Modification of the acute respiratory effects of ambient air pollution by ambient temperature

Abstract

Introduction: Ambient air pollution health effects may not be uniform across ambient temperatures due to differences in pollutant mixtures, physiologic processes, and exposure patterns. We investigated whether risk ratios for air pollutant concentrations and acute respiratory disease varied by temperature in four US cities. Methods: Using data from Atlanta (2002-2008), Birmingham (2002-2008), Dallas (2006-2008) and St. Louis (2002-2007), we estimated temperature-specific associations between five ambient air pollutants (ozone, nitrogen dioxide, fine particulate matter (PM2.5), PM2.5 organic and elemental carbon) and emergency department (ED) visits among people of all ages for asthma (n=425,064 ED visits) and upper respiratory infection (URI; n=1,309,025 ED visits). For each city, daily average temperature was assigned using central monitors and daily pollutant concentrations were estimated by population-weighting concentrations from a fused monitoring-Community Multiscale Air Quality model. In city-specific Poisson generalized linear models we modeled interaction between 3-day moving average temperature and 3-day moving average pollutant concentrations in relation to respiratory ED counts using product terms between pollution and smooth functions of temperature. Results: We generally observed higher estimated risk ratios for air pollution and respiratory ED visits when ambient temperatures were in the 15-25° Celsius range compared to colder or warmer temperatures. For example, the risk ratio for URI for a 2 µg/m3 increase in PM2.5 organic carbon in Atlanta was 1.05 (95%CI=1.03-1.07) at 20° Celsius and 1.00 (95%CI=0.98-1.01) at 10° Celsius. Conclusions: Results suggest that relationships of ambient air pollution and acute respiratory outcomes vary by outdoor temperatures. We speculate that higher risk ratios observed during comfortable temperatures reflect higher exposure to ambient pollution via increased time spent outdoors and/or higher air exchange rates.