Subcortical Brain‐Age Gaps Reveal Asymmetric Aging Patterns in Parkinson's Disease With Cognitive Impairment

ABSTRACT Aim The study utilized MRI‐derived brain structure age (BSA) to compare global and regional subcortical BSA among healthy controls (HCs), Parkinson's disease (PD) patients with normal cognition (PD‐NC), and mild cognitive impairment (PD‐MCI), identifying regions with accelerated aging and linking altered BSA to native volumes. Methods We analyzed structural MRI data from 55 participants (22 HCs, 18 PD‐NC, 15 PD‐MCI) using the volBrain platform to estimate global and regional subcortical BSA. Group differences in age, global, and regional BSA were tested via Kruskal‐Wallis. Follow‐up analyses included Pearson correlations for significant regions and ANOVAs where assumptions were met. Results No significant group differences were found for chronological age ( p = 0.111) or global BSA ( p = 0.143). However, at the regional level, non‐parametric analyses revealed significant group differences in the predicted age of the left amygdala (H = 6.42, p = 0.040) and the left basal forebrain (H = 6.01, p < 0.05), though effect sizes were small (ε 2 ≤ 0.07). The predicted ages of these two regions were highly collinear ( r = 0.992). Subsequent parametric tests and Bonferroni‐corrected pairwise comparisons on other subcortical regions did not yield any significant differences. Conclusion Accelerated aging appears to be a localized and asymmetric process confined to the limbic‐cholinergic network, specifically involving the left amygdala and basal forebrain. Accelerated brain aging in PD is not global but a localized, asymmetric process in the left limbic‐cholinergic network. Regional brain‐age metrics offer a sensitive biomarker for detecting the specific neurodegeneration linked to cognitive decline.