基于孟德尔随机化法探究血脂四项与慢性肾脏病之间的因果关系

Causality between blood lipid tests and chronic kidney diseases:a Mendelian randomization study

目的利用孟德尔随机化(MR)研究方法验证血脂四项与慢性肾脏病(CKD)之间是否存在因果关系。方法从IEU OPen GWAS Project网站获取总胆固醇(TC)、甘油三酯(TG)、高密度脂蛋白(HDL)、低密度脂蛋白(LDL)血脂四项和CKD的基因组关联研究(GWAS)数据,筛选出合格的单核苷酸多态性(SNP),通过逆方差加权法(IVW)、加权中位数法(WM)和MR-Egger法,分别判断血脂四项与CKD的因果关系。绘制工具变量相关的血脂四项与CKD发病风险的散点图、漏斗图、森林图。采用留一法(Leave-one-out法)进行敏感性分析。结果IVW模型结果显示,TG(P=0.012,OR=1.16)为CKD的危险因素,TC(P=0.101,OR=0.87)、HDL(P=0.422,OR=1.06)、LDL(P=0.334,OR=0.92)与CKD发病风险之间的关系均无统计学意义;Cochran Q检验结果显示,TC、TG、HDL、LDL的P值均>0.05,表明本研究无异质性;MR-Egger回归截距P值均>0.05,即不存在水平多效性;敏感性分析显示,未发现工具变量中具有对结果有强影响的SNP位点。结论用MR分析血脂四项与CKD发病风险之间的因果关系,排除混杂因素和反向因果关系后,发现仅TG与CKD之间存在因果关系。

Objective To examine the causality between blood lipid tests and chronic kidney disease(CKD)using Mendelian randomization.Methods Total cholesterol(TC),triglyceride(TG),high-density lipoprotein(HDL),low-density lipoprotein(LDL)and genome-wide association study(GWAS)data were retrieved from IEU Open GWAS Project,and eligible single-nucleotide polymorphisms(SNPs)were screened.The causality between TC,TG,HDL and LDL and CKD was evaluated using the inverse variance weighted(IVW)method,weighted median(WM)method and the MR-Egger method.Scatter plots,funnel plots and forest plots were created for instrumental variables-associated TC,TG,HDL and LDL,and the risk of CKD,and sensitivity analysis was performed with the leave-one-out method.Results IVW analysis showed that TG(P=0.012,OR=1.16)was a risk factor of CKD,and TC(P=0.101,OR=0.87),HDL(P=0.422,OR=1.06)or LDL(P=0.334,OR=0.92)had no statistical associations with the risk of CKD.Cochran′s Q test showed that the P values were all>0.05 for TC,TG,HDL and LDL,indicating no heterogeneity in the study.All P value were>0.05 for the regression intercept of the MR-Egger method,indicating no horizontal pleiotropy.In addition,sensitivity analysis showed that there were no SNP sites in instrumental variables that had a strong impact on the results.Conclusions Based on a Mendelian randomization study of the causality between blood lipid tests and risk of CKD,the causality is only found between TG and risk of CKD after excluding confounding factors and reverse causality.