Multipollutant Mixtures

This report is Part 3 of HEI Research Report 183, Development of Statistical Methods for Multipollutant Research. It describes a study to develop and apply statistical methods to analyze the effects of multipollutant exposures on health, expanding beyond the two-pollutant approaches used in many studies to date. HEI funded three innovative studies in recent years to improve the tools for analyzing complex multipollutant exposures. In this last report from these studies, John Molitor and colleagues describe a Bayesian framework to identify spatial clusters of air pollution exposures — and other covariates such as socioeconomic status — and estimated pregnancy outcomes associated with those clusters, using a data set for Los Angeles county. 

Contents: HEI's Strategic Plan for 2015–2020 Now Under Way;  Greenbaum Chairs NRC Review of U.S. Climate, Health Assessment; O'Keefe Closes Out Successful Run on Key Air Advisory Committee; Communicating Results - HEI Science Highlighted at EPA Workshops HEI in the News - Wide Audience for ACES Findings; Novel Statistical Methods for Studying Pollutant Mixtures; Journal Spotlights NPACT Studies; Eyes on the Future - HEI Sponsors Meet with Research Committee to Discuss the Path Forward;  Walter A. Rosenblith New Investigator Award Announced

This report describes a study to investigate the relationships among daily variations in air pollution in Ho Chi Minh City (HCMC), Vietnam, hospital admissions for acute lower respiratory infections in children under age 5, and poverty. The study was part of HEI's Public Health and Air Pollution in Asia (PAPA) program and is the first study of air quality and health to be performed in Vietnam. The team of investigators, led by Drs. Le Truong Giang, Long Ngo, and Sumi Mehta, collected daily pollutant data for PM₁₀, sulfur dioxide, nitrogen dioxide, and ozone at multiple locations throughout the city and obtained admissions data from the two pediatric hospitals in HCMC. They then performed statistical analysis on the data using Poisson time-series and case–crossover methods.