ABSTRACT
Objectives To analyze the features on temperature and mortality of Changsha in 2009-2019, and to explore the association between temperatures variation between neighboring days (TVN) and mortality by using time-series analysis. Methods A Poisson generalized linear regression model combined with a distributed lag non-linear model was used to analyse the association between TVN and mortality. Results A total of 404 328 deaths were studied in Changsha during 2010-2019，the proportion of people aged over 65 years, males respiratory disease, and cardiovascular disease were 74.1761%, 58.9842%, 11.11% and 54.4671%, respectively. During the 3652-day study period, the daily mean maximum and minimum temperature were 35.8℃ -2.8℃. The TVN varied from -12.30℃ to 10.8℃,and a significant correlation was found between TVN and mortality risk, with 1.12% (RR=1.0112，95% CI：1.0061~1.0164) mortality risk increased for 1℃ rise in TVN, and the greatest effect of TVN on mortality was at 4 days lag. Based on age, gender and group study For cardiovascular disease and respiratory disease，a 1℃ increased in TVN were associated with 2.97% and 1.52% death risk increase respectively. The effect appeared on the first day after exposure and lasted for 7 days, the maximum affection came on the fourth day. According to the analysis on age, gender and death-cause, the elderly man over 65 years old, respiratory disease people were more vulnerable to the temperature change between day by day. Conclusion This study provides a comprehensive picture of the non-linear associations between temperature variation and mortality, and there is a certain lag effect. The findings on vulnerability characteristics can help improve clinical and public health practices to reduce disease burden associated with current and future abnormal weather.