Therapeutic alliance in a randomized clinical trial for bulimia nervosa.

OBJECTIVE This study examined the temporal relation between therapeutic alliance and outcome in two treatments for bulimia nervosa (BN). METHOD Eighty adults with BN symptoms were randomized to 21 sessions of integrative cognitive-affective therapy (ICAT) or enhanced cognitive-behavioral therapy (CBT-E). Bulimic symptoms (i.e., frequency of binge eating and purging) were assessed at each session and posttreatment. Therapeutic alliance (Working Alliance Inventory) was assessed at Sessions 2, 8, 14, and posttreatment. Repeated-measures analyses using linear mixed models with random intercepts were conducted to determine differences in alliance growth by treatment and patient characteristics. Mixed-effects models examined the relation between alliance and symptom improvement. RESULTS Overall, patients in both treatments reported strong therapeutic alliances. Regardless of treatment, greater therapeutic alliance between (but not within) subjects predicted greater reductions in bulimic behavior; reductions in bulimic behavior also predicted improved alliance. Patients with higher depression, anxiety, or emotion dysregulation had a stronger therapeutic alliance in CBT-E than ICAT, while those with more intimacy problems had greater improvement in therapeutic alliance in ICAT compared to CBT-E. CONCLUSIONS Therapeutic alliance has a unique impact on outcome, independent of the impact of symptom improvement on alliance. Within- and between-subjects effects revealed that changes in alliance over time did not predict symptom improvement, but rather that individuals who had a stronger alliance overall had better bulimic symptom outcomes. These findings indicate that therapeutic alliance is an important predictor of outcome in the treatment of BN.