Graph theoretical analyses and network science are powerful tools to study functional organization of the brain.

To investigate functional neural pathway organization and changes over-time in patients with Parkinson’s disease (PD) using advanced network-based techniques.

146 PD patients performed clinical and cognitive evaluations and resting-state functional MRI at baseline and every year for 4 years. Hierarchical cluster analysis identified two PD subtypes: 86 “early” and 60 “mild-to-severe” patients. Within the “early” subtype, two clinical groups were identified: “early-motor-predominant” and “early-diffuse”, the latter having greater cognitive deficits and more frequent non-motor manifestations. 60 age- and sex-matched controls performed baseline assessments. Graph analysis and connectomics assessed global and local topological network properties and regional functional connectivity (FC) at baseline and changes over-time.

“Early” PD patients showed a relatively preserved global functional architecture at baseline and over-time. At baseline, “mild-to-severe” patients showed altered functional topological properties of the sensorimotor and parietal areas relative to controls. Longitudinal analysis showed a progressive deterioration of global functional topological features in “mild-to-severe” patients, with a specific involvement of the sensorimotor, frontal, parietal and temporal regions. At baseline, a widespread pattern of decreased FC involving the basal ganglia, sensorimotor, frontal, parietal and temporal networks was found in “mild-to-severe” and “early-diffuse” patients relative to controls and in “mild-to-severe” relative to “early-motor-predominant” cases. When FC changes over-time were compared between subtypes, five distinct patterns of progression were identified: 1) different trend of change between subtypes (increase vs decrease); 2 & 3) similar trend of change (increase or decrease), with or without FC difference between the groups; 4) different but stable FC values over time in the two subtypes; 5) stable FC with no difference between groups.

Graph analysis and connectomics might represent a powerful approach to understand the pathophysiological process associated with PD progression.