Preventing Depression in Children of Depressed Parents: Predictors and Moderators of Intervention Response

Introduction:  We aimed to evaluate predictors and moderators of response to two family-based prevention programs aimed at promoting resilience in children of depressed parents: a clinician-facilitated Intervention (Family Talk) and a lecture group intervention. We focused on understanding individual and family-level variables that account for a weaker intervention response.  Such an approach is important to our understanding of how best to match people to different intervention strategies. 

Method:  We enrolled a sample of 105 families in which there was at least 1 parent who suffered from a mood disorder and at least 1 nondepressed child within the 8- to 15-year age range (mean age=11.6 years).  All family members were randomly assigned to one of the intervention groups. Following the completion of the interventions, families were reassessed immediately and 9-12 months thereafter for 4^1/2 years.   The key outcome was the number of changes adults reported on a measure of behavior and attitude change that was developed for this study.  Parent- and family-level variables were examined as possible predictors and moderators, using regression analyses with generalized estimating equations.

Results:  Although positive changes in child-related behaviors and attitudes post-intervention were found for both intervention programs, overall, the lecture-based intervention was less beneficial.  Parent characteristics associated with weaker intervention response for both groups include being older (p<.001), male (p<.001), and reporting a less chronic and severe depression history (p<01).  Moderator effects revealed that relative to other parent groups (i.e., nondepressed parents or depressed/poorly functioning parents), currently depressed but functioning parents reported the most changes in the clinician condition, and currently depressed but poorly functioning parents reported the most changes in the lecture group (p<.05).  Family-level analyses indicated family characteristics associated with a weaker intervention response in both groups (e.g. higher SES, p<.05, dual-parent household, p.001).  Family-level moderator analyses revealed that when both parents are depressed at baseline, there was no difference in the clinician versus lecture-based approach. Furthermore, when the father only was depressed, the clinician condition was better than the lecture condition (p<.05).

Conclusions: These findings demonstrate the role that parent and family characteristics play in determining the effectiveness of preventive intervention programs aimed at promoting resilience in the children of affectively ill parents. Findings from this study can help direct the intervention strategies that are most appropriate for different types of at-risk individuals and families.