Doctoral Dissertation Research: Spatial Inequality in Birth Outcomes - Testing Classes of Proximate Mechanisms

Harvard University, Cambridge MA

Investigators

Abstract

Low birth weight and preterm birth have lasting and negative implications for healthy child development, which influences their later educational and socioeconomic attainment. These birth outcomes also occur more frequently in some neighborhoods than in others, but we do not understand the mechanisms through which neighborhood characteristics may be influential. This project will analyze a wide range of mechanisms that vary by neighborhood and may contribute to adverse birth outcomes. These mechanisms include access to prenatal care, such as transportation time to hospitals or neighborhood social connectedness, as well as sources of stress such as public disorder, public conflicts or private conflicts. The project will control for a wide range of individual level determinants, including prior maternal health. It will also analyze a large sample of pregnancies from the same hospital, thus controlling for health care delivery, which can also implicate low birthweight and preterm birth outcomes. The project will provide implications for how neighborhood factors can influence birth outcomes, thus suggesting how social policies may be constructed to reduce birth outcome variation. This study will be the first to determine whether and why birth outcome inequality persists including both individual and neighborhood level determinants, holding sorting into hospitals constant. The project will use a novel dataset comprised of a cohort of 4,324 women who were recruited by their first trimester and who delivered live singleton births at the Brigham and Women's Hospital in Boston, MA., between the years 2006 and 2015. The project will use multivariate analysis to determine whether spatial inequality in birth outcomes occurs net of a wide range of individual determinants. The project will use clinical measures of maternal health status and exact addresses to capture neighborhood characteristics. Transportation times are computed using Google Maps API in R. Contextual sources of stress are merged with individual-level pregnancy data from two sources: (1) the 2006, 2008 and 2010 BNS (Boston neighborhood) Surveys, with linear imputations on non-survey years; and (2) annual reports from the City of Boston's 911 database and 311 databases. Contextual variables are merged with the pregnancy data in order to form a spatial weights matrix and data are analyzed using hierarchical linear modeling. Lastly, the project incorporates bio-markers of angiogenesis, which are available at three points in pregnancy for a subset of the data, to perform causal mediation analyses. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.