MRI-derived choroid plexus volume and longitudinal cognitive decline in Parkinson's disease

Magnetic resonance imaging (MRI)-derived choroid plexus (CP) volume has been associated with cognition in Parkinson's disease, but its close relationship with ventricular anatomy complicates interpretation.

We analyzed Parkinson's Progression Markers Initiative (PPMI) data. Baseline FreeSurfer-derived CP and structural volumes were examined using linear models, and post-baseline Montreal Cognitive Assessment (MoCA) trajectories were assessed using mixed-effects models with participant-specific correlated random intercepts and time slopes. Models accounted for demographic and disease variables together with ventricular and global brain anatomy, with time interactions for age, lateral ventricular volume and global brain volume.

Among 1557 baseline participants with Parkinson's disease, 779 had CP measurements and 778 entered the structural complete-case analyses. CP volume was strongly associated with lateral ventricle volume (standardized beta = 0.472, 95% confidence interval [CI] 0.415 to 0.530; P = 1.49 × 10^-50) and showed smaller, region-specific subcortical associations. The longitudinal analysis included 4196 post-baseline MoCA observations from 738 participants. Each 1-standard deviation higher baseline CP volume was associated with 0.076 points faster annual MoCA decline (time-by-CP beta = -0.076, 95% CI -0.128 to -0.024; P = 0.0048). Effect magnitude varied with model specification, and the five-year-restricted estimate was imprecise (beta = -0.052, 95% CI -0.136 to 0.032).

Higher MRI-derived CP volume was associated with faster longitudinal cognitive decline in Parkinson's disease. Its strong relationship with ventricular anatomy may partly reflect shared anatomical or measurement-related variation; tissue specificity and clinical utility remain unestablished.