孟德尔随机化分析失眠与神经系统疾病及耳鸣的因果关系

Causal Relationship of Insomnia and Neurological Diseases and Tinnitus analyzed by Mendelian Randomization

目的:采用两样本孟德尔随机化研究方法(Mendelian randomization,MR)评估失眠与神经系统疾病及耳鸣发生风险之间的潜在因果关系。方法:利用IEU OpenGWASproject网站获取失眠、神经系统疾病及耳鸣的全基因组关联研究(GWAS)数据,筛选与失眠密切相关且独立的单核苷酸多态性(SNPs)作为工具变量(IVs),与神经系统疾病相关的SNPs作为结局。其中失眠数据包含462341例样本,神经系统疾病数据包含148552例样本,耳鸣数据包含117882例样本,从中选取42个与失眠密切相关的SNP作为工具变量。分别采用逆方差加权法(inverse varianceweighted,IVW)、加权中位数(Weightedmedian,WME)和MR-Egger法进行孟德尔随机化分析。同时采用Egger intercept检测水平多效性,留一法(Leave-one-out)进行敏感性分析,保证结果稳健性。结果:失眠与神经系统疾病发生之间呈显著正相关(IVW:OR=2.032,95%CI 1.567~2.636,P=9.011×10^(-8));Egger intercept的值和0没有统计学差异(P=0.803),表明没有水平多效性;Leave-one-out法显示结果稳定,未检测到对结果影响大的SNP;全部F值大于10,提示不存在弱IVs偏倚。失眠与耳鸣发生之间呈正相关(IVW:OR=1.056,95%CI 1.009~1.104,P=0.019);Egger intercept的值和0没有统计学差异(P=0.985),表明没有水平多效性;Leave-one-out法显示结果稳定,未检测到对结果影响大的SNP;全部F值大于10,提示不存在弱IVs偏倚。结论:失眠与神经系统疾病发生之间存在正向因果关系;失眠与耳鸣发生之间也存在正向因果关系。

Objective:A two-sample Mendelian randomization(MR)study was used to evaluate the potential causal relationship between insomnia and the risk of neurological disorders and tinnitus.Methods:The IEU OpenGWASproject website was used to obtain genome-wide association study(GWAS)data of insomnia and neurological diseases,and the independent single nucleotide polymorphisms(SNPs)closely related to insomnia were screened as instrumental variables(IVs),and SNPs related to neurological diseases were selected as outcomes.Among them,The insomnia data included 462341 samples,the neurological disease data included 148552 samples,and the tinnitus data contained 117882 samples.A total of 42 SNPs closely related to insomnia were selected as instrumental variables.Inverse variance weighted(IVW),weighted median(WME)and MR-Egger methods were used for Mendelian randomization analysis.At the same time,the Egger intercept detection level pleiotropy and the leave-one-out method were used for sensitivity analysis to ensure the robustness of the results.Results:IVW analysis showed a significant positive correlation between insomnia and the occurrence of neurological diseases(OR=2.032,95%CI 1.567~2.636,P=9.011×10^(-8)).There was no significant difference between the value of Egger intercept and 0(P=0.803),indicating that there was no horizontal pleiotropy.The leave-one-out method showed that the results were stable,and no SNPs with a large impact on the results were detected.All F-scores greater than 10 indicate no weak IVs bias.IVW analysis showed that there was a positive correlation between insomnia and tinnitus(OR=1.056,95%CI 1.009~1.104,P=0.019).There was no significant difference between the value of Egger intercept and 0(P=0.985).The leave-one-out method showed that the results were stable,and no SNPs with a large impact on the results were detected.All F-scores greater than 10 indicate no weak IVs bias.Conclusion:There is a positive causal relationship between insomnia and the occurrence of neurological diseases,and there is also a positive causal relationship between insomnia and tinnitus.