两样本孟德尔随机化分析幽门螺杆菌感染与慢性乙型肝炎的因果关系

Two-sample Mendelian randomization analysis of the causal relationship between Helicobacter pylori infection and chronic hepatitis B

目的研究幽门螺杆菌(Helicobacter pylori,H.pylori)感染与慢性乙型肝炎(chronic hepatitis B,CHB)的因果关系。方法利用大规模人群全基因组关联研究(genome-wide association studies,GWAS)汇总数据,筛选与H.pylori感染密切相关且独立的单核苷酸多态性(single nucleotide polymorphisms,SNPs)作为工具变量(instrumental variables,IVs),运用逆方差加权法(inverse variance weighted,IVW)、MR-Egger回归、加权中值法(weighted median,WM)、简单模式和加权模式进行孟德尔随机化(Mendelian randomization,MR)分析,用效应值(b)或OR值及P值来评价H.pylori感染与CHB之间是否存在关联,Cochran′s Q检验统计数据用于量化异质性,MR-Egger回归评估SNPs的多水平效应检验,留一法确定是否具有潜在影响的SNPs,评估结果的稳定性,最后绘制可视化散点图、森林图、漏斗图来综合判断H.pylori感染与CHB的因果关系。结果IVW显示H.pylori感染与CHB之间无因果关系(b=12.8904,OR>1,P=0.371),其他MR-Egger、WM、简单模式和加权模式4种分析方法也不能证明H.pylori感染与CHB之间有因果关系。Cochran′s Q检验结果分别为P=0.894和P=0.899,即不存在异质性;Egger-intercept结果为P=0.424,即不存在多效性;敏感性分析显示结果稳定;全部F值>10,提示不存在弱IVs偏倚。结论基于分析结果,目前暂缺乏强有力的证据支持H.pylori感染与CHB之间的因果关系。

Objective To study the causal relationship between Helicobacter pylori(H.pylori)infection and chronic hepatitis B(CHB).Methods By using data from genome-wide association studies(GWAS)of large-scale population,single nucleotide polymorphisms(SNPs)closely related to H.pylori infection were selected as instrumental variables(IVs),and Mendelian randomization(MR)analysis was conducted using inverse variance weighted method(IVW),MR-Egger regression,weighted median method(WM),simple mode and weighted mode.The effect value(b)or odd ratio(OR)and P value were used to assess if there is a correlation between H.pylori infection and CHB,Cochran's Q test statistics were used to quantify heterogeneity,MR-Egger regression was used to evaluate the multi-level effect test of SNPs,leave-one-out analysis was used to determine whether there was a potential SNPs,and the stability of the results was evaluated.Finally,create a visual scatter plot,forest plot and funnel plot to comprehensively judge the causal relationship between H.pylori infection and CHB.Results IVW showed that there was no causal relationship between H.pylori infection and CHB(b=12.8904,OR>1,P=0.371),and other four analysis methods such as MR-Egger,WM,simple mode,and weighted mode could not prove the causal relationship between H.pylori infection and CHB.Cochran's Q test results were P=0.894 and P=0.899,respectively,indicating no significant heterogeneity.The Egger-intercept result was P=0.424,indicating no pleiotropic effect.The sensitivity analysis showed that the results were stable.All F values>10,indicating no weak IVs bias.Conclusion Based on the analysis results,there is no strong evidence to support the causal relationship between H.pylori infection and CHB.