Effectiveness of VR-based cognitive training and games on cognitive rehabilitation in patients with MCI：A Systematic Review and Meta-Analysis (Preprint)

<sec> <title>BACKGROUND</title> Mild Cognitive Impairment (MCI), characterized by cognitive decline exceeding age-related norms without significant functional impairment, represents a critical prodromal stage for dementia. Given the limited efficacy of pharmacological interventions and the escalating global burden of aging populations, effective non-pharmacological strategies are urgently needed. Virtual Reality (VR), offering immersive and interactive environments, shows promise for cognitive rehabilitation in MCI, but evidence regarding its efficacy remains inconsistent due to heterogeneity in intervention types (training vs. gaming) and technical parameters, particularly immersion level. </sec> <sec> <title>OBJECTIVE</title> This systematic review and meta-analysis synthesized evidence from randomized controlled trials (RCTs) to evaluate the efficacy of VR-based cognitive training and gaming interventions on cognitive function in older adults with MCI and to investigate the moderating role of immersion level. </sec> <sec> <title>METHODS</title> We systematically searched four electronic databases (PubMed, Web of Science, Embase, Scopus) from inception to July 20, 2025, for RCTs investigating VR interventions (cognitive training or games) in individuals aged ≥ 55 years diagnosed with MCI. Two independent reviewers performed study selection, data extraction (including intervention characteristics, implementation details, and behavior change techniques), and risk-of-bias assessment using the Cochrane Risk of Bias tool (RevMan 5.4.1). Standardized mean differences ((Hedges’s g) with 95% confidence intervals (CI) were pooled using random-effects models in Stata 18.0. Heterogeneity was quantified using I². Publication bias was assessed via funnel plots and Egger’s test. Pre-specified meta-regression explored immersion level as a potential moderator. </sec> <sec> <title>RESULTS</title> Of the 1,411 articles retrieved in total, 11 studies were included in the analysis. VR demonstrated a statistically significant improvement in the efficacy of cognitive rehabilitation among patients with MCI (Hedges’s g = 0.6, 95% CI: 0.29 to 0.90, p &lt; 0.05). Specifically, VR-based games (Hedges’s g = 0.68, 95% CI: 0.12 to 1.24, p = 0.02) showed greater advantages in improving cognitive impairments compared to VR-based cognitive training (Hedges’s g = 0.52, 95% CI: 0.15 to 0.89, p = 0.05). The immersive level of VR interventions emerged as a significant moderator of heterogeneity across the included studies. </sec> <sec> <title>CONCLUSIONS</title> VR-based interventions, including cognitive training and games, effectively improve cognitive function in MCI patients, with VR games showing a trend toward greater efficacy. Critical to these outcomes is VR immersion level, highlighting the need to optimize sensory integration (e.g., multi-modal feedback) while balancing individual tolerance. Moderate heterogeneity across studies underscores the importance of standardized protocols for intervention design and immersion assessment. These findings support VR as a valuable non-pharmacological tool for MCI cognitive rehabilitation, with future research needed to refine personalized approaches and enhance generalizability. </sec>