Graph theory reveals functional connectome disruptions in adolescent major depressive disorder with childhood trauma.
Authors
Affiliations (9)
Affiliations (9)
- College of Artificial Intelligence Medicine, Chongqing Medical University, Chongqing, China.
- Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China.
- Department of Psychiatry, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China.
- Department of Clinical Neurosciences, University of Cambridge, Cambridge, United Kingdom.
- Department of Psychiatry and Behavioral Neuroscience, University of Cincinnati College of Medicine, Cincinnati, OH, USA.
- Department of Psychiatry, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China. [email protected].
- Key Laboratory of Major Brain Disease and Aging Research (Ministry of Education, Chongqing Medical University, Chongqing, China. [email protected].
- College of Artificial Intelligence Medicine, Chongqing Medical University, Chongqing, China. [email protected].
- Key Laboratory of Major Brain Disease and Aging Research (Ministry of Education, Chongqing Medical University, Chongqing, China. [email protected].
Abstract
Childhood trauma (CT) is a major risk factor for adolescent major depressive disorder (MDD), yet its neurobiological underpinnings and longitudinal treatment effects remain poorly characterized. Leveraging graph theory and resting-state fMRI, we analyzed in 343 adolescents with MDD aged 10 - 18 years, including 211 with a history of childhood trauma (MDD-CT) and 106 without childhood trauma (MDD-NCT), as well as 149 healthy controls. Machine learning models were applied to baseline functional network data to distinguish between treatment responders and non-responders. We identified CT-associated functional connectome disruptions marked by increased network randomness and topological deficits in default mode network (DMN) hubs (left parahippocampal gyrus, posterior cingulate gyrus, temporal pole). Longitudinal neuroimaging revealed post-treatment normalization of these abnormalities, particularly in the left precuneus and amygdala, paralleling symptom improvement. Machine learning models using baseline connectomes predicted antidepressant response with 82% accuracy. Our findings establish CT-driven connectome disturbances in adolescent MDD, map dynamic network reorganization to therapeutic recovery, and position functional connectivity as a clinically actionable biomarker. This work bridges neurobiological mechanisms of trauma-related depression with precision treatment strategies, offering a path toward biomarker-guided interventions.