Article
Repositioning of the global epicentre of
non-optimal cholesterol
https://doi.org/10.1038/s41586-020-2338-1
NCD Risk Factor Collaboration (NCD-RisC)*
Received: 18 October 2019
Accepted: 2 April 2020
Published online: 3 June 2020
Open access
Check for updates
High blood cholesterol is typically considered a feature of wealthy western
countries1,2. However, dietary and behavioural determinants of blood cholesterol are
changing rapidly throughout the world3 and countries are using lipid-lowering
medications at varying rates. These changes can have distinct effects on the levels of
high-density lipoprotein (HDL) cholesterol and non-HDL cholesterol, which have
different effects on human health4,5. However, the trends of HDL and non-HDL
cholesterol levels over time have not been previously reported in a global analysis.
Here we pooled 1,127 population-based studies that measured blood lipids in
102.6 million individuals aged 18 years and older to estimate trends from 1980 to 2018
in mean total, non-HDL and HDL cholesterol levels for 200 countries. Globally, there
was little change in total or non-HDL cholesterol from 1980 to 2018. This was a net
effect of increases in low- and middle-income countries, especially in east and
southeast Asia, and decreases in high-income western countries, especially those in
northwestern Europe, and in central and eastern Europe. As a result, countries with
the highest level of non-HDL cholesterol—which is a marker of cardiovascular risk—
changed from those in western Europe such as Belgium, Finland, Greenland, Iceland,
Norway, Sweden, Switzerland and Malta in 1980 to those in Asia and the Pacific,
such as Tokelau, Malaysia, The Philippines and Thailand. In 2017, high non-HDL
cholesterol was responsible for an estimated 3.9 million (95% credible interval
3.7 million–4.2 million) worldwide deaths, half of which occurred in east, southeast
and south Asia. The global repositioning of lipid-related risk, with non-optimal
cholesterol shifting from a distinct feature of high-income countries in northwestern
Europe, north America and Australasia to one that affects countries in east and
southeast Asia and Oceania should motivate the use of population-based policies and
personal interventions to improve nutrition and enhance access to treatment
throughout the world.
Blood cholesterol is one of the most important risk factors for ischaemic
heart disease (IHD) and ischaemic stroke4–6. Consistent and comparable
information on cholesterol levels and trends in different countries can
help to benchmark national performance in addressing non-optimal
cholesterol, investigate the reasons behind differential trends and
identify countries in which interventions are needed the most.
A previous global analysis7 reported trends in total cholesterol
from 1980 to 2008, but did not analyse important lipid fractions—
including HDL and non-HDL cholesterol—that are key to understanding the cardiovascular disease risk associated with non-optimal
cholesterol. Dietary and behavioural determinants of cholesterol
have changed throughout the world in the past decades, including
a worldwide rise in adiposity8,9, divergent global trends in alcohol
use10, a rise in the intake of animal-source foods in middle-income
countries (especially in east Asia)3,11, and a replacement of saturated
fats and trans fats with unsaturated fats in some high-income countries3,11,12. There is also considerable variation in how much different
countries have adopted lipid-lowering medications13. These changes
are likely to have influenced cholesterol levels substantially in the
decade since the last estimates were made. Furthermore, HDL and
non-HDL cholesterol, which have opposite associations with cardiovascular diseases4,5, respond differently to diet and treatment, and
may therefore have different geographical patterns and trends over
time14. Information on these major lipid fractions, which were not
included in the previous global estimates, is essential for priority
setting and intervention choice.
Here we pooled 1,127 population-based studies that measured blood
lipids in 102.6 million individuals aged 18 years and older (Extended
Data Figs. 1, 2 and Supplementary Table 1) and used a Bayesian hierarchical model to estimate trends from 1980 to 2018 in mean total, non-HDL
and HDL cholesterol levels for 200 countries. We also estimated
the number of deaths caused by IHD and ischaemic stroke that were
attributable to high levels of non-HDL cholesterol using information
on its hazards from epidemiological studies.
*A list of participants and their affiliations appears in the online version of the paper.
Nature | Vol 582 | 4 June 2020 | 73
Article
a
b
Eastern Europe
Southwestern Europe
Central Europe
High-income Asia–Pacific
Northwestern Europe
Southeast Asia
Central Latin America
Caribbean
High-income English-speaking countries
Southern Latin America
Andean Latin America
East Asia
Central Asia
World
Melanesia
Middle East and north Africa
Polynesia and Micronesia
Southern Africa
South Asia
East Africa
Central Africa
West Africa
Eastern Europe
Central Europe
High-income Asia–Pacific
Southwestern Europe
Northwestern Europe
Southeast Asia
Southern Latin America
High-income English-speaking countries
Central Latin America
East Asia
Andean Latin America
Caribbean
World
Melanesia
Central Asia
Middle East and north Africa
Polynesia and Micronesia
South Asia
Southern Africa
Central Africa
East Africa
West Africa
3.8
4.2
4.6
5.0
5.4
5.8
Age-standardized mean total cholesterol (mmol l–1)
Central and eastern Europe
Central Asia, Middle East and north Africa
3.8
4.2
4.6
5.0
5.4
5.8
Age-standardized mean total cholesterol (mmol l–1)
East and southeast Asia
High-income Asia–Pacific
Oceania
South Asia
Sub-Saharan Africa
World
The start of the arrow shows the level in 1980 and the head indicates the level in
2018. See Extended Data Fig. 3 for age-standardized mean HDL cholesterol.
One mmol l−1 is equivalent to 38.61 mg dl−1.
Fig. 1 | Change in age-standardized mean total cholesterol between 1980
and 2018 by region for women and men. a, Age-standardized mean total
cholesterol in women. b, Age-standardized mean total cholesterol in men.
Trends in total cholesterol
−1
In 2018, global age-standardized mean total cholesterol was 4.6 mmol l
(95% credible interval, 4.5–4.7) for women and 4.5 mmol l−1 (4.3–4.6)
for men. Global age-standardized mean total cholesterol changed
little over these nearly four decades, decreasing by 0.03 mmol l−1 per
decade (−0.02–0.08) in women and 0.05 mmol l−1 per decade (0.00–
0.11) in men (posterior probability of the observed declines being true
declines = 0.90 for women and 0.98 for men) (Fig. 1). Regionally, total
cholesterol decreased the most in high-income western regions and
in central and eastern Europe. The decrease was the largest (around
0.3 mmol l−1 per decade; posterior probability >0.9999) in northwestern Europe, where mean total cholesterol levels had been the highest in 1980. The decrease in total cholesterol in high-income western
regions and central and eastern Europe was largely due to a decline
in non-HDL cholesterol (Extended Data Fig. 4), which among women
was offset partly by an increase in mean HDL cholesterol levels. Mean
total cholesterol changed little in most of the other regions, with the
notable exception of east and southeast Asia, where it increased by
more than 0.1 mmol l−1 per decade in both women and men (posterior
probability ≥0.95). The increase in east and southeast Asia was largely
due to an increase in non-HDL cholesterol.
Trends in non-HDL and HDL cholesterol
In 2018, global age-standardized mean non-HDL cholesterol was
3.3 mmol l−1 (3.2–3.4) for women and 3.3 mmol l−1 (3.3–3.4) for men;
global age-standardized mean HDL cholesterol was 1.3 mmol l−1 (1.2–1.3)
for women and 1.1 mmol l−1 (1.1–1.2) for men. Global age-standardized
mean non-HDL cholesterol remained almost unchanged from 1980 to
2018, decreasing by only 0.02 mmol l−1 per decade (−0.02–0.06; posterior probability = 0.80) in women and 0.01 mmol l−1 per decade (−0.03–
0.06; posterior probability = 0.72) in men. Global age-standardized
mean HDL cholesterol remained unchanged for women and decreased
slightly for men (by 0.02 mmol l−1 per decade, posterior probability = 0.91).
Regionally, non-HDL cholesterol decreased substantially in
high-income western regions and central and eastern Europe. The
largest decrease occurred in northwestern Europe (>0.3 mmol l−1
per decade; posterior probability >0.9999) (Fig. 2). By contrast, it
increased in east and southeast Asia, parts of sub-Saharan Africa and
Melanesia. The increase was the largest in southeast Asia, increasing by
74 | Nature | Vol 582 | 4 June 2020
High-income western countries
Latin America and Caribbean
approximately 0.2 mmol l−1 per decade (posterior probability >0.9999).
Mean HDL cholesterol increased in the high-income Asia–Pacific region,
by as much as 0.1 mmol l−1 per decade in women (posterior probability >0.9999) but decreased in Melanesia, Polynesia and Micronesia
(Extended Data Fig. 3).
Belgium, Finland, Greenland, Iceland, Norway, Sweden, Switzerland
and Malta had some of the highest non-HDL cholesterol levels in 1980
(>4.5 mmol l−1 in women and >4.7 mmol l−1 in men) but experienced
some of the largest declines (Figs. 3, 4). At the extreme, mean non-HDL
cholesterol declined by around 0.45 mmol l−1 per decade or more in
Belgian and Icelandic women and men, changing their ranks from being
in the top 10 countries in terms of non-HDL cholesterol in 1980 to being
ranked in the lower half of the countries in 2018—below countries in
southwestern Europe such as France and Italy. The largest increases
were found in east Asian countries (for example, China) and southeast
Asian countries (for example, Indonesia, Thailand, Malaysia, Cambodia
and Lao PDR). In these countries, age-standardized mean non-HDL
cholesterol increased by as much as 0.23 mmol l−1 per decade. As a result
of these opposite trends, countries with the highest age-standardized
mean non-HDL cholesterol levels in 2018 were all outside northwestern
Europe: Tokelau, Malaysia, The Philippines and Thailand, all of which
had mean non-HDL cholesterol around or above 4 mmol l−1. China, which
had one of the lowest mean non-HDL cholesterol levels in 1980, reached
or surpassed non-HDL cholesterol levels of many high-income western
countries in 2018. Sub-Saharan African countries had the lowest mean
non-HDL cholesterol in 2018, as low as 2.6 mmol l−1 in some countries,
as they had in 1980. Not only did high-income countries benefit from
decreasing non-HDL cholesterol levels, they had higher mean HDL cholesterol than low- and middle-income countries (Extended Data Fig. 6).
Deaths attributable to non-optimal cholesterol
In 2017, high non-HDL cholesterol was responsible for an estimated
3.9 million (3.7–4.2 million) worldwide deaths from IHD and ischaemic
stroke (Fig. 5), accounting for a third of deaths from these causes. From
1990 to 2017, the number of deaths caused by IHD and ischaemic stroke
that were attributable to high non-HDL cholesterol increased by around
910,000 globally. This increase was a net result of a large decrease in
western countries, from 950,000 (890,000–990,000) to 480,000
(430,000–530,000), and a large increase throughout Asia. In particular,
the number of deaths attributable to high non-HDL cholesterol more
a
b
Southeast Asia
Melanesia
Central Latin America
Polynesia and Micronesia
Caribbean
Eastern Europe
Andean Latin America
Southwestern Europe
Central Europe
Central Asia
World
Southern Latin America
East Asia
Northwestern Europe
Middle East and north Africa
High-income Asia–Pacific
High-income English-speaking countries
South Asia
Southern Africa
East Africa
West Africa
Central Africa
Central Europe
Southeast Asia
Eastern Europe
Central Latin America
Melanesia
Southwestern Europe
High-income Asia–Pacific
Andean Latin America
Polynesia and Micronesia
Southern Latin America
Caribbean
Northwestern Europe
East Asia
World
High-income English-speaking countries
Central Asia
Middle East and north Africa
South Asia
Southern Africa
East Africa
Central Africa
West Africa
2.2
2.6
3.0
3.4
3.8
4.2
4.6
2.2
Age-standardized mean non-HDL cholesterol (mmol l–1)
Central and eastern Europe
Central Asia, Middle East and north Africa
2.6
3.0
3.4
3.8
4.6
4.2
Age-standardized mean non-HDL cholesterol (mmol l–1)
East and southeast Asia
High-income Asia–Pacific
High-income western countries
Latin America and Caribbean
Oceania
South Asia
Sub-Saharan Africa
World
Fig. 2 | Change in age-standardized mean non-HDL cholesterol between
1980 and 2018 by region for women and men. a, Age-standardized mean
non-HDL cholesterol in women. b, Age-standardized mean non-HDL
cholesterol in men. The start of the arrow shows the level in 1980 and the head
indicates the level in 2018. See Extended Data Fig. 3 for age-standardized mean
HDL cholesterol. One mmol l−1 is equivalent to 38.61 mg dl−1.
than tripled in east Asia, from 250,000 (230,000–270,000) to 860,000
(770,000–940,000), and more than doubled in southeast Asia, from
110,000 (100,000–120,000) to 310,000 (290,000–330,000). As a
result, by 2017 east, southeast and south Asia accounted for half of
all deaths attributable to high non-HDL cholesterol, compared with
a quarter in 1990.
a
b
Caribbean
American Samoa
Bahrain
Bermuda
Brunei Darussalam
Cabo Verde
Comoros
Cook Islands
Fiji
Montenegro
French Polynesia
Nauru
Kiribati
Niue
Maldives
Palau
Marshall Islands
Samoa
Mauritius
Sao Tome and Principe
Federated States of Micronesia
Seychelles
Solomon Islands
Tokelau
Tonga
Tuvalu
Vanuatu
Caribbean
American Samoa
Bahrain
Bermuda
Brunei Darussalam
Cabo Verde
Comoros
Cook Islands
Fiji
Montenegro
Nauru
French Polynesia
Kiribati
Niue
Maldives
Palau
Marshall Islands
Samoa
Mauritius
Sao Tome and Principe
Federated States of Micronesia
Seychelles
Solomon Islands
Tokelau
Tonga
Tuvalu
Vanuatu
American Samoa
Bahrain
Bermuda
Brunei Darussalam
Cabo Verde
Comoros
Cook Islands
Montenegro
Fiji
Nauru
French Polynesia
Niue
Kiribati
Palau
Maldives
Samoa
Marshall Islands
Sao Tome and Principe
Mauritius
Federated States of Micronesia
Seychelles
Solomon Islands
Tokelau
Tonga
Tuvalu
Vanuatu
d
c
Caribbean
American Samoa
Bahrain
Bermuda
Brunei Darussalam
Cabo Verde
Comoros
Cook Islands
Montenegro
Fiji
Nauru
French Polynesia
Niue
Kiribati
Palau
Maldives
Samoa
Marshall Islands
Sao Tome and Principe
Mauritius
Federated States of Micronesia
Seychelles
Solomon Islands
Tokelau
Tonga
Tuvalu
Vanuatu
Caribbean
Age-standardized mean non-HDL cholesterol (mmol l–1)
2.0
2.5
3.0
Fig. 3 | Age-standardized mean non-HDL cholesterol by country in 1980 and
2018 for women and men. a, Age-standardized mean non-HDL cholesterol in
women in 1980. b, Age-standardized mean non-HDL cholesterol in women in
2018. c, Age-standardized mean non-HDL cholesterol in men in 1980.
3.5
4.0
4.5
5.0
d, Age-standardized mean non-HDL cholesterol in men in 2018. See Extended
Data Fig. 5 for age-standardized mean total cholesterol and Extended Data
Fig. 6 for age-standardized mean HDL cholesterol. One mmol l−1 is equivalent to
38.61 mg dl−1.
Nature | Vol 582 | 4 June 2020 | 75
Article
a
b
American Samoa
Bahrain
Bermuda
Brunei Darussalam
Cabo Verde
Comoros
Cook Islands
Caribbean
Fiji
Montenegro
French Polynesia
Nauru
Niue
Kiribati
Palau
Maldives
Marshall Islands
Samoa
Mauritius
Sao Tome and Principe
Federated States of Micronesia
Seychelles
Solomon Islands
Tokelau
Tonga
Tuvalu
Vanuatu
American Samoa
Bahrain
Bermuda
Brunei Darussalam
Cabo Verde
Comoros
Cook Islands
Caribbean
Fiji
Montenegro
French Polynesia
Nauru
Kiribati
Niue
Maldives
Palau
Marshall Islands
Samoa
Mauritius
Sao Tome and Principe
Federated States of Micronesia
Seychelles
Solomon Islands
Tokelau
Tonga
Tuvalu
Vanuatu
Change in age-standardized mean non-HDL cholesterol (mmol l–1 per decade)
−0.49 −0.40 −0.30 −0.20 −0.10
0.10 0.20
for change per decade in age-standardized mean total cholesterol and
Extended Data Fig. 8 for change per decade in age-standardized mean HDL
cholesterol. One mmol l−1 is equivalent to 38.61 mg dl−1.
Fig. 4 | Change in age-standardized mean non-HDL cholesterol per decade
by country for women and men. a, Change per decade in age-standardized
mean non-HDL cholesterol in women. b, Change per decade in
age-standardized mean non-HDL cholesterol in men. See Extended Data Fig. 7
men using data from high-income western countries and countries in
east and southeast Asia, the two regions that experienced the largest
decrease and increase, respectively, in non-HDL cholesterol levels.
Finally, changes in diet, especially a decrease in carbohydrate and an
increase in fat intake28–31, may have contributed to the large increase
in HDL cholesterol observed in the high-income Asia–Pacific region,
where there was little increase in overweight and obesity relative to
other regions8,9. By contrast, the large increase in diabetes32 and adiposity8 in Oceania may have contributed to the decrease in HDL cholesterol in this region. The Pearson correlation coefficient between
the change in HDL cholesterol and the change in body-mass index8 was
−0.87 for women and −0.69 for men using countries in the high-income
Asia–Pacific region and Oceania, the two regions that had the largest
increase and decrease, respectively, in HDL cholesterol; the Pearson
correlation coefficient for the change in HDL cholesterol and change
in diabetes prevalence32 was −0.84 for women and −0.69 for men. In the
same regions, the Pearson correlation coefficient between the change
in non-HDL cholesterol and the change in body-mass index8 was 0.77
for women and 0.62 for men; for the change in non-HDL cholesterol
and the change in diabetes prevalence32, the Pearson correlation coefficient was 0.54 for women and 0.40 for men.
Although it has previously been documented that the prevalence of
adiposity8,9, diabetes32 and high blood pressure33 is now higher in lowand middle-income countries than in high-income countries, higher
cholesterol is commonly considered to be a feature of affluent western nations1,2. We show that, when focusing on non-HDL cholesterol,
Implications
Our results show that over the past nearly four decades, there has been
a major global repositioning of lipid-related risk, with non-optimal cholesterol patterns shifting from being a distinct feature of high-income
countries in northwestern Europe, north America and Australasia to one
that affects middle-income countries in east and southeast Asia, as well
as some countries in Oceania and central Latin America. This transition
is especially noticeable for non-HDL cholesterol, which had not been
quantified previously in a global analysis. This global repositioning has
occurred as a consequence of opposing trends in high-income western
countries and in Asia, which has led to some Asian countries having the
highest worldwide non-HDL cholesterol levels in 2018.
The decrease in non-HDL cholesterol in western countries started
in the 1980s, before statins were widely used15,16. This indicates that
changes in diet, especially the replacement of saturated with unsaturated fats3,17–21 and reduction in trans fats12,17,22, are major contributors
to this decline. Nonetheless, the increased use of statins from the
late 1990s onwards15,16, may explain up to one half of the decrease in
those countries in which statins are widely used19,23,24. In contrast to
high-income western countries, the consumption of animal-source
foods, refined carbohydrates and palm oil has increased substantially
in east and southeast Asia3,25,26, where statin use remains low13,27. For
example, the Pearson correlation coefficient between the change in
non-HDL cholesterol and the change in a multi-dimensional score
of animal-source foods and sugar3 was 0.69 for women and 0.67 for
a
0
b
1990
1990
2017
2017
0
0.5
1.0
1.5
2.0
Number of attributable deaths (millions)
Eastern Europe
Central Europe
Southwestern Europe
Northwestern Europe
High-income English-speaking countries
Southern Latin America
Central Latin America
Andean Latin America
Caribbean
Polynesia and Micronesia
0
0.5
1.0
1.5
2.0
Number of attributable deaths (millions)
Melanesia
High-income Asia–Pacific
East Asia
Southeast Asia
South Asia
Central Asia
Middle East and north Africa
East Africa
Southern Africa
Central Africa
West Africa
Fig. 5 | Deaths from IHD and ischaemic stroke attributable to high non-HDL cholesterol by region in 1990 and 2017 for women and men. a, Deaths in women
attributable to high non-HDL cholesterol. b, Deaths in men attributable to high non-HDL cholesterol.
76 | Nature | Vol 582 | 4 June 2020
middle-income countries have emerged as the new global epicentre
of non-optimal cholesterol as they did for other major cardiovascular
disease risk factors, indicating that there is no such a thing as a western
risk factor. At the same time, the populations of high-income countries
would also benefit from further lowering non-HDL cholesterol. Therefore, population-based policies and personal interventions to improve
nutrition and enhance treatment are now needed in all countries, especially as a part of the movement towards universal health coverage.
15.
16.
17.
18.
19.
20.
Online content
Any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information,
acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code
availability are available at https://doi.org/10.1038/s41586-020-2338-1.
21.
1.
24.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Danaei, G. et al. The global cardiovascular risk transition: associations of four metabolic
risk factors with national income, urbanization, and Western diet in 1980 and 2008.
Circulation 127, 1493–1502 (2013).
Ezzati, M. et al. Rethinking the “diseases of affluence” paradigm: global patterns of
nutritional risks in relation to economic development. PLoS Med. 2, e133 (2005).
Bentham, J. et al. Multi-dimensional characterisation of global food supply from 1961 to
2013. Nat. Food 1, 70–75 (2020).
Prospective Studies Collaboration. Blood cholesterol and vascular mortality by age, sex,
and blood pressure: a meta-analysis of individual data from 61 prospective studies with
55,000 vascular deaths. Lancet 370, 1829–1839 (2007).
The Emerging Risk Factors Collaboration. Major lipids, apolipoproteins, and risk of
vascular disease. J. Am. Med. Assoc. 302, 1993–2000 (2009).
Global Burden of Metabolic Risk Factors for Chronic Diseases Collaboration.
Cardiovascular disease, chronic kidney disease, and diabetes mortality burden of
cardiometabolic risk factors from 1980 to 2010: a comparative risk assessment. Lancet
Diabetes Endocrinol. 2, 634–647 (2014).
Farzadfar, F. et al. National, regional, and global trends in serum total cholesterol since
1980: systematic analysis of health examination surveys and epidemiological studies
with 321 country-years and 3·0 million participants. Lancet 377, 578–586 (2011).
NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in body-mass index,
underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416
population-based measurement studies in 128·9 million children, adolescents, and
adults. Lancet 390, 2627–2642 (2017).
NCD Risk Factor Collaboration (NCD-RisC). Rising rural body-mass index is the main
driver of the global obesity epidemic in adults. Nature 569, 260–264 (2019).
Manthey, J. et al. Global alcohol exposure between 1990 and 2017 and forecasts until
2030: a modelling study. Lancet 393, 2493–2502 (2019).
Micha, R. et al. Global, regional and national consumption of major food groups in 1990
and 2010: a systematic analysis including 266 country-specific nutrition surveys
worldwide. BMJ Open 5, e008705 (2015).
Micha, R. et al. Global, regional, and national consumption levels of dietary fats and oils
in 1990 and 2010: a systematic analysis including 266 country-specific nutrition surveys.
Br. Med. J. 348, g2272 (2014).
Roth, G. A. et al. High total serum cholesterol, medication coverage and therapeutic
control: an analysis of national health examination survey data from eight countries. Bull.
World Health Organ. 89, 92–101 (2011).
NCD Risk Factor Collaboration (NCD-RisC). National trends in total cholesterol obscure
heterogeneous changes in HDL and non-HDL cholesterol and total-to-HDL cholesterol
ratio: a pooled analysis of 458 population-based studies in Asian and Western countries.
Int. J. Epidemiol. 49, 173–192 (2020).
22.
23.
25.
26.
27.
28.
29.
30.
31.
32.
33.
Carroll, M. D., Kit, B. K., Lacher, D. A., Shero, S. T. & Mussolino, M. E. Trends in lipids and
lipoproteins in US adults, 1988–2010. J. Am. Med. Assoc. 308, 1545–1554 (2012).
Walley, T., Folino-Gallo, P., Stephens, P. & Van Ganse, E. Trends in prescribing and
utilization of statins and other lipid lowering drugs across Europe 1997–2003. Br. J. Clin.
Pharmacol. 60, 543–551 (2005).
Vartiainen, E., Laatikainen, T., Tapanainen, H. & Puska, P. Changes in serum cholesterol
and diet in North Karelia and all Finland. Glob. Heart 11, 179–184 (2016).
Miller, J. C. et al. Trends in serum total cholesterol and dietary fat intakes in New Zealand
between 1989 and 2009. Aust. N. Z. J. Public Health 40, 263–269 (2016).
Eriksson, M. et al. Greater decreases in cholesterol levels among individuals with high
cardiovascular risk than among the general population: the northern Sweden MONICA
study 1994 to 2014. Eur. Heart J. 37, 1985–1992 (2016).
Arnett, D. K. et al. Twenty-year trends in serum cholesterol, hypercholesterolemia, and
cholesterol medication use: the Minnesota Heart Survey, 1980–1982 to 2000–2002.
Circulation 112, 3884–3891 (2005).
Houterman, S., Verschuren, W. M., Oomen, C. M., Boersma-Cobbaert, C. M. & Kromhout,
D. Trends in total and high density lipoprotein cholesterol and their determinants in The
Netherlands between 1993 and 1997. Int. J. Epidemiol. 30, 1063–1070 (2001).
Leth, T., Jensen, H. G., Mikkelsen, A. A. & Bysted, A. The effect of the regulation on trans
fatty acid content in Danish food. Atheroscler. Suppl. 7, 53–56 (2006).
Kypridemos, C. et al. Quantifying the contribution of statins to the decline in population
mean cholesterol by socioeconomic group in England 1991 - 2012: a modelling study.
PLoS ONE 10, e0123112 (2015).
Ford, E. S. & Capewell, S. Trends in total and low-density lipoprotein cholesterol among
U.S. adults: contributions of changes in dietary fat intake and use of cholesterol-lowering
medications. PLoS ONE 8, e65228 (2013).
He, Y. et al. The dietary transition and its association with cardiometabolic mortality
among Chinese adults, 1982–2012: a cross-sectional population-based study. Lancet
Diabetes Endocrinol. 7, 540–548 (2019).
Wolmarans, P. Background paper on global trends in food production, intake and
composition. Ann. Nutr. Metab. 55, 244–272 (2009).
Zhang, M. et al. Prevalence of dyslipidemia and achievement of low-density lipoprotein
cholesterol targets in Chinese adults: a nationally representative survey of 163,641 adults.
Int. J. Cardiol. 260, 196–203 (2018).
Song, Y. & Joung, H. A traditional Korean dietary pattern and metabolic syndrome
abnormalities. Nutr. Metab. Cardiovasc. Dis. 22, 456–462 (2012).
Yoshiike, N., Matsumura, Y., Iwaya, M., Sugiyama, M. & Yamaguchi, M. National Nutrition
Survey in Japan. J. Epidemiol. 6, S189–S200 (1996).
Matsumura, Y. Nutrition trends in Japan. Asia Pac. J. Clin. Nutr. 10, S40–S47 (2001).
Kim, S., Moon, S. & Popkin, B. M. The nutrition transition in South Korea. Am. J. Clin. Nutr.
71, 44–53 (2000).
NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in diabetes since 1980: a
pooled analysis of 751 population-based studies with 4·4 million participants. Lancet 387,
1513–1530 (2016).
NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in blood pressure from 1975
to 2015: a pooled analysis of 1479 population-based measurement studies with
19·1 million participants. Lancet 389, 37–55 (2017).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution
4.0 International License, which permits use, sharing, adaptation, distribution
and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made. The images or other third party material in this article are
included in the article’s Creative Commons license, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons license and your
intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this license,
visit http://creativecommons.org/licenses/by/4.0/.
© The Author(s), under exclusive licence to Springer Nature Limited 2020
Nature | Vol 582 | 4 June 2020 | 77
Article
NCD Risk Factor Collaboration (NCD-RisC)
Cristina Taddei1, Bin Zhou1, Honor Bixby1, Rodrigo M. Carrillo-Larco1, Goodarz Danaei2, Rod T.
Jackson3, Farshad Farzadfar4, Marisa K. Sophiea1, Mariachiara Di Cesare5, Maria Laura
Caminia Iurilli1, Andrea Rodriguez Martinez1, Golaleh Asghari6, Klodian Dhana7, Pablo
Gulayin8, Sujay Kakarmath9, Marilina Santero8, Trudy Voortman10, Leanne M. Riley11, Melanie
J. Cowan11, Stefan Savin11, James E. Bennett1, Gretchen A. Stevens1,12, Christopher J.
Paciorek13, Wichai Aekplakorn14, Renata Cifkova15,16, Simona Giampaoli17, Andre Pascal
Kengne18, Young-Ho Khang19, Kari Kuulasmaa20, Avula Laxmaiah21, Paula Margozzini22,
Prashant Mathur23, Børge G. Nordestgaard24, Dong Zhao25, Mette Aadahl26, Leandra
Abarca-Gómez27, Hanan Abdul Rahim28, Niveen M. Abu-Rmeileh29, Benjamin
Acosta-Cazares30, Robert J. Adams31, Imelda A. Agdeppa32, Javad Aghazadeh-Attari33,
Carlos A. Aguilar-Salinas34, Charles Agyemang35, Tarunveer S. Ahluwalia36, Noor Ani
Ahmad37, Ali Ahmadi38, Naser Ahmadi4, Soheir H. Ahmed39, Wolfgang Ahrens40, Kamel
Ajlouni41, Monira Alarouj42, Fadia AlBuhairan43, Shahla AlDhukair44, Mohamed M. Ali11,
Abdullah Alkandari42, Ala’a Alkerwi45, Eman Aly46, Deepak N. Amarapurkar47, Philippe
Amouyel48,49, Lars Bo Andersen50, Sigmund A. Anderssen51, Ranjit Mohan Anjana52, Alireza
Ansari-Moghaddam53, Hajer Aounallah-Skhiri54, Joana Araújo55, Inger Ariansen56, Tahir
Aris37, Raphael E. Arku57, Nimmathota Arlappa21, Krishna K. Aryal58, Thor Aspelund59, Maria
Cecília F. Assunção60, Juha Auvinen61,62, Mária Avdicová63, Ana Azevedo64, Fereidoun Azizi65,
Mehrdad Azmin4, Nagalla Balakrishna21, Mohamed Bamoshmoosh66, Maciej Banach67, Piotr
Bandosz68, José R. Banegas69, Carlo M. Barbagallo70, Alberto Barceló71, Amina Barkat72, Iqbal
Bata73, Anwar M. Batieha74, Assembekov Batyrbek75, Louise A. Baur76, Robert Beaglehole3,
Antonisamy Belavendra77, Habiba Ben Romdhane78, Mikhail Benet79, Marianne Benn24, Salim
Berkinbayev80, Antonio Bernabe-Ortiz81, Gailute Bernotiene82, Heloisa Bettiol83, Santosh K.
Bhargava84, Yufang Bi85, Asako Bienek86, Mukharram Bikbov87, Bihungum Bista88, Peter
Bjerregaard89, Espen Bjertness39, Marius B. Bjertness39, Cecilia Björkelund90, Katia V. Bloch91,
Anneke Blokstra92, Simona Bo93, Bernhard O. Boehm94, Jose G. Boggia95, Carlos P.
Boissonnet96, Marialaura Bonaccio97, Vanina Bongard98, Rossana Borchini99, Herman
Borghs100, Pascal Bovet101,102, Imperia Brajkovich103, Juergen Breckenkamp104, Hermann
Brenner105, Lizzy M. Brewster35, Graziella Bruno93, Anna Bugge106, Markus A. Busch107,
Antonio Cabrera de León108, Joseph Cacciottolo109, Günay Can110, Ana Paula C. Cândido111,
Mario V. Capanzana32, Eduardo Capuano112, Vincenzo Capuano112, Viviane C. Cardoso83,
Joana Carvalho113, Felipe F. Casanueva114, Laura Censi115, Charalambos A. Chadjigeorgiou116,
Snehalatha Chamukuttan117, Nish Chaturvedi118, Chien-Jen Chen119, Fangfang Chen120,
Shuohua Chen121, Ching-Yu Cheng122, Bahman Cheraghian123, Angela Chetrit124, Shu-Ti
Chiou125, María-Dolores Chirlaque126, Belong Cho127, Yumi Cho128, Jerzy Chudek129, Frank
Claessens130, Janine Clarke131, Els Clays132, Hans Concin133, Susana C. Confortin134, Cyrus
Cooper135, Simona Costanzo97, Dominique Cottel136, Chris Cowell76, Ana B. Crujeiras137,
Semánová Csilla138, Liufu Cui121, Felipe V. Cureau139, Graziella D’Arrigo140, Eleonora d’Orsi141,
Jean Dallongeville136, Albertino Damasceno142, Rachel Dankner124, Thomas M. Dantoft26, Luc
Dauchet48,49, Kairat Davletov75, Guy De Backer132, Dirk De Bacquer132, Giovanni de Gaetano97,
Stefaan De Henauw132, Paula Duarte de Oliveira60, David De Ridder143, Delphine De Smedt132,
Mohan Deepa52, Alexander D. Deev144, Abbas Dehghan1, Hélène Delisle145, Elaine
Dennison135, Valérie Deschamps146, Meghnath Dhimal88, Augusto F. Di Castelnuovo147, Zivka
Dika148, Shirin Djalalinia149, Annette J. Dobson150, Chiara Donfrancesco17, Silvana P. Donoso151,
Angela Döring152, Maria Dorobantu153, Nico Dragano154, Wojciech Drygas67,155, Yong Du107,
Charmaine A. Duante32, Rosemary B. Duda156, Vilnis Dzerve157, Elzbieta
Dziankowska-Zaborszczyk67, Ricky Eddie158, Ebrahim Eftekhar159, Robert Eggertsen90, Sareh
Eghtesad4, Gabriele Eiben160, Ulf Ekelund51, Jalila El Ati161, Denise Eldemire-Shearer162, Marie
Eliasen26, Roberto Elosua163, Rajiv T. Erasmus164, Raimund Erbel165, Cihangir Erem166, Louise
Eriksen89, Johan G. Eriksson167, Jorge Escobedo-de la Peña30, Saeid Eslami168, Ali Esmaeili169,
Alun Evans170, David Faeh171, Caroline H. Fall135, Elnaz Faramarzi172, Mojtaba Farjam173,
Mohammad Reza Fattahi174, Francisco J. Felix-Redondo175, Trevor S. Ferguson162, Daniel
Fernández-Bergés176, Daniel Ferrante177, Marika Ferrari115, Catterina Ferreccio22, Jean
Ferrieres98, Bernhard Föger133, Leng Huat Foo178, Ann-Sofie Forslund179, Maria Forsner179,
Heba M. Fouad46, Damian K. Francis162, Maria do Carmo Franco180, Oscar H. Franco10,
Guillermo Frontera181, Yuki Fujita182, Matsuda Fumihiko183, Takuro Furusawa183, Zbigniew
Gaciong184, Fabio Galvano185, Jingli Gao121, Manoli Garcia-de-la-Hera186, Sarah P. Garnett76,
Jean-Michel Gaspoz143, Magda Gasull187, Andrea Gazzinelli188, Johanna M. Geleijnse189, Ali
Ghanbari4, Erfan Ghasemi4, Oana-Florentina Gheorghe-Fronea153, Anup Ghimire190,
Francesco Gianfagna147,191, Tiffany K. Gill192, Jonathan Giovannelli48,49, Glen Gironella32,
Aleksander Giwercman193, David Goltzman194, Helen Gonçalves60, David A.
Gonzalez-Chica192, Marcela Gonzalez-Gross195, Juan P. González-Rivas196, Clicerio
González-Villalpando197, María-Elena González-Villalpando198, Angel R. Gonzalez199,
Frederic Gottrand48, Sidsel Graff-Iversen56, Dušan Grafnetter200, Ronald D. Gregor73, Tomasz
Grodzicki201, Anders Grøntved202, Giuseppe Grosso185, Gabriella Gruden93, Dongfeng Gu203,
Pilar Guallar-Castillón69, Ong Peng Guan204, Elias F. Gudmundsson205, Vilmundur
Gudnason59, Ramiro Guerrero206, Idris Guessous143, Johanna Gunnlaugsdottir205, Rajeev
Gupta207, Laura Gutierrez8, Felix Gutzwiller171, Seongjun Ha208, Farzad Hadaegh209, Rosa
Haghshenas4, Hamid Hakimi169, Ian R. Hambleton210, Behrooz Hamzeh211, Sari Hantunen212,
Rachakulla Hari Kumar21, Seyed Mohammad Hashemi-Shahri53, Jun Hata213, Teresa
Haugsgjerd214, Alison J. Hayes76, Jiang He215, Yuna He216, Marleen Elisabeth Hendriks217, Ana
Henriques55, Sauli Herrala62, Ramin Heshmat218, Allan G. Hill135, Sai Yin Ho219, Suzanne C.
Ho220, Michael Hobbs221, Albert Hofman10, Reza Homayounfar173, Wilma M. Hopman222,
Andrea R. V. R. Horimoto223, Claudia M. Hormiga224, Bernardo L. Horta60, Leila Houti225,
Christina Howitt210, Thein Thein Htay226, Aung Soe Htet227, Maung Maung Than Htike227,
José María Huerta228, Ilpo Tapani Huhtaniemi1, Martijn Huisman229, Monica L. Hunsberger90,
Abdullatif S. Husseini29, Inge Huybrechts230, Nahla Hwalla231, Licia Iacoviello97,191, Anna G.
Iannone112, Mohsen M. Ibrahim232, Norazizah Ibrahim Wong37, Iris Iglesia233, Nayu Ikeda234, M.
Arfan Ikram10, Violeta Iotova235, Vilma E. Irazola8, Takafumi Ishida236, Muhammad Islam237,
Aziz al-Safi Ismail178, Masanori Iwasaki238, Jeremy M. Jacobs239, Hashem Y. Jaddou74, Tazeen
Jafar122, Kenneth James162, Konrad Jamrozik192,448, Imre Janszky240, Edward Janus241,
Marjo-Riitta Jarvelin1,61,62, Grazyna Jasienska201, Ana Jelakovic242, Bojan Jelakovic243, Garry
Jennings244, Gorm B. Jensen24, Seung-lyeal Jeong208, Anjani Kumar Jha88, Chao Qiang
Jiang245, Ramon O. Jimenez246, Karl-Heinz Jöckel165, Michel Joffres247, Jari J. Jokelainen62, Jost
B. Jonas248, Torben Jørgensen26, Pradeep Joshi249, Farahnaz Joukar250, Jacek Józwiak251, Anne
Juolevi20, Anthony Kafatos252, Eero O. Kajantie20, Ofra Kalter-Leibovici124, Nor Azmi
Kamaruddin253, Pia R. Kamstrup24, Khem B. Karki254, Joanne Katz255, Jussi Kauhanen212,
Prabhdeep Kaur256, Maryam Kavousi10, Gyulli Kazakbaeva87, Ulrich Keil257, Sirkka
Keinänen-Kiukaanniemi62, Roya Kelishadi258, Maryam Keramati168, Alina Kerimkulova259,
Mathilde Kersting260, Yousef Saleh Khader74, Davood Khalili6, Mohammad Khateeb41,
Motahareh Kheradmand261, Alireza Khosravi262, Ursula Kiechl-Kohlendorfer263, Stefan
Kiechl263, Japhet Killewo264, Hyeon Chang Kim265, Jeongseon Kim266, Yeon-Yong Kim208,
Jurate Klumbiene82, Michael Knoflach263, Stephanie Ko86, Hans-Peter Kohler267, Iliana V.
Kohler267, Elin Kolle51, Patrick Kolsteren132, Jürgen König268, Raija Korpelainen61,269, Paul
Korrovits270, Jelena Kos242, Seppo Koskinen20, Katsuyasu Kouda271, Sudhir Kowlessur272,
Wolfgang Kratzer273, Susi Kriemler171, Peter Lund Kristensen202, Steiner Krokstad240, Daan
Kromhout274, Urho M. Kujala275, Pawel Kurjata155, Catherine Kyobutungi276, Fatima Zahra
Laamiri277, Tiina Laatikainen20, Carl Lachat132, Youcef Laid278, Tai Hing Lam219,
Christina-Paulina Lambrinou279, Vera Lanska200, Georg Lappas280, Bagher Larijani281, Tint Swe
Latt282, Lars E. Laugsand240, Maria Lazo-Porras81, Jeannette Lee283, Jeonghee Lee266, Nils
Lehmann165, Terho Lehtimäki284,285, Naomi S. Levitt286, Yanping Li2, Christa L. Lilly287, Wei-Yen
Lim283, M. Fernanda Lima-Costa288, Hsien-Ho Lin289, Xu Lin290, Yi-Ting Lin291, Lars Lind291, Allan
Linneberg26, Lauren Lissner90, Jing Liu25, Helle-Mai Loit292, Esther Lopez-Garcia69, Tania
Lopez293, Paulo A. Lotufo83, José Eugenio Lozano294, Dalia Luksiene82, Annamari Lundqvist20,
Robert Lundqvist295, Nuno Lunet113, Guansheng Ma296, George L. L. Machado-Coelho297,
Aristides M. Machado-Rodrigues298, Suka Machi299, Ahmed A. Madar39, Stefania Maggi300,
Dianna J. Magliano301, Emmanuella Magriplis302, Gowri Mahasampath77, Bernard Maire303,
Marcia Makdisse304, Fatemeh Malekzadeh174, Reza Malekzadeh4, Kodavanti Mallikharjuna
Rao21, Yannis Manios279, Jim I. Mann305, Fariborz Mansour-Ghanaei250, Enzo Manzato306,
Pedro Marques-Vidal307, Reynaldo Martorell308, Luis P. Mascarenhas309, Ellisiv B.
Mathiesen310, Tandi E. Matsha311, Christina Mavrogianni279, Shelly R. McFarlane162, Stephen T.
McGarvey312, Stela McLachlan313, Rachael M. McLean305, Scott B. McLean131, Breige A.
McNulty314, Sounnia Mediene-Benchekor225, Parinaz Mehdipour4, Kirsten Mehlig90, Amir
Houshang Mehrparvar315, Aline Meirhaeghe316, Christa Meisinger152, Ana Maria B. Menezes60,
Geetha R. Menon317, Shahin Merat4, Alibek Mereke75, Indrapal I. Meshram21, Patricia Metcalf3,
Haakon E. Meyer39, Jie Mi120, Nathalie Michels132, Jody C. Miller305, Cláudia S. Minderico318, G.
K. Mini319, Juan Francisco Miquel22, J. Jaime Miranda81, Mohammad Reza Mirjalili315, Erkin
Mirrakhimov259, Pietro A. Modesti320, Sahar Saeedi Moghaddam4, Bahram Mohajer4, Mostafa
K. Mohamed321, Kazem Mohammad4, Zahra Mohammadi4, Noushin Mohammadifard322, Reza
Mohammadpourhodki168, Viswanathan Mohan52, Salim Mohanna81, Muhammad Fadhli Mohd
Yusoff37, Iraj Mohebbi33, Farnam Mohebi4, Marie Moitry323,324, Line T. Møllehave26, Niels C.
Møller202, Dénes Molnár325, Amirabbas Momenan6, Charles K. Mondo326, Eric
Monterrubio-Flores197, Mahmood Moosazadeh261, Alain Morejon327, Luis A. Moreno233, Karen
Morgan328, Suzanne N. Morin194, George Moschonis329, Malgorzata Mossakowska330, Aya
Mostafa321, Jorge Mota113, Mohammad Esmaeel Motlagh123, Jorge Motta331, Kelias P.
Msyamboza332, Maria L. Muiesan333, Martina Müller-Nurasyid152, Jaakko Mursu212, Norlaila
Mustafa253, Iraj Nabipour334, Shohreh Naderimagham4, Gabriele Nagel335, Balkish M. Naidu37,
Farid Najafi211, Harunobu Nakamura336, Jana Námešná63, Ei Ei K. Nang283, Vinay B. Nangia337,
Matthias Nauck338, William A. Neal287, Azim Nejatizadeh159, Ilona Nenko201, Flavio Nervi22,
Nguyen D. Nguyen339, Quang Ngoc Nguyen340, Ramfis E. Nieto-Martínez341, Thomas Nihal77,
Teemu J. Niiranen20,342, Guang Ning85, Toshiharu Ninomiya213, Marianna Noale300, Oscar A.
Noboa95, Davide Noto70, Mohannad Al Nsour343, Irfan Nuhoğlu166, Terence W. O’Neill344,
Dermot O’Reilly170, Angélica M. Ochoa-Avilés151, Kyungwon Oh128, Ryutaro Ohtsuka345, Örn
Olafsson205, Valérie Olié146, Isabel O. Oliveira60, Mohd Azahadi Omar37, Altan Onat346,448, Sok
King Ong347, Pedro Ordunez71, Rui Ornelas348, Pedro J. Ortiz81, Clive Osmond349, Sergej M.
Ostojic350, Afshin Ostovar4, Johanna A. Otero224, Ellis Owusu-Dabo351, Fred Michel
Paccaud352, Elena Pahomova157, Andrzej Pajak201, Luigi Palmieri17, Wen-Harn Pan119,
Songhomitra Panda-Jonas248, Francesco Panza353, Winsome R. Parnell305, Nikhil D. Patel354,
Nasheeta Peer355, Sergio Viana Peixoto288, Markku Peltonen20, Alexandre C. Pereira223,
Annette Peters152, Astrid Petersmann338, Janina Petkeviciene82, Niloofar Peykari149, Son Thai
Pham356, Rafael N. Pichardo357, Iris Pigeot358, Aida Pilav359, Lorenza Pilotto360, Aleksandra
Piwonska155, Andreia N. Pizarro113, Pedro Plans-Rubió361, Silvia Plata362, Hermann
Pohlabeln358, Miquel Porta163, Marileen L. P. Portegies10, Anil Poudyal88, Farhad Pourfarzi363,
Hossein Poustchi4, Rajendra Pradeepa52, Jacqueline F. Price313, Rui Providencia118, Jardena J.
Puder307, Soile E. Puhakka61,269, Margus Punab270, Mostafa Qorbani364, Tran Quoc Bao365,
Ricardas Radisauskas82, Salar Rahimikazerooni174, Olli Raitakari342, Sudha Ramachandra
Rao256, Ambady Ramachandran366, Elisabete Ramos64, Rafel Ramos367, Lekhraj Rampal368,
Sanjay Rampal369, Josep Redon370, Paul Ferdinand M. Reganit371, Luis Revilla293, Abbas
Rezaianzadeh174, Robespierre Ribeiro372,448, Adrian Richter338, Fernando Rigo373, Tobias F.
Rinke de Wit374, Fernando Rodríguez-Artalejo69, María del Cristo Rodriguez-Perez375, Laura A.
Rodríguez-Villamizar376, Ulla Roggenbuck165, Rosalba Rojas-Martinez197, Dora Romaguera137,
Elisabetta L. Romeo377, Annika Rosengren90,378, Joel G. R. Roy131, Adolfo Rubinstein8,
Jean-Bernard Ruidavets379, Blanca Sandra Ruiz-Betancourt30, Paola Russo380, Petra Rust268,
Marcin Rutkowski68, Charumathi Sabanayagam204, Harshpal S. Sachdev381, Alireza Sadjadi4,
Ali Reza Safarpour174, Saeid Safiri172, Olfa Saidi382, Nader Saki123, Benoit Salanave146, Diego
Salmerón228, Veikko Salomaa20, Jukka T. Salonen167, Massimo Salvetti333, Jose
Sánchez-Abanto383, Susana Sans384, Alba M. Santaliestra-Pasías233, Diana A. Santos385, Maria
Paula Santos113, Rute Santos113, Jouko L. Saramies386, Luis B. Sardinha385, Nizal
Sarrafzadegan387, Kai-Uwe Saum105, Savvas C. Savva116, Norie Sawada388, Mariana Sbaraini139,
Marcia Scazufca389, Beatriz D. Schaan139, Herman Schargrodsky390, Christa Scheidt-Nave107,
Anja Schienkiewitz107, Sabine Schipf338, Carsten O. Schmidt338, Ben Schöttker105, Sara
Schramm165, Sylvain Sebert61, Aye Aye Sein227, Abhijit Sen391, Sadaf G. Sepanlou4, Jennifer
Servais131, Ramin Shakeri4, Svetlana A. Shalnova144, Teresa Shamah-Levy197, Maryam
Sharafkhah4, Sanjib K. Sharma190, Jonathan E. Shaw301, Amaneh Shayanrad4, Zumin Shi28,
Kenji Shibuya392, Hana Shimizu-Furusawa393, Dong Wook Shin394, Youchan Shin204, Majid
Shirani38, Rahman Shiri395, Namuna Shrestha88, Khairil Si-Ramlee347, Alfonso Siani380,
Rosalynn Siantar204, Abla M. Sibai231, Diego Augusto Santos Silva141, Mary Simon366, Judith
Simons396, Leon A. Simons397, Michael Sjöström398, Tea Skaaby399, Jolanta
Slowikowska-Hilczer67, Przemyslaw Slusarczyk330, Liam Smeeth400, Marieke B. Snijder35,
Stefan Söderberg179, Agustinus Soemantri401, Reecha Sofat118, Vincenzo Solfrizzi402,
Mohammad Hossein Somi172, Emily Sonestedt193, Thorkild I. A. Sørensen403, Charles Sossa
Jérome404, Aïcha Soumaré405, Kaan Sozmen406, Karen Sparrenberger139, Jan A. Staessen407,
Maria G. Stathopoulou408, Bill Stavreski244, Jostein Steene-Johannessen51, Peter Stehle409,
Aryeh D. Stein308, Jochanan Stessman239, Ranko Stevanović410, Jutta Stieber152,448, Doris
Stöckl152, Jakub Stokwiszewski411, Karien Stronks35, Maria Wany Strufaldi180, Ramón
Suárez-Medina412, Chien-An Sun413, Johan Sundström291, Paibul Suriyawongpaisal14, Rody G.
Sy371, René Charles Sylva414, Moyses Szklo255, E. Shyong Tai283, Abdonas Tamosiunas82, Eng
Joo Tan76, Mohammed Rasoul Tarawneh415, Carolina B. Tarqui-Mamani383, Anne Taylor192,
Julie Taylor118, Grethe S. Tell214, Tania Tello81, K. R. Thankappan416, Lutgarde Thijs407, Betina H.
Thuesen26, Ulla Toft26, Hanna K. Tolonen20, Janne S. Tolstrup89, Murat Topbas166, Roman
Topór-Madry201, María José Tormo417, Michael J. Tornaritis116, Maties Torrent418, Laura
Torres-Collado186, Pierre Traissac303, Oanh T. H. Trinh339, Julia Truthmann107, Shoichiro
Tsugane388, Marshall K. Tulloch-Reid162, Tomi-Pekka Tuomainen212, Jaakko Tuomilehto20,
Anne Tybjaerg-Hansen24, Christophe Tzourio405, Peter Ueda398, Eunice Ugel419, Hanno
Ulmer263, Belgin Unal420, Hannu M. T. Uusitalo421, Gonzalo Valdivia22, Damaskini Valvi422, Rob
M. van Dam283, Yvonne T. van der Schouw423, Koen Van Herck132, Hoang Van Minh424, Lenie
van Rossem425, Natasja M. Van Schoor229, Irene G. M. van Valkengoed35, Dirk
Vanderschueren130, Diego Vanuzzo360, Anette Varbo24, Patricia Varona-Pérez412, Senthil K.
Vasan135, Lars Vatten240, Tomas Vega294, Toomas Veidebaum292, Gustavo
Velasquez-Melendez188, Silvia J. Venero-Fernández412, Giovanni Veronesi191, W. M. Monique
Verschuren92, Cesar G. Victora60, Dhanasari Vidiawati426, Lucie Viet92, Salvador
Villalpando197, Jesus Vioque427, Jyrki K. Virtanen212, Sophie Visvikis-Siest408, Bharathi
Viswanathan101, Tiina Vlasoff428, Peter Vollenweider307, Ari Voutilainen212, Alisha N. Wade429,
Aline Wagner323, Janette Walton430, Wan Mohamad Wan Bebakar178, Wan Nazaimoon Wan
Mohamud431, Ming-Dong Wang432, Ningli Wang433, Qian Wang434, Ya Xing Wang435, Ying-Wei
Wang125, S. Goya Wannamethee118, Niels Wedderkopp202, Wenbin Wei435, Peter H.
Whincup436, Kurt Widhalm437, Indah S. Widyahening426, Andrzej Wiecek129, Alet H. Wijga92,
Rainford J. Wilks162, Johann Willeit263, Peter Willeit263, Tom Wilsgaard310, Bogdan Wojtyniak411,
Roy A. Wong-McClure27, Andrew Wong118, Tien Yin Wong122, Jean Woo220, Mark
Woodward397,438, Frederick C. Wu344, Shouling Wu121, Haiquan Xu439, Liang Xu433, Weili Yan440,
Xiaoguang Yang216, Tabara Yasuharu183, Xingwang Ye290, Toh Peng Yeow441, Panayiotis K.
Yiallouros442, Moein Yoosefi4, Akihiro Yoshihara238, San-Lin You413, Novie O.
Younger-Coleman162, Ahmad Faudzi Yusoff37, Ahmad A. Zainuddin37, Seyed Rasoul Zakavi168,
Mohammad Reza Zali6, Farhad Zamani443, Sabina Zambon306, Antonis Zampelas302, Ko Ko
Zaw282, Tomasz Zdrojewski68, Tajana Zeljkovic Vrkic242, Zhen-Yu Zhang407, Wenhua Zhao216,
Shiqi Zhen444, Yingfeng Zheng445, Bekbolat Zholdin446, Baurzhan Zhussupov80, Nada
Zoghlami54, Julio Zuñiga Cisneros331, Edward W. Gregg1 & Majid Ezzati1,447 ✉
Mexico. 35University of Amsterdam, Amsterdam, The Netherlands. 36Steno Diabetes Center
Copenhagen, Gentofte, Denmark. 37Ministry of Health Malaysia, Kuala Lumpur, Malaysia.
38
Shahrekord University of Medical Sciences, Shahrekord, Iran. 39University of Oslo, Oslo,
Norway. 40University of Bremen, Bremen, Germany. 41National Center for Diabetes,
Endocrinology and Genetics, Amman, Jordan. 42Dasman Diabetes Institute, Kuwait City,
Kuwait. 43Aldara Hospital and Medical Center, Riyadh, Saudi Arabia. 44King Abdullah
International Medical Research Center, Riyadh, Saudi Arabia. 45Luxembourg Institute of
Health, Strassen, Luxembourg. 46World Health Organization Regional Office for the Eastern
Mediterranean, Cairo, Egypt. 47Bombay Hospital and Medical Research Centre, Mumbai, India.
48
University of Lille, Lille, France. 49Lille University Hospital, Lille, France. 50Western Norway
University of Applied Sciences, Sogndal, Norway. 51Norwegian School of Sport Sciences,
Oslo, Norway. 52Madras Diabetes Research Foundation, Chennai, India. 53Zahedan University
of Medical Sciences, Zahedan, Iran. 54National Institute of Public Health, Tunis, Tunisia.
55
Institute of Public Health of the University of Porto, Porto, Portugal. 56Norwegian Institute of
Public Health, Oslo, Norway. 57University of Massachusetts, Amherst, MA, USA. 58Abt
Associates, Kathmandu, Nepal. 59University of Iceland, Reykjavik, Iceland. 60Federal University
of Pelotas, Pelotas, Brazil. 61University of Oulu, Oulu, Finland. 62Oulu University Hospital, Oulu,
Finland. 63Regional Authority of Public Health, Banska Bystrica, Slovakia. 64University of Porto
Medical School, Porto, Portugal. 65Research Institute for Endocrine Sciences, Tehran, Iran.
66
University of Science and Technology, Sana’a, Yemen. 67Medical University of Lodz, Lodz,
Poland. 68Medical University of Gdansk, Gdansk, Poland. 69Universidad Autónoma de Madrid/
CIBERESP, Madrid, Spain. 70University of Palermo, Palermo, Italy. 71Pan American Health
Organization, Washington, DC, USA. 72Mohammed V University de Rabat, Rabat, Morocco.
73
Dalhousie University, Halifax, Nova Scotia, Canada. 74Jordan University of Science and
Technology, Irbid, Jordan. 75Al-Farabi Kazakh National University, Almaty, Kazakhstan.
76
University of Sydney, Sydney, New South Wales, Australia. 77Christian Medical College,
Vellore, India. 78University Tunis El Manar, Tunis, Tunisia. 79Cafam University Foundation,
Bogota, Colombia. 80Kazakh National Medical University, Almaty, Kazakhstan. 81Universidad
Peruana Cayetano Heredia, Lima, Peru. 82Lithuanian University of Health Sciences, Kaunas,
Lithuania. 83University of São Paulo, São Paulo, Brazil. 84Sunder Lal Jain Hospital, Delhi, India.
85
Shanghai Jiao-Tong University School of Medicine, Shanghai, China. 86Public Health Agency
of Canada, Ottawa, Ontario, Canada. 87Ufa Eye Research Institute, Ufa, Russia. 88Nepal Health
Research Council, Kathmandu, Nepal. 89University of Southern Denmark, Copenhagen,
Denmark. 90University of Gothenburg, Gothenburg, Sweden. 91Universidade Federal do Rio de
Janeiro, Rio de Janeiro, Brazil. 92National Institute for Public Health and the Environment,
Bilthoven, The Netherlands. 93University of Turin, Turin, Italy. 94Nanyang Technological
University, Singapore, Singapore. 95Universidad de la República, Montevideo, Uruguay.
96
Centro de Educación Médica e Investigaciones Clínicas, Buenos Aires, Argentina. 97IRCCS
Neuromed, Pozzilli, Italy. 98Toulouse University School of Medicine, Toulouse, France.
99
University Hospital of Varese, Varese, Italy. 100University Hospital KU Leuven, Leuven,
Belgium. 101Ministry of Health, Victoria, Seychelles. 102University of Lausanne, Lausanne,
Switzerland. 103Universidad Central de Venezuela, Caracas, Venezuela. 104Bielefeld University,
Bielefeld, Germany. 105German Cancer Research Center, Heidelberg, Germany. 106University
College Copenhagen, Copenhagen, Denmark. 107Robert Koch Institute, Berlin, Germany.
108
Universidad de La Laguna, Tenerife, Spain. 109University of Malta, Msida, Malta. 110Istanbul
University – Cerrahpasa, Istanbul, Turkey. 111Universidade Federal de Juiz de Fora, Juiz de Fora,
Brazil. 112Gaetano Fucito Hospital, Mercato San Severino, Italy. 113University of Porto, Porto,
Portugal. 114Santiago de Compostela University, Santiago, Spain. 115Council for Agricultural
Research and Economics, Rome, Italy. 116Research and Education Institute of Child Health,
Nicosia, Cyprus. 117Dr. A. Ramachandran’s Diabetes Hospital, Chennai, India. 118University
College London, London, UK. 119Academia Sinica, Taipei, Taiwan. 120Capital Institute of
Pediatrics, Beijing, China. 121Kailuan General Hospital, Tangshan, China. 122Duke-NUS Medical
School, Singapore, Singapore. 123Ahvaz Jundishapur University of Medical Sciences, Ahvaz,
Iran. 124The Gertner Institute for Epidemiology and Health Policy Research, Ramat Gan, Israel.
125
Ministry of Health and Welfare, Taipei, Taiwan. 126Murcia Health Council, Murcia, Spain.
127
Seoul National University College of Medicine, Seoul, Republic of Korea. 128Korea Centers
1
Imperial College London, London, UK. 2Harvard T. H. Chan School of Public Health, Boston,
MA, USA. 3University of Auckland, Auckland, New Zealand. 4Tehran University of Medical
Sciences, Tehran, Iran. 5Middlesex University, London, UK. 6Shahid Beheshti University of
Medical Sciences, Tehran, Iran. 7Rush University Medical Center, Chicago, IL, USA. 8Institute
for Clinical Effectiveness and Health Policy, Buenos Aires, Argentina. 9Harvard Medical
School, Boston, MA, USA. 10Erasmus Medical Center Rotterdam, Rotterdam, The Netherlands.
11
World Health Organization, Geneva, Switzerland. 12Independent researcher, Los Angeles,
CA, USA. 13University of California Berkeley, Berkeley, CA, USA. 14Mahidol University, Nakhon
Pathom, Thailand. 15Charles University in Prague, Prague, Czech Republic. 16Thomayer
Hospital, Prague, Czech Republic. 17Istituto Superiore di Sanità, Rome, Italy. 18South African
Medical Research Council, Cape Town, South Africa. 19Seoul National University, Seoul,
Republic of Korea. 20Finnish Institute for Health and Welfare, Helsinki, Finland. 21ICMR–
National Institute of Nutrition, Hyderabad, India. 22Pontificia Universidad Católica de Chile,
Santiago, Chile. 23ICMR–National Centre for Disease Informatics and Research, Bengaluru,
India. 24Copenhagen University Hospital, Copenhagen, Denmark. 25Capital Medical University
Beijing An Zhen Hospital, Beijing, China. 26Bispebjerg and Frederiksberg Hospital,
Copenhagen, Denmark. 27Caja Costarricense de Seguro Social, San José, Costa Rica. 28Qatar
University, Doha, Qatar. 29Birzeit University, Birzeit, Palestine. 30Instituto Mexicano del Seguro
Social, Mexico City, Mexico. 31Flinders University, Adelaide, South Australia, Australia. 32Food
and Nutrition Research Institute, Taguig, The Philippines. 33Urmia University of Medical
Sciences, Urmia, Iran. 34Instituto Nacional de Ciencias Médicas y Nutricion, Mexico City,
for Disease Control and Prevention, Cheongju-si, Republic of Korea. 129Medical University of
Silesia, Katowice, Poland. 130Katholieke Universiteit Leuven, Leuven, Belgium. 131Statistics
Canada, Ottawa, Ontario, Canada. 132Ghent University, Ghent, Belgium. 133Agency for
Preventive and Social Medicine, Bregenz, Austria. 134Federal University of Maranhão, São Luís,
Brazil. 135University of Southampton, Southampton, UK. 136Institut Pasteur de Lille, Lille,
France. 137CIBEROBN, Madrid, Spain. 138University of Debrecen, Debrecen, Hungary.
139
Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil. 140National Council of
Research, Reggio Calabria, Italy. 141Federal University of Santa Catarina, Florianópolis, Brazil.
142
Eduardo Mondlane University, Maputo, Mozambique. 143Geneva University Hospitals,
Geneva, Switzerland. 144National Research Centre for Preventive Medicine, Moscow, Russia.
145
University of Montreal, Montreal, Québec, Canada. 146French Public Health Agency, St
Maurice, France. 147Mediterranea Cardiocentro, Naples, Italy. 148University of Zagreb, Zagreb,
Croatia. 149Ministry of Health and Medical Education, Tehran, Iran. 150University of Queensland,
Brisbane, Queensland, Australia. 151Universidad de Cuenca, Cuenca, Ecuador. 152Helmholtz
Zentrum München, Munich, Germany. 153Carol Davila University of Medicine and Pharmacy,
Bucharest, Romania. 154University Hospital Düsseldorf, Düsseldorf, Germany. 155National
Institute of Cardiology, Warsaw, Poland. 156Beth Israel Deaconess Medical Center, Boston, MA,
USA. 157University of Latvia, Riga, Latvia. 158Ministry of Health and Medical Services, Gizo,
Solomon Islands. 159Hormozgan University of Medical Sciences, Bandar Abbas, Iran.
160
University of Skövde, Skövde, Sweden. 161National Institute of Nutrition and Food
Technology, Tunis, Tunisia. 162The University of the West Indies, Kingston, Jamaica. 163Institut
Article
Hospital del Mar d’Investigacions Mèdiques, Barcelona, Spain. 164University of Stellenbosch,
Cape Town, South Africa. 165University of Duisburg-Essen, Duisburg, Germany. 166Karadeniz
Technical University, Trabzon, Turkey. 167University of Helsinki, Helsinki, Finland. 168Mashhad
University of Medical Sciences, Mashhad, Iran. 169Rafsanjan University of Medical Sciences,
Rafsanjan, Iran. 170Queen’s University of Belfast, Belfast, UK. 171University of Zurich, Zurich,
Switzerland. 172Tabriz University of Medical Sciences, Tabriz, Iran. 173Fasa University of Medical
Sciences, Fasa, Iran. 174Shiraz University of Medical Sciences, Shiraz, Iran. 175Centro de Salud
Villanueva Norte, Badajoz, Spain. 176Servicio Extremeño de Salud, Badajoz, Spain. 177Ministry of
Health, Buenos Aires, Argentina. 178Universiti Sains Malaysia, Kelantan, Malaysia. 179Umeå
University, Umeå, Sweden. 180Federal University of São Paulo, São Paulo, Brazil. 181Hospital
Universitario Son Espases, Palma, Spain. 182Kindai University, Osaka-Sayama, Japan. 183Kyoto
University, Kyoto, Japan. 184Medical University of Warsaw, Warsaw, Poland. 185University of
Catania, Catania, Italy. 186CIBER en Epidemiología y Salud Pública, Alicante, Spain. 187CIBER en
Epidemiología y Salud Pública, Barcelona, Spain. 188Universidade Federal de Minas Gerais,
Belo Horizonte, Brazil. 189Wageningen University, Wageningen, The Netherlands. 190B. P. Koirala
Institute of Health Sciences, Dharan, Nepal. 191University of Insubria, Varese, Italy. 192University
of Adelaide, Adelaide, South Australia, Australia. 193Lund University, Lund, Sweden. 194McGill
University, Montreal, Québec, Canada. 195Universidad Politécnica de Madrid, Madrid, Spain.
196
St Anne’s University Hospital, Brno, Czech Republic. 197National Institute of Public Health,
Cuernavaca, Mexico. 198Centro de Estudios en Diabetes A.C., Mexico City, Mexico.
199
Universidad Autónoma de Santo Domingo, Santo Domingo, Dominican Republic.
200
Institute for Clinical and Experimental Medicine, Prague, Czech Republic. 201Jagiellonian
University Medical College, Kraków, Poland. 202University of Southern Denmark, Odense,
Denmark. 203National Center of Cardiovascular Diseases, Beijing, China. 204Singapore Eye
Research Institute, Singapore, Singapore. 205Icelandic Heart Association, Kopavogur, Iceland.
Uppsala University, Uppsala, Sweden. 292National Institute for Health Development,
Tallinn, Estonia. 293Universidad San Martín de Porres, Lima, Peru. 294Consejería de Sanidad
Junta de Castilla y León, Valladolid, Spain. 295Norrbotten County Council, Luleå, Sweden.
296
Peking University, Beijing, China. 297Universidade Federal de Ouro Preto, Ouro Preto, Brazil.
298
University of Coimbra, Coimbra, Portugal. 299The Jikei University School of Medicine, Tokyo,
Japan. 300Institute of Neuroscience of the National Research Council, Padua, Italy. 301Baker
Heart and Diabetes Institute, Melbourne, Victoria, Australia. 302Agricultural University of
Athens, Athens, Greece. 303French National Research Institute for Sustainable Development,
Montpellier, France. 304Hospital Israelita Albert Einstein, São Paulo, Brazil. 305University of
Otago, Dunedin, New Zealand. 306University of Padua, Padua, Italy. 307Lausanne University
Hospital, Lausanne, Switzerland. 308Emory University, Atlanta, GA, USA. 309Universidade
Estadual do Centro-Oeste, Guarapuava, Brazil. 310UiT The Arctic University of Norway, Tromsø,
Norway. 311Cape Peninsula University of Technology, Cape Town, South Africa. 312Brown
University, Providence, RI, USA. 313University of Edinburgh, Edinburgh, UK. 314University
College Dublin, Dublin, Ireland. 315Shahid Sadoughi University of Medical Sciences, Yazd, Iran.
316
Institut National de la Santé et de la Recherche Médicale, Lille, France. 317ICMR–National
Institute of Medical Statistics, New Delhi, India. 318Lusófona University, Lisbon, Portugal.
319
Women’s Social and Health Studies Foundation, Trivandrum, India. 320Università degli Studi
di Firenze, Florence, Italy. 321Ain Shams University, Cairo, Egypt. 322Isfahan Cardiovascular
Research Center, Isfahan, Iran. 323University of Strasbourg, Strasbourg, France. 324Strasbourg
University Hospital, Strasbourg, France. 325University of Pécs, Pécs, Hungary. 326Mulago
Hospital, Kampala, Uganda. 327University of Medical Sciences of Cienfuegos, Cienfuegos,
Cuba. 328Royal College of Surgeons in Ireland Dublin, Dublin, Ireland. 329La Trobe University,
Melbourne, Victoria, Australia. 330International Institute of Molecular and Cell Biology,
Warsaw, Poland. 331Instituto Conmemorativo Gorgas de Estudios de la Salud, Panama City,
206
Panama. 332World Health Organization Country Office, Lilongwe, Malawi. 333University of
Universidad Icesi, Cali, Colombia. 207Eternal Heart Care Centre and Research Institute,
Jaipur, India. 208National Health Insurance Service, Wonju, Republic of Korea. 209Prevention of
Metabolic Disorders Research Center, Tehran, Iran. 210The University of the West Indies, Cave
Hill, Barbados. 211Kermanshah University of Medical Sciences, Kermanshah, Iran. 212University
of Eastern Finland, Kuopio, Finland. 213Kyushu University, Fukuoka, Japan. 214University of
Bergen, Bergen, Norway. 215Tulane University, New Orleans, LA, USA. 216Chinese Center for
Disease Control and Prevention, Beijing, China. 217Joep Lange Institute, Amsterdam, The
Netherlands. 218Chronic Diseases Research Center, Tehran, Iran. 219University of Hong Kong,
Hong Kong, China. 220The Chinese University of Hong Kong, Hong Kong, China. 221University of
Western Australia, Perth, Western Australia, Australia. 222Kingston Health Sciences Centre,
Kingston, Ontario, Canada. 223Heart Institute, São Paulo, Brazil. 224Fundación Oftalmológica de
Santander, Bucaramanga, Colombia. 225University Oran 1, Oran, Algeria. 226Independent
Public Health Specialist, Nay Pyi Taw, Myanmar. 227Ministry of Health and Sports, Nay Pyi Taw,
Myanmar. 228CIBER en Epidemiología y Salud Pública, Murcia, Spain. 229VU University Medical
Center, Amsterdam, The Netherlands. 230International Agency for Research on Cancer, Lyon,
France. 231American University of Beirut, Beirut, Lebanon. 232Cairo University, Cairo, Egypt.
233
University of Zaragoza, Zaragoza, Spain. 234National Institutes of Biomedical Innovation,
Health and Nutrition, Tokyo, Japan. 235Medical University Varna, Varna, Bulgaria. 236The
University of Tokyo, Tokyo, Japan. 237The Hospital for Sick Children, Toronto, Ontario, Canada.
238
Niigata University, Niigata, Japan. 239Hadassah University Medical Center, Jerusalem, Israel.
240
Norwegian University of Science and Technology, Trondheim, Norway. 241University of
Melbourne, Melbourne, Victoria, Australia. 242University Hospital Centre Zagreb, Zagreb,
Croatia. 243University of Zagreb School of Medicine, Zagreb, Croatia. 244Heart Foundation,
Melbourne, Victoria, Australia. 245Guangzhou 12th Hospital, Guangzhou, China. 246Universidad
Eugenio Maria de Hostos, Santo Domingo, Dominican Republic. 247Simon Fraser University,
Burnaby, British Columbia, Canada. 248Ruprecht-Karls-University of Heidelberg, Heidelberg,
Germany. 249World Health Organization Country Office, Delhi, India. 250Guilan University of
Medical Sciences, Rasht, Iran. 251University of Opole, Opole, Poland. 252University of Crete,
Heraklion, Greece. 253Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia. 254Maharajgunj
Medical Campus, Kathmandu, Nepal. 255Johns Hopkins Bloomberg School of Public Health,
Baltimore, MD, USA. 256National Institute of Epidemiology, Chennai, India. 257University of
Münster, Münster, Germany. 258Research Institute for Primordial Prevention of
Non-communicable Disease, Isfahan, Iran. 259Kyrgyz State Medical Academy, Bishkek,
Kyrgyzstan. 260Research Institute of Child Nutrition, Dortmund, Germany. 261Mazandaran
University of Medical Sciences, Sari, Iran. 262Hypertension Research Center, Isfahan, Iran.
263
Medical University of Innsbruck, Innsbruck, Austria. 264Muhimbili University of Health and
Allied Sciences, Dar es Salaam, Tanzania. 265Yonsei University College of Medicine, Seoul,
Republic of Korea. 266National Cancer Center, Goyang-si, Republic of Korea. 267University of
Pennsylvania, Philadelphia, PA, USA. 268University of Vienna, Vienna, Austria. 269Oulu
Deaconess Institute Foundation, Oulu, Finland. 270Tartu University Clinics, Tartu, Estonia.
271
Kansai Medical University, Hirakata, Japan. 272Ministry of Health and Quality of Life, Port
Louis, Mauritius. 273University Hospital Ulm, Ulm, Germany. 274University of Groningen,
Groningen, The Netherlands. 275University of Jyväskylä, Jyväskylä, Finland. 276African
Population and Health Research Center, Nairobi, Kenya. 277Higher Institute of Health Sciences
of Settat, Settat, Morocco. 278Ministry of Health, Algiers, Algeria. 279Harokopio University,
Athens, Greece. 280Sahlgrenska Academy, Gothenburg, Sweden. 281Endocrinology and
Metabolism Research Center, Tehran, Iran. 282University of Public Health, Yangon, Myanmar.
283
National University of Singapore, Singapore, Singapore. 284Tampere University Hospital,
Tampere, Finland. 285Tampere University, Tampere, Finland. 286University of Cape Town, Cape
Town, South Africa. 287West Virginia University, Morgantown, WV, USA. 288Oswaldo Cruz
Foundation Rene Rachou Research Institute, Belo Horizonte, Brazil. 289National Taiwan
University, Taipei, Taiwan. 290University of Chinese Academy of Sciences, Shanghai, China.
291
Brescia, Brescia, Italy. 334Bushehr University of Medical Sciences, Bushehr, Iran. 335Ulm
University, Ulm, Germany. 336Kobe University, Kobe, Japan. 337Suraj Eye Institute, Nagpur,
India. 338University Medicine of Greifswald, Greifswald, Germany. 339University of Medicine
and Pharmacy at Ho Chi Minh City, Ho Chi Minh City, Vietnam. 340Hanoi Medical University,
Hanoi, Vietnam. 341Miami Veterans Affairs Healthcare System, Miami, FL, USA. 342University of
Turku, Turku, Finland. 343Eastern Mediterranean Public Health Network, Amman, Jordan.
344
University of Manchester, Manchester, UK. 345Japan Wildlife Research Center, Tokyo, Japan.
346
Istanbul University, Istanbul, Turkey. 347Ministry of Health, Bandar Seri Begawan, Brunei.
348
University of Madeira, Funchal, Portugal. 349MRC Lifecourse Epidemiology Unit,
Southampton, UK. 350University of Novi Sad, Novi Sad, Serbia. 351Kwame Nkrumah University
of Science and Technology, Kumasi, Ghana. 352Institute for Social and Preventive Medicine,
Ottawa, Ontario, Canada. 353IRCCS Ente Ospedaliero Specializzato in Gastroenterologia S. de
Bellis, Bari, Italy. 354Jivandeep Hospital, Anand, India. 355South African Medical Research
Council, Durban, South Africa. 356Vietnam National Heart Institute, Hanoi, Vietnam. 357Clínica
de Medicina Avanzada Dr. Abel González, Santo Domingo, Dominican Republic. 358Leibniz
Institute for Prevention Research and Epidemiology – BIPS, Bremen, Germany. 359University of
Sarajevo, Sarajevo, Bosnia and Herzegovina. 360Cardiovascular Prevention Centre, Udine, Italy.
361
Public Health Agency of Catalonia, Barcelona, Spain. 362Observatorio de Salud Pública de
Santander, Bucaramanga, Colombia. 363Ardabil University of Medical Sciences, Ardabil, Iran.
364
Alborz University of Medical Sciences, Karaj, Iran. 365Ministry of Health, Hanoi, Vietnam.
366
India Diabetes Research Foundation, Chennai, India. 367Institut Universitari d’Investigació en
Atenció Primària Jordi Gol, Girona, Spain. 368Universiti Putra Malaysia, Serdang, Malaysia.
369
University of Malaya, Kuala Lumpur, Malaysia. 370University of Valencia, Valencia, Spain.
371
University of the Philippines, Manila, The Philippines. 372Minas Gerais State Secretariat for
Health, Belo Horizonte, Brazil. 373CS S. Agustín Ibsalut, Palma, Spain. 374Amsterdam Institute
for Global Health and Development, Amsterdam, The Netherlands. 375Canarian Health
Service, Tenerife, Spain. 376Universidad Industrial de Santander, Bucaramanga, Colombia.
377
Associazione Calabrese di Epatologia, Reggio Calabria, Italy. 378Sahlgrenska University
Hospital, Gothenburg, Sweden. 379Toulouse University Hospital, Toulouse, France. 380Institute
of Food Sciences of the National Research Council, Avellino, Italy. 381Sitaram Bhartia Institute
of Science and Research, New Delhi, India. 382Faculty of Medicine of Tunis, Tunis, Tunisia.
383
National Institute of Health, Lima, Peru. 384Catalan Department of Health, Barcelona, Spain.
385
Universidade de Lisboa, Lisbon, Portugal. 386South Karelia Social and Health Care District,
Lappeenranta, Finland. 387Cardiovascular Research Institute, Isfahan, Iran. 388National Cancer
Center, Tokyo, Japan. 389University of São Paulo Clinics Hospital, São Paulo, Brazil. 390Hospital
Italiano de Buenos Aires, Buenos Aires, Argentina. 391Center for Oral Health Services and
Research Mid-Norway, Trondheim, Norway. 392King’s College London, London, UK. 393National
Center for Global Health and Medicine, Tokyo, Japan. 394Sungkyunkwan University, Seoul,
Republic of Korea. 395Finnish Institute of Occupational Health, Helsinki, Finland. 396St Vincent’s
Hospital, Sydney, New South Wales, Australia. 397University of New South Wales, Sydney, New
South Wales, Australia. 398Karolinska Institutet, Stockholm, Sweden. 399Research Centre for
Prevention and Health, Glostrup, Denmark. 400London School of Hygiene & Tropical Medicine,
London, UK. 401Diponegoro University, Semarang, Indonesia. 402University of Bari, Bari, Italy.
403
University of Copenhagen, Copenhagen, Denmark. 404Institut Régional de Santé Publique,
Ouidah, Benin. 405University of Bordeaux, Bordeaux, France. 406Izmir Katip Çelebi University,
Izmir, Turkey. 407University of Leuven, Leuven, Belgium. 408Institut National de la Santé et de la
Recherche Médicale, Nancy, France. 409Bonn University, Bonn, Germany. 410Croatian Institute
of Public Health, Zagreb, Croatia. 411National Institute of Public Health–National Institute of
Hygiene, Warsaw, Poland. 412National Institute of Hygiene, Epidemiology and Microbiology,
Havana, Cuba. 413Fu Jen Catholic University, Taipei, Taiwan. 414National Statistic Office of Cabo
Verde, Praia, Cabo Verde. 415Ministry of Health, Amman, Jordan. 416Central University of Kerala,
Kasaragod, India. 417Health Service of Murcia, Murcia, Spain. 418Institut d’Investigacio Sanitaria
Illes Balears, Menorca, Spain. 419Universidad Centro-Occidental Lisandro Alvarado,
Barquisimeto, Venezuela. 420Dokuz Eylul University, Izmir, Turkey. 421University of Tampere Tays
Eye Center, Tampere, Finland. 422Icahn School of Medicine at Mount Sinai, New York City, NY,
USA. 423Utrecht University, Utrecht, The Netherlands. 424Hanoi University of Public Health, Hanoi,
Vietnam. 425University Medical Center Utrecht, Utrecht, The Netherlands. 426Universitas
Indonesia, Jakarta, Indonesia. 427Instituto de Investigación Sanitaria y Biomédica de Alicante,
Alicante, Spain. 428North Karelian Center for Public Health, Joensuu, Finland. 429University of the
Witwatersrand, Johannesburg, South Africa. 430Cork Institute of Technology, Cork, Ireland.
431
Institute for Medical Research, Kuala Lumpur, Malaysia. 432Health Canada, Ottawa, Ontario,
Canada. 433Beijing Institute of Ophthalmology, Beijing, China. 434Xinjiang Medical University,
Urumqi, China. 435Capital Medical University, Beijing, China. 436St George’s, University of
London, London, UK. 437Medical University of Vienna, Vienna, Austria. 438University of Oxford,
Oxford, UK. 439Institute of Food and Nutrition Development of Ministry of Agriculture and Rural
Affairs, Beijing, China. 440Children’s Hospital of Fudan University, Shanghai, China. 441Penang
Medical College, Penang, Malaysia. 442University of Cyprus, Nicosia, Cyprus. 443Iran University of
Medical Sciences, Tehran, Iran. 444Jiangsu Provincial Center for Disease Control and Prevention,
Nanjing, China. 445Sun Yat-sen University, Guangzhou, China. 446West Kazakhstan State Medical
University, Aktobe, Kazakhstan. 447University of Ghana, Accra, Ghana. 448Deceased: Konrad
Jamrozik, Altan Onat, Robespierre Ribeiro, Jutta Stieber. ✉e-mail: majid.ezzati@imperial.ac.uk
Article
Methods
Our aim was to estimate trends in mean total, HDL and non-HDL cholesterol for 200 countries and territories (Supplementary Table 2). We
used non-HDL cholesterol rather than low-density lipoprotein (LDL)
cholesterol because most studies in our analysis had measured total
cholesterol and HDL cholesterol, from which non-HDL cholesterol can
be calculated through subtraction. By contrast, LDL cholesterol was
directly measured in only around 14% of studies. When LDL cholesterol
is not directly measured, its calculation requires data on triglycerides,
which were available in approximately 64% of the studies. Furthermore,
the most-commonly used estimation method—that is, the Friedewald
equation—can be inaccurate, particularly at high levels of triglycerides34. Non-HDL and LDL cholesterol were highly correlated (Pearson
correlation coefficient = 0.94) in studies with data on both variables
(Extended Data Fig. 9), because LDL cholesterol constitutes most of
non-HDL cholesterol. Furthermore, non-HDL cholesterol predicts
IHD risk at least as well as LDL cholesterol5,35, and can be measured at
a lower cost than LDL cholesterol, which is relevant for how widely it
can be used in low- and middle-income countries. Although non-HDL
cholesterol is now commonly used in clinical guidelines36–38, LDL cholesterol continues to be a key target for treatment36,37, possibly because
the interpretation of non-HDL cholesterol is more complex than LDL
cholesterol alone. Specifically, an increase in non-HDL cholesterol
could be due to the increase in LDL cholesterol or very-low-density
lipoprotein cholesterol39. Furthermore, there is some evidence that
triglyceride levels are high in Asian populations, compared to levels
seen in high-income western countries40. Therefore, data on non-HDL
cholesterol can motivate dietary interventions to both reduce LDL
cholesterol (for example, reducing saturated and trans fat intake)
and triglyceride levels (for example, reducing refined carbohydrates
and increasing omega-3 fatty acids) as well as treatments that lower
LDL cholesterol (statins), alongside those that lower triglycerides (for
example, fibrates).
Data sources
We used a database of population-based data on cardiometabolic risk
factors collated by the NCD Risk Factor Collaboration (NCD-RisC),
a worldwide network of health researchers and practitioners that
systematically monitors the worldwide trends and variations in
non-communicable disease (NCD) risk factors. The database was
collated through multiple routes for identifying and accessing data.
We accessed publicly available population-based multi-country and
national measurement surveys (for example, Demographic and Health
Surveys and surveys identified through the Inter-University Consortium for Political and Social Research and European Health Interview
& Health Examination Surveys Database). We requested, via the World
Health Organization (WHO) and its regional and country offices, from
ministries of health and other national health and statistical agencies to identify and access population-based surveys. Requests were
also sent via the World Heart Federation to its national partners. We
made a similar request to the co-authors of an earlier pooled analysis
of cardiometabolic risk factors7,41–43, and invited the co-authors of the
analysis to reanalyse data from their studies and join NCD-RisC. Finally,
to identify major sources that were not accessed through the above
routes, we searched and reviewed published studies as described in
the Supplementary Information and invited all eligible studies to join
NCD-RisC.
For each data source, we recorded the available information about
the study population, start year and duration of measurement, sampling approach and measurement methods. The information about
study population was used to establish that each data source was
population-based, and to assess whether it covered the whole country,
multiple subnational regions or one or a small number of communities,
and whether it was rural, urban or combined.
We carefully checked all data sources in terms of how they met our
inclusion and exclusion criteria listed below. We identified duplicate
data sources by comparing studies from the same country and year.
Additionally, all NCD-RisC members are asked periodically to review
the list of sources from their country, to suggest additional sources not
in the database, and to verify that the included data meet the inclusion
criteria listed below and are not duplicates. The NCD-RisC database is
continuously updated through the above routes and through regular
contact with NCD-RisC members.
Anonymized individual record data from sources included in
NCD-RisC were reanalysed according to a common protocol. Within
each survey, we included participants aged 18 years and older who
were not pregnant. We removed participants with implausible total
cholesterol levels (defined as total cholesterol levels of <1.75 mmol l−1
or >20 mmol l−1, or total cholesterol values that were lower than HDL
cholesterol values) (<0.05% of all participants with total cholesterol
measurements) or HDL cholesterol levels (defined as HDL cholesterol
levels of <0.4 mmol l−1 or >5 mmol l−1, or total cholesterol values that
were lower than HDL cholesterol values) (<0.15% of all participants with
HDL cholesterol measurements). When data on LDL cholesterol were
also available, we removed individuals for whom the sum of LDL and
HDL cholesterol level surpassed total cholesterol level by more than is
plausible based on the limits to errors in their measurement (following
the CDC Cholesterol Reference Method Laboratory Network (CRMLN)
standards, these errors were set at 8.9% for total cholesterol, 13% for
HDL cholesterol and 12% for LDL cholesterol) (<0.06% of all participants
with total cholesterol and HDL cholesterol measurements)44–46.
We calculated mean total cholesterol, mean HDL cholesterol and
mean non-HDL cholesterol, and associated standard errors and sample sizes, by sex and age group (18–19 years, 20–29 years, followed by
10-year age groups and 80+ years). All analyses incorporated appropriate sample weights and complex survey design in calculating
age–sex-specific means when applicable. To ensure summaries were
prepared according to the study protocol, computer code was provided
to NCD-RisC members who requested assistance. All submitted data
were checked independently by at least two researchers. Questions and
clarifications were discussed with NCD-RisC members and resolved
before the data were incorporated in the database.
Finally, we obtained data not accessed through the above routes
by extracting data from published reports of all additional national
health surveys identified through the above-described strategies, as
well as eight sites of the WHO Multinational MONItoring of trends and
determinants in CArdiovascular disease (MONICA) project that were
not deposited in the MONICA Data Centre. Data were extracted from
published reports only when reported by sex and in age groups no wider
than 20 years. We also used data from a previous pooling study7 when
such data did not overlap with those accessed through the above routes.
Data inclusion and exclusion
Data sources were included in NCD-RisC database if: (1) measured data
on total, LDL, HDL cholesterol and/or triglycerides were available; (2)
study participants were 10 years of age or older; (3) data were collected
using a probabilistic sampling method with a defined sampling frame;
(4) data were from population samples at the national, subnational
(covering one or more subnational regions, more than three urban
communities or more than five rural communities) or community (one
or a small number of communities) level; (5) data were collected in or
after 1950; and (6) data were from the countries and territories listed
in Supplementary Table 2.
We excluded all data sources that included only hypercholesterolaemia or dyslipidaemia diagnosis history or medication status without
measurement of cholesterol levels. We also excluded data sources
on population subgroups for which the lipid profile may differ systematically from the general population, including: (1) studies that
had included or excluded people on the basis of their health status or
cardiovascular risk; (2) studies for which the participants were only
from ethnic minorities; (3) studies that had recruited only specific educational, occupational or socioeconomic subgroups, with the exception
noted below; and (4) studies that had recruited participants through
health facilities, with the exception noted below.
We used school-based data in countries and for age–sex groups, for
which secondary school enrolment was 70% or higher. We used data for
which the sampling frame was health insurance schemes in countries
in which at least 80% of the population was insured. Finally, we used
data collected through general practice and primary-care systems in
high-income and central European countries with universal insurance,
because contact with the primary-care systems tends to be as good as or
better than response rates for population-based surveys. We used data
sources regardless of fasting status, because the differences between
fasting and non-fasting measurements are negligible for total, non-HDL
and HDL cholesterol39, and therefore non-fasting lipid profiles are now
widely endorsed for the estimation of cardiovascular risk36,37.
Data used in the analysis
For this paper, we used data from the NCD-RisC database for years
1980 to 2018 and individuals aged 18 years and older. A list of the data
sources that we used in this analysis and their characteristics is provided
in Supplementary Table 1. The data comprised 1,127 population-based
measurement surveys and studies that included measurements of
blood lipids on 102.6 million participants aged 18 years and older.
We had at least one data source for 161 of the 200 countries that we
made estimates for, covering 92.4% of the world’s population in 2018
(Extended Data Fig. 1); and at least two data sources for 104 countries
(87.5% of the world population). Of these 1,127 sources, 409 (36.3%)
sampled from national populations, 250 (22.2%) covered one or more
subnational regions, and the remaining 468 (41.5%) were from one or a
small number of communities. Regionally, data availability ranged from
around 2 data sources per country in sub-Saharan Africa to approximately 35 sources per country in the high-income Asia–Pacific region.
In total, 454 data sources (40.3%) were from years before 2000 and the
remaining 673 (59.7%) were collected from 2001 onwards.
Adjusting for the differences in mean cholesterol between
portable device and laboratory measurements
In 112 (10%) of the 1,127 data sources used in our analysis (11.5% and 5.8%
of age–sex-specific data points for total and HDL cholesterol, respectively) lipids were measured using a portable device. Some portable
devices have narrower analytical ranges than laboratory methods,
which results in truncations of blood cholesterol data that are outside
their range (Supplementary Table 3). This may in turn affect the population mean. Although cholesterol concentrations that fall outside the
analytical range are displayed as ‘high’ (above the measurement range)
or ‘low’ (below the measurement range) by these devices, different surveys record and code cholesterol concentrations outside the analytical
range in different ways, for example using ‘too low’, ‘too high’ and ‘error’
codes; assigning the minimum or maximum value to individuals whose
cholesterol was below or above the analytical range, respectively; setting values outside the analytical range to missing; and so on. We used
an approach that treated surveys with such data consistently.
Specifically, we first dropped all participants with cholesterol levels
below and at the minimum, and at and above the maximum, values of
the analytical range of each portable device before calculating the mean
cholesterol (Supplementary Table 3). We then developed conversion
regressions to adjust the mean cholesterol levels measured using a
portable device (calculated over the restricted range, Supplementary
Table 3) to the levels expected using laboratory measurements. The
dependent variable in each regression was mean total, non-HDL or HDL
cholesterol for the full range, and the main independent variable was
mean total, non-HDL or HDL cholesterol over the above-mentioned
restricted cholesterol range of the portable devices. The regression
coefficients were estimated from data sources for which lipids were
measured in a laboratory, and thus had the full range of measurement
and could be used to calculate both dependent and independent variables. When estimating the regression coefficients, we constructed the
dependent variable using the full data, and the independent variable
by dropping the values outside the above-mentioned restricted cholesterol range of each device, mimicking those that would be expected
if a portable device had been used. Separate models were developed
according to the specific range of the different portable devices. All
regressions included terms for age and sex, as well as interactions
between predictors and age and sex, based on the Bayesian information
criterion47. The regressions for mean non-HDL cholesterol also included
mean total cholesterol and mean HDL cholesterol because non-HDL
cholesterol is calculated from total cholesterol and HDL cholesterol.
We excluded data points for which there were fewer than 25 individuals for the purpose of estimating the coefficients of these regressions.
All sources of uncertainty in the conversion—including the sampling
uncertainty of the original data, the uncertainty of the regression coefficients and residuals—were carried forward by using repeated draws
from their respective distributions. The regression coefficients and
number of data points used to estimate the coefficients are shown in
Supplementary Table 4.
Statistical analysis
We used a statistical model to estimate mean total, non-HDL and HDL
cholesterol by country, year, sex and age using all of the available data.
The model is described in detail in a statistical paper and related substantive papers8,32,33,48; the computer code is available at http://www.
ncdrisc.org/. In summary, we organized countries into 21 regions,
mainly based on geography and national income; these regions were
further aggregated into 9 ‘super-regions’ (Supplementary Table 2). The
model had a hierarchical structure in which estimates for each country
and year were informed by its own data, if available, and by data from
other years in the same country and from other countries, especially
countries in the same region or super-region with data for similar
time periods. The extent to which estimates for each country-year
are influenced by data from other years and other countries depends
on whether the country has data, the sample size of data, whether or
not they are national, and the within-country and within-region data
variability. The model incorporated nonlinear time trends comprising
linear terms and a second-order random walk. The age association of
blood lipids was modelled using a cubic spline to allow nonlinear age
patterns, which might vary across countries. The model accounted for
the possibility that blood lipids in subnational and community samples
might systematically differ from nationally representative ones; and/
or have larger variation. These features were implemented by including data-driven fixed-effect and random-effect terms for subnational
and community data. The fixed effects adjust for systematic differences between subnational or community studies and national studies.
The random effects allow national data to have larger influence on the
estimates than subnational or community data with similar sample
sizes. The model also accounted for rural–urban differences in blood
lipids, through the use of data-driven fixed effects for rural-only and
urban-only studies. These rural and urban effects were weighted by the
difference between study-level and country-level urbanization in the
year in which the study was done. The proportion of the national population living in urban areas was also included as a predictor (covariate)
in the model. The model for mean non-HDL and HDL cholesterol also
used age-standardized mean total cholesterol as a covariate.
We fitted the statistical model with the Markov chain Monte Carlo
(MCMC) algorithm, and obtained 5,000 post-burn-in samples from
the posterior distribution of model parameters, which were in turn
used to obtain the posterior distributions of mean total, non-HDL and
HDL cholesterol. We calculated average change in mean total, HDL and
non-HDL cholesterol across the 39 years of analysis (reported as change
Article
per decade). Age-standardized estimates were generated by taking
weighted averages of age–sex-specific estimates, using the WHO standard population. Estimates for regions and the world were calculated
as population-weighted averages of the constituent country estimates
by age group and sex. The reported credible intervals represent the
2.5–97.5th percentiles of the posterior distributions. We also report
the posterior probability that an estimated increase or decrease represents a truly increasing or decreasing trend as opposed to a chance
observation. We performed all analyses by sex, because blood lipids
levels and trends are different in men and women.
Validation of statistical model
We tested how well our statistical model predicts missing data, known
as external predictive validity, in two different tests. In the first test, we
held out all data from 10% of countries with data (that is, created the
appearance of countries with no data where we actually had data). The
countries for which the data were withheld were randomly selected
from the following three groups: data rich (5 or more data sources, with
at least one data source after the year 2000), data poor (1 data source)
and average data availability (2–4 data sources). In the second test, we
assessed other patterns of missing data by holding out 10% of our data
sources, again from a mix of data-rich, data-poor and average-data
countries, as defined above. For a given country, we either held out
a random half of the data of a country or all of the 2000–2018 data
of the country to determine, respectively, how well we filled in the
gaps for countries with intermittent data and how well we estimated in
countries without recent data. In both tests, we then fitted the model
to the remaining 90% of the countries (test 1) or data sources (test 2)
and made estimates of the held-out observations. We repeated each
test five times, holding out a different subset of data in each repetition.
In both tests, we calculated the differences between the held-out data
and the estimates. We also calculated the 95% credible intervals of the
estimates; in a model with good external predictive validity, 95% of
held-out values would be included in the 95% credible intervals.
Our statistical model performed well in the external validation tests,
that is, in estimating mean cholesterol when data were missing. The
estimates of mean total, non-HDL and HDL cholesterol were unbiased,
as evidenced with median errors that were very close to zero globally
for every outcome and test, and less than ±0.30 mmol l−1 in every subset
of withheld data except for women in the high-income Asia–Pacific
region in test 1 for non-HDL cholesterol (median error 0.47 mmol l−1)
and men in south Asia in test 2 for non-HDL cholesterol (median error
−0.33 mmol l−1) (Supplementary Table 5). The 95% credible intervals
of estimated means covered 83–92% and 75–83% of true data globally
in the first and second tests, respectively. In subsets, coverage ranged
from 47% to 100%, but was mostly greater than 75%, with coverage
generally lower in test 2 than test 1. Median absolute errors ranged
from 0.07 to 0.23 mmol l−1 globally for different outcomes and sexes,
and were no more than 0.45 mmol l−1 in all subsets of withheld data,
except for women in the high-income Asia–Pacific region for non-HDL
cholesterol in test 1 (median absolute error 0.47 mmol l−1).
Calculation of the number of deaths attributable to high
cholesterol
We estimated the number of deaths from IHD and ischaemic stroke
attributable to high non-HDL cholesterol. For each country, year, sex
and age group, we first calculated the population attributable fractions—that is, the proportion of deaths from IHD and ischaemic stroke
that would have been prevented if non-HDL cholesterol levels were at
an optimal level (defined as a mean of 1.8–2.2 mmol l−1) in the population6,49. For these calculations, we used age-specific relative risks from
meta-analyses of prospective cohort studies4,5,50. The number of IHD
and ischaemic stroke deaths attributable to high non-HDL cholesterol
was calculated for each country–year–age–sex group by multiplying the
cause-specific population attributable fractions by the cause-specific
deaths from the Global Burden of Disease study in 1990 and 2017 (the
earliest and latest years with cause-specific mortality data).
Strengths and limitations
The strengths of our study include its scope in making consistent and
comparable estimates of trends in blood cholesterol and its cardiovascular disease mortality burden, over almost four decades for all of
the countries in the world, including global estimates of non-HDL and
HDL cholesterol. We used a large amount of population-based data,
which came from countries in which 92% of the global adult population
lives. We used only data from studies that had measured blood lipids
to avoid bias in self-reported data. Data were analysed according to a
consistent protocol, and the characteristics and quality of data from
each country were rigorously verified through repeated checks by
NCD-RisC members. We pooled data using a statistical model that
took into account the epidemiological features of cholesterol, including nonlinear time trends and age associations. Our statistical model
used all available data while giving more weight to national data than
to subnational and community sources.
Similar to all global analyses, our study is affected by some limitations. Despite our extensive efforts to identify and access worldwide
population-based data, some countries had no or few data sources,
especially those in sub-Saharan Africa, the Caribbean, central Asia
and Melanesia. Estimates for these countries relied mostly or entirely
on the statistical model, which shares information across countries
and regions through its hierarchy. Data scarcity is reflected in wider
uncertainty intervals of our estimates for these countries and regions,
highlighting the need for national NCD-oriented surveillance. The
distribution of lipids measured in a population using a portable device,
which was used in 10% of our studies, may be truncated and may therefore affect the population mean. To overcome this issue, we developed
conversion regressions to adjust mean cholesterol levels measured
using a portable device to the levels expected in laboratory measurements; the conversion regressions used for this purpose had good
predictive accuracy. Although most studies had measured cholesterol
in serum samples, around 7% had used plasma samples. As cholesterol measured in plasma and serum samples differ51 by only about 3%,
adjusting for plasma-serum differences would have little effect on our
results, as seen in a previous analysis14. Although methods to measure
total and HDL cholesterol have evolved over time, since the 1950s there
have been systematic efforts to standardize lipid measurements that
have resulted in increased comparability between different methods.
In our analysis, 90% of studies measured lipids in a laboratory; of these
studies more than 60% for total cholesterol and more than 70% for
HDL cholesterol participated in a lipid standardization programme
or quality control scheme. We did not analyse emerging lipid markers such as apolipoprotein B and apolipoprotein A-I, because they are
neither commonly measured in population-based health surveys, nor
routinely used in clinical practice36.
Comparison with other studies
There are no global analyses on trends in lipid fractions for comparison
with our results. Our findings for total cholesterol were largely consistent with the only other previous analysis7, but we estimated a larger
decrease in mean total cholesterol in high-income western countries
and central Europe, and a larger increase in southeast Asia, because
we had an additional decade of data compared with the earlier global
analysis. Therefore, although the highest mean total cholesterol levels
reported previously7, for 2008, were still in high-income western countries, we estimated that in 2018 total cholesterol was equally high or
higher in southeast Asia. Our findings on mean total cholesterol trends
are also largely consistent with previous multi- and single-country
reports14,15,17–21,52–73. Differences from previous studies—for example,
in Italy61, Lithuania63, the Netherlands65, Russian Federation69 and in
some countries that participated in the MONICA Project52—mostly arise
because our study covered a longer period and used a larger number
of data sources. Studies15,18,54,63,66,70,74–77 that have reported trends in
lipid fractions for a period longer than 15 years have found changes
in non-HDL cholesterol (or in LDL cholesterol for some studies) that
were consistent with our results.
Reporting summary
Further information on research design is available in the Nature
Research Reporting Summary linked to this paper.
Data availability
Estimates of mean total, non-HDL and HDL cholesterol by country,
year and sex are available at http://www.ncdrisc.org/. Input data from
publicly available sources can also be downloaded from http://www.
ncdrisc.org/. For other data sources, contact information for data
providers can be obtained from http://www.ncdrisc.org/.
Code availability
The computer code for the Bayesian hierarchical model used in this
work is available at http://www.ncdrisc.org/.
34. Martin, S. S. et al. Friedewald-estimated versus directly measured low-density lipoprotein
cholesterol and treatment implications. J. Am. Coll. Cardiol. 62, 732–739 (2013).
35. Cui, Y. et al. Non-high-density lipoprotein cholesterol level as a predictor of
cardiovascular disease mortality. Arch. Intern. Med. 161, 1413–1419 (2001).
36. Grundy, S. M. et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/
NLA/PCNA guideline on the management of blood cholesterol: a report of the American
College of Cardiology/American Heart Association Task Force on clinical practice
guidelines. Circulation 139, e1082–e1143 (2019).
37. Mach, F. et al. 2019 ESC/EAS guidelines for the management of dyslipidaemias: lipid
modification to reduce cardiovascular risk. Eur. Heart J. 41, 111–188 (2020).
38. Expert Dyslipidemia Panel of the International Atherosclerosis Society. An International
Atherosclerosis Society Position Paper: global recommendations for the management of
dyslipidemia—full report. J. Clin. Lipidol. 8, 29–60 (2014).
39. Nordestgaard, B. G. et al. Fasting is not routinely required for determination of a lipid
profile: clinical and laboratory implications including flagging at desirable concentration
cut-points—a joint consensus statement from the European Atherosclerosis Society and
European Federation of Clinical Chemistry and Laboratory Medicine. Eur. Heart J. 37,
1944–1958 (2016).
40. Bilen, O., Kamal, A. & Virani, S. S. Lipoprotein abnormalities in South Asians and its
association with cardiovascular disease: current state and future directions. World J.
Cardiol. 8, 247–257 (2016).
41. Danaei, G. et al. National, regional, and global trends in systolic blood pressure since
1980: systematic analysis of health examination surveys and epidemiological studies
with 786 country-years and 5·4 million participants. Lancet 377, 568–577 (2011).
42. Danaei, G. et al. National, regional, and global trends in fasting plasma glucose and
diabetes prevalence since 1980: systematic analysis of health examination surveys and
epidemiological studies with 370 country-years and 2·7 million participants. Lancet 378,
31–40 (2011).
43. Finucane, M. M. et al. National, regional, and global trends in body-mass index since
1980: systematic analysis of health examination surveys and epidemiological studies
with 960 country-years and 9·1 million participants. Lancet 377, 557–567 (2011).
44. Cholesterol Reference Method Laboratory Network. Total Cholesterol Certification
Protocol for Manufacturers. https://www.cdc.gov/labstandards/pdf/crmln/
RevisedTCprotocolOct04.pdf (Cholesterol Reference Method Laboratory Network, 2004).
45. Cholesterol Reference Method Laboratory Network. HDL Cholesterol Certification
Protocol for Manufacturers. https://www.cdc.gov/labstandards/pdf/crmln/HDL_
Certification_Protocol-508.pdf (Cholesterol Reference Method Laboratory Network, 2018).
46. Cholesterol Reference Method Laboratory Network. LDL Cholesterol Certification
Protocol for Manufacturers. https://www.cdc.gov/labstandards/pdf/crmln/LDL_
Certification_Protocol-508.pdf (Cholesterol Reference Method Laboratory Network, 2018).
47. Schwarz, G. Estimating the dimension of a model. Ann. Stat. 6, 461–464 (1978).
48. Finucane, M. M., Paciorek, C. J., Danaei, G. & Ezzati, M. Bayesian estimation of
population-level trends in measures of health status. Stat. Sci. 29, 18–25 (2014).
49. Ezzati, M., Lopez, A. D., Rodgers, A., Vander Hoorn, S. & Murray, C. J. Selected major risk
factors and global and regional burden of disease. Lancet 360, 1347–1360 (2002).
50. Singh, G. M. et al. The age-specific quantitative effects of metabolic risk factors on
cardiovascular diseases and diabetes: a pooled analysis. PLoS ONE 8, e65174 (2013).
51. National Cholesterol Education Program. Recommendations on Lipoprotein
Measurement. From the Working Group on Lipoprotein Measurement. NIH Publication No.
95-3044 (National Institutes of Health, National Heart, Lung, and Blood Institute, 1995).
52. Evans, A. et al. Trends in coronary risk factors in the WHO MONICA project. Int. J.
Epidemiol. 30, S35–S40 (2001).
53. Bennett, S. A. & Magnus, P. Trends in cardiovascular risk factors in Australia. Results from
the National Heart Foundation’s Risk Factor Prevalence Study, 1980–1989. Med. J. Aust.
161, 519–527 (1994).
54. Cífková, R. et al. Longitudinal trends in major cardiovascular risk factors in the Czech
population between 1985 and 2007/8. Czech MONICA and Czech post-MONICA.
Atherosclerosis 211, 676–681 (2010).
55. Sun, J. Y. et al. The changing trend of serum total cholesterol in Beijing population aged
25 - 64 years during 1984 - 1999 (article in Chinese). Zhonghua Nei Ke Za Zhi 45, 980–984
(2006).
56. Afzal, S., Tybjærg-Hansen, A., Jensen, G. B. & Nordestgaard, B. G. Change in body mass
index associated with lowest mortality in Denmark, 1976–2013. J. Am. Med. Assoc. 315,
1989–1996 (2016).
57. Ferrières, J. et al. Trends in plasma lipids, lipoproteins and dyslipidaemias in French
adults, 1996–2007. Arch. Cardiovasc. Dis. 102, 293–301 (2009).
58. Truthmann, J. et al. Changes in mean serum lipids among adults in Germany: results from
National Health Surveys 1997–99 and 2008–11. BMC Public Health 16, 240 (2016).
59. Sigfusson, N. et al. Decline in ischaemic heart disease in Iceland and change in risk factor
levels. Br. Med. J. 302, 1371–1375 (1991).
60. Gupta, R. et al. Twenty-year trends in cardiovascular risk factors in India and influence of
educational status. Eur. J. Prev. Cardiol. 19, 1258–1271 (2012).
61. Giampaoli, S. et al. Cardiovascular health in Italy. Ten-year surveillance of cardiovascular
diseases and risk factors: Osservatorio Epidemiologico Cardiovascolare/Health
Examination Survey 1998–2012. Eur. J. Prev. Cardiol. 22, 9–37 (2015).
62. Iso, H. Changes in coronary heart disease risk among Japanese. Circulation 118,
2725–2729 (2008).
63. Luksiene, D. et al. Trends in prevalence of dyslipidaemias and the risk of mortality in
Lithuanian urban population aged 45–64 in relation to the presence of the dyslipidaemias
and the other cardiovascular risk factors. PLoS ONE 9, e100158 (2014).
64. Uusitalo, U. et al. Fall in total cholesterol concentration over five years in association with
changes in fatty acid composition of cooking oil in Mauritius: cross sectional survey.
Br. Med. J. 313, 1044–1046 (1996).
65. Koopman, C. et al. Trends in risk factors for coronary heart disease in the Netherlands.
BMC Public Health 16, 835 (2016).
66. Metcalf, P. et al. Trends in major cardiovascular risk factors in Auckland, New Zealand:
1982 to 2002–2003. N. Z. Med. J. 119, U2308 (2006).
67. Sharashova, E., Wilsgaard, T. & Brenn, T. Resting heart rate on the decline: the Tromsø
Study 1986–2007. Int. J. Epidemiol. 44, 1007–1017 (2015).
68. Pajak, A. et al. Changes over time in blood lipids and their correlates in Polish rural and
urban populations: the Poland–United States Collaborative Study in cardiopulmonary
disease epidemiology. Ann. Epidemiol. 7, 115–124 (1997).
69. Vlasoff, T. et al. Ten year trends in chronic disease risk factors in the Republic of Karelia,
Russia. Eur. J. Public Health 18, 666–673 (2008).
70. Bovet, P. et al. Divergent fifteen-year trends in traditional and cardiometabolic risk factors
of cardiovascular diseases in the Seychelles. Cardiovasc. Diabetol. 8, 34 (2009).
71. Serra-Majem, L. et al. Trends in blood lipids and fat soluble vitamins in Catalonia, Spain
(1992–2003). Public Health Nutr. 10, 1379–1388 (2007).
72. Berg, C. M. et al. Trends in blood lipid levels, blood pressure, alcohol and smoking habits
from 1985 to 2002: results from INTERGENE and GOT-MONICA. Eur. J. Cardiovasc. Prev.
Rehabil. 12, 115–125 (2005).
73. Wietlisbach, V., Paccaud, F., Rickenbach, M. & Gutzwiller, F. Trends in cardiovascular risk
factors (1984–1993) in a Swiss region: results of three population surveys. Prev. Med. 26,
523–533 (1997).
74. Leiviskä, J. et al. What have we learnt about high-density lipoprotein cholesterol
measurements during 32 years? Experiences in Finland 1980–2012. Clin. Chim. Acta 415,
118–123 (2013).
75. Yokoyama, S. et al. High-density lipoprotein levels have markedly increased over the past
twenty years in Japan. J. Atheroscler. Thromb. 21, 151–160 (2014).
76. Hardoon, S. L. et al. Assessing the impact of medication use on trends in major coronary
risk factors in older British men: a cohort study. Eur. J. Cardiovasc. Prev. Rehabil. 17,
502–508 (2010).
77. Carroll, M. D. et al. Trends in serum lipids and lipoproteins of adults, 1960–2002. J. Am.
Med. Assoc. 294, 1773–1781 (2005).
Acknowledgements This study was funded by a Wellcome Trust (Biomedical Resource &
Multi-User Equipment grant 01506/Z/13/Z) and the British Heart Foundation (Centre of
Research Excellence grant RE/18/4/34215). C.T. was supported by a Wellcome Trust Research
Training Fellowship (203616/Z/16/Z). The authors alone are responsible for the views
expressed in this Article and they do not necessarily represent the views, decisions, or policies
of the institutions with which they are affiliated.
Author contributions M.E. and G.D. designed the study and oversaw research. C.T., B.Z., H.B.
and R.C.L. led the data collection. The other authors contributed to study design; and
collected, reanalysed, checked and pooled data. C.T. analysed pooled data and prepared
results. C.T., E.G. and M.E. wrote the first draft of the manuscript with input from the other
authors.
Competing interests M.E. reports a charitable grant from the AstraZeneca Young Health
Programme, and personal fees from Prudential, Scor and Third Bridge, outside the submitted
work. The other authors declare no competing interests.
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/s41586-0202338-1.
Correspondence and requests for materials should be addressed to M.E.
Peer review information Nature thanks Frank Hu and Pekka Jousilahti for their contribution to
the peer review of this work.
Reprints and permissions information is available at http://www.nature.com/reprints.
Article
Extended Data Fig. 1 | Number of data sources by country. The colour indicates the number of data sources for each country used in the analysis. Countries and
territories that were not included in the analysis are coloured in grey.
Extended Data Fig. 2 | Number of data sources by region and year. The size of each circle shows the number of data sources for each region and year, and the
colours indicate the relative size of national, subnational and community data sources.
Article
Extended Data Fig. 3 | Change in age-standardized mean HDL cholesterol between 1980 and 2018 by region for women and men. The start of the arrow
shows the level in 1980 and the head shows the level in 2018. One mmol l−1 is equivalent to 38.61 mg dl−1.
Extended Data Fig. 4 | Change in age-standardized mean HDL and non-HDL cholesterol between 1980 and 2018 by region for women and men. One mmol l−1
is equivalent to 38.61 mg dl−1.
Article
Extended Data Fig. 5 | Age-standardized mean total cholesterol by country in 1980 and 2018 for women and men. One mmol l−1 is equivalent to 38.61 mg dl−1.
Extended Data Fig. 6 | Age-standardized mean HDL cholesterol by country in 1980 and 2018 for women and men. One mmol l−1 is equivalent to 38.61 mg dl−1.
Article
Extended Data Fig. 7 | Change per decade in age-standardized mean total cholesterol by country for women and men. One mmol l−1 is equivalent to
38.61 mg dl−1.
Extended Data Fig. 8 | Change per decade in age-standardized mean HDL cholesterol by country for women and men. One mmol l−1 is equivalent to
38.61 mg dl−1.
Article
Extended Data Fig. 9 | The association between mean LDL and non-HDL cholesterol in studies that measured lipids in a laboratory that had data for both
variables. Each data point is one study–age–sex group (n = 6,864). One mmol l−1 is equivalent to 38.61 mg dl−1.
Last updated by author(s): Mar 19, 2020
Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist.
Statistics
nature research | reporting summary
Corresponding author(s): Majid Ezzati
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested
A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.
Software and code
Policy information about availability of computer code
Data collection
Processing of secondary data was conducted using the statistical software R (version 3.6.0).
Data analysis
All analyses were conducting using the statistical software R (version 3.6.0). The code for estimation of mean risk factor trends is
available at www.ncdrisc.org.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers.
We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
- A description of any restrictions on data availability
October 2018
This is a data-pooling study that brings together more than 1000 disparate data sources and uses a Bayesian hierarchical model to estimate population risk factor
trends. Estimates of mean total, non-HDL and HDL cholesterol by country, year, and sex will be available from www.ncdrisc.org upon the publication of the paper.
Some of the input data sources are publicly available, for which we will add links in the final version of the paper. Others are the property of specific research groups
and agencies, for which we will provide contact information.
1
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences
Behavioural & social sciences
Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf
Behavioural & social sciences study design
All studies must disclose on these points even when the disclosure is negative.
Study description
We pooled and re-analysed population-based data that had measured blood lipids in adults to estimate trends in mean total, non-HDL
and HDL cholesterol from 1980 to 2018 for 200 countries and territories, using a Bayesian hierarchical model.
Research sample
We pooled data from 1,127 population-based studies of blood lipids conducted in 161 countries, with measurement of blood lipids in
over 102 million adults aged 18 years and older. Studies were representative of a national, subnational or community population.
Sampling strategy
We included data collected using a probabilistic sampling method with a defined sampling frame. We therefore included studies with
simple random and complex survey designs but excluded convenience samples.
Data collection
We used data on measured blood lipids to calculate mean total, non-HDL and HDL cholesterol. We excluded self-reported data.
Timing
We pooled data collected from 1980 to 2018. We also included national studies for the 3 years prior to 1980 (n=1), assigning them to
1980, so that they can inform the estimates in countries with slightly earlier national data.
Data exclusions
We excluded all data sources that included only hypercholesterolemia or dyslipidaemia diagnosis history or medication status without
measurement of cholesterol levels. We also excluded data sources on population subgroups whose lipid profile may differ systematically
from the general population, including:
• studies that had included or excluded people based on their health status or cardiovascular risk;
• studies whose participants were only ethnic minorities;
• specific educational, occupational, or socioeconomic subgroups, with the exception noted below;
• those recruited through health facilities, with the exception noted below.
nature research | reporting summary
Field-specific reporting
We used school-based data in countries, and in age-sex groups, where secondary school enrollment was 70% or higher. We used data
whose sampling frame was health insurance schemes in countries where at least 80% of the population were insured. Finally, we used
data collected through general practice and primary care systems in high-income and central European countries with universal
insurance, because contact with the primary care systems tends to be as good as or better than response rates for population-based
surveys. Our exclusion criteria were established at the initiation of the study to ensure all data were representative.
Non-participation
Our inclusion/exclusion criteria were designed to ensure participants of the surveys included were representative of the general
population from which each sample was drawn.
Randomization
Our study is descriptive, and we did not carry out experiments.
Reporting for specific materials, systems and methods
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
Materials & experimental systems
Methods
n/a Involved in the study
n/a Involved in the study
Antibodies
ChIP-seq
Eukaryotic cell lines
Flow cytometry
Palaeontology
MRI-based neuroimaging
Animals and other organisms
Clinical data
October 2018
Human research participants
2