologic studies were limited in number and quality (e.g., small studies, ecologic
studies, limited control of potential confounders), had inconsistent results, or showed little or no evidence of exposure–risk relationships.
b Ventura
et al. (1999).
c Ventura
et al. (2001).
d Birth
e Based
defect birth prevalence rates are from California unless otherwise indicated (California Birth Defects Monitoring Program (2002).
on 3% of the approximately 4 million births annually in the United States.
f Microcephaly
g Urinary
birth prevalence rate from U.S. Metropolitan Atlanta Congenital Defects Program (International Clearinghouse for Birth Defect Monitoring Systems, 1999).
tract birth prevalence rate from Canadian Congenital Anomaly Surveillance System (International Clearinghouse for Birth Defect Monitoring Systems, 1999).
NA not available
370
CHILD HEALTH
AND THE
ENVIRONMENT
Cancer
There have been numerous well-designed epidemiologic studies of relationships between prenatal and childhood exposures and risks of childhood and young adult cancers. There is sufficient evidence for causal roles
for prenatal or childhood ionizing radiation exposure in childhood leukemia, brain cancer, and thyroid cancer and in adult cancers including
breast and thyroid cancers and leukemia (Table 13–2). The evidence for
these relationships came mainly from groups with high-level exposure
(atomic bomb, Chernobyl nuclear reactor accident, therapeutic radiation).
The lifetime excess cancer risk from a given dose of ionizing radiation
during childhood appears to be about double that for an equivalent exposure during adulthood. The attributable risk of environmental sources
of ionizing radiation is unknown but likely to be quite small because of
the low prevalence of high exposures. Intense sun exposure during childhood is a major cause of malignant melanoma and basal cell skin cancers
that occur mainly during adulthood.
Limited epidemiologic evidence precludes firm conclusions about the
potential causal roles of pesticides, paternal smoking/ETS, ionizing radiation (parental occupational exposure, radioactive fallout from atmospheric nuclear tests), certain parental occupational exposures, and powerfrequency magnetic fields in childhood cancers. Polychlorinated biphenyls
have been linked to adult cancers including breast and non-Hodgkin’s
lymphoma but the role of childhood PCB exposures is unknown.
Neurobehavioral Outcomes
Neurobehavioral problems among children range from subtle deficits detectable only through specialized testing to severely disabling conditions
(Table 13–3). High-dose prenatal maternal exposures to methylmercury,
PCBs/dioxin-like compounds, and ionizing radiation are known causes
of severe neurobehavioral effects including motor, sensory, and cognitive
deficits. Early childhood exposure to lead is a known cause of sensory,
motor, and cognitive deficits, and there is suggestive/limited evidence
for a causal role in attention deficit hyperactivity disorder. There is limited evidence that relatively low-level exposure in early childhood to
methylmercury or PCBs/dioxin-like compounds from contaminated
foods (especially fish) can cause cognitive and other neurobehavioral
deficits. Occupational exposure of youth and young adults to OP and
carbamate insecticides has been linked to persistent sensorimotor
deficits.
TABLE 13–2. Role of Prenatal or Childhood Exposure to Environmental Agents in Childhood and Adult Cancersa
Environmental Risk Factors and Level of Epidemiologic Evidenceb
Outcome
Cases per year
(United States)c
Sufficient
Limited
Inadequate
Childhood cancers
(Age 0–19 years)
12,800
Leukemia
3,250
Lymphomas
Brain
PCBs, other PHAHs, RF radiation,
indoor air VOCs
Ionizing radiation
(prenatal diagnostic
x-rays)
Pesticides, paternal smoking,
childhood ionizing radiation
exposure (diagnostic X-rays),
radioactive fallout from nuclear
testing, child exposure to powerfrequency magnetic fields, paternal
occupational exposure to solvents
and paints
Cadmium, ambient air pollution (traffic
density, benzene), paternal preconceptual ionizing radiation exposure, childhood environmental radiation exposure
(radon, other), paternal exposure to
power-frequency magnetic fields, radiofrequency radiation, indoor air VOCs,
THMs non-THM VOCs in drinking
water
1,700
Pesticides, paternal smoking (? ETS
or other mechanism), parental
occupational exposure to solvents
and other petroleum products
Ambient air pollution, ionizing radiation
(prenatal diagnostic)
2,200
Pesticides, paternal smoking (? ETS
or other mechanism), paternal
occupational exposure to solvents
and paints
Childhood environmental radiation
exposure (radon, other), powerfrequency magnetic or electric fields
or radiofrequency radiation, paternal
exposure to power-frequency magnetic
fields, drinking water nitrite
(continued)
TABLE 13–2. Role of Prenatal or Childhood Exposure to Environmental Agents in Childhood and Adult Cancersa (continued)
Environmental Risk Factors and Level of Epidemiologic Evidenceb
Outcome
Cases per year
(United States)c
Wilms’ tumor
500
Sufficient
Limited
Pesticides
Germ cell tumors
900
Thyroid tumors
350
Melanoma
325
Intense sun exposure
(see also adult melanoma below)
Ewing’s sarcoma
(bone)
200
Pesticides
Inadequate
Prenatal maternal ionizing radiation
(diagnostic X-rays)
Childhood ionizing
radiation exposure
Childhood environmental radiation
exposure (radium in drinking water)
Adult cancers
(Age ⱖ 20)
Breast
1,285,000
205,000
Testicular
7,500
Brain
17,000
Ionizing radiation
during childhood
(atomic bomb
survivors, highdose X-rays)
Role of childhood exposure to PCBs and
other PHAHs
PCBs, ETS
Cadmium, HAAs
Ionizing radiation
during childhood
(therapeutic radiation)
Thyroid
20,700
Ionizing radiation
during childhood
(therapeutic radiation)
Leukemia
30,800
Ionizing radiation
during childhood
(atomic bomb
survivors)
Non-Hodgkin’s
lymphoma
53,900
Melanomad
53,600
Kidney
31,800
Bladder
56,500
Lung
169,400
Stomach
21,600
Ionizing radiation during childhood
(radioactive fallout from nuclear
testing)
HAAs
PCBs
Sun exposure during
childhood
Lead
THMs (or related DBPs)
ETS during childhood
TCDD
H. pylori (infection during
childhood from
contaminated drinking
water is a probable source)
Note: exposure refers to childhood exposures unless otherwise indicated.
a Limited
b See
to prenatal and childhood exposures.
relevant chapters for supporting epidemiologic evidence.
c Children:
d Basal
cases per year, U.S SEER program, 1990–1995 (SEER, 1999); adults: estimated cases, 2002, American Cancer Society, 2002.
cell carcinomas are far more common than melanoma but are not included in most cancer registries (or data are very incomplete).
TABLE 13–3. Role of Prenatal or Childhood Exposure to Environmental Agents in Childhood Neurobehavioral Conditions
Environmental Risk Factors and Level of Epidemiologic Evidencea
Number of affected children
(United States)b
Outcome
Sufficient
Limited
Cognitive deficits
760,000 (mental retardation)c
Lead, methylmercury,
ionizing radiation,
PCBs/dioxin-like
compounds
Motor deficits
270,000 (complete or partial
paralysis of limbs)d
Lead, methylmercury,
elemental mercury
PCBs/dioxin-like compounds,
certain organophosphate
and carbamate pesticides
Visual or hearing deficits
1,300,000 (visual or
hearing impairment)d
Lead, methylmercury
Certain organophosphate and
carbamate pesticides
Cerebral palsy (congenital)
160,000e
Methylmercury
Attention deficit
hyperactivity disorder
360,000c
Learning disabilities
200,000c
a See
relevant chapters for supporting epidemiologic evidence.
b Estimated
by multiplying prevalence rates times the number of persons age 18 years in United States (2000).
c Parent-reported,
d United
e Boyle
aged 18 years, United States, 1992–1994 (Halfon and Newacheck, 1999).
States, 1996, aged 18 years (Adams et al., 1999).
et al. (1996).
Pesticides
Lead
Ionizing radiation
Inadequate
Lead
Conclusion
375
Respiratory Diseases
The only known environmental cause of incident asthma is house-dust
mite antigen (Table 13–4). Factors known to exacerbate asthma include
cat, cockroach, and house-dust mite antigens, ETS (in preschool-age children), and ambient air pollution. Environmental tobacco smoke is a
known cause of acute respiratory conditions (bronchiolitis, pneumonia)
and middle ear infections, and a probable cause of incident asthma in preschool-age children. There is suggestive evidence that incident asthma can
be caused by cockroach antigens, personal NO2, and outdoor activities in
high-ozone areas. Exacerbation of existing asthma has been linked to dog,
rodent, fungal, and soybean antigens, and NO2 or formaldehyde from indoor sources. There is inadequate evidence of a causal role for several indoor contaminants in incident asthma or as triggers of asthma episodes.
Other Diseases
ETS is a probable cause of SIDS independent of prenatal maternal smoking, a known cause (Table 13–5). Moderately high childhood lead exposure is a known cause of anemia. Other known causal relationships
include TCDD and PCBs/dioxin-like compounds and chloracne, intense
sun exposure and benign skin nevi (moles), acute high-dose exposure
to pesticides, lead, or CO and symptoms and signs of acute poisoning,
and fecally contaminated drinking water and gastrointestinal infections.
Lead and various forms of mercury have also been linked to increased
urinary protein excretion suggestive of renal tubular damage. Perinatal
PCB exposure has been linked to increased risk of infections during
infancy.
Knowledge Development Policy Issues
This section is restricted to knowledge gaps and the need for improved
environmental health research programs and tracking systems. Policy
needs related to environmental risk management are beyond the scope of
this book.
Information Gaps
The information summarized above and material presented throughout
this book point to several key information gaps that limit our ability to
TABLE 13.4. Role of Prenatal or Childhood Exposure to Environmental Agents in Childhood Respiratory and Related Diseases
Outcome
Environmental Risk Factors and Level of Epidemiologic Evidencea
Prevalence or number
of events per year
(United States)b
Sufficient
Limited
Inadequate
Asthma
4.4 million children aged
18 years
House-dust mite antigens
Prenatal and postnatal maternal
smoking (in preschool-age
children), cockroach antigen,
personal air NO2, outdoor
activities in high-ozone areas
ETS (in school-age children),
animal and fungal antigens,
endotoxins, pollen, Alternaria,
VOCs, formaldehyde,
pesticides, plasticizers
Asthma episodes
14,000,000 school absence
days for asthma among
children aged 5–17 years
ETS in preschool-age children,
cat, cockroach, and housedust mite antigens, outdoor
air pollutants (SO2, ozone,
PM10)
ETS (in school-age children),
dog, rodent, fungal and
soybean antigens, NO2
from indoor sources,
formaldehyde
Cow and horse antigens,
insects other than
cockroaches or house-dust
mites, endotoxins, pollens,
pesticides, VOCs, plasticizers
Acute respiratory
conditions (mainly
infections)
207 million restricted
activity days per year
among children aged
18 years
ETS, outdoor PM10
Indoor air NO2, outdoor air
pollutants (SO2, ozone,
PM10)
Idiopathic pulmonary
hemosiderosis
NA
Middle ear infections
34 million restricted
activity days per year
among children
aged 18 years
a See
relevant chapters for supporting epidemiologic evidence.
bUnited
States, 1996 (Adams et al., 1999).
Stachybotrys chartarum
(fungus)
ETS
Conclusion
377
understand and prevent adverse health effects of environmental hazards,
including knowledge of
• The role of prenatal and childhood exposures to environmental, genetic, and other factors in the etiology of adverse developmental outcomes and childhood and adult diseases
• The distribution of environmental contaminant exposure levels among
reproductive-age persons, infants, and children in the general population
• Given the above two information gaps, and with some exceptions, it is
not yet possible to quantify with confidence
° The proportions of adverse pregnancy outcomes and childhood and
adult diseases attributable to environmental contaminants
° The benefits of interventions to reduce exposure levels
° Progress in reducing environmental threats to child health
Policy Needs
To address the knowledge gaps described above, there is an urgent need
for countries and international agencies to invest in population and laboratory research on the role of environmental hazards in fetal and child
health and development, including (Bennett and Waters, 2000; Carroquino
et al., 1998; Lucier and Schecter, 1998; Needham and Sexton, 2000; Pirkle
et al., 1995; Suk et al., 1996; U.S. General Accounting Office, 2000; Warren
and Shields, 1997; Weaver et al., 1998):
• Research infrastructure—scientists, laboratories, childhood cancer
registries
• Large-scale epidemiologic research programs and projects that incorporate strong statistical power, exposure assessment (including the
periconceptual period), biomarkers of exposure and susceptibility, and
control of potential confounders
• Tracking systems—purpose-defined surveys and special studies to
measure baselines and time trends for exposure to environmental contaminants and the occurrence of adverse health effects including structural and functional abnormalities and diseases
• The United States has taken promising steps in these directions (Centers for Disease Control and Prevention, 2001; U.S. Environmental Protection Agency, (2000a, 2002b). These should be maintained, and other
countries and international agencies should take up the challenge.
TABLE 13–5. Role of Prenatal or Childhood Exposure to Environmental Agents in Other Childhood Conditions
Outcome
Environmental Risk Factors and Level of Epidemiologic Evidencea
Number of
events per year
(United States)
Sufficient
Limited
Inadequate
Sudden infant death
syndrome (SID)
2,600 deaths
per year b
Prenatal maternal smoking
Anemia
NA
Lead
Renal function
abnormalities
NA
Lead, mercury (all three forms),
fecally contaminated drinking
water—E. coli 0157:H7)
Immune function
(susceptibility to
infections) during
infancy
NA
Hypothyroidism
(neonatal)
NA
NA
PCBs (prenatal
hypothyroidism)
Premature menarche
NA
NA
HAAs
Reduced sperm quality
NA
NA
HAAs
Chloracne (severe acnelike skin condition)
NA
TCDD, PCBs/dioxin-like
compounds
ETS (independent of prenatal
maternal smoking)
Cadmium
PCBs/dioxin-like
compounds
Other nonmalignant
skin abnormalities
conditions
NA
Tooth abnormalities
(hypomineralization)
NA
Poisonings
54,500c
Lead, arsenic, pesticides, CO
Gastrointestinal viral
infections, aged
18 years
20 million
restricted
activity days
per year d
NA
Fecally contaminated drinking
water
Infantile
methemoglobinemia
Cataracts
a See
PCBs/dioxin-like compounds
(skin pigmentation),
inorganic mercury (skin rash),
sun exposure (nevi or “moles”)
PCBs/dioxin-like
compounds
NA
relevant chapters for supporting epidemiologic evidence.
b United
States, 1999 (National Center for Health Statistics, 2002).
c United
States, 2000 (Litovitz et al., 2001).
d United
States, 1996 (Adams et al., 1999).
Nitrate in drinking
water (especially
farm wells)
Sun exposure during
childhood
380
CHILD HEALTH
AND THE
ENVIRONMENT
Epilogue
Until recently, countries and international health agencies have generally
assigned lower priority to addressing environmental threats to fetal and
child health than to more immediate or obvious threats, particularly infectious diseases. Certainly, the dramatic increases in life expectancy observed during the twentieth century were mainly attributable to the control of childhood infectious diseases. History, however, has shown
repeatedly the potential for serious adverse health effects after prenatal
or childhood exposures to environmental toxicants such as lead and ionizing radiation.
Much of the evidence for the proven environmental health hazards
described in this book comes from studies of pregnant women or children who were highly exposed accidentally or deliberately at a time when
the health risks were incompletely understood. In several instances, exposure of children and/or pregnant women has continued even after
demonstration of likely health risks. This drives home the need to avoid
arrogance and to adopt a precautionary approach that protects children
from exposure to hazards in the face of scientific uncertainty and opposition from vested interests.
In the past, epidemiologic studies had a limited ability to quantify
the health risks of low-level environmental contaminant exposure. Recent
advances in environmental exposure and genetic susceptibility assessment promise to enable substantially improved knowledge of health risks
including exposure–risk relationships and gene–environment interactions. Assuming that this materializes, it will help ensure evidence-based
public health decisions that improve child health protection within a sustainable economy.
This book will be supplemented by a website at the R. Samuel
McLaughlin Centre for Population Health Risk Assessment, Institute of
Population Health, University of Ottawa, Ottawa, Canada. The website
will include supplementary tables and citations relevant to the topics covered in this book. The website address is http://www.mclaughlincentre.ca.
References
Adams PF, Hendershot GE, Marano MA. (1999). Current estimates from the National Health Interview Survey. Vital Health Stat 10:212 p.
American Cancer Society. (2002). Facts and figures 2002. Located at http://www.
cancer.org/
Bennett DA, Waters MD. (2000). Applying biomarker research. Environ Health
Perspect 108:907–10.
Conclusion
381
Boyle CA, Yeargin-Allsopp M, Doernberg NS, Holmgreen P, Murphy CC, Schendel DE. (1996). Prevalence of selected developmental disabilities in children
3–10 years of age: the Metropolitan Atlanta Developmental Disabilities Surveillance Program, 1991. MMWR CDC Surveill Summ 45:1–14.
California Birth Defects Monitoring Program. (2002). Birth defects in California.
Located at http://www.cbdmp.org/index.htm
Carroquino MJ, Galson SK, Licht J, Amler RW, Perera FP, Claxton LD, Landrigan
PJ. (1998). The U.S. EPA conference on preventable causes of cancer in children: a research agenda. Environ Health Perspect 106(Suppl 3):867–73.
Centers for Disease Control and Prevention. (2001). National report on human exposure to environmental chemicals. Atlanta: Centers for Disease Control and
Prevention.
Halfon N, Newacheck PW. (1999). Prevalence and impact of parent-reported disabling mental health conditions among U.S. children. J Am Acad Child Adolesc Psychiatry 38:600–9.
International Clearinghouse for Birth Defect Monitoring Systems. (1999). Annual
report (1999). Rome: International Centre for Birth Defects.
Litovitz TL, Klein-Schwartz W, White S, Cobaugh DJ, Youniss J, Omslaer JC, Drab
A, Benson BE. (2001). 2000 annual report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med
19:337–95.
Lucier GW, Schecter A. (1998). Human exposure assessment and the National Toxicology Program. Environ Health Perspect 106:623–27.
National Center for Health Statistics. (2002). Fast stats A to Z. Infant mortality. Located at http://www.cdc.gov/nchs/fastats/infmort.htm
Needham LL, Sexton K. (2000). Assessing children’s exposure to hazardous environmental chemicals: an overview of selected research challenges and complexities. J Expo Anal Environ Epidemiol 10:611–29.
Pirkle JL, Sampson EJ, Needham LL, Patterson DG, Ashley DL. (1995). Using biological monitoring to assess human exposure to priority toxicants. Environ
Health Perspect 103(Suppl 3):45–8.
SEER. (1999). Cancer Incidence and Survival among Children and Adolescents:
United States SEER Program 1975–1995. National Cancer Institute, SEER Program (NIH Pub. No. 99-4649). Ries LAG, Smith MA, Gurney JG, Linet M,
Tamra T, Young JL, Bunin GR (eds). Bethesda, MD.
SEER (Surveillance, Epidemiology, and End Results). (2002). Childhood cancer by
site incidence, survival and mortality. Located at http://seer.cancer.gov/
Suk WA, Collman G, Damstra T. (1996). Human biomonitoring: research goals
and needs. Environ Health Perspect 104(Suppl 3):479–83.
U.S. Environmental Protection Agency. (2002a). Centers for Children’s Environmental Health and Disease Prevention Research. Located at http://es.epa.gov/ncer/
centers/cecehdpr/01/.
U.S. Environmental Protection Agency. (2002b). The National Children’s Study.
Located at http://nationalchildrensstudy.gov/
U.S. General Accounting Office. (2000). Toxic chemicals. Long-term coordinated
strategy needed to measure exposures in humans (GAO-HEHS-00-80). Washington, DC: U.S. General Accounting Office.
Ventura SJ, Martin JA, Curtin SC, Menacker F, Hamilton BE. (2001). Births: final
data for 1999. Natl Vital Stat Rep 49:16–7.
382
CHILD HEALTH
AND THE
ENVIRONMENT
Ventura SJ, Mosher WD, Curtin SC, Abma JC, Henshaw S. (1999). Highlights of
trends in pregnancies and pregnancy rates by outcome: estimates for the
United States, 1976–96. Natl Vital Stat Rep 47:1–9.
Warren AJ, Shields PG. (1997). Molecular epidemiology: carcinogen-DNA adducts
and genetic susceptibility. Proc Soc Exp Biol Med 216:172–80.
Weaver VM, Buckley TJ, Groopman JD. (1998). Approaches to environmental exposure assessment in children. Environ Health Perspect 106(Suppl 3):827–32.
Index
Aboriginal populations, 2, 108–10, 154
Acid aerosols
exposure, sources, 315, 320, 326
health effects, 308, 310
Acrodynia (pink disease), 9, 106, 112
Acute lymphoblastic leukemia. See
Leukemia
Adolescence
brain remodeling, 8, 64
cell phone use, 254–55
environmental tobacco smoke
exposure, 281
general, 6–11
ionizing radiation (breast cancer), 235
nitrogen dioxide exposure (hockey
arenas), 290
obesity, 79
sexual maturation, 193, 199, 210–11
sun exposure (melanoma), 259
Aeroallergens. See also Cat or dog
allergens; Cockroaches; Housedust mites; Fungi; Sensitization;
Soybean allergen
asthma, bronchial hyperreactivity, 274,
281–84, 305, 310–11
exposure, sources, 284–86
intervention, 286–88
summary information, 376
Agent Orange, 143, 170
Aggregate or cumulative risk, 51, 56–58,
181, 220
Agricultural chemicals. See Pesticides
Air intake, child, 12
Air pollution. See Indoor air; Outdoor air
Alachlor, 180
Alcohol, 8–9
Aldehydes (including formaldehyde)
indoor air, 271, 274, 276, 288–93
outdoor air, 305–6, 316, 325–26
Aldicarb, 175–176, 179
Aldrin, 180–82
Allergens. See Aeroallergens
Allergies. See also Aeroallergens
asthma, 272, 281–87
inflammatory reaction, 304
PCBs, 147
sulfur dioxide, 310
Alternaria, 286, 311
Androgen receptor. See Endocrine
system
383
384
Anemia
arsenic, 119
benzene, 289
lead, 73, 79
Anticholinesterase, 56, 165, 169–72
Arsenic
cancer, other chronic disease, 118–19
developmental effects, 118
exposure levels, 119
health-based exposure limits, 52–57, 120
intervention, 119–21
molecular mechanisms, 117–18
Aryl hydrocarbon receptor (AhR)
AhR-mediated toxicity, 138–40, 145–46
AhR-independent toxicity, 141, 145
dioxin toxic equivalency, 140
Asbestos, 271, 281
Asthma
biologic agents, 281–84
burden, 1, 3–5
diagnosis, risk factors, 272–73
environmental tobacco smoke, 275,
278–79
indoor environmental risk factors, 274
lung development, 10–11
outdoor air contaminants, 307–11
pesticides, 170–71
summary information, 375–76
Atomic bomb survivors, 231–35, 238
Atopy. See Allergies
Atrazine, 5, 173–74, 177–80
Attention deficit, attention deficit
hyperactivity disorder (ADHD)
lead, 76–78
methylmercury, 103–4, 113
research need, 9, 20, 64
summary information, 370, 374
Auditory function, hearing loss
importance, 5
lead, 52–54, 72–73, 76–78
methylmercury, 102, 104
PCBs, 145
summary information, 370, 374
thyroid hormone receptor, 199–200
Autism
burden, 64
research need, 5–6, 9, 20
Basal cell carcinoma, skin, 257–61
Behavior, 7. See also Hand-mouth
behavior
Benchmark dose
definition, regulatory use, 52–56, 67
INDEX
methylmercury neurotoxicity, 113
PCB neurotoxicity, 144
Benomyl, 208
Benzene
activation, by cytochromes, 13–14
cancer, 289–91, 312
exposure, sources, 276, 289–92, 316–17,
320–21
genotoxicity, 289, 336
health-based exposure limits, 55
Benzo[a]pyrene, 14, 35, 140, 203
Bias, 28–32, 36–42
Biomarkers of exposure (biomonitoring)
arsenic, 119
cadmium, 123
carbon monoxide, 291, 317
diesel particulate matter, 317
environmental tobacco smoke, 280
exposure assessment, 15, 31–34, 51
hormonally active agents, 214–15
ionizing radiation, 232
lead, 79–82
mercury, 106–8
PCBs, dioxin-like compounds, 149–50,
214
pesticides, 171–73
trihalomethanes, 343
ultraviolet light, 257
volatile organic carbons, 290–91, 317,
343–45
Biomarkers of susceptibility. See
Polymorphisms
Birth defects. See also specific birth defects
arsenic, 118
burden, 1–5
cadmium, 122
clusters, 35–36
drinking water, 337–40
electromagnetic fields, 246–47
hormonally active agents, 208–11
ionizing radiation, 234
lead, 79
manganese, 127
outdoor air pollution, 306–7
pathogenesis, 8, 17
PCBs, dioxin-like compounds, 140–43
pesticides, 167–68
reproductive tract, 209–10. See also
Cryptorchidism; Hypospadias
risk assessment, 62–64
summary information, 367–69
trends, 190–91
Birth weight. See Intrauterine growth
retardation; Preterm birth
Index
Bisphenol A, related compounds
ER agonist activity, 201, 204–5
hormonal activity in mice, 210–12
production, uses, 216, 220–21
Blood–brain barrier, 9, 13
Brain cancer
burden, 4
disinfection by-products, 341
electromagnetic fields, 247–51
environmental tobacco smoke, 279
ionizing radiation, 237–38
outdoor air contaminants, 312
pesticides, 169
summary information, 371–72
Brain development, 8–9
Breast cancer
diethylstilbestrol, 213
electromagnetic fields, melatonin,
245–46
environmental tobacco smoke, 279
ionizing radiation, 235
PCBs, 146
summary information, 372
Breast development, premature, 211
Breast milk
arsenic, 119
methylmercury, 106
monitoring, 22
PCBs, other organochlorine
compounds, 149–50, 173
Bromodichloromethane. See also
Trihalomethanes
cancer, 56
fetal death, 337
health-based exposure limits, 53, 57
Bronchial hyperreactivity
asthma, 308–9
environmental tobacco smoke, 278
house-dust mites, 287
mixed outdoor air pollutants, 311
nitrogen oxides, 305, 310
ozone, 305, 309
Bronchitis, bronchiolitis
environmental tobacco smoke, 272–75
PCBs, dioxin-like compounds, 147
summary information, 376
Butyrylcholinesterase, 171–72
Cadmium, 53–55, 117, 121–25
Cancer. See also specific cancers
burden, 1–6
summary information, risk factors,
371–73
385
Carbamate pesticides. See Aldicarb;
Carbaryl
Carbaryl, 171–77
Carbon monoxide
developmental effects, 306–7
exposure, sources, 291–93, 314, 317–8
poisoning, 289–90, 305
Carcinogenicity
arsenic, 117–19
cadmium, 121–22
drinking water contaminants, 335–36,
341–42
electromagnetic fields, 247–51
environmental tobacco smoke, 276
hormonally-active agents, 213, 220
ionizing radiation, 232–38
lead, 74
methylmercury, 106
outdoor air contaminants, 302–4,
311–12
pesticides, 169–70, 173, 180
PCBs, dioxin-like compounds, 140,
145–46
risk assessment, 48–50, 56–62
summary information, 372–73
sunlight, ultraviolet light, 257–59
testing, 18
volatile organic carbons, 289
Cardiac birth defects
carbon monoxide, 306–7
di-(2-ethylhexyl) phthalate (DEHP),
210–11
disinfection by-products, 338–39
ionizing radiation, 234
lead, 79
mutagens, 17
pesticides, 168
summary information, 369
Case-control studies, 40
Cat or dog allergens, asthma, 271–74,
282–87, 375–76
Causal relationships, criteria, 48–50, 58
Cell phones, 244–46, 251–56. See also
Electromagnetic fields
Cerebral palsy
burden, 3, 5
methylmercury, 101–3
pesticides, 169
summary information, 374
Chemical toxicity testing. See also
Carcinogenicity; Developmental
toxicity; Genotoxicity;
Neurotoxicity
cumulative risks, 56
386
Chemical toxicity testing (continued)
general, 7, 14–21
hormonally active agents, 201–8, 215
mixtures, 18, 48
number of commercial chemicals,
15–19
pesticides, 179–82
priorities, 19
Chernobyl, 230–36
Child health
environmental risk factors, summary,
366–82
leading conditions, 1–5
Chloracne, 137–42, 147–48
Chlordane
breast milk, 173
cancer, 170
home uses, 171, 177
persistent organochlorine pesticides,
181–82
schools, 178–79
Chlordecone
infertility, males, 168
persistent organochlorine pesticides,
181–82
hormonal activity, 201–2, 205
Chlorpyrifos
baby foods, 175
developmental neurotoxicity, 165
exposure, sources, 171–79
health-based exposure limits, 53–55
home uses, 171, 177–78
intervention, 181
schools, 178–79
Chromosome abnormalities
arsenic, 117
benzene, 289
biomonitoring, 33–34
cadmium, 121–22
disinfection by-products, 335
fetal deaths, 3, 8, 17
ionizing radiation, 232–33
pesticides, 168, 170–72
radiofrequency radiation, 246
ultraviolet radiation, 257
Clean Air Act (United States)
dioxin (TCDD), 154
lead, 16, 84
mercury, 109–10
MMT, 129
national ambient air quality standards,
322, 326
Climate change, 1, 5, 19–20
Cluster investigation, 35–37
INDEX
Cockroaches, 283–85
Cognitive deficits
cadmium, 122
ionizing radiation, 238
lead, 39, 72–73, 76–78,
manganese, 126–27
mercury, 101–6
neurotoxicity testing, 17–18
PCBs, dioxin-like compounds, 138,
144–45, 211–12
risk assessment, 64–66,
summary information, 370, 374
Cohort studies, 38–39
Confounding, 20, 28–31, 38, 41–42, 49
Congenital anomalies. See Birth defects
Cotinine, 276, 280, 317
Coumestrol, 198, 205–6
Cross-sectional studies, 38–39
Cryptorchidism
antiandrogenic pesticides, 11
dioxin (TCDD), 143
incidence trends, 190–91
pesticides, 168, 209–10
summary information, 369
testicular cancer, 213
Cryptosporidium parvum, 350, 356–57,
360–61
Cumulative or aggregate risk, 51, 56–58,
181, 220
Cyanobacterial toxins, 356
Cyclospora cayetanensis, 358
Cystic fibrosis, 12
Cytochromes
activation, detoxification, 8, 13–14, 140
intrauterine growth retardation, 277
lead toxicity, 74
PCB, dioxin-like compounds, toxicity,
140, 146, 207–8
pesticides, childhood leukemia, 170
trihalomethanes, childhood leukemia,
341
2,4-D (2,4-dichlorophenoxyacetic acid)
cancer, 170
dioxin contamination, 155
exposure, 171–73
health-based exposure limits, 180
indoor air, dust, surface levels, 177–78
sales, 173–74
Daidzein, 205–6
DBCP (1,2-dibromo-3-chloropropane)
groundwater contamination, 179
male infertility, 66, 168
Index
testicular toxicity, 11
DDT, DDE
(dichlorodiphenyltrichloroethane,
dichlorodiphenyldichloroethylene)
breast milk, 150, 173, 214
cancer, 146, 213
cryptorchidism, hypospadias, 168,
190–91, 209–10
developmental effects, 167–68
drinking water guideline, 180
feminization of males, 11, 190
food residues, 176
history, 162–63, 216
hormonal activity, 201–2, 205–10
house-dust, surface levels, 177
immune system effects, 170–71
infections, 47
malaria control, 181–82
neurotoxicity, 169
serum, 214
DES. See Diethylstilbestrol
Detoxification, 7–14, 74, 117–18, 165, 341
Developmental neurotoxicity, conditions
ionizing radiation, 238
lead, 75–78
methylmercury, 100–106
PCBs, dioxin-like compounds, 141–45
pesticides, 165, 168–69, 176, 179–81
risk assessment, 64–66
summary information, 374
Developmental toxicity, conditions
burden of developmental effects, 2–6
history, 14–15
developmental toxicity, 17–18, 52–53
risk assessment, 62–64
summary information, 368–69
Diarrhea, 350–59
Dibenzo[a,h]anthracene, 207
-Dichlorobenzene, 172–73, 290–92
Dieldrin,
breast milk, 173
drinking water, 179–80
hormonal activity, 201
leukemia, 146–47
neurotoxicity, 165
risk management, 182
Diesel particulate matter
biomarkers, 317
cancer, 311–12, 318
hormonal activity, 204, 211
nitrogen oxides, 318, 320
particulate matter, 301–5, 311, 318
sulfur dioxide, 315
Diethylstilbestrol (DES)
387
birth defects, 209, 211
history, 14
cancer, 61, 190, 213
reproductive effects, 66, 190, 209, 211
hormonal activity, 201, 205
immune system effects, 212
Dioxin (2,3,7,8-tetrachlorodibenzo-dioxin or TCDD). See also
Polychlorinated dibenzo()dioxins
carcinogenicity, 145–46
female reproductive tract development,
203, 211
feminization of males, 203, 210
health-based exposure limits, 54–55,
152–56
high exposure groups, 143, 145–48
immune system effects, 10
sources, 154–55
testicular function, 207–8
thyroid modulation, 208
toxic equivalency (TEQ), 138–40, 211
toxic mechanisms, 138–41, 203, 207–8
Disinfection by-products (DBPs). See
Drinking water, disinfection
by-products
DNA repair
arsenic, cadmium, 117, 162
diethylstilbestrol, 205
ionizing radiation, 232–34, 237
ultraviolet light, 257
Dose-response assessment, 30, 42–51,
58–66
Drinking water, chemical contaminants
arsenic, 119–20
carcinogens, genotoxins, 335–36
consumption, 12
developmental effects, 340–41
exposures above MCLs, prevalence, 344
hormonally active agents, 218
health-based exposure limits, 347–49
lead, 89
levels in water, 344–45
monitoring, 15, 21
PCBs, dioxin (TCDD), 154–55
pesticides, 179
radionuclides, 238, 240–41
volatile organic carbons, 325–26
Drinking water, disinfection by-products
cancer, 341–42, 370–73
developmental effects, 337–41, 367–69
exposure levels, 342–45
genotoxicity, 335–36
health-based exposure limits, 345–49
sources, 342–47
388
Drinking water, microbial agents
bacteria, 350–56
protozoa, 356–58
viruses, 358–59
Dyslexia, 9, 64
Ear infections. See Otitis media
Ecologic studies. See Epidemiology
Economic costs, 1–2, 48, 275
Electromagnetic fields, 243–56
cancer, 247–51
cell phones, 244–46, 251–56
developmental effects, 246–47
electromagnetic spectrum, 230
exposure levels, 252–53
health-based exposure limits, 253
intervention, 253–56
microwaves, 244, 252–54
molecular mechanisms and biological
effects, 245–46
power-frequency electromagnetic
fields, 6, 29–30, 243–56
power lines, 29–30, 244, 247–55
radiofrequency electromagnetic fields,
244–47, 251–56
sources, 253–54
Endocrine disruptors. See Hormonally
active agents
Endocrine system
androgen receptor, 193–203, 207–8
estradiol-17, 193–95, 200–207, 211–12
estrogen receptor, 198–207
follicle stimulating hormone, 192–97
luteinizing hormone, 192–97
Mullerian inhibitory hormone, 193
normal function, 192–201
progesterone, 193–97
progesterone receptor, 199
sex hormone binding globulin, 200
sex steroid synthesis, 196–97
testosterone, 117, 193–200, 203, 209–11
thyroid hormone receptor, 195, 199–200
thyroid stimulating hormone, 192,
195–98, 208
thyrotropin releasing hormone, 192,
196
thyroxine
brain development, 199–200
cadmium, 122–23
function, 195
PCBs, neurotoxicity, 145, 208
production, 196–98
transthyretin, 141, 208
INDEX
triiodothyronine, 195–201
Endosulfan, 175–77, 201, 205
Endotoxin, 274, 283–88
Endrin, 182
Enteric infections. See Drinking water,
microbial agents
Environmental tobacco smoke
biomonitoring, 15
carcinogens, 276–77
health effects, 274–81
importance, 5
summary information, 367–68, 375–76
Epidemiologic evidence
bias, 29, 36–40
confounding, 29
inconsistency, 34
level of evidence, 58–61
limitations, 28–31
strengths, 27–28
temporal ambiguity, 38–39
Epidemiology study types
case-control studies, 40
case reports, 34–35
cluster investigation, 35–37
cohort studies, 40–41
cross-sectional studies, 38–39
ecologic studies, 37–38
experimental studies, 41–42
Epilepsy, 9, 76
Escherichia coli, 351–52, 355, 360–61
Estradiol-17, 193–95, 200–7, 211–12
Estrogen receptor, 198–207
Ethics, 30–31
Exposure assessment. See also Biomarkers
bias, 31–32
environmental contaminant levels,
32–33
general, 15, 28–34, 51, 61
internal dose estimation, 32–34
monitoring, 21–22
Fallout, radioactive, 235–36, 239
Feminization of males
diethylstilbestrol (DES), 209
dioxin (TCDD), 210
pesticides, 11, 209–10
phthalates, 210–11
Fertility
androgen receptor mutations,
polymorphisms, 199
pesticides, 168
reproductive toxicants, 66–67, 202–4, 211
semen quality, time trends, 191
Index
Fetal deaths
arsenic, 118
burden, 2–3, 8
DES, 211
disinfection by-products, 337–38
electromagnetic fields, 246–47
environmental tobacco smoke, 277
fetal abnormalities, 8
ionizing radiation, 234
lead, 79
mutagens, 17
outdoor air contaminants, 306
PCBs, dioxin-like compounds, 143
pesticides, 167–68
risk assessment, 62–64
summary information, 367–68
Fire retardants, 216, 221. See also
Polybrominated diphenyl ethers
Folic acid, 8
Follicle stimulating hormone. See
Endocrine system
Food consumption, child, 12
Food contaminants
food consumption, 11–12
dietary habits, 22
pesticides, 20
Food Quality Protection Act
aggregate and cumulative risks, 51,
181, 220
uncertainty factors, developmental risk
assessment, 52, 176–77
Formaldehyde
asthma, 274, 288–89
cancer, 276
exposure levels, 291–92, 316
indoor air, 271
outdoor air sources, 325
Fuels, fuel additives (motor vehicle)
alcohol, 325–26
benzene, 317, 320–21, 325–26
lead, 16, 83
MMT, 125–29, 320–21, 325–26
particulate matter, 301–4
sulfur dioxide, strong acidity, 301, 315
volatile organic carbons, 305–6, 320–21,
325–26
Fungi
asthma, 283–84, 311
exposure levels, 286
Fungicides. See also specific fungicides
baby foods, pesticide residues, 175
cumulative toxicity, 220
sales, 174
Furans. See Polychlorinated dibenzofurans
389
Gene-environment interaction. See also
Polymorphisms
arsenic, 117–18
cadmium, 121–22
cancer, 341
detection, 34
dioxin-like compounds, 140
general, 12–13
lead toxicity, 73–74
Genistein, 18, 198, 201, 205–6, 212. See
also Phytoestrogens
Genotoxicity
diesel exhaust particulate, 304
drinking water, 335–36
electromagnetic fields, 245–46
general, 8, 13–14, 17
ionizing radiation, 232–34
volatile organic carbons, 305–6
German Environmental Survey, 15, 89
Giardia intestinalis, 357
Glucans, 286
Glutathione, 13, 100, 117, 272
Growth, prenatal. See Intrauterine growth
retardation; Preterm birth
Growth, postnatal. See Stature
Hand-mouth behavior, enhanced
exposure to toxicants
general, 11
lead, 82, 86
pesticides, 165–66, 178
phthalates, 219
Hazard identification, 48–50
Hazardous air pollutants, 301, 316, 322
Hazardous wastes
cadmium, 124
importance, 5
PCBs, dioxin-like compounds, 150–51
priority contaminants, 335–36
volatile organic carbons, 339–49
Health-based exposure limits, 52–56, 67.
See also specific toxicants
Helicobacter pylori, 355
Hepatitis, 359
Heptachlor
breast milk, 173
cryptorchidism, 209
intervention, 181–82
exposure, home environment, 171
Herbicides
birth defects, 168
cancer, 170
dioxin contamination, 151, 155
390
Herbicides (continued)
drinking water, 179–80
exposure, home environment, 173–74,
177–78
fetal deaths, 167
sales, 171–74
Hexachlorobenzene
breast milk, 140, 173
cord blood, 149
cryptorchidism, 209
ear infections, 147
poisoning, 166
Hexachlorocyclohexane, 146
High production volume (HPV)
chemicals, 16, 19
Hodgkin’s disease, 312
Home environment. See Drinking water;
Electromagnetic fields; Indoor Air;
Lead; Mercury; Pesticides; Radon
Hormonally active agents. See also
Endocrine system
androgen modulators, 207–8
bioassays, 201, 205–8
cancer, 213
developmental effects, 209–11
estrogen modulators, 201, 205–7
exposure levels, 214–15
immune system modulators, 212
intervention, 215–23
mechanisms, 201–8
reproductive effects, 209–11
sources, 215–21
thyroid modulators, 208
Hormones. See Endocrine system
House-dust contaminants, 15
House-dust mites, 273, 282–85
Hypospadias
antiandrogenic pesticides, 11
diethylstilbestrol (DES), 209
incidence trends, 190–91
pesticides, 168, 209–10
summary information, 369
Idiopathic pulmonary hemosiderosis, 11,
284
Immune system
general, 10
inflammation, 272, 281–82, 304
sensitization, 281–86
Immunotoxicants
benzene, 55
fungi, 11, 283
hormonally active agents, 201–3, 212
lead, 79
INDEX
mercury, 114
PCBs, dioxin-like compounds, 140–47
pesticides, 162, 170–71
sunlight, ultraviolet light, 260
Incinerators
dioxins, furans, 145–46, 151, 154–55
mercury, 108–10
Indoor air. See also Environmental
tobacco smoke
biologic agents, 281–88
contaminant categories, 271
monitoring, 15, 21
volatile organic carbons and gases,
288–95. See also Volatile organic
carbons
Infections. See Otitis media; Drinking
water, microbial agents
Insecticides, 162–66. See also Pesticides
Intrauterine growth retardation
arsenic, 118
cadmium, 122
developmental toxicants, 62–64
drinking water contaminants, 337,
340–41
environmental tobacco smoke, 277
ionizing radiation, 234
mutagens, 17
outdoor air contaminants, 306
PCBs, dioxin-like compounds, 141
pesticides, 167
summary information, 367–68
Ionizing radiation, 229–43, 367–73
background exposure levels, 239–40
cancer, 234–38, 371–73
Chernobyl, 230–36
developmental effects, 234
emissions, nuclear power production
reactors, 239
exposure levels, 239–40
exposure units, 232
fallout, radioactive, 235–36, 239
genotoxicity, 232–34
health-based exposure limits, 241
intervention, 240–43
neurotoxicity, 238
nuclear weapon tests, 235–36, 239–40
parental exposure, prenatal, 236–37
radon, 231–32, 238–43
X-rays, 230–38
Issues, child health and the environment,
2–5
Kidney conditions. See Renal conditions,
toxicity
Index
Lead
developmental effects, 78–79
dose-reponse relationships, 72–73,
76–81, 85
exposure levels, 79–83
health-based exposure limits, 52, 54,
88–90
history, 72
intervention, 82–92
mechanisms, toxicity, 73–75
neurotoxicity, 75–78
nutritional factors, 74, 79, 89–90
poisoning, 2, 71–78, 90
Learning disabilities
general, 9, 20
importance, 5
ionizing radiation, 238
lead, 74–78
PCBs, dioxin-like compounds, 144
pesticides, 165
Leukemia
burden, 4
clusters, 36
disinfection by-products, 341
electromagnetic fields, 247–51
environmental tobacco smoke, 279
ionizing radiation, 234–38
motor vehicle emissions, 312
PCBs, dioxin-like compounds, 145–46
pesticides, 169–70
polymorphisms, 14
summary information, 370–71
Level of evidence
carcinogens, 58–62
developmental toxicants, 62–64
environmental risk factors, summary,
366–82
reproductive toxicants, 63, 66–67
Limb reduction birth defects
mutagens, 17
pesticides, 168
Linuron, 11, 207, 210
Low birth weight, burden, 2–5. See also
Intrauterine growth retardation;
Preterm birth
Lowest observed adverse effect level
(LOAEL), 52, 55, 67
Lung development, 10–11
Lung function deficits, 277–78, 284, 307–11
Lung function, spirometry, 277, 307
Luteinizing hormone, 192–97
Magnetic fields. See Electromagnetic
fields
391
Malathion, 167, 171, 177
Manganese, 125–30. See also MMT
Melanoma, 146, 258–59, 372–73
Melatonin, 245–46
Mental retardation. See Cognitive deficits
Mercury
elemental mercury, 53–55, 100, 113–16
emissions, 108–11
environmental levels, 108–12
food levels, 110–12
health-based exposure limits, 53–55,
107, 111–15
inorganic mercury, 9, 100–101, 113–14
methylmercury
exposure biomarkers, 106–8
intervention, 107–13
major cohort studies, summary, 104–5
neurotoxicity, 101–6
poisoning, 9, 100–103
sources, 2, 108–12
Metallothioneins, 117, 122
Metals. See Arsenic; Cadmium; Lead;
Manganese; Mercury
Methoxychlor, 177, 201, 205, 212
Methylmercury. See Mercury
Microwaves, 244, 252–54. See also
Electromagnetic fields
Middle ear infections, 146–47, 170–71,
275, 279
Minamata, 101–3, 107, 110–12. See also
Mercury
Minimal risk level, 54–55. See also risk
management sections of other chapters
Miscarriage. See Fetal deaths
MMT (methylcyclopentadienyl
manganese tricarbonyl), 125–29
Moles (nevi), 258, 375, 379. See also
Sunlight
Mullerian inhibitory hormone, 193
Mutations. See also Genotoxicity;
Polymorphisms
general, 7, 12
germ cell mutations, 199, 232–33, 279
somatic cell mutations, 121, 232–33,
257–58, 336
MX, 335, 341–43, 347
National ambient air quality standards
(NAAQS)
compliance, monitoring, 312–17
criteria air pollutants, 301
lead, 85
review process, 322–23
standards, 323
392
National Health and Nutrition
Examination Survey (NHANES)
asthma, 273, 282–83
biomonitoring, 15, 92
cadmium, 123
hazard identification, 39
lead, 74, 77, 80–84, 90
mercury, 114
PCBs, dioxin-like compounds, 149
pesticides, 172
phthalates, 214
Neural tube birth defects
arsenic, 118
di-(2-ethylhexyl) phthalate (DEHP),
210–11
disinfection by-products, 337–40
ionizing radiation, 234
lead, 79
mutagens, 17
pesticides, 168
summary information, 368
trichloroethylene, tetrachloroethylene,
339
Neuropathy
arsenic, 118–19
elemental mercury, 113–14
lead, 78
methylmercury, 102
pesticides, 169
Neurotoxicity
general, 1–2, 8–9
risk assessment, 16–18, 64–66
toxicity testing, 7–9, 16–18, 64–66, 165,
180–81
Nevi, 258, 379. See also Sunlight
Nitrate, 334, 342, 367–68, 379
Nitrogen oxides
cancer, 312
health-based exposure limits, 322–24,
327
monitoring, 312–17
ozone formation, 318–19
respiratory effects, 308–11
sources, 317–26
summary information, 376
time trends, air levels, 314, 321
toxicity, 305
No observed adverse effect level
(NOAEL), 52, 67
Non-Hodgkin’s lymphoma
paternal smoking, 279
PCBs, 146
pesticides, 169–70
Nonmelanoma skin cancer, 257–61
INDEX
Non-threshold toxicants, 56–57
Nonylphenol, 201, 205, 212, 216–18
Norwalk virus, 359
Nuclear power production reactors, 229,
235–36, 239. See also Ionizing
radiation
Nuclear weapon tests, 235–36, 239–40. See
also Ionizing radiation
Nutrition
arsenic, 117
general, 1, 8, 12
lead, 78
mercury, 111
pesticides, 174
puberty, onset, 211
Organisation for Economic Cooperation
and Development (OECD), 16, 19,
52
Organochlorine compounds. See also
Dioxin; Hormonally active agents;
Pesticides; Polychlorinated
biphenyls; Polychlorinated
dibenzo()dioxins; Polychlorinated
dibenzofurans
breast milk levels, 16, 22, 149–50
international intervention, 155
Organophosphate pesticides
aggregate and cumulative risks, 56
biomonitoring, 172
intervention, 176–82
neurotoxicity, 165, 169, 181
poisoning, 35
sources, 171–79
Organotin compounds, 208, 218, 221
Orofacial cleft birth defects
disinfection by-products, 338–39
dioxin (TCDD), 143
lead, 79
mutagens, 17
pesticides, 168
Otitis media, 146–47, 170–71, 275, 279, 376
Outdoor air
air quality index, 326–28
cancer, 311–13, 371
criteria air pollutants, 301–6, 312–28
developmental effects, 306–7, 367–69
exposure levels, 312–17
intervention, 322–28
monitoring, 312–17
national ambient air quality standards
(NAAQS), 85, 301, 309, 312, 315,
322–24
Index
respiratory effects, 307–12, 376
smog, 300–301, 305–6, 315–19
sources, 317–22
toxic mechanisms, 302–6
trends, 321–22
Ovarian development, 11, 193–96
Ozone, 304–309, 312–19, 322–24
p53, 121, 232, 257
PCBs. See Polychlorinated biphenyls
Pentachlorophenol
dioxin contamination, 137, 151, 155
exposure levels, 172–73
food levels, 176–77
health-based exposure limits, 180
intrauterine growth retardation,
167
Perinatal conditions, burden, 2–5. See also
Birth defects; Fetal deaths;
Intrauterine growth retardation;
Preterm birth
Permethrin, 175, 177
Personal air
outdoor air pollutants, 307–10, 315–17,
321, 325
pesticides, 177
volatile organic carbons, 290
Pesticides. See also specific pesticides
asthma, 170–71
cancer, 169–70
carbamates, 166–76
classes, 164, 174
cumulative risk, 56, 58
developmental effects, 167–68
developmental neurotoxicity, 165,
168–69, 176, 180–81
exposure levels, 171–73
foods, residues, 169, 174–77, 181,
220–21
herbicides, 151, 155, 163, 167–75, 177–81
immunotoxicity, 170–71
intervention, 173–83
neurotoxicity, 168–69, 370, 374
organophosphates, 163–69, 171–83
poisonings, 35, 166, 379
reproductive toxicity, 66, 168–69
rodenticides, 120, 166
sales, 173–74
Pet allergens, asthma, 271–74, 282–87
Phthalates
biomonitoring, 214
di-(2-ethylhexyl) adipate (DEHA),
217–19
393
di-(2-ethylhexyl) phthalate (DEHP),
208–211, 214, 219–20
diethylphthalate (DEP), 214–16, 220
di-isononyl phthalate (DINP), 210, 214,
219–20
di-n-butyl phthalate (DBP), 208–10,
214, 217–20
di-(n-hexyl) phthalate (DHP), 210
hormonal activity, 201, 204, 207–8
reproductive toxicity, 11, 210–11
sources, 216–20
Phytoestrogens, 201, 205–6, 212, 218. See
also Coumestrol; Daidzein;
Genistein; Zearalenone
Placenta, permeability to toxicants, 14,
100, 113, 118, 149
Poisoning
burden, 3
carbon monoxide, 289–93
elemental mercury, 113–14
inorganic mercury, 113–15
lead, 71–78, 90–92
methylmercury, 101–3, 111
pesticides, 35, 166, 175–76, 179, 182
summary information, 379
surveillance, 35, 90–92
Policy issues
economic considerations, 48
electromagnetic fields, 254–55
general, 19–22, 30–31, 39
lead, 90–91
scientific evidence base, 47, 375–80
Polybrominated biphenyls, 143, 211
Polybrominated diphenyl ethers, 150, 212
Polychlorinated biphenyls
biomonitoring, 16, 33, 149–50, 214
carcinogenicity, 140, 145–46, 203
health-based exposure limits, 53–55,
152–56
hormonal activity, 203, 206
infections, 170–71
intervention, 150–57
neurotoxicity, 103
production, uses, 216
sources, 137–38, 141–43, 150–56, 218
summary information, 367–69, 370–73
thyroid effects, 208
uses, 216
Polychlorinated dibenzo()dioxins. See
also Dioxin
carcinogenicity, 145–46, 170
exposure biomarkers, 149–50
immune system modulation, 212
intervention, 150–57
394
Polychlorinated dibenzo()dioxins
(continued)
poisoning, 137–38, 141–43
sources, 137–38, 143, 150–56, 173,
203–209
summary information, 367–69, 370–73
Polychlorinated dibenzofurans, 63,
136–43, 147–48, 208
Polycyclic aromatic hydrocarbons
environmental tobacco smoke, 276
general, 12–14
hormonal activity, 207
immune system effects, 10
intrauterine growth retardation, 306, 314
lung development, 10
outdoor air particulate matter, 302, 314
Polyhalogenated aromatic hydrocarbons
(PHAHs), 136–37
Polymorphisms
disinfection by-products, leukemia, 341
environmental tobacco smoke, low
birth weight, 277
general, 12–14
ionizing radiation, leukemia, 237–38
lead toxicity, 73–74
pesticides, genotoxicity, 172
pesticides, leukemia, 170
Poverty, 2, 5, 7
Power-frequency electromagnetic fields,
6, 29–30, 243–56
Power lines, 29–30, 244, 247–55
Precautionary principle, 19–20, 71, 115,
129, 255, 262
Precision, 29, 42
Preterm birth
environmental tobacco smoke, 277
lead, 78
nitrate (drinking water), 341
outdoor air contaminants, 306
PCBs, dioxin-like compounds, 143
pesticides, 167
power-frequency magnetic fields, 246
summary information, 367–68
Priorities, 4–5
Procymidone, 207, 210, 220
Progesterone, 193–97, 205
Puberty. See Adolescence
Pyrethrins, 163–64, 171, 179
Pyrethroids, 163–68, 174, 179, 182. See also
Permethrin
Radiation. See Electromagnetic fields;
Ionizing radiation; Sunlight
INDEX
Radiofrequency electromagnetic fields,
244–47, 251–56
Radon, 38, 231–32, 238–43, 371
Renal conditions, toxicity
cadmium, 55, 121–25
E. coli O157:H7, 351–55
lead, 73, 75, 79
mercury, 106, 113–14
summary information, 378
Reproductive toxicity risk assessment,
66–67
Reproductive tract development
germ cells, 17, 56
Mullerian duct, 193
sexual development, 11, 192–96
testicles
cadmium, 122
normal development, 192–95
reproductive toxicants, 66–67, 209–10
Wolffian duct, 193, 195
Research, U.S. children’s environmental
health research centers, 20
Respiratory disease. See also Asthma;
Indoor air pollution; Outdoor air
pollution
burden, 2–5
summary information, risk factors, 376
Risk assessment
carcinogens, 58–62
developmental toxicants, 62–64
epidemiologic evidence, 28, 42–44
framework, 47–58
neurotoxins, 64–66
reproductive toxicants, 66–67
Risk characterization, 51–58
Risk estimates. See also risk management
sections of other chapters
aggregate or cumulative, 51, 56–58,
181, 220
genotoxic carcinogens, 56–57
noncancer health effects, 52–56
Rotavirus, 358
Schools
indoor air contaminants, 281, 285, 288
pesticides, 169, 177–81
power-frequency electromagnetic
fields, 254
radon, 232–33, 240–42
sunlight, 262
waterborne infections, 351, 355
Secondhand smoke. See Environmental
tobacco smoke
Index
Sensitization. See also Aeroallergens
asthma, 272
immunologic sensitization and
inflammation, 281–89, 311
Sertoli cell
barrier, 67
benomyl, 208
phthalates, 207–211
testicular development, 193–94
Seveso, 143–48, 154
Sexual development, 11, 192–96
Shigella sonnei, 350–52
Skin conditions. See also Acrodynia (pink
disease); Basal cell carcinoma,
skin; Chloracne; Melanoma; Nevi;
Non-melanoma skin cancer
hyperpigmentation, 141
porphyria cutanea tarda, 166
xeroderma pigmentosum, 257
Smog, 300–301, 305–6, 315–19
Soybean allergen, asthma, 311
Sperm quality
pesticides, 67, 168, 210–11
phthalates, other hormonally active
chemicals, 210–11
trends over time, 191
Spermatogenesis, 11
Spina bifida. See Birth defects
Stachybotrys chartarum, 11
Stature
environmental tobacco smoke, 277
lead, 79
PCBs, dioxin-like compounds, 141–43
pesticides, 167
summary information, 369
Steroid synthesis, 196–97
Stillbirth. See Fetal deaths
Sudden infant death syndrome (SIDS)
burden, 3
environmental tobacco smoke, 279–80
outdoor air contaminants, 306
Sulfur dioxide, 301–6, 310–29, 375–76
Sunburn
nevi (moles), 257–58
prevalence, 260
skin cancer, 258–59
ultraviolet light exposure, 256–57
Sunlight. See also Sunburn
exposure levels, 260
intervention, 260–63
melanoma, 258–59, 370, 372
molecular mechanisms, 257–58
nevi, 258, 375, 379
non-melanoma skin cancer, 259
395
Sunscreens, 261–63
Surveillance
carbon monoxide poisoning, 289–90
descriptive epidemiology, 35
drinking water, 356–61
general, 20–22
lead exposure, 81, 90–92
pesticide poisonings, 166, 179
phthalate exposure, 214
Susceptibility, 73–74, 163–66. See also
Polymorphisms
Swimming
disinfection by-products, 292, 343–45
waterborne infections, 352, 355–57
Teratogenicity, 8, 11, 14–17
Testicular cancer
cadmium, 122
cryptorchidism, relation to, 213
time trends, 190–91, 213
Testosterone, 117, 193–203, 209–11
Tetrachloroethylene, 53
Three Mile Island, 230
Threshold toxicants, 52–56, 62–64
Thyroid cancer
hormonally active agents, 213
ionizing radiation, 234–37
PCBs, dioxin-like compounds, 145
summary information, 373
time trends, 4, 236
Thyroid hormones. See Endocrine system
Tobacco. See Environmental tobacco smoke
Toxaphene
hormonal activity, 201
persistent organochlorine pesticides,
181–82
Toxicity testing
chemical mixtures, 48
developmental toxicity, 17–18, 62–64
hormonal activity, 201–8, 215
neurotoxicity, 64–66
number of commercial chemicals,
16–19
pesticides, 179–82
priorities, 19
reproductive toxicity, 66–67
screening information data set, 19
Toxoplasma gondii, 357–58
Tracking systems. See Surveillance
Trichloroethylene, 334–48
2,4,5-Trichlorophenoxyacetic acid, 151,
155, 170
Trihalomethanes, 334–50
INDEX
396
Uncertainty factors. See also Food Quality
Protection Act
sources of uncertainty, 20
health-based exposure limits, 52–56, 67
Undescended testicles. See
Cryptorchidism
Ultraviolet light. See Sunlight
Urinary tract birth defects
disinfection by-products, 337–40
mutagens, 17
summary information, 369
Vaginal cancer, 201, 205, 213
Validity, 29
Vibrio cholerae, 350–52, 355–56
Vinclozolin, 201–2, 207, 210
Visual abnormalities, deficits
lead, 77–78
manganese, 126
methylmercury, 102–4
PCBs, dioxin-like compounds, 144
pesticides, 169
summary information, 374
visual recognition memory test, 65
Volatile organic carbons
asthma, 288–89
cancer, 289, 312, 342
developmental effects, 337–41
environmental levels, 316, 344–45
exposure levels, 290–92, 316–17
intervention, 291–93, 325–26, 347–48
ozone formation, 305–6, 318–19
sources, 291–93, 305–6, 317–22
Vulnerability, 1, 7–14
Water consumption, child, 12
Waterborne infections. See Drinking
water, microbial agents
Wireless phones. See Electromagnetic
fields, cell phones
Xeroderma pigmentosum, 257
X-rays. See Ionizing radiation
Yucheng, Yusho, 141–43, 146–49
Zearalenone, 205–6