Brain Functional Connectivity in de novo Parkinson's Disease Patients Based on Clinical EEG

In Parkinson's disease (PD), cortical-subcortical interplay plays a relevant role in affecting clinical performance. Functional MRI sequences described changes in functional connectivity at different stages of disease. Scarce are, instead, the investigations examining brain connectivity in patients with PD at early stages of disease. For this aim, here we analyzed the differences in functional connectivity between de novo, never treated, PD patients and healthy controls. The analyses were based upon custom-written scripts on the Matlab platform, combined with high-level functions of Fieldtrip, Brainstorm, and Brain Connectivity toolboxes. First, we proceeded to the spectral analysis of the EEG data in the five frequency bands (δ-θ-α-β-γ). Second, we calculated functional connectivity matrices based on both coherency (COH) and imaginary part of coherency (iCOH), in the δ-θ-α-β-γ frequency bands. Then, four network measures (density, transitivity, global efficiency, and assortativity) were computed in identified connectivity matrices. Finally, we compared the spectral density, functional connectivity matrices, and network measured between healthy controls and de novo PD patients through two-samples T-test. A total of 21 de novo PD patients and 20 healthy subjects were studied. No differences were observed in spectral analysis between the two groups, with the exception of the γ band where a significant increase in power density was found in PD patients. A reduced connectivity in the main EEG frequency bands (α-β frequency bands) was observed in PD patients compared to controls, while a hyperconnectivity was found in PD patients in γ band. Among the network measures, a reduced assortativity coefficient was found in de novo PD patients in α frequency band. Our results show the occurrence of early EEG functional connectivity alterations from the initial stages of PD. From this point of view, connectivity analysis may ease a better understanding of the complexity of PD physiopathology.