气温对山东省四城市卒中死亡影响归因风险评估  

Attributable risk assessment of the effect of temperature on stroke mortality in four cities of Shandong Province

作　　者：何芬芬 赵珂 董意麟 刘超 刘成榕 张珮瑶 刘浩宇 郭晓雷[2] 薛付忠 贾贤杰 HE Fenfen;ZHAO Ke;DONG Yilin;LIU Chao;LIU Chengrong;ZHANG Peiyao;LIU Haoyu;GUO Xiaolei;XUE Fuzhong;JIA Xianjie(Department of Epidemiology and Statistics,School of Public Health,Bengbu Medical University,Bengbu,Anhui 233030,China;不详)

机构地区：[1]蚌埠医科大学公共卫生学院流行病与卫生统计学教研室,安徽蚌埠233030 [2]山东省疾病预防控制中心 [3]山东大学齐鲁医学院公共卫生学院生物统计学系 [4]国家健康医疗大数据研究院

出　　处：《中国预防医学杂志》2024年第12期1543-1549,共7页Chinese Preventive Medicine

基　　金：“512人才培育计划”(by51201202)。

摘　　要：目的探讨山东省中部地区不同温度暴露对卒中及亚型死亡的相对和归因贡献。方法收集山东省中部地区四区(县)2013—2019年逐日卒中死亡、气象和空气污染资料。应用基于quasi-Poisson回归的分布滞后非线性模型定量分析各区县温度对卒中死亡影响的滞后与累积相对风险(relative risks,RRs)。采用随机效应模型结合极大似然估计合并效应值并进行最佳线性无偏预测;定量估算不同温度区间造成的人群归因分数(attributable fraction,AF),并按性别和年龄分层,以确定敏感人群。结果2013—2019年山东省中部地区4区县因卒中死亡30721例,其中缺血性卒中(ischemic stroke,IS)19594例(63.78%),出血性卒中(hemorrhagic stroke,HS)11127例(26.22%)。日均气温与卒中及亚型死亡均呈非线性关系,并存在一定滞后性;以最低死亡风险温度为参照,极冷与极热对HS死亡无显著影响;而极热对卒中和IS死亡的累积效应在Lag 0~7 d达最高值,RR分别为1.54(95%CI:1.24~1.91)和1.87(95%CI:1.46~2.40),尚未发现极冷对两者的影响;在Lag 0~14 d卒中和IS死亡归因于非适宜温度暴露的AF分别为19.78%(95%CI:6.44%~20.20%)和24.56%(95%CI:7.47%~34.95%);中冷造成大部分卒中死亡负担(AF=15.42%,95%eCI:3.19%~23.44%)。分层结果显示,男性和<65岁人群对极冷更敏感,在Lag 0~21 d时RR分别为2.09(95%CI:1.02~4.30)和2.71(95%CI:1.06~6.90),女性和≥65岁人群对极热更敏感,在Lag 0~7 d时RR分别为2.28(95%CI:1.55~3.63)和1.64(95%CI:1.27~2.11);与极冷相比,中冷对男性影响更大,在Lag 0~14 d时AF为24.98%(95%CI:14.62%~31.92%)。结论低温和高温均可增加卒中死亡负担,但两者滞后模式不同;与其他温度范围相比,中等低温效应的可归因比例最高,男性更易受到低温影响,而高温对老年人影响更大,应予以适时重点预防。Objective To investigate the relative and attributable contributions of different temperature exposuresto stroke and its subtype mortality in the central region of Shandong Province,China.Methods Daily stroke mortality,meteorological,and air pollutants data from four districts/counties in the central region of Shandong Province from 2013 to 2019 were collected.A distributional lag non-linear quasi-Poisson regression model was applied to estimate the lagged and cumulative relative risks(RRs)of temperature on stroke mortality across each district/county.A random-effects meta-analysis combined with maximum likelihood estimation was used to pool county-specific effects and perform the best linear unbiased prediction.The attributable fractions(AF)were further quantitatively estimated,and subgroup analyses stratified by gender and age were conducted to identify sensitive populations.Results From 2013 to 2019,there were 30721 deaths due to stroke in the four districts/counties,of which ischemic(IS)and hemorrhagic stroke(HS)accounted for 19594(63.78%)and 11127(26.22%),respectively.A non-linear relationship between daily average temperature and stroke/subtypes mortality,with significant lag effects.Compared to the temperature of minimum mortality risk,extreme cold and heat had no significant impact on HS mortality,while extreme heat had the highest cumulative effects on stroke and IS mortality at 0-7 d,with RRs of 1.54(95%CI:1.24-1.91)and 1.87(95%CI:1.46-2.40),respectively.The AF of non-optimal temperature exposure on stroke and IS mortality at Lag 0-14 days were 19.78%(95%CI:6.44%-20.20%)and 24.56%(95%CI:7.47%-34.95%),respectively,with moderate cold contributing the most to stroke mortality burden(AF=15.42%,95%CI:3.19%-23.44%).Stratified results showed that males and those aged<65 years old were more sensitive to extreme cold,with RRs of 2.09(95%CI:1.02-4.30)and 2.71(95%CI:1.06-6.90)at lag 0-21 d,respectively,and females and those≥65 years old were more sensitive to extreme heat,with RRs of 2.28(95%CI:1.55-3.63)and 1.6

关 键 词：温度 缺血性卒中 出血性卒中 死亡率 人群归因分数 

分 类 号：R122[医药卫生—环境卫生学]