Concentration Response Curve for Ozone related Mortality at High Concentrations

Abstract

Background Rising temperatures and decreased global circulation in the upcoming decades are expected to have a detrimental impact on air quality, particularly with respect to ozone. The largest impacts are anticipated to occur in the form of the frequency and duration of summertime regional pollution episodes. These events are likely to induce nonlinear increases in ozone levels. Objective We examine concentration-response curve for ozone related mortality to evaluate the sensitivity of the health burden with respect to high end of ozone distribution. Methods A flexible Bayesian hierarchical model is developed to allow for the nonlinear ozone risk curve with a shape parameter controlling the prior belief about the monotonicity of the risk. We use polynomial spline basis functions and their derivatives to evaluate relative at different quantiles of the ozone distribution. The model was applied to the data from the US National Morbidity, Mortality, and Air Pollution Study for 95 major US urban centers between 1987 and 2000. Results were examined with respect to those obtained under the assumption of linear effect of ozone without positivity constraint (Bell et al, JAMA 2004). Results Under the assumption of linear effect of ozone relative risk is 0.12+-0.03%. Under moderate and strong monotonicity priors we estimate average log relative rate as 0.25+-0.06% and 0.56+-0.04% per 10ppb, respectively, across the interquartile range of maximum hourly Ozone concentrations. Above the 95% percentile of the ozone distribution these estimates double (0.57+-0.09% and 1.2+-0.05%). Conclusion Our preliminary results indicate that concentration-response relationship is not linear across all levels of ozone. Air pollution episodes are likely to impact high end of ozone distribution and thus contribute larger health burden per unit increase of ozone. Disclaimer: This abstract does not necessarily reflect U.S. EPA policy.