Differential predictors of motor and cognitive progression in Parkinson's disease: a 2-year prospective multimodal...

The clinical and biochemical heterogeneity of Parkinson's disease (PD) complicates accurate progression prediction. To address this, we developed a multimodal biomarker panel incorporating transcranial sonography (TCS), oxidative stress/iron indices, and serum neurofilament light chain (NfL) to predict motor and cognitive decline rates in PD patients.

In this prospective cohort, 130 PD patients and 50 healthy controls were enrolled from Xianning Central Hospital between 2023 and 2025. All participants underwent transcranial sonography (TCS) at baseline to measure the area of substantia nigra hyperechogenicity (SN+ Area) and the width of the third ventricle (V3 Width). Oxidative stress markers (malondialdehyde, reduced glutathione), iron metabolism indices (ferritin, transferrin saturation), and serum NfL were assessed. PD patients were assessed at baseline and 2-year follow-up using MDS-UPDRS Part III and Montreal Cognitive Assessment (MoCA). Multivariable logistic regression identified baseline predictors of 2-year motor and cognitive decline.

Baseline biomarkers differed significantly between PD and controls. Over 2 years, 22.3% of patients exhibited motor progression and 31.5% cognitive progression. The motor prediction model (AUC 0.86) identified baseline MDS-UPDRS III, serum NfL (OR 1.20, p = 0.017), and plasma MDA (OR 3.31, p = 0.001) as independent predictors. In contrast, the cognitive prediction model (AUC 0.88) identified older age, lower baseline MoCA, and larger V3 width (OR 2.38, p = 0.045) as independent predictors, with no significant contribution from fluid markers.

Motor worsening in mild-to-moderate PD is driven by neuroaxonal injury (NfL) and oxidative stress (MDA), whereas cognitive decline is encoded in structural atrophy (V3) and baseline function.