Interpersonal problems as a predictor of treatment outcome in adult depression: An individual participant data...

Interpersonal problems are a fundamental feature of depression, but study-level meta-analyses of their association with treatment outcome have been limited by heterogeneity in primary studies' analyses and reported results. We conducted a pre-registered individual participant data meta-analysis (IPD-MA) to examine this relationship for adult depression. This meta-analytic strategy can reduce variability by standardizing data analysis across primary studies.

We included studies examining the efficacy of five treatments for adult depression and assessing interpersonal problems at baseline. One-stage IPD-MA was conducted with three-level mixed models to determine whether baseline overall interpersonal distress, agency, and communion predicted depressive symptom level at post-treatment, 12-month, and 24-month follow-up. The moderating effect of treatment type was also investigated.

Ten studies (including n = 1282 participants) met inclusion criteria. Only overall interpersonal distress was negatively related with outcomes at post-treatment (γ = 0.11, CI₉₅[0.06, 0.16], r = 0.11), 12-month follow-up (γ = 0.17, CI₉₅[0.08, 0.25], r = 0.17), and 24-month follow-up (γ = 0.16, CI₉₅[0.05, 0.26], r = 0.16), indicative of smaller effect sizes. The agency and communion dimensions were not significantly related to outcome. Treatment type did not significantly moderate interpersonal distress-outcome associations.

Results show a small association between patient baseline overall interpersonal distress and subsequent depression treatment outcome in brief treatments for depression. Further studies might require to account for therapist effects. Registration number osf.io/u46t7.