血浆代谢物与冠心病的因果关系:双向两样本孟德尔随机化分析

Causal Relationship between Plasma Metabolites and Coronary Heart Disease:a Bidirectional Two-Sample Mendelian Randomization Analysis

目的通过双向两样本孟德尔随机化(MR)分析探索血浆代谢物与冠心病(CHD)的因果关系。方法采用双向两样本MR分析方法,正向MR分析时将血浆代谢物作为暴露因素,CHD作为研究结局;反向MR分析时则将CHD作为暴露因素,血浆代谢物作为研究结局。从GWAS Catalog数据库获取1400种血浆代谢物,其均为欧洲人群样本,样本量为8299例,单核苷酸多态性(SNP)数量约1540万个。从IEU OpenGWAS project网站获取CHD数据集,其为欧洲人群样本,样本量为86995例,SNP数量为2420361个。使用R 4.3.3软件、采用Two Sample MR包(版本号:0.4.22)在Windows 11系统上进行MR分析,以逆方差加权法(IVW)分析结果为主,以MR-Egger回归、加权中位数法(WME)、简单模型、加权模型分析结果为补充。采用Cochran'sQ检验判断SNP间是否存在统计学异质性,采用MR-Egger回归的截距项、MR-PRESSO Outlier检验、MR-PRESSO Global检验分析SNP的水平多效性,采用留一法分析单个SNP对IVW分析结果的影响,绘制漏斗图以评估SNP的潜在偏倚。结果共筛选出62个与血浆代谢物强相关的SNP,15个与CHD强相关的SNP。最终共纳入6种血浆代谢物。正向IVW分析结果显示:1-棕榈酰-GPG(16∶0)升高是CHD的保护因素[OR=0.842,95%CI(0.766~0.926)],N-乙酰腐胺[OR=1.125,95%CI(1.067~1.186)]、N-乙酰天冬氨酸[OR=1.068,95%CI(1.029~1.109)]、5-乙酰氨基-6-甲酰氨基-3-甲基尿嘧啶[OR=1.095,95%CI(1.048~1.144)]、N-乙酰瓜氨酸[OR=1.079,95%CI(1.034~1.126)]、花生四烯酸(20∶4n6)至油酸至反式十八碳一烯酸的比率[OR=1.316,95%CI(1.136~1.524)]升高是CHD的危险因素(P<0.05);反向IVW分析结果显示:CHD是1-棕榈酰-GPG(16∶0)[OR=1.089,95%CI(1.014~1.169)]、花生四烯酸(20∶4n6)至油酸至反式十八碳一烯酸的比率[OR=1.094,95%CI(1.013~1.182)]升高的危险因素(P<0.05),CHD与N-乙酰腐胺、N-乙酰天冬氨酸、5-乙酰氨基-6-甲酰氨基-3-甲基尿嘧啶、N-乙酰瓜氨酸无因果关系(P>0.05)。MR-Egger回归、WME、简单模型、加权模型分析的β值均与正向和反向IVW的β值方向一致。Cochran'sQ检验分析结果显示,与上述6种血浆代谢物、CHD强相关的SNP间不存在统计学异质性(P>0.05)。MR-Egger回归的截距项、MR-PRESSO Outlier检验、MR-PRESSO Global检验分析结果显示,与上述6种血浆代谢物、CHD强相关的SNP不存在水平多效性(P值均>0.05)。留一法分析结果显示,剔除单个SNP后,MR分析结果无明显改变。漏斗图分析结果显示,与上述6种血浆代谢物、CHD强相关的SNP基本对称分布,无潜在偏倚。结论1-棕榈酰-GPG(16∶0)、花生四烯酸(20∶4n6)至油酸至反式十八碳一烯酸的比率与CHD互为因果关系,N-乙酰腐胺、N-乙酰天冬氨酸、5-乙酰氨基-6-甲酰氨基-3-甲基尿嘧啶、N-乙酰瓜氨酸升高是CHD的危险因素。

Objective To explore the causal relationship between plasma metabolites and coronary heart disease(CHD)through bidirectional two-sample Mendelian randomization(MR)analysis.Methods A bidirectional two-sample MR analysis was conducted.In the forward MR analysis,plasma metabolites were considered as the exposure factors and CHD was considered as the outcome.Conversely,in the reverse MR analysis,CHD was considered as the exposure factor and plasma metabolites were considered as the outcome.A total of 1400 plasma metabolites were obtained from the GWAS Catalog database,which were all European population samples with a sample size of 8299 cases and the number of single nucleotide polymorphism(SNP)was about 15.4 million.The CHD dataset was obtained from the IEU OpenGWAS project website.It was a European population sample with a sample size of 86995 cases and the number of SNP was 2420361.The MR analysis was performed using R 4.3.3 software and the Two Sample MR package(version 0.4.22)on a Windows 11 system.The analysis results were mainly based on the inverse variance weighting(IVW),supplemented by MR-Egger regression,weighted median estimator(WME),simple model,and weighted model analysis results.Cochran'sQtest was used to determine whether there was statistical heterogeneity between SNPs.The intercept term of MR-Egger regression,MR-PRESSO Outlier test,and MR-PRESSO Global test were used to analyze the horizontal pleiotropy of SNPs.The leave-one-out method was used to analyze the impact of a single SNP on the IVW analysis results,and the funnel plot was drawn to evaluate the potential bias of SNPs.Results A total of 62 SNPs strongly associated with plasma metabolites and 15 SNPs strongly associated with CHD were screened out.Six plasma metabolites were ultimately included.The positive IVW analysis results showed that the increased 1-palmitoyl-GPG(16∶0)was a protective factor for CHD[OR=0.842,95%CI(0.766-0.926)],the increased N-acetylputrescine[OR=1.125,95%CI(1.067-1.186)],N-acetylasparagine[OR=1.068,95%CI(1.029-1.109)],5-acetylamino-6-formylamino-3-methyluracil[OR=1.095,95%CI(1.048-1.144)],N-acetylcitrulline[OR=1.079,95%CI(1.034-1.126)]and arachidonate(20∶4n6)to oleate to vaccenate(18∶1)ratio[OR=1.316,95%CI(1.136-1.524)]were risk factors for CHD(P<0.05).The reverse IVW analysis results showed that CHD was a risk factor for the increase of 1-palmitoyl-GPG(16∶0)[OR=1.089,95%CI(1.014-1.169)]and arachidonate(20∶4n6)to oleate to vaccenate(18∶1)ratio[OR=1.094,95%CI(1.013-1.182)](P<0.05),there was no causal relationship between CHD and N-acetylputrescine,N-acetylasparagine,5-acetylamino-6-formylamino-3-methyluracil,and N-acetylcitrulline(P>0.05).The direction ofβvalues of MR-Egger regression,WME,simple model,and weighted model analysis were consistent with theβvalue of positive and reverse IVW.The results of Cochran'sQtest showed that there was no statistical heterogeneity among the SNPs strongly associated with the above 6 plasma metabolites and CHD(P>0.05).The analysis results of intercept term of MR-Egger regression,MR-PRESSO Outlier test and MR-PRESSO Global test showed that there was no horizontal pleiotropy of SNPs strongly related to the above 6 plasma metabolites and CHD(Pvalues were all>0.05).Leave-one-out analysis results showed that there was no significant change in MR analysis results after single SNP was excluded.The funnel plot analysis results showed that SNPs strongly related to the above 6 plasma metabolites and CHD were basically symmetrical without potential bias.Conclusion There is a bidirectional causal relationship between 1-palmitoyl-GPG(16∶0),arachidonate(20∶4n6)to oleate to vaccenate(18∶1)ratio and CHD.Elevated N-acetylputrescine,N-acetylasparagine,5-acetylamino-6-formylamino-3-methyluracil,and N-acetylcitrulline are risk factors for CHD.