Skin and gut bacteria may predict eczema before it starts and point to new microbiome therapies
A review of how skin and gut microbes drive atopic dermatitis, which early biomarkers predict it, and how probiotics and live bacterial transplants may help.
Summary
Atopic dermatitis (eczema) involves more than a faulty skin barrier. This review explains how an imbalance of microbes on the skin and in the gut, called dysbiosis, feeds a two-way loop of inflammation. On eczema skin, Staphylococcus aureus overgrows and breaks down protective proteins, while helpful bacteria decline. In the gut, low diversity and fewer Bifidobacteria weaken the intestinal lining and increase inflammatory signaling. Early-life markers, including a high Klebsiella-to-Bifidobacterium ratio at 3 months and certain skin proteins at 2 months, may predict who develops eczema. Standard treatments, moisturizers and dupilumab all shift the microbiome toward a healthier profile. Probiotics, fecal transplantation and live skin bacteria such as Roseomonas mucosa are being tested to reset the microbiome rather than just suppress inflammation.
Detailed Summary
Atopic dermatitis (AD) is a common, chronic inflammatory skin disease driven by genetics, immune dysregulation, environment and microbes. This narrative review by Hou, Shao, Yuan and Tham pulls together evidence on how the skin and gut microbiomes shape AD onset, severity and treatment response. It matters for healthspan because the review frames AD as part of a gut-skin-immune axis, a network of barrier and immune interactions that also applies to other inflammatory conditions, including psoriasis.
On the skin, AD is marked by low microbial diversity, overgrowth of Staphylococcus aureus and loss of commensals such as Propionibacterium and Corynebacterium. S. aureus toxins (delta- and alpha-toxin) and proteases disrupt keratinocyte tight junctions, degrade filaggrin and suppress antimicrobial peptides. Commensal S. epidermidis, by contrast, supports barrier repair. Fungal diversity shifts too: an 8-versus-8 pilot study found more non-Malassezia fungi in AD. Healthy skin also carries more short-chain fatty acids (SCFAs), and in a Korean study Propionibacterium abundance correlated negatively with S. aureus.
In the gut, AD patients show reduced alpha diversity and fewer Bifidobacterium, Lactobacillus and Faecalibacterium. A cross-sectional study of 172 children (49 with eczema, 123 controls) found markedly lower diversity, especially of Bifidobacteria, in eczema patients aged 2 to 3 years. Dysbiosis is linked to reduced tight junction proteins (ZO-1, occludin) and a "leaky gut." In 50 AD patients versus 25 controls, permeability markers (LBP, Reg3A, I-FABP) were higher and correlated with EASI and SCORAD severity scores. Skin cytokines (TSLP, IL-25, IL-33) can in turn activate gut type 2 immunity, completing a feedback loop. Microbial metabolites, including SCFAs, tryptophan-derived AhR ligands and secondary bile acids, mediate much of this cross-talk.
The review also catalogs early-life predictors. Free phytosphingosine in skin tape strips at 2 months reached 76% prediction accuracy, the best of the single biomarkers. High TARC/CCL17, IL-8 and IL-18 at 2 months raised the odds of moderate-to-severe AD by up to five times. In a cohort of 75 high-risk infants, higher TSLP predicted AD by 24 months. Gut markers include low early diversity (100-infant cohort), low Bifidobacteria and high Klebsiella at 3 months (21 allergic infants versus 18 controls), and elevated fecal calprotectin at 2 months, which predicted AD and asthma at age 6.
Treatment shifts the microbiome. After 84 days of moisturizer, 72% of patients improved clinically and their skin microbiota resembled unaffected skin. Topical corticosteroids and calcineurin inhibitors lowered S. aureus and raised diversity, and in 51 patients the S. aureus-dominant group had the greatest severity and the clearest recovery. Dupilumab reduced S. aureus and Malassezia. Oral Lactobacillus and Bifidobacterium probiotics lowered SCORAD, and an 8-week synbiotic formulation increased gut diversity in step with EASI improvement.
Looking ahead, the authors highlight microbiota replacement. Fecal transplantation moved the gut microbiota of AD-model dogs toward a healthy donor profile, with better engraftment tracking clinical improvement. Roseomonas mucosa from healthy donors improved AD-like disease in mice, while AD-derived strains worsened it. In the BACTERiAD I/II trials (10 adults, 5 children), topical R. mucosa reduced SCORAD, itch and S. aureus load. Caveats: this is a narrative review, many biomarker studies are small and need validation, effects of JAK and IL-31 inhibitors on the microbiome are undocumented, and bile acid findings remain preclinical.
Key Findings
- In 172 children, gut microbiota diversity, particularly Bifidobacteria, was significantly lower in the 49 with eczema than in the 123 controls at ages 2-3 years
- Reduced free phytosphingosine in skin tape strips at 2 months predicted later AD with 76% accuracy, the best single biomarker cited
- Elevated TARC/CCL17, IL-8 and IL-18 at 2 months raised the likelihood of moderate-to-severe AD by up to five times
- A higher Klebsiella/Bifidobacterium ratio at 3 months was linked to greater risk of allergic disease by age 3 (21 allergic infants vs 18 controls)
- After 84 days of moisturizer use, 72% of patients improved clinically and skin microbiota shifted toward that of unaffected skin
- In 50 AD patients vs 25 controls, gut permeability markers (LBP, Reg3A, I-FABP) correlated with EASI and SCORAD severity scores
- Topical live Roseomonas mucosa in 10 adults and 5 children (BACTERiAD I/II) reduced SCORAD, itch and S. aureus load
Methodology
This is a narrative review of published human, animal and in vitro studies on skin and gut microbiota in atopic dermatitis; no formal systematic search, PRISMA protocol or meta-analysis is described. The cited studies range from small pilot sequencing studies (8 patients vs 8 controls) to prospective infant cohorts (100 infants) and early clinical trials (15 participants). Evidence is organized by mechanism, early-life biomarkers, treatment-related microbiome change and emerging therapeutics. No pooled statistics are reported.
Study Limitations
The authors note that many predictive biomarkers come from small cohorts and need validation in other populations and larger studies. Effects of JAK and IL-31 inhibitors on the microbiome are largely inferred rather than documented, and bile acid and fecal transplantation evidence is mostly from animal models. Being a narrative review, it may be subject to selection bias; no conflicts of interest were available in the supplied text.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: