PDF(Pubmed)
Abstract：
BACKGROUND: Numerous observational studies have investigated on the correlation of whole, semi-skimmed, and skimmed milk with coronary artery disease (CAD) and myocardial infarction (MI) risk; However, no consensus has been reached and evidence on any causal links between these exposures and outcomes remains unclear. This study aimed to conduct univariate and multivariate Mendelian randomization (MR) analyses, using publicly released genome-wide association study summary statistics (GWAS) from the IEU GWAS database, to ascertain the causal association of milk with various fat content with CAD and MI risk.
METHODS: For the exposure data, 29, 15, and 30 single-nucleotide polymorphisms for whole milk, semi-skimmed milk, and skimmed milk, respectively, obtained from 360,806 Europeans, were used as instrumental variables. CAD and MI comprised 141,217 and 395,795 samples, respectively. We used inverse variance weighted (IVW), weighted median, MR-Egger regression, and MR Pleiotropy Residual Sum and Outlier analyses to determine whether pleiotropy and heterogeneity could skew the MR results. Sensitivity tests were conducted to verify the robustness of the results.
RESULTS: After adjusting for false discovery rates (FDR), we discovered proof that skimmed milk intake is a genetically predicted risk factor for CAD (odds ratio [OR] = 5.302; 95% confidence interval [CI] 2.261-12.432; P < 0.001; FDR-corrected P < 0.001) and MI (OR = 2.287; 95% CI 1.218-4.300; P = 0.010; FDR-corrected P = 0.009). Most sensitivity assessments yielded valid results. Multivariable MR for CAD and MI produced results consistent with those obtained using the IVW method. There was no causal relationship between whole or semi-skimmed milk, and CAD or MI.
CONCLUSIONS: Our findings indicate that the consumption of skimmed milk may increase the risk of CAD and MI. This evidence may help inform dietary recommendations for preventing cardiovascular disease. Further studies are required to elucidate the underlying mechanisms.
METHODS: For the exposure data, 29, 15, and 30 single-nucleotide polymorphisms for whole milk, semi-skimmed milk, and skimmed milk, respectively, obtained from 360,806 Europeans, were used as instrumental variables. CAD and MI comprised 141,217 and 395,795 samples, respectively. We used inverse variance weighted (IVW), weighted median, MR-Egger regression, and MR Pleiotropy Residual Sum and Outlier analyses to determine whether pleiotropy and heterogeneity could skew the MR results. Sensitivity tests were conducted to verify the robustness of the results.
RESULTS: After adjusting for false discovery rates (FDR), we discovered proof that skimmed milk intake is a genetically predicted risk factor for CAD (odds ratio [OR] = 5.302; 95% confidence interval [CI] 2.261-12.432; P < 0.001; FDR-corrected P < 0.001) and MI (OR = 2.287; 95% CI 1.218-4.300; P = 0.010; FDR-corrected P = 0.009). Most sensitivity assessments yielded valid results. Multivariable MR for CAD and MI produced results consistent with those obtained using the IVW method. There was no causal relationship between whole or semi-skimmed milk, and CAD or MI.
CONCLUSIONS: Our findings indicate that the consumption of skimmed milk may increase the risk of CAD and MI. This evidence may help inform dietary recommendations for preventing cardiovascular disease. Further studies are required to elucidate the underlying mechanisms.
摘要：
背景：许多观察性研究已经调查了整体的相关性，半脱脂,和脱脂乳与冠状动脉疾病（CAD）和心肌梗死（MI）的风险;然而，尚未达成共识,这些暴露与结局之间的因果关系的证据仍不清楚.本研究旨在进行单变量和多变量孟德尔随机化（MR）分析，使用IEUGWAS数据库中公开发布的全基因组关联研究摘要统计数据（GWAS），确定不同脂肪含量的牛奶与CAD和MI风险的因果关系。
方法：对于暴露数据，29、15和30个单核苷酸多态性的全乳，半脱脂牛奶,和脱脂牛奶,分别,从360,806欧洲人那里获得，被用作工具变量。CAD和MI包括141,217和395,795个样本，分别。我们使用逆方差加权(IVW)，加权中位数，MR-Egger回归，以及MR多效性残差和和离群值分析，以确定多效性和异质性是否会扭曲MR结果。进行了灵敏度测试以验证结果的稳健性。
结果：在调整错误发现率(FDR)后，我们发现有证据表明,脱脂奶摄入量是CAD的遗传预测危险因素(比值比[OR]=5.302;95%置信区间[CI]2.261-12.432;P<0.001;FDR校正P<0.001)和MI(OR=2.287;95%CI1.218-4.300;P=0.010;FDR校正P=0.009).大多数敏感性评估都产生了有效的结果。CAD和MI的多变量MR产生的结果与使用IVW方法获得的结果一致。全脂或半脱脂牛奶之间没有因果关系，和CAD或MI。
结论：我们的研究结果表明，食用脱脂奶可能会增加CAD和MI的风险。这些证据可能有助于提供预防心血管疾病的饮食建议。需要进一步的研究来阐明潜在的机制。
方法：对于暴露数据，29、15和30个单核苷酸多态性的全乳，半脱脂牛奶,和脱脂牛奶,分别,从360,806欧洲人那里获得，被用作工具变量。CAD和MI包括141,217和395,795个样本，分别。我们使用逆方差加权(IVW)，加权中位数，MR-Egger回归，以及MR多效性残差和和离群值分析，以确定多效性和异质性是否会扭曲MR结果。进行了灵敏度测试以验证结果的稳健性。
结果：在调整错误发现率(FDR)后，我们发现有证据表明,脱脂奶摄入量是CAD的遗传预测危险因素(比值比[OR]=5.302;95%置信区间[CI]2.261-12.432;P<0.001;FDR校正P<0.001)和MI(OR=2.287;95%CI1.218-4.300;P=0.010;FDR校正P=0.009).大多数敏感性评估都产生了有效的结果。CAD和MI的多变量MR产生的结果与使用IVW方法获得的结果一致。全脂或半脱脂牛奶之间没有因果关系，和CAD或MI。
结论：我们的研究结果表明，食用脱脂奶可能会增加CAD和MI的风险。这些证据可能有助于提供预防心血管疾病的饮食建议。需要进一步的研究来阐明潜在的机制。
