Therapeutic Alliance With Depressed Adolescents: Predictor or Outcome? Disentangling Temporal Confounds to Understand Early Improvement

Document Type


Publication Date


Digital Object Identifier (DOI)



Psychotherapy research reveals consistent associations between therapeutic alliance and treatment outcomes in the youth literature; however, past research frequently suffered measurement issues that obscured temporal relationships between alliance and symptomatology by measuring variables later in therapy, thereby precluding examination of important early changes. The current study aimed to explore the directions of effect between alliance and outcome early in therapy with adolescents by examining associations between first- and fourth-session therapeutic alliance and symptomatology. Thirty-four adolescents (∼63% female, 38% ethnic/racial minority) participated in a school-based cognitive-behavioral therapy for adolescents with depression. Participants completed the Beck Depression Inventory at baseline and Session 4, and therapeutic alliance was coded from audiotapes of Sessions 1 and 4 by objective coders using the Alliance Observation Coding System. Autoregressive path analyses determined that first-session therapeutic alliance was a strong significant predictor of Session 4 depression symptoms, but pretreatment depression scores were not significantly predictive of subsequent therapeutic alliance. Adding reciprocal effects between alliance and depression scores did not adversely affect model fit, suggesting that reciprocal effects may exist. Early therapeutic alliance with adolescents is critical to fostering early gains in depressive symptomatology. Knowing alliance's subsequent effect on youth outcomes, clinicians should increase effort to foster a strong relationship in early sessions and additional research should be conducted on the reciprocal effects of therapeutic alliance and treatment outcome in adolescence.

Was this content written or created while at USF?


Citation / Publisher Attribution

Journal of Clinical Child & Adolescent Psychology, v. 46, issue 4, p. 600-610