Abstract
Virtual reality (VR) has emerged as an innovative platform for delivering cognitive and physical training to individuals with cognitive impairment. However, the differential effectiveness of fully immersive versus partially immersive VR interventions remains unclear. This network meta-analysis aimed to evaluate how immersion level influences cognitive, motor, and functional outcomes in neurodegenerative populations. A systematic search of PubMed, Embase, Cochrane Library, and Web of Science up to October 2025 identified 20 randomized controlled trials involving 1382 participants with mild cognitive impairment (MCI) or dementia. Interventions were categorized into four groups: (1) fully immersive VR (head-mounted displays), (2) partially immersive VR (screen-based or motion-capture systems), (3) active control (traditional cognitive or physical training), and (4) passive control (usual care or health education). Outcomes included the Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Trail Making Test (TMT), Digit Span Test (DST), Timed Up and Go (TUG), and Instrumental Activities of Daily Living (IADL). Standardized mean differences (SMDs) and surface under the cumulative ranking curve (SUCRA) values were calculated using RevMan 5.4 and Stata 18.0. Fully immersive VR significantly improved global cognition compared to passive control (MMSE: SMD = 0.51, 95% CI [0.06, 0.96]), while partially immersive VR showed superior effects on executive function versus active control (TMT-B: SMD = -1.29, 95% CI [-2.62, -0.93]) and on motor function (TUG: SMD = -0.59, 95% CI [-1.11, -0.08]). In MoCA performance, both VR modalities outperformed traditional interventions (SUCRA: fully immersive = 76.0%; partially immersive = 84.8%). SUCRA rankings suggest that fully immersive VR is optimal for memory and foundational cognition (81.7%), whereas partially immersive VR performs best for executive function (98.9%). These findings indicate that the efficacy of VR-based cognitive or physical-cognitive interventions is modulated by immersion level. Tailoring VR modality to specific cognitive domains may optimize rehabilitation outcomes in MCI and dementia care.