Faculty Bibliography

BACKGROUND:Research surrounding the built environment (BE) and health has resulted in inconsistent findings. Experts have identified the need to examine methodological choices, such as development and testing of BE indices at varying spatial scales. We sought to examine the impact of construction method and spatial scale on seven measures of the BE using data collected at two time points. METHODS:The Children's Environmental Health Initiative conducted parcel-level assessments of 57 BE variables in Durham, NC (parcel N = 30,319). Based on a priori defined variable groupings, we constructed seven mutually exclusive BE domains (housing damage, property disorder, territoriality, vacancy, public nuisances, crime, and tenancy). Domain-based indices were developed according to four different index construction methods that differentially account for number of parcels and parcel area. Indices were constructed at the census block level and two alternative spatial scales that better depict the larger neighborhood context experienced by local residents: the primary adjacency community and secondary adjacency community. Spearman's rank correlation was used to assess if indices and relationships among indices were preserved across methods. RESULTS:Territoriality, public nuisances, and tenancy were weakly to moderately preserved across methods at the block level while all other indices were well preserved. Except for the relationships between public nuisances and crime or tenancy, and crime and housing damage or territoriality, relationships among indices were poorly preserved across methods. The number of indices affected by construction method increased as spatial scale increased, while the impact of construction method on relationships among indices varied according to spatial scale. CONCLUSIONS:We found that the impact of construction method on BE measures was index and spatial scale specific. Operationalizing and developing BE measures using alternative methods at varying spatial scales before connecting to health outcomes allows researchers to better understand how methodological decisions may affect associations between health outcomes and BE measures. To ensure that associations between the BE and health outcomes are not artifacts of methodological decisions, researchers would be well-advised to conduct sensitivity analysis using different construction methods. This approach may lead to more robust results regarding the BE and health outcomes.

BACKGROUND:Previous research has linked polybrominated diphenyl ether (PBDE) exposure to poor birth outcomes and altered thyroid hormone levels. OBJECTIVES/OBJECTIVE:We examined whether maternal PBDE serum levels were associated with infant birth weight (g), head circumference (cm), birth length (cm), and birth weight percentile for gestational age. We explored the potential for a mediating role of thyroid hormone levels. METHODS:During 2008-2010, we recruited 140 pregnant women in their third trimester as part of a larger clinical obstetrics study known as Healthy Pregnancy, Healthy Baby. Blood samples were collected during a routine prenatal clinic visit. Serum was analyzed for PBDEs, phenolic metabolites, and thyroid hormones. Birth outcome information was abstracted from medical records. RESULTS:In unadjusted models, a two-fold increase in maternal BDE 153 was associated with an average decrease in head circumference of 0.32cm (95% CI: -0.53, -0.12); however, this association was attenuated after control for maternal risk factors. BDE 47 and 99 were similarly negatively associated but with 95% confidence intervals crossing the null. Associations were unchanged in the presence of thyroid hormones. CONCLUSIONS:Our data suggest a potential deleterious association between maternal PBDE levels and infant head circumference; however, confirmatory studies are needed in larger sample sizes. A mediating role of thyroid hormones was not apparent.

BACKGROUND:Racial residential segregation has been associated with preterm birth. Few studies have examined mediating pathways, in part because, with binary outcomes, indirect effects estimated from multiplicative models generally lack causal interpretation. We develop a method to estimate additive-scale natural direct and indirect effects from logistic regression. We then evaluate whether segregation operates through poor-quality built environment to affect preterm birth. METHODS:To estimate natural direct and indirect effects, we derive risk differences from logistic regression coefficients. Birth records (2000-2008) for Durham, North Carolina, were linked to neighborhood-level measures of racial isolation and a composite construct of poor-quality built environment. We decomposed the total effect of racial isolation on preterm birth into direct and indirect effects. RESULTS:The adjusted total effect of an interquartile increase in racial isolation on preterm birth was an extra 27 preterm events per 1000 births (risk difference = 0.027 [95% confidence interval = 0.007 to 0.047]). With poor-quality built environment held at the level it would take under isolation at the 25th percentile, the direct effect of an interquartile increase in isolation was 0.022 (-0.001 to 0.042). Poor-quality built environment accounted for 35% (11% to 65%) of the total effect. CONCLUSION/CONCLUSIONS:Our methodology facilitates the estimation of additive-scale natural effects with binary outcomes. In this study, the total effect of racial segregation on preterm birth was partially mediated by poor-quality built environment.

Motivated by a study examining geographic variation in birth outcomes, we develop a spatial bivariate probit model for the joint analysis of preterm birth and low birth weight. The model uses a hierarchical structure to incorporate individual and areal-level information, as well as spatially dependent random effects for each spatial unit. Because rates of preterm birth and low birth weight are likely to be correlated within geographic regions, we model the spatial random effects via a bivariate conditionally autoregressive prior, which induces regional dependence between the outcomes and provides spatial smoothing and sharing of information across neighboring areas. Under this general framework, one can obtain region-specific joint, conditional, and marginal inferences of interest. We adopt a Bayesian modeling approach and develop a practical Markov chain Monte Carlo computational algorithm that relies primarily on easily sampled Gibbs steps. We illustrate the model using data from the 2007-2008 North Carolina Detailed Birth Record.

BACKGROUND:Research has documented the adverse relationship of child cognitive development with maternal prenatal smoking and poor birth outcomes. The potential, however, for maternal prenatal smoking to modify the association between birth outcomes and cognitive development is unclear. METHODS:We linked statewide North Carolina birth data for non-Hispanic white and non-Hispanic black children to end-of-grade test scores in reading and mathematics at fourth grade (n = 65 677). We fit race-stratified multilevel models of test scores regressed on maternal smoking, birth outcomes (as measured by continuous and categorical gestational age and birthweight percentile for gestational age), and their interaction, controlling for maternal- and child-level socio-economic factors. RESULTS:Smoking was consistently associated with decrements in test scores, and better birth outcomes were associated with improvements in test scores, even in clinically normal ranges. Test scores increased quadratically with improving birth outcomes among smoking and non-smoking mothers. Among non-Hispanic white children, the magnitude of the association between gestational age and test scores was larger for children whose mothers smoked during pregnancy compared with the non-smoking group. However, among non-Hispanic black children, birth outcomes did not appear to interact with maternal smoking on test scores. CONCLUSIONS:Maternal prenatal smoking may interact with birth outcomes on reading and mathematics test scores, particularly among non-Hispanic white children. Improvements in birth outcomes, even within the clinically normal range, may be associated with improved academic performance. Pregnancy-related exposures and events exert a significant and long-term impact on cognitive development.

IMPORTANCE/OBJECTIVE:One in 88 children in the United States is diagnosed as having autism spectrum disorder. Significant interest centers on understanding the environmental factors that may contribute to autism risk. OBJECTIVE:To examine whether induced (stimulating uterine contractions prior to the onset of spontaneous labor) and/or augmented (increasing the strength, duration, or frequency of uterine contractions with spontaneous onset of labor) births are associated with increased odds of autism. DESIGN, SETTING, AND PARTICIPANTS/METHODS:We performed an epidemiological analysis using multivariable logistic regression modeling involving the North Carolina Detailed Birth Record and Education Research databases. The study featured 625,042 live births linked with school records, including more than 5500 children with a documented exceptionality designation for autism. EXPOSURES/METHODS:Induced or augmented births. MAIN OUTCOMES AND MEASURES/METHODS:Autism as assessed by exceptionality designations in child educational records. RESULTS:Compared with children born to mothers who received neither labor induction nor augmentation, children born to mothers who were induced and augmented, induced only, or augmented only experienced increased odds of autism after controlling for potential confounders related to socioeconomic status, maternal health, pregnancy-related events and conditions, and birth year. The observed associations between labor induction/augmentation were particularly pronounced in male children. CONCLUSIONS AND RELEVANCE/CONCLUSIONS:Our work suggests that induction/augmentation during childbirth is associated with increased odds of autism diagnosis in childhood. While these results are interesting, further investigation is needed to differentiate among potential explanations of the association including underlying pregnancy conditions requiring the eventual need to induce/augment, the events of labor and delivery associated with induction/augmentation, and the specific treatments and dosing used to induce/augment labor (e.g., exogenous oxytocin and prostaglandins).