White Matter Diffusion Anisotropy and the Effects of Inpatient Multidisciplinary Rehabilitation in Parkinson's Disease: A Diffusion Tensor Imaging Analysis.

Abstract

In Parkinson's disease (PD), compensatory increases in corticospinal tract (CST) fractional anisotropy (FA) occur early; whether pretreatment white matter integrity predicts rehabilitation responsiveness is unknown. To determine whether pretreatment CST FA predicts inpatient rehabilitation response in PD. Ninety-four PD patients and 13 age-matched healthy controls underwent DTI before an 8-week inpatient multidisciplinary rehabilitation program. Patients were classified as responders (UPDRS Part III improvement ≥2.5 points; n = 61) or non-responders (n = 33). FA across 20 white matter ROIs was compared (Kruskal-Wallis; Mann-Whitney U post-hoc) and voxel-wise TBSS performed. Logistic regression adjusted for age, disease duration, H-Y stage, and levodopa dose. Responders had higher bilateral CST FA than non-responders (right: P < .001, d = 0.99; left: P < .001, d = 1.14; both FDR-corrected) and maintained CST FA comparable to healthy controls, whereas non-responders showed significantly lower CST FA. TBSS confirmed bilateral CST clusters of higher FA in responders (P < .05, TFCE-corrected). CST FA correlated inversely with H-Y stage (ρ = -0.307, P = .003) and motor improvement (ρ = -0.417, P < .001). Logistic regression identified CST FA as an independent predictor (OR = 5.51 per SD; 95% CI 2.50-12.06; Nagelkerke R² = 0.408; P < .001). Pretreatment CST FA independently predicts rehabilitation responsiveness in PD, with responder integrity comparable to healthy controls. With good internal discrimination (AUC 0.79), DTI-derived CST FA is a promising marker for stratifying rehabilitation potential, though external validation is required before clinical use. UMIN Clinical Trials Registry, https://www.umin.ac.jp/, UMIN000023641.