Abstract
Hallucinations negatively impact quality of life in Parkinson's disease, yet their neural mechanisms remain poorly understood, particularly in early disease stages. This study aimed to identify functional connectivity differences associated with visual hallucinations in early Parkinson's disease and to validate these findings across independent datasets. Resting-state functional MRI data from two independent studies were used: the 'Parkinson's Progression Markers Initiative' dataset was used as a discovery cohort (N = 25 hallucinators, N = 56 non-hallucinators) and the 'Incidence of Cognitive Impairments in Cohorts with Longitudinal Evaluation' dataset as replication (N = 49 hallucinators, N = 55 non-hallucinators overall). Group differences in functional connectivity were assessed within predefined cytoarchitectonic cortical classes and functional networks, followed by whole-brain analysis using Network-Based Statistics. This method identified a subnetwork of reduced functional connectivity in hallucinators, connecting regions involved in the default mode, somatomotor and attentional networks. Associations with clinical measures-including hallucination severity, motor symptoms, cognition and attention-were evaluated. Reduced functional connectivity in hallucinators was significantly associated with baseline and future motor symptoms, cognition and attention in the main cohort and with hallucination severity in the independent cohort. The identified functional subnetwork offers a potential direction for future research on Parkinson's disease psychosis.