COGNITIVE-BEHAVIORAL GROUP TREATMENT OF ADOLESCENT DEPRESSION - PREDICTION OF OUTCOME

This paper attempts to identify variables which distinguish depressed adolescents who were recovered versus not recovered at post treatment in a controlled outcome study. Fifty-nine adolescents meeting Research Diagnostic Criteria for major and/or intermittent depression were randomly assigned to one of three conditions: (a) a cognitive-behavioral, psycho-educational treatment group for adolescents (n = 21); (b) an identical protocol for adolescents, but with their parents enrolled in a separate parent group (n = 19); and (c) a waitlist control condition (n = 19). Thirty-seven of the 40 actively treated subjects were retained for prediction analyses examining treatment recovery via two assessment methods: (a) a discriminant function analysis predicting membership in KSADS/RDC diagnosis-based Recovered (n = 16) and Not Recovered (n = 21) groups; and (b) a hierarchical block regression analysis predicting residual pre-topost change on the Beck Depression Inventory. Because of the small sample, potential predictor variables were limited to those which demonstrated significant univariate correlations with the respective outcome variables. Recovery in the discriminant function analysis was associated with lower intake levels on the BDI, lower intake state anxiety, higher enjoyment and frequency of pleasant activities, and more rational thoughts on the Subjective Probability Questionnaire. The overall canonical correlation was .6277. In the hierarchical block regression analysis predicting residual BDI scores, better outcome was associated with a greater number of past psychiatric diagnoses, parent involvement in treatment, and younger age at onset of first depressive episode. The overall regression solution yielded a multiple R = .842. The discussion focuses on the continuity between the current findings and results obtained from similar studies conducted with depressed adults, as well as the degree to which the observed results are consistent with the "capitalization" model of depression therapy outcome.