The effects of virtual reality training on cognitive function in older adults: a systematic review and meta-analysis

Objectives Virtual reality (VR) has emerged as a promising non-pharmacological approach for supporting cognitive health in older adults. However, the overall magnitude of its effects and the extent to which outcomes vary by cognitive status, intervention duration, and immersion level remain uncertain. Methods Randomized controlled trials (RCTs) were searched in Web of Science, PubMed, Scopus, Embase, Cochrane Library, and Wanfang Database from database inception to September 24, 2025, without language restrictions. Eligible studies included older adults receiving VR-based training compared with routine rehabilitation, conventional training, no intervention, or non-VR controls. Outcomes were classified as global cognition, TMT-based executive/attention-related performance, and functional mobility. Risk of bias was assessed using the Cochrane RoB 2 tool. Random-effects meta-analyses were conducted using standardized mean differences (SMDs) with 95% confidence intervals (CIs). Subgroup analyses were interpreted as exploratory. Results Ten RCTs involving 749 participants were included. VR-based training significantly improved global cognition (SMD = 0.49, 95% CI: 0.25–0.73, p 0.0001), with moderate heterogeneity ( I 2 = 50%). Functional mobility, analyzed as a secondary outcome, also improved (SMD = 0.39, 95% CI: 0.05–0.73, p = 0.02; I 2 = 56%). Only one trial reported TMT-based executive/attention-related outcomes using TMT-A and TMT-B. Because these outcomes were derived from the same study population, they were not interpreted as evidence from independent studies; the findings suggested a favorable direction of effect but were considered preliminary. Exploratory subgroup analyses suggested possible variation by cognitive status, intervention duration, and immersion level, but several comparisons were based on small numbers of studies. Conclusion VR-based training may improve global cognition in older adults and may also benefit functional mobility. Evidence for TMT-based executive/attention-related performance remains preliminary. Current evidence is insufficient to support firm conclusions regarding optimal intervention duration, preferred immersion level, or subgroup superiority. Larger, well-designed RCTs with standardized outcome classification and longer follow-up are needed. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420251208123 , Identifier: CRD420251208123.