Abstract
OBJECTIVE: To test the validity and clinical utility of the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5) severity specifiers for binge-eating disorder (BED) in predicting treatment outcomes. METHOD: Participants (N = 521) were patients in randomized controlled trials (RCTs) at 1 medical center testing treatments for BED; data were aggregated from RCTs testing cognitive-behavioral therapy (CBT), behavioral weight loss (BWL), and/or multimodal (i.e., CBT or BWL plus pharmacotherapy) treatment. Participants were categorized according to DSM-5 severity specifiers for BED: "mild" (n = 273; 52.5%), "moderate" (n = 182; 34.9%), and "severe/extreme" (n = 58; 11.1%). Participants had their weight and height measured and were assessed using established interviews and self-report measures at baseline, throughout treatment, and post treatment. RESULTS: Mixed models revealed that severity category did not significantly predict treatment response. However, there were main effects of BED severity category: "Severe/extreme" BED had greater binge-eating frequency and had greater global eating-disorder psychopathology than did "mild" BED across all time points. "Severe/extreme" BED was less likely to have remission from binge eating than was "mild" BED. Weight loss and depression scores did not significantly differ by severity category. CONCLUSION: In a large series of treatment-seeking individuals with BED aggregated across RCTs at 1 medical center testing psychological and pharmacological treatments for BED, DSM-5 severity specifiers for BED had limited validity and utility predicting response to treatments. Future research is needed to identify more robust severity indicators with clinical utility to inform future DSM revisions and clinical practice. (PsycINFO Database Record (c) 2020 APA, all rights reserved).