Exploring the heterogeneity of IgE-mediated food allergy through latent class analysis

Abstract

Introduction: Food allergy (FA) is a heterogeneous disease with multiple morbidities and a huge burden for patients and healthcare systems. Variable manifestations, comorbidities (atopic dermatitis [AD], asthma, and/or allergic rhinitis [AR]), severity (anaphylaxis), and outcomes suggest the existence of different endotypes that cluster analyses may reveal. In this study, we aimed to investigate distinct subgroups among patients with FAs using data from 524 children/adolescents. Methods: 524 patients with IgE-mediated FA (353 male [67%]; median age 4.4 years [IQR:3.0-6.8]), 354 (68%) had multiple FA. The history of AD, asthma, AR, and anaphylaxis was recorded in 59.4%, 35.5%, 24.2%, and 51.2% of the patients, respectively. Latent class analysis was carried out to distinguish clinical FA phenotypes using five potential markers of allergy severity (single/multiple FA, never/inactive/current asthma and AD, AR, and anaphylaxis). Results: Three distinct phenotypes were identified: (1) multiple FA with eczema and respiratory multimorbidity (42%), (2) multiple FA with persistent eczema (34%), and (3) single FA with respiratory multimorbidity without eczema (24%). Compared with the single FA cluster, the prevalence of AD was significantly higher in multiple FA groups. Cluster 1 had the highest frequency of AR and allergic asthma, and the lowest rate of total tolerance of FA. Discussion: We put forward the hypothesis of underlying pathogenesis according to the clinical phenotypes. While skin barrier defect may play a dominant role in the pathogenesis in Cluster 2, immune dysregulation may be dominant in Cluster 3. In Cluster 1, the most severe group, a combination of both skin barrier defects and immune dysregulation may be responsible for the clinical features.