Anxiety and Depression During Childhood and Adolescence: Testing Theoretical Models of Continuity and Discontinuity

Abstract

The present study sought to clarify the trajectory (i.e., continuous vs. discontinuous) and expression (i.e., homotypic vs. heterotypic) of anxiety and depressive symptoms across childhood and adolescence. We utilized a state-of-the-science analytic approach to simultaneously test theoretical models that describe the development of internalizing symptoms in youth. In a sample of 636 children (53% female; M _age = 7.04; SD _age = 0.35) self-report measures of anxiety and depression were completed annually by youth through their freshman year of high school. For both anxiety and depression, a piecewise growth curve model provided the best fit for the data, with symptoms decreasing until age 12 (the "developmental knot") and then increasing into early adolescence. The trajectory of anxiety symptoms was best described by a discontinuous homotypic pattern in which childhood anxiety predicted adolescent anxiety. For depression, two distinct pathways were discovered: A discontinuous homotypic pathway in which childhood depression predicted adolescent depression and a discontinuous heterotypic pathway in which childhood anxiety predicted adolescent depression. Analytical, methodological, and clinical implications of these findings are discussed.

Keywords: Anxiety; Depression; Discontinuity; Growth curve modeling; Heterotypic.

Figure 1

Model A shows the combined piecewise model without cross-lagged paths between symptoms. Model B shows the combined piecewise model with cross-lagged paths. In all models, correlations were examined between symptoms or their residuals within each measurement period. Per modeling conventions, all latent variables were allowed to correlate, but are not presented here for display clarity. Intercept and slope terms are unique for anxiety and depression. Use of the term “child” and “adolescent” is based on age benchmarks as recommended by the National Institute of Child Health and Human Development (NICHD) Pediatric Terminology Harmonization Initiative.

Figure 2

Estimated and actual anxiety means from the Revised Child Manifest Anxiety Scale (RCMAS). Estimated means were derived from a piecewise longitudinal model with a “knot” placed at age 12. A second intercept and slope were estimated at the “knot”. Individual data from a random selection of 10 cases is also displayed and was generated using Mplus version 7.2

Figure 3

Estimated and actual depression means from the Short Mood and Feelings Questionnaire (SMFQ) Depression scale. Estimated means were derived from a piecewise longitudinal model with a “knot” placed at age 12. A second intercept and slope were estimated at the “knot”. Individual data from a random selection of 10 cases is also displayed and was generated using Mplus version 7.2

Figure 4

Developmental model of the expression of anxiety and depressive symptoms. Significant paths (p < .05) are represented by solid black lines and nonsignificant paths (p > .05) are represented by dashed gray lines. Paths predicting anxiety and depressive symptoms are presented. In this model, specific paths were examined between all child symptom variables and all adolescent symptom variables. Additionally, correlations were examined between variables representing the same growth function (e.g., child intercepts) or, for endogenous variables, their disturbances (e.g., adolescent slopes). These correlations are not shown in order to enhance figure clarity. Correlations between disturbances of the same symptom type and period of growth (e.g., childhood depression intercept and slope) were also examined. Covariates of gender and race are not displayed in order to enhance figure clarity. Use of the term “child” and “adolescent” is based on age benchmarks as recommended by the National Institute of Child Health and Human Development (NICHD) Pediatric Terminology Harmonization Initiative.