Ozone Modifies the Association Between Temperature and Mortality in 12 U.S. Cities, 1987–2000

TAA3-O-06 Introduction: A few studies have shown that temperature is associated with health outcomes after adjustment for the confounding effects of air pollution. However, little information is available on whether air pollution modifies the temperature-disease relationship. This study examines whether ozone modifies the association between 3-day moving averages of maximum temperature and total nonexternal mortality in 12 U.S. cities using the National Morbidity, Mortality and Air Pollution Study (NMMAPS) data. Methods: We selected the 6 largest cities from the northeast and southeast U.S. regions, respectively, and then used 3 Poisson-generalized additive models to examine whether ozone modified temperature-mortality associations after adjustment for covariates, such as seasonality, day of the week, and dew point temperature. We first used an independent nonparametric model to characterize the temperature-mortality pattern for each of these cities. Then a bivariate response surface model was applied to explore joint effects of temperature and ozone on mortality. Finally, we used a stratification model to quantitatively examine the effect modification of ozone in different regions categorising ozone into different levels. Results: The independent model shows that the temperature-mortality associations were heterogeneous across the regions. In the northeast region, J- or U-shaped temperature-mortality patterns were usually observed, ie, negative association in the cold season, and positive in the hot season. However, in the southeast areas, the temperature-mortality associations were approximately linear, and generally negative through the whole year except for Dallas/Fort Worth. The joint effects of ozone and temperature illustrated that ozone modified the temperature-mortality associations in the northeast region but not obviously in the southeast region. The stratification model indicated that the higher the ozone concentrations, the stronger the temperature-mortality associations in the northeast region. This pattern was not observed in the southeast areas. Conclusion: Ozone modified temperature-mortality associations in the northeast U.S. region, but not in the southeast region. It is important to evaluate the modifying role of air pollution in the assessment of temperature-related health impacts.