Time-Series Study on the Association between Daily Temperature Variation and Mortality in Guangzhou from 2011 to 2016

Objectives To assess the effect of average daily temperature variation on mortality.

Methods The delayed and cumulative effect of temperature change on mortality were examined by using a distributed lag non-linear model based on the data of daily death toll and the meteorological data in Guangzhou from 2011 to 2016. The extreme cold and thermal effect was estimated by the P_2.5 and P_97.5 of the temperature in Guangzhou. Take the optimum temperature as a reference, the relative risk was estimated by the average temperature going up or down each one degree Celsius (1℃).

Results The effect of higher temperature was strong and immediate, which could last four days, and the effect of lower temperature could last up to 14 days. For cold and thermal effect, the cumulative relative risk (RR) of mortality associated with 1℃ below and over the minimum mortality temperature (MMT) at 0~21 days lagged behind was 1.013 (95%CI:1.003~1.027) and 1.023(95%CI:1.006~1.050) respectively for all causes of death. The effect of higher temperature and lower temperature was statistically significant for male over 65 year-old; and the effect of lower temperature was statistically significant for females at 21 days lagged behind.

Conclusions Both higher temperature and lower temperature might have impact on mortality. Particularly, the effect of lower temperature lagged longer than higher temperature. The result indicated that the public awareness and the ability of cold protection should be improved for residents in sub-tropical region. Elder people were more sensitive to the effect of lower temperature and higher temperature, and female was more sensitive than male to the effect of lower temperature.