Comparative effectiveness and exploratory dose-response of exercise for adolescent depression: a network meta-analysis of randomized controlled trials

Background: Adolescent depression is a global mental health concern, with significant functional impairment and long-term consequences. Exercise shows therapeutic potential, but the comparative efficacy of different modalities and parameters remains unclear.

Objectives: To evaluate the comparative effectiveness of exercise modalities for adolescent depression and examine potential moderators and dose–response patterns.

Method: This systematic review and Bayesian network meta-analysis were conducted across seven databases. Eligible studies were RCTs that examined the effects of exercise interventions on depression in adolescents with depression. The primary analysis employed a Bayesian network meta-analysis to estimate the relative efficacy of different exercise modalities compared to controls. Certainty of evidence was assessed using the CINeMA framework (PROSPERO: CRD420251063361).

Results: Twenty-two RCTs (49 arms; 1327 participants) were included. Compared with active control, aerobic plus resistance training showed the largest effects (Hedges' g = −0.71, 95 % CrI: −0.98 to −0.45), followed by mindbody interventions (g = −0.64, 95 % CrI: −0.99 to −0.29), mixed aerobic exercise (g = −0.56, 95 % CrI: −0.87 to −0.24), and walking/jogging (g = −0.54, 95 % CrI: −1.00 to −0.09), with the overall quality was predominantly low to very low. SUCRA supported this ordering, and global inconsistency was not significant. Effects were significantly moderated by gender, age and comorbidity with no significant dose–response relationships.

Conclusions: Aerobic plus resistance training is likely the optimal exercise prescriptions for adolescent depression, and its optimization may require addressing adolescent heterogeneity rather than merely increasing the dose. These findings warrant cautious interpretation given the low certainty of evidence.