Associated factors and abnormal dorsal raphe nucleus connectivity patterns of freezing of gait in Parkinson’s disease

Background Freezing of gait (FOG) is a common, disabling symptom of Parkinson’s disease (PD), and its exact pathophysiological mechanism is still poorly understood. The control of gait is a complex process that may be influenced by emotions modulated by serotonergic networks. Therefore, this study aimed to determine factors associated with FOG in PD patients and to evaluate the importance of the dorsal raphe nucleus (DRN; central node in the serotoninergic system) in FOG pathophysiology. Methods We combined cross-sectional survey data from 453 PD patients. According to the Freezing of Gait Questionnaire (FOGQ), patients were divided into two groups: the “PD with frozen gait (PD-FOG)” and “PD without frozen gait (PD-nFOG)” groups. Demographic characteristics, clinical features, and motor and nonmotor symptoms (NMS) assessments of PD patients were recorded. Univariate statistical analysis was performed between the two groups, and then regression analysis was performed on related factors. We also acquired resting-state functional MRI (rs-fMRI) data from 20 PD-FOG, 21 PD-nFOG, and 22 healthy controls (HCs) who were randomly chosen. We defined seeds in the DRN to evaluate functional connectivity (FC) patterns. Results The overall frequency of FOG was 11.9% patients in the PD-FOG group were older, had a longer disease duration, had a higher levodopa equivalent daily dose, had more severe motor symptoms and worse quality of life, had a higher proportion of dyskinesia, wearing-off and postural instability/gait difficulty (PIGD) clinical phenotype, and experienced more depression and impaired sleep function than those in the PD-nFOG group. Logistic regression analysis showed that H&Ystage ≥ 3, UPDRS-III scores, PIGD clinical phenotype and excessive daytime sleepiness were associated with FOG. In addition, there was significantly lower FC between the DRN and some cortical structures, including the supplementary motor area (SMA), left superior frontal gyrus (SFG), and left median cingulated cortex (MCC) in PD-FOG patients than HCs and PD-nFOG patients. Conclusions These results demonstrate that the severity of PD and PIGD clinical phenotype are associated factors for freezing and that DRN dysfunction may play a key role in PD-related NMS and FOG. An abnormal cortical and brainstem networks may contribute to the mechanisms underlying FOG.