阻塞性睡眠呼吸暂停低通气综合征相关危险因素——基于全表现关联分析(PheWas)和全基因组关联分析(GWAS)

Risk factors associated with obstructive sleep apnea syndrome-based on Phenomewide association study(PheWas)and Genome-wide association study(GWAS)

目的目前关于睡眠呼吸暂停综合征(Obstructive sleep apnea syndrome,OSA)的危险因素尚不清楚,特别是可改变的危险因素。此研究利用全表现关联分析(Phenome-wide association study,PheWas)和孟德尔随机化(Mendelian randomization,MR)从基因角度更进一步探究OSA的可改变变量。方法此研究使用了OSA的基因多态性(SNPs)作为工具变量(Instrumental variable,IVs),研究了遗传倾向的呼吸暂停综合征对英国生物库中所有表型的影响,这些IVs具有稳定性且与OSA独立相关。结果采用逆方差加权法(inverse-variance weighted,IVW)评估危险因素与阻塞性睡眠呼吸暂停综合征(OSA)之间是否具有因果关系;同时采用Cochrane Q检验评估异质性,MR_Egger截距测试检验水平多效性。结果PheWas结果经多重矫正后,共6个与OSA相关的人体特征,随后采用两样本MR验证特征与OSA之间的因果关系。MR(IVW分析法)结果显示:BMI(IVW,OR=1,95%CI=1.003-1.005,P=5.93E-20),身体脂肪含量(IVW,OR=1,95%CI=1.003-1.005,P=3.46E-12),体重(IVW,OR=1,95%CI=1.003-1.005,P=1.86E-07),臀围(IVW,OR=1,95%CI=1.003-1.005,P=0.002),腰围(IVW,OR=1,95%CI=1.003-1.005,P=0.02)与OSA存在因果关系。结论此研究的PheWas和MR研究证实了可改变人体特征与OSA之间的因果关联,为将来的临床决策和公共卫生政策制定提供可循依据。

Objective The risk factors for Obstructive sleep apnea syndrome(OSA)are currently unknown,especially the modifiable risk factors.We used phenome-wide association study(PheWas)and Mendelian randomization(MR)to explore the modifiable variables of OSA further from a genetic perspective.Methods We used genetic polymorphisms(SNPs)in OSA as instrumental variables(IVs)to investigate the effect of genetic predisposition to apnea syndrome on all phenotypes in a UK biobank that were stable and independently associated with OSA.The results were evaluated using the inverse-variance weighted(IVW)was major way to assess whether there was a causal relationship between risk factors and obstructive sleep apnea(OSA);Heterogeneity was also assessed using the Cochrane Q-test.The MR_Egger intercept test was used to test for horizontal pleiotropy.Results PheWas results were corrected for FDR multiple corrections for a total of six human characteristics associated with OSA,and two-sample MR was used to validate the causal relationship between the characteristics and OSA.The MR(IVW)results showed that BMI(IVW,OR=1,95%CI=1.003-1.005,P=5.93E-20),whole body fat(IVW,OR=1,95%CI=1.003-1.005,P=3.46E-12),body weight(IVW,OR=1,95%CI=1.003-1.005,P=1.86E-07),hip circumference(IVW,OR=1,95%CI=1.003-1.005,P=0.002),waist circumference(IVW,OR=1,95%CI=1.003-1.005,P=0.02)were causally associated with OSA.Conclusions The PheWas and MR analysis in this study confirmed the causal association between modifiable human characteristics and OSA.It provides evidence for future clinical decision which can make health policy.