Predictors of 12-Month Outcome in Bulimia Nervosa and the Influence of Attitudes to Shape and Weight Christopher G. Fairburn Department of Psychiatry Oxford University Robert C. Peveler Department of Psychiatry Oxford University Rosemary Jones Department of Psychiatry Oxford University R. A. Hope Department of Psychiatry Oxford University Helen A. Doll Department of Psychiatry Oxford University ABSTRACT Seventy-five patients with bulimia nervosa were treated with 1 of 3 short-term psychological treatments and were then entered into a closed 1-year period of follow-up. Pretreatment predictors of 3 measures of outcome were sought. Only 2 variables were significantly associated with outcome: attitudes toward shape and weight, and self-esteem. The nature of the relation between attitudinal disturbance and outcome was complex and unexpected. The data set was also used to test the major prediction of the cognitive view of bulimia nervosa, namely that among patients who have responded to treatment, the residual level of attitudinal disturbance will predict subsequent outcome. This prediction was confirmed. This research was supported by the Wellcome Trust. The study itself was funded by a project grant from the Wellcome Trust and three of the authors received personal support from the Trust at some stage during its course (a Senior Lectureship award to Christopher G. Fairburn and Research Training Fellowship awards to Robert C. Peveler and R. A. Hope). We are grateful to Jenny Burton, Sally Carr, Marianne O'Connor, and Ruth Solomon for their help with the study. Correspondence may be addressed to Christopher G. Fairburn, Department of Psychiatry, Oxford University, Warneford Hospital, Oxford, United Kingdom, OX3 7JX. Received: June 10, 1992 Revised: September 8, 1992 Accepted: September 27, 1992 Clinical evidence suggests that bulimia nervosa tends to run a chronic course because most patients present with a history of many years of symptoms, and often there have been prior attempts at treatment. Therefore, when evaluating treatments for bulimia nervosa, long-term response is the outcome of interest. Of the studies of short-term psychological treatment for bulimia nervosa, to our knowledge only one has systematically sought predictors of longer term outcome ( Fairburn, Kirk, O'Connor, Anastasiades, & Cooper, 1987 ). Fairburn et al. found that the pretreatment level of self-esteem was the sole consistent predictor of outcome, with patients having low self-esteem faring the least well. Because Journal of Consulting and Clinical Psychology © 1993 by the American Psychological Association August 1993 Vol. 61, No. 4, 696-698 For personal use only--not for distribution. Page 1 of 611/6/2000http://spider.apa.org/ftdocs/ccp/1993/august/ccp614696.htmlthis finding was based on data from a small sample, replication is necessary. The first aim of this study was to conduct a second examination of predictors of longer term outcome following short-term psychological treatment for bulimia nervosa. The data used were from a treatment trial in which three such treatments were compared. The second aim was to use this data set to test the central prediction of the cognitive view of bulimia nervosa: that in patients who have responded to treatment, the residual level of attitudinal disturbance will predict subsequent outcome with those patients having the greatest attitudinal disturbance doing the least well ( Fairburn, Cooper, & Cooper, 1986 ). Method Seventy-five patients with bulimia nervosa were randomly allocated to one of three short-term psychological treatments: cognitive—behavior therapy (CBT), a purely behavioral version of CBT, and a specific adaptation of interpersonal psychotherapy (IPT). 1 After treatment (which lasted 18 weeks and involved 19 treatment sessions), patients judged not to need immediate further treatment were entered into a closed 1-year period of follow-up. This article is concerned with predictors of their 1-year outcome. Patients were assessed before treatment, at the end of treatment, and at 4-, 8-, and 12-month follow-up. The 10th edition of the Eating Disorder Examination (EDE; Cooper & Fairburn, 1987 ; Fairburn & Cooper, 1993 ) was used to measure the eating behavior and attitudes characteristic of patients with eating disorders. The Symptom Checklist—90 (SCL—90); ( Derogatis, Lipman, & Covi, 1973 ) was used to assess general psychiatric symptoms. Self-esteem and personality disturbance were measured using the Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1965 ) and the Personality Diagnostic Questionnaire (PDQ; Hyler, Rieder, Spitzer, & Williams, 1978 ), respectively. The following pretreatment predictor variables were chosen on the basis of prior research findings and clinical reports: duration of the eating disorder and age at onset; history of anorexia nervosa; frequency of bulimic episodes (objective bulimic episodes on the EDE); severity of dietary restraint (EDE Restraint subscale score); severity of attitudinal disturbance (sum of the two EDE items that directly address the overvalued ideas of these patients ["importance of shape" and "importance of weight"]); overall severity of eating disorder psychopathology (mean of the five EDE subscale scores); level of general psychiatric disturbance (SCL-90 Global Severity Index); level of self-esteem (RSES score); and overall degree of personality disturbance (PDQ total score). For the test of the cognitive view of bulimia nervosa, the predictor variable chosen was the posttreatment level of attitudinal disturbance (measured as before). Three indexes of 12-month outcome were used because there is no satisfactory single measure of the severity of bulimia nervosa. Outcome Index 1 was the overall severity of eating disorder psychopathology (measured as previously indicated). Outcome Index 2 was whether each subject's overall level of eating disorder psychopathology fell within 1 SD of the mean of young women from the same geographical area. 2 Outcome Index 3 was whether the subject had not only ceased to have episodes of uncontrolled overeating (both objective and subjective bulimic episodes on the EDE) but had also ceased to vomit and take laxatives. This represents a particularly good outcome in behavioral terms. Results Twenty-five (33%) of the original 75 patients either dropped out of treatment or follow-up or were Page 2 of
View Full Document
Unlocking...