Disorganization of Small-World Functional Brain Networks in First-Episode, Treatment-Naïve Adolescents With Major Depressive Disorder.

Background And Aims: Adolescent major depressive disorder (MDD) is prevalent globally but often goes unnoticed due to differences in symptoms compared to adult criteria. Analyzing the brain from a network perspective provides new insights into higher-level brain functions and its pathophysiology. This study aimed to investigate changes in the topological organization of functional networks in adolescents with first-episode, treatment-naïve MDD.

Method: The study included 23 adolescents with depression and 27 matched healthy controls (HCs). Resting-state functional MRI (rs-fMRI) was conducted, and whole-brain functional networks were constructed. Graph theory analysis was used to evaluate network topological properties. A machine-learning multivariate diagnostic model was developed using network metrics associated with depression severity.

Results: Both the MDD and HC groups displayed small-world topology, with male MDD patients showing reduced global clustering efficiency (Cp). The nodal Cp (NCp) and local efficiency (NLE) in the bilateral pallidum were significantly positively correlated with depression severity. In contrast, nodal efficiency (NE) in the left medial orbital superior frontal gyri (ORBsupmed) showed a negative correlation with disease severity. A machine-learning multivariate model using regional network topological features produced an AUROC of 0.71 (95% CI: 0.54-0.92) and an F1 score of 0.65, successfully differentiating adolescent MDD from HCs.

Conclusion: Our findings suggest disruptions in small-world topology in both global and local brain networks in adolescent depression. These abnormal nodal properties may serve as novel neural markers of the disorder.