Purpose: A meta-analysis comparing the efficacy and acceptability of selective serotonin reuptake inhibitors (SSRIs) versus tricyclic antidepressants (TCAs) in depressed children, adolescents, and young adults was performed.

Methods: A comprehensive literature search of the PubMed, Cochrane, Embase, Web of Science, and PsycINFO databases was conducted from 1970 to December 2013. Only clinical trials that randomly assigned one SSRI or TCA to patients aged 7 to 25 years who met the diagnostic criteria for unipolar depressive disorder were included. Primary efficacy was determined by the pooling of standardized mean differences (SMDs) calculated from the difference in the reduction in mean depression rating scale scores for the 2 antidepressants. Acceptability was determined by pooling the risk ratios (RRs) of dropouts for all reasons and for adverse effects as well as the suicide-risk outcome.

Findings: Five trials with a total of 422 patients were considered to be eligible for inclusion. SSRIs were significantly more effective than TCAs in primary efficacy (SMD = -0.52; 95% CI, -0.81 to -0.24; P = 0.0003). Patients taking SSRIs had a significantly greater response to depressive symptoms than patients taking TCAs (RR = 1.55; 95% CI, 1.04 to 2.29; P = 0.03). On an individual SSRI basis, fluoxetine had a significantly greater efficacy than TCAs (SMD = -0.82; 95% CI, -1.34 to -0.29; P = 0.003). On an individual TCA basis, only imipramine was not significantly worse than SSRIs (SMD = -0.27; 95% CI, -0.56 to 0.02; P = 0.06). Significantly more patients taking TCAs discontinued treatment than patients taking SSRIs (35.8% vs 25.1%; RR = 0.70; 95% CI, 0.52 to 0.93; P = 0.02).

Implications: SSRI therapy has a superior efficacy and is better tolerated compared with TCA therapy in young patients.