High ambient temperature and mortality in California: exploring the roles of age, disease, and mortality displacement.

Investigators have consistently demonstrated associations between elevated temperatures and mortality worldwide. Few have recently focused on identifying vulnerable subgroups, and far fewer have determined whether at least some of the observed effect may be a manifestation of mortality displacement. We examined mean daily apparent temperature and mortality in 13 counties in California during the warm season from 1999 to 2006 to identify age and disease subgroups that are at increased risk, and to evaluate the potential effect of mortality displacement. The time-series method using the Poisson regression was applied for data analysis for single lag days of 0-20 days, and for cumulative average lag days of five and ten days. Significant associations were observed for the same-day (excess risk=4.3% per 5.6 °C increase in apparent temperature, 95% confidence interval: 3.4, 5.2) continuing up to a maximum of three days following apparent temperature exposure for non-accidental mortality. Similar risks were found for mortality from cardiovascular diseases, respiratory diseases, and among children zero to 18 years of age, and adults and the elderly 50 years and older. Since no significant negative effects were observed in the following single or cumulative days, evidence of mortality displacement was not found. Thus, the effect of temperature on mortality appears to be an event that occurs within three days following exposure, and requires immediate attention for prevention.