Comparative effectiveness of non-pharmacological interventions for social anxiety disorder in adults: a systematic review and meta-analysis.

Background: Social anxiety disorder (SAD) is a prevalent and disabling mental health condition in adults, associated with marked functional impairment, reduced quality of life, and substantial personal and societal burden. Previous meta-analyses have mainly evaluated single non-pharmacological interventions for SAD, but few evidence syntheses have compared the relative effectiveness of multiple non-pharmacological approaches within a unified analytic framework. As a result, the comparative efficacy of these interventions remains unclear. Methods: We searched the Cochrane Library, EBSCO, PubMed, Web of Science, and Embase from inception to 3 November 2025 for randomized controlled trials (RCTs) of non-pharmacological interventions for adults with SAD or clinically relevant social anxiety symptoms. Risk of bias was assessed using the Cochrane RoB 2 tool, and certainty of evidence was evaluated using GRADE. A frequentist random-effects network meta-analysis was conducted, and results were expressed as standardized mean differences (SMDs) with 95% confidence intervals (CIs). Results: A total of 104 RCTs involving 10,708 participants were included. Compared with control conditions, most non-pharmacological interventions were associated with significant reductions in social anxiety severity. Based on SUCRA values, cognitive behavioral therapy (CBT; 86.4%; SMD = -0.80, 95% CI: -0.92 to -0.68), combination therapy (CT; 71.7%; SMD = -0.74, 95% CI: -1.00 to -0.47), and psychotherapy (PT; 70.7%; SMD = -0.73, 95% CI: -0.92 to -0.54) ranked among the more effective interventions. Exploratory subgroup analyses suggested that CBT remained among the higher-ranked interventions across the country's development level, intervention duration, baseline severity, and delivery format. However, the certainty of evidence was predominantly low to moderate, and subgroup findings should be interpreted cautiously. Conclusion: This network meta-analysis integrates direct and indirect evidence on non-pharmacological interventions for adult social anxiety. CBT, CT, and PT appear to be among the more effective approaches, although confidence in comparative estimates remains limited. These findings may inform shared decision-making and implementation planning, underscoring the need for well-designed head-to-head trials and more consistent reporting to strengthen comparative conclusions. Systematic review registration: PROSPERO (CRD420251179037).