Virtual reality interventions and their impact on physical function and quality of life in multiple sclerosis: A systematic review with meta-analysis.

BACKGROUND: Impairments in gait, balance and upper limb function are among the most disabling consequences of multiple sclerosis (MS), and virtual reality (VR) has been proposed as a means of delivering repetitive, task-oriented and feedback-driven rehabilitation. Previous syntheses have been confined to single devices, to balance outcomes, or to mixed neurological populations, and none has examined physical function and health-related quality of life jointly across the full range of VR modalities used in MS. METHODS: We searched seven electronic databases from inception to July 30, 2026, for randomized controlled trials of structured VR-based rehabilitation in adults with MS (PROSPERO CRD420261332329). Two reviewers independently screened records, extracted data, appraised risk of bias with RoB 2 and rated certainty of evidence with GRADE. Standardized mean differences (Hedges' g) were pooled using random-effects models, and heterogeneity was quantified with the I² statistic. RESULTS: Nineteen trials were included, comprising 781 randomized participants, of whom 702 (89.9%) contributed outcome data. VR was associated with improvements in gait speed (10MWT: g = 0.99, 95% CI 0.39 to 1.58; I² = 51.8%), functional mobility (TUG: g = 1.00, 95% CI 0.47 to 1.54; I² = 52.4%), balance (BBS: g = 0.59, 95% CI 0.14 to 1.03; I² = 57.7%) and health-related quality of life (MSQOL-54: g = 0.87, 95% CI 0.01 to 1.73; I² = 86.0%; MusiQoL: g = 1.19, 95% CI 0.60 to 1.78; I² = 57.5%). Upper limb outcomes, assessed in eight trials, were synthesized narratively because of differences in instruments and reporting. Certainty of evidence was low for functional mobility and balance, and very low for gait speed, MSQOL-54, and MusiQoL. CONCLUSIONS: VR-based rehabilitation may improve gait speed, functional mobility, balance and quality of life in adults with MS. Given the low to very low certainty of the evidence and the moderate to considerable heterogeneity across trials, these estimates should be regarded as preliminary and interpreted by outcome rather than as a general effect of VR. Adequately powered trials with clearly defined VR modalities, standardized dosing, active dose-matched comparators and longer follow-up are required.