Effects of Varied Virtual Reality (VR) Treatment Parameters on Balance Function Rehabilitation in Stroke Patients: A Meta-Analytic Review

Objective: To evaluate the efficacy of virtual reality (VR) interventions on the balance function in post-stroke patients based on a comprehensive analysis of an array of parameters. Methods: Systematic searches were conducted across different databases, including PubMed, The Cochrane Library, EMBASE, Web of Science, CNKI, and Wanfang, from their inception time to March 15, 2025. The resulting literature and extracted information was screened by two independent researchers, and they also evaluated the risk of bias of the included studies using the Cochrane Collaboration’s Risk of Bias Assessment Tool for Randomised Controlled Trials (Version 5.1.0). Meta-analyses were performed using RevMan 5.4.1 software. The quality of evidence for each outcome metric was evaluated using the GRADE approach. Assessment of publication bias was carried out based on funnel plots developed using Stata 15.1, in combination with Begg test results. Results: Nineteen studies involving a total of 703 patients were included. Meta-analysis showed that VR interventions significantly improved static balance most (standardize mean difference, SMD = 0.80; 95% confidence interval, CI: 0.33-1.28, P < 0.01). In addition, VR training also effectively improved balance as shown by the Berg Scale scores (mean difference, MD = 2.62, 95% CI: 1.92-3.33, P < 0.01), Timed Up and Go (TUG) test scores (MD = 1.93, 95%CI: 0.14-3.72, P < 0.05), and Tinetti scale scores (MD = -2.26, 95%CI: -3.82-0.69, P < 0.01). Yet, VR intervention showed no effect on Functional Reach Test (FRT) (MD = 1.45, 95%CI: -2.75-5.65, P = 0.50 > 0.05). In terms of intervention modality, VR combined with other therapies demonstrated greater improvement (SMD = 0.66, 95%CI: 0.37-0.95, P < 0.01). In terms of treatment duration, session lasting no more than 30 minutes (SMD = 0.70, 95%CI: 0.49-0.91, P < 0.01) and an intervention of no more than 6 weeks (SMD = 0.68, 95%CI: 0.42-0.94, P < 0.01) were associated with better outcomes. Conclusion VR-based rehabilitation effectively improves balance function in post-stroke patients. VR combined training sessions lasting for less than 30 minutes and an intervention duration of no more than 6 weeks are recommended to maximize intervention outcomes.