α-Synuclein Oligomers in Skin Biopsies Predict the Worsening of Cognitive Functions in Parkinson’s Disease: A Single-Center Longitudinal Cohort Study

α-synuclein oligomers within synaptic terminals of autonomic fibers of the skin reliably discriminate Parkinson's disease (PD) patients from healthy controls. Nonetheless, the prognostic role of oligomers for disease progression is unknown. We explored whether α-synuclein oligomers evaluated as proximity ligation assay (PLA) score may predict the worsening of cognitive functions in patients with Parkinson's disease. Thirty-four patients with PD and thirty-four healthy controls (HC), matched 1:1 for age and sex, were enrolled. Patients with PD underwent baseline skin biopsy and an assessment of cognitive domains including Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Clock Drawing Test, and Frontal Assessment Battery. At the last follow-up visit available, patients were either cognitively stable (PD-CS) or cognitively deteriorated (PD-CD). α-synuclein oligomers were quantified as PLA scores. Differences between groups were assessed, controlling for potential confounders. The relationship between skin biopsy measures and cognitive changes was explored using correlation and multivariable regression analyses. The discrimination power of the PLA score was assessed via ROC curve. To elucidate the relationship between skin biopsy and longitudinal cognitive measures, we conducted multivariable regression analyses using delta scores of cognitive tests (Δ) as dependent variables. We found that PD-CD had higher baseline PLA scores than PD-CS (<i>p</i> = 0.0003), and they were correctly identified in the ROC curve analysis (AUC = 0.872, <i>p</i> = 0.0003). Furthermore, ANCOVA analysis with Bonferroni correction, considering all groups (PD-CS, PD-CD, and HC), showed significant differences between PD-CS and PD-CD (<i>p</i> = 0.003), PD-CS and HC (<i>p</i> = 0.002), and PD-CD and HC (<i>p</i> < 0.001). In the regression model using ΔMMSE as the dependent variable, the PLA score was found to be a significant predictor (β = -0.441, <i>p</i> = 0.016). Similar results were observed when evaluating the model with ΔMoCA (β = -0.378, <i>p</i> = 0.042). In conclusion, patients with Parkinson's disease with higher α-synuclein burden in the peripheral nervous system may be more susceptible to cognitive decline.