茶摄入量与结直肠癌发病因果关系的孟德尔随机化研究  

Mendelian randomization study of the causal relationship between tea intake and colorectal cancer

作　　者：周应舒 李鑫磊[1] 李汶龙 刘培慧[2] 刘菲[3] ZHOU Ying-shu;LI Xin-lei;LI Wen-long;LIU Pei-hui;LIU Fei(The Second Department of General Surgery,Daqing People’s Hospital of Heilongjiang Province,Daqing,Heilongjiang 163711,China;不详)

机构地区：[1]黑龙江省大庆市人民医院普外二科,黑龙江大庆163711 [2]辽宁省葫芦岛市中心医院神经介入科 [3]中国医科大学基础研究院免疫学教研室

出　　处：《现代预防医学》2024年第4期613-618,639,共7页Modern Preventive Medicine

基　　金：辽宁省教育厅科学研究课题(QN2019038)。

摘　　要：目的 利用两样本孟德尔随机化研究茶摄入量与结直肠癌发病的因果关系。方法 利用IEU OPEN GWAS公共数据库中的结直肠癌全基因组关联分析数据(GWAS ID:ieu-b-496 5)和茶摄入量全基因组关联分析数据(GWAS ID:ukb-b-606 6),汇总两个数据集,筛选与茶摄入量密切相关且独立的单核苷酸多态性(SNPs)作为工具变量(IVs),通过采用两样本孟德尔随机化中的MR-Egger、逆方差加权(IVW)、加权中位数法(Weighted Median)、简单模型(Simple Mode)和加权模型(Weighted Mode)五种方法进行茶摄入量与结直肠癌发病的因果分析的判断,采用OR以及95%CI判断茶摄入量与结直肠癌发病是否存在因果关系,利用MR-PRESSO进行离群值检验,采用IVW(Cochran Q检验)以及MR-Egger(Rücker Q检验)进行异质性分析,采用Egger-intercept进行多效性检验,通过留一法进行敏感性分析,采用漏斗图分析是否存在潜在偏倚。结果 最终纳入39个SNPs, MR-PRESSO未发现离群SNPs,P=0.237>0.05;在因果关系的判定上:IVW(OR=0.996,95%CI:0.990~1.002,P=0.293),MR-Egger(OR=0.992,95%CI:0.977~1.007,P=0.293),Weighted Median(OR=0.997,95%CI:0.987~1.007,P=0.479),Simple Mode(OR=0.996,95%CI:0.976~1.017,P=0.724),Weighted Mode(OR=0.996,95%CI:0.985~1.007,P=0.473),上述结果显示:茶摄入量与结直肠癌发病无因果关系,在异质性检测中Cochran Q检验中P=0.181>0.05,Rücker Q检验中P=0.167>0.05,无显著异质性,Egger-intercept进行多效性检验P=0.489>0.05未见多效性,留一法逐个剔除,未见明显影响结果的SNPs,漏斗图基本对称,显示主结果可靠。结论 茶摄入量与结直肠癌发病无因果关系。Objective To study the causal relationship between tea intake and the incidence of colorectal cancer using two-sample Mendelian randomization.Methods Using the colorectal cancer genome-wide association analysis data(GWAS ID:ieu-b-4965)and tea intake genome-wide association analysis data(GWAS ID:ukb-b-6066)in the IEU OPEN GWAS public database,we summarized and screened the two data sets for single nucleotide polymorphisms(SNPs)which were closely related to and independent of tea intake as instrumental variables(IVs).By using MR-Egger and inverse variance weighting in two-sample Mendelian randomization,Five methods(IVW),weighted median method(Weighted Median),simple model(Simple Mode)and weighted model(Weighted Mode)were used to determine the causal analysis between tea intake and the incidence of colorectal cancer.OR and 95%CI were used to examine the causal relationship between tea intake and the incidence of colorectal cancer.MR-PRESSO was used for outlier testing.IVW(Cochran Q test)and MR-Egger(Rücker Q test)were used for heterogeneity analysis.Egger-intercept was used for pleiotropy test,sensitivity analysis was performed by leave-one-out method,and funnel plot was used to analyze the presence of potential bias.Results 39 SNPs were finally included,and MR-PRESSO found no outlier SNPs,P=0.237>0.05;in terms of causality determination:IVW(OR=0.996,95%CI:0.990-1.002,P=0.293),MR-Egger(OR=0.992,95%CI:0.977-1.007,P=0.293),Weighted Median(OR=0.997,95%CI:0.987-1.007,P=0.479),Simple Mode(OR=0.996,95%CI:0.976-1.017,P=0.724),Weighted Mode(OR=0.996,95%CI:0.985-1.007,P=0.473).The above results showed that there was no causal relationship between tea intake and the incidence of colorectal cancer.In the heterogeneity test,P=0.181>0.05 in Cochran Q test and P=0.167>0.05 in Rucker Q test.There was no significant heterogeneity.The Egger-intercept test showed that P=0.489>0.05 and was not pleiotropic.The leave-one-out method was removed one by one,and no SNPs significantly affecting the results were found.The funnel plot was bas

关 键 词：孟德尔随机化 茶摄入量 结直肠癌 因果关系 预后 

分 类 号：R735.34[医药卫生—肿瘤]