Birth defects are the leading known cause of infant mortality and one of the leading sources of years of potential life lost. In fact, one out of every 33 live births in the United States results in a birth defect. The lack of known, modifiable environmental factors associated with birth defects remains a significant barrier to progress. In prior work, our team examined exposure to arsenic, manganese, cadmium, and lead in relation to birth defects. We have demonstrated that exposure to a toxic metals mixture through private well-water is associated with increased risk for birth defects. Specifically, in areas where arsenic and manganese co-occur, we observed higher-than-expected prevalence of birth defects. Current filtration technology is sufficient to reduce exposures to many toxic metals. Unfortunately, widespread adoption of filtration is infeasible due to cost, and there is a substantial gap in knowledge about how to best intervene. Ideally, controlled experiments of water filtration would be used to guide decisions about where to intervene, but, critically, obtaining such data is prohibitively expensive. Thus, public health would be well served by using existing, observational data to estimate the effects of such interventions. This innovative project will use state-of-the-art statistical methods to directly quantify the impact of potential interventions on toxic metal mixtures exposure as a strategy for reducing the risk of birth defects. This approach can be used with varying levels of sophistication to adapt to local public health needs. The overarching hypothesis of this proposal is that we can use routinely collected surveillance data to identify the public health burden of birth defects in North Carolina due to toxic metals exposures, as well as identify interventions to maximize reductions in this burden under realistic constraints on cost and feasibility. We will test this hypothesis in three specific aims.
In Aim 1, we will estimate the risk of birth defects attributable to toxic metal mixtures using data on well water contamination and 1.2 million NC births from 2003-2013 from the NC Department of Health and Human Services and the NC Birth Defects Monitoring Program. We will apply a cutting-edge framework that combines Bayesian methodology with a causal inference framework to estimate attributable risks from highly correlated exposures.
In Aim 2, we apply our framework to estimate the reductions in the attributable risk of birth defects under potential interventions including filtration or changing water sources. We will contrast birth defects risks under interventions that target areas of concern, such as highly exposed areas, as a way to maximize reductions in birth defects.
In Aim 3, we will conduct a cost-effectiveness analysis in order to optimize available resources to reduce birth defects. This work is a paradigm shift in how environmental mixtures can be addressed. The results will provide stakeholders with data for effective decision making. Importantly, our new approach to the analysis of environmental mixtures provides a template for identifying priority exposures, areas, or groups to maximize the public health benefit of policies on exposure mixtures in resource-limited settings.
This project identifies optimal public health strategies to reduce birth defects resulting from exposure to a mixture of toxic metals in well water with a Bayesian causal inference framework. We apply this framework to evaluate targeted well water intervention strategies in North Carolina. This method holds exceptional promise for accelerating public health decisions about interventions related to exposure mixtures in the broader United States.
