Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Jennifer . Laurent


Objective: Internalizing problems are commonly diagnosed during adolescence, and are associated with distress, impairment, and negative mental health outcomes in adulthood. Thus, there is a critical need to characterize adolescents who are at the highest risk for escalating to clinical levels of internalizing problems while extending current literature and incorporating both biological and environmental predictors. This study aimed to characterized risk profiles for fourteen-year-old adolescents who developed clinical levels of internalizing (High Internalizing [HI]) problems by age nineteen, using brain, genetic, personality, cognitive, life history, psychopathology, and demographic measures. The study also examined whether there were functional and structural brain differences in three groups of adolescents on select regions of interest (ROIs) on the Faces Task, Stop Signal Task, and Modified Incentive Delay Task.

Method: Participants were 91 adolescents who met clinical criteria for at least one Anxiety and/or Depressive Disorder by age 19 and 1,244 controls who varied in symptom level but did not reach clinically-diagnostic criteria. Ten-fold cross-validated logistic regression using elastic net regularization was used to identify risk profiles associated with high levels of internalizing symptomatology. To examine group differences in regions of interest on three fMRI tasks and in gray matter volume, ANCOVAs were conducted. The three groups were: 1) adolescents who never met HI criteria (Controls), 2) those who met HI criteria in middle adolescence (Middle Onset), and 3) those who met HI criteria in late adolescence (Late Onset).

Results: Logistic regression identified 13 variables from personality, psychopathology, life events, and functional brain variables to predict High Internalizing symptoms (mean AUC 0.78, p<.0001). ANCOVAs showed there were several ROIs that demonstrated main effects of Time, and one main effect of Group during response inhibition in the left inferior frontal gyrus, triangular part (pars triangularis), with participants in the Middle Onset group showing increased activation levels compared with the Control group. There were no other significant main effects of Group or Time x Group interactions.

Conclusions: These findings give insight into personality, psychopathological, and brain-related factors that are associated with high levels of internalizing symptoms, highlighting the importance of including biological variables in conjunction with psychosocial variables when examining risk factors for internalizing problems. Results also suggest an association between activation in frontal cortex and parietal lobe regions during response inhibition and higher internalizing symptoms in late adolescence. Between-group activation and volumetric ROI comparisons generally yielded main effects of time, confirming prior evidence that activation levels and GMV continue to change over the course of adolescence.



Number of Pages

73 p.