Dynamic functional connectivity assesses the progression of Parkinson’s disease

Abstract

<p>Parkinson’s disease (PD) induces functional connectivity (FC) changes during its course. However, the impact of PD progression on the temporal properties of FC remains ambiguous. In the current study, we aimed to uncover longitudinal shifts in dynamic FC (DFC) temporal properties of brain networks during PD progression, proposing a novel biomarker for PD progression evaluation. We conducted a longitudinal study on 45 PD patients from the Parkinson’s Progression Markers Initiative database. Patients underwent dual-timepoint neurological assessments and resting-state fMRI scans at baseline and 1-4 years of subsequent follow-up. The sliding-window technique and k-means clustering were employed to scrutinize DFC patterns of the entire brain network, including individual cortical subnetworks and subcortical nuclei (SN) at every timepoint. From this analysis, DFC analyses revealed two predominant states: a high-frequency sparse FC state and a low-frequency intense FC state. For the entire brain network, the mean dwell time (MDT) in the sparse FC state diminished with PD progression, and this decrease was closely tied to motor deterioration. Concerning cortical subnetworks and SN, MDTs in the sparse FC state reduced at the second timepoint in both visual (VN) and limbic networks (LN) linked with the SN. The MDT reduction in LN-SN positively correlated with cognitive decline, while the MDT reduction in VN-SN showed a strong link with motor degradation. These results emphasize that DFC might offer insights into the evolving brain dynamics in PD patients over the disease's course, underscoring its prospective utility as a progression biomarker.</p>