摘要
目的 系统评价长期暴露于细颗粒物(PM2.5)对痴呆发生率的影响.方法 检索PubMed、EMbase、Cochrane Library、Web of Science、中国知网(CNKI)、万方(Wan Fang Data)等数据库,纳入关于长期暴露于PM2.5后痴呆发生率的队列研究,利用NOS量表进行文献质量评价,采用Stata 16.0软件对符合纳入标准的文献进行Meta分析、绘制森林图、漏斗图、检验异质性、敏感性及发表偏倚.结果 共纳入30篇文献,其中以痴呆、阿尔茨海默病(AD)和血管性痴呆(VaD)发病为结局的文献分别有27、16和9篇.Meta分析结果显示,中老年人群中,长期PM2.5暴露(浓度每增加10 μg/m3),增加痴呆发病率(HR=1.32,95%CI:1.20～1.44),包括AD(HR=1.32,95%CI:1.16～1.50)和VaD(HR=1.99,95%CI:1.32～2.98).异质性检验提示所纳入研究之间存在异质性.漏斗图及Egger检验提示存在发表偏倚的可能性极小.敏感性分析提示结果可靠性高.结论 长期暴露于PM2.5可能是痴呆发生的重要危险因素,需加强对以PM2.5为主空气污染的关注与管理.
Abstract
Aim To systematically evaluate the influence of long-term particulate matter 2.5(PM2.5)exposure on dementia.Methods Several prevailing databases,including PubMed,EMbase,Cochrane Library,Web of Science(WOS),CNKI and Wan Fang Data,were comprehensively searched for cohort researches involving the dementia occurrence after long-term exposure to fine particular matter 2.5(PM2.5).The Newcastle Ottawa Scale(NOS)for cohort studies was employed to estimate the quality of the literature.Afterwards,a commercial software Stata 16.0 was utilized to conduct the Meta-analysis based on the selected articles.Both forest plots and funnel plots were generated,while heterogeneity,sensitivity and publication bias of these articles were examined.Results Thirty articles are included in this Meta-analysis totally.Among them,twenty-seven articles related to long-term PM2.5 effects on dementia,sixteen on Alzheimer's Disease(AD)and five on vascular dementia(VaD)were extracted.The analysis results indicated that long term PM2.5 exposure could increase the risk of dementia(HR=1.32,95%CI:1.20-1.44),including two main subtypes,AD(HR=1.32,95%CI:1.16-1.50)and VaD(HR=1.99,95%CI:1.32-2.98).Although some heterogeneity existed in the selected articles,the publication bias remains extremely low,the analysis results possessed high reliability.Conclusion Long-term exposure to PM2.5 could be a clear marker for an elevated risk of dementia.It can be deduced that an urgent concern and control should be attached to the air pollution issue.
NETL
NSTL
万方数据