Objective: To explore the impact of the socio-economic factors on the temperature-mortality association in different cities in southern China.
Methods: Daily mortality registration data, meteorological data and air pollution data of the cities as Changsha and Kunming during 2006-2009, and cities as Guangzhou and Zhuhai during 2006-2010, were collected to explore modifying effects, stratified by age, gender, education and place of death, of socio-economic factors on the association between temperature and mortality, by distributed lag non-linear model. The accumulative effect of temperature-mortality were separately analyzed in each city, under the high temperature (0-3 days) and low temperature (0-20 days) situation. The association between temperature and mortality was evaluated by general linear threshold model. The above process was firstly adopted to analyze the impact in single city and then Meta analysis was applied to analyze the impact in several cities by effect-combine.
Results: The relationship between temperature and mortality in the four cities showed nonlinearity. The minimum mortality risk was separately 23.5 °C, 20.5 °C, 25.0 °C and 26.0 °C in Changsha, Kunming, Guangzhou and Zhuhai. The results of effect-combine showed that low-temperature (RR = 1.67, 95%CI:1.54-1.80) has a higher gross effect than high-temperature (RR = 1.11, 95%CI:1.01-1.18) on population. With the age increasing, risk of death increased both under high and low temperature situation, and the effect of low temperature was greater (RR = 1.83, 95%CI:1.65-2.04) for the elderly than it of high temperature (RR = 1.17, 95%CI:1.03-1.33). The mortality risk among females (cold and hot effects(95%CI) were 1.75(1.57-1.97) and 1.11(0.99-1.25), respectively)was higher than it among males (cold and hot effects(95%CI) were 1.59(1.45-1.77) and 1.11(1.03-1.19), respectively). Whereas the mortality risk on higher education population was significantly higher (cold and hot effects (95%CI) were 1.89(1.48-2.45)and 1.34(1.19-1.48), respectively) than it on other educated people.
Conclusion: Age, gender, educational level and place of death showed modifying effects on the association between temperature and mortality. The elderly, women and highly educated people were vulnerable to the temperature influence on mortality.