Specific Virtual Reality (VR) systems designed for rehabilitation and non-specific VR systems intended for entertainment are used in Parkinson's disease (PD) rehabilitation, but their effects are unclear. The extent to which these systems integrate neurorehabilitation principles for PD rehabilitation is unknown. Previous meta-analyses exist but data are lacking on the impact of VR on mobility.

Primary aim: to perform a systematic review with meta-analysis comparing the effects of VR and conventional therapy (CT) on balance and mobility in people with PD. Secondary aim: to perform subgroup analyses on VR type, disease severity, and treatment duration.

Randomized controlled trials comparing VR and CT effects on balance and mobility in adults with PD were selected from Pubmed and EMBASE until September 2024. Data were synthesized qualitatively and quantitatively using a standardized mean difference (SMD) with random-effects model. Subgroup analyses (VR type, disease severity, and treatment duration) and analysis of fulfilled neurorehabilitation principles were conducted. Risk of bias was assessed (PEDro checklist and Cochrane RoB-2).

Twenty-eight studies (12 countries) were included: 1151 participants, mean Hoehn & Yahr stage between 1.4 and 3.4, mean (SD) treatment duration 18.8 (11.5) hours in the qualitative analysis and 23 in the quantitative analysis. Overall risk of bias was high (10 studies), some concerns (13 studies), or low (5 studies). VR was more effective than CT for balance (630 participants; 11 studies; SMD 0.42; 95% CI, 0.19–0.65; P < 0.001) and as effective as CT for mobility (591 participants; 10 studies; SMD 0.18; 95% CI, -0.03 to 0.40; P = 0.09). Balance and mobility outcomes did not differ between specific and non-specific VR. Subgroup analyses found no significant differences.

VR improved balance in people with PD more than CT (low-certainty evidence). VR improved mobility similarly to CT (moderate-certainty evidence). VR games should integrate neurorehabilitation principles.