欧洲人种新型冠状病毒感染与结核病发病风险因果关系:一项双样本孟德尔随机化研究  

Causality between coronavirus disease 2019 and tuberculosis in Europeans:a two-sample Mendelian randomization study

作　　者：王雷 陈迟到 宿连政 李凌玮 王欣淼 王鹏 黄忠浩 Wang Lei;Chen Chidao;Su Lianzheng;Li Lingwei;Wang Xinmiao;Wang Peng;Huang Zhonghao(Department of Medical Imaging Technology,College of Medical Imaging,Qilu Medical University,Shandong Province,Zibo 255000,China)

机构地区：[1]齐鲁医药学院医学影像学院医学影像技术教研室,淄博255000

出　　处：《中国防痨杂志》2025年第5期653-659,共7页Chinese Journal of Antituberculosis

基　　金：山东省艺术科学协会重点课题(L2024Q05100079);山东省教育发展促进会2024年度教育规划课题(JCHKT2024209)。

摘　　要：目的:利用孟德尔随机化分析方法探寻新型冠状病毒感染(coronavirus disease 2019,COVID-19)与结核病之间的因果关系。方法:从IEU Open GWAS数据库(https://gwas.mrcieu.ac.uk/)获取COVID-19 GWAS汇总数据,从UKB数据库(https://www.ukbiobank.ac.uk/)获取结核病GWAS数据。使用5种孟德尔随机化分析方法,逆方差加权法(IVW)为主要方法,加权中位数法、MR-Egger回归、简单众数法、加权众数法作为补充(COVID-19样本:626151例,病例组:3886例,对照组:622265例;结核病样本:462933例,病例组:2277例,对照组:460656例),采用Cochran Q检验、MR-Egger截距、留一法检测多效性和敏感度,最后再做反向孟德尔随机化分析。结果:从GWAS数据库中筛选出14个与COVID-19显著相关的独立的单核苷酸多态性(single nucleotide polymorphism,SNP)作为工具变量,SNP位点编号分别为rs1064213、rs10875713、rs113488799、rs11634857、rs13050728、rs1566837、rs2269899、rs2277732、rs2597569、rs2781267、rs28815269、rs4782434、rs622568、rs787642。逆方差加权分析显示,COVID-19与结核病发病风险没有相关性(OR_(IVW)=1.001,95%CI:0.995~1.002,P=0.527)。MR-Egger回归显示,水平多效性影响结果的可能性小(截距=0.0002435,P=0.101),逆方差加权和MR-Egger回归没有发现异质性(Q=13.205,P=0.432;Q=10.040,P=0.612)。“留一法”分析显示,孟德尔随机化整体估计结果较为可靠,未出现显著偏差,漏斗图未显示不对称。反向孟德尔随机化分析表明,结核病与COVID-19发病风险无因果关系[P(IVW)=0.805]。结论:基于欧洲人群遗传数据的双向孟德尔随机化分析表明,欧洲人群COVID-19与结核病之间可能没有双向因果关系。Objective:To investigate the causality between coronavirus disease 2019(COVID-19)and tuberculosis(TB)using Mendelian randomization analysis.Methods:Summary data on COVID-19 was extracted from the Integrative Epidemiology Unit(IEU)Open Genome-Wide Association Studies(GWAS)pool(https://gwas.mrcieu.ac.uk/).GWAS data on TB was also retrieved from the UK Biobank database(https://www.ukbiobank.ac.uk/).The relationship between COVID-19 and TB was examined using five methods,the major method being inverse variance weighting(IVW),while additional methods included weighted median,MR-Egger regression,simple mode and weighted mode(626151 cases of COVID-19,including 3886 cases and 622265 controls;462933 cases of TB,including 2277 cases and 460656 controls).Cochran’s Q statistic,MR-Egger regression analysis,and the“leave-one-out-method”analysis were used to evaluate heterogeneity and sensitivity,and finally inverse Mendelian randomization was done.Results:Fourteen independent single nucleotide polymorphisms(SNPs)associated with COVID-19 were selected as instrumental variables from GWAS(SNP IDs:rs1064213,rs10875713,rs113488799,rs11634857,rs13050728,rs1566837,rs2269899,rs2277732,rs2597569,rs2781267,rs28815269,rs4782434,rs622568,and rs787642).IVW showed no significant correlation between COVID-19 and TB(IVW odds ratio=1.001,95%confidence interval=0.995-1.002,P=0.527).MR-Egger regression showed that horizontal pleiotropy was unlikely to influence the results(intercept=0.0002435,P=0.101),neither IVW nor MR-Egger regression revealed heterogeneity(Q=13.205,P=0.432;Q=10.040,P=0.612).The“leave-one-out-method”analysis revealed that Mendelian randomization overall estimates were reliable,with no significant bias,and the funnel plot did not exhibit asymmetry.Reverse MR analysis also confirmed the absence of a causality between COVID-19 and TB risk(P(IVW)=0.805).Conclusion:Bidirectional MR analysis based on genetic data of European population revealed no significant bidirectional relationship between COVID-19 and TB in this popul

关 键 词：冠状病毒感染 结核 孟德尔随机化 因果律 

分 类 号：R52[医药卫生—内科学]