基于孟德尔随机化研究房颤/房扑与失眠的因果关系

Exploring the causal relationship between atrial fibrillation/atrial flutter and insomnia based on Mendelian randomization

目的通过双样本孟德尔随机化(Mendelian randomization,MR)研究方法探究心房颤动(简称“房颤”)/心房扑动(简称“房扑”)与失眠潜在的双向因果关系。方法从IEU OpenGWAS project数据库中获取房颤/房扑和失眠的全基因组关联研究汇总数据,提取显著且独立的单核苷酸多态性(single nucleotide polymorphism,SNPs)用于MR分析。采用逆方差加权法(inverse-variance weighted,IVW)、加权中位数、MREgger回归、简单众数法和加权众数法五种方法进行因果关系评估。Cochran's Q检验评估IVW的异质性,MR-Egger检测水平多效性,MR-PRESSO方法检测离群值以检验基因多效性,留一法(leave-one-out)进行敏感性分析。结果以房颤/房扑为暴露,失眠为结局,共筛选获得47个SNPs作为工具变量,IVW显示房颤/房扑与失眠间存在正向因果关系[OR=1.103,95%CI(1.008,1.208),P<0.05],加权中位数、MR-Egger回归及简单众数法和加权众数法的β值与IVW的β值方向均一致。敏感性检测方面,Cochran's Q检验P>0.05说明不存在异质性,MR-Egger检验P>0.05说明不存在水平多效性,同时MR-PRESSO方法未检测到离群值,表明没有基因多效性,留一法未检测到对结果影响大的SNP,表明结果稳固无偏倚。反向,以失眠为暴露,房颤/房扑为结局,筛选到12个SNPs作为工具变量,IVW显示失眠与房颤/房扑间不存在因果关系[OR=0.989,95%CI(0.951,1.029),P>0.05)]。结论房颤/房扑会增加失眠的发生风险,在失眠患者的健康管理中应积极预防和治疗房颤、房扑等常见心律失常疾病。

Objective To explore the potential bidirectional causality between atrial fibrillation(AF)/atrial flutter(AFL)and insomnia by two sample Mendelian randomization(MR).Methods The summary data for AF/AFL and insomnia were obtained from the genome-wide association study(GWAS)in IEU OpenGWAS project databases respectively.Significant and independent single nucleotide polymorphisms(SNPs)were extracted for MR analysis.Inversevariance weighted(IVW),weighted median(WM),MR-Egger,simple mode and weighted mode were used to assess causality.In addition,the heterogeneity of IVW was examined by Cochran's Q test,MR-Egger test was used to detect horizontal pleiotropy,the outliers were detected by MRPRESSO method to test gene pleiotropy and the leave-one-out sensitivity analysis was conducted to ensure result robustness.Results With AF/AFL as exposure and insomnia as outcome,a total of 47 SNPs were screened as instrumental variables,and IVW showed a positive causal relationship between AF/AFL and insomnia[OR=1.103,95%CI(1.008,1.208),P<0.05].Theβvalues were in the same direction of WM,MR-Egger,simple mode and weighted mode with that of IVW.Sensitivity analysis shows that there was no heterogeneity in the results of Cochran's Q test(P>0.05),and MR-Egger test(P>0.05)indicated that there was no horizontal pleiotropy.At the same time,no outlier value was detected by MR-PRESSO method,and there was no gene pleiotropy.No SNP was tested with leave-one-out method,indicating that the results were robust and unbiased.When insomnia as exposure and AF/AFL as outcome,a total of 12 SNPs were screened as instrumental variables,and IVW showed there was no causal relationship between insomnia and AF/AFL[OR=0.989,95%CI(0.951,1.029),P>0.05].Conclusion AF/AFL will increase the risk of insomnia.In the management of insomnia patients,common arrhythmia diseases such as AF and AFL should be actively prevented and treated.