Altered Neural Network Connectivity Predicts Depression in de novo Parkinson’s Disease

• DOI: 10.3389/fnins.2022.828651
• Publication Date: 2022-03-04T05:30:13Z
• Authors: Jianxia Xu; Yubing Chen; Hui Wang; Yuqian Li; Lanting Li; Jingru Ren; Yu Sun; Weiguo Liu
• Keywords: Depression

Background Depression, one of the most frequent non-motor symptoms in Parkinson’s disease (PD), was proposed to be related neural network dysfunction advanced PD patients. However, underlying mechanisms early stage remain unclear. The study aimed explore alterations large-scale networks de novo patients with depression. Methods We performed independent component analysis (ICA) on data resting-state functional magnetic resonance imaging from 21 depression (dPD), 34 without (ndPD), and 43 healthy controls (HCs) extract networks. Intranetwork internetwork connectivity calculated for comparison between groups, correlation analysis, predicting occurrence PD. Results observed an ordered decrease among groups within ventral attention (VAN) (dPD < ndPD HCs), mainly located left middle temporal cortex. Besides, dPD exhibited hypoconnectivity auditory (AUD) default mode (DMN) or VAN compared controls. Correlation revealed that severity negatively correlated value positively AUD-VAN patients, respectively. Further showed area under curve (AUC) prediction 0.863 when combining intranetwork AUD-DMN AUD-VAN. Conclusion Our results demonstrated may associated abnormality bias especially processing. Altered is expected a potential neuroimaging biomarker predict