Linking VR performance to cognitive ability: The significance of ACC-PCL connectivity in aging populations

Neuropsychological tests provide standardized cognitive assessment but have limited ecological validity. Virtual reality (VR) creates immersive environments, better reflecting real-world cognitive performance. Although impaired VR performance correlates with early cognitive decline, its neural mechanisms remain unclear in non-demented elders. This study investigates fMRI biomarkers linked to VR performance to identify neural predictors of cognitive decline.

24 non-demented older adults, including cognitively normal (CN) and mild cognitive impairment (MCI), completed baseline resting-state fMRI and immersive VR tasks combining physical and cognitive demands (dog walking, mountain climbing, drone protection). VR performance score (VRS) was quantified via entropy weight method. We analyzed relationships between VRS, neuropsychological scores, and resting-state functional connectivity (FC).

VRS significantly correlated with MoCA visuospatial/executive scores (p = 0.015). Whole-brain FC analysis revealed a strong association between VRS and FC between the left anterior cingulate cortex (ACC) and left paracentral lobule (PCL) in the overall sample (adjusted p = 1.94 × 10⁻⁶), present in CN but absent in MCI. Stepwise regression confirmed this FC as the significant VRS predictor in CN (R² = 0.599, p < 0.001).

Immersive VR performance reflects visuospatial/executive function and is predicted by left ACC-PCL connectivity, serving as a neuroimaging biomarker for real-world cognition that complements traditional assessments. Practically, this biomarker requires validation in larger longitudinal cohorts.