Efficacy of three virtual reality interventions for social anxiety disorder: A meta-analysis

OBJECTIVE: To evaluate the efficacy of virtual reality exposure therapy (VRET), virtual reality cognitive behavioral therapy (VR-CBT), and virtual reality exposure combined with biofeedback (VRE + Biofeedback) for social anxiety disorder (SAD) through meta-analysis, and to explore variations according to outcome measures, hardware type, population, region, and control conditions. METHOD: Randomized controlled trials (RCTs) of virtual reality interventions for SAD were retrieved from Web of Science, PubMed, Scopus, ScienceDirect, ProQuest, CINAHL, SpringerLink, PsycINFO, and Embase. A total of 17 studies were included, and meta-analysis was conducted using Comprehensive Meta-Analysis V3. RESULTS: VR interventions were effective for SAD (SMD = 0.477, p < 0.001). VRET had a moderate effect (SMD = 0.521, p < 0.001), VR-CBT had a substantial effect (SMD = 0.618, p = 0.042), whereas VRE + Biofeedback did not show a significant pooled effect (SMD = 0.110, p > 0.05). Subgroup analyses indicated that VRET had comparable effects on global and dimension-specific social anxiety (SMD = 0.528 and 0.527, respectively), with significant effects across Eastern (SMD = 0.606) and Western regions (SMD = 0.441), immersive-device settings (SMD = 0.585), and non-student populations (SMD = 0.600). VR-CBT showed significant effects for global social anxiety measures, immersive-device settings, non-student populations, and Western regions (SMD = 0.618); however, VR-CBT has not been tested in the East. CONCLUSION: VR interventions alleviate SAD. VRET and VR-CBT performed well in immersive settings and non-student populations, with VRET showing comparable global and dimension-specific effects across regions, while VR-CBT affected Western regions and global measures. VRE + Biofeedback showed no significant effect.