Effectiveness of non-immersive virtual reality exercises for balance and gait improvement in older adults: A meta-analysis.

Virtual reality (VR)-based physical exercise is an innovative and effective intervention strategy for healthcare in older adults. This meta-analysis aimed to clarify the effects of VR-based balance exercise programs on various balancing abilities of older adults. In addition, the effect size of each variable was computed by total exercise time, sensor type, avatar presence, and feedback type to determine influencing factors that lead to the success of VR-based rehabilitation programs. The databases searched were PubMed/Medline, CINAHL, NDSL, and Google Scholar. Inclusion criteria were: (1) independent older adults; (2) non-immersive VR exercise; (3) randomized controlled design; (4) both balance and gait data; and (5) written in English and Korean. The studies without information to compute effect sizes were excluded. Standardized mean difference was used to analyze the effect size (d). Twenty-five studies were finally included in this study. The main findings of this meta-analysis were as follows: (1) Non-immersive VR-based balance exercises are moderately and largely effective for improving overall balance function, (2) VR balance exercise was more effective for static balance than for gait, (3) VR exercise is more effective when avatars are presented and KP is provided as feedback. Total exercise time and mode of feedback are influencing factors that affect the effectiveness of VR-based balance exercises.