2012 Progress Report: Novel Estimates of Pollutant Mixtures and Pediatric Health in Two Birth CohortsEPA Grant Number: R834799C003
Subproject: this is subproject number 003 , established and managed by the Center Director under grant R834799
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
Center: The Southeastern Center for Air Pollution and Epidemiology: Multiscale Measurements and Modeling of Mixtures
Center Director: Tolbert, Paige
Title: Novel Estimates of Pollutant Mixtures and Pediatric Health in Two Birth Cohorts
Investigators: Strickland, Matthew J , Chang, Howard , Darrow, Lyndsey , Davis, Robert , Guensler, Randy , Klein, Mitchel , Liu, Yang , Mulholland, James , Russell, Armistead G. , Waller, Lance
Current Investigators: Strickland, Matthew J , Chang, Howard , Darrow, Lyndsey , Guensler, Randy , Klein, Mitchel , Liu, Yang , Mulholland, James , Russell, Armistead G. , Waller, Lance
Institution: Emory University , Georgia Institute of Technology
Current Institution: Emory University , Georgia Institute of Technology , University of Nevada - Reno
EPA Project Officer: Chung, Serena
Project Period: January 1, 2011 through 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
In utero and early life experiences affect physiological development and can influence sensitivity to environmental factors throughout life. In this Project we explore the interplay between certain early life events, characterizations of air pollutant mixtures developed as part of the Center’s Mixtures Characterization Toolkit, and a range of pediatric health outcomes using two large, population-based birth cohorts. One cohort consists of roughly 1.7 million Georgia birth records that have been geocoded to the Census block level and linked with pediatric emergency department visits by staff at the Georgia Department of Human Resources. Using this statewide birth cohort, we are investigating acute effects of air pollution mixtures on respiratory health outcomes and ear infections in children, and we are assessing whether children who were born premature or low birth weight are more sensitive to ambient air pollutant concentrations than their counterparts. Further, we are using the statewide birth cohort to investigate whether ambient air pollutant mixtures during pregnancy are associated with the risk of preterm delivery or reduced birth weight. The second birth cohort is comprised of children who were members of the Kaiser Permanente Georgia Health Maintenance Organization in metropolitan Atlanta. In this birth cohort, where comprehensive medical and residential histories are available for each study subject, we will examine whether air pollutant mixtures during the first year of life are associated with the incidence of childhood asthma.
We have 1,705,130 individual-level birth records from 1994-2006 for Georgia. We have 8,252,996 individual-level emergency department visits among children age 0-18 years during 2002-2010 for Georgia and we have 2,458,950 individual-level hospital records for children age 0-18 years during 1999-2010 for Georgia. We have linked the birth record data with Census 2000 data to bring in a variety of socioeconomic variable that can be used in the analyses. The birth data and the ED/hospitalization data are linked via a unique longitudinal ID, and we have devoted significant effort to examine the quality of this linkage. Although the ID should be unique (it is based on letters from the first name, last name, birth date, and sex), for 18,921 (1.1%) of the birth records this longitudinal ID is not unique. For 6,880 records this is because twins had similar first names. The remaining 12,041 instances appear to be unrelated individuals born on the same day with similar names. We then linked these data with the ED data based on longitudinal ID. When an ED record linked with multiple birth records we used ZIP code to determine which record to link to; if the zip code was the same (as in the case for twins) then we randomly selected which twin to assign that ED visit to. In total 3,700,003 ED records link with a birth record. We evaluated whether there were inconsistencies in the longitudinal ID that might have caused a problem with the linkage (e.g., a missing apostrophe, such as ONeill instead of O’Neill, would cause all of the characters to be shifted to the right one spot) but did not find any major problems. There were several potential links that were suggestive but not definitive (e.g., the longitudinal ID matched on all characters save for one), but ultimately we chose not to create many new linkages based on longitudinal IDs that did not match.
Significant progress has also been made on the development of satellite-derived estimates of PM2.5 at 10 km resolution in Georgia, with calibration to the fixed site monitors using a Bayesian statistical model. Multiple years of satellite remote sensing data, gridded meteorological and land use variables have been collected and processed as input data to the Bayesian model. Work is ongoing to estimate biomass burning events using high resolution remote sensing data from multiple NASA satellites. Through collaboration with Georgia EPD and the Tall Timbers Research Station in Florida, we have collected ground measurements of prescribed burn areas in southwestern and southern Georgia. Customized satellite data processing procedures are being developed to better extract burn scar signals.
Analytic datasets have been prepared, and an abstract on the association between outdoor pollutant concentrations (population-weighted average from fixed site monitors) has been accepted for the August 2012 International Society for Environmental Epidemiology conference (Gass et al., 2012). We have also been working on classification and regression tree approaches for examining complex joint effects of mixtures, which was presented at the annual CLARC meeting and has also been accepted as a presentation at the October 2012 International Society for Exposure Science conference (Gass et al., 2012). Work on the effects of measurement error on the risk ratio estimates from time-series studies has been ongoing, with a new publication appearing in Atmospheric Environment (Goldman et al., 2012) and a follow-up paper that is in draft form that will be submitted later this summer.
The subcontract with Kaiser Permanente begins in 2013, and this fall we will begin working with them on development of that birth cohort. Various epidemiologic analyses of the statewide data will be conducted, with some analyses being more or less “conventional” (e.g., single pollutant and two-pollutant models) and others focusing on innovative approaches for characterizing mixtures (e.g., regression trees and/or cluster analyses). Continued work on measurement error in time-series studies will be conducted, with a view towards extending our current work (which focused on pollutants one-at-a-time) to the situation where there are two pollutants in the model.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
|Other subproject views:||All 91 publications||39 publications in selected types||All 37 journal articles|
|Other center views:||All 334 publications||136 publications in selected types||All 132 journal articles|
Supplemental Keywords:ambient air, atmosphere, health effects, human health, susceptibility, vulnerability, sensitive populations, infants, children, risk, dose-response, cumulative effects, epidemiology, exposure, public policy, air quality modeling, monitoring, measurement methods, aerosol, particulates, PM2.5, organics, elemental carbon, metals, ozone, oxidants, PAH, sulfates, source characterization, mobile sources, Georgia, GA, Southeast, Health, Scientific Discipline, Health Risk Assessment, Risk Assessments, Biochemistry, Environmental Monitoring, Atmospheric Sciences, particulate matter, children's health, ambient air monitoring, climate change, air pollution, susceptibility, ambient particle health effects, airshed modeling, human health risk
Progress and Final Reports:Original Abstract
Main Center Abstract and Reports:R834799 The Southeastern Center for Air Pollution and Epidemiology: Multiscale Measurements and Modeling of Mixtures
Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
R834799C001 Development and Deployment of an Instrumentation Suite for Comprehensive Air Quality Characterization Including Aerosol ROS
R834799C002 Examining In-Vehicle Pollution and Oxidative Stress in a Cohort of Daily Commuters
R834799C003 Novel Estimates of Pollutant Mixtures and Pediatric Health in Two Birth Cohorts
R834799C004 A Multi-City Time-Series Study of Pollutant Mixtures and Acute Morbidity