Structure-Function Coupling In Human Brain Networks From Birth To Adolescence And Its Relationship To Transdiagnostic Psychopathology

Half of all lifetime mental health disorders are present by adolescence, but the neural mechanisms underlying transdiagnostic risk across the preceding developmental period remain poorly understood. This dissertation leveraged the UNC Early Brain Development Study (EBDS) to examine structure–function coupling (SFC) from birth through early adolescence in relation to the general psychopathology factor (p-factor). In Study 1, lower SFC within the default mode network (DMN) in early adolescence was associated with higher p-factor. The DMN supports self-referential and internally directed cognition, processes broadly implicated in psychopathology; reduced structure-function alignment within this network may therefore be particularly relevant to transdiagnostic risk. In Study 2, earlier timepoints were used to predict p-factor at ages 12 or 14. Latent growth curve models failed to converge, so three alternative approaches were employed: 1) time-windowed linear models tested whether SFC at each developmental window was associated with p-factor and whether age moderated this association; 2) pairwise change score models examined whether within-individual change in SFC across adjacent age intervals predicted p-factor; and 3) timepoint-specific cross-sectional models replicated Study 1 methods at each age. Windowed and timepoint-specific analyses yielded largely null findings, with one exception: a significant interaction between salience network (SAL)–somatomotor network (SMN) coupling and age in late childhood indicated that SAL–SMN SFC was more positively associated with higher p-factor at younger ages within this window, consistent with less differentiated network architecture and greater reliance on bottom-up, stimulus-driven processing. Pairwise change score analyses revealed temporally specific associations between ΔSFC and p-factor, particularly involving the frontoparietal network (FPN). These associations shifted in direction across developmental windows: negative in early-to-middle childhood, positive in late childhood, and negative again approaching adolescence. These shifts may reflect developmentally contingent consequences of structural constraint: during consolidation, higher FPN SFC may reflect adaptive maturation of cognitive control, whereas during reorganization, equivalent constraint may become maladaptive. Given small pairwise sample sizes, this oscillating pattern warrants cautious interpretation and replication. Together, these findings suggest that transdiagnostic risk is characterized not only by SFC at a single timepoint, but by disruptions in timing of structure–function alignment across large-scale networks throughout development.