Biol Psychiatry. 2026 Mar 12:S0006-3223(26)00099-5. doi: 10.1016/j.biopsych.2026.03.002. Online ahead of print.
ABSTRACT
BACKGROUND: Adolescence is a vulnerable period for the onset of anxiety and depression, yet their neurodevelopmental origins remain unclear.
METHODS: In this 7-year prospective longitudinal study, we recorded resting-state electroencephalogram (EEG) data at ages 7, 9, and 11, followed by fMRI scanning and symptom assessments at age 13. Using connectome-based predictive modeling, we examined whether childhood EEG patterns could predict adolescent symptoms, with rigorous control analyses and external validation in the Healthy Brain Network dataset (HBN, n = 384). We further characterized the developmental trajectories of these predictive networks. To mechanistically ground these electrophysiological markers, we conducted mediation analyses to test whether the amygdala-seeded circuits mediate the link between childhood EEG dynamics and adolescent symptom severity.
RESULTS: We identified dissociable EEG indicators emerging at age 9 that predicted adolescent anxiety (alpha, 8-12 Hz) and depression (beta1, 12-18 Hz). Importantly, the dynamic maturation of these EEG networks highlighted distinct neurodevelopmental susceptibilities, in which longitudinal EEG shifts between ages 9 and 11 predicted symptom severity in adolescence. The divergent developmental trajectories of EEG-based networks were characterized by opposing hemispheric lateralization: leftward for anxiety and rightward for depression. Mechanistically, these predictive associations were mediated by lateralized amygdala-ventrolateral prefrontal cortex (vlPFC) circuits, with the right and left vlPFC pathways selectively mediating anxiety risk and depression risk, respectively. These models generalized robustly to the independent HBN cohort.
CONCLUSIONS: Our findings highlight early neurobiological indicators of distinct developmental trajectories and susceptibilities for anxiety and depression, providing a foundation for early risk stratification and targeted precision prevention.
PMID:41831747 | DOI:10.1016/j.biopsych.2026.03.002
Share Evidence Blueprint
Save to Google Notes
Search Google Scholar
Save as PDF
⭐ My Revision List
