The Association Between Short-Term Exposure to Ozone and Risk of Mortality in the United States

• Activity Number: 116
• Type: Invited
• Date/Time: Monday, August 4, 2008 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Michelle L. Bell*+ and Roger D. Peng and Aidan McDermott and Scott L. Zeger and Jonathan M. Samet and Francesca Dominici
• Companies: Yale University and Johns Hopkins Bloomberg School of Public Health and Johns Hopkins Bloomberg School of Public Health and Johns Hopkins Bloomberg School of Public Health and Johns Hopkins Bloomberg School of Public Health and Johns Hopkins University
• Keywords: air pollution ; human health ; Bayesian modeling ; epidemiology ; mortality ; ozone

Abstract

We explored effects of ozone on mortality risk for 95 U.S. urban communities for the period 1987--2000. We applied Bayesian hierarchical models that first estimated associations within each community with single and distributed lag Poisson models, adjusted for weather, day of the week, and seasonality and long-term trend. In a second stage we combined community-specific estimates to generate national effects, accounting for the within-county and between-county statistical variances. Threshold effects were evaluated by estimating the nonlinear exposure-response in multiple model forms. A 10 ppb increase in the previous week's ozone was associated with a 0.52% (95% posterior interval 0.27-0.77%) in mortality risk. Community-specific and national estimates were robust to adjustment by PM10 and PM2.5. Results from threshold models consistently show that effects persist at low ozone levels.