6 种代谢产物与特应性皮炎的因果关联:双向双样本孟德尔随机化研究

Objective We made a bidirectional two-sample Mendelian randomization (MR) analysis to investigate the causal connection between metabolites and atopic dermatitis (AD) . Methods Single nucleotide polymorphic (SNP) sites associated with metabolites were extracted from the aggregated data of a genome-wide association study (GWAS) as instrumental variables. Causal association between six metabolites and AD was analyzed using the TwoSampleMR package in R software. The primary methods employed included inverse variance weighted (IVW), MR Egger, and weighted median. Heterogeneity testing was conducted using Cochran's Q statistics. MR Egger intercept was employed to test for level pleiotropy. Additionally, sensitivity analysis was carried out using the "leave-one-out" approach. Results The IVW analysis results indicated that ascorbic acid (OR =0.861, 95% CI: 0.751-0.987), arachidonic acid (OR =0.363, 95% CI: 0.193-0.683), and cortisone (OR = 0.447, 95% CI; 0.221 — 0.906) were negatively correlated with the occurrence of AD, while uridine (OR =3.473, 95% CI: 1.043 — 11.562), serotonin (OR = 1.896, 95% CI ¦. 1.007-3.571), and 2-hydroxyglutaric acid (OR = 2.158, 95% CI ¦. 1.186-3.924) were positively correlated with the occurrence of AD, and the differences were statistically significant. Conclusion There is no significant evidence supporting a causal association between AD and metabolic disorder, but this study identified potential evidence for a causal effect of six metabolic factors on an increased risk of AD.