UCLA

Although growing evidence suggests that biologically-based factors, such as variations in dopaminergic genes and early reactive temperament, may explain ‘differential susceptibility’ to both social stress and support, little is known about whether patterns of environmental sensitivity change across development. Thus, this dissertation employed a developmental psychopathology framework to test differential susceptibility theory in three longitudinal studies of youth at risk for externalizing behaviors (EB; e.g., aggression, substance use), while directly considering the developmental impact of previous adversity. Study 1 was based on a longitudinal study of children adopted from foster care. Generalized estimating equations examined whether reactive temperament heightened sensitivity to both pre-adoption maltreatment and later adoptive family support. Children with reactive temperament did not show heightened vulnerability to pre-adoption maltreatment; instead, maltreatment directly predicted more substance use by late-adolescence. Reactive temperament did heighten sensitivity to the protective effects of family cohesion in early adoption, although effects were not maintained in late-adolescence. Study 2 used latent growth curve analysis to model EB trajectories in a longitudinal sample of adolescents transitioning into adulthood. Employing a polygenic dopaminergic risk score (DRD4, DRD2, DAT1) previously linked to EB, we compared patterns of gene-environment interactions (GxE) between adolescents with and without histories of maltreatment. Polygenic risk moderated the association between parental closeness and concurrent adolescent EB, but these GxE effects did not predict EB changes over time. Furthermore, adolescents with maltreatment histories showed an overall blunted sensitivity to both parental closeness and friendship involvement, and these effects were largely unaffected by dopaminergic genes. Study 3 employed a longitudinal sample of children oversampled for ADHD, an early EB risk factor. Based on prenatal programming theory, a moderated mediation model tested how birth weight (a proxy for global prenatal stress) and dopaminergic genes independently and interactively influence formation of early reactive temperament. Dopaminergic genes interacted with birth weight to predict negative emotionality in a pattern consistent with differential susceptibility. Negative emotionality, in turn, directly predicted adolescent EB, beyond the direct and interactive effects of positive and negative parenting behavior. Overall, results partially support that genetic and temperamental variations influence sensitivity to the environment, although sensitivity was primarily limited to early developmental periods (prenatal period, childhood). We discuss these results in the context of prevailing developmental theories of EB and emphasize the importance of considering the dynamic nature of the environment when investigating individual differences in sensitivity to stress and support.