Effects of virtual reality-based exercise on biomechanical and functional outcomes: a narrative review of randomized controlled trials

This study was conducted as a structured narrative review of randomized controlled trials (RCTs) investigating virtual reality (VR)-based exercise and rehabilitation. PubMed and the Research Information Sharing Service were searched for peer-reviewed studies published between 2015 and 2025. Predefined search terms and eligibility criteria based on the population, intervention, comparison, and outcome framework were applied to identify relevant studies for narrative synthesis. Search terms were organized into three conceptual domains: VR, exercise or rehabilitation, and biomechanical outcomes. Of 2,138 records initially identified, 19 RCTs were included in the final analysis. Whole-body movement was the most common intervention (11 studies), followed by center-of-mass shifting exercise (three studies). Gait-related outcomes were investigated in 13 of the 19 RCTs, balance and postural control in 12 studies, and direct or instrumented biomechanical outcomes in 10 studies. Across the trials, VR-based exercise was generally associated with improvements in gait, balance, postural control, and selected neuromuscular outcomes; however, the direction, magnitude, and statistical significance of these effects varied across studies. Because substantial heterogeneity was observed in study populations, intervention duration and dose, VR modality, and outcome assessment, pooled effect sizes were not estimated, and definitive conclusions regarding the magnitude of treatment effects could not be drawn.