Electroencephalography Functional Network Responses to Immersive Virtual Reality in Alzheimer Disease and Mild Cognitive Impairment: Exploratory Single-Session 3-Group Repeated-Measures Study.

BACKGROUND: Alzheimer disease (AD) is increasingly conceptualized as a disorder of large-scale brain network disruption rather than isolated regional dysfunction. Immersive virtual reality (VR) may provide a controlled sensory challenge for neurophysiological assessment, but its association with whole-brain electroencephalography (EEG) functional network organization across AD and mild cognitive impairment (MCI) remains insufficiently characterized. OBJECTIVE: This study examined whether immersive VR exposure was associated with state-dependent changes in EEG-derived functional network organization among older adults with AD, MCI, and normal cognition (NC). METHODS: This exploratory single-session, 3-group repeated-measures EEG study included 60 older adults: 20 with AD, 20 with MCI, and 20 participants with NC. During a single visit, participants underwent repeated within-session EEG recordings during pre-VR resting, immersive VR exposure, and post-VR resting states. VR was delivered through a head-mounted display presenting a passive first-person roller-coaster environment. Functional connectivity was quantified using phase-locking value (PLV) across multiple frequency bands, and global efficiency (GE), modularity, and nodal strength were calculated to characterize network topology. Between-group differences within each state were assessed using 1-way ANOVA with post hoc pairwise comparisons, whereas within-group state changes were evaluated using paired-sample 2-tailed t tests. Benjamini-Hochberg false discovery rate (FDR) correction was applied to inferential graph-theoretical analyses. PLV matrices, nodal strength maps, and connectivity graphs were interpreted descriptively. RESULTS: In the broadband range, GE was higher in the AD group than in the NC group during the pre-VR resting state (raw P=.009; FDR-adjusted P=.04), whereas the AD-NC difference was not statistically significant during VR exposure or the post-VR resting state. Baseline modularity was also higher in AD than in NC and remained significant after FDR correction; no significant between-group modularity differences were detected during VR exposure. Within-group analyses showed an FDR-corrected increase in GE from pre-VR to VR in NC participants and from pre-VR to post-VR in participants with MCI. No within-group GE change in AD remained significant after FDR correction. Descriptive PLV, nodal strength, and connection maps suggested state-associated changes in hemispheric distribution and high-frequency connectivity patterns. However, high-frequency GE findings did not remain significant after FDR correction and were therefore interpreted as exploratory. CONCLUSIONS: In this exploratory single-session, 3-group repeated-measures EEG study, immersive VR exposure was associated with short-term, state-dependent changes in PLV-derived EEG functional network organization among older adults with AD, MCI, and NC. Graph-theoretical analysis demonstrated the feasibility of using repeated within-session VR-EEG measurements to characterize acute network responsiveness across cognitive groups. However, descriptive connectivity maps were not treated as confirmatory evidence, and the absence of a non-VR comparison condition and concurrent cognitive or clinical outcomes precludes conclusions regarding causality or therapeutic efficacy. Larger controlled and longitudinal studies are needed to establish the reproducibility and clinical relevance of these findings.