二甲双胍与结直肠癌发病关系的两样本孟德尔随机化研究  

A two-sample Mendelian randomized study on the relationship between metformin and colorectal cancer

作　　者：邱汉波 唐东昕 杨兵 白媛媛 李洲强 杨柱 QIU Han-bo;TANG Dong-xin;YANG Bing;BAI Yuan-yuan;LI Zhou-qiang;YANG Zhu(Guizhou University of Traditional Chinese Medicine,Clinical Medical Research Center of the First Affiliated Hospital of Guizhou University of Traditional Chinese Medicine,Guizhou Cancer inheritance and Science and Technology Innovation Talent Base,Guizhou Cancer Inheritance and Innovation Research Team,Guiyang,Guizhou 550000,China;不详)

机构地区：[1]贵州中医药大学、贵州中医药大学第一附属医院临床医学研究中心、贵州省中医肿瘤传承与科技创新人才基地、贵州省中医肿瘤传承与创新研究团队,贵州贵阳550000 [2]广西中医药大学 [3]柳州市中医院

出　　处：《现代预防医学》2024年第13期2484-2490,共7页Modern Preventive Medicine

基　　金：贵州省高等学校中医药民族医药防治肿瘤医学转化工程研究中心(黔教技[2023]037号);贵州省中医肿瘤传承与科技创新人才基地(黔人领发[2018]3号)。

摘　　要：目的采用两样本孟德尔随机化(MR)分析评价二甲双胍改善结直肠癌是否存在潜在的因果关系。方法从遗传变异全基因关联研究(genome-wide association analyses,GWAS)数据库获取二甲双胍及结直肠癌数据样本及单核苷酸多态性(SNP),以MR Egger回归、IVW(inverse variance weighted)两种方法进行异质性检验,利用MR-Egger回归截距进行多效性分析,最后采用逆方差加权(IVW)作为主要分析方法,MR-Egger回归、加权中位值法(WM)、简单模式法(SM)、加权模式法为补充方法,计算OR值评估两者的潜在因果关系,以散点图、森林图的形式进行可视化。结果异质性检验MR Egger回归(P=0.61)、IVW(P=0.55)表明不存在异质性;多效性检验回归截距(MR Egger intercept=0.02,P=0.14)表明不存在多效性;IVW(OR=0.0050,95%CI:0.0004~0.0703,P<0.001)、MR Egger(OR=5.47×10-6,95%CI:5.05×10-10~0.0592,P=0.01)、加权中位值(OR=0.0034,95%CI:6.68×10-5~0.1697,P=0.0044)、SM(OR=0.0137,95%CI:2.46×10-5~7.5991,P=0.19)、加权模式(OR=0.0003,95%CI:2.26×10-5~0.0367,P=0.002)方法OR值均<1,且IVW、MR-Egger回归、加权中位值、加权模式均P<0.05。结论基于本次MR分析结果表明,二甲双胍能降低结直肠癌发病率。Objective To evaluate the potential causality of metformin in the improvement of colorectal cancer by two-sample Mendelian randomized(MR)analysis.Methods The data of metformin and colorectal cancer and single nucleotide polymorphism(SNP)were obtained from Genome-wide association analyses(GWAS)database.The heterogeneity was tested by MR Egger regression and IVW(Inverse variance weighted),and the pleiotropy was analyzed by MR-Egger regression intercept.Finally,inverse variance weighting(IVW)was used as the main analysis method.MR-Egger regression,weighted median method(WM),simple model method(SM),and weighted model method(WM)were used as supplementary methods to calculate the OR value to evaluate the potential causal relationship between them,and to visualize them in the form of scatter map and forest plot.Results The heterogeneity test MR Egger regression(P=0.61)IVW(P=0.55)showed that there was no heterogeneity,and the regression intercept(MR Egger intercept=0.02,P=0.14)showed no pleiotropy.The OR values of the above five methods were all less than 1:IVW[OR(95%CI)=0.0050(0.0004-0.0703),P<0.001],MR Egger[OR(95%CI)=5.47×10-6(5.05×10-10-0.0592),P=0.01],weighted median[OR(95%CI)=0.0034(6.68×10-5-0.1697),P=0.0044],SM[OR(95%CI)=0.0137(2.46×10-5 to 7.5991),P=0.19],and the OR values of IVW,MR-Egger regression,weighted median and weighted model were all(P<0.05),and the weighted model[OR(95%CI)=2.26×10-5-0.0367,P=0.002].Conclusion Based on the results of this Mendelian randomized analysis,metformin can reduce the incidence of colorectal cancer.

关 键 词：结直肠癌 二甲双胍 孟德尔随机化 

分 类 号：R735.3[医药卫生—肿瘤]