Short and long-term exposures to fine particulate matter (PM2.5) have been associated with increased risks of cardiovascular disease in large urban areas. Although previous studies have included increasingly larger populations and sophisticated analytical methods, many gaps of knowledge remain. One limitation is that by relying on the Environmental Protection Agency air pollution monitoring network only studies will: 1) exclude less urban areas where PM2.5 is not monitored; 2) not be able to differentiate health effects associated with regional/ long-range transport versus local pollution sources; and 3) not be able to estimate health effects of air pollution at low concentrations, which requires large populations in that exposure range. In addition, with the mounting threat of climate change, evidence regarding a potential synergistic effect of air pollution and heat is totally lacking.
In Aim 1 we will integrate satellite-based aerosol optical depth (AOD) data with land use regression variables to predict and validate daily PM2.5 levels at 1x1 km grid cell resolution for the entire US. This data will be used to calculate zip-code exposures that vary by location and date, which will be linked to the entire Medicare cohort (90% of the US elderly population) over the period 2000 to 2013.
In Aim 2 we will develop computationally efficient algorithms to analyze these linked data sets and provide evidence of acute and chronic risks of cardiovascular disease of unprecedented accuracy by using improved exposure to PM2.5 at a very local scale, and including rural populations from locations in which pollution monitors are not available. We will estimate the shape of the exposure-response relationship between cardiovascular admissions and PM2.5 to address questions of effects below the air quality standards; and evaluate effect modification by rural/ urban area, by multiple socioeconomic measures, by individual and by land use characteristics.
In Aim 2. 1 we will decompose the estimated daily pollution surfaces into different spatial scales, representing regional/long- range transport and locally-generated pollution.
In Aim A.2.2 we will estimate the effects of PM2.5 components and gases in urban and rural areas in New England.
In Aim A.2.2 we will develop approaches to adjust for confounding bias using auxiliary information.
In Aim 3 we will estimate the acute and chronic health risks associated with simultaneous exposure to heat and PM2.5.
In Aim 4, we will make available to the research community maps of the output obtained in the previous aims. We will also disseminate software for the efficient implementation of the newly developed statistical methods needed for the analysis of these big data sets. In summary, without a substantial improvement in the accuracy of the exposure data, sample size, methodology, and characterization of the exposure response relationship and the joint effects of heat and air pollution, it will be challenging to provide the necessary evidence for future air quality regulations. The results will provide a much needed evidence base for developing the most cost effective and beneficial air quality interventions.
This project will quantify the acute and chronic cardiovascular effects of PM2.5 and other pollutants using improved exposure at a smaller scale, including rural populations that are not currently monitored, and at low concentrations; will isolate the health effects of regionally and locally-generated pollution; identify sources that explain effect heterogeneity; develop statistical approaches to adjust for confounding using auxiliary data; characterize potential synergism between PM2.5 and temperature; develop methods for the analysis of massive spatio-temporal data sets, and disseminate software and data to the research community.
Showing the most recent 10 out of 35 publications
