Causal association between allergic diseases and celiac disease: a bidirectional two-sample and multivariable Mendelian-randomization study

This study aimed to assess the causality of allergic diseases and celiac disease. We collected summary-level data from publicly available genome-wide association studies to conduct our bidirectional two-sample and multivariable Mendelian randomization analysis. Furthermore, a series of sensitivity analyses were applied to validate our findings. In bidirectional two-sample MR analyses, we found a significant causal effect of atopic dermatitis (AD) on CD (Inverse-variance weighted (IVW): odds ratio [OR] = 1.302, 95% confidence interval [CI] = 1.152–1.471, <i>p</i> &lt; 0.001). We also found a significant causal effect of allergic rhinitis (AR) on CD (IVW: OR = 4.181, 95% CI = 1.495–11.697, <i>p</i> = 0.006). However, the MR-Egger method indicated a different causal effect direction compared to the IVW and weighted median method. After Bonferroni correction, the result of asthma on CD is suggestive of a causal effect (IVW: OR = 1.186, 95% CI = 1.021–1.378, <i>p</i> = 0.026). No causal effects were found when CD was considered as an exposure variable. In MVMR analyses, after separately and jointly adjusting for the influence of smoking and BMI, the causal effect of AD on CD remained robust. Our study suggests that AD is a risk factor for CD and it is considered suggestive of a causal relationship between asthma and CD. Further research is needed to explore the potential mechanisms underlying this causal effect.

本研究旨在评估过敏性疾病与乳糜泻（celiac disease, CD）之间的因果关联。我们从公开可得的全基因组关联研究（genome-wide association studies, GWAS）中获取汇总级数据，以开展双向两样本孟德尔随机化（Mendelian randomization, MR）与多变量孟德尔随机化（multivariable Mendelian randomization, MVMR）分析。此外，本研究采用一系列敏感性分析以验证所得结果。在双向两样本孟德尔随机化分析中，我们发现特应性皮炎（atopic dermatitis, AD）对乳糜泻存在显著因果效应（逆方差加权法（Inverse-variance weighted, IVW）：比值比（odds ratio, OR）=1.302，95%置信区间（confidence interval, CI）=1.152~1.471，P<0.001）。我们还发现过敏性鼻炎（allergic rhinitis, AR）对乳糜泻存在显著因果效应（IVW：OR=4.181，95%CI=1.495~11.697，P=0.006）。然而，MR-Egger法所提示的因果效应方向与IVW法及加权中位数法存在差异。经Bonferroni校正后，哮喘对乳糜泻的分析结果提示存在潜在因果效应（IVW：OR=1.186，95%CI=1.021~1.378，P=0.026）。当以乳糜泻作为暴露变量时，未检测到显著因果效应。在多变量孟德尔随机化分析中，在分别校正与联合校正吸烟及体质量指数（body mass index, BMI）的影响后，特应性皮炎对乳糜泻的因果效应仍保持稳健。本研究表明，特应性皮炎是乳糜泻的危险因素，且提示哮喘与乳糜泻之间存在潜在因果关联。仍需开展进一步研究以阐明该因果效应背后的潜在机制。