Objective

Internet-delivered psychotherapies show promise for adolescent major depression (MDD), but little is known about which baseline characteristics predict improvement or moderate differential effects between internet-delivered cognitive-behavioral therapy (ICBT) and psychodynamic therapy (IPDT). This study examined pretreatment predictors and moderators of symptom change in a randomized trial comparing the treatments.

Method

This secondary analysis used data from a non-inferiority trial (n = 272) in which adolescents with MDD received 10 weeks of guided ICBT or IPDT. Mixed-effects models tested predictors and moderators of change in self-rated depression. Anxiety symptoms, length of depressive episode, emotion regulation, personality disorder severity, attachment, self-compassion, suicidal ideation, and baseline depression were examined as predictors and moderators.

Results

Higher self-compassion predicted steeper improvement across treatments, as the only significant predictor. Comorbid anxiety moderated outcome where higher anxiety was associated with significantly greater improvement in IPDT relative to ICBT, with no difference at average levels. At the lowest levels of anxiety, ICBT showed significantly better outcomes. No other significant moderators emerged.

Conclusions

Findings suggest that baseline variables may influence the rate of improvement in internet-delivered therapy for adolescent MDD, as well as help guide treatment selection. Replication is needed to establish the clinical utility of these variables.