Virtual Reality Training Improves Balance for Patients With Parkinson Disease

A recent literature review found that balance training conducted via virtual reality is beneficial at improving symptoms in Parkinson disease.

Virtual reality (VR) training is increasingly being used for a variety of health care applications, such as for pain relief, in multiple sclerosis, or for ophthalmology uses.

While there is no cure for Parkinson disease (PD), levodopa and deep brain electrical stimulation can ease motor symptoms, and physical rehabilitation is used as well. VR is increasingly seen as an adjunct to physical exercise and may change neuroplasticity in the brains of patients with PD and other neurodegenerative diseases.

A recent systematic review and meta-analysis examined the use of VR to improve balance in patients with PD. The authors conducted the study because previous reviews have found inconsistent conclusions about the efficacy of VR for this population.

Databases, including Web of Science, PubMed, Scopus, China National Knowledge Infrastructure, and Wanfang, were searched to identify all relevant studies published in English or Chinese from September 15, 2010, to September 15, 2020.

Three sets of keywords were used for literature retrieval:

  • Virtual reality, VR, Kinect, Wii, Xbox
  • Parkinson, parkinsonian disorders, Parkinson, parkinsonism, Parkinson disease, PD
  • Balance, equilibrium, dynamic postural control

The studies were either randomized controlled trials (RCTs) or nonrandomized controlled trials. Studies without a comparison group or those that did not report intergroup comparisons were excluded. Subjects in the studies were adults aged 18 or older.

VR interventions were classified as nonimmersive, semi-immersive, and fully immersive. The control group consisted of a wait group, routine physical therapy, or other types of treatment such as drug therapy.

Balance ability was measured by the Berg Balance Scale (BBS), which is widely used in movement disorders; it includes 14 balance-related activities. Higher scores indicate better ability; the maximum score is 56 points, and a score below 40 indicates declining balance ability.

A total of 491 articles were retrieved, and 10 were added manually. After exclusions, 16 RCTs were analyzed (n = 583 patients with PD). The methodological quality evaluation score ranged from 5 to 8 points, with an average score of 6.56. Due to the nature of VR, it was largely difficult to implement any sort of blinding on the part of the participants, evaluators, or providers.

The standardized mean difference (SMD) was selected as the index of effect scale, and the effect size indicated the degree of impact of VR training on PD balance ability (ranging from SMD < 0.20, indicating a negligible effect, up to an SMD > 0.80 showing a large effect). A random effects model was selected to combine effect sizes and I2 evaluated the heterogeneity of the studies, where a larger I2 statistic displayed greater heterogeneity.

Results showed that VR training significantly improved balance ability in PD, compared with the control group, with a high degree of heterogeneity (SMD, 2.127; 95% CI, 1.202-3.052; P <.001; I2 = 95.1; degrees of freedom = 15).

The authors said the choice of VR platform may be the main reason for heterogeneity.

However, further regression analysis showed that no training program could predict the actual impact of VR training on PD balance ability when looking at the time spent on VR or the number of training sessions.

A subgroup analysis of each training program to look at the magnitude of the effect indicated that the best results were obtained when a VR session took place just once and lasted up to 20 minutes, 4 to 6 times per week for 3 to 5 weeks. The authors noted this dose-response finding needs further study.

Reference

Wu J, Zhang H, Chen Z, et al. Benefits of virtual reality balance training for patients with Parkinson disease: systematic review, meta-analysis, and meta-regression of a randomized controlled trial. JMIR Serious Games. Published online March 1, 2022. doi:10.2196/30882