腰围和冠心病因果关系的孟德尔随机化研究  

A Mendelian randomization study on the causal relationship between waist circumference and incidence of coronary heart disease

作　　者：刘新辉 李洪凯 李明卓 于媛媛 司书成[1,2] 侯蕾 刘璐 李文超[1,2] 袁同慧 李云霞[1,2] 周宇畅 薛付忠[1,2] LIU Xinhui;LI Hongkai;LI Mingzhuo;YU Yuanyuan;SI Shucheng;HOU Lei;LIU Lu;LI Wenchao;YUAN Tonghui;LI Yunxia;ZHOU Yuchang;XUE Fuzhong(Department of Biostatistics,School of Public Health,Shandong University,Jinan 250012,Shandong,China;Healthcare Big Data Institute,Shandong University,Jinan 250002,Shandong,China)

机构地区：[1]山东大学公共卫生学院生物统计学系,山东济南250012 [2]山东大学健康医疗大数据研究院,山东济南250002

出　　处：《山东大学学报（医学版）》2019年第11期103-109,共7页Journal of Shandong University:Health Sciences

基　　金：国家自然科学基金(81773547)

摘　　要：目的采用两样本孟德尔随机化方法探索腰围与冠状动脉粥样硬化性心脏病发病风险间的因果关系。方法利用大规模全基因组关联研究数据库,选择相互独立且与腰围密切相关的遗传位点作为工具变量,通过Wald比值法、逆方差加权法、MR-Egger回归法以及加权中位数法等两样本孟德尔随机化方法,以比值比(OR值)作为评价指标对腰围与冠心病间的因果关系进行研究。结果本研究利用样本量为232 101关于腰围的欧洲人群数据库,与样本量为86 995关于冠心病的欧洲人群数据库,选择其中的39个SNP作为工具变量。运用上述两样本孟德尔随机化方法得出的因果效应估计值相近,其中IVW法的OR值为1.531(95%CI:1.248~1.877;P<0.001),MR-Egger回归结果表明遗传多效性不会对结果造成偏倚(截距=0.003,P=0.768)。结论两样本孟德尔随机化分析结果显示,腰围与冠心病的发病风险间呈正向因果关系,即腰围每增加一个标准差(SD=12.5 cm)会导致冠心病发病风险增加约50%。Objective To explore the causality between waist circumference(WC) and incidence of coronary heart disease(CHD) using 2-sample mendelian randomization(2 MR). Methods Mutually independent genetic variants which were closely related to WC were chosen as instrumental variable(IV) from Genome Wide Association Study(GWAS) databases. By using odds ratio(OR) as outcome indicator, the causal relationship between WC and CHD was analyzed using 2 MR methods, including Wald ratio method, inverse-variance weighted(IVW) method, mendelian randomization-Egger(MR-Egger) regression and weight median method. Results WC dataset of 232 101 Europeans and CHD dataset of 86 995 Europeans were used. Totally 39 SNPs were selected as IV. Similar effect estimations were obtained using 2 MR methods, of which IVW results showed that the OR between WC and CHD was 1.531(95%CI:1.248-1.877;P<0.001). Besides, results of MR-Egger regression indicated that genetic pleiotropy did not bias the effect estimation(slope=0.003, P=0.768). Conclusion This large study using 2 MR, shows there is a causal relationship between WC and incidence of CHD, and the risk of CHD will increase by approximately 50% when WC increases one standard deviation(SD=12.5 cm).

关 键 词：腰围 冠心病 两样本孟德尔随机化 因果推断 

分 类 号：R541.4[医药卫生—心血管疾病]