Source-specific multi-pollutant exposures and the neighborhood context in disparities in stillbirth

点击翻译按钮获取中文摘要

ABSTRACT The origin of racial and ethnic disparities in obstetric outcomes does not appear to be genetic nor fully explained by individual-level factors. Hence, we must look toward the broader context in the physical, social and built environments to understand and mitigate disparities in maternal and child health. In the physical, social and built environments, air pollution, neighborhood disadvantage, and residential greenness may play important roles. As compared to low birth weight, fetal growth restriction and preterm birth, considerably less is known about air pollution exposure's impact on stillbirth, defined as an intrauterine fetal death that occurs at or after 20 weeks of gestation. Urban air pollution is a complex mixture of particle and gaseous pollutants, often correlated because they are influenced by common nearby sources and meteorology and thus, simultaneously studying exposure to multiple air pollutants is challenging. Previous studies on the risk of stillbirth associated with air pollution exposure have largely focused on pollutant-specific effects, including our studies examining ozone and metal constituents of PM in the fourth largest city in the U.S., Houston, Texas. Using source apportionment to assess air pollutant exposures takes advantage of the highly correlated nature of the air pollutant mixture, quantifies the contributions of different source categories and provides information that allows more targeted control strategies to be developed. To date, we are aware of only one study that applied source apportionment for fine particulate matter, and it used the Positive Matrix Factorization (PMF) receptor modeling. Yet, that investigation used estimated source contributions as if they were true exposures without accounting for spatial misalignment error and inherent uncertainty in the method itself. Addressing these shortcomings, we will apply an innovative approach that integrates Bayesian spatial multivariate receptor models (BSMRM) with land-use and traffic data as well as accounting for exposure measurement errors in health effects estimation. Like air pollution, the social and built environments cluster neighborhood-level exposures that adversely impact health. These stressors are often greater among minority and low-income populations who suffer what has been termed “double jeopardy”. The objective of the proposed study is three-fold. First, we wish to evaluate the impact of source-specific multipollutant exposures on stillbirth using an innovative approach that integrates BSMRM with land-use and traffic data and accounting for exposure measurement errors in health effects estimation. We hypothesize that increased air pollutant source-apportioned concentrations will be associated with increased risk of stillbirth and that effect sizes will be different for different source categories. Second, we wish to examine the impact of the neighborhood context on risk of stillbirth. Third, given the priority of health disparities for the NIH and the nation, another objective is to evaluate the hypothesis that minorities will suffer disproportionate stillbirth risks associated with air pollutant source-apportioned concentrations and with living in deprived neighborhoods, as assessed by neighborhood disadvantage and residential greenness.