The role of virtual reality in balance training for Parkinson's disease: a systematic review and meta-analysis.

Only six trials and 214 people were included, and everyone was at Hoehn and Yahr stage II or III and well enough to take part in exercise training, so this tells you nothing about people with more advanced disease or those who cannot walk. The trials disagreed so much with each other that 93% of the variability between them was due to real differences in the studies rather than chance, which makes a single pooled number hard to trust. The VR devices, programmes and doses varied widely, from 10 to 60 sessions. The authors describe the blinding in the six trials as inconclusive, so some of the measured improvement may reflect expectation rather than the treatment. The review only accepted trials that used the Berg Balance Scale, so it says nothing about other balance, gait or falls measures, and the authors note the scale is limited by a ceiling effect and by scoring that depends on the tester's judgment. The VR programmes were predominantly non-immersive, so these results do not tell you about immersive headset-based VR. Average follow-up was 3.5 months and only one trial followed participants for 12 months, so longer-term effects are unknown. The authors suggest the training duration may simply have been too short.

The authors declare that no financial support was received for this work or its publication. No other conflicts of interest are reported in the text provided.

Do not read the pooled standardised mean difference of 1.89, or the VR group's average gain of 2.85 points on the Berg Balance Scale, as evidence that virtual reality improves balance. The difference between the groups was not statistically significant (p = 0.269), the pooled confidence interval ran from -0.62 to 4.4 and crossed zero, and the overall test found no effect. Both groups' gains were below the 4 to 7 points the scale needs before a change can be detected, the six trials disagreed strongly with each other, and the authors' suggestion that VR adds something on top of conventional physiotherapy was not tested here.