Efficacy of different virtual reality-based interventions on activities of daily living, depression and anxiety in stroke survivors: a systematic review and network meta-analysis

Virtual reality (VR)-based interventions have been recognized as promising nonpharmacological therapies for improving the ability to perform daily activities, as well as for reducing depression and anxiety, among stroke survivors. However, the comparative effectiveness of different types of VR interventions remains unclear. Therefore, this study aimed to compare and rank the efficacy of various VR interventions for improving the ability to perform daily activities, as well as for treating depression and anxiety, among stroke survivors. A frequentist network meta-analysis was conducted to compare and rank the relative effects of various VR interventions on improving the ability to perform daily activities, as well as for treating depression and anxiety, among stroke survivors. We systematically searched six English-language and three Chinese-language databases from their inception to July 2026. Randomized controlled trials (RCTs) investigating the effectiveness of VR interventions in stroke survivors were included. A random-effects model was used for pairwise meta-analyses to directly assess the effects of individual VR interventions. A consistency model was employed for the network meta-analysis to evaluate the relative effects and probability of being the best model for the different VR interventions. This analysis included 21 RCTs involving four different non‑pharmacological interventions. Preliminary evidence suggests that virtual reality-based comprehensive cognitive games (VR-CCG) may be an effective intervention for improving activities of daily living (ADLs) in stroke survivors. Additionally, VR-based team therapy (VR-TM) showed potential advantages in alleviating depressive symptoms, whereas none of the VR interventions demonstrated statistically significant improvements in anxiety outcomes. However, this evidence base remains considerably weak—the finding for VR-TM on depression was supported by only one study, and that for anxiety outcomes was based on only five studies, with both confidence and prediction intervals being notably wide and the overall certainty of evidence rated as very low. Therefore, these results should be regarded as exploratory findings with hypothesis‑generating value. Meta‑regression analyses revealed no significant associations between the predefined covariates (including age, country, session duration, intervention frequency, intervention period, and level of VR immersion) and any of the outcome measures. Nevertheless, this negative finding may be constrained by the limited number of included studies and the narrow range of covariate distributions. This study is the first network meta-analysis to simultaneously evaluate and rank the comparative efficacy of different VR interventions on ADLs, depression, and anxiety in stroke patients. Preliminary evidence suggests that VR-based interventions may improve ADLs and alleviate depressive symptoms in post-stroke patients; however, definitive evidence supporting their effectiveness against anxiety remains unavailable. Nevertheless, given the limitations of this study, including the small number of included studies and sparse network structures, these findings should be interpreted with caution. Future well‑powered comparative trials are warranted to validate the differential efficacy among various VR interventions and to determine the optimal intervention parameters (i.e., dosage, frequency, and duration), thereby providing more robust and direct evidence for their comparative effectiveness.