Altered resting state brain networks in Parkinson's disease

Abstract

Parkinson's disease (PD) is a neurodegenerative disorder affecting dopaminergic neurons in the substantia nigra leading to dysfunctional cortico-striato-thalamic-cortical loops. In addition to the characteristic motor symptoms, PD patients often show cognitive impairments, affective changes and other non-motor symptoms, suggesting system-wide effects on brain function. Here, we used functional magnetic resonance imaging and graph-theory based analysis methods to investigate altered whole-brain intrinsic functional connectivity in PD patients (n = 37) compared to healthy controls (n = 20). Global network properties indicated less efficient processing in PD. Analysis of brain network modules pointed to increased connectivity within the sensorimotor network, but decreased interaction of the visual network with other brain modules. We found lower connectivity mainly between the cuneus and the ventral caudate, medial orbitofrontal cortex and the temporal lobe. To identify regions of altered connectivity, we mapped the degree of intrinsic functional connectivity both on ROI- and on voxel-level across the brain. Compared to healthy controls, PD patients showed lower connectedness in the medial and middle orbitofrontal cortex. The degree of connectivity was also decreased in the occipital lobe (cuneus and calcarine), but increased in the superior parietal cortex, posterior cingulate gyrus, supramarginal gyrus and supplementary motor area. Our results on global network and module properties indicated that PD manifests as a disconnection syndrome. This was most apparent in the visual network module. The higher connectedness within the sensorimotor module in PD patients may be related to compensation mechanism in order to overcome the functional deficit of the striato-cortical motor loops or to loss of mutual inhibition between brain networks. Abnormal connectivity in the visual network may be related to adaptation and compensation processes as a consequence of altered motor function. Our analysis approach proved sensitive for detecting disease-related localized effects as well as changes in network functions on intermediate and global scale.

Figure 1. Comparison of the two parcellation approaches.

A) The mean voxel correlation (Fisher transformed) within the ROIs for the parcellation according to the AAL atlas and the parcellation derived from the modular structure within the AAL regions (sub-AAL). B) The number of voxels per ROI for AAL and sub-AAL parcellation is plotted. C) ROIs significantly correlated to the left posterior cingulate cortex (PCC) ROI for the AAL parcellation. D) ROIs significantly correlated to the seed ROI in the left PCC for the sub-AAL ROIs.

Figure 2. Brain regions with high degree centrality in controls.

Shown are the results from voxel-level (red to white scale) and ROI-level (blue) analyses.

Figure 3. Global network properties for Parkinson’s disease patients (PD, full circles) and healthy controls (squares) as a function of the sparsity.

A) mean clustering coefficient B) characteristic path length C) clustering coefficient normalized to a reference random network D) characteristic path length normalized to a reference random network. The asterisks indicate significant between-group differences (permutation test, α = 0.05).

Figure 4. Modular structure of the brain network.

A) Overview of the different network modules. B) mean z-degree of nodes comprising module C) mean participation nodes comprising module for patients (red squares) and controls (blue circles).

Figure 5. Between group effects in z-degree.

Voxel-level (assessed for cluster-wise significance using a cluster defining threshold of p = 0.005; 0.05 FDR-corrected) and ROI-level data (blue areas; 0.05 FDR corrected; sparsity S = 0.2) are presented. A) Regions with a larger degree in healthy controls compared to patients. B) Regions with a larger degree for PD patients compared to controls.

Figure 6. Altered connectivity of the visual network.

A) Altered connectivity of nodes in module 4 (visual network). We observe stronger occipital-frontal and occipital-temporal connectivity for healthy controls. B) Altered connectivity of the cuneus node CunL1. C) Coronal view showing the anatomical location of the two caudate nodes CauL3 and CauR5. Abbreviations: Amy: Amygdala; Cal: Calcarine; Cau: Caudate; Cun: Cuneus; CCA: anterior cingulate cortex; FMO: medial orbitofrontal cortex; Hes: Heschl gyrus; OI: inferior occipital cortex; Par: parahippocampus; RO: rolandic operculum; TM: medial temporal gyrus; TS: superior temporal gyrus.