Aberrant cortical morphology and brain structure similarity networks in first-episode, treatment-naive adolescents with major depressive disorder

Macroscopic structural hallmarks of the cerebral cortex and the topological organization of its structural networks are emerging as cross-scale evidence for analyzing pathological mechanisms of mental disorders. This study aims to construct individual-based morphological similarity brain networks using gray matter (GM) morphological characteristics of first-episode adolescent patients with major depressive disorder (FEA-MDD) and to explore alterations in GM structure and network topology, as well as their correlations with clinical characteristics. A total of 49 FEA-MDD patients and 50 healthy controls (HCs) were enrolled. Three - Dimensional T1 - Weighted Imaging (3D-T1WI) was collected from all subjects. Freesurfer was used to compute morphological indices, including cortical thickness (CT), gray matter volume (GMV) and surface area (SA), for constructing individual-level morphological brain networks (MBN). Graph theory analysis was employed to calculate network topological properties. Group differences were assessed using two-sample t -tests. Spearman correlation analysis was performed to examine the relationship between differential cortical structures, network topological properties and the clinical scores in FEA-MDD. Compared with the HCs, the FEA-MDD exhibited cortical thinning in the left middle frontal gyrus, lingual gyrus, bilateral postcentral gyrus and calcarine cortex, while cortical thickening was observed in the left insula and inferior temporal gyrus. The SA of the right precuneus and fusiform gyrus was significantly reduced. Additionally, GMV was decreased in the bilateral calcarine cortex, left postcentral gyrus, left temporal pole and right precuneus. Notably, CT-based morphometric networks demonstrated significant topological remodeling, characterized by increased global efficiency (Eg) and altered nodal properties in brain regions associated with emotional processing and higher-order cognitive functions, including the prefrontal-cortical-limbic regions and visual networks. In contrast, SA- and GMV- based networks remained largely intact. Importantly, pericallosal sulcus nodal efficiency (Ne) and degree centrality (DC) negatively correlated with rumination and self-esteem, while calcarine Ne was linked to anhedonia, and cuneus DC correlated with self-esteem. FEA-MDD patients have differences in brain structure and topological properties of cortical similarity network. The structural and network alterations are interrelated and jointly influence brain function, offering novel insights into the neuropathological mechanisms underlying adolescent depression. • Structural alterations in first-episode adolescent MDD: Reduced cortical thickness in frontal, visual, and somatosensory regions (e.g., left middle frontal gyrus, lingual gyrus, bilateral postcentral gyrus) and increased thickness in limbic and temporal areas (left insula, inferior temporal gyrus). • Gray matter volume and surface area changes: Decreased volume in bilateral calcarine cortex, left postcentral gyrus, and right precuneus; reduced surface area in right precuneus and fusiform gyrus. • Network topology remodeling: Cortical thickness-based brain networks showed enhanced global efficiency and altered nodal properties in prefrontal-limbic and visual networks, linking to emotional processing and cognitive dysfunction. • Clinical correlations: Nodal efficiency/centrality in pericallosal sulcus, calcarine cortex, and cuneus correlated with rumination, anhedonia, and self-esteem, highlighting their roles in depressive symptomology. • Integrative insights: Structural and network alterations in adolescent MDD suggest disrupted cross-brain-region connectivity, providing novel neuroimaging evidence for understanding the pathological mechanisms of early-life depression.