Longitudinal associations between depressive symptoms and brain structure across late childhood and adolescence: A panel network analysis study

Abstract Background Major depressive disorder (MDD) is the leading cause of non-fatal disability in youth and disproportionately affects adolescent females. Structural MRI studies of adolescent depression have yielded inconsistent findings, potentially reflecting symptom heterogeneity and rapid developmental changes in brain morphology. Methods In this longitudinal study, we examined associations between specific depressive symptoms and structural brain MRI measures in 9,722 youth (53% male, age range = 10.0–17.7, 24,378 observations) from the Adolescent Brain Cognitive Development (ABCD) Study. A four-wave panel graphical vector autoregression (GVAR) model was estimated to separate within-person (contemporaneous and temporal networks) from stable between-person effects. Brain measures included cortical thickness in the insula, cingulate, medial orbitofrontal cortex (mOFC) and fusiform gyrus, as well as hippocampal volume. Depressive symptoms included parent-reported depressed mood, anhedonia, lethargy, and worthlessness. Additionally, sex-differences in network structures were tested. Results Strong within-domain associations were observed among brain measures and among symptoms, with the largest effects in the symptom domain. Cross-domain (brain-symptom) associations emerged only at the within-person level, where elevated depressed mood was associated with contemporaneous and subsequent reductions in cingulate and fusiform gyrus thickness (partial r = [-0.02 – 0.04]). No cross-domain associations were detected in the between-person networks. Sex-differences emerged only in the within-person networks. Conclusions Associations between brain structure and depressive symptoms were subtle, symptom-specific, and dynamic rather than reflecting stable individual differences. Longitudinal within-person approaches are therefore important for understanding neurodevelopmental contributions to adolescent depression risk.