Can Gut Bacteria Predict Allergy Shot Success? Early Data

A stool test that could predict whether allergen immunotherapy will help would save patients years of uncertainty, appointments, and expense. A 2025 prospective study raises that possibility in adults with allergic rhinitis. It does not establish a usable test.

The study found that baseline Proteobacteria abundance was associated with symptom and quality-of-life improvement after six months of allergen immunotherapy (AIT). But the authors reported only P<0.05 for that association, without an effect size, confidence interval, taxonomic detail below the phylum level, or a stated correction for the many microbial features that sequencing can test. Worse, Proteobacteria expansion usually appears in allergic-rhinitis literature as a sign of dysbiosis, not immune tolerance.

This is an interesting clinical signal. It is not a reason to buy a microbiome test before starting allergy shots or sublingual immunotherapy.

What the 450-person cohort actually found

In the Huo et al. study, researchers enrolled 300 adults with allergic rhinitis who were eligible for AIT and 150 healthy controls. At baseline, they profiled fecal bacteria with 16S rRNA V4 sequencing, measured circulating short-chain fatty acids (SCFAs) by LC-MS/MS, counted regulatory T cells (Tregs) by flow cytometry, and measured cytokines including IL-10 and IL-4.

Compared with healthy controls, the allergic-rhinitis group had lower gut microbial diversity (P=0.028; the reported summary does not identify the diversity metric), lower circulating SCFAs, fewer Tregs (P<0.001), lower IL-10 (P<0.05), and higher IL-4 and Th2-cell measures (P<0.001). Those results describe an immune pattern consistent with allergic inflammation: less regulatory signaling and more type 2, or Th2-skewed, activity.

After six months of AIT, the patient group improved on both the Total Nasal Symptom Score (TNSS) and Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ), with P<0.001 for each outcome. The study then linked higher baseline Proteobacteria abundance with greater TNSS reduction and RQLQ improvement (P<0.05).

The clinical question is excellent. AIT can require years of treatment, and response varies. A baseline marker that reliably separated likely responders from nonresponders could change how clinicians counsel patients. This study cannot yet tell us how well its marker would perform because it does not report the size of the association, a confidence interval, or predictive accuracy measures such as sensitivity, specificity, or area under the curve.

The association has a major confounding problem

Body mass index is the most obvious concern. The allergic-rhinitis group had a significantly higher BMI than the control group (P=0.0006). BMI strongly relates to gut microbial diversity and broad bacterial composition. The available report does not show whether the diversity analysis or the microbiome-response analysis adjusted for BMI.

That gap matters. A difference in diversity between patients and controls could reflect allergic rhinitis, body size, diet, medication use, or a mixture of these factors. The same problem can distort an apparent relationship between a bacterial group and treatment response. The publication also leaves key details unclear, including sequencing depth, rarefaction thresholds, the statistical model, dietary intake, antibiotic history, and allergen exposure at enrollment.

Observational follow-up creates another limit: the study can identify associations, but it cannot show that gut bacteria caused a better or worse AIT response. The gut microbiome may influence immune tolerance. It may also reflect the immune state, diet, adiposity, or behaviors that influence treatment outcomes.

Why the broader allergic-rhinitis evidence does not settle the question

The cohort's lower-diversity finding has weak support as a general feature of allergic rhinitis. Li et al.'s 2025 systematic review and meta-analysis pooled 10 observational studies, totaling about 550 people with allergic rhinitis and 385 healthy controls. Using random-effects analyses, it found no statistically significant pooled difference in Shannon, Simpson, or Chao1 diversity measures. It also found no consistent phylum-level differences, with substantial heterogeneity between studies.

That meta-analysis did not test AIT response, so it does not directly refute the Huo study's prediction hypothesis. It asked a different question: whether people with allergic rhinitis consistently differ from healthy people before treatment. Still, a predictive biomarker is easier to trust when the underlying disease signature reproduces across populations. Here, it does not.

Differences in geography, diet, allergic-rhinitis phenotypes, sequencing platforms, laboratory methods, and analysis pipelines can all contribute to microbiome heterogeneity. That is a reason to standardize future research, not a license to treat any one positive P value as a clinical test.

The Proteobacteria result runs against the usual biological story

The proposed gut-airway mechanism is plausible in broad outline. Gut microbes produce SCFAs, which can affect immune cells and mucosal immunity beyond the intestine. A review of the gut-lung axis describes this communication through microbial metabolites, circulating immune cells, and cytokine signaling.

For allergic rhinitis specifically, a 2025 review by Yang et al. describes lower abundance of SCFA-associated taxa, including Faecalibacterium and Bacteroides, alongside increased Proteobacteria in clinical and animal studies. In that framework, SCFA depletion can weaken Treg induction and favor Th2 responses, IgE production, and eosinophilic inflammation.

The Huo cohort's apparent responder-associated Proteobacteria signal therefore needs independent replication before anyone gives it a biological interpretation. A 2025 pediatric observational study of children with allergic rhinitis and functional constipation likewise linked expansion of Proteobacteria, including Klebsiella and Escherichia/Shigella, with inflammatory signaling and impaired barrier function. These studies concern different populations and endpoints, but they make a protective phylum-level interpretation hard to accept without much better data.

A mouse study offers a mechanism, not a treatment answer

The most coherent mechanistic clue points in the opposite direction: toward specific Bacteroidetes-related organisms and Treg activity. In a 2021 mouse study by Pang et al., researchers found Bacteroidetes depletion in six people with allergic rhinitis and in ovalbumin-challenged mice. The human sample was far too small to establish a reliable patient signature, but the mouse experiment tested a causal intervention.

Researchers gave sensitized BALB/c mice oral Bacteroides thetaiotaomicron at 10⁸ CFU/mL on days 14 through 20. Treated mice had less eosinophil infiltration in nasal and bronchoalveolar lavage fluid, lower IgE, IgG1, IL-4, IL-5, and IL-13, and more ICOS⁺CD4⁺Foxp3⁺ Tregs in mesenteric and cervical lymph nodes and spleen. The findings support a possible route from specific gut bacteria to immune tolerance.

They do not show that giving this organism helps people with allergic rhinitis, and they do not test whether it improves AIT response. Ovalbumin-sensitized mice raised under specific-pathogen-free conditions are useful for mechanism work, but they do not reproduce human diets, environmental exposures, microbiomes, or immunotherapy protocols.

What to do with this finding now

No validated gut microbiome test predicts AIT success in allergic rhinitis. Patients should choose AIT based on established clinical factors—documented allergen sensitization, symptoms, exposure pattern, treatment preferences, access, and guidance from an allergy clinician—not a consumer stool report.

The next useful study would prospectively pre-specify candidate taxa and prediction metrics, adjust for BMI, diet, antibiotics, and seasonal exposure, standardize sequencing, and replicate the result in an independent cohort. A randomized trial of at least 500 patients that profiles the microbiome before AIT and follows patient-centered outcomes could determine whether a baseline microbial pattern predicts response rather than merely traveling alongside it.

Verdict: The gut-airway immune connection has credible biological support. A phylum-level Proteobacteria finding from one observational cohort does not yet provide a clinical predictor of allergy-shot success.