How do cognitive behavioral therapy and interpersonal psychotherapy improve youth depression? Applying meta-analytic structural equation modeling to three decades of randomized trials.

ABSTRACT

Investigating the mechanisms through which psychotherapy brings about desired change can inform efforts to improve therapies. We applied meta-analytic structural equation modeling (MASEM) to assess putative change mechanisms for cognitive behavioral therapy (CBT) as mediators of youth depression treatment outcome. Then, we tested whether candidate mediators (CMs) showed evidence of treatment-specificity to CBT versus interpersonal psychotherapy (IPT). Literature searches identified 34 randomized trials (27 CBT, 6 IPT, 1 CBT/IPT, 3,868 participants, published 1982-2020) that measured seven CMs negative cognition, social engagement, family functioning, pleasant activity engagement, problem solving, reframing, or avoidance. We assessed mediational pathways and whether posttreatment CMs mediated treatment effects on posttreatment depression symptoms, covarying pretreatment CMs, and symptoms. Treatment type was tested as a moderator of mediational pathways. Results show that negative cognition (24 trials) and pleasant activities (3 trials) mediated depression symptom outcome in CBT. Social engagement and family functioning showed stronger mediation in IPT (5 and 6 trials) than in CBT (14 and 13 trials). We conclude that negative cognition is a robust mediator of CBT but may not be treatment-specific; pleasant activities may also be a mediator of CBT. However, the lack of treatment or mediation effects involving problem solving and reframing contradicts CBT theory. In contrast, social and family mechanisms appear to be IPT-specific mediators. These conclusions are provisional due to small samples examining IPT and several CMs, limitations in CM measurement (i.e., posttreatment retrospective report), and assumptions of MASEM-and will need to be confirmed when more and better evidence accumulates. (PsycInfo Database Record (c) 2023 APA, all rights reserved).