Modeling trait depression amplifies the effect of childbearing on postpartum depression

Background: The literature on the relative risk for depression in the postpartum period has largely focused on state (or episodic) depression, and has not addressed trait depression (a woman's general tendency to experience depressed mood). The present study evaluates the association between childbirth and depression in the postpartum period, taking into account the role of stable differences in women's vulnerability for depression across a 10-year span. Methods: Data from the National Longitudinal Survey of Youth 1997 Cohort (N = 4385) were used. The recency of childbirth was used as a predictor of state depression in two models: one that modeled stable depressive symptoms over time (a multi-state single-trait model; LST), and one that did not (an autoregressive cross-lagged model; ARM). Results: Modeling trait depression, in addition to state depression, improved model fit and had the effect of increasing the magnitude of the association between childbirth and state depression in the postpartum period. Limitations: The secondary nature of the data limited the complexity of analyses (e.g., models with multivariate predictors were not possible), as the data were not collected with the present study in mind. Conclusions: These findings may reflect the fact that some of the covariance between childbirth and episodic depression is obscured by the effect of trait depression, and it is not until trait depression is explicitly modeled that the magnitude of the relationship between childbirth and depression becomes clear.