Prevention of Depression in At-Risk Adolescents: Predictors and Moderators of Acute Effects.

OBJECTIVE

To assess predictors and moderators of a cognitive-behavioral prevention (CBP) program for adolescent offspring of parents with depression.

METHOD: This 4-site randomized trial evaluated CBP compared to usual community care (UC) in 310 adolescents with familial (parental depression) and individual (youth history of depression or current subsyndromal symptoms) risk for depression. As previously reported by Garber and colleagues, a significant prevention effect favored CBP through 9 months; however, outcomes of CBP and UC did not significantly differ when parents were depressed at baseline. The current study expanded on these analyses and examined a range of demographic, clinical, and contextual characteristics of families as predictors and moderators and used recursive partitioning to construct a classification tree to organize clinical response subgroups.

RESULTS

Depression onset was predicted by lower functioning (hazard ratio [HR] = 0.95, 95% CI = 0.92-0.98) and higher hopelessness (HR = 1.06, 95% CI = 1.01-1.11) in adolescents. The superior effect of CBP was diminished when parents were currently depressed at baseline (HR = 6.38, 95% CI = 2.38-17.1) or had a history of hypomania (HR = 67.5, 95% CI = 10.9-417.1), or when adolescents reported higher depressive symptoms (HR = 1.04, 95% CI = 1.00-1.08), higher anxiety (HR = 1.05, 95% CI = 1.01-1.08), higher hopelessness (HR = 1.10, 95% CI = 1.01-1.20), or lower functioning (HR = 0.94, 95% CI = 0.89-1.00) at baseline. Onset rates varied significantly by clinical response cluster (0%-57%).

CONCLUSION

Depression in adolescents can be prevented, but programs may produce superior effects when timed at moments of relative wellness in high-risk families. Future programs may be enhanced by targeting modifiable negative clinical indicators of response.

CLINICAL TRIAL REGISTRATION INFORMATION: Prevention of Depression in At-Risk Adolescents; http://clinicaltrials.gov/; NCT00073671.