Abstract
BACKGROUND: Major depressive disorder (MDD) and bipolar disorder (BD) commonly emerge during adolescence, a critical stage of brain development. Characterizing their shared and distinct neurobiological features is essential for early differential diagnosis. METHODS: In this study, we investigated adolescents with major depressive disorder (MDD) (n = 48), bipolar disorder (BD) (n = 37), and healthy controls(n = 40). We acquired diffusion tensor imaging (DTI) data and applied a combined framework of graph-theoretical analysis and fixel-based analysis (FBA). We reconstructed structural networks using deterministic tractography and calculated global and nodal graph metrics. We then used FBA to measure fiber density (FD), fiber cross-section (FC), and combined FDC across major white matter tracts. We tested group differences and examined correlations with clinical severity after multiple-comparison correction. RESULTS: Global graph metrics did not differ across groups, whereas nodal alterations were observed in regions assigned to the default mode, salience, and central executive networks, with disorder specific patterns. Fixel based analysis indicated widespread tract involvement in major depressive disorder but more circumscribed abnormalities in bipolar disorder, particularly in the corpus callosum and fornix. Several metrics were associated with symptom severity. CONCLUSIONS: These findings suggest disorder-related alterations in nodal topological properties and fiber-specific microstructure in adolescent MDD and BD, providing insight into the neurobiological basis of mood symptoms during this developmental stage. While these structural features may assist future efforts in differential characterization or individualized risk assessment, their potential clinical utility requires further validation. CLINICAL TRIAL NUMBER: Not applicable. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12888-026-07961-x.