基于孟德尔随机化的血脂六项与冠心病因果关系研究

目的 探究血脂六项指标（甘油三酯、总胆固醇、高密度脂蛋白、低密度脂蛋白、载脂蛋白A、载脂蛋白B）与冠心病之间的因果关系,为血脂六项指标与冠心病发病风险关联提供遗传学证据支持。方法 研究采用MR-Egger回归、加权中位数法（weight median estimator, WME）、随机效应逆方差加权法（inverse-variance weighted, IVW）、加权模型4种回归模型对血脂六项指标与冠心病的因果关系进行孟德尔随机化分析。结果 IVW模型显示甘油三酯（OR=1.402）、总胆固醇（OR=1.414）、低密度脂蛋白（OR=1.867）、载脂蛋白B（OR=1.738）为冠心病的危险因素,高密度脂蛋白（OR=0.791）为冠心病的保护因素,IVW模型则显示载脂蛋白A为有统计学意义的影响因素;MR-Egger回归截距项P值均>0.05,即筛选出的SNP不存在基因多效性,因此孟德尔随机化在本研究中为因果推断的有效方法;异质性检验结果P值均<0.001,故应重点关注随机效应IVW模型。结论 利用两样本孟德尔随机化方法排除混杂因素和反向因果关联后,得到无偏估计的结果,据此可以确定甘油三酯、总胆固醇、高密度脂蛋白、低密度脂蛋白、载脂蛋白B与冠心病之间存在因果关系。

Causality between six items of blood lipids and coronary heartdisease based on Mendelian randomization

Objective To explore the causal relationship between the six blood lipid indicators (triglyceride, total cholesterol, high-density lipoprotein, low-density lipoprotein, apolipoprotein A, apolipoprotein B) and coronary heart disease, and to provide genetic evidence to support the association between the six blood lipid indicators and the risk of coronary heart disease. Methods Four regression models including MR Egger regression, weighted median (WME), random effect inverse variance weighted (IVW) and weighted model were used to conduct Mendelian randomization analysis on the causal relationship between six blood lipid indicators and coronary heart disease. At the same time, quality control was carried out to ensure the accuracy of the results. Results IVW model showed that triglyceride (OR=1.402), total cholesterol (OR=1.414), low-density lipoprotein (OR=1.867) and apolipoprotein B (OR=1.738) were risk factors for coronary heart disease. High-density lipoprotein (OR=0.791) was a protective factor for coronary heart disease. IVW model showed that apolipoprotein A was statistically significantly related to coronary heart disease. The screened SNP did not have gene pleiotropy because the MR Egger regression intercept was not statistically significant (P>0.05), Therefore, the Mendelian randomization method was an effective method for causal inference in this study. Due to the heterogeneity (P<0.001), the IVW model result should be mentioned. Conclusion After excluding confounding factors and reversing causality by using the two-sample Mendelian randomization method, unbiased estimation results were obtained, according to which there is a causal relationship between triglycerides, total cholesterol, high-density lipoprotein, low-density lipoprotein and apolipoprotein B with coronary heart disease.