White matter connectivity networks predict early dementia conversion in Parkinson’s disease

Abstract

AbstractBackgroundSeveral clinical and neuroimaging biomarkers have been proposed to identify individuals at high risk for dementia conversion in Parkinson’s disease (PD). This study aimed to explore whether white matter (WM) connectivity disruption can predict the dementia conversion in patients with newly diagnosed PD with mild cognitive impairment (PD‐MCI).MethodNeuroimaging analyses of WM structural connectivity were performed in 75 patients with drug‐naïve PD‐MCI who underwent serial cognitive assessments during the follow‐up period (>5 years). The patients were classified into either the PD with dementia (PDD) high‐risk group (PDD‐H, n = 38) or low‐risk group (PDD‐L, n = 37), depending on whether they converted to dementia within 5 years of PD diagnosis. We conducted degree‐based statistics (DBS) analyses based on a graph‐theoretical concept to identify the subnetworks whose WM connectivity was disrupted in the PDD‐H group compared with the PDD‐L group. We also performed partial correlation analyses to investigate whether the network connectivity strength was correlated with either the cognitive composite scores or the estimated risk score for PDD conversion within 5 years.ResultThe PDD‐H group showed poorer cognitive performance on frontal/executive, visual memory/visuospatial, and attention/working memory/language function than the PDD‐L group at baseline assessment, even though there was no difference in parkinsonian motor deficits severity, disease duration, and years of education. The PDD‐H group exhibited more severely disrupted WM connectivity in both frontal and posterior cortical regions with eight hub nodes in the DBS analysis. The strength of structural connectivity within the identified subnetworks was correlated with the composite scores of the frontal/executive function domain (γ = 0.393, p = 0.010) and the risk score of PDD conversion within 5 years (γ = ‐0.480, p = 0.001).ConclusionThe present study demonstrated that disrupted WM connectivity in frontal and posterior cortical regions, which was correlated with frontal/executive dysfunction, could be used as a marker for early dementia conversion in patients with PD‐MCI.