Dynamic functional connectivity reveals hyper-connected pattern and abnormal variability in freezing of gait of Parkinson's disease

Abstract

Background: Freezing of gait (FOG) is an intractable and paroxysmal gait disorder that seriously affects the quality of life of Parkinson's disease (PD) patients. Emerging studies have reported abnormal brain activity of distributed networks in FOG patients, whereas ignoring the intrinsic dynamic fluctuations of functional connectivity. The purpose of this study was to examine the dynamic functional network connectivity (dFNC) of PD-FOG. Methods: In total, 52 PD patients with FOG (PD-FOG), 73 without FOG (PD-NFOG) and 38 healthy controls (HCs) received resting state functional magnetic resonance imaging (rs-fMRI). Sliding window method, k-means clustering and graph theory analysis were employed to retrieve dynamic characteristics of PD-FOG. Partial correlation analysis was conducted to verify whether the dFNC was related to freezing gait severity. Results: Seven brain networks were identified and configured into seven states. Compared to PD-NFOG, significant spatial pattern was identified for state 2 in freezers, showing increased functional coupling between default mode network (DMN) and basal ganglia network (BG), as a concrete manifestation of increased precuneus-caudate coupling. The mean dwell time and fractional window of state 2 had a positive correlation with FOG severity. Furthermore, PD-FOG group exhibited lower variance in nodal efficiency of independent components (IC) 7 (left precuneus). Conclusions: Our study suggested that aberrant coupling of precuneus-caudate and disrupted variability of precuneus efficiency might be associated to the neural mechanisms of FOG.

Keywords

• Freezing of gait
• Parkinson's disease
• Dynamic functional network connectivity
• Dynamic graph theory property
• Resting-state functional connectivity