Abstract
Cognitive decline in older adults is a growing global concern, with over 55 million people living with dementia worldwide. Exercise games, which combine physical exercise with cognitive training in an interactive digital environment, represent a promising approach for mitigating this decline in community settings. Here we present a systematic review and meta-analysis examining the cognitive effects of exercise games on community-dwelling older adults, including those with mild cognitive impairment and dementia. Across 10 randomized controlled trials with 690 participants (mean age 74 years), exercise games produced a small but statistically significant overall cognitive benefit compared to control conditions (Hedges' g = 0.135, 95% CI [0.037, 0.229], p = 0.012). Domain-specific analyses revealed memory showing a trend toward improvement (g = 0.179, p = 0.060), while global cognition, executive function, and attention did not reach statistical significance. A single study examining processing speed showed a positive effect, though this finding requires replication. Multicomponent exercise programs incorporating aerobic elements demonstrated numerically larger, though non-significant, effects (g = 0.291, p = 0.138) compared to aerobic-only interventions (g = 0.119, p = 0.142). The evidence base for clinical populations remains limited and inconclusive, with only two studies examining MCI (g = 0.256, p = 0.168) and three studies examining dementia (g = 0.005, p = 0.975), neither showing statistically significant effects. These findings provide preliminary evidence supporting exercise games as a modest, accessible supplementary intervention for cognitive health in community settings, though the small effect sizes and limited evidence in clinical populations indicate that substantially more research with larger samples and standardized protocols is needed to establish efficacy and optimize implementation. SYSTEMATIC REVIEW REGISTRATION: https://doi.org/10.17605/OSF.IO/AM8QT.