Characterizing temperature and mortality in nine California counties

Background: Elevated temperature has been associated with increased mortality. Few epidemiologic studies, however, have considered air pollutants as potential confounders or effect modifiers. None has focused on California, where the climate is generally mild and pollution levels tend to be high-an ideal setting to examine the independent effect of temperature from air pollution. Methods: We examined the association between mean daily apparent temperature and nonaccidental mortality in 9 counties throughout California from May to September 1999-2003. Data were obtained from the National Climatic Data Center (temperature and relative humidity), the California Department of Health Services (mortality), and the California Air Resources Board (particulate matter, ozone, carbon monoxide, and nitrogen dioxide). We conducted a time-stratified case-crossover study, with a time-series analysis as a sensitivity analysis, adjusting for day of the week using both methods and adjusting for time trend in the time-series analysis. We first obtained county-specific estimates and then combined them using meta-analytic methods. Results: A total of 248,019 deaths were included. Each 10 degrees (Fahrenheit) increase in same-day mean apparent temperature corresponded to a 2.3% increase in mortality (95% confidence interval = 1.0%-3.6%) in the case-crossover analysis for all 9 counties combined, with nearly identical results produced from the time-series analysis. No air pollutant examined was found to be a significant confounder or effect modifier. Conclusions: Even without extremes in apparent temperature, we observed an association between temperature and mortality in California that was independent of air pollution.