Immunological differences in atopic dermatitis across age groups: insights from single-cell multi-omics

Atopic dermatitis (AD) occurs across all ages but presents distinct clinical and immunologic features between children, adults, and older adults. The molecular programs underlying these age-specific immune differences remain poorly understood.

We performed single-cell multi-omics profiling of peripheral blood mononuclear cells (PBMCs) from 29 AD patients and 29 matched healthy controls (HC), spanning pediatric (0-17 years), adult (18-59 years), and geriatric (≥60 years) groups. Using Cellular Indexing of Transcriptomes and Epitopes by sequencing (CITE-seq), we simultaneously quantified transcriptomic (RNA) and surface proteomic (ADT) profiles across ~280,000 immune cells. Integrated analyses identified 30 immune subsets for cell-type proportion and differential expression analyses. Machine-learning classifiers were trained on significant gene and protein features to distinguish AD subgroups by age.

Compared with HC, AD blood showed enrichment of CD14+ monocytes, plasmacytoid dendritic cells, and CD4+ proliferating T cells, and differential gene expression analysis of AD vs HC revealed downstream Th2-associated signatures shared across all age groups. Within AD, pediatric patients had increased γδ T cells, naïve CD4+, and naïve CD8+ T cells, while geriatric patients exhibited more CD4+ cytotoxic and CD8+ central memory T cells, indicating a shift from naive to effector predominance with aging. Transcriptomic and proteomic analyses revealed distinct programs: pediatric AD was enriched for IL-10 and cytokine-cytokine receptor signaling; adult AD demonstrated activation of metabolic and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB)/Th1/Th17 pathways; and geriatric AD exhibited reduced adaptive immune activity but increased innate signaling. Machine-learning models based on differentially expressed genes and proteins accurately classified AD age groups (transcript-based F1 = 0.70, AUC = 0.79), identifying stable markers such as IRF2, PDK4, ZFP90, CD21, CD94, and CD122.

Single-cell multi-omics profiling revealed age-specific transcriptional and immunological programs overlaid on a shared Th2-driven inflammatory foundation in AD. Rather than discrete disease states, pediatric, adult, and geriatric AD each exhibited distinct molecular signatures: developmental and cytoskeletal in children, stress-response and chronic inflammatory in adults, and innate and metabolic in geriatric individuals. These findings support age-group molecular subtyping and age-tailored therapeutic strategies across the AD lifespan.