基于孟德尔随机化分析探讨类风湿关节炎与肺动脉高压的因果关系

Study on causal relationship between rheumatoid arthritis and pulmonary hypertension based on Mendelian randomization

目的通过孟德尔随机化设计方法探讨类风湿关节炎(RA)与肺动脉高压(PH)的因果关系。方法使用公开的全基因组关联研究获取暴露与结局的单核苷酸多态性(SNPs)数据进行汇总分析。使用逆方差加权法(IVW)作为主要分析方法评估暴露因素(RA)对结局(PH)的因果效应。以MR-Egger回归法、加权中位数法(WM)、加权模式和简单模式作为回归补充说明,进行敏感性分析评估结果的稳健性;采用“heterogeneity”函数计算P值检验异质性,采用“horizontal pleiotropy”函数计算P值检验水平多效性。结果在GWAS数据库中收录的“FINNGEN数据”中选择出的SNPs经孟德尔随机分析去除连锁不平衡且满足P<5×10^(-8),具有强相关性;通过整合暴露因素与结局获取了有效工具变量。IVW结果显示,RA是PH的风险因素(OR=1.295,95%CI:1.053~1.593,P=0.014)。“heterogeneity”函数检验表明结果不存在异质性(P=0.221);“horizontal pleiotropy”函数检验表明结果不存在水平多效性(P=0.877),总体结果稳健可靠。结论RA是PH的危险因素,RA与PH存在正向关联。

Objective To investigate the causal relationship between rheumatoid arthritis(RA)and pulmonary hypertension(PH)by the Mendelian randomization design method.Methods The data on single nucleotide polymorphisms(SNPs)of exposure and outcome were obtained by using publicly available genome-wide association studies and the summary analysis was conducted;the inverse variance-weighted(IVW)method as the primary analysis method was used to assess the causal effect of exposure factors(RA)on the outcomes(PH);the MR-Egger regression method,weighted median method(WM),weighted model and simple model were used as supplementary regression explanations to conduct the sensitivity analysis for evaluating the robustness of results;the“heterogeneity”function was used to calculate the“P value”for testing the heterogeneity,and the“horizontal pleiotropy”function was used to calculate the“P value”to test the level pleiotropy.Results In the“FINNGEN data”included in the GWAS database,the SNPs had the strong correlation after removing the linkage imbalance by Mendelian random analysis and satisfying“P<5×10^(-8)”were selected,the effective instrumental variables were obtained by integrating the exposure and outcome.The IVW results showed that RA was a risk factor for PH(OR=1.295,95%CI:1.053-1.593,P=0.014).“heterogeneity”function test showed that the results had no heterogeneity(P=0.221);“horizontal pleiotropy”function test showed that the results had no horizontal pleiotropy(P=0.877),and the total results were steady and reliable.Conclusion RA is a risk factor for PH,and RA is positively associated with PH.