Longitudinal internalizing psychopathology structure in a diverse community sample of Los Angeles adolescents from 9th to 12th grade.

Purpose: Manifestations of internalizing symptoms vary developmentally, particularly during adolescence, but most taxonomies, including modern structural models of psychopathology, do not adequately account for this. Understanding variation in the structure and components of internalizing psychopathology may improve diagnosis. Methods: Self-reported internalizing symptom data were collected from a diverse sample of LA-county adolescents (N = 3368) during each of the four years of high school (grades 9-12). DSM-based symptoms of generalized anxiety, major depression, social phobia, specific phobia, and obsessive-compulsive problems were assessed. Covariance symptom structures were modeled at the four waves using exploratory and confirmatory factor analysis. Results: Based on model fit, parsimony, and meaningfulness, a five-factor model with one factor for each of the DSM-defined internalizing disorders was selected at freshman year (CFI = 0.94), a two-factor fear-and-distress model at sophomore year (CFI = 0.91), and a one-factor model at both junior (CFI = 0.91) and senior years (CFI = 0.92). In freshman year, the nature of the structure is aligned with DSM conceptions of internalizing problems, in sophomore year, the nature of the structure is aligned with fear-distress conceptualizations of internalizing problems as compared to anxiety-depression, and in junior and senior year, the nature of the structure was aligned with a single internalizing dimension. Conclusion: Findings suggest that internalizing problems become increasingly unidimensional across high school. Findings may indicate changes in symptom presentation; accounting for these changes may improve the identification and treatment of internalizing problems among adolescents.