[Association between temperature and daily mortality in Guangzhou, 2006-2009: a time-series study]

To study the association between temperature and daily mortality from June 1, 2006 to December 31, 2009 in Guangzhou.

Time series approach was used to estimate the impact of temperature on the rates of total and cause-specific daily mortality. We fitted generalized additive Poisson regression using non-parametric smooth functions to control for the long-term time trend, day of week, air pollution and other weather variables.

A slight sloping U-like relationship between the total mortality and temperature was found, with an optimum average temperature (temperature with lowest mortality risk) value of 19.7°C in Guangzhou. For temperature above the optimum value, the relative risk of total mortality increased by 3.0% (RR=1.030, 95%CI: 1.011-1.050) for each increase of degree in Celsius. For average temperature below the optimum value, the relative risk of total mortality and diseases of circulatory system had a 3.3% (RR=0.967, 95%CI: 0.936-0.997) decrease and a 3.6% (RR=0.964, 95%CI: 0.935-0.994) increase, for each degree of Celsius increase, respectively.

Our findings showed that the temperature had an impact on the daily mortality in Guangzhou. Countermeasures needed to be taken to reduce the temperature related mortality.