Postpartum depression in adolescent and adult mothers: comparing prenatal risk factors and predictive models

Our objective was to assess whether postpartum depression risk factors differ between adolescent and adult mothers and to evaluate the need for adolescent specific screening instruments. We performed a retrospective cohort study using data from the Rhode Island Pregnancy Risk Assessment Monitoring System, 2004-2008. We identified maternal age specific risk factors using weighted logistic regression and developed predictive models using a forward selected weighted logistic regression. Notable differences in odds ratios were observed for risk factors such as maternal race (OR Hispanic vs. White: 0.99, 95 % CI 0.49-1.99 among adolescents; 3.32, 95 % CI 2.01-5.49 among adults), pre-pregnancy alcohol use (OR use vs. non-use: 2.04, 95 % CI 1.08-3.86 among adolescents; 0.49, 95 % CI 0.33-0.73 among adults), and pregnancy intention (OR unintended vs. intended: 1.05, 95 % CI 0.37-2.97 among adolescents; 2.67, 95 % CI 1.51-4.74 among adults). In predictive models, adolescent postpartum depressive symptoms were most influenced by prior depression and social support while adult postpartum depressive symptoms were associated with risk factors including maternal race, pregnancy intention, SES, prior depression, mental health during pregnancy, stressors, and social support. We were able to identify similarities and dissimilarities in risk factors for postpartum depressive symptoms among adolescents and adults. Predictive models developed in the general population of pregnant women performed poorly among adolescents relative to age specific predictive models, suggesting that current screening tools may not adequately identify high risk adolescents.