International Journal of Obesity (2005) 29, 381–387

& 2005 Nature Publishing Group All rights reserved 0307-0565/05 $30.00
www.nature.com/ijo

PEDIATRIC HIGHLIGHT
Schema and parental bonding in overweight and
nonoverweight female adolescents
HM Turner1*, KS Rose2 and MJ Cooper2
1
School of Medicine, University of Southampton, UK; and 2Isis Education Centre, University of Oxford, Warneford
Hospital, Oxford, UK

OBJECTIVE: To investigate whether family functioning and cognitions in a group of overweight female adolescents differ
significantly from those in a group of normal weight female adolescents.
DESIGN: Cross-sectional study.
SUBJECTS: In all, 23 overweight female adolescents (mean age: 17.6 y, mean body mass index (BMI: 27.8 kg/m2), and 23
normal weight female adolescents (mean age: 17.7 y, mean BMI: 20.2 kg/m2).
MEASUREMENTS: The following self-report measures were completed: the Parental Bonding Inventory,1 the Young Schema
Questionnaire-short version,2 the Eating Attitudes Test,3 the Beck Depression Inventory4 and the Eating Disorder Belief
Questionnaire.5
RESULTS: Overweight female adolescents reported more negative self-beliefs and greater belief in schema relating to emotional
deprivation, fears of abandonment, subjugation and insufficient self-control. They also perceived their fathers as being
significantly more overprotective and significantly less caring. Within this group perceived level of maternal care correlated
negatively with negative self-beliefs and schema.
CONCLUSIONS: Overweight female adolescents show some of the cognitive features associated with the development of an
eating disorder. However, positive parent–child relationships may serve to protect overweight adolescents from developing
clinical eating disorders and from psychological distress later in life.
International Journal of Obesity (2005) 29, 381–387. doi:10.1038/sj.ijo.0802915

Keywords: adolescents; schema; negative self-beliefs

Introduction sectional school-based survey exploring psychosocial con-

The prevalence of obesity is on the increase worldwide.6 cerns and health-compromising behaviours, Neumark-Sztai-
Studies examining the prevalence of obesity in childhood ner and colleagues7 found that psychosocial concerns such
and adolescence have reported a significant increase in the as emotional well-being, suicidal ideation, peer concerns and
number of children and young people who are overweight or future job concerns did not differ significantly between
obese. In the USA, it has been reported that the number of overweight and nonoverweight adolescents. However, their
overweight children has doubled and the number of over- results did suggest that overweight adolescents were more
weight adolescents has trebled since 1980.6 likely than their normal weight counterparts to express
Alongside the many physical complications of obesity, weight-specific concerns and engage in behaviours such as
researchers have explored whether overweight children and chronic dieting and binge eating. These results support the
adolescents are more at risk of adverse psychological, social earlier findings of Pastore et al,8 who found that those who
and economic outcomes. This includes the presence of were overweight did not differ from their normal weight
eating disorder-related symptoms, concerns and behaviours. peers in relation to reported levels of self-esteem or anxiety,
To date the findings remain ambiguous. In their cross- although they did report more abnormal eating attitudes.
However, Mellin et al9 report contrasting findings from
their cross-sectional school-based survey. They found that
*Correspondence: Dr H Turner, Hampshire Partnership NHS Trust Eating overweight adolescents reported experiencing higher levels
Disorders Service, Eastleigh Community Enterprise Centre, Unit 3, Barton of psychological distress compared with their normal weight
Park, Eastleigh, Hampshire, UK.
peers. In particular, they rated their school performance and
E-mail: hmt@soton.ac.uk
Received 1 June 2004; revised 30 November 2004; accepted 24 January educational future lower and reported higher levels of
2005 emotional distress. Importantly, Mellin and her colleagues
 Schema and parental bonding in adolesecents
HM Turner et al
382
also found that familial connectedness and parental expec- as risk factors for the develop of an eating disorder. In
tations served as important protective factors against the particular, we were unable to find a study in overweight
negative stereotyping and discrimination that overweight adolescents that explored the cognitions that have been
teenagers often face. These findings led the authors to linked to eating disorders including schema that may be
conclude that positive relationships between parents and associated with eating disorder psychopathology. As noted
adolescents who are overweight may have a positive impact by Waller et al21 in their review of the literature on schemas
on the health-related behaviours and psychosocial well- in the eating disorders, children and adolescents represent a
being of the adolescent. group whose cognitions are particularly in need of further
Mellin and colleagues’ findings were confirmed in a study research attention.
exploring family functioning and psychological adjustment Schema have been investigated in, and associated with,
in overweight children aged between 10 and 16 y.10 This anorexia nervosa, bulimia nervosa and mixed groups of
study found that overweight youngsters reported lower eating disordered patients.22–24 High levels of belief in
scores on physical appearance, athletic competence, social dysfunctional schema are typically associated with eating
acceptance, body esteem and global self-worth compared to disorder diagnosis. Two measures of schema have been used
their normal weight peers. Their findings also revealed that in studies: the negative self-beliefs subscale of the Eating
weight-specific parental overconcern negatively correlated Disorder Belief Questionnaire5 and the Young Schema
with social and emotional adjustment of overweight chil- Questionnaire.2 Assumptions have also been measured using
dren. Consequently, the authors concluded that parental the three remaining subscales of the EDBQ. Perceived
overconcern may place overweight children at increased risk parental relationships have been measured using the Par-
of experiencing psychosocial problems and called for further ental Bonding Instrument,1 and there is a relationship
investigation into the relationship between family function- between PBI scores and schema in studies investigating
ing and the social and emotional adjustment of the eating disorder-related attitudes and behaviour (eg Turner,
overweight child. Rose and Cooper, unpublished data).
Alongside the adverse psychosocial outcomes already The current study therefore had a number of aims: first, to
commented upon, a history of obesity has been identified explore the psychological characteristics (including eating
as a specific risk factor for the development of a clinical disorder-related symptoms, depressive symptoms and sche-
eating disorder.11 Some studies12 have suggested that girls ma) of a group of overweight female adolescents and
who are physically larger than their peers are at a higher risk compare these findings to a group of normal weight female
of developing problems with their eating, while another adolescents; second, to explore whether group differences
study13 found that a higher body mass index and early exist in relation to perceived parental relationships; and
pubertal development predict disordered eating in a com- third, to explore whether any within-group associations exist
munity sample of adolescent girls. between perceived parental relationships and psychological
Set within the wider context, there is now an extensive factors including some of the schema that have previously
theoretical and empirical literature documenting risk factors been found to be associated with an eating disorder.
for the development of eating disorders.14 In their integrated
community-based case–control series, Fairburn and collea-
gues identified a number of risk factors for the development
of eating disorders including perfectionism, low self-esteem, Methods
eating and weight concerns, and affective disorder.15–17 Participants
Obesity in childhood was a significant risk factor for the In all, 367 17- to 18-y-old female adolescents were recruited
development of bulimia nervosa.15 It is also widely docu- through local secondary schools. Of this group, 345 gave
mented that unhealthy family relationships may be of information relating to their weight and height. Within this
aetiological importance in eating disorders.18,19 group, 232 had a body mass index (BMI) of 20 kg/m2 or over.
Burrows and Cooper20 set out explicitly to explore the The top 10% and the bottom 10% of this group were selected
presence of possible risk factors in the development of eating and used in comparison analyses.
disorders in preadolescent girls. In their study comparing a
group of overweight girls with a group of average weight
girls, the findings indicated that overweight girls showed
some of the psychological features associated with the Index group. The index group consisted of the top 10%
development of an eating disorder including increased (n ¼ 23) of the group. Participants within this group were
concerns about weight, shape and eating. This group also overweight, that is, had a BMI of 25 kg/m2 or over.
had lower levels of self-esteem and more symptoms of
depression than average weight girls. However, to date no
studies have specifically explored whether overweight Control group. The control group consisted of the bottom
female adolescents, as opposed to children, report higher 10% (n ¼ 23) of the group. Participants in this group had a
levels of concern regarding factors that have been identified BMI within the normal range.

International Journal of Obesity

 Schema and parental bonding in adolesecents
HM Turner et al
383
Measures levels of internal consistency, reliability and validity com-
Demographics. Girls were asked to give their date of birth, pared with the longer version.
from which their age was calculated.

Height and weight. Self-reports of height and weight were Parental Bonding Instrument (PBI)
used to calculate BMI. The PBI1 measures perception of parental bonding during
the first 16 y of life. Each of the 25 items is rated on a four-
point Likert scale from ‘very like my parent’ to ‘very unlike
Eating Attitudes Test (EAT) my parent’. The items are grouped into two scales: ‘care’ and
This widely used 40-item Self-Report Questionnaire3 pro- ‘overprotection’, and the measure is completed twice, once
vides a global score of eating disturbance. Each item is rated for each parent. Unhealthy parent–child bonding is indi-
on a six-point Likert scale (1 ¼ never, 6 ¼ always). Garner and cated by lower ‘care’ scores and higher ‘overprotection’
Garfinkel3 suggest that scores Z30 may indicate a clinical scores.1 The PBI has good psychometric properties.29
eating disorder. The EAT has good psychometric properties
including reliability (Cronbach’s alpha coefficient ¼ 0.94 for
combined anorexia nervosa and control samples) and Procedure
validity (range of correlations with other eating disorder Local Research Ethics Committees’ approval was obtained in
measures ¼ 0.42–0.81).25 the areas from which the participants were recruited. Letters
and an information sheet outlining the research were sent to
19 schools and a follow-up phone call determined whether
Beck Depression Inventory (BDI) or not they were willing to take part. Participating schools
The BDI4 is a 21-item self-report measure giving a global were visited and the study was presented to the 17- to 18-y-
score of depressive symptomatology. Each item is answered old females. The girls were provided with an information
on a four-point Likert scale, with higher scores indicating sheet to keep and, where required by local research ethics
more severe symptomatology. The BDI has good psycho- committees, a letter informing their parents of the nature of
metric properties including high concurrent validity and the study. Girls were informed of where and when to return
internal consistency (alpha coefficients ¼ 0.73 and 0.86, should they wish to complete the questionnaires. At the
respectively).26 appointed time girls were each given an unmarked envelope
containing the pack of questionnaires, which they were
asked to return to a box at the front of the room on
Eating Disorder Belief Questionnaire (EDBQ) completion. The researcher was available during and follow-
The EDBQ5 is a 32-item Self-Report Questionnaire that ing each episode of data collection to deal with any concerns
assesses assumptions and beliefs associated with eating or questions. All girls were also provided with an informa-
disorders. Each item is rated on a scale from 0 ¼ ‘I do not tion sheet giving relevant organisations’ addresses and
usually believe this at all’ to 100 ¼ ‘I am usually completely phone numbers, should the girls experience any concerns
convinced that this is true’. The EDBQ has four subscales: regarding eating, weight and shape.
negative self-beliefs, weight and shape as a means to self-
acceptance, weight and shape as a means to acceptance by
others and control over eating. Research with women5
Data analysis
revealed that the EDBQ has good reliability (Cronbach
The data were analysed using SPSS.30 Checks for normality of
coefficient alphas for each factor ranged from 0.86 to 0.94)
distribution was made by the group using the Kolmogrov–
and construct validity (correlations between the four
Smirnov test. Where data failed to meet the criteria for
subscales and other eating disorders measures were signifi-
parametric tests, nonparametric equivalents were applied.
cant at Po0.01).

Young Schema Questionnaire-Short Form (YSQ-S) Results

The YSQ-S2 is a 75-item self-report measure developed to There was no significant difference between the two groups
identify 15 early maladaptive schemas. Each item is rated on in relation to age (t ¼ 0.756, df ¼ 43, NS). However, as
a Likert scale from 1 ¼ ‘completely untrue of me’ to expected, the groups did differ significantly in relation to
6 ¼ ‘describes me perfectly’. A total YSQ-S score can be BMI; the index group had a significantly higher BMI than
obtained by summing all scores on the YSQ-S, whereas the control group (t ¼ 14, df 22, Po0.0001). Mean age and
adding all scores in each schema can identify individual BMI are shown in Table 1.
schema scores. Young2 suggests that a schema with two or There were no significant differences between the two
more items rated as 5 or 6 is ‘clinically significant’. groups on the EAT, BDI, EDBQ-SA, EDBQ-OA or EDBQ-C.
Research27,28 suggests that the short version has very similar However, the index group scored significantly higher than

International Journal of Obesity

 Schema and parental bonding in adolesecents
HM Turner et al
384
Table 1 Mean age and BMI of the index and control group Table 2 Summary data for the BDI; EAT, EDBQ, PBI and YSQ-S scales

Index group (n ¼ 23) Control group (n ¼ 23) Index group (n ¼ 23) Control group (n ¼ 23)

Mean (s.d.) Mean (s.d.) Mean (s.d.) Mean (s.d.)

Age 17.6 (0.42) 17.7 (0.43) BDI 14.7 (11.6) 10.0 (9.0)
BMI (kg/m2) 27.8 (2.6) 20.2 (0.10) EAT 14.91 (3.0) 14.13 (12.8)

EDBQ subscales
EDBQ-NS 310.87 (246.1) 161.74 (152.0)
EDBQ-SA 328.2 (142.2) 318 (160)
the control group on the negative self-beliefs subscales of the EDBQ-AO 233.48 (250.9) 173.48 (211.3)
EDBQ (t ¼ 2.4, df ¼ 36.6, Po0.018). EDBQ-C 123.04 (146) 135 (138.0)
Significant differences between the two groups were found
PBI subscales
on the following PBI subscales: father care (t ¼ 2.4, df ¼ 31.8,
Maternal care 28.8 (7.0) 30.7 (4.1)
Po0.02) and father overprotection (t ¼ 2.5, df ¼ 34.8, Paternal care 23.39 (8.7) 28.38 (4.1)
Po0.016). Girls in the overweight group perceived their Maternal overprotection 12.91 (7.6) 9.55 (4.3)
fathers to be significantly more overprotecting and signifi- Paternal overprotection 13.52 (8.2) 8.48 (4.5)
cantly less caring than those in the normal weight group.
YSQ subscales
In relation to the YSQ-S, the groups scored significantly Emotional deprivation 11.43 (6.3) 7.43 (2.5)
different on the following subscales: emotional depriva- Abandonment 13.91 (7.3) 8.87 (4.2)
tion (t ¼ 2.7, df ¼ 29, Po0.009), abandonment (t ¼ 2.8, Mistrust/abuse 11.52 (6.1) 9.35 (4.2)
df ¼ 35.2, Po0.007), subjugation (t ¼ 2.1, df ¼ 30.2, Social isolation 11.83 (7.1) 8.83 (4.6)
Defectiveness/shame 11.48 (7.5) 8.2 (3.9)
Po0.037), insufficient self-control/self-discipline (t ¼ 1.9, Failure 12.28 (6.7) 9.78 (4.2)
df ¼ 42, Po0.059) and the YSQ-S total scores (t ¼ 2.4, Dependence/ 10.65 (6.0) 8.39 (3.6)
df ¼ 32, Po0.020). Girls in the index group scored signifi- incompetence
Vulnerability to harm/ 10.13 (5.6) 8.35 (3.6)
cantly higher on these subscales compared with the control
illness
group. Enmeshment 8.91 (4.7) 6.89 (2.0)
Mean scores and standard deviations on questionnaire Subjugation 11.00 (5.5) 8.26 (2.4)
measures are shown in Table 2. Self-sacrifice 15.83 (5.2) 13.0 (5.0)
Emotional inhibition 10.13 (4.7) 10.43 (5.1)
Unrelenting standards 17.35 (6.3) 16.35 (4.7)
Entitlement 11.39 (4.3) 10.39 (3.3)
Insufficient self-control/ 14.70 (5.7) 11.7 (4.7)
self-discipline
Within-group correlations
Total YSQ-S score 182.54 (63.6) 146.2 (31.3)
The results of the within-group correlations are shown in
Table 3. BDI ¼ Beck Depression Inventory; EAT ¼ Eating Attitudes Test; EDBQ-
Within the control group, the ‘mother care’ subscale of the NSB ¼ Eating Disorder Beliefs Questionnaire-Negative Self-Belief; EDBQ-
SA ¼ Eating Disorder Beliefs Questionnaire-self-acceptance; EDBQ-AO ¼ Eating
PBI correlated negatively with the EAT (Po0.01), EDBQ-AO Eating Disorder Beliefs Questionnaire-Acceptance by Others; EDBQ-C ¼ Eating
(Po0.05), EDBQ-C (Po0.01), and the abandonment subscale Disorder Beliefs Questionnaire-Control Overeating; PBI ¼ Parental Bonding
of the YSQ-S (Po0.05). Within the index group, BMI cor- Instrument; YSQ-S ¼ Young Schema Questionnaire-Short Form.
related negatively with insufficient self-control (Po0.05).
The ‘mother care’ subscale correlated negatively with the
EDBQ-NSB, (Po0.05) and the following YSQ subscales:
emotional deprivation (Po0.01), mistrust/abuse (Po0.01), Discussion
emotional inhibition (Po0.01), unrelenting standards This study aimed to explore psychological characteristics,
(Po0.01) and entitlement (Po0.05). However, positive including factors related to eating disorders, and perceived
correlations were found between the ‘mother overprotec- parental bonding in a group of overweight female adoles-
tion’ subscale of the PBI and the EDBQ-NSB, EDBQ-SA, cents. The study also aimed to compare these findings with a
EDBA-AO and the BDI (all, Po0.05), as well as with the group of non-overweight female adolescents, and explore
following YSQ-S subscales: emotional deprivation (Po0.01), whether any within-group correlations existed between
abandonment (Po0.05), mistrust/abuse (Po0.01), social these two sets of variables.
isolation (Po0.05), defectiveness/shame (Po0.05), vulner- The findings indicate that there were no significant
ability to harm (Po0.05), subjugation(Po0.01), emotional differences between the two groups in relation to eating
inhibition (Po0.05) and unrelenting standards (Po0.05). disorder symptoms, depressive symptoms and cognitions
There were no significant correlations between ‘father care’ relating to weight and shape as a means of acceptance by self
or ‘father overprotection’ and the psychological variables, or others. However, those in the index group obtained
including schema. higher score on negative self-beliefs, and reported greater

International Journal of Obesity

 Schema and parental bonding in adolesecents
HM Turner et al
385
Table 3 Correlations between the EAT, BDI, EDBQ YSQ-S, and the PBI and BMI

BMI (kg/m2) PBIFmaternal overprotection PBIFmaternal care

Index Control Index Control Index Control

EAT 0.316 0.017 0.303 0.027 0.003 0.625**

BDI 0.110 0.151 0.522* 0.067 0.383 0.019

EDBQ subscales
EDBQ-NSB 0.241 0.068 0.486* 0.057 0.453* 0.106
EDBQ-SA 0.169 0.193 0.456* 0.033 0.302 0.392
EDBQ-AO 0.206 0.165 0.437* 0.130 0.296 0.432*
EDBQ-C 0.299 0.078 0.377 0.071 0.145 0.547**

YSQ-S subscales
Emotional deprivation 0.225 0.112 0.608** 0.067 0.640** 0.183
Abandonment 0.282 0.242 0.455* 0.051 0.359 0.535*
Mistrust/abuse 0.102 0.177 0.624** 0.333 0.620** 0.036
Social isolation 0.230 0.215 0.418* 0.398 0.381 0.203
Defectiveness/shame 0.231 0.020 0.486* 0.030 0.350 0.351
Failure 0.324 0.037 0.349 0.248 0.213 0.246
Dependence/incompetence 0.329 0.321 0.340 0.009 0.196 0.012
Vulnerability to harm/illness 0.230 0.022 0.477* 0.321 0.219 0.184
Enmeshment 0.014 0.270 0.156 0.042 0.181 0.134
Subjugation 0.124 0.405 0.549** 0.173 0.276 0.259
Self-sacrifice 0.213 0.226 0.120 0.087 0.154 0.251
Emotional inhibition 0.197 0.040 0.467* 0.207 0.650** 0.126
Unrelenting standards 0.257 0.014 0.460* 0.041 0.604** 0.177
Entitlement 0.204 0.312 0.253 0.227 0.433* 0.038
Insufficient self-control/self-discipline 0.496* 0.192 0.278 0.170 0.381 0.407

n ¼ 23 per group. *Po0.05 level, **Po0.01 level, two-tailed. EAT ¼ Eating Attitudes Test; BDI ¼ Beck Depression Inventory; EDBQ-NSB ¼ Eating Disorder Belief
Questionnaire-Negative Self-Belief; EDBQ-SA ¼ Eating Disorder Belief Questionnaire-Self-Acceptance; EDBQ-AO ¼ Eating Disorder Belief Questionnaire-Acceptance
by Others; EDBQ-C ¼ Eating Disorder Belief Questionnaire-Control Overeating; YSQ-S ¼ Young Schema Questionnaire-Short Form; BMI ¼ body mass index;
PBI ¼ parental bonding instrument.

belief in schema relating to emotional deprivation, fears of that eating disorder symptoms are not increased or elevated
abandonment, subjugation and insufficient self-control. The in this group. However, the current findings do indicate that
overweight adolescents also perceived their fathers as being some of the cognitions associated with eating disorders (and
significantly more overprotective and significantly less perhaps with other psychopathology) are present in the
caring than those in the control group. These findings overweight group and not in the normal weight group. This
suggest that while being overweight is not necessarily suggests that some potential risk factors for eating disorders
associated with increased distress and increase in eating are indeed present in an overweight group. One reason for
disorder-related symptoms, it is associated with increased the lack of overt behavioural symptoms and distress in our
presence of some of the cognitions that have been associated sample might be that these have not yet had time to develop,
with eating disorders including negative self-beliefs and and that cognitions act as a precursor. This suggestion would
schema. Perceived poorer family functioning, particularly in be compatible with the development of disorder and distress
relation to fathers, is also associated with being overweight. in cognitive theory.31 It would be useful to conduct some
In the normal weight group, perceived level of maternal longitudinal research to follow-up groups such as these and
care correlated negatively with eating disorder symptoms assess whether the cognitions do develop into overt distress
and abandonment schema on the YSQ-S. In the overweight and eating disorder-related behaviour. The findings concern-
group, a positive relationship was found between maternal ing perceived parent–child relationship add further to the
overprotection, and the presence of cognitions associated literature concerned with parental relationships in over-
with eating disorders (on both EDBQ and YSQ-S subscales). weight adolescents. Previous research findings9,10 have
The level of maternal care appeared to correlate negatively suggested that a positive parent–child relationship may
with the level of negative self-beliefs, as well as with YSQ-S serve to reduce psychosocial risk in overweight adolescents.
subscales relating to schemas such as unrelenting standards The present study supports this and suggests that in the
and emotional deprivation. overweight group cognitions may also be associated with
The findings are consistent with those who have failed to perceived parental care and protection. In addition, the
find increased distress in overweight adolescents7 and negative relationship found between perceived maternal
contrary to findings in overweight children,20 suggesting care, and eating attitudes and fears of abandonment, in the

International Journal of Obesity

 Schema and parental bonding in adolesecents
HM Turner et al
386
control group, lends support to the idea that positive potential to further our understanding of how eating
parental relationships may serve to protect adolescent disorders might develop.
females from developing an eating disorder including its To conclude, the findings from the present study suggest
cognitive as well as its more behavioural features. This that overweight adolescents show some of the cognitive
suggestion is consistent with the idea that early experiences features associated with the development of an eating
within the family are important in the development of disorder. The study also highlights the potential role of
schema.32 parent–child relationships in protecting overweight adoles-
The findings of the present study have a number of cents from experiencing such cognitions. Future research
potential implications. Firstly, if attempts are to be made to exploring the precise nature of the relationship between
reduce the risk of psychological distress in overweight these variables is required if we are to minimise the potential
adolescents, it is important that professionals and parents risk of overweight children developing clinical eating
are aware that overweight adolescents may be at increased disorders and psychological distress in later life.
risk of developing a poor self-view, characterised by negative
beliefs regarding the self, even in the absence of overt signs
of psychological distress. To this end, it is important that
issues of self-esteem are addressed at an early age in this References
population. Linked to this is the potentially complex nature
1 Parker G, Tupling H, Brown L. A Parental Bonding Instrument. Br
of the parental role. Although the present study does not J Med Psychol 1979; 52: 1–10.
allow firm conclusions regarding causality to be drawn, it is 2 Young JE. The Young Schema Questionnaire: Short Form. http://
possible that poor family relationships, in particular, high home.sprynet.com/sprynet/schema/ysqs1.htm. 1998.
levels of paternal overprotection and low levels of paternal 3 Garner DM, Garfinkel PE. The Eating Attitudes Test: an index of
the symptoms of anorexia nervosa. Psychol Med 1979; 9: 273–279.
care, may lead to the development of negative self-beliefs 4 Beck AT. Beck Depression Inventory. The Psychological Corporation:
and schema in overweight adolescents. For example, a child San Antonio, TX; 1996.
growing up in an environment characterised by low levels of 5 Cooper M, Cohen-Tovee E, Todd G, Wells A, Tovee M. The Eating
paternal care may develop a sense of emotional deprivation, Disorder Belief Questionnaire: preliminary development. Behav
Res Ther 1997; 35 (4): 381–388.
that is, a sense of feeling uncared for and unloved. In line 6 World Health Organisation. Global Strategy on Diet, Physical
with this, it is also possible that positive parental relation- Activity and Health http//www.who.int/dietphysicalactivity/pub-
ships, such as high levels of maternal care, may serve to lications/facts/obesity/en/ 2003.
protect adolescent females from such an outcome. This 7 Neumark-Sztainer D, Story M, French S, Hanna PJ, Resnick MD,
Blum RW. Psychological concerns and health-compromising
(tentative) causal link suggests that it may be crucial to behaviours among overweight and non-overweight adolescents.
increase parental awareness of the possible pathways by Obes Res 1997; 5: 237–249.
which both mothers and fathers might serve to increase or 8 Pastore DR, Fisher M, Friedman S. Abnormalities in weight status,
decrease the risk of overweight children developing psycho- eating attitudes and eating behaviors among urban high school
students. J Adolesc Health 1996; 18: 312–319.
logical disorders later in life. Perhaps most importantly, it is 9 Mellin AE, Neumark-Sztainer D, Story M, Ireland M, Resnick M.
possible that positive parent–child relationships may serve to Unhealthy behaviours and psychosocial difficulties among over-
protect overweight females from developing unhealthy weight adolescents: the potential impact of familial factors.
negative self-beliefs and schema. J Adolesc Health 2002; 31: 145–153.
10 Stradmeijer M, Bosch J, Koops W, Seidell J. Family functioning
The present study highlights a number of areas for future and psychosocial adjustment in overweight youngsters. Int J
research. As previously mentioned, the use of correlations in Eating Disord 2000; 27: 110–114.
the present study prevents any firm conclusions regarding 11 National Institute of Clinical Excellence. Predicting the Outcome of
causation being drawn. One line of further work would be to Treatment and Recovery Review. Appendix 11. Eating Disorders: Core
Interventions in the Treatment and Management of Anorexia Nervosa,
explore the causative nature of the relationships between Bulimia Nervosa and Related Eating Disorders. National Collaborat-
schemas, parent–child relationships and eating disorder ing Centre for Mental Health Commissioned by the National
psychopathology in overweight female adolescents, possibly Institute of Clinical Excellence, The British Psychological Society
through drawing upon models of mediating and moderating and Gaskell: London; 2004.
12 Attie I, Brooks-Gunn J. Development of eating problems in
influences.33 This would enable the role of schema in adolescent girls: a longitudinal study. Dev Psychol 1989; 25:
relation to parental bonding and eating disorder-related 70–79.
symptoms to be explored more fully. Longitudinal studies 13 Keel PK, Fulkerson JA, Leon GR. Disordered eating precursors in
have a particularly valuable contribution to make and would pre and early adolescent girls and boys. J Youth Adolesc 1997; 26:
203–216.
permit rigorous testing of some of the ideas highlighted 14 Streigel-Moore RH, Cachelin FM. Etiology of eating disorders in
including the suggestion that cognitions may be precursors women. Counsel Psychologist 2001; 29: 635–661.
of later distress and eating disorder psychopathology. 15 Fairburn CG, Welch SL, Doll HA, Davies BA, O’Connor ME. Risk
Schema have an important role in recent theoretical writings factors for bulimia nervosa: a community-based case–control
study. Arch Gen Psychiatry 1997; 54: 509–517.
on eating disorders21,34 and their presence and nature in 16 Fairburn CG, Cooper Z, Doll HA, Welch SL. Risk factors for
related conditions, including overweight adolescents and anorexia nervosa: three integrated case–control comparisons.
children, warrants further investigation. Such work has the Arch Gen Psychiatry 1999; 56: 468–476.

International Journal of Obesity

 Schema and parental bonding in adolesecents
HM Turner et al
387
17 Fairburn CG, Cooper Z, Doll HA, Welch SL. Risk factors for binge Allison DB (ed) Handbook of Assessment Methods for Eating
eating disorder: a community-based case–control study. Arch Gen Behaviours and Weight-Related Problems. Measures, Theory &
Psychiatry 1999; 56: 468–476. Research. Sage Publications: London; 1995.
18 Mallinckrodt B, McCreary BA, Robertson AK. Co-occurrence of 26 Beck AT, Steer RA, Garbin MG. Psychometric properties of the
eating disorders and incest: the role of attachment, family Beck Depression Inventory: twenty-five years of evaluation. Clin
environment, and social competencies. J Counsel Psychol 1995; Psychol Rev 1988; 8: 77–100.
42: 178–186. 27 Waller G, Meyer C, Ohanian V. Psychometric properties of the
19 Minuchin S, Roseman BL, Baker L. Psychosomatic Families: long and short versions of the Young Schema Questionnaire: core
Anorexia Nervosa in Context. Harvard University Press: Cambridge, beliefs among bulimic and comparison women. Cogn Ther Res
MA; 1978. 2001; 25: 137–147.
20 Burrows A, Cooper MJ. Possible risk factors in the development of 28 Stopa L, Thorne P, Waters A, Preston J. Are the short and long
eating disorders in overweight pre-adolescent girls. Int J Obes Relat forms of the Young Schema Questionnaire comparable and how
Metab Disord 2002; 26: 1–6. well does each version predict psychopathology scores? J Cogn
21 Waller G, Kennerley H, Ohanian V. Schema-focused cognitive Psychother 2001; 15: 253–272.
behaviour therapy with the eating disorders. In: LP Riso, PT du 29 Mackinnon AJ, Henderson AS, Scott R, Duncan-Jones P. The
Toit and JE Young (eds) Cognitive Schemas and Core Beliefs in Parental Bonding Instrument (PBI): an epidemiological study in a
Psychiatric Disorders: A Scientist-Practitioner Guide. American general population sample. Psychol Med 1989; 19: 1023–1034.
Psychological Association: New York, (in press). 30 SPSS. Advanced Statistics 7.5. SPSS Inc.: Chicago; 1997.
22 Cooper M, Turner H. Underlying assumptions and core beliefs in 31 Beck AT, Rush AJ, Shaw BF, Emery G. Cognitive Therapy of
anorexia nervosa and dieting. Br J Clin Psychol 2000; 39: 215–218. Depression. Guildford: New York; 1979.
23 Cooper M, Hunt J. Core beliefs and underlying assumptions 32 Young JE, Klosko JE, Weishaar ME. Schema Therapy: A Practitioner’s
in bulimia nervosa and depression. Behav Res Ther 1998; 36: Guide. The Guildford Press: New York; 2003.
895–898. 33 Baron RM, Kenny DA. The moderator–mediator variable distinc-
24 Leung N, Waller G, Thomas G. Core beliefs in anorexic and tion in social psychological research: conceptual, strategic and
bulimic women. J Nerv Ment Dis 1999; 187: 736–741. statistical considerations. J Pers Soc Psychol 1986; 51: 1173–1182.
25 Williamson DA, Anderson DA, Jackman LP, Jackson SR. Assess- 34 Cooper MJ, Wells A, Todd G. A cognitive model of bulimia
ment of eating disordered thoughts, feelings and behaviours. In: nervosa. Br J Clin Psychol 2004; 16: 1–16.

International Journal of Obesity