Changes in emotion regulation are associated with treatment outcomes in web-based eating disorder interventions: analyses of two randomized controlled trials.

OBJECTIVE: Web-based interventions effectively reduce eating disorder symptoms; however, the associated mechanisms remain largely unknown. This study examined whether changes in emotion regulation processes were linked to treatment response in two web-based cognitive-behavioral self-help interventions for bulimia nervosa (BN) and binge eating disorder (BED), and tested pre-treatment emotion regulation as a moderator. METHOD: Data from two randomized controlled trials (N = 308, age: M = 32.76, SD = 10.12; female: 96.43%) were analyzed. Latent-change models and network intervention analyses were used to estimate indirect associations between treatment and changes in behavioral symptoms (primary outcome), global eating disorder psychopathology, and clinical impairment (secondary outcomes) through changes in emotion regulation difficulties and strategy use. Moderator effects were tested using linear mixed models. RESULTS: Significant indirect pathways from treatment allocation to pre- to post-treatment outcome changes through concurrent decreases in emotion regulation difficulties emerged for BN and BED (partially standardized indirect paths: |ab| ≥ 0.10). Treatment was linked with changes in global eating disorder psychopathology and clinical impairment through concurrent, combined decreases in rumination, suppression, and avoidance in BN (partially standardized overall indirect paths: |ab| ≥ 0.17) and combined increases in acceptance, reappraisal, and problem-solving in BED (partially standardized overall indirect paths: |ab| ≥ 0.19). No significant differences in the indirect paths between BN and BED were found. Baseline-to-mid-treatment changes in emotion regulation did not predict subsequent changes in outcomes. No consistent moderator effects were detected. CONCLUSION: Reductions in emotion regulation deficits were associated with concurrent symptom declines in web-based eating disorder treatments, pointing to emotion regulation as a candidate change mechanism. The absence of temporal effects precludes causal claims.