1 Division of Public Health & Primary Health Care, University of Oxford, Institute of Health Sciences, Old Road, Headington, Oxford OX3 7LF, 2 Department of Reproductive Medicine, The General Infirmary, Leeds LS2 9NS and 3 Department of Paediatrics, University of Cambridge, Addenbrookes Hospital, Cambridge CB2 2QQ, UK
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: bulimia nervosa/eating disorders/polycystic ovaries/polycystic ovarian syndrome
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The investigation of a potential relationship between PCOS and bulimia nervosa has been prompted by the observation that menstrual irregularity and acne are features which occur commonly in both (Franks, 1989; Fairburn and Beglin, 1990
; Gupta et al., 1992
; Fairburn and Cooper, 1993
; Balen et al., 1995
). Several studies have reported an association between the two conditions, but an aetiological link has never been satisfactorily explained (McSherry, 1990
; McCluskey et al., 1991
, 1992
; Raphael et al., 1995
; Jahanfar et al., 1995
). Interpretation of existing data is confusing due to limitations of the diagnostic tools used to detect bulimia nervosa, varying criteria applied for diagnosing PCOS, and the influence of selection biases in studies of women recruited from specialist clinics. The possibility that a spurious association may have been identified between the two relatively common conditions of polycystic ovaries and disordered eating habits must also be considered.
The reported link between polycystic ovaries, PCOS, and eating disorders was explored using data gathered in our previously reported study of the prevalence of polycystic ovaries in a group of young, post-menarcheal women in the normal population (Michelmore et al., 1999). An interviewer-based questionnaire was used for the assessment of behavioural eating patterns. The relationship of abnormal eating behaviours with ultrasound evidence of polycystic ovary morphology and features of PCOS was examined.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
All participants completed the eating disorder examination (EDE) (Fairburn and Cooper, 1993). This is an interviewer-based questionnaire which is considered to be the `gold standard' diagnostic tool for assessment of anorexia nervosa, bulimia nervosa, and their variants (Fairburn and Beglin, 1990
). The EDE enables the key aspects of eating disorders, (i) overeating and (ii) the use of extreme measures of weight control, to be assessed and provides frequency ratings for their occurrence over the previous 3 months. The series of EDE questions is divided into four main groups which represent different areas of eating disorder psychopathology. Scores for questions within these four groups were added and divided by the total number of questions within each group to produce a `subscale' score. These subscales included eating restraint, eating concern, shape concern, and weight concern and are described in more detail in Table I
, and `normal values' as described by Fairburn et al. for women with body mass index (BMI) <30 kg/m2 are also presented in Table I
(Fairburn and Cooper, 1993
). The four subscale scores were added and divided by four to produce a `global score' which provides a measure of overall severity of eating disorder psychopathology.
|
A total of 224 women attended for a transabdominal ultrasound scan. In almost all cases, EDE questionnaires were completed before scanning took place. The presence or absence of polycystic ovaries was determined using the criteria of Adams et al. (Adams et al., 1986). PCOS was defined as the presence of polycystic ovaries on ultrasound plus one additional feature including: menstrual irregularity, acne, hirsutism, BMI >25 kg/m2, raised serum testosterone (>3 nmol/l), or raised LH (>10 IU/l) (Franks, 1989
; Conway et al., 1989
; Balen et al., 1995
; Balen, 1999
).
As the subscale and global score distributions for EDE data were positively skewed, non-parametric MannWhitney U tests were used to compare the mean scores between women with and without polycystic ovaries.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The EDE interview was completed by all 230 participants. Seventy women (30%) described episodes of overeating which could be subdivided into: (i) `objective overeating' where a large amount of food was consumed, but where the participant did not feel any loss of control over eating, reported by 56 (24%); (ii) `objective bulimic episodes' where a large amount of food was consumed and where the participant described a feeling of loss of control over eating, reported by 10 (4%); and (iii) `subjective bulimic episodes' where the participant described a loss of control over eating while consuming amounts which were not objectively large, reported in 14 (6%). Extreme methods of weight control had been employed by nine (4%) participants. These included: (i) self-induced vomiting in four women, (ii) laxative misuse in two women, (iii) diuretic misuse in one woman, and (iv) intense exercising in six women. None of the 230 participants fulfilled the criteria for anorexia nervosa. Two women (1%) fulfilled the criteria for bulimia nervosa, and five women (2%) fulfilled the criteria for binge eating disorder. Neither of the women with bulimia nervosa was overweight, defined as a BMI >25 kg/m2, but three of the women with binge eating disorder were overweight.
As frank eating disorders (bulimia nervosa and binge eating disorder) were identified in only 3% of the total study population, comparisons between women with polycystic and those with normal ovaries were made using the quantitative EDE data. These are summarized in Table II. Comparisons were made for three features of disordered eating patterns to determine if polycystic ovarian morphology was associated with (i) binge eating or overeating, (ii) dieting behaviour, or (iii) global eating disorder symptoms.
|
`Overeating' was assessed by adding the number of objective bulimic episodes and the number of objective overeating episodes, as both involved consumption of a large amount of food. Sixty-two women reported episodes which would be included in this category. This included 27% (20/74) of women with polycystic ovaries, and 28% (42/150) of women with normal ovaries. The mean frequency of objective bulimic and overeating episodes was 5.5 episodes in the 20 women with polycystic ovaries compared with 4.6 episodes in the 42 women with normal ovaries. Again this difference was not statistically significant.
Dieting behaviour
The restraint subscale of the EDE was used as a measure of dietary restraint to compare women with polycystic and normal ovaries. The mean value for this subscale was higher in women with polycystic ovaries compared with normal ovaries (0.9 versus 0.8) but this difference was not statistically significant.
Global eating disorder symptoms
The global score for the EDE was used as an indicator of overall severity of eating disorder psychopathology. Mean global scores were noted to be higher in women with polycystic ovaries (0.9 versus 0.8), but this finding did not achieve statistical significance.
PCOS and EDE scores
Comparisons between women with PCOS versus non-PCOS women for the eating disorder examination data were similar to those reported for women with polycystic ovaries versus women with normal ovaries (Table III). Women with PCOS were found to have reported more frequent binge eating and overeating episodes than non-PCOS women, and also scored more highly for dieting behaviour and eating disorder symptomatology, but these differences were not statistically significant.
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The overall prevalence of bulimia nervosa detected in the current study is lower than that described by McCluskey et al. (1991), who used a self-completed questionnaire, the bulimia investigation test, Edinburgh (BITE), to assess the prevalence of bulimia in a group of 375 women presenting to a specialist endocrine clinic (McCluskey et al., 1991). A total of 44% of their study population was diagnosed as having PCOS, and bulimia nervosa was diagnosed by a high BITE score in 6% of these women as compared with 1% in their control group. Jahanfar et al. (1995) using the BITE questionnaire to diagnose sub-clinical eating disorder and bulimia nervosa, reported that 21% of women with polycystic ovaries (7/34) had abnormal questionnaire scores compared with 2.5% (1/40) of women with normal ovaries (Jahanfar et al., 1995
). The prevalence of bulimia nervosa in the current group closely resembles that of larger population based studies (Cooper and Fairburn, 1983
; Fairburn and Beglin, 1990
; Whitehouse et al., 1992
). The discrepancy with the above studies may possibly be explained by the observation that self-reported questionnaires (e.g. BITE), have a lower threshold for diagnosing bulimia and binge-eating than the more consistent and reliable interview-based techniques such as the EDE (Fairburn and Beglin, 1990
; Fairburn and Cooper, 1993
).
Although the prevalence of eating disorders in this population is in keeping with other community based studies, the possibility that a degree of selection bias may have occurred must still be considered. The study information literature distributed to potential volunteers clearly indicated that participants would be asked to complete a questionnaire about their eating habits. Women with eating disorders are often very secretive about their eating problems and weight control behaviour, and in cases of bulimia nervosa, commonly have strong feelings of shame or guilt about their eating habits (Fairburn and Beglin, 1990; Fairburn and Cooper, 1993
). Therefore, women with eating disorders may have been reluctant to participate in a study which would require them to expose and discuss their eating behaviour. In addition, the possibility that the prevalence of eating disorders within the study group was in fact higher than detected must be considered. Some of the study participants may not have felt comfortable with the interviewer-based EDE setting or questionnaire, and may have therefore provided some inaccurate responses.
Several theories have been advanced to explain the proposed link between polycystic ovaries, PCOS, and bulimia nervosa. It has been suggested that emotional distress associated with adverse symptoms of PCOS (menstrual irregularity, obesity, hirsutism, and acne), might act to promote the development of disordered eating habits. This theory is supported by the work of Fairburn et al. who performed a large case-control study to investigate risk factors for bulimia nervosa (Fairburn et al., 1997). They identified two broad classes of risk factors for bulimia nervosa; those which are general risk factors for psychiatric disorder (e.g. premorbid psychiatric disorder, behavioural problems, parental psychiatric disorder, disruptive life events), and those which increase the risk of dieting (e.g. critical comments by family about shape or weight, teasing about shape, weight, eating, or appearance, childhood or parental obesity). It seems logical to consider that some symptoms of PCOS (e.g. acne, hirsutism, and weight gain) might contribute towards an increased risk of dieting and a negative self image, and hence may act as risk factors for the development of bulimia nervosa or sub-clinical eating disorders. In the population of young women studied, however, there were no significant differences in acne, hirsutism, or BMI between women with polycystic and women with normal ovaries (Michelmore et al., 1999
) and no significant differences in EDE scores for dieting or overall eating disorder symptoms were detected.
It has also been suggested that bulimia nervosa itself creates a hormonal environment which predisposes towards polycystic ovarian changes, as polycystic ovaries have been more commonly identified in bulimic women when compared with controls in some small studies (McCluskey et al., 1992; Raphael et al., 1995
). Altered insulin secretion and insulin resistance in bulimic women have been proposed as potential mechanisms (Schweiger et al., 1987
; McCluskey et al., 1991
) but several studies have failed to detect significant differences in fasting insulin and in insulin secretion in bulimic women compared to controls (Weingarten et al., 1988
; Blouin et al., 1993
; Johnson et al., 1994
; Raphael et al., 1995
). The prevalence of frank bulimia nervosa in the current study was too low to confirm or refute the association with polycystic ovaries, however it can be noted that the frequency of binge-eating and overeating was not significantly greater in women with polycystic ovaries or PCOS when compared with women with normal ovaries.
In summary, the current study, using a reliable diagnostic tool for the assessment of eating disorders, does not indicate that bulimia nervosa or other binge-eating disorders occur more commonly in women with polycystic ovaries or PCOS. In addition, women with polycystic ovaries do not demonstrate significantly higher scores for dieting and other features of shape and weight concern when compared with women with normal ovaries. It has been suggested that screening for abnormal eating behaviour in women with PCOS should be adopted as common clinical practice, before dieting is recommended as treatment for symptom control (Morgan, 1997). The suggestion that polycystic ovaries predispose towards development of eating disorders, however, is not supported by the current study.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
American Psychiatric Association (1993) DSM-IV draft criteria (3/1/93). Diagnostic and Statistical Manual of Mental Disorders. American Psychiatric Association, Washington, DC.
Balen, A. (1999) Pathogenesis of polycystic ovary syndrome the enigma unravels? Lancet, 354, 966967.[ISI][Medline]
Balen, A.H., Conway, G.S., Kaltsas, G. et al. (1995) Polycystic ovary syndrome: the spectrum of the disorder in 1741 patients. Hum. Reprod., 10, 21072111.[Abstract]
Blouin, A.G., Blouin, J., Bushnik, T. et al. (1993) A double-blind placebo-controlled glucose challenge in bulimia nervosa: psychological effects. Biol. Psychiatry, 33, 160168.[ISI][Medline]
Clayton, R.N., Ogden, V., Hodgkinson, J. et al. (1992) How common are polycystic ovaries in normal women and what is their significance for the fertility of the population? Clin. Endocrinol. Oxf., 37, 127134.[ISI][Medline]
Conway, G.S., Honour, J.W. and Jacobs, H.S. (1989) Heterogeneity of the polycystic ovary syndrome: clinical, endocrine and ultrasound features in 556 patients. Clin. Endocrinol. Oxf., 30, 459470.[ISI][Medline]
Conway, G.S. (1990) Insulin resistance and the polycystic ovary syndrome. Contemp. Rev. Obstet. Gynaecol., 2, 3439.
Cooper, P.J. and Fairburn, C.G. (1983) Binge-eating and self-induced vomiting in the community. A preliminary study. Br. J. Psychiatry, 142, 139144.[Abstract]
Fairburn, C.G. (1997) Eating disorders. In Oxford Textbook of Medicine. Oxford University Press, Oxford, pp. 42124218.
Fairburn, C.G. and Beglin, S.J. (1990) Studies of the epidemiology of bulimia nervosa. Am. J. Psychiatry, 147, 401408.[Abstract]
Fairburn, C.G. and Cooper, Z. (1993) The eating disorder examination, 12th edn. In Binge eating: nature, assessment and treatment. Guilford Press, New York, pp. 317331.
Fairburn, C.G., Welch, S.L., Doll, H.A. et al. (1997) Risk factors for bulimia nervosa. A community-based case-control study. Arch. Gen. Psychiatry, 54, 509517.[Abstract]
Farquhar, C.M., Birdsall, M., Manning, P. et al. (1994) The prevalence of polycystic ovaries on ultrasound scanning in a population of randomly selected women. Aust. N. Z. J. Obstet. Gynaecol., 34, 6772.[ISI][Medline]
Fox, R., Corrigan, E., Thomas, P.A. and Hull, M.G. (1991) The diagnosis of polycystic ovaries in women with oligo-amenorrhoea: predictive power of endocrine tests. Clin. Endocrinol. Oxf., 34, 127131.[ISI][Medline]
Franks, S. (1989) Polycystic ovary syndrome: a changing perspective. Clin. Endocrinol. Oxf., 31, 87120.[ISI][Medline]
Gupta, M.A., Gupta, A.K., Ellis, C.N. and Voorhees, J.J. (1992) Bulimia nervosa and acne may be related: a case report. Can. J. Psychiat., 37, 5861.[ISI][Medline]
Jahanfar, S., Eden, J.A. and Nguyent, T.V. (1995) Bulimia nervosa and polycystic ovary syndrome. Gynecol. Endocrinol., 9, 113117.[ISI][Medline]
Johnson, W.G., Jarrell, M.P., Chupurdia, K.M. and Williamson, D.A. (1994) Repeated binge/purge cycles in bulimia nervosa: role of glucose and insulin. Int. J. Eat. Disord., 15, 331341.[ISI][Medline]
McCluskey, S., Evans, C., Lacey, J.H. et al. (1991) Polycystic ovary syndrome and bulimia. Fertil. Steril., 55, 287291.[ISI][Medline]
McCluskey, S., Lacey, J.H. and Pearce, J.M. (1992) Binge-eating and polycystic ovaries. Lancet, 340, 723.[ISI][Medline]
McSherry, J.A. (1990) Bulimia nervosa and polycystic ovary syndrome: evidence for an occasional causal relationship. Med. Therapist, 6, 1011.
Michelmore, K.F., Balen, A.H., Dunger, D.B. and Vessey, M.P. (1999) Polycystic ovaries and associated clinical and biochemical features in young women. Clin. Endocrinol. Oxf., 51, 779786.[ISI][Medline]
Morgan, J.F. (1997) Polycystic ovary syndrome (letter). Brit. J. Hosp. Med., 57,
Patton, G.C., Selzer, R., Coffey, C. et al. (1999) Onset of adolescent eating disorders: population based cohort study over 3 years. Brit. Med. J., 318, 765768.
Polson, D.W., Adams, J., Wadsworth, J., and Franks, S. (1988) Polycystic ovaries a common finding in normal women. Lancet, 1, 870872.[ISI][Medline]
Raphael, F.J., Rodin, D.A., Peattie, A. et al. (1995) Ovarian morphology and insulin sensitivity in women with bulimia nervosa. Clin. Endocrinol. Oxf., 43, 451455.[ISI][Medline]
Schweiger, U., Poellinger, J., Laessle, R. et al. (1987) Altered insulin response to a balanced test meal in bulimic patients. Internat. J. Eat. Disord., 6, 551556.[ISI]
Weingarten, H.P., Hendler, R. and Rodin, J. (1988) Metabolism and endocrine secretion in response to a test meal in normal-weight bulimic women. Psychosom. Med., 50, 273285.[Abstract]
Whitehouse, A.M., Cooper, P.J., Vize, C.V. et al. (1992) Prevalence of eating disorders in three Cambridge general practices: hidden and conspicuous morbidity. Br. J. Gen. Pract., 42, 5760.[ISI][Medline]
Submitted on June 7, 2000; accepted on January 2, 2001.