The earliest clues indicating whether an infant will develop eczema accompanied by food sensitization or food allergy may reside within the microscopic ecosystem inhabiting their skin, according to a study published in the journal Allergy. Led by researchers at Washington State University’s College of Veterinary Medicine in collaboration with international institutions, the investigation found microbial differences months before these allergic conditions were diagnosed. The findings raise the possibility that skin microbiome signatures could eventually help identify infants at elevated risk, potentially opening the door to earlier preventive interventions.
The study centers on atopic dermatitis (AD)—a common form of eczema affecting up to 20% of children—alongside food sensitization and food allergy. These conditions often represent some of the earliest manifestations of the “atopic march,” in which allergic diseases can emerge sequentially during childhood. By investigating whether the skin microbiome changes before clinical onset, researchers sought to better understand the biological events occurring at the interface between the skin barrier and the developing immune system.
The research team analyzed 1,078 skin samples from infants and their mothers. The study included 429 infants overall, with 80 sampled at 2 to 3 months and 425 sampled around 12 months. Using shotgun metagenomic sequencing, the scientists characterized both the composition of the microbial communities and their genetic functional potential. The infants were clinically evaluated for atopic dermatitis, food sensitization, and food allergy. Importantly, the researchers distinguished food sensitization—evidence of an immune response despite clinical tolerance—from food allergy, which was determined using oral food challenges, skin-prick testing, and ingestion and reaction history.
The most intriguing findings emerged from the samples collected at 2 to 3 months, before the infants had been diagnosed with these conditions. Early skin dysbiosis was detected in infants who later developed AD together with food sensitization or food allergy, but not in those who subsequently developed AD alone. Increased Staphylococcus species were associated with later AD accompanied by food sensitization or allergy, while Staphylococcus aureus was specifically enriched in infants who later developed AD with food allergy. As lead author Zeyang Shen, an assistant professor in WSU’s School of Molecular Biosciences, explained:
We actually saw skin microbiome changes in infants who hadn’t even been diagnosed with these diseases yet. That was one of the most exciting findings because those changes could potentially be used as biomarkers to help clinicians diagnose disease earlier.
The researchers also uncovered important differences among clinical phenotypes. By 12 months, infants with AD alone exhibited different skin microbiome profiles from those with AD accompanied by food allergy or food sensitization. The study also examined loss-of-function, or null, mutations in the FLG gene, which encodes filaggrin, a structural protein critical to maintaining the skin barrier. Among infants with AD, those carrying FLG mutations had significantly different microbial profiles from non-carriers, supporting an association between genetic skin-barrier defects and microbial colonization.
Beyond genetics, the study found evidence that the immediate environment may help shape an infant’s skin microbiome, documenting strain-level microbial sharing between mothers and their babies. Underscoring this relationship, Shen noted:
The microbial strains are very much shared between infants and their mothers. That gives us another layer to this story, suggesting that the people we live around and continuously exchange microbes with could also be contributing to our skin microbiome.
The researchers plan to continue following and sampling the children as they grow to determine how these early microbial patterns relate to later allergic diseases, including asthma. The early-life analysis was relatively small, involving 80 infants sampled at 2 to 3 months, so the findings will require confirmation in larger and independent populations. Nevertheless, the study lays the groundwork for investigating whether skin microbiome signatures could eventually be used for early risk prediction. As Shen summarized:
Today, these diseases are diagnosed after symptoms appear. If we can validate these microbial signals in other groups of children, they could potentially serve as early biomarkers that help identify disease risk much earlier.
