Cognitive Behavioral Therapy in Treating Depression: A Meta-Analytic Review

Abstract

Objective: To synthesize and critically evaluate the meta-analytic evidence for the efficacy of cognitive behavioral therapy (CBT) in adults and special populations with depression, including comparisons with control conditions, other psychotherapies, pharmacotherapies, combined treatments, and digital CBT formats.

Methods: We performed an umbrella review of major meta-analyses, individual participant data (IPD) network meta-analyses, and systematic reviews published up to 2023, focusing on pooled effect sizes, long-term outcomes, format (individual, group, guided and unguided internet-based CBT), moderators (age, baseline severity, setting), risk of bias, heterogeneity, and publication bias. Representative meta-analytic effect sizes were tabulated and visualized.

Results: Across large syntheses, CBT demonstrates moderate to large effects versus control conditions (Hedges’ g ranging ~0.47 after bias adjustment to ~0.79 unadjusted) and comparable effects to pharmacotherapy in the short term but superior maintenance at 6–12 months. Internet-delivered CBT (iCBT) is effective; guided iCBT outperforms unguided iCBT for moderate to severe depressions (MD in PHQ-9 ≈ −0.8 at post-treatment in IPD-NMA), though unguided formats retain value for milder cases. Effects in special populations (perinatal depression, inpatients, adolescents) are generally positive but heterogeneous. Publication bias and trial risk of bias reduce pooled estimates in sensitivity analyses.

Conclusions: CBT is an evidence-based, effective treatment for depression across formats and populations. Future work should prioritize high-quality trials with low risk of bias, head-to-head comparisons using standardized outcomes, and research on precision allocation (which patients benefit most from which CBT format).