Parenting Self-Efficacy, Childhood Trauma and Trajectory of Perinatal Depressive Symptoms: A Hierarchical Linear Model

Background and Purpose: Perinatal depression (PND) is one of the most common complications of pregnancy, and consequences of PND can be devastating. To effectively address PND, we need a better understanding of its natural course. Correlation between parenting self-efficacy (PSE) and PND is endorsed by previous research. However, empirical evidence is built on either cross-sectional or longitudinal studies where the self-efficacy construct was treated as a time invariant covariate. Both designs failed to capture the fluctuation of PSE as well as the dynamic between PSE and PND. Previous studies mainly focus on the influence of maternal or paternal support on PND; relation between community support and PND is rarely discussed, yet we know that community support is critical in buffering depression for mothers. In addition, there is no consensus whether childhood trauma predicts onset of PND.

To fill the knowledge gap in literature, research questions for this study are:

1. How do depressive symptoms change over time from pregnancy to 24-month after childbirth?

2. How do PSE and community support, as time varying covariates, predict depressive symptoms within individual?

3. How do depressive symptoms vary across individuals with different levels of childhood trauma and demographic characteristics? 

Methods: The dataset “Predicting and Preventing Neglect in Teen Mothers (2001-2007)”, available by National Data Archive on Child Abused and Neglect, was used to explore the research questions. In total, 682 pregnant women were recruited in the primary study and 676 cases were used in this analysis. Fifty eight percent of participants were adolescents and 65% were black. The Beck Depression Inventory II, the Borkowski Parenting self-efficacy and the Childhood Trauma Questionnaire were employed to measure PND, PSE and childhood trauma, respectively. The number of unmet community needs was counted, as indicator of community support (lower values indicate higher community support). Depression, PSE and community support were measured four times from pregnancy to 24-month after childbirth. Hierarchical Linear Model (HLM) was applied. A level-1 model explores trajectory of depressive symptoms and its correlations with PSE and community support within individuals. Level-2 model predicts and explains the variance of trajectories from between-individual level.

Results: Level-1 model shows a quadratic change: over time depressive symptoms decrease (β=-1.72, p< .001), but the decrease become slower and slower (β=0.71, p< .001). After controlling for the inherent tendency to diminish over time, PSE negatively predicts (β= -0.80, p< .001) and unmet community needs positively predicts (β= 0.46, p< .001) depressive symptoms within individuals. Level-2 model shows childhood trauma significantly predicts depressive symptoms (β= 0.16, p< .001). Average level and trajectories of depression do not vary across age, ethnicity, education, and monthly income.

Conclusion and Implications: Our study contributes to existing research by providing evidence that parenting self-efficacy and community support are steady predictors of perinatal depression. Social workers should consider holistic approaches that increase a mother’s community support and her self-efficacy. The integration of trauma-informed care for mothers suffering from perinatal depression is discussed.