Abstract
Executive Function (EF) deficits are a putative risk factor for adolescent depression, but neural correlates of this association remain unclear. Imbalances between externally and internally oriented attention underlying depression risk may be reflected in altered temporal dynamics of large-scale functional brain networks. This longitudinal study evaluated neural correlates of EF-related depression risk in an adolescent sample (n=154, ages 13-19). We applied co-activation pattern analysis to identify transient network states occurring during a working memory (WM) task, and dynamic structural equation modeling to model depressive symptom trajectories over two years. Results failed to support general EF as predictor of symptoms, but exploratory analyses revealed that WM deficits during low-load conditions predicted greater depression variability. WM-related risk for depression was partially mediated by longer persistence of a transient network involving co-activation of frontoparietal regions. Findings suggest that WM deficits and related temporal dynamics of frontoparietal regions may contribute to depression risk.