广西医学

目的基于“肺胃相合”理论，采用两样本双向孟德尔随机（MR）分析方法探讨功能性消化不良（FD）与肺癌的潜在因果关系。方法从FinnGen联盟获取FD和肺癌的全基因组关联研究数据，以FD和肺癌互为暴露因素和结局因素，以与暴露因素显著相关的单核苷酸多态性（SNP）为工具变量。以逆方差加权法作为主要分析方法，MR⁃Egger、加权众数法、加权中位数法（WME）、简单模式分析方法作为补充进行双向MR分析，并采用Cochran's Q检验评估工具变量的异质性，MR⁃Egger截距评估工具变量的水平多效性，采用MR⁃PRESSO法评估离群异常SNP，采用“留一法”评估单个SNP对结果的偏倚影响。结果正向MR结果显示，FD与肺癌发生风险之间具有正向因果关系（OR>1，P<0.05），WME亦得到同样的因果关系（OR>1，P<0.05）。反向MR结果显示，肺癌与FD无因果关联（P>0.05），其他4种分析方法亦得到同样的结果（P>0.05）。正向MR分析中工具变量不存在显著异质性，反向MR分析中工具变量存在显著异质性，但两种MR分析均不存在水平多效性。正向MR留一法分析中，逐一剔除单个SNP后没有显著改变FD对肺癌的总体影响。结论 FD可增加肺癌的发生风险，而肺癌不会反向影响FD的发生。

Objective To investigate the potential causal relation between functional dyspepsia (FD) and pulmonary carcinoma using a two⁃sample bidirectional Mendelian randomization (MR) approach based on the “lung⁃stomach connection” theory. Methods Genome⁃wide association study data for FD and pulmonary carcinoma were obtained from the FinnGen consortium. FD and pulmonary carcinoma were used as both exposure and outcome variables in turn, with single nucleotide polymorphism (SNP) significantly associated with the exposure variables selected as instrumental variables. Inverse variance weighted was used as the primary analytical method, supplemented by MR⁃Egger, weighted mode, weighted median (WME), and simple mode methods for bidirectional MR analysis. Cochran's Q test was adopted to evaluate heterogeneity among instrumental variables, the MR⁃Egger intercept was used to evaluate horizontal pleiotropy of instrumental variables, and MR⁃PRESSO was applied to detect outlier SNP. The leave⁃one⁃out was performed to identify the bias influence of individual SNP on the overall results. Results The forward MR analysis exhibited a positive causal relation between FD and the risk of pulmonary carcinoma occurred (OR>1, P<0.05), which was also supported by the WME (OR>1, P<0.05). The reverse MR analysis indicated no causal association between pulmonary carcinoma and FD (P>0.05), and this result was consistent across the other four analytical methods (P>0.05). No significant heterogeneity was observed among instrumental variables in the forward MR analysis, whereas significant heterogeneity of instrumental variables was present in the reverse MR analysis; however, no horizontal pleiotropy was detected in either MR analysis. The leave⁃one⁃out of forward MR demonstrated that the overall effect of FD on pulmonary carcinoma was not significantly altered after sequentially excluding each SNP. Conclusion FD may increase the risk of pulmonary carcinoma occurred, whereas pulmonary carcinoma does not appear to reversely influence the occurrence of FD.

ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref