中文摘要：

目的研究日均气温对人群死亡的影响。方法利用广州市2003-2007年全人群逐日死亡人数的时间序列资料，结合同期气象资料，采用分布滞后非线性模型分析气温对人群死亡的滞后效应和累积效应。以最适宜温度为参照，分别估计高温和低温情况下平均气温每上升或下降1℃的相对危险度。结果高温影响快速且持续一周，低温影响相对缓慢但持续时间长达14d；低温对于全死因、心血管疾病、呼吸系统疾病以及其他死因的滞后14d累计相对危险度分别为1．012（95％CI：1．000-1．024）、1．025（1．006～1．043）、1．020（1．001—1．037）；高温对于全死因、心血管疾病的滞后14d累计相对危险度分别为1．019（1．004～1．047），1．035（1．004—1．068），对于呼吸系统疾病与其他病因的影响无统计学意义。结论高温和低温均造成人群死亡率上升，但低温影响持续时间更长。

英文摘要：

Objective To assess the effects of daily mean temperature on mortality. Methods Based on the data of daily number of deaths and meteorological data in Guangzhou from 2003 to 2007, the delayed and cumulative effects of temperature on mortality were examined using a distributed lag non-linear model. Results It was found that the effects of higher temperature were strong and immediate, while the effects of lower temperature persisted up to 14 days. For cold effects over lag0-14 days, the overall relative risk （RR） of mortality associated with 1 ℃ below the minimum mortality temperature （MMT） was 1.012（95%CI： 1.000-1.024）, 1.025（1.006-1.043） and 1.020（1.001-1.037） for all-cause mortality, cardiovascular（CVD） and respiratory disease（RD）, respectively. For hot effects over lag0-14 days, the overall RR was 1.019（1.004-1.047）, 1.035（1.005- 1.068） for all-cause mortality and CVD, respectively. But the effects of higher temperature were not statistical significant for RD and all-other mortality. Conclusion Both of higher temperature and lower temperature may have impacts on mortality. Particularly, the effects of lower temperature lagged longer than higher temperature. The results indicate that the public awareness and ability of cold protection should be improved for residents in sub-tropical regions.