Developmental trajectories of atopic dermatitis with multiomics approaches in the infant gut: COCOA birth cohort

Abstract

Background: An understanding of the phenotypes and endotypes of atopic dermatitis (AD) is essential for developing precision therapies. Recent studies have demonstrated evidence for the gut-skin axis in AD.

Objective: We sought to determine the natural course and clinical characteristics of AD phenotypes and investigate their mechanisms on the basis of multiomics analyses.

Methods: Latent class trajectory analysis was used to classify AD phenotypes in 2247 children who were followed until age 9 years from the COhort for Childhood Origin of Asthma and allergic diseases birth cohort study. Multiomics analyses (microbiome, metabolites, and gut epithelial cell transcriptome) using stool samples collected at age 6 months were performed to elucidate the underlying mechanisms of AD phenotypes.

Results: Five AD phenotypes were classified as follows: never/infrequent, early-onset transient, intermediate transient, late-onset, and early-onset persistent. Early-onset persistent and late-onset phenotypes showed increased risks of food allergy and wheezing treatment ever, with bronchial hyperresponsiveness evident only in the early-onset persistent phenotype. Multiomics analyses revealed a significantly lower relative abundance of Ruminococcus gnavus and a decreased gut acetate level in the early-onset persistent phenotype, with potential associations to ACSS2, Janus kinase-signal transducer and activator of transcription signaling, and systemic T_H2 inflammation. The early-onset transient phenotype was associated with adenosine monophosphate-activated protein kinase (AMPK) and/or chemokine signaling regulation, whereas the late-onset phenotype was linked with IL-17 and barrier dysfunction.

Conclusions: Multiomics profiling in early life may offer insights into different mechanisms underlying AD phenotypes in children.

Keywords: Atopic dermatitis; microbiome; multiomics; phenotype; transcriptome.