基于孟德尔随机化的空气污染物与肺部疾病的因果关系研究OACHSSCD

Objective To explore the causal relationship between air pollutants and lung diseases using the Mendelian randomization(MR)method,and to consider the possible mediating effects of upper respiratory tract diseases such as throat diseases and cough in it.Methods Based on the data from genome-wide association studies,single nucleotide polymorphisms closely related to air pollutants(PM2.5,PM2.5-10,NO2,etc.)and in appropriate amounts were extracted as instrumental variables.The inverse variance weighted method was adopted as the main analysis method.MR analysis was conducted with the weighted median method,weighted mode method and MR-Egger regression method,etc.as aids.Sensitivity analysis was performed using Cochran's Q test,MR-Egger intercept,MR-PRESSO test and leave-one-out(LOO)method,and the possible mediating effects of air pollutants on upper respiratory tract diseases such as throat diseases and cough were further examined.Results Two-sample MR analysis showed that PM2.5 and PM2.5-10 in air pollutants would increase the risk of bacterial pneumonia[PM2.5∶OR=1.356,95％CI(1.060,1.736),P=0.015;PM2.5-10∶OR=1.469,95％CI(1.044,2.068),P=0.027].In the subsequent sensitivity analysis,it was found that PM2.5 showed heterogeneity but no horizontal pleiotropy,and neither PM2.5-10 showed heterogeneity nor horizontal pleiotropy.The LOO test also conformed to the above results.Conclusion Through MR analysis,it is concluded that there is a certain degree of causal relationship between some air pollutants and certain lung diseases,and there is no mediating effect for upper respiratory tract diseases such as throat diseases and cough.