Group cognitive-behavior therapy for bulimia nervosa: Statistical versus clinical significance of changes in symptoms across treatment

Background: Cognitive-behavior therapy (CBT) is the most effective treatment to date for bulimia nervosa. The current study investigated the effects of group CBT treatment (including some interpersonal elements) for bulimic clients. Method: Twenty-nine patients completed the Stirling Eating Disorder Scales, the Beck Depression Inventory, and the Beck Anxiety Inventory at assessment, pretreatment, end of treatment, and at 6 months follow-up. Symptom change was explored in two ways. Statistically significant change was determined using repeated-measures analyses of variance and clinically significant change was determined using criteria proposed by Jacobson & Truax (1991, journal of Consulting and Clinical Psychology, 59, 12-19). Results: There was an overall improvement in dimensional measures of bulimic and restrictive attitudes and behaviors (maintained at the 6-month follow-up), which was most closely matched by clinically significant changes in bulimic behaviors. Depression (but not anxiety) was also targeted effectively. Statistically significant improvements in psychological functioning were evident only for assertiveness, but the analysis of clinical significance showed improvement for some participants in self-evaluation (self-directed hostility). Discussion: Group CBT (including interpersonal elements) is broadly effective when treating bulimia nervosa, but it does not work in all cases (and may lead to enhancement of restrictive characteristics in some cases). Tests of statistical and clinical significance provide different information, which can inform practice and aid in the development of treatments for patients who respond less well to current best practice. (C) 2004 by Wiley Periodicals, Inc.