Zhou Jianfeng, Fang Pinhao, Liu Yixin, Liang Zhiwen, Luan Siyuan, Xiao Xin, Li Xiaokun, Shang Qixin, Zhang Hanlu, Zeng Xiaoxi, Yang Yushang, Yuan Yong
Department of Thoracic Surgery, Med+X Center for Informatics, West China Hospital, Sichuan University, Chengdu, China.
Biomedical Big Data Center of West China Hospital, Med+X Center for Informatics, Sichuan University, Chengdu, China.
Eur J Nutr. 2024 Dec 21;64(1):49. doi: 10.1007/s00394-024-03562-0.
Previous studies have indicated a potential correlation between cheese intake and risk of various diseases. However, establishing a causal relationship is challenging. To address this, we employed Mendelian randomization (MR) to simulate randomized trial groups and to investigate whether there is a causal link between cheese intake and the risk of gastroesophageal reflux disease (GERD) and Barrett's esophagus.
We conducted a multivariable MR analysis using individual-level data on GERD and Barrett's esophagus from the published datasets. Univariable and multivariable MR investigations were carried out to explore and substantiate the causal association between genetically predicted cheese intake and esophageal diseases. Additionally, a network MR analysis was executed to identify potential intermediate variables.
Based on the primary causal effects model using MR analyses with the inverse-variance weighted (IVW) method, the genetically predicted that cheese intake demonstrated a protective factor of GERD (OR = 0.356; 95% CI 0.256-0.495; P = 8.22E-10) and Barrett's esophagus (OR = 0.223; 95% CI 0.114-0.437; P = 1.19E-5). These effects remained consistent after adjusting for potential confounders such as tobacco smoking (GERD: OR = 0.440; 95% CI 0.347 - 0.558; P = 1.17E-11; Barrett's esophagus: OR = 0.263; 95% CI 0.160 - 0.432; P = 1.33E-7) and BMI (GERD: OR = 0.515; 95% CI 0.424 - 0.626; P = 2.49E-11; Barrett's esophagus: OR = 0.402; 95% CI 0.243 - 0.664; P = 3.72E-4). Furthermore, the network MR showed that BMI mediated 28.10% and 27.50% of the causal effect of cheese intake on GERD and Barrett's esophagus, respectively, with statistically significant mediation effects.
The multivariable MR analysis conducted in this study revealed a reverse causal relationship between cheese intake and GERD and Barrett's esophagus. Furthermore, BMI was potential mediating factor of the cheese intake effects on GERD and Barrett's esophagus. This finding provides causal evidence for the potential protective role of cheese intake in the prevention of esophageal diseases. The mediating effect of BMI suggests that dietary interventions combined with weight management may help reduce the risk of these diseases.
先前的研究表明，奶酪摄入量与各种疾病风险之间可能存在关联。然而，确定因果关系具有挑战性。为解决这一问题，我们采用孟德尔随机化（MR）方法来模拟随机试验组，并研究奶酪摄入量与胃食管反流病（GERD）和巴雷特食管风险之间是否存在因果联系。
我们使用已发表数据集中关于GERD和巴雷特食管的个体水平数据进行多变量MR分析。进行单变量和多变量MR研究，以探索和证实基因预测的奶酪摄入量与食管疾病之间的因果关联。此外，还进行了网络MR分析以识别潜在的中间变量。
基于使用逆方差加权（IVW）方法进行MR分析的主要因果效应模型，基因预测奶酪摄入量显示为GERD（OR = 0.356；95% CI 0.256 - 0.495；P = 8.22E - 10）和巴雷特食管（OR = 0.223；95% CI 0.114 - 0.437；P = 1.19E - 5）的保护因素。在调整吸烟（GERD：OR = 0.440；95% CI 0.347 - 0.558；P = 1.17E - 11；巴雷特食管：OR = 0.263；95% CI 0.160 - 0.432；P = 1.33E - 7）和BMI（GERD：OR = 0.515；95% CI 0.424 - 0.626；P = 2.49E - 11；巴雷特食管：OR = 0.402；95% CI 0.243 - 0.664；P = 3.72E - 4）等潜在混杂因素后，这些效应仍然一致。此外，网络MR显示BMI分别介导了奶酪摄入量对GERD和巴雷特食管因果效应的28.10%和27.50%，具有统计学显著的中介效应。
本研究进行的多变量MR分析揭示了奶酪摄入量与GERD和巴雷特食管之间的反向因果关系。此外，BMI是奶酪摄入量对GERD和巴雷特食管影响的潜在中介因素。这一发现为奶酪摄入量在预防食管疾病中的潜在保护作用提供了因果证据。BMI的中介作用表明，饮食干预与体重管理相结合可能有助于降低这些疾病的风险。
