Brain functional connectome defines a transdiagnostic dimension associated with cognition and psychopathology in youth

Abstract

ABSTRACTThe search for psychiatric biomarkers has remained elusive, in part, due to high comorbidity, low specificity, and poor concordance between neurobiological abnormalities and existing diagnostic categories. Developmental shifts in symptom expression and brain function across the lifespan further complicate biomarker identification. Recent studies suggest that focusing on cognition may be a pathway forward: Cognitive dysfunction is a common feature across psychiatric disorders. Individual differences in cognition may reflect variability in the connectivity of underlying neurocognitive brain networks, and predict psychopathology at different developmental periods. In the present study we identified brain-based dimensions of general cognitive capacity and psychopathology using sparse canonical correlation analysis (sCCA) in a sample of 7,383 preadolescents from the Adolescent Brian Cognitive Development (ABCD) study. This analysis revealed correlated patterns of functional connectivity with cognitive control capacity and psychopathology. Specifically, the results identified a single connectome-based latent brain variate that was positively correlated with performance on cognitive measures across domains and negatively correlated with parent-reported psychopathology across diagnoses and domains. Functional connectivity loadings for the brain variate were across distributed cortical and subcortical brain networks and a dose-dependent relationship with psychopathology based upon the cumulative number of psychiatric diagnoses was observed. These findings provide preliminary evidence for a connectome-based biomarker that indexes individual differences in cognitive control and predicts transdiagnostic psychopathology in a dose-dependent fashion.