The prevalence of binge-eating and bulimia in 1063 college students

OOU-3956/H SJ.OO+ .OO Pcrgamon Press Ltd.

J.psychiot. Res., Vol. 19, No. 2/3. pp. 161-166.1985 Primed in Great Britain

THE PREVALENCE OF BINGE-EATING AND BULIMIA IN 1063 COLLEGE STUDENTS KEVINHEALY, RONANM. CONROYand NOELWALSH Department of Psychiatry, St. Vincent’s Hospital, Elm Park, Dublin 4, Ireland Summary-We used a purpose-designed questionnaire to survey the prevalence of binge-eating and bulimia in a sample of 1063 Dublin third-level students aged 17-25 yr. There were 361 males and 701 females. The questionnaire was based on DSM-III. and included a written definition of a binge and cross-check questions. Although 17.7% of males and 37Qaof females claimed to have had an eating binge, cross-check items reduced this to 1.1% of males and 10.8% of females who met the DSM-III definition. No male and only 7.7% of females also met the behavioural criteria under item B of DSM-III, and only 5% of females reported dysphoric mood. Excluding those experiencing fewer than one episode per week gave a prevalence of 2.8Qa in females and 0% on males. Previously-reported prevalences using questionnaire may be inflated due to poor respondent understanding of the psychiatric terms being used.

INTRODUCTION

IN THE last decade there has been an increased interest in binge-eating and bulimia. Bulimia has been described as a feature of anorexia nervosa (GARF~NKEL et al., 1980), but symptoms of binge-eating and vomiting have also been reported in subjects of normal weight (STRANGLERand PRINTZ, 1980). As yet there is no consensus as to whether bulimia is a distinct psychiatric syndrome or an ominous variant of anorexia. Binge-eating, the episodic and rapid consumption of large quantities of easily-ingested food, has been described as an essential feature of a number of syndromes described in recent years‘thin fat people’ (BRUCH, 1974), ‘bulimarexics’ (B~SKIND-LODAHL,1976), ‘binge-eating’ (NOGAMIand YOBANA,1977), dietary chaos syndrome (PALMER,1979), bulimia nervosa (RUSELL, 1979), bulimia (DSM-III, 1980) and ‘pigging-out’ (LUCU, 1982). The prevalence of any of these syndromes in the general population is not known. HAwru~s and CLEMENT (1980) reported 79% of females and 49% of males as having experienced episodes of ‘uncontrolled excessive eating’ and CARPERand FA~BURN (1983) found that 20.9% of females attending family planning services currently experienced such episodes. These figures cannot, however, be taken to represent the prevalence of a definable psychiatric disorder. HALMI ef al. (1981), found a prevalence of the syndrome of bulimia, as defined by DSM-III (1980), of 19% in females and 5% in males in a sample of 355 college students. No information on the prevalence of bulimia, as defined by DSM-III, was available in the Irish population, but our own experience led us to believe that the proportion of Irish female college students with diagnosable bulimia was less than the one in five reported by HALMI (1981). We therefore decided to survey a large sample of Irish third-level students, using a questionnaire based on DSM-III. 161

162

KEVIN HEALY

etaI.

SUBJECTS AND METHODS

We obtained the co-operation of a number of third-level institutions in Dublin: three primary-teacher training colleges and the departments of medicine and engineering of University College, Dublin. With their permission we surveyed students immediately after scheduled lectures. Students were free not to complete the questionnaire, which was distributed by one of the two senior authors. The questionnaire used in the study was written specially by the senior authors. Without wishing to add yet another questionnaire to the psychiatric literature, we felt that no existing instrument allowed the rating of all DSM-III criteria for bulimia. In addition, we felt that any questionnaire to elicit bulimic symptoms should present a clear description of an eating binge, as this concept may not be well-understood by the general public. The resulting instrument, the Conroy-Healy Eating Questionnaire (CHEQ), operationalised the criteria in DSM-III as shown in Table 1. We also included two items to rate morbid fear of weight gain, which is proposed by RUSSELL (1979) as a diagnostic criterion for ‘bulimia nervosa’ in addition to those specified in DSM-III. The criterion of ‘frequent weight variation of more than 10 pounds due to alternating binges and fasts’ was operationalised as ‘weight variation of more than 5 kg (11 pounds) in the past year’. ‘Recurrent episodes’ of binge eating was defined as episodes once a week or more for the past three months. Body mass index (BMI) was calculated for each subject. This is the TABLE1. DSM-III CRITERIA FORBULIMIA AND CHEQ DEFINITIONS

CHEQ

DSM A: Recurrent binge-eating episodes.

At least 1per week for last 3 months.

B: At least three of: (1) Consumption of high caloric. easily ingested food during binge.

Same as DSM.

(2) Inconspicuous eating during binge.

Respondant feels that others do not notice.

(3) Termination of episodes by abdominal pain, sleep, social interruption or self-induced vomiting.

Any one of DSM list.

(4) Repeated attempts to lose weight by harsh diets, selfinduced vomiting cathartics and/or diuretics.

Tries to control weight by harsh diets OR use of laxatives, diet pills of diuretics once a week or more.

(5) Frequent weight fluctuations of more than 10 Ibs due to alternating binges and fasts.

Weight fluctuation of at least 5 kg (1I lb) in last year (cause unspecified).

c: Awareness that the eating pattern is not normal and fear of being unable to stop.

Used in binge defbtition and cross-checked.

D: Depressed mood and self-deprecating

Same as DSM-III. Both features required.

thoughts after binge.

E: Bulimic episodes not due to anorexia or organic causes.

Order in which criteria are applied to potential cases: (1) Control (C above). (2) Three items from B. (3) Mood criteria. (4) Current (A above). (5) Not anorexic.

Exclude subjects who are under normal weight and admit harsh dieting to control weight.

BULIMIAIN COLLECIE STUDENTS

163

ratio of the subject’s weight in kilograms to the square of his/her height in metres. A cutoff point for BMI was established below which 12% of each sex scored, and this, together with claiming to use harsh dieting to control weight, was used as an operational definition of anorexia. The questionnaire was divided into three sections: the first page recorded routine demographic data and information on weight, weight variation, age and height. The second section contained items relating to binge-eating. These were prefaced by a description of an eating-binge. The final section contained miscellaneous symptoms of eating disorders. Before completing the questionnaire the participants listened to standard instructions read by one of the authors. These included reading aloud the definition of a binge-eating episode. The questionnaire was piloted on a group of 30 student nurses, revised on the basis of their comments and repiloted on a further 26 student nurses before attaining its present form. RESULTS

A total of 1096 completed questionnaires were returned. The non-participation rate could not be calculated for some settings, but in those settings where head-counts could be checked against numbers of questionnaires returned, response rate was in excess of 95%. Seventeen questionnaires were spoiled and a further 16 completed by subjects aged over 25 yr. These were excluded from the analysis. Table 2 shows the age, sex and parental occupation distribution of the resulting sample of 1063 students. Table 3 shows the individual prevalence of the five items which make up criterion (B) of DSM-III, three or more of which constitute a fulfillment of the criterion. Table 4 shows the prevalence of each of the criteria in DSM-III, as operationalised by the CHEQ. A total of 320 subjects (30.4Vo) claimed to have had an eating binge. Despite written and aural instructions to the effect that a necessary feature of a binge was the TABLE

Under 18

2. AGE, SEXAND FATHER’S OCCUPATION OFSAMPLE Age group 19

18

20 to 25

Male Female

129 (35.7%) 170 (24.2%)

98 (27.1 @IO) 237 (33.8Vo)

47 (13.0%) 195 (27.8%)

87 (24.1070) 100 (14.2@70)

Total

299(28.1%)

335 (31.5%)

242 (22.8%)

187 (17.6%)

Total 361 (34%) 701 (66Vo) 1063 (1OOVo)

Parent’soccupation Group 1 Male Female Total

87 (25.4010) 81 (12.0%) 168 (16.5’I’o) Missing: Parental

Group 2

Group 3

60 (17.5%) 112 (32.7%) 129(19.1%) 190(28.1%) 189(18.58) 302 (29.6%) occupation: 44

Group 4

Group 5

Group 6

27 (7.9%) 54 (8.0%) 81 (7.9Vo)

12 (3.51) 49 (7.22) 61 (6.01)

45 (13.1%) 173 (25.6%) 218 (21.4Vo)

Group 1: Higher professional. Group 4: Skilled manual. Group 2: Intermediate. Group 5: Semi or unskilled manual. Group 3: Routine non-manual. Group 6: Agricultural. Chi-square=50.5,df=3,p<0.0001. Sex by age: Sex by occupation: Chi-squarel48.5, df = 5,~ < 0.0001.

164

KEVIN HEALY et al.

fear of being unable to stop, only 80 of these (7.6%) reported being afraid that they could not stop eating voluntarily. Table 4 shows that successive application of DSM-III criteria to subjects who reported a binge. Of those who admitted binging and feared loss of control, 54 females (7.7% of females) and no males fulfilled three criteria from section B of DSM-III. Dysphoric mood-depression and either guilt or anger-terminating binge episodes was absent in 19 of these, and of the remaining 35, 20 had one or more episodes per week in the past three months. Of these, one subject met the CHEQ definition of anorexia, leaving a final prevalence of 19 females (2.7% of females) and no males. TMLE 3.

PRKVALENCE OF DSB-III

Criterion High-caloric, easily-ingested food: Feels others do not notice: Ending as specified: (see Table 1) Weight control by harsh dieting:

TABLE4.

Females

52 (14.4%)

239 (34.0%)

17 22

117 (16.7%) 148(21.1%)

(4.7%) (6.1%)

56 (8.OQo) 127 (18.1%)

total 1063.

PR6VALENCK OF BtJLtbUA DEFINED ON DSM-III Percent meeting criterion Male Female

Claimed binge Met definition (DSM C) 3 Items from list (B) Dysphoric mood (D) Current binge eating (A) Not due to anorexia

(‘70)

Males (To)

4 (1.1%) 19 (5.3%)

Weight fluctuations > 5 kg in last year: Subjects = 361 males, 702 females,

(B) LIST CRITERIA

17.7% 1.1% 4.4% 2.8% 2.8%

37.0% 10.8% 19.2% 16.0% 14.0%

BY SEX

Cumulative percent meeting all criteria

Male

Female

17.7% 1.1% 0.0% 0.0% 0.0% 0.0%

37.0% 10.8% 7.7% 5.0% 2.8% 2.7%

Prevalence of: experience of binge-eating Male = I. 1% Female = 10.8% current Bulimia (DSM-III) Male = 0.0% Female= 2.7%.

Of the DSM-III criteria (A-E in Table I), 89 subjects met one criterion, 69 met two, 50 three and 21 four criteria. Females were more likely to meet each DSM criterion than males, and the ratio of males to females became more extreme with increasing numbers of criteria met. DISCUSSION

The criteria provided in DSM-III make it difficult to operationaiise a definition of bulimia suited to questionnaire screening. Control of weight by harsh dieting is a DSM-III bulimia criterion (see Table I), but also a feature of anorexia. Our exclusion of low-weight subjects who admitted harsh dieting to control weight might conceivably have resulted in

BULIMIA IN COLLEGESTUDENTS

165

the exclusion of a large number of the subjects selected on previous criteria. In fact only one of the subjects was thus excluded, but this area remains problematic, both for the methodology of prevalence studies and, more generally, for the issue of whether bulimia is to be considered as subsumed under, overlapping with, or distinct from anorexia. These issues are unlikely to be resolved until we are in a position to define these disorders in terms of psychopathology rather than of external characteristics such as behaviour and weight. Given the current limitations of the diagnostic criteria for bulimia, we believe that the results quoted provide a face-valid prevalence rate. The maximum prevalence rate for any binge-eating disorder which uses the definition of an eating binge cited in DSM-III is 1.1070of males and 10.8% of females in the population studied. The prevalence of bulimia, as specified in DSM-III, is 2.7% in females and two low to be estimated in males. RUSSELL(1979) specifies a morbid fear of weight gain as a criterion for bulimia nervosa. Two items on the CHEQ were designed to elicit this construct: 16 of the 19 bulimics identified reported being “terrified of being overweight” and 13 reported usually becoming anxious before eating. Only one reported neither, indicating a considerable overlap between Russell’s bulimia nervosa and DSM-III bulimia. The prevalence found in this study is considerably lower than reported in the other study which used DSM-III criteria (HALMIet al., 1981). It is possible that the prevalence of bulimia in New York is considerably higher than in Dublin. Nevertheless, Halmi’s prevalence (19% of females) would give bulimia a higher prevalence than even minor affective disorders, which is counter-intuitive. There are a number of other factors which might explain the lower prevalences found in the Dublin sample. We found a considerable degree of misunderstanding among students about the nature of an eating binge. Despite a written description and verbal instructions the majority of those who claimed a binge had something in mind other than the symptom defined by DSM-III. The concept of an eating binge may be hard for even well-educated respondants to grasp, and high prevalences may reflect a difference in semantics between psychiatrist and respondant. Just as not everyone who experiences breathlessness is dyspnoeic, not everyone who eats too much occasionally is a bulimic. It is likely that few respondants understand the compulsive quality of eating binges unless they actually experience them. In addition, the CHEQ uses only extreme responses to rate symptoms. Definitionally, we believe, many symptoms such as preoccupations with weight, fears of weight gain and being conscious of trying to control weight are only of psychiatric significance if they occur with a frequency and intensity that makes them interfere with normal cognition and behaviour. Positive responses to these items on the CHEQ may not indicate severe symptoms, but are intended to identify significant psychopathology, even at the risk of missing mild phenomena. Finally, Halmi’s analysis presents prevalences of some of the features of DSM-III, but does not give a prevalence based on all DSM criteria. Although the prevalences found in this study are low by comparison with the international literature, they nevertheless represent a significant absolute quantity of psychiatric morbidity. It was not possible, given the anonymous format used in the study, to interview subjects who were potentially clinically bulimic. We did, however, examine the questionnaires of the 80 subjects who claimed to have had a binge as defined by the CHEQ. Of

166

KEVIN HEALY et

al.

these, 16 were considered by both senior authors to have bulimia of probably clinical intensity on the basis of their written descriptions. We are currently comparing the responses on the CHEQ of a group of 50 bulimics and 50 matched normal controls in an attempt to identify the best item subset to identify clinical cases.

REFERENCES M. (1976) Cinderella’s sisters: a feminist perspective on anorexia nervosa and bulimia. S&u: J. Women Cuit. Sot. 2.342-356. Bnuca, H. (1974) Eating Disonlrrs. Routkdge & Kegan Paul, London. COOPER, P. J. and FAIRBURN,C. G. (1983) Binge eating and self-induced vomiting in the community. Er. J. Psychiut. 142,139-144. Diagnostic and Statisricol ManuaI of Mental Disorders, 3rd Edn. (1980) American Psychiatric Association, Washington. DC. Hawtct~s, R. C. and C~~MBNTS,P. F. (1980) Development and construct validation of a self report measure of binge eating tendencies. Addict. Behav. $219-226. GMFWREL. P. E., MOLDOPSKY,H. and GARNER, D. M. (1980) The hetrogeneity of anorexia nervosa: bulimia as a distinct subgroup. Br. J. Psychiat. 144,238-246. HALMI,K. H., FALIT,J. R. and %XWARTS. E. (1981) Binge-eating and vomiting: a survey in a college population. Psychol. Med. 11,697-706. Lucw. A. R. (1982) Pigging out. J. Am. med. Ass. 1,247. NO~AW, Y. and YOBANA.F. (1977) On kibarashi-gui (binge eating). Foliapsychiat. neural. Japonica 31, 159-166. PUMJZR,R. L. (1979) The dietary chaos syndrome: a useful new term? Br. J. med. Psycho/. 52, 187-190. RUSSELL.G. (1979) Bulimia nervosa: an ominous variant of anorexia nervosa. Psychol. Med. 9.429-448. STRANOLER, R. S. and PRINTZ, A. M. (1980) DSM-III: psychiatric diagnosis in a university population. Am. J. Psychiat. 137.937-940. BOSKZND-LODAHL,