Addictive Behaviors, Vol. 20, No. 4, pp. 471-480, 1995 Copyright 0 1995 Elsevier Science Ltd
Pergamon
Printed in the USA. All rights reserved 0306-4603195 $9.50 + .oa
03064603(95)00002-X
THE CORRELATES OF BINGE EATING IN TWO NONPATIENT SAMPLES JOHN F. O’MAHONY
and SARAH HOLLWEY
University College, Dublin Binge eating in two nonpatient samples was examined as a function of actual body weight (expressed as BMI), attempts to control eating, food/weight preoccupation,
Abstract -
and neuroticism. All of these factors were univariately associated with binge eating, but multivariate analyses indicated that food/weight preoccupation was the most powerful predictor, accounting for much of the common variance shared by the various correlates. This replicates previous work and suggests that the dynamics of subclinical bingeing are substantially similar to that found in the full DSM-III-R
syndrome of bulimia nervosa.
Bingeing, roughly understood as the repeated consumption of excessive amounts of food in a short period of time, accompanied by a sense of loss of control, is a common enough problem. When a full syndrome of DSM-III-R-defined bulimia ner-
vosa (frequent bingeing, accompanied by various forms of purging, with a sense of loss of control and overconcern with body shape and weight; American Psychiatric Association, 1987) is considered, some 6% to 8% of the young female population are sufferers (Schlundt & Johnson, 1990). When the behavior of bingeing alone is studied, about a quarter of the young female population will admit to it, with perhaps ten percent engaging in this behavior once a week or more (Schlundt & Johnson, 1990). The nosological status of the various eating patterns involving bingeing is not fully settled. Bingeing behavior can occur as part of various more or less well-recognised clinical syndromes (anorexia nervosa, bulimia nervosa, obesity, and binge eating disorder (Spitzer et al., 1992). Our major focus in this paper is on the correlates of bingeing behavior rather than questions of nosology. Risk factors for binge eating are well established. Severe food restriction can lead to subsequent binge eating among those with previously normal eating patterns (Hsu, 1990a; Keys, Brozek, Henschel, Michelson, & Taylor, 1950). Dieting to lose weight is an accepted cultural practice among women in the developed world. After a literature survey of data gatherered mainly in developed countries, Schlundt and Johnson (1990) estimate that some 40% of young women diet, while 14% or so diet “harshly.” Dieting is very common among those who meet DSM definitional criteria for bulimia nervosa (Hsu, 1990a; Schlundt & Johnson, 1990). Patton, Johnson-Sabine, Wood, Mann, & Wakeling (1990) have prospectively shown that earlier dieting considerably increases the risk of the development of a latter bulimic disorder. There is also a substantial association between dietary restraint and binge eating in student samples (Greenberg, 1986; Hawkins & Clement, 1980; Wardle, 1987). Weight status is also important. Those who binge are often concurrently overweight or have a history of being overweight and of dieting (Spitzer et al., 1992; Turnbull, Freeman, Barry, & Henderson, 1989). Thanks are due to the participating women and to Ciaran Dolphin, Ronan Conroy, and Mary Darby. Requests for reprints should be sent to John F. O’Mahony, Eastern Health Board, Psychology, Rathdown Road, Dublin 7, Ireland. 471
412
J. F. O’MAHONY and S. HOLLWEY
Persons who binge-eat also have personality disturbances. Those with eating problems and concerns of all kinds tend to have more personality, mood, and social abnormalities than those without such problems. These include: impulsivity, substance abuse, depression, anxiety, lability of mood, interpersonal sensitivity, low self-esteem, and social maladjustment (Garner, Olmsted, Davis, Rockert, Goldbloom, & Eagle, 1990). This is true of people meeting formal DSM criteria (Garner et al., 1990; Hsu, 1990a) for bulimia nervosa and for those binge eaters who do not meet DSM criteria for bulimia nervosa (e.g., Cooper & Fairburn, 1983; Ruderman & Besbeas, 1992; Vanderheyden & Boland, 1987). These risk factors have all been assigned primary causal status in the emergence of bingeing problems. Bingeing has been conceptualised as a natural physiologically based response to calorie or carbohydrate restriction (Wurtman, 1989), a behavioral rebound effect of excessive control over the normally relatively unregulated function of eating (Polivy & Herman, 1985), or a method of regulating moods and aversive self-cognition (Heatherton & Baumeister, 1991; Johnson & Connors, 1987; Williamson, 1990). However, not all dieters binge; some who binge are not nor ever have been overweight; and not all persons with personality and mood problems binge. Only some people possessed of these “risk” factors do. A consensus has emerged, therefore, that the causes of eating disturbances, bingeing included, are multifarious (e.g. Hsu, 1990a; Schlundt & Johnson, 1989; Williamson, 1990). Causes can work at different and interacting levels and in an intertwined way: Starvation leads to hunger and food/eating preoccupation (Keys et al., 1950); dissatisfied persons may seek satisfaction by attempting to mould their physiques into socially desirable (i.e., thin) shapes (Cattanach & Rodin, 1988); or they may distract themselves in eating behavior from their aversive moods and self-concerns (Elmore & de Castro, 1990; Heatherton & Baumeister, 1991; Hsu, 1990b; Kaye, Gwirtsman, George, Weiss, & Jimerson, 1986); bingeing often leads to guilt and self deprecation (Abraham & Beumont, 1982) and may contribute to social isolation (Herzog, Keller, Lavori, & Ott, 1987). The dynamics of bingeing behavior may also change with time, and a mechanism can come to serve purposes other than those that originally motivated it. Thus, a person who discovers that relief from one set of distressing thoughts can be achieved through bingeing may adopt bingeing as a general, short-term quieting mechanism a response to a variety of unpleasantness and worries. Additions to the behavioral sequence may also be important. Bingeing usually begins as just bingeing. It often then progresses to include vomiting and other forms of compensatory purging (Turnbull et al., 1989). The elaboration of simple bingeing into bingeingjpurging may change and enrich the meaning and function of bingeing. The urge to binge, once vigorously resisted because it would lead to weight gain, may now be freely indulged as its natural and normal consequences can apparently be easily remedied. It seems, therefore, that an inquiry into the correlates of bingeing of those with mild bingeing manifestations might be worthwhile. One might get a clearer view in such a population before the various interrelated factors mentioned above become too intertwined (see Patton et al., 1990 and Steiger, Puentes-Neuman, & Leung, 1991 on this point). Our study did this. We took two samples: one quite unselected for eating, or any other, problems; another selected for a likely elevated level of mild bingeing behavior.
Correlates of bingeing
473
METHOD Subjects Participants
in the study were in two groups. The first comprised 96 women, available persons who agreed to take part in the study, otherwise unselected. These were nurses and undergraduate students. We will call this the “ordinary” group. The second group of women was drawn from populations whose interests or occupations suggested that they would have an especial concern about weight, appearance, or bodily condition: members of health clubs, models, dancers, and athletes. There were 25 models, 26 dancers, 20 competitive athletes, and 34 health club attendees in all, 105. These populations were chosen as they are known to have an elevated level of eating problems (e.g., Szmukler, Eisler, Gillis, & Hayward, 1985; Yates, Leehey, & Shisslak, 1983). We will call this body-focussed group the body group. The mean age for the two groups (ordinary and body, respectively) was 21.1 and 25.6 years (SDS = 3.3 and 8.0), mean heights 65 in. in both cases (SDS, 2.8 in. and 2.8 in.), mean weight 129 (SD = 17.6) and 126 (SD = 17.6) pounds, and body mass index (BMI) 21.7 (SD = 2.6) and 20.9 (SD = 2.8). Participants were recruited in several ways. Some were approached personally by one author, a female psychologist, while others were given questionnaires by a third party, such as a teacher, trainer, or employer. Almost all (95%) of those invited to participate in the study agreed to do so. Measures
On the basis of pilot work a questionnaire composed of 49 questions about food, eating, and weight was constructed. It had three main sections: demographic and personal data; a section investigating bulimic behavior, bingeing, purging, and guilt about these; and anorexic-type behavior such as the reduction of food intake, food fads and rituals, body-image disturbance, amenorrhoea, and excessive exercise. The response format varied somewhat between items. Binge frequency, for example, was measured as “not at all,” “once a month,” “once a week,” “once a day,” and “more often.” Most of the eating-related issues were, however, coded as present or absent. The time frame for the questions was defined as over the previous year. Because the concept of binge can be differently understood (Wilson, 1992), it was explicitly defined at the beginning of the questionnaire, according to DSM-UZ-R terms, as rapid consumption of a large amount of food in a discrete (that is, relatively brief) period of time, combined with feelings of loss of control. Two composite measures were computed. The first was a measure of the behavioral control of eating/weight and was formed by combining the items: “have rigid rules about eating, ” “consciously try to control weight,” “diet harshly to control weight,” “ eat the same food daily, ” “keep a daily record of food,” and “exercise to control weight,” with a weighing of 1 for each affirmative answer. These items were chosen for face validity, and this six-item scale had an Alpha coefficient of .62, for the whole sample. A measure of preoccupation with eating/weight issues was formed by combining the items “think of food, eating, and weight most of the time,” “have a terror of being overweight, ” “constantly desire to be thinner,” “have intrusive thoughts about food and weight sometimes or often,” and “am preoccupied with fear of putting on weight.” Again, the items were chosen for face validity and this five-item scale had an Alpha coefficient of .76 for the whole sample.
474
J. F. O’MAHONY and S. HOLLWEY
In order to assess the temporal stability of this questionnaire, two groups of persons not involved in the present study - 31 formerly anorexic patients and 53 student nurses - completed the questionnaire on two occasions, separated by 5 months for the formerly anorexic patients and 3 weeks for the nurses. Reliability was computed as a Pearson correlation coefficient between the composite scales on the two occasions. Coefficients were Y = .77 and .82 for the patients and nurses, respectively, for the behavioral control scale and Y = .84 and .87 for the food/weight preoccupation composite, indicating that the behavior and attitudes measured by the questionnaire are temporally quite stable. One validation study was carried out. A group of 60 female undergraduates, not involved in this study, completed the questionnaire and the 26-item version of the Eating Attitudes Test (Garner & Garfinkel, 1979; Halmi, 1985) on separate occasions. A composite measure of severity of eating-related problems based on 20 items from the questionnaire correlated highly at r = .86 with the overall Eating Attitudes Test score. Body Mass Index (i.e., weight in kg./height in cm?), computed from self-reported weight and height, was used as a measure of relative body weight as it has a low correlation with height (Keys, Fidanza, Karvonen, Kimura, & Taylor, 1972). Maximum and minimum-ever BMI were computed using reported highest- and lowestever weight. All participants also completed the Neuroticism scale of the Eysenck Personality Questionnaire (EPQ; Eysenck, & Eysenck, 1975). This scale was used as a general measure of psychological well-being, or conversely, maladjustment, with high scores indicating strong emotions, lability, and emotional overactivity. Since, as mentioned earlier, bulimia and bingeing are associated with a large variety of measures of abnormality and distress, we felt that a general measure of maladjustment such as this scale would be the most appropriate approach to abnormality outside of the eating sphere (see Ruderman & Besbeas (1992) on this point). The EPQ is a well known measure and adequate reliability and validity are documented (Eysenck & Eysenck, 1975). RESULTS
Bingeing was common enough. Of the ordinary group, 25% binged once a month, 16% once a week, and none once a day. The corresponding figures for the body group were 18%, 14%, and 2%. However, these distributions were not significantly different (x2 (3, N = 201) = 3.3, p = .34). The distribution of eating-related issues in the two groups is shown in Table 1. Since daily bingeing was so infrequent, binge frequency was trichotomised to “not at all,” “once a month or more,” and “once a week or more” for most analyses. Of the 14 items in Table 1, there were differences on four items (by x2 (1, N = 201), p < .05) between the ordinary and body groups, generally reflecting more food/weight concerns in the body group. The groups also differed significantly on the composite measures of behavioral control and food/ weight preoccupation, F( 1,199) = 5.9, p < .05, and F( 1,199) = 7.0, p < .05 for the two measures, respectively. Accordingly, the two groups were analysed separately. We attempted an approximation of DSM-ZZZ-Rcriteria for bulimia nervosa and found none with problems sufficiently severe to warrant such a designation. This was mainly due to the fact that only two respondents out of a total of 201 admitted to bingeing more than once a week, an essential criterion for DSM caseness.
2
Note. aDefined *p < .05. **p < .Ol. ***p < ,001.
as fewer
Consciously tries to control weight Has rigid rules about eating Diets harshly to lose weight Eats same food daily Keeps a daily record of food Exercises to control weight Thinks of food 50% or more of the time Has intrusive thoughts about food and weight sometimes or often Terrified of gaining weight Constantly desires to be thinner Constantly preoccupied by fear of weight gain Amenorrhoea” Takes laxatives to lose weight sometimes or often Takes diuretics to lose weight sometimes or often
ITEM
32 16
26
79
32
63
68
68
74 5
21
21
4 9
9
55
10
35
33
32
25 4
9
7 year
18
24
59 12
77
65
82
65
65
24
18 29
47
82
Once a week or more (n = 17)
in the preceding
63
20
than three
95
58
Once a month (n = 19)
Binge frequency
NS
NS
*** NS
***
*
***
***
NS
NS
** NS
***
*
Sig. of x’
of eating-related
(n = 105)
as percentage,
Body group
I. Prevalence,
periods
Not at all (n = 69)
Table
2
2
11
16 0
16
16
19
4
46
5
2 7
16
42
Not at all (n = 57)
in two groups
13
39 6
46
45
48
23
61
14
II 13
32
67
Whole body group
issues, Ordinary
status
0
0
42 4
50
54
58
4
54
4
8 13
33
88
Once a month (n = 24)
Binge frequency
by binge (n = 96)
7
13
60 7
73
80
60
13
60
0
20 7
47
80
Once a week or more (n = 15)
group
NS
*
** NS
***
***
***
NS
NS
NS
* NS
*
***
Sig. of x2
2
3
29 2
33
35
35
5
50
4
6 8
25
59
Whole ordinary Group
476
J. F. O’MAHONY and S. HOLLWEY
Table 2. Means and standard deviations for descriptive variables for the two groups at different levels of bingeing severity Binge frequency Body group
Variable Food/weight preoccupation (maximum = 5) Behavioral control (maximum = 6) Neuroticism Body Mass Index Maximum ever BMI Minimum ever BMI
Not at all (n = 69)
Once a month or more (n = 19)
1.3 (1.4)
3.1 (2.0)
1.6 (1.3) 9.6 (6.4) 20.4 (2.3)
21.4 (3.0)
21.9 (3.0) 18.8 (2.1)
Ordinary group
Not at all (n = 57)
Once a month or more (n = 24)
3.5 (1.4)
0.7 (1.1)
2.1 (1.4)
2.9 (1.5)
3.1 (1.2)
2.6 (1.7)
11.5(5.7)
1I .9 (5.0) 22.0 (4.0)
1.2 (1.2) 8.8 (4.2) 21.2 (2.4)
2.0 (1.1) 11.5 (3.0) 22.2 (2.6)
2.1 (1.4) 10.7 (3.9) 22.7 (3.1)
22.8 (3.0)
23.7 (4.3)
22.4 (2.8)
22.8 (2.8)
23.7 (3.6)
19.5 (1.7)
20.1 (3.3)
19.8 (2.3)
20.6 (2.7)
20.4 (2.5)
Once a week or more (n = 17)
Once a week or more (n = 15)
Table 2 shows the scores on the various measures for persons who binge to a greater or lesser degree: neuroticism, behavioral control, food/weight preoccupation, BMI, maximum-ever BMI (a measure of past weight, which might have given cause for concern), and minimum-ever BMI. MANOVA conducted separately for the two groups shows that persons of different binge status differ on these variables (for the ordinary group, Wilks’s lambda = 58, approximate F(12,176) = 4.5, p < .OOl; for the body group Wilks’s lambda = .63, approximate F(12,194) = 4.3, p < .OOl). Follow-up unadjusted univariate F tests showed that in the ordinary group persons of different bingeing status differed on behavioral control (F(2,93) = 6.5, p < .005), food/weight preoccupation (F(2,93) = 22.2, p < .OOl), and neuroticism (F(2,93) = 4.0, p < .05), with BMI binge-group differences approaching significance (F(2,93) = 2.4, p = .094). For the body group there were differences on behavioral control (F(2,102) = 11.6, p < .OOl), food/weight preoccupation (F(2,102) = 19.3, p < .OOl), and neuroticism (F(2,102) = 3.4, p < .05), with differences for BMI again approaching conventional significance levels (F(2,102) =2.7, p = .07). Intercorrelations among the various variables and binge frequency are shown in Tables 3 and 4. Since maximum- and minimum-ever BMI are very highly correlated with BMI itself, and because they were not univariately significant in the previous analysis, they were not considered further. Direct discriminant analyses with frequency of bingeing as the dependent variable were conducted separately for both groups with all significant and nearly significant variables: for the body group, Behavioral control, Food/weight preoccupation and BMI; for the ordinary group, Behavioral control, Food/weight preoccupation, Neuroticism, and BMI. Sample sizes were used to estimate prior probabilities for group membership. These analyses yielded reliable associations (x2 (8) = 39.01, p < .OOl and x2 (8) = 45.0, p < .OOl, with classification efficiencies of 65% (expected chance
477
Correlates of bingeing
Table 3. Pearson product-moment
1
2
3
.33*** .25*
.54*** .32**
.25*
.22* .15 .12
.34*** .31** .09
.32*** .30** .09
Variable 1. Binge frequency 2. Preoccupation with food/weight 3. Behavioral control over eating 4. Neuroticism 5. Body mass index (BMI) 6. Maximum-ever BMI 7. Minimum-ever BMI
correlations among binge frequency and the descriptive the ordinary group
variables for
4
5
6
.lO .I0 .I0
.94*** .77***
.74***
7
.56***
Notes. All probabilities are two-tailed. N = 96. *p < .05. **p < .Ol. ***p < ,001.
classification 49%) and 69% (expected chance classification 44%) for the ordinary We conducted multiple step-down analyses (see and body groups, respectively. Tabachnick & Fidell, 1989) involving all combination of significant and marginally significant variables in order to evaluate the relative significance of variables. For the body group: all analyses incorporating food/weight preoccupation were significant (p < .05, here and following) and continued to be so even with behavioral control and/or BMI covaried out; with food/weight preoccupation in any analysis behavioral control added nothing further but BMI did; when BMI was covaried out, both behavioral control and food/weight preoccupation continued to add. Thus it seems that food/weight preoccupation is the major contributor, with BMI contributing something extra. The effect of behavioral control is largely accounted for by food/weight preoccupation. For the ordinary group: food/weight preoccupation was a significant predictor of bingeing when any or all of the variables - behavioral control, neuroti-
Table 4. Pearson product-moment Variable 1. Binge frequency 2. Preoccupation with food/weight 3. Behavioral control over eating 4. Neuroticism 5. Body mass index (BMI) 6. Maximum-ever BMI 7. Minimum-ever BMI
correlations among binge frequency and the descriptive the body group
I
2
4
5
6
-.09 .oo -.06
.90*** .83***
.78***
.51*** .35*** .I6 .22* .21* .21*
.62*** .36** -.oo .07** .Ol
Notes. All probabilities are two-tailed. N = 105. *p < .05.
**p i .Ol. ***p < ,001.
3
variables for
.23* -.Ol .05 -.05
7
478
J. F. O’MAHONY and S. HOLLWEY
cism, or BMI - were covaried; with food/weight preoccupation covaried no further stepdown equations were significant; BMI and neuroticism seemed to explain some nonoverlapping variance in that, when each was covaried in stepdown analysis, the other accounted for a significant amount of variance. DISCUSSION
As the univariate analyses show, binge frequency is significantly associated with the well-known risk factors of personality variables (neuroticism), weight (in the form of BMI), attempts to control eating, and preoccupation with food/weight issues. By a considerable margin, however, the largest association is with food/weight preoccupation. Multivariate analyses indicate that food/weight preoccupation mostly accounts for the effects of the other variables. While, as mentioned earlier, multifactorial models of the aetiology of eating disorders enjoy considerable popularity, the exact causal links in the genesis of eating problems in general and binge eating in particular are still very speculative. Our findings suggest that, among those with modest bingeing problems (we take the view that any bingeing is a problem), a key element in the bingeing process is food/weight preoccupation. This is consistent with the literature; the most consistent and strongest (compared with body weight, personality, and other variables) predictor of bingeing at all levels of severity of problem are such concerns (Bennett, Spoth, & Borgen, 1991; Lowe & Caputo, 1991). In this study, those who binge were not psychologically disturbed in any very marked or general way. Their neuroticism scores (see Table 2) were below the adult manual norms of Eysenck & Eysenck (1975). Their weight is in the normal range (Beumont, Al-Alami, & Toyuz, 1988). Eating behaviors/attitudes are not very abnormal: The vast majority try to control their weight, but only about 20% diet harshly; only a few used laxatives or diuretics. The major abnormality seems to be food/ weight preoccupation. It seems that bingeing can occur at quite low levels of personality abnormality, at reasonably normal and stable average weights, and in the absence of severe dietary restriction. What variance these factors contribute is largely accounted for by concerns about food, body, and weight. This is consistent with the near unanimous view (Hsu, 1980a; Schlundt & Johnson, 1990) that concern with body shape and weight are core features of the syndrome of bulimia nervosa and the addition of such concerns as defining features of the concept of bulimia nervosa in the DSM-III-R (American Psychiatric Association, 1987). These concerns had not been included in the original definition of bulimia, promulgated in the original DSM-III (American Psychiatric Association, 1980). Of course, this kind of correlational evidence in no way establishes such concerns as primary causes of bingeing; the sequence could begin with bingeing and proceed to food/weight concerns. There are also obvious weaknesses in the study. The data are entirely based on pencil-and-paper self-report, with all the inherent problems of this method. The eating measures used were not standard, though they had substantial face validity and were shown to be associated with other, more conventional measures. And the study was cross-sectional. Nonetheless, the evidence suggests that bingeing can occur without severe dietary restriction or general psychological abnormality, requiring merely a mindset fo-
Correlates of bingeing
479
cussed on weight, food, and eating. If such a mindset is sufficient to initiate the bingeing process, then the modern relentless portrayal of ideal body types in the media and the celebration of what is for many (Garner & Wooley, 1991) an unrealisable standard of thinness may well have its costs.
REFERENCES Abraham, S. F., & Beumont, P. J. V. (1982). How patients describe bulimia or binge eating. Psychologicul Medicine,
12, 625-635.
American Psychiatric Association (1980). Diagnostic and sratistical manual qf menral disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association (1987). Diagnostic and sturisticul manual ofmental disorders (3rd ed., rev.). Washington, DC: Author. Bennett, N. A. M., Spoth, R. L., & Borgen, F. H. (1991). Bulimic symptoms in high school females: Prevalence and relationship with multiple measures of psychological health. Journal of Community Psychology, 19, 13-28. Beumont, P., Al-Alami, M., & Toyuz, S. (1988). Relevance of a standard measurement of undernutrition to the diagnosis of anorexia nervosa: Use of Quetelet’s Body Mass Index. fnternationul Journal of Eating Disorders,
Cattanach,
7, 399-405.
L., & Rodin, J. (1988). Psychosocial
tionul Journal
of Eating Disorders,
components
of the stress process in bulimia. Internu-
7, 75-88.
Cooper, P. J., & Fairbum, C. G. (1983). Binge-eating and self-induced vomiting in the community: A preliminary study. British Journal of Psychiutry, 142, 139-144. Elmore, D. K., & de Castro, J. M. (1990). Self-related moods and hunger in relation to spontaneous eating behavior in bulimics, recovered bulimics, and normals. International Journal of Eating Disorders, 9. 55, 179-190.
Eysenck, H., & Eysenck, S. (1975). Manual ofthe Eysenck Personality Questionnaire. Sevenoaks, U.K.: Hodder and Stoughton. Garner, D. M., & Garfinkel, P. E. (1979). The Eating Attitudes Test: An index of the symptoms of anorexia nervosa. Psychological Medicine, 9, 273-279. Garner, D. M., & Wooley, S. C. (1991). Confronting the failure of behavioral and dietary treatments for obesity. Clinical Psychology Review, 11, 729-780. Garner, D. M., Olmsted, R., Davis, R., Rockert, W., Goldbloom, D., & Eagle, M. (1990). The association between bulimic symptoms and reported psychopathology. Internutionul Journul of Euting Disorders, 9, I-16.
Greenberg,
B. R. (1986). Predictors
Journal of Earing Disorders,
of binge eating in bulimic and nonbulimic women. Internutionul
5, 269-284.
Halmi, K. A. (1985). Rating scales in the eating disorders. Psychopharmacology Bulletin, 21, 1001-1003. Hawkins, R. C., & Clement, P. F. (1980). Development and construct validation of a self-report measure of binge eating tendencies. Addictive Behaviors, 5, 219-226. Heatherton, T. F., & Baumeister, R. F. (1991). Binge eating as escape from self-awareness. Psychologicul Bulletin. 110.86-108. Herzog, D. B., Keller, M. B., Lavori, P. W., & Ott. I. L. (1987). Social impairment in bulimia. Inrernutional Journal
of Eating Disorders,
6, 741-747.
Hsu, L. K. G. (1990a). Eating disorders. New York: Guilford Press. Hsu, L. K. G. (1990b). Experiential aspects of bulimia nervosa: Implications for cognitive behavioral therapy. Behavior Modijcation, 14, 50-65. Johnson, C. L., & Connors, M. E. (1987). The etiology und treatment of bulimia neruosu: A hiopsychosociul perspecfiue. New York: Basic Books. Kaye, W. H., Gwirtsman, H. E., George, D. T., Weiss, S. R., & Jimerson, D. C. (1986). Relationship of mood alterations to bingeing behaviour in bulimia. British Journul of Psychiatry, 149, 479-485. Keys, A., Brozek, J., Henschel, A., Michelson, O., & Taylor, H. L. (1950). The biology of human sturuution (Vol. 1). Minneapolis: University of Minnesota Press. Keys, A., Fidanza, F., Karvonen, M. J., Kimura, N., & Taylor. H. L. (1972). Indices of relative weight and obesity. Journal of Chronic Diseases, 25, 329-343. Lowe, M. R., & Caputo, G. C. (1991). Binge eating in obesity: Towards the specification of predictors. International Journal of Eating Disorders, 10,49-55. Patton, G. C., Johnson-Sabine, E., Wood. K., Mann, A. H., & Wakeling, A. (1990). Abnormal eating attitudes in London schoolgirls-a prospective epidemiological study: Outcome at twelve-month follow-up. Psychological Medicine, 20, 383-394. Polivy, J., & Herman, C. P. (1985). Dieting and bingeing: A causal analysis. Americun Psychologist, 40, 193-201.
J. F. O’MAHONY and S. HOLLWEY
480
Ruderman, A. J., & Besbeas, M. (1992). Psychological characteristics of dieters and bulimics. Journal of Abnormal Psychology, 101, 383-390. Schhmdt, D. G., & Johnson, W. G. (1990). Earing disorders: Assessment and treatment. Boston: Allyn and Bacon. Spitzer, R. L., Devlin, M. J., Walsh, B. T., Hasin, D., et al. (1992). Binge eating disorder: A multisite field trial of the diagnostic criteria. Internafional Journal ofEating Disorders, 11, 191-203. Steiger, H., Puentes-Newman, G., & Leung, F. Y. K. (1991). Personality and family features of adolescent girls with eating symptoms: Evidence for restricter/binger differences in a nonclinical population. Addictive Behaviors, 16, 303-314. Szmukler, G. I., Eisler, I., Gillis, C. & Hayward, M. E. (1985). The implications of anorexia nervosa in a ballet school. Journal of Psychiatric Research, 19, 177-181. Tabachnick, B. G., & Fidell, L. S. (1989). Using multivariate statistics 2nd ed.). HarperCollins: New York. Turnbull, J., Freeman, C. P. L., Barry, F., & Henderson, A. (1989). The clinical characteristics of bulimic women. International Journal of Eating Disorders, 8, 399-409. Vanderheyden, D. A., & Boland, F. J. (1987). A comparison of normals, mild, moderate, and severe binge eaters and binge vomiters using discriminant function analysis. International Journal ofEaring Disorders, 6, 331-337.
Wardle, J. (1987). Compulsive eating and dietary restraint. British Journal ofclinical 55.
Psychology, 26,47-
Williamson, D. A. (1990). Assessment of eating disorders: Obesity, anorexia, and bulimia nervosa. New York: Pergamon Press. Wilson, G. T. (1992). Diagnostic criteria for bulimia nervosa. International Journal ofEating Disorders, 11,315-319. Wurtman, R. J. (1989). Neurotransmitters, control of appetite and obesity. In M. Winick (Ed.), Control of appetite (pp. 27-34). New York: Wiley. Yates, A., Leehey, K., & Shisslak, C. M. (1983). Running-An analogue of anorexia? New England Journal
of Medicine,
308, 251-255.