Parkinson\'s disease (PD) patients with postural gait abnormalities exhibit poorer motor function scores, more severe non-motor symptoms, faster cognitive function deterioration, and a less favorable response to drugs and surgery compared to PD patients with tremor. This discrepancy is believed to be associated with more pronounced gray matter atrophy and abnormal functional connectivity. To investigate the distinctive pathological mechanisms between PD subtypes, we examined gray matter volume (GMV) and functional connectivity in patients with Parkinson\'s disease presenting with postural instability/gait difficulty (PD-PIGD), patients with tremor-dominant Parkinson\'s disease (PD-TD), and healthy controls. Voxel-based morphometry (VBM) of T1-weighted images was conducted to compare GMV among 64 PD-PIGD patients, 44 PD-TD patients, and 32 controls. Subsequently, functional connectivity within regions showing reduced GMV was compared across the groups. We analyzed whether differences among the groups were associated with clinical characteristics and neuroimaging biomarkers using partial correlation and binary logistic regression. Our comparison between PD-PIGD and PD-TD patients revealed a link between PD-PIGD and more extensive frontotemporal atrophy, potentially indicating increased basal ganglia activity accompanied by decreased cerebellum activity. Furthermore, in addition to the smaller GMV in the left middle temporal gyrus, the increased functional connectivity between this brain region and the right caudate was also the independent risk factor for PD-PIGD. In addition, we compared brain network connectivity between the PIGD and TD subtypes, using an independent component analysis (ICA). We found that Compared to PD-TD, PD-PIGD patients showed an enhanced sensorimotor network (SMN) around the left supplementary motor area. These findings suggest that severe gray matter atrophy and abnormal functional connectivity and brain networks may serve as pathophysiological mechanisms distinguishing PD-PIGD patients from other subtypes.

BACKGROUND: Tremor-dominant (TD) and postural instability/gait difficulty (PIGD) are common subtypes of Parkinson\'s disease, each with distinct clinical manifestations and prognoses. The neural mechanisms underlying these subtypes remain unclear. This study aimed to investigate the altered connectivity of the frontal cortex and supplementary motor area (SMA) in different types of Parkinson\'s disease.
METHODS: Data of 173 participants, including 41 TD patients, 65 PIGD patients, and 67 healthy controls, were retrospectively analyzed. All subjects underwent resting-state functional magnetic resonance imaging (rs-fMRI) and clinical assessments. Differences in amplitude of low frequency fluctuation (ALFF), voxel-wise functional connectivity (FC), and functional network connectivity (FNC) among the three groups were compared, followed by partial correlation analysis.
RESULTS: Compared to healthy controls, the left dorsolateral superior frontal gyrus (DLSFG) ALFF was significantly increased in both PIGD and TD patients. The FC between the left DLSFG and the left SMA, as well as between the left paracentral lobule and the right DLSFG, was significantly decreased. Similarly, the FNC between the visual network and the auditory network was reduced. Compared to TD patients, PIGD patients showed a significantly higher ALFF in the left DLSFG and a notably reduced FC between the left DLSFG and left SMA. Additionally, the FC of the left DLSFG-SMA was inversely correlated with the PIGD score exclusively in PIGD patients. The FNC of the visual-auditory network was inversely associated with the tremor score only in TD patients.
CONCLUSIONS: Decreases in the left DLSFG-SMA connectivity may be a key feature of the PIGD subtype, while reduced VN-AUD connectivity may characterize the TD subtype.

The Postural Instability/Gait Difficulty (PIGD) subtype of Parkinson\'s disease (PD) has a faster disease progression, a higher risk of cognitive and motor decline, yet the alterations of structural topological organization remain unknown. Diffusion Tensor Imaging (DTI) and 3D-TI scanning were conducted on 31 PD patients with PIGD (PD-PIGD), 30 PD patients without PIGD (PD-non-PIGD) and 35 Healthy Controls (HCs). Structural networks were constructed using DTI brain white matter fiber tractography. A graph theory approach was applied to characterize the topological properties of complex structural networks, and the relationships between significantly different network metrics and motor deficits were analyzed within the PD-PIGD group. PD-PIGD patients exhibited increased shortest path length compared with PD-non-PIGD and HCs (P < 0.05, respectively). Additionally, PD-PIGD patients exhibited decreased nodal properties, mainly in the cerebellar vermis, prefrontal cortex, paracentral lobule, and visual regions. Notably, the degree centrality of the cerebellar vermis was negatively correlated with the PIGD score (r = -0.390; P = 0.030) and Unified Parkinson\'s Disease Rating Scale Part III score (r = -0.436; P = 0.014) in PD-PIGD patients. Furthermore, network-based statistical analysis revealed decreased structural connectivity between the prefrontal lobe, putamen, supplementary motor area, insula, and cingulate gyrus in PD-PIGD patients. Our findings demonstrated that PD-PIGD patients existed abnormal structural connectomes in the cerebellar vermis, frontal-parietal cortex and visual regions. These topological differences can provide a topological perspective for understanding the potential pathophysiological mechanisms of PIGD in PD.