An Investigation of the Tripartite Model in Three Age Cohorts of Children and Youth

Abstract

A large body of literature exists to show that childhood anxiety and depression are significantly interrelated, both at the disorder and the symptom level. Clark and Watson (1991) proposed that a tripartite model could account for the observed relations between anxiety and depression. The tripartite model holds that while anxiety and depression share a substantial component of general affective distress or negative affectivity (NA), they can be meaningfully differentiated on the basis of specific symptoms. Anhedonia, or low positive affect (PA), is seen as specific to depression, and physiological arousal (PH), or somatic tension, is seen as unique to anxiety. The studies presented in this thesis were designed to investigate the tripartite model of anxiety and depression in children and youth. Although previous studies provide some support for the validity of the tripartite model in children, they have not systematically examined the generalisability of the model across the developmental spectrum. Where age differences have been explored, there is some evidence to suggest that anxiety and depression show greater differentiation across development. Therefore, the first study sought to test unitary, dual, and tripartite models of anxiety and depression in a cross-sectional design, using three distinct age cohorts of non-referred children and youth recruited from grade 3 (mean age 7.74 years), grade 6 (mean age 10.59 years), and grade 9 (mean age 13.52 years). A confirmatory factor analytic strategy was used with selected items from the Revised Children’s Manifest Anxiety Scale and the Children’s Depression Inventory. The study found little evidence to support the idea of increasing differentiation with age. All models provided a moderate fit to the data, although analyses indicated that a correlated three-factor tripartite model was the preferred model in all age cohorts. These results suggest that the tripartite model may be a clinically useful tool in differentiating between depressive and anxious symptoms and disorders in children of all ages.

The second study then sought to examine the predictive utility of the tripartite model, using a subsample of participants from the first study. In line with the tripartite theory, it was anticipated that the dimensions of NA and PA would predict concurrent and future symptoms of hopelessness (depression). Further, it was expected that the dimensions of NA and PH would predict concurrent and future symptoms of panic (anxiety). Self-report symptoms of hopelessness and panic were measured at baseline and again 12 months later. Results revealed some consistencies and inconsistencies with respect to the predictions of the model. In accordance with predictions, results indicated that NA and PA were significant predictors of concurrent symptoms of hopelessness for grade 3 and grade 6 students, but not for grade 9 students. In addition, NA and PH were significant predictors of concurrent symptoms of panic for grade 6 students, but not for grade 3 or grade 9 students. The tripartite dimensions of NA and PA predicted future hopelessness symptoms for grade 6 students only, however the NA and PH dimensions did not predict future symptoms of panic for any students. Methodology of the study is examined in interpreting the obtained results, and limitations and directions for future research are discussed.