Decreased Dorsal Attention Network Homogeneity As a Potential Neuroimaging Biomarker for Major Depressive Disorder

Yujun Gao,Xin Guo,Yi Zhong,Xiaoxin Liu,Shanshan Tian,Jiahui Deng,Xiao Lin,Yanpin Bao,Lin Lu,Gaohua Wang
DOI: https://doi.org/10.1016/j.jad.2023.03.080
IF: 6.533
2023-01-01
Journal of Affective Disorders
Abstract:Background: Gaining insight into abnormal functional brain network homogeneity (NH) has the potential to aid efforts to target or otherwise study major depressive disorder (MDD). The NH of the dorsal attention network (DAN) in first-episode treatment-naive MDD patients, however, has yet to be studied. As such, the present study was developed to explore the NH of the DAN in order to determine the ability of this parameter to differentiate between MDD patients and healthy control (HC) individuals.Methods: This study included 73 patients with first-episode treatment-naive MDD and 73 age-, gender-, and educational level-matched healthy controls. All participants completed the attentional network test (ANT), Hamilton Rating Scale for Depression (HRSD), and resting-state functional magnetic resonance imaging (rs-fMRI) analyses. A group independent component analysis (ICA) was used to identify the DAN and to compute the NH of the DAN in patients with MDD. Spearman's rank correlation analyses were used to explore relationships between significant NH abnormalities in MDD patients, clinical parameters, and executive control reaction time.Results: Relative to HCs, patients exhibited reduced NH in the left supramarginal gyrus (SMG). Support vector machine (SVM) analyses and receiver operating characteristic curves indicated that the NH of the left SMG could be used to differentiate between HCs and MDD patients with respective accuracy, specificity, sensitivity, and AUC values of 92.47 %, 91.78 %, 93.15 %, and 65.39 %. A significant positive correlation was observed between the left SMG NH values and HRSD scores among MDD patients.Conclusions: These results suggest that NH changes in the DAN may offer value as a neuroimaging biomarker capable of differentiating between MDD patients and healthy individuals.
What problem does this paper attempt to address?