Depressive symptom trajectories among girls in the juvenile justice system: 24-month outcomes of an RCT of Multidimensional Treatment Foster Care

Abstract

Youth depression is a significant and growing international public health problem. Youth who engage in high levels of delinquency are at particularly high risk for developing problems with depression. The present study examined the impact of a behavioral intervention designed to reduce delinquency (Multidimensional Treatment Foster Care; MTFC) compared to a group care intervention (GC; i.e., services as usual) on trajectories of depressive symptoms among adolescent girls in the juvenile justice system. MTFC has documented effects on preventing girls' recidivism, but its effects on preventing the normative rise in girls' depressive symptoms across adolescence have not been examined. This indicated prevention sample included 166 girls (13-17 years at T1) who had at least one criminal referral in the past 12 months and who were mandated to out-of-home care; girls were randomized to MTFC or GC. Intent-to-treat analyses examined the main effects of MTFC on depression symptoms and clinical cut-offs, and whether benefits were greatest for girls most at risk. Depressive symptom trajectories were specified in hierarchical linear growth models over a 2 year period using five waves of data at 6 month intervals. Depression clinical cut-off scores were specified as nonlinear probability growth models. Results showed significantly greater rates of deceleration for girls in MTFC versus GC for depressive symptoms and for clinical cut-off scores. The MTFC intervention also showed greater benefits for girls with higher levels of initial depressive symptoms. Possible mechanisms of effect are discussed, given MTFC's effectiveness on targeted and nontargeted outcomes.

Figures

Figure 1

Consort diagram of participant flow for girls who completed any portion of the assessment at a given time point. Note.aSome intervention services were received by all youth, although intervention length varied. In addition, this diagram differs from Kerr et al. (2009), who used a cumulative method of computing participation at any assessment across the study period.

Figure 2

Observed depressive symptom means, their 95th percent confidence intervals, and mean trajectories by group condition over time.

Figure 3

a. The HLM estimated linear growth rates in depressive symptoms by group condition, controlling for initial status and baseline covariates and risk factors. b. The HGLM estimated nonlinear growth rates in depression clinical cut-off scores by group condition, controlling for initial status and baseline covariates and risk factors.