Dynamic associations between maternal depressive symptoms and adolescents' depressive and externalizing symptoms

Abstract

The current prospective study investigated transactional relations between maternal depressive symptoms and children's depressive and externalizing symptoms. Participants included 240 children (M age = 11.86 years, SD = 0.56; 53.9% female) and their mothers who were part of a 6-year longitudinal study. Measures of maternal depression (Beck Depression Inventory), child depression (Children's Depression Inventory), and children's externalizing symptoms (Youth Self-Report Form) were assessed annually. Data analyses using dynamic latent difference score structural equation models indicated that the observed relations between mothers' and adolescents' symptoms were stable across the 6 years. Higher levels of maternal depressive symptoms predicted subsequent elevations in children's depressive symptoms and in their externalizing problems over time. Among mothers with high initial levels of depression, children's depressive symptoms predicted subsequent declines in mothers' depressive symptoms. Children's externalizing problems were not related to subsequent change in maternal symptoms.

Figure 1

Dynamic bivariate latent difference score (LDS) model of mothers’ and children’s depressive symptoms. Note: CDI = Child Depression Inventory, BDI = Mothers’ Beck Depression Inventory. Unlabeled one-arrow paths were set equal to 1. Error variances of observed variables were set equal across time for maternal and child depressive symptoms, respectively, and the error variances of mother and child depressive symptoms within each time point were correlated (not shown in model). * indicates the latent variable score; Δ indicates the latent difference score (i.e., latent change), β_C and β_M represent time-varying proportional coefficients for children and mothers, respectively; γ_MC represents unidirectional longitudinal coupling from mothers to children, such that mothers’ depressive symptoms predict subsequent change in children’s symptoms over time; γ_CM represents unidirectional longitudinal coupling from children to mothers, such that children’s depressive symptoms predict subsequent change in mothers’ symptoms over time.

Figure 2

Panel A Expected change in children’s depressive symptoms over time. Note: Scores transformed back to original scale. Mean Time 1 child depressive symptoms scores used. BDI = Mothers’ Beck Depression Inventory. Low BDI = Time 1 BDI score was less than 1 SD below mean. High BDI = Time 1 BDI score more than 1 SD above mean. Panel B. Expected change in mothers’ depressive symptoms over time. Note: Scores were transformed back to original scale. BDI = Mothers’ Beck Depression Inventory; CDI = Child Depression Inventory. Low CDI and Low BDI = Time 1 score was less than 1 SD below mean. High CDI and High BDI = Time 1 score was 1 SD above mean. Average BDI trajectory is plotted based on mean Time 1 BDI and CDI scores.