The effects of virtual reality on anxiety and depression in cancer patients: a systematic review with meta-analysis

Anxiety and depression are prevalent yet undertreated in cancer patients, impairing quality of life, treatment adherence, and outcomes. Virtual reality, with its immersive and interactive features, offers a promising non-pharmacological intervention. However, a rigorous synthesis of randomized evidence is needed to establish its efficacy and identify key moderators of effectiveness, guiding targeted clinical implementation. A comprehensive search was conducted across seven electronic databases (PubMed, Web of Science, Cochrane Library, Embase, CNKI, WanFang, VIP) from inception to December 10, 2025. Eligible studies were RCTs involving cancer patients, comparing a VR intervention targeting emotional distress to a non-VR control. Study selection, data extraction, and risk-of-bias assessment using the Cochrane RoB 2 tool were performed independently by reviewers. Primary outcomes were anxiety and depression. Data were synthesized using random-effects meta-analysis in R (meta package). Pooled effects were estimated as Standardized Mean Differences with 95% Confidence Intervals. Heterogeneity was assessed via I² and τ² statistics, and pre-specified subgroup analyses were conducted for the proposed moderators. Sensitivity analyses and assessment for small-study effects were performed. Eighteen RCTs (total participants: N = 1303; VR group: n = 658, Control: n = 645) met the inclusion criteria. The meta-analysis demonstrated statistically significant beneficial effects of VR interventions on anxiety (17 RCTs, n = 1287; SMD = -1.16, 95% CI: -1.60 to -0.71; P < 0.001) and depression (7 RCTs, n = 543; SMD = -1.28, 95% CI: -2.52 to -0.04; P = 0.04), with both outcomes favoring the VR groups. Substantial heterogeneity was observed. Risk of bias assessment indicated that most studies had some concerns regarding bias, with three studies rated low risk, twelve with some concerns, and three high risk. Subgroup analyses suggested variations in effect sizes across potential effect modifiers, though findings were exploratory. This review suggests that VR interventions may alleviate anxiety and depression in cancer patients, with potentially clinically meaningful effects. The findings advance the field by shifting focus from overall efficacy toward contextual factors influencing outcomes, providing insights for personalized application. However, moderate-certainty evidence and substantial heterogeneity underscore the need for future large-scale, rigorous RCTs that prioritize standardized reporting, mechanistic exploration, and optimize future clinical implementation. PROSPERO (Registration No. CRD420251089141).