Exploring the Impact of the Home Environment as a Moderator in School Performance Outcomes of Low-birth-weight Survivors: a Social-epidemiologic Perspective

There has been substantial research on low birth weight (LBW) as a predictor of adverse educational and cognitive outcomes. LBW infants perform more poorly on cognitive battery tests than children born full term and at normal birth weight, however children exposed to similar risks do not all share the same experiences. The complex, interrelated factors responsible for these long-term developmental problems vary for different populations, but researchers hypothesize that neighborhood conditions and the home environment may influence the infants’ long-term health outcomes. This research seeks to examine the home environment as a moderator in the causal pathway from neonatal brain injury to school performance in a secondary analysis of a prospectively-studied, geographically-defined cohort from the Neonatal Brain Hemorrhage Study (NBHS). The secondary analysis sample included 543 infants with birth weights of 501g-2000g who were born consecutively in three community hospitals in New Jersey between September 1, 1984, and June 30, 1986. The data that were needed came from three different time points (Birth-NICU stay, 6 year-old follow-up, and 9 year-old follow-up). These data were gathered by medical record abstraction, head ultrasound results, maternal interview, and home visits. The dependent variable studied was school performance, as measured by the Woodcock-Johnson Tests of Achievement reading and math scores. The home environment variables/moderators tested included: Discipline, Toys, Home Observations, Neighborhood Observations, and Family Cohesion & Conflict. Perinatal brain injury was the predictor variable of interest. These data were analyzed using multi-step hierarchical regression modeling. A moderating effect between Neighborhood Observations and brain injury was demonstrated for the outcome math score. The moderating relationship was found for Neighborhood Observations in the category of children with no brain injury (β=1.76 p=.005). While there appears to be some statistically significant and potentially clinical meaningful models when looking at Discipline, Toys, Neighborhood Observations, and Observations in the Home as they relate to reading and math scores, a moderating effect is only present in these data with the variable Neighborhood Observations. The findings suggest that at least one variable (Neighborhood Observations) within a LBW child’s socioenvironmental milieu can moderate the effects of perinatal brain injury on school performance outcomes. Degree Type Dissertation Degree Name Doctor of Philosophy (PhD) This dissertation is available at ScholarlyCommons: http://repository.upenn.edu/edissertations/259 Graduate Group Nursing First Advisor Jennifer Pinto-Martin Second Advisor Barbara Medoff-Cooper Third Advisor Terri Lipman