2012 Progress Report: A National Study to Assess Susceptibility, Vulnerability, and Effect Modification of Air Pollution Health RisksEPA Grant Number: R834798C005
Subproject: this is subproject number 005 , established and managed by the Center Director under grant R834798
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
Center: Air Pollution Mixtures: Health Effects across Life Stages
Center Director: Koutrakis, Petros
Title: A National Study to Assess Susceptibility, Vulnerability, and Effect Modification of Air Pollution Health Risks
Investigators: Dominici, Francesca , Bell, Michelle L. , Zanobetti, Antonella , Schwartz, Joel
Institution: Harvard University
EPA Project Officer: Chung, Serena
Project Period: January 1, 2011 through December 31, 2015 (Extended to December 31, 2016)
Project Period Covered by this Report: August 1, 2011 through July 31,2012
RFA: Clean Air Research Centers (2009) RFA Text | Recipients Lists
Research Category: Health Effects , Air
This national study is aimed at identifying factors that explain the heterogeneity of health risks associated with air pollution exposure. We hypothesize that such factors include medical and social conditions, conditions that modify exposure, and differences in pollution composition that modify exposure toxicity. Moreover, we hypothesize that the relevant factors vary among different health outcomes. Our research will be fully interactive with the other Center projects. Our previous results (e.g., diabetic susceptibility) have guided their analyses, and their results have generated specific hypotheses that we will test. We have three objectives. In Aims 1 and 2, we will conduct national studies of short-term and long-term exposures to individual pollutants, sources, and mixtures. A main focus of our Center is to study established cohorts (NAS, Framingham, and Viva) in Massachusetts and surrounding states using novel, validated approaches to assess exposure. In Aim 3, we will complement those cohort studies, by establishing a cohort of 2.3 million Medicare enrollees residing in the same region and following its members prospectively for cause-specific hospital admissions and mortality for the period 2000-2014, and also by studying all live births in Eastern Massachusetts, geo-coded to exact address and followed for adverse birth outcomes.
Project 5 has three aims; progress is reported for each aim.
Aim 1. Develop statistical methods and conduct national studies to estimate mortality and hospitalization risks associated with short-term exposures to individual pollutants, source types and air pollution mixtures.
Temporal changes in cardiovascular disease management in the US as a potential effect modifier in our air pollution studies. In a recent paper ("Geographical disparities in hospitalized myocardial infarction incidence and outcomes: does a rising tide lift all boats?" Yeh, Normand, Wang, Barr, and Dominici, published in Circ. Cardiovascular Quality and Outcomes) we have estimated temporal and geographical trends in the incidence and treatment of cardiovascular diseases in the Medicare population for the period 1998 to 2008. This is the period where Medicare data Part A is available to us for 100% of the Medicare population. Now, the Medicare data has been extended to the year 2009 and 2010. We sought to examine: 1) whether there have been significant regional disparities in myocardial infarction (MI) incidence, treatment and outcomes during this time period; and 2) whether these regional disparities have changed over time. In our study population, which includes nearly 300 million Medicare fee-for-service beneficiary-years, we found: 1) statistically significant declines in the incidence of MI in all U.S. Census Divisions; 2) statistically significant differences in the incidence of MI between geographic regions at the start of the study period which increased over time between 2000 and 2008; and 3) strong evidence of a wide and persistent variation in rates of cardiac catheterization and revascularization after MI across regions. To our knowledge, this is the largest study to date that examines geographic differences of trends in acute MI in the U.S. elderly, and the first that examines geographic differences in MI incidence, procedures and 30-day mortality. As always, it will require some serious considerations on how to incorporate the temporal changes in cardiovascular disease management into our models of air pollution exposure (short- and long-term) and mortality and morbidity outcomes. We will continue to investigate this issue in year 3.
Variable selection for multi-pollutant risk estimation. We have submitted a manuscript entitled “Stochastic search variable selection in multi-pollutant models: a study of fine particulate matter components in 118 U.S. counties, 2000-2008” (Barr CD, Diez, DM, Dou Y, Wang Y, Bell M, Dominici F, Peng RD). In this paper, we apply Bayesian stochastic search variable selection (SSVS) to ambient PM2.5 components in multi-pollutant models of emergency hospital admissions for cardiovascular disease (CVD) in a population of Medicare enrollees older than 65 in each of 118 U.S. counties, 2000-2008. Our SSVS method allows us to evaluate a large number of possible statistical models in each county (we consider six of the most common PM2.5 components: sulfate, nitrate, silicon, elemental carbon, organic carbon matter, and sodium ion) to identify, within each county, the combination of PM2.5 components most associated with CVD hospital admissions. Complete results and extensive tools for reproducibility are made freely available online at www.ddiez.com/epissvs (login: map, password: Review65). We also have published a paper by Levy, et al., 2012, "A meta-analysis and multisite time-series analysis of the differential toxicity of major fine particulate matter constituents," American Journal of Epidemiology 175 (11): 1091-1099) where we performed a literature review and conducted a multi-site time-series analysis of hospital admissions and exposure to PM2.5 constituents (elemental carbon, organic carbon matter, sulfate, and nitrate) in a population of 12 million U.S. Medicare enrollees for the period 2000 to 2008. The literature review illustrated a general lack of multi-constituent models or insight about probabilities of differential impacts per unit concentration change. Consistent with previous results, the multi-site time-series analysis found statistically significant associations between short-term changes in elemental carbon and cardiovascular hospital admissions. Posterior probabilities from multi-constituent models provided evidence that some individual constituents were more toxic than others. Ratios of constituent toxicities, commonly used in risk assessment to describe differential toxicity, were extremely uncertain for all comparisons. These analyses emphasize the subtlety of the statistical techniques and epidemiological studies necessary to inform risk assessments of particle constituents.
Adjustment for confounding in time series analyses. In Wang, et. al., 2012, "Bayesian effect estimation accounting for adjustment uncertainty," published in Biometrics, we have developed an innovative Bayesian approach to account for the uncertainty in how to adjust for confounders in time series studies.
Aim 2. Develop statistical methods and conduct national studies to estimate mortality and hospitalization risks associated with long-term exposures to individual pollutants, source types and air pollution mixtures.
Long term exposure to PM2.5 and life expectancy. In Correia, et al. ("The effect of air pollution control on life expectancy in the United States: an analysis of 545 US counties for the period 2000 to 2007," published in Epidemiology (under revision), we investigate whether more recent and slower declines in PM2.5 levels continue to improve life expectancy. We assembled a dataset for 545 U.S. counties consisting of yearly county-specific average PM2.5, yearly county-specific life expectancy, and several potentially confounding variables related to both socioeconomic status and demographic characteristics for the years 2000 and 2007. We used regression models to estimate the adjusted effect of reductions in PM2.5 on changes in life expectancy for the period 2000 to 2007. We found that reductions in PM2.5 were significantly associated with improvements in life expectancy for the period 2000 to 2007. These results demonstrated that air pollution control in the last decade continues to have a positive impact to public health.
Causal inference methods for assessing the public health impact of air quality regulations. In Zigler, et al. ("Estimating causal effects of air quality regulations using principal stratification for spatially-correlated multivariate intermediate outcomes" published in Biostatistics 13:289-302), we have developed a new approach to evaluate the effect of air quality intervention on ambient levels of air pollution and on health outcomes. We apply our method to examine whether the 1990 Clean Air Act Amendments causally affected Medicare mortality through affecting ambient concentrations of particulate matter and ozone. We also published a commentary in JAMA (Dominici F and Mittleman M. "China’s dilemma: reconciling economic growth with environmental protection, 2012;307(19):2100-2102), doi:10.1001/jama.2012.4601) to discuss the paper by Zhang, et al., 2012, a study of the relationship between air pollutants and biomarkers of inflammation and thrombosis in 125 medical students, before, during, and after the Olympics. Zhang, et al., report substantial reductions in the mean concentration of SO2 (-60%), CO (-48%), NO2 (-43%), EC (-36%), PM2.5 (-27%), OC (-22%), and sulfate (-13%). These changes in air quality were accompanied by statistically significant improvements in biomarkers related to platelet adhesion and activation including a 34.0% decrease in sCD62P and a 13.1% decrease in von Willebrand's Factor during the period of the Olympics that returned toward baseline after the air pollution controls were removed.
Aim 3. Conduct two cohort studies in Massachusetts and surrounding states to estimate health risks associated with long-term exposures to individual pollutants, sources, and air pollution mixtures.
We currently are developing the data set and conducting exploratory analyses to estimate the long-term effects of PM2.5 on cardiovascular and lung cancer hospital admissions. In close collaboration with the Exposure Core, we have obtained estimates of ambient exposure to PM2.5 from satellite data and we are now developing statistical modeling for estimating health effects.
Aims 1 and 2: Continue analysis of national short- and long-term exposures to individual pollutants, sources, and mixtures. We are continuing to develop methods for multi-pollutant models (Bobb JF, Dominici F, Peng R. “Reduced Bayesian hierarchical models: estimating health effects of simultaneous exposure to multiple pollutants” Journal of the Royal Statistical Society, Series C, under revision) and we are extending multi-pollutant analyses of metals.
Aim 3: Continue development and analysis of a cohort of Medicare enrollees in Massachusetts and surrounding states for cause-specific hospital admissions and mortality; studying all live births in Eastern Massachusetts, geo-coded to exact address and followed for adverse birth outcomes.
Journal Articles on this Report : 7 Displayed | Download in RIS Format
|Other subproject views:||All 60 publications||59 publications in selected types||All 59 journal articles|
|Other center views:||All 410 publications||347 publications in selected types||All 347 journal articles|
Supplemental Keywords:National studies, air pollution, heterogeneity, vulnerability, susceptibility, Scientific Discipline, Air, air toxics, Health Risk Assessment, Air Pollution Effects, Biochemistry, Biology, ambient air quality, complex mixtures, health effects, sensitive populations, children's health, air pollutants, biological sensitivities, exposure and effects, lung epithelial cells, susceptible populations, chemical composition, neurotoxicity, toxicity, coronary artery disease, cardiopulmonary, cardiotoxicity, environmental effects, mortality, human health
Progress and Final Reports:Original Abstract
Main Center Abstract and Reports:R834798 Air Pollution Mixtures: Health Effects across Life Stages
Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
R834798C001 Relative Toxicity of Air Pollution Mixtures
R834798C002 Cognitive Decline, Cardiovascular Changes, and Biological Aging in Response to Air Pollution
R834798C003 Identifying the Cognitive and Vascular Effects of Air Pollution Sources and Mixtures in the Framingham Offspring and Third Generation Cohorts
R834798C004 Longitudinal Effects of Multiple Pollutants on Child Growth, Blood Pressure and Cognition
R834798C005 A National Study to Assess Susceptibility, Vulnerability, and Effect Modification of Air Pollution Health Risks