Clinical moderators and predictors of cognitive-behavioral therapy by guided-self-help versus therapist-led for binge-eating disorder: Analysis of aggregated clinical trials.

Abstract

Cognitive-behavioral therapy (CBT)-therapist-led (CBTth) and guided-self-help (CBTgsh)-has efficacy for binge-eating disorder (BED) but many patients do not benefit sufficiently. We examined predictors and moderators for these two CBT methods. Data were aggregated from randomized controlled trials (RCTs) testing psychosocial treatments for BED in the U.S. Predictors and moderators of outcomes (treatment completion and binge-eating remission) were examined in N=457 participants who received either CBTgsh (N=164) or CBTth (N=293). Analyses, adjusting for demographic/clinical variables, indicated CBTth was significantly superior to CBTgsh for treatment completion (odds ratio [OR]=20.0) and remission (OR=14.6). For remission, analyses revealed significant predictors (age, treatment length, Weight Concern), a moderator (weight concern [OR=5.13]), and a significant interaction between CBT-type and treatment length (OR=2.66). For CBTgsh, longer treatment was associated with less remission, whereas for CBTth, longer treatment was associated with greater remission. For CBTgsh, 44.1% with low weight concern versus 56.3% with high weight concern achieved remission whereas for CBTth, 43.5% with high weight concern and 61.0% with low weight concern achieved remission. Analyses of aggregated RCT BED data, adjusting for demographic/clinical characteristics, indicated superiority (large effect-sizes) in treatment outcomes of CBTth over CBTgsh and that Weight Concern moderated outcomes.