Prevention of Depression in at-Risk Adolescents: Spillover Effects on Non-Targeted Domains

Introduction: Depression is a highly prevalent, disabling, and recurrent disorder. Nearly a quarter of the population will experience clinically impairing depression, with half of onsets occurring in adolescence conferring a high risk for chronic recurrence throughout the lifespan. Adolescent depression also is associated with a range of negative sequelae including disrupted relationships and high conflict, substance use, and other psychiatric comorbidities including anxiety disorders. In the current presentation, we report on the “spillover” outcomes of the prevention of depression in adolescents (POD) trial. 

Methods: POD was a multi-site (Boston, Nashville, Pittsburgh, Portland), randomized trial aimed at preventing depression in adolescents (ages 13-17) at elevated risk for depression due to parental depression and individual factors (i.e., past depression history and/or subsyndromal depressive symptoms). These high-risk adolescents (N=316) were randomly assigned to either a cognitive behavioral prevention program (CBP) or usual care (UC). CBP involved eight 90-minute weekly group meetings, and then six monthly 90-minute meetings. Across the four sites, recruitment goals were successfully met, and there were no site x randomization differences in baseline demographic or clinical characteristics. In POD, CBP demonstrated a significant effect on time to depressive episodes detectable at post-intervention, 3-year, and 6-year follow-up assessments. Measures of secondary outcomes included the Screen for Child Anxiety Related Emotional Disorders (SCARED), Patient Health Questionnaire (PHQ), and Disruptive Behavior Disorders (DBD).

Results: Hierarchical linear modeling (HLM) was used to test for intervention differences in adolescents’ trajectories of symptoms, controlling for child sex and age. Analyses indicated that CBP had significant post-intervention spillover effects on other important symptom domains, including total anxiety symptoms (b = .71, SE = .31, t = 2.27, p = .02), generalized anxiety (γ101 = -.10, SE = .05, t = 2.09, p = .04), social phobia (b = .18, SE = .09, t = 1.96, p = .05), eating problems (b = -.05, SE = .02, t = 2.28, p < .001), defiant behavior (b = .01, SE = .01, t = 2.04), and alcohol and substance abuse (b = -.08, SE = .04, t = 2.15, p = .03).  Duration of these effects and new findings will be examined for later follow-up time-points and implications for the efficient prevention of depression and comorbid conditions discussed. Significant condition by sex interactions were found in predicting change in anxiety. Boys in CBP showed a significant decline in anxiety symptoms as compared to boys in the UC group; boys in the CBP group declined significantly faster than girls in CBP group.

Conclusions: Duration of these effects and new findings will be examined for later follow-up time-points and implications for the efficient prevention of depression and comorbid conditions discussed.