Network-level and motor consequences of longitudinal perivascular diffusivity alterations indexed by DTI-ALPS in Parkinson’s disease
摘要
Parkinson’s disease (PD) is increasingly recognized as a large-scale network disorder extending beyond nigrostriatal degeneration. We investigated longitudinal changes in ALPS-derived perivascular diffusivity and their associations with structural connectivity and clinical outcomes in PD. Multimodal MRI data from PD participants and healthy controls drawn from PPMI, HABS, and ADNI datasets were analyzed longitudinally using mixed-effects models, network-based regression with family-wise error correction, and latent growth curve modeling. The ALPS-derived index declined significantly in PD but remained stable in controls, with a significant group-by-time interaction (β = −0.032, p < 0.001), indicating a disease-specific trajectory independent of normal aging. The ALPS index was consistently lower in males and associated with greater motor severity and REM sleep behavior disorder burden. Structural connectome analysis revealed progressive reductions within default mode, somatomotor, and subcortical networks alongside increases in cerebellar–sensorimotor and interhemispheric pathways. Lower ALPS values were associated with widespread connectivity alterations spanning subcortical, cerebellar, and higher-order association networks. These findings support DTI-ALPS as a sensitive marker of disease-related biological change and highlight the value of integrating perivascular diffusivity metrics with connectomic analyses to characterize PD progression.