应用Meta分析研究中国不同区域PM2.5污染与人群非意外总死亡率的关系

大部分研究在进行空气污染健康危险度评价工作时,较少考虑中国空气污染的区域差异性。通过搜索多个数据库,以"PM2.5"、"死亡"等为关键词,收集了2003-2017年发表的所有符合研究设定要求的文献,筛选出15篇文献中的21组数据,使用Stata 12.0软件进行Meta分析,利用固定或随机效应模型合并效应值,并对结果进行敏感性分析和发表性偏倚检验。结果显示,PM2.5的日均质量浓度每上升10μg/m3,京津冀、长三角、珠三角以及中国居民每日非意外总死亡率分别上升0.433%(95%置信区间(CI):0.317%~0.550%),0.344%(95%CI:0.195%~0.493%),1.347%(95%CI:1.000%~1.695%)和0.730%(95%CI:0.575%~0.886%)。PM2.5日均质量浓度对人群死亡健康效应的影响不是简单线性关系,具有较大的区域性差异特点。因此,在进行地区性大气污染相关的健康危险度评价工作时,要充分考虑地区差异性问题。

Meta-analysis of the association between PM2.5 pollution and total non-accidental mortality of residents in different regions in China

Regional differences in China were rarely taken into account in the assessment of the health risk of air pollution in most studies.Multiple databases were searched and conducted using keywords"PM2.5","death"and so on.All the articles published in 2003-2017 that met the research setting requirements were collected,15 studies and 21 sets of data were included.Additional residents of total non-accidental mortality in association with a 10μg/m3 increase of daily average PM2.5 were included in the study to perform Meta-analysis by Stata 12.0 software.A fixed or random effect model was used to combine effect values,the sensitivity analysis and adjust publication were also performed.In terms of short-terms effects,the combined excess risk of total non-accidental mortality of BeijingTianjin-Hebei,Yangtze River Delta,Pearl River Delta and China were 0.433%(95% CI:0.317%-0.550%),0.344%(95% CI:0.195%-0.493%),1.347%(95% CI:1.000%-1.695%)and 0.730%(95% CI:0.575%-0.886%).This study further confirmed the relationship between the crowd death health effects and concentrations of PM2.5 was not a simple linear relation,so the air pollution concentration-response parameters was not the same because of the regional differences.Therefore,it was necessary to fully consider the regional differences in the assessment of health risks associated with regional air pollution.