The Causal Connection Pattern Of The Default Mode Network For First-time Patients With Major Depression

Major depressive disorder (MDD) is a common mental disorder, typically characterized by persistent and overwhelming feelings of sadness, guilt, the decrease of cognitive capacity, insomnia and loss of appetite. There is no doubt that depression makes huge damage on the patients’ daily life and the economy state of society and nation. These years, cognitive psychology focused on the negative influence from depression on cognitive ability, especially working memory and central executive function. Otherwise, neuroimaging tried to explore pathophysiology of depression based on the abnormal structure and functional connectivity of brain. Resting-state fMRI has been widely used to in these researches, during which there is no stimulus for patients to make action. This method can reflect spontaneous neural activity, from which we can infer the inner activity of brain. Previous resting-state fMRI studies showed that abnormal activity of default mode network (DMN) has a strong relationship with major depressive disorder (MDD). Some methods like functional connectivity analysis and independent component analysis have only been used to test undirected connectivity but not directional influence among DMN which can reflect the deep pathophysiology about MDD. The present study aims to discover the causal or directional influence between multivariate DMN regions and the relationship between these and clinical symptoms.Resting-state fMRI was performed on 38 first-episode, treatment-naive patients with MDD and 38 matched healthy subjects. Seed-based functional connectivity (FC) analysis was performed to identify DMN regions. Subsequently, multivariate Granger Causality Analysis (mGCA) was conducted to estimate cross-predictions between all nodes deprived from FC. Pearson correlation analysis was also computed between significant cross-predictions and the score of Clinical Hamilton Depression Rating Scale (HDRS) in MDD patients. In MDD group, we observed two kinds of causal influence:positive and negative. Firstly, right thalamus predicted the increased activity in ventral medial prefrontal cortex (vmPFC) and left medial temporal cortex (MTC) respectively. Secondly, right inferior parietal cortex (IPC) and vmPFC forecasted subsequent deactivation of the left MTC. vmPFC not only received the influence from right thalamus but also exerted an effect on left MTC. In addition, cross-prediction of vmPFC-left MTC correlated with the score of HDRS significantly in MDD.We discovered both positive and negative influences among DMN regions in the two groups. Compared with control group, we emphasized the abnormal effective influence among thalamus, vmPFC and left MTC. Also, our study can be regarded as the applicability of GCA in resting-state study.