体重指数与骨质疏松症之间的因果关系:一项孟德尔随机化研究

Causal relationship between body mass index and osteoporosis:A Mendelian randomization study

目的应用孟德尔随机化(Mendelian randomization,MR)研究方法,探究体重指数(BMI)与骨质疏松症之间的因果关系。方法在全基因组关联研究(genome-wide association study,GWAS)汇总数据中选取与BMI密切相关的遗传变异作为工具变量,采用MR-Egger回归法、加权中位数法、逆方差加权法、简单模式、加权模式进行MR分析,通过计算OR值和95%CI评价BMI与骨质疏松症之间的因果关系,应用MR-APSS方法使基于MR的因果推断结果更加可靠,应用连锁不平衡评分(LDSC)回归评估遗传相关性,并进行水平多效性检验、异质性检验、留一法评价结果是否可靠,通过MR-PRESSO异常值测试减少异质性和水平多效性的影响。结果共纳入421个SNP,逆方差加权法为主要分析方法,计算的OR值和95%CI为0.994(95%CI 0.992~0.997),表明BMI对骨质疏松症具有保护作用,MR-APSS方法显示BMI对骨质疏松症的影响具有统计学意义,LDSC回归显示BMI和骨质疏松症之间存在遗传相关性,MR-Egger回归截距显示工具变量不存在水平多效性,漏斗图显示工具变量不存在偏倚,留一法显示结果稳健。结论BMI和骨质疏松症之间可能存在负向因果关系,BMI是骨质疏松症的保护因素。

Objective To investigate the relationship between body mass index(BMI)and osteoporosis using Mendelian randomization analysis.Methods The genetic variation strongly related to BMI was selected as the instrumental variables in the collection data set of the genome-wide association study(GWAS).The MR-Egger regression,weighted median estimator(WME),inverse variance weighted(IVW),simple mode and weighted mode were used for Mendelian randomization(MR)analysis.The causal association between BMI and osteoporosis was evaluated by odds ratio and 95%confidence interval.The MR-APSS method was applied to make the causal inference results based on MR more reliable.The Linkage disequilibrium score regression was applied to evaluate the genetic correlation,and the horizontal pleiotropy test,heterogeneity test,and leave-one-out method were used to evaluate whether the results were reliable,The influence of heterogeneity and horizontal pleiotropy were reduced by the MR-PRESSO outlier test.Results A total of 421 SNPs were included,with inverse variance-weighted method as the main analysis approach.The calculated OR value and 95%CI were 0.994(95%CI 0.992-0.997),indicating a protective effect of BMI on osteoporosis.The MR-APSS method showed that the effect of BMI on osteoporosis was statistically significant.Linkage disequilibrium score regression demonstrated a genetic correlation between BMI and osteoporosis.MR-Egger regression intercept showed no horizontal pleiotropy of instrumental variables,and the funnel plot showed no bias in instrumental variables.Leave-one-out analysis confirmed robust results.Conclusion There may be a negative causal relationship between BMI and osteoporosis and BMI is a protective factor for osteoporosis.