Nasal DNA methylation at three CpG sites predicts childhood allergic disease

Childhood allergic diseases, including asthma, rhinitis and eczema, are prevalent conditions that share strong genetic and environmental components. Diagnosis relies on clinical history and measurements of allergen-specific IgE. We hypothesized that a multi-omics model could accurately diagnose childhood allergic disease. We show that nasal DNA methylation has by far the strongest predictive power to diagnose childhood allergy, surpassing blood DNA methylation, genetic risk scores, and environmental factors. DNA methylation at only three nasal CpG sites classifies allergic disease in Dutch children, with an area under the curve (AUC) of 0.86. This was replicated in US Hispanic children (AUC 0.82). DNA methylation at these CpGs additionally detects allergic multimorbidity and symptomatic IgE sensitization. Using nasal single-cell RNA-sequencing data, we map these three CpG sites to reflect the influx of T cells and macrophages that contribute to allergic inflammation. Our study offers a simple, non-invasive diagnostic test for childhood allergy.