Functional neural networks stratify Parkinson's disease patients across the spectrum of cognitive impairment

Abstract

Introduction: Cognitive impairment (CI) is a significant non-motor symptoms in Parkinson's disease (PD) that often precedes the emergence of motor symptoms by several years. Patients with PD hypothetically progress from stages without CI (PD-normal cognition [NC]) to stages with Mild CI (PD-MCI) and PD dementia (PDD). CI symptoms in PD are linked to different brain regions and neural pathways, in addition to being the result of dysfunctional subcortical regions. However, it is still unknown how functional dysregulation correlates to progression during the CI. Neuroimaging techniques hold promise in discriminating CI stages of PD and further contribute to the biomarker formation of CI in PD. In this study, we explore disparities in the clinical assessments and resting-state functional connectivity (FC) among three CI stages of PD.

Methods: We enrolled 88 patients with PD and 26 healthy controls (HC) for a cross sectional clinical study and performed intra- and inter-network FC analysis in conjunction with comprehensive clinical cognitive assessment.

Results: Our findings underscore the significance of several neural networks, namely, the default mode network (DMN), frontoparietal network (FPN), dorsal attention network, and visual network (VN) and their inter-intra-network FC in differentiating between PD-MCI and PDD. Additionally, our results showed the importance of sensory motor network, VN, DMN, and salience network (SN) in the discriminating PD-NC from PDD. Finally, in comparison to HC, we found DMN, FPN, VN, and SN as pivotal networks for further differential diagnosis of CI stages of PD.

Conclusion: We propose that resting-state networks (RSN) can be a discriminating factor in distinguishing the CI stages of PD and progressing from PD-NC to MCI or PDD. The integration of clinical and neuroimaging data may enhance the early detection of PD in clinical settings and potentially prevent the disease from advancing to more severe stages.

Keywords: MCI; Parkinson's disease; cognitive impairment; dementia; fMRI; resting-state.

FIGURE 1

Analysis pipeline.

FIGURE 2

Flowchart of the study.

FIGURE 3

Significant DR analysis between Parkinson's disease (PD)‐normal cognition (NC), PD‐Mild cognitive impairment (MCI), PD‐dementia (PDD), and healthy control (HC). p < .05; corrected for multiple comparisons with threshold‐free cluster enhancement (TFCE). Resting‐state networks are shown in green. Pairwise comparisons show the functional connectivity (FC) differences in the transitions between stages. Peak regions are shown with arrows. More information on cluster sizes and their anatomical locations is given in Table 3. All images are shown in the radiological convention. a‐DMN, anterior default mode network; l‐FPN, left frontoparietal network; l‐VN, lateral visual network; m‐VN, medial visual network; p‐DMN, posterior default mode network; SMN, sensory‐motor network; SN, salience network; TPN, temporoparietal network.

FIGURE 4

Significant differences in edge strength (Z transformed) across every two groups based on Schaefer2018 with 100 parcellation, showing significant differences in the connectivity between two nodes and related violin plots (p < .05, FWE‐corrected). More information on the anatomical locations is given in Table 4.

FIGURE 5

Differences between edge‐bundling connectograms in the inter‐network comparisons. Only four nodes are compared here as example. All connections (edges) of the selected node on the connectograms (e.g., L precuneus) are shown in the inset maps right to the connectograms. Positions of the nodes on the connectograms are according to their hierarchical clustering (Supplementary file 2). Red connections show positive correlations, and blue connections show anticorrelations. Darker and thicker lines indicate stronger connections. The interactive connectograms can be reached at https://osf.io/f4avz/.

FIGURE 6

The continuum of the cognitive impairment (CI) stages of Parkinson's disease (PD).