PSYCHOSOCIAL DOSE RESPONSE AND MATERNAL/NEWBORN OUTCOMES

Purpose/Aims: The purpose of this population-based study was to use hierarchical stepwise regression and complex survey design to evaluate the effects of increasing levels of psychosocial stress on maternal/newborn outcomes, specifically gestational age, birth weight, newborn admission to ICU, and postpartum depression. We hypothesized that increasing levels of maternal stress, depression and abuse would predict increasing risk for adverse maternal/newborn outcomes, in a dose response fashion.  Rationale/Conceptual Basis/Background: It is well documented that maternal psychosocial factors contribute to adverse pregnancy outcomes, but no studies have analyzed the dose response effects of stress, depression and abuse on maternal/newborn outcomes using population-based data. Methods: We analyzed data previously collected by the Utah Department of Health Pregnancy Risk Assessment and Monitoring System (PRAMS) and linked birth certificates for 4682 live births, reflecting a total population size of 143,373 live births between 2009-2011. Exposures of interest were self-reported experiences of maternal stress, depression and abuse before and during pregnancy. Outcomes were gestational age, birth weight, newborn admission to the intensive care unit and postpartum depression. Analysis: Three sets of predictor variables (demographic/behavior, psychosocial, SES/history) were explored for each outcome variable in a hierarchical fashion. We used a forward stepwise approach at each hierarchy with a p-value < .05 as criteria for entry into the model. To deal with the complex survey design of data collection a SAS macro was implemented within SAS 9.3 to obtain corrected p-values. The complex survey design PRAMS weight statements accounted for oversampling of women with lower education and low birth weight in Utah, as well as response rate.  Results: After controlling for maternal demographics, body mass index and smoking, and in a dose-response fashion, women with increasing levels of depression prior to and during pregnancy demonstrated a corresponding increase in newborn admissions to NBICU (OR 1.66 to 2.48; p < .001), postpartum depression symptoms (OR: 3.94 to 9.13; p < .001) and postpartum depression diagnosis (OR: 7.72 to 59.60; p < .001). Maternal stress was associated with increased odds of postpartum depression symptoms (OR 1.34 to 5.51; p < .001), but not postpartum depression diagnosis or NBICU admissions. Implications: Increasing maternal depression and stress are associated with an increasing risk for newborn NBICU admission and/or postpartum depression. These findings suggest that screening for maternal stress and depression as early as the first trimester of pregnancy can identify a group of women who may be at greatest risk for postpartum depression and their babies who may be more vulnerable at the time of birth. Identification of at risk women in early pregnancy could open a greater window of opportunity for initiating interventions with potential for reducing adverse maternal/newborn outcomes.  Future prospective studies could evaluate the usefulness of early prenatal risk identification and introduction of psychosocial interventions for improving maternal/newborn outcomes for high risk women.