To examine levels of eating disorder behaviours and cognitions of young women with obesity in the Australian Capital Territory, Australia and assess the impact upon psychological status.
General population cross-sectional survey.
A total of 4891 young women from the community aged 18–42 years, of which 630 were in the obese weight range.
Body mass index (BMI), eating disorder psychopathology (eating disorder examination questionnaire), and psychological distress (K-10).
Women with obesity had significantly higher levels of dietary restraint, eating concern, weight concern, shape concern, binge eating, misuse of diuretics, use of diet pills and fasting compared to other women in the community. These eating disorder cognitions and behaviours were associated with increased levels of psychological distress. In women with obesity, eating concern, weight concern, shape concern, dietary restraint and decreased age predicted psychological distress in a multivariate model. Among other women in the community, behaviours such as laxative misuse, ‘hard’ exercise and subjective bulimic episodes also contributed to the model predicting psychological distress.
As disordered eating psychopathology is high in young obese women and negatively impacts upon psychological status, obesity prevention and treatment should consider eating disorder psychopathology and mental health outcomes.
Both obesity and eating disorders (EDs) are significant health problems in the Australian society. It has been shown that levels of obesity are currently increasing, with the recent 2001 Australian National Health Survey finding approximately 14% of Australians over the age of 15 are obese.1 The lifetime community prevalence of EDs has remained relatively stable with the lifetime prevalence of anorexia nervosa (AN) estimated to be approximately 0.5% and bulimia nervosa (BN) 1–2%. However, the prevalence of EDs not meeting the formal criteria for AN or BN, such as ‘eating disorders not otherwise specified’ (EDNOS) that includes binge eating disorder (BED), is considerably higher (up to 5%) and these disorders are associated with marked impairment in functioning.2, 3, 4, 5, 6 While often regarded as distinct problems, EDs and weight disorders have many common characteristics, including dieting behaviour, binge eating, poor body image and psychosocial difficulties.4, 7, 8, 9, 10, 11, 12, 13, 14
It has been well established that EDs are associated with high levels of psychiatric co-morbidities including depression, anxiety and personality disorders.7 In contrast, research has generally found that levels of general psychopathology, such as depression and anxiety are no different for people who are obese compared to those who are not obese, in the community. However, this differs for various subgroups within the obese population, such as those seeking treatment, those with co-morbid binge eating and poor body image and those with severe levels of obesity, all whom experience higher levels of psychological distress.15, 16, 17, 18
Many previous studies on the interface between weight and EDs have relied on treatment seeking, often small clinical samples of people with obesity, for example,15, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 or have focused on BED20, 21, 22, 23, 24, 25, 26, 27 or ED behaviours only.4 They have found that binge eating contributes to obesity, that obese people with BED have higher depression, anxiety, eating and weight concerns and poorer quality of life than in other obese subjects; also that poor body image evaluation and negative self schemas relate to depression in obese people and that body image beliefs relate to level of obesity (heavier women are less satisfied with appearance). Furthermore, although there is good evidence that ED symptoms are related to depression and anxiety, we are not aware of any study which identifies specific ED behaviours and cognitions that most affect the mental health status of community women with obesity. Thus, this study aims to examine ED psychopathology comprehensively (i.e. assessment of binge eating, compensatory weight control behaviours and cognitions) in a large community-based sample representative of young adult obese women in the Australian Capital Territory (ACT), Australia. It also aimed to examine the way in which the psychological status of women with obesity may be influenced by eating psychopathology by identifying specific ED behaviours and cognitions that contribute to psychological distress. We hypothesized that ED behaviours and cognitions would be more frequent in women with obesity compared to other women in the sample, and that in women with obesity, higher levels of ED psychopathology would be associated with increasing psychological distress. We had no specific hypotheses concerning which particular aspects of ED psychopathology will be contributing to the psychological distress as this part of the study was exploratory.
Design and participants
The research was conducted as part of the Health and well-being of female ACT residents study, a large-scale epidemiological study of disability and health-service utilization associated with the more commonly occurring (bulimic-type) EDs among young adult women in the community. At the first phase of the study, self-report questionnaires were posted to a sample of 10 000 female residents of the ACT region of Australia (population 323 000), a highly urbanized region which includes the capital city of Australia, Canberra. Participants aged 18–42 were selected randomly from the electoral roll.28 A total of 5255 individuals responded to the phase 1 questionnaire representing a response rate of 57.1%.28 The ACT Human Health Research Ethics Committee provided ethics approval for the project and all participants provided informed written consent.
Body mass index (BMI, kg/m2) was calculated from self-reported height and weight. Previously, we found a very high correlation (r=0.97) between BMI calculated in this way and BMI calculated according to measured height and weight.29 Obesity was classified as BMI30.0 using the classification scheme outlined by the World Health Organization and utilized in the Clinical Practice Guidelines for the Management of Overweight and Obesity in Adults.30
Eating disorder examination questionnaire
The eating disorder examination questionnaire (EDE-Q)31 is a 36-item self-report measure derived from the eating disorders examination interview (EDE).32 The EDE-Q focuses on the past 28 days and is scored using a seven-point, forced-choice, rating scheme to measure individual items of attitudinal aspects of EDs. Subscale scores – relating to dietary restraint, eating concerns, concerns about weight and concerns about shape – and a global score, are derived from the 22 items addressing these attitudinal aspects of ED psychopathology. Frequencies of ED (overeating and compensatory) behaviours are also assessed in terms of the number of episodes occurring during the past 4 weeks. Reliability and validity of the EDE-Q has been demonstrated in both community and clinical samples (with the exception of overestimation of the binge eating item).29, 33
Kessler-10 item distress scale
General psychological distress (depression and anxiety) were assessed with the Kessler-10 item distress scale (K-10). The K-10 has robust psychometric properties, and is designed to detect cases of anxiety and affective disorders in the general population,34 and has been used in our previous ED research.35 The frequency of each of the 10 depressive or anxiety symptoms is measured on a scale from one to five. In the present study coding of the response options was such that total scale scores ranged from 10 to 50, with higher scores indicating greater symptomatology.
Data were inspected for normality, and parametric and non-parametric tests were used accordingly. Differences between obese and non-obese groups on categorical variables were compared using χ2 tests, whereas between-group differences on continuous variables were compared using the Mann–Whitney U-test (Z). To assess the relationship between variables Spearman's ρ correlation coefficient (rs) was utilized, and following the normalization of scores using the method of BLOM, backward selection linear regression was performed to examine the predictive value of disordered eating variables in psychological status. To correct multiple tests, the significance level was lowered (P<0.01). Data analysis was performed using SPSS 12.0.1 for Windows (2003).
Completed questionnaires were received, following reminder letters, from 5255 individuals, which represented a response rate of 57.1% after incorrectly listed addresses (n=684) and individuals away from home at the time of the survey (n=112) were taken into account. This is a conservative estimate as only a proportion of individuals with incorrectly listed addresses will be identified.28 Information concerning age was available only for non-respondents. The age distribution of respondents did not differ significantly from that of non-respondents.28 The demographic profile of the phase one sample of 5255 women was compared to the 2001 census data and found to be representative of the total population of young women residing in the ACT with respect to marital and employment status, education, children and first language. The sample consisted of approximately 10% of the female population aged 18–42 in the ACT.28
Of the 5255 women, 4891 women provided heights and weights from which BMI could be calculated. The 364 women who did not supply heights and weights were excluded from further analysis. The total number of obese women was 639 or 13% of the 4891 women. A comparison of demographic profiles between obese and non- obese women is displayed in Table 1.
Obese and non-obese women were compared in regards to psychological distress, EDE-Q global and EDE subscale scores using the Mann–Whitney U-test (see Table 2). Obese women had a small but significant increase in psychological distress. Obese women experienced higher levels in all the EDE-Q subscale scores.
There were 78 women in the obese subgroup (12.2% of all obese women) that had global EDE-Q scores two standard deviations above normal (a score 4.02).28 This level is extremely high and would indicate a possible clinical ED. Only 159 or 3.7% of the 4252 non-obese women had EDE-Q scores two standard deviations above normal.
The difference between obese and non-obese subsamples of women in terms of the presence of any and regular ED behaviours was investigated using χ2 test. As shown in Table 3, obese women were significantly more likely to have regular bulimic episodes, diuretic use, use of diet pills, and fasting behaviours.
To assess the association between disordered eating psychopathology and psychological distress in obese and non-obese subsamples, Spearman's rank–order correlations were performed between K-10, EDE-Q subscales and levels of specific ED behaviours. As shown in Table 4, the K-10 significantly correlated with EDE-Q subscales and bulimic episodes in both subsamples of women; however, significant correlations between the K-10 other ED behaviours differed between the groups. To ensure the results in the non-obese subsample were not affected by women of low weight, the analysis was repeated for women who did not fall into either the underweight or obese categories (i.e. those women with BMI18 and <30) and there were no changes in the results.
Age and weight status (BMI) were considered as possible confounding variables that may have affected linear regression results, therefore Spearman's rank–order correlations between these variables and psychological distress were conducted for the obese and non-obese subsamples of women. Age correlated significantly with the K-10 in both the obese and non-obese subsamples (obese sample: rs=−0.156, P<0.0001; Non-obese subsample rs=−0.150, P<0.0001). However, BMI did not correlate with the K-10 in either group.
Following normalization of all variables using the method of BLOM, backward selection linear regression was used to ascertain which ED factors in the obese subsample of women, that correlated with psychological distress (refer to Table 4) were the best predictors of psychological distress in a multivariate model. In addition to correlated ED features, age was also entered into the regression calculation. Dietary restraint, eating concern, weight concern, shape concern, level of subjective bulimic episodes, level of objective episodes and age were entered into the regression calculation; decreased age, dietary restraint, weight concern, eating concern and shape concern emerged as predicting variables (see Table 5) which explained 24% of the variance in the model (R2=0.238, F=33.892, P<0.0001), the other variables were excluded from the model.
For comparison, backward selection linear regression was used to ascertain which ED factors that correlated with psychological distress in the non-obese sub-sample (refer to Table 4) were the best predictors of psychological distress in a multivariate model. The normalized variables and backward selection linear regression method were again used; and age was added as a variable. Dietary restraint, eating concern, weight concern, shape concern, level of subjective bulimic episodes, objective bulimic episodes, self-induced vomiting, laxative misuse, use of diet pills and hard exercise to control shape or weight, as well as age were entered into the regression calculation. Decreased age, shape concern, eating concern, restraint, hard exercise, laxative misuse and subjective bulimic episodes emerged as predicting variables (see Table 5), which explained 21% of the variance in the model (R2=0.208, F=140.167, P<0.0001), the other variables were excluded from the model.
Although there are many studies on obesity, EDs and mental health, previous studies have often been in clinical settings or focused mainly on BED, and have not examined the relationship of obesity, disordered eating psychopathology and psychological status in a large community-based representative sample of young Australian women with obesity. This study indicated high levels of ED behaviours and cognitions in young obese women in the community, and when this occurred it significantly increased psychological distress. Furthermore, it has identified the specific ED predictors of psychological distress in this sample of obese women to be dietary restraint, weight concern, eating concern and shape concern.
Research has indicated that obesity is more common among those with poor socio-economic status.36 Although the questionnaire used in the study did not ask direct questions regarding level of income, it seems to be in agreement with past research, with the obese group of women having lower levels of university higher education and greater likelihood of reporting seeking full time work as their main activity. The obese sample were also slightly older, more likely to be born in Australia, be married or living as married, and more likely to class home duties or caring for children as their main activity. In this sample of women with obesity there was a very small but significant increase in psychological distress (depression and anxiety) compared to non-obese young women.
Many previous studies investigating disordered eating and obesity have focused on the behaviour of binge eating and BED. When comparing levels of binge eating and compensatory behaviours of the obese subsample to the non-obese subsample, the women with obesity were more likely to have experienced objective and subjective bulimic episodes. However, it is important to note in this comprehensive assessment of ED psychopathology that we also found regular use of diuretics, fasting and diet pills; and eating, shape and weight concerns and dietary restraint; were also significantly higher in the obese sub-sample. Furthermore, self-induced vomiting, laxative misuse, and hard exercise for shape or weight reasons occurred at similar or higher frequencies in the obese group, but not at levels that reached statistical significance. This comparison highlights that obese woman in the community experience a range of ED psychopathology, not limited to those expected in BED.
This study differs from a South Australian community epidemiological study of ED behaviour that found increased weight was associated with binge eating, but not purging (use of laxatives, diuretics and self-induced vomiting), or strict dieting or fasting (‘going on a very strict diet’ or ‘hardly eating anything at all for a time’).4 This present study however, consisted of young women only. Young women are known to be at high risk of EDs37 and it appears that young women with obesity may be particularly vulnerable to ED behaviours and cognitions compared to other segments of the Australian population.
All EDE-Q subscale scores were higher in the obese subgroup and all correlated strongly within the obese subgroup to psychological distress. This confirms previous studies that have shown dieting, weight and shape concerns or body image problems, eating concerns and binge eating to be high in obese people8, 15, 18, 24, 38, 39 and those that indicate a relationship with these factors with psychological health.8, 15, 18, 24 Although the presence of ED psychopathology increased psychological distress in all women irrespective of weight status, this study indicates that the levels of ED psychopathology are much higher in women with obesity, and therefore an important factor contributing to the poor health of this population group. Furthermore this study indicates that among women with obesity, disordered cognitions and high concerns with eating, weight and shape and dietary restraint play a significant role in predicting psychological distress (rather then other ED behaviours which play a greater role in other women) and thus these particular ED features may possibly be important to target in obesity treatment.
Similar to our work, a study on obese binge eaters compared to obese non binge eaters by Nauta et al.,24 showed that those who binge eat had increased eating concern, weight concern, shape concern, depression and decreased self esteem. They also found that independent of binge eating status people with obesity who had negative self schemas (cognitive generalizations about self which were based on shape, weight and eating), for example, ‘If I eat too much then I am good for nothing’ had increased depression and decreased self-esteem.24 Although binge eating is a common amongst people with obesity, this study highlights that disordered eating cognitions in obese people play a role in mental health, irrespective of whether they display the behaviour of binge eating.
An article by Schwartz and Brownell9 discusses the argument that in obese people life dissatisfaction, social liabilities and body image distress (as shown in our sample by high weight and shape concern), may be helpful in gaining motivation to lose weight.9 However, they suggest that these are more likely to be barriers to emotional regulation, that may lead to increased eating and not to weight loss.9 This present study supports this idea as the obese women in this sample display significant concern regarding weight and shape have a strong desire to lose weight, with many attempting various levels of dietary restraint and some attempting dangerous weight control measures associated with severe EDs. Despite these concerns and attempts at weight loss, these women are still obese and among those with high ED psychopathology there was greater psychological distress. Therefore considerable care and effort is required to change people's attitudes and behaviours in terms of effective and safe obesity prevention and weight loss methods, along with ways of enhancing mental well-being. Longitudinal studies on the interaction of life variables, disordered eating, mental health and obesity needs to occur so that a thorough assessment of the directional and causal links between these factors can be made and how they may interact as barriers or promoting factors to weight loss or maintenance, improved eating psychopathology and mental health.
There are a number of limitations to this present study. Firstly, it would have been beneficial to have an increased response rate, as it can not be ascertained if the women who did not provide contact details, or did not provide height and weight measures differed in ED psychopathology, weight status, or psychological status. Another problem is the reliance on self-report instruments, which are an effective way of gathering data in large population studies, but often have problems in validity. Previous studies, in particular, have also found validity problems in the bulimic episode measures in the self report EDE-Q.29
A recent Cochrane systematic review has recently found that psychological interventions particularly behavioural and cognitive behavioural strategies enhance weight reduction when combined with dietary and exercise strategies.40 The Cochrane systematic review on psychotherapy for bulimia nervosa and binging also supports cognitive behavioural therapy for the treatment of bulimia nervosa and similar EDs.41 Both reviews however, acknowledge that cognitive behavioural therapies are unlikely to be effective alone in reducing body weight.40, 41 Given the level of disordered eating in a community sample of young women with obesity and the considerable impairment this has on the psychological status, it would be prudent to screen for EDs in all young women seeking help for a weight problem. Further development and evaluation of combined exercise, dietary and psychological therapies (such as behavioural or cognitive behavioural therapy) for young women with obesity and co-morbid clinical or subclinical disordered eating psychopathology needs to occur with measurable outcomes including not only weight reduction, but changes in disordered eating, and other related social, physiological and psychological variables. Preventive efforts in obesity also need to consider the prevention of disordered eating cognitions and behaviours.
Levels of disordered eating behaviours and cognitions in young Australian women with obesity in the community are high. These problems have a negative impact on psychological (depression and anxiety) status. The findings of this study act as an important reminder to ensure prevention and treatment efforts for young obese women include psychological work on altering disordered attitudes in relation to eating, shape, weight and also altering behaviours and thoughts about effective methods of weight control and maintenance.
This research was assisted with funding from the NSW Institute of Psychiatry in the form of a research training fellowship for Jonathan Mond. A funding grant was also received from the Australian Capital Territory Department of Health and Community Care.
About this article
Quality of life in Brazilian obese adolescents: effects of a long-term multidisciplinary lifestyle therapy
Health and Quality of Life Outcomes (2009)