摘要
[目的]分析上海市卢湾区大气污染急性暴露对居民每日死亡数的影响。[方法]分别采用时间序列的半参数广义相加模型(GAM)和广义线性模型(GLM),在控制死亡的长期趋势、气象因素、“星期几效应”等混杂因素的基础上,分析上海市卢湾区2001年1月1日-2004年12月31日大气污染与居民每日死亡数的关系。[结果]在GAM中大气可吸入颗粒物(PM10),SO2和NO2,48h平均浓度每增加10μg/m^3,居民死亡的相对危险度分别为1.0003(95%CI 1.0000-1.0007)、1.0009(95%C10.9998-1.0019)和1.0012(95%CI 1.0003-1.0021);在GLM中大气PM10,SO2和NO2,48h平均浓度每增加10μg/m^3,居民死亡的相对危险度分别为1.0004(95%CI 1.0001~1.0008)、1.0008(95%CI 0.9998-1.0019)和1.0012(95%CI 1.0002~1.0022)。[结论]GAM、GLM拟和结果相似,上海市卢湾区目前的PM10和NO2水平对居民日死亡数有影响。
[Objective] To assess the association between air pollution and daily mortality in Luwan District, Shanghai. [Methods] All death records of the deceased who lived in Luwan district from January 1,2001 to December 31,2004 were collected. A time-series approach (GAM and GLM ) was used to study the acute effects of air pollution on daily mortality after controlling for long-term trends, weather variables, and day of the week. [Results] An increase of 10μg/ m^3 of PM10,SO2 and NO2 corresponded to 1. 0003 ( 95% CI 1. 0000-1. 0007 ), 1. 0009 ( 95% CI 0. 9998-1. 0019 ) and 1. 0012 ( 95% CI 1. 0003-1. 0021 ) relative risk of daily mortality in GAM and 1. 0004 (95% CI 1. 0000-1. 0008), 1. 0008( 95% CI 0.9998-1. 0019) and 1.0012(95% CI 1.0002-1.0022)in GLM, respectively in Shanghai. [Conclusion]Two analysis models have similar result that air pollution has an effect on mortality.
出处
《上海预防医学》
CAS
2006年第6期264-266,共3页
Shanghai Journal of Preventive Medicine
关键词
大气污染
死亡率
时间序列
广义相加模型
广义线性模型
Air pollution
Mortality
Time-series
Generalized additive model
Generalized linear mode
作者简介
俞冰(1980-)女,医师。在读硕士。