German Journal of Psychiatry

ISSN 1433-1055

Body Image Dissatisfaction in Bulimia Nervosa and Atypical Bulimia Nervosa

Francisco J. Vaz, Eva M. Peñas, and María I. Ramos

Department of Pharmacology and Psychiatry, University of Extremadura Medical School (Badajoz, Spain). Corresponding Author: Prof. Francisco J. Vaz, Área de Psiquiatría, Facultad de Medicina de la UEX, Avda. de Elvas s/n, 06071 Badajoz, Tel./Fax: +34 924 28 94 56, E-mail: fjvaz@unex.es



The current study was undertaken to provide descriptive data regarding the body image dissatisfaction of two clinical populations: 39 patients with bulimia nervosa, and 22 patients with atypical/subclinical bulimia nervosa. A control group of 50 individuals without eating disorders was included. Three different techniques for assessing body image dissatisfaction were used (difference between real and ideal Body Mass Index, selection of real and ideal drawn images, and selection of dissatisfactory parts of the body). The three methods produced similar results, confirming the central role of body dissatisfaction in the psychopathology of bulimia nervosa. The study also showed that some differences existed among the patients with complete and atypical/subclinical forms of bulimia nervosa, affecting the way they experienced their bodies. The paper discusses the relevance of these differential clinical findings in the development of bulimia nervosa and the utility of identifying and isolating the atypical/subclinical forms of bulimia (German Journal of Psychiatry, 1999;2:59-74)

Key words: Bulimia nervosa, atypical bulimia nervosa, subclinical bulimia nervosa, eating disorders, body image assessment, body image dissatisfaction

Received: 17.9.99

Published: 18.10.99



The term "body image" has been used in literature to refer to at least three related items: (1) the mental image that a person has of him/herself on a physical level; (2) the evaluation that the person makes of his/her own physical appearance; and (3) the influence that such perceptions and attitudes have on behavior (Rosen, 1990). Body image and its disturbances have for many years been related to eating disorders (Rosen, 1995; Thompson, 1990). Hilde Bruch (1962) highlighted the role that body image disturbance had in the etiology of anorexia nervosa. Since Bruch, other authors have examined this topic, and at present there is a general assumption that the distortion of the body image is a primordial element in the psychopathology of anorexia and bulimia nervosa. In this way, DSM-IV (American Psychiatric Association, 1994) considers that "disturbance in the way in which one’s body weight or shape is experienced, undue influence of body weight or shape on self-evaluation, and denial of the seriousness of the current low body weight" are basic diagnostic criteria for anorexia nervosa. For bulimia nervosa, however, DSM-IV is rather less precise, pointing out only that "self-evaluation is unduly influenced by body shape and weight". The ICD-10 (World Health Organization, 1992) is much more explicit, considering the alteration of the body image of patients with anorexia nervosa an overvalued idea. In this sense, ICD-10 clearly specifies that a specific distortion of the body image exists in patients with anorexia nervosa. For bulimia nervosa, on the other hand, the manual of the WHO is also more explicit than DSM-IV, stressing that the central psychopathological features of bulimia nervosa and anorexia nervosa are very similar, and states that bulimic patients have a morbid fear of gaining weight, choosing for themselves a target body weight significantly lower than the premorbid one.

In accordance with current theories on eating disorders, distortion of the body image can be considered a multidimensional process that includes elements of a perceptive, attitudinal and behavioral nature (Kearney-Cooke and Striegel-Moore, 1994, 1997; Rosen, Saltzberg and Srebnik, 1989; Slade, 1990; Williamson, 1990). It is also associated with other factors, such as cultural patterns of feminine beauty, learning experiences that take place in the bosom of the family, the development of ego-identity and sense of self-effectiveness, psychosexual development and consolidation of self-esteem (Rosen, 1995).

Two basic mechanisms have been proposed to explain the genesis of body image disturbance, not only in patients with eating disorders, but also in other clinical groups, such as patients with mood or dysmorphic disorders: internalization and projection. The first should be related to negative past experiences involving the body, such as sexual aggression or severe surgical interventions, and also to interpersonal relationships that cause physical appearance to have great importance on self-evaluation, such as the existence of a history of family obsession with obesity or body appearance (Kearney-Cooke and Striegel-Moore, 1994). The second mechanism has to do with the confusion between body and self, the individual using the body like a "battlefield" in his/her psychological and interpersonal conflicts, associating the control of the body with a sense of self-control and self-effectiveness (Heatherton and Baumeister, 1991). These two mechanisms can combine to generate "negative" body schemata (in the sense given classically to this term in cognitive theory), body perceptions, affects and cognitions being equally "negative", which prepare the way for the appearance of "compensatory" eating behaviors. In this sense, the assessment of body image disturbance seems to be a basic procedure for the understanding and the treatment of eating disorders since, if its contribution to the etiology of such disorders is as decisive as most clinical studies indicate, its persistence could condition a negative outcome and an increased risk of recurrence (Kearney-Cooke and Striegel-Moore, 1997; Vaz, 1998).

A clinical differentiation between body image disturbance and body image dissatisfaction should be established. Both terms are often identified, and this identification produces some important clinical problems, because body dissatisfaction only can be considered as a manifestation of body image disturbance, which is a much wider concept and includes other important elements (Thompson, 1995).

In this context, our investigation was basically aimed at determining how body image dissatisfaction could contribute to the psychopathology of eating disorders and, more specifically, to the psychopathology of bulimia nervosa. We assessed the intensity of body image dissatisfaction in two clinical populations (patients with full-blown and atypical forms of bulimia nervosa), comparing them with a control group. The assessment was made using three clinical methods which, as will be seen, gave concordant results and allowed us to draw some conclusions that could open lines of work for future investigations.



The details of the constitution of the group under study and of the control group are presented, together with the assessment methods and the statistical procedures used in the investigation.


The group under study was composed of 61 female bulimic patients requesting treatment at a university eating disorders unit. Thirty-nine of these patients (64%) met DSM-IV criteria for the diagnosis of bulimia nervosa [307.51], and they were included in the "bulimia nervosa group" (BN group). In the other twenty-two patients, one or more of the diagnostic criteria were absent (for example, frequency of vomiting or binging episodes), and they were included in the "subclinical group" (subBN group). The diagnosis for the patients of the subBN group was "eating disorder not specified" [307.50], in accordance with the terminology of DSM-IV, or "atypical bulimia nervosa" [F50.3], according to the categories of ICD-10. In order to avoid excessive fragmentation of the groups, the patients with purgative and non-purgative forms were not broken up into different subgroups. Neither group presented significant differences in age or time of evolution of the eating problem.

Control group

This was composed of 50 female students without present/past history of eating disorders who voluntarily accepted to participate in the study. The control group individuals presented similar characteristics regarding age, educational and socio-economic level to those of the group of patients.


Patients and controls were evaluated with the help of the same diagnostic clinical tools. The instruments used were as follows:

a) Clinical assessment

Interview for Diagnosis of Eating Disorders —IDED— (Williamson, 1990), which was applied at the start of the consultation for the diagnosis, positive (in the patients) and negative (in the controls). The IDED is a semi-structured interview, composed of four blocks, aimed at the diagnosis of anorexia nervosa, bulimia nervosa and binge eating disorder, using DSM criteria.

b) Evaluation of body dissatisfaction

Body Image Assessment (BIA; Williamson et al, 1985), which was applied equally on initial contact with the subjects. It is a set of nine female silhouettes, drawn in black on white cards of 16x11.5 cm, representing women of increasing body size. The cards were presented randomly and the subjects were requested to select the card that in their opinion best matched their real body size (RBS). Then the process was repeated, offering the cards and requesting the subjects to select their ideal body size (IBS). On the back of each card there was a number from 1 to 9, and it was thus possible to obtain, by calculating the difference RBS—IBS, an index of body image dissatisfaction.

Real/Ideal Body Mass Index. These were calculated after measuring the height and weight of the individuals. The real Body Mass Index (RBMI) was obtained from the real weight and height, while the ideal Body Mass Index (IBMI) was obtained from the ideal weight (the weight that the individuals desired for themselves) and the real height. By calculating the difference RBMI—IBMI another index of body dissatisfaction was obtained.

Body areas that produced dissatisfaction. Two drawn female images (front and back) were offered to the individuals, and they were requested to select on the drawing the parts of their bodies with which they did not feel satisfied. One of the images used in the study is reproduced at the top of Figure 1 (front figure).

Figure 1. Ideal representation of body image dissatisfaction in the three populations studied

Statistical procedures

Comparison of the data was carried out by one-way analysis of variance (ANOVA) using the program StatView 512+. The results of the overall test are given in each table; nevertheless, and in order to avoid a Type I error, only the comparisons that were significant at 99% in the post-hoc analysis (Scheffé test) will be discussed.


Table 1 shows the items related to eating behavior and attitudes toward the body in the studied sample. It can be observed that significant differences were obtained for all the items, with highly significant values in most of the cases.

The real BMI was significantly higher in the patients of the subBN group than in the patients of the BN group and in the controls. These last two groups presented very similar values in this parameter. It is remarkable, however, that the patients of the BN group presented an ideal BMI significantly lower than those in the other two groups, which were very similar. By calculating the index of dissatisfaction RBMI—IBMI it was possible to observe a gradation, with values that were very high for the subBN group, intermediate for the BN group and lower for the control group.

Regarding the data obtained when applying the BIA, we can see that the patients of the subBN group identified a real body image significantly more obese than those of the other two groups (SubBN vs. controls significant at 99%), but we also see that in the patients of the BN group the values were also somewhat higher than in the controls. On the other hand, when choosing the ideal body image, the patients of the BN group selected the thinner images, so that the differences were compensated (BN vs. controls significant at 99%), the levels of dissatisfaction (RBS-IBS) being very similar in the BN and subBN groups (there is a significant difference between SubBN and controls, and also between BN and controls, but not between SubBN and BN).

Table 2 shows the dissatisfaction produced by the different parts of the body in the three groups studied. As we can see, both groups of patients (BN and subBN) felt more dissatisfied with almost all the areas of their bodies, especially with their thighs, than did the controls. The patients of the BN group were especially dissatisfied with their faces, arms and backs, showing also higher levels of global dissatisfaction.

Table 1. Items related to attitudes toward the body. Abbrevations in the text





F value







.0003 (a)(b)






.0056 (c)






.0001 (a)(b)(c)






.0012 (b)






.0006 (c)






.0001 (a)(c)

(a) Significant at 99% SubBN vs. BN (Scheffé F-test)

(b) Significant at 99% SubBN vs. Controls (Scheffé F-test)

(c) Significant at 99% BN vs. Controls (Scheffé F-test)


Table 2. Levels of dissatisfaction referring to diverse areas of the body





F value













.0085 (c)
























.0059 (c)












.0004 (c)
























.0001 (b)(c)
























.0003 (b)(c)

(a) Significant at 99% SubBN vs. BN (Scheffé F-test)

(b) Significant at 99% SubBN vs. Controls (Scheffé F-test)

(c) Significant at 99% BN vs. Controls (Scheffé F-test)



The results presented confirm, in our opinion, the hypothesis that body image dissatisfaction is a basic symptom in patients with bulimia nervosa, and is also a symptom that makes it possible to differentiate between patients with complete and atypical/subclinical forms of bulimia nervosa. As we pointed out in a recently published paper (Vaz, Peñas and Guisado, 1998), the study of subclinical and atypical forms of bulimia nervosa is very interesting, since they often correspond to the first evolutional phases of the disorder or have their origin in the use of specific methods of weight control (Vaz and Peñas, in press). According to this point of view, the assessment of body image dissatisfaction could provide some interesting topics for discussion.

One of the most interesting findings in our study is that the patients of the subBN group had higher average weight than the patients of the BN group (higher BMI). Although the average value of the BMI of these patients fell inside the parameters that are considered normal (and even healthy), they were above the values of the control group and in clear opposition to the current social ideal for women. This fact could contribute to maintaining high levels of body dissatisfaction and these patients’ struggle with their weight. On this level, they contrasted with the patients of the BN group, who presented very similar BMI values to those of the control group. If we keep in mind the idea that the subclinical forms of bulimia nervosa can represent either a precocious evolutional phase of the disorder or a residual form (Abraham, Mira and Llewellyn-Jones, 1983; Fairburn et al, 1995; Herzog, Hopkins and Burns, 1993; Steinhausen and Seidel, 1993), we could consider that the passage from the incomplete forms to the complete forms might be conditioned by the success of the behaviors aimed at weight control (diet, vomiting, use of laxatives, etc.) and that the transition from the clinical to the subclinical forms might be conditioned by the failure of such mechanisms. In this way, the patients with complete forms would be able to maintain a body weight similar to that of the population without eating disorders, in accordance with the social ideal, attributing the "success" to the maintenance of the symptoms. Nevertheless, we think that it is also possible to elaborate a more "positive" point of view regarding the patients with subclinical bulimia nervosa, not simply considering them "unsuccessful" bulimic patients, since, according to our results, they have more realistic expectations concerning their body weight and they want to achieve a body size similar to that of the control population. The high dissatisfaction detected through the BMI may be conditioned by the fact that their weight was higher, but in general they wished for a weight quite consistent with a healthy one, differing from the patients with complete forms, who pursued a much lower body weight. They seemed also to be more realistic when evaluating their body size, selecting higher sized silhouettes. This selection did not take place in the patients with complete forms, who in spite of having a body weight similar to the controls, looked for a much thinner body size. We think that the tendency to select thinner silhouettes reflects not only the existence of high body dissatisfaction, but also a distortion in the perception of the body image, since they tended to overestimate the body size and to pursue an especially low weight. Returning to the need for a differentiation between body image dissatisfaction and body image distortion, as we discussed in the introduction, it seems evident that body dissatisfaction was the basic feature in our subclinical population, whereas dissatisfaction and distortion were complementary features in the population with complete forms. This finding could confirm the proximity of the core psychopathology of anorexia nervosa and bulimia nervosa, along the lines pointed out in ICD-10.

Finally, another finding should be commented on. It has to do with what we could denominate "topography" of body dissatisfaction. Figure 1 aims to represent graphically the intensity of body dissatisfaction in the three studied populations, by means of the deformation of the areas of the body according to the levels of dissatisfaction referring to each of them (using a procedure similar to that used by Penfield and Rasmussen in their representation of the sensorial and motoric brain areas). As we can see, in our culture of today, young women seem to be dissatisfied with the body area between the chest and the knees (abdomen, hips, buttocks, thighs...), that is to say, the area where the fat tends to accumulate after puberty. Something similar, although amplified, happens in the patients with subclinical bulimia nervosa, this finding being in accordance with the higher levels of body dissatisfaction detected in our study. In the patients with complete forms of bulimia nervosa, however, the dissatisfaction, apart from being higher, reaches to specific areas (such as the face, the arms or the back) that are not easily modifiable by dieting. In our opinion, the specificity of body dissatisfaction may be another factor responsible for the maintenance of the disorder, because body dissatisfaction continues in spite of the "success" in weight control that the compensatory behaviors provide.


Abraham SF, Mira M, Llewellyn-Jones C. Bulimia: a study of outcome. International Journal of Eating Disorders 1983;2:175-180

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV). Washington DC: American Psychiatric Press; 1994

Bruch H. Perceptual and conceptual disturbances in anorexia nervosa. Psychosomatic Medicine 1962;24:187-194

Fairburn CG, Norman PA, Welch SL, O’Connor ME, Doll HA, Peveler RC. A prospective study of outcome in bulimia nervosa and the long-term effects of three psychological treatments. Archives of General Psychiatry 1995;52:304-312

Heatherton TF, Baumeister RF. Binge eating as escape from self awareness. Psychological Bulletin 1991;110:86-108

Herzog DB, Hopkins JD, Burns CD. A follow-up study of 33 subdiagnostic eating disordered women. International Journal of Eating Disorders 1993;14:261-267

Kearney-Cooke A, Striegel-Moore R. Treatment of childhood sexual abuse in anorexia nervosa and bulimia nervosa. A feminist psychodynamic approach. International Journal of Eating Disorders 1994;15:305-319

Kearney-Cooke A, Striegel-Moore R. The etiology and treatment of body image disturbance. In: Garner DM, Garfinkel PE (eds) Handbook of treatment for eating disorders, 2nd edition. New York: Guilford Press; 1997:295-306

Rosen JC. Body image disturbances in eating disorders. In: Cash T, Pruzinsky T (eds) Body images: development, deviance, and change. New York: Guilford Press; 1990:190-214

Rosen JC. Assessment and treatment of body image disturbance. In: Brownell KD, Fairburn CG (eds) Eating disorders and obesity. A comprehensive handbook. New York: Guilford Press; 1995:369-373

Rosen JC, Saltzberg E, Srebnik D. Cognitive behavior therapy for negative body image. Behavior Therapy 1989;20:393-404

Slade PD. What is body image? Behaviour Research and Therapy 1994;32:497-502

Steinhausen HC, Seidel R. Outcome in adolescents with eating disorders. International Journal of Eating Disorders 1993;14:487-496

Thompson JK. Body image disturbance. Assessment and treatment. Elmsford, NJ: Pergamon Press; 1990

Thompson JK. Assessment of body image. In: Allison DB (ed) Handbook of assessment methods for eating behaviors and weight related problems. Measures, theory, and research. Thousand Oaks, CA: Sage Publications; 1995:119-148

Vaz FJ. Outcome of bulimia nervosa: prognostic indicators. Journal of Psychosomatic Research 1998;45: 391-400

Vaz FJ, Peñas EM, Guisado JA. Identidad clínica y relevancia de los trastornos alimentarios atípicos y subclínicos. Anales de Psiquiatría 1998;14:8-14

Vaz FJ, Peñas EM. Estudio diferencial de las formas completas y subclínicas de bulimia nerviosa. Actas Españolas de Psiquiatría (in press)

Williamson DA. Assessment of eating disorders. Obesity, anorexia, and bulimia nervosa. New York: Pergamon Press; 1990

Williamson DA, Kelley ML, Davis CJ, Ruggiero L, Blouin DC. Psychopathology of eating disorders: A controlled comparison of bulimic, obese, and normal subjects. Journal of Consulting and Clinical Psychology 1985;53:161-166

World Health Organization. The ICD-10 Classification of Mental and Behavioural Disorders: Clinical descriptions and diagnostic guidelines. Geneve: WHO; 1992