Baseline Regional Cholinergic Denervation Predicts Cognitive Trajectories in Moderate Parkinson Disease

ABSTRACT Objective Cognitive decline is a disabling and variable feature of Parkinson disease (PD). While cholinergic system degeneration is linked to cognitive impairments in PD, most prior research reported cross‐sectional associations. We aimed to fill this gap by investigating whether baseline regional cerebral vesicular acetylcholine transporter ligand [ 18 F]‐fluoroethoxybenzovesamicol ([ 18 F]‐FEOBV) binding predicts longitudinal cognitive changes in mild–moderate, non‐demented PD subjects. Methods Seventy‐five non‐demented, mild–moderate PD subjects underwent baseline [ 18 F]‐FEOBV PET and standardized cognitive evaluations, with repeat cognitive testing after 2 years. Participants were classified into four cognitive classes: persistent normal (no MCI at either time point; N = 41), persistent MCI (MCI at both; N = 21), MCI conversion (normal to MCI; N = 6), and MCI reversion (MCI to normal; N = 7). We performed whole‐brain voxel comparisons (controls and between classes) and used linear and multinomial regression to predict follow‐up cognitive status. Results Whole‐brain voxel analyses revealed class‐specific [ 18 F]‐FEOBV binding deficits. Persistent MCI showed the most widespread reductions, extending into frontal regions. Compared with MCI reverter and persistently normal groups, persistent MCI had lower binding in occipito‐temporal and temporo‐parieto‐frontal regions, respectively, while MCI reverters showed higher occipital binding than converters. Binding in occipital, temporo‐parietal, and medial frontal regions predicted follow‐up cognitive decline, with medial frontal binding showing the strongest association (esp. global cognition, memory, and visuospatial domains). Interpretation Baseline cholinergic system changes associate with varying cognitive trajectories in mild–moderate PD. Assessment of regional cholinergic deficits may be useful for prediction of cognitive trajectories in this population, enhancing subject selection and stratification for intervention trials aimed at improving cognition in PD.