The impact of symptom severity and interpersonal problems in psychotherapy for depression with or without comorbid personality disorders: a sub-analysis of a randomised clinical trial

Short-term psychotherapies are effective for depression, but remission rates remain modest. Given the frequent co-occurrence of depression, personality disorders, and interpersonal problems, this study examined the association of comorbid personality disorders (PD) on depression outcome while accounting for baseline symptom severity and interpersonal problems. Data were derived from a randomized clinical trial comparing two 16-session psychotherapies for depression. Multilevel regression analyses tested whether baseline symptom severity (composite index of standardized scores of symptom measures), interpersonal problems (Outcome Questionnaire interpersonal relations scale), comorbid personality disorders (International Personality Disorder Examination – Self Report), predicted outcome (Hamilton Depression Rating Scale). Depressed patients with comorbid PD had higher baseline symptom severity and interpersonal problems. They showed a trend toward one-year follow-up ( EMMs = 13.44 vs. 10.44, t = -1.84, P = .07). However, after controlling for baseline symptom severity, neither comorbid PD nor interpersonal problems predicted depression outcome, whereas baseline symptom severity remained a robust predictor of follow-up depression ( B = 3.17, SE = 0.44, P < .01). The negative association of comorbid PD on depression outcome is explained by baseline symptom severity. Baseline severity, rather than PD or interpersonal problems, is the most robust predictor of treatment response. • Depressed patients with and without comorbid PD improved over one year, but patients with comorbid PD consistently reporting higher symptom levels. • Comorbid PD was not associated with differential depressed symptom change over time, but showed a trend before adjustment of baseline severity index and interpersonal problems. • Higher follow-up depression in patients with comorbid PD was explained by baseline severity index and interpersonal problems. After additional adjustment, baseline symptom severity remained the primary predictor of outcome