Childhood trauma and postpartum care use, estimating mediation by posttraumatic stress disorder and depressive symptoms

Abstract

To estimate the effect of childhood trauma on postpartum visit attendance and explore mediation by posttraumatic stress disorder (PTSD) and depressive symptoms. We analyzed data from Grady Trauma Project surveys linked to electronic medical records. We measured childhood trauma using the Childhood Trauma Questionnaire and 4-12 week postpartum visit attendance using outpatient records. We fit log binomial causal mediation models to estimate risk ratios (RR) and 95% confidence intervals (CIs) for the total effect of childhood trauma (categorized as any moderate to severe trauma or not) on postpartum visit attendance and the controlled direct effect, eliminating PTSD or depressive symptoms, controlling for age, parity, income, education, and relationship status. Among 493 birthing people, 54.3% (268) attended the postpartum visit. Individuals without childhood trauma were more likely to attend the visit than those with childhood trauma (56.9% [173/304] v. 50.3% [95/189]). In mediation models, childhood trauma remained associated with reduced likelihood of attendance (adjusted RR: 0.83 [0.68, 1.38]) with possible mediation by PTSD symptoms (controlled direct effect: 0.98 [0.70,1.42]) but not depressive symptoms (controlled direct effect: RR: 0.86 [0.56, 1.38]). Screening and treating PTSD symptoms during pregnancy may slightly increase postpartum care engagement.