https://www.theguardian.com/lifeandstyle/2026/apr/12/at-home-food-intolerance-blood-tests
Rebecca Seal’s piece in The Guardian raises concerns that deserve to be taken seriously. The direct-to-consumer food testing market is genuinely poorly regulated. Hair-based bioresonance testing has no credible scientific basis whatsoever. There are practitioners in this space who either don’t understand the tests they’re ordering or are motivated primarily by commission. Unnecessary dietary restriction in children is a real clinical harm, with published evidence behind it. These are legitimate problems, and saying so clearly matters.
What the article also does, however, is extend those legitimate concerns into a blanket dismissal of food-specific IgG testing as a category — and in doing so, it makes a number of immunological claims that are factually incorrect, significantly oversimplified, or drawn from a deliberately narrow reading of a much more complex evidence base.
That distinction is worth making carefully, because conflating genuinely pseudoscientific tests with a misrepresentation of the immunology does not protect patients. It just replaces one kind of misinformation with another.
The claim that IgG “has nothing to do with food intolerances” is wrong
The British Dietetic Association’s framing — that IgG antibodies are simply a normal immune response to eating food — is partially correct, and partially a category error. IgG antibodies do form as part of a normal physiological response to dietary antigen exposure. That much is true. What that statement cannot do is explain why patients with Crohn’s disease, ulcerative colitis, eosinophilic esophagitis, and irritable bowel syndrome consistently show food-specific IgG positivity rates and multi-antigen reactivity that are dramatically and statistically significantly higher than healthy controls eating comparable diets. If elevated IgG were proportional only to dietary frequency, you would not see that pattern. You would see roughly similar distributions across healthy and disease populations, modulated only by eating habits. That is not what the data shows.
A 2014 study in PLOS ONE found food-specific IgG antibodies in 75.9% of Crohn’s disease patients and 63.6% of ulcerative colitis patients, compared with 33.1% of healthy controls — with more than a third of IBD patients reacting to three or more food antigens simultaneously, compared with 0.8% of controls. A 2024 study in the European Journal of Clinical Nutrition found that Crohn’s patients with measurable small bowel inflammation had significantly higher food-specific IgG positivity rates and more IgG-reactive foods than those without. These are not marginal findings.
The argument that IgG simply tracks exposure cannot account for this. People with IBD and healthy controls are not eating radically different diets such that the IBD population would be consuming eggs, wheat, and dairy at many times the frequency of healthy controls. The more immunologically coherent explanation is that intestinal barrier dysfunction — which is well-established in IBD — changes the context in which food antigens encounter the immune system, and that this context determines whether the immune response is tolerogenic or inflammatory. That is a mechanistic distinction the BDA statement does not engage with at all.
The oral tolerance and IgG4 conflation
Much of the science cited to dismiss IgG testing is specifically about IgG4, and specifically derived from allergen immunotherapy research. That context matters enormously.
IgG4 is the smallest of the four IgG subclasses — representing approximately 5% of total circulating IgG — and it has genuinely distinctive structural properties. It does not activate complement. It has poor affinity for most Fc gamma receptors, and instead preferentially binds the inhibitory receptor FcγRIIb. Through a process called Fab-arm exchange, IgG4 antibodies form bispecific molecules that effectively cross-link allergen without triggering downstream effector responses. In the context of allergen immunotherapy, where patients are exposed to incremental doses of a known allergen under controlled immunological conditions, IgG4 rises robustly and correlates with clinical desensitisation. Beekeepers exposed repeatedly to bee venom over a season show rising IgG4 to phospholipase A alongside a phenotype of B cells expressing high IL-10 and suppressed antigen-specific T cell proliferation. This is a well-characterised regulatory pathway, and it genuinely represents tolerance.
The problem is that this IgG4-as-tolerance story has been lifted out of that context and applied wholesale to all food-specific IgG elevation in all clinical situations. That is not immunologically justified. IgG1 and IgG3 — the subclasses that most commercial food-specific IgG panels actually capture alongside IgG4, or as total IgG — activate complement, form immune complexes, and drive pro-inflammatory signalling through different Fc receptors. The tolerance argument does not apply to them in the same way, and applying the IgG4 immunotherapy story to total IgG reactivity in an IBD patient is comparing a well-characterised regulatory process to a genuinely different immunological scenario.
Eosinophilic esophagitis has become the most important example of why this conflation fails. EoE is a condition in which the esophagus becomes chronically inflamed in response to food allergens — most commonly cow’s milk and wheat — in the absence of IgE sensitisation. Serum food-specific IgG4 is elevated in EoE patients compared to controls. Esophageal biopsies show significantly higher IgG4 titres in EoE than in non-EoE controls with dysphagia. In the SOFEED trial, baseline food-specific IgG4 levels predicted dietary response. A 2025 study found elevated IgG and IgA subclasses beyond IgG4 in active EoE, correlating with disease activity and varying with treatment response. This is a condition in which IgG4 to food proteins is not functioning as a benign tolerance marker — it is a marker of active, clinically relevant, food-driven tissue pathology. The immune biology is more contextually dependent than the “IgG4 equals tolerance” framing allows for.
The Shayla Love anecdote, and what it actually demonstrates
The Shayla Love story — in which her IgG test “intolerances” shifted to reflect her new dietary pattern after she eliminated her old ones — is presented in the article as definitive proof that IgG testing is meaningless. It is presented as a gotcha. And on its own, in that context, as a description of how a healthy person’s IgG profile changes with their diet, it is not actually surprising or alarming. That is expected. What it tells you in a healthy person with no pathological immune activation is, yes, the foods you have been eating most recently. That reading is correct in that context.
But the critical immunological question is whether dietary frequency is the only variable governing IgG food antibody titres, and the answer is clearly that it is not — not in patients with increased intestinal permeability, not in patients with active gut inflammation, not in patients with autoimmune disease. The mechanism matters. Food antigens presented to the immune system across an intact, healthy mucosal barrier, in a context of functional secretory IgA, healthy dendritic cell regulation, and appropriate Treg activity, will generate a qualitatively different immune response than the same food antigen crossing a compromised barrier into an inflamed lamina propria. Both scenarios may produce IgG. They are not producing it for the same reasons, or through the same immunological pathways.
The assertion that Seal’s IgG results “simply reflected what she had eaten most recently” was probably accurate for her, a person who took the test specifically after a fancy birthday lunch, with no report of any underlying gut pathology. That is exactly what you would expect, and it is exactly why context — both clinical context and the context of appropriate test interpretation — is what determines whether an IgG panel is useful clinical information or background noise.
The clinical trial evidence
The article does not engage with the controlled clinical trial literature at all. That is a significant omission in a piece that makes strong claims about evidence.
The Atkinson et al. 2004 randomised controlled trial, published in Gut — not a functional medicine journal, not an alternative health publication — enrolled 150 IBS outpatients and randomised them to either a true elimination diet guided by their IgG antibody results, or a sham diet that eliminated the same number of foods but not the IgG-reactive ones. The sham-controlled design is important: it means any benefit of the true diet was specifically attributable to eliminating the IgG-reactive foods, rather than to the general effects of dietary change, therapeutic attention, or the restriction of calories. The true diet produced a statistically significant reduction in IBS symptom severity scores compared to the sham diet. Patients who reintroduced excluded foods worsened by 24% relative to those who maintained the diet, which is a dose-response relationship that further supports biological plausibility.
A double-blind randomised crossover trial in patients with comorbid migraine and IBS found that IgG-guided elimination significantly reduced attack frequency, duration, and severity compared to the sham arm. A 2025 randomised sham-controlled trial extended this finding by measuring IL-6, TNF-α, and calcitonin gene-related peptide (CGRP) — not just subjective symptom scores. The true IgG-elimination group showed reductions in these inflammatory and neuropeptide biomarkers that were not seen in the sham group. These are mechanistic findings, not patient-reported outcomes.
None of this trial evidence appears in a piece that claims there is “absolutely no evidence” that any of these tests provide useful information. The evidence is modest in volume and requires more and larger trials. The mechanism remains incompletely understood. But “absolutely no evidence” is not a scientifically defensible characterisation of the published literature.
What the legitimate concerns actually are
The article is correct that the direct-to-consumer IgG testing market is poorly regulated and prone to misuse. It is correct that hair-based testing, applied kinesiology, and bioresonance testing should not be in the same conversation as immunological biomarker panels — conflating them, as the article does by testing all three simultaneously, muddies rather than clarifies the picture. It is correct that untrained or unethical practitioners using any test as a revenue driver rather than a clinical tool is a genuine problem. It is correct that aggressive, unguided elimination diets — especially in children — carry real clinical risks, and that the evidence base for IgG-guided intervention is not yet strong enough to support widespread prescriptive use outside of clinical supervision.
These concerns can all be held simultaneously with the recognition that food-specific IgG testing, interpreted within the correct immunological framework by a competent clinician, in the right clinical context, provides information that reflects a real immunological signal. The question of when that signal indicates clinically meaningful immune reactivity versus when it reflects normal dietary exposure is precisely the interpretive question that requires clinical expertise — which is an argument for appropriate clinical oversight of these tests, not for their categorical abolition.
The distinction between a graduate of a 2.5-hour online nutrition course ordering an IgG panel and collecting a commission, and a clinical practitioner using the same test within a comprehensive assessment of intestinal barrier function, immune phenotype, and multi-system symptoms, is real and meaningful. Collapsing that distinction in the name of simplicity does not protect patients from the first scenario. It just removes access to the second.
A note on the immunology education gap
The immunological argument in the article rests largely on the BDA statement and on comments from a medical anthropologist, not from immunologists. That is worth noting. The nuance between IgG subclasses, the context-dependence of tolerogenic versus inflammatory IgG responses, the role of intestinal permeability in shifting antigen presentation conditions, and the eosinophilic esophagitis literature — none of this is touched. The paper’s own framing of food intolerances as non-immune mediated is also incorrect: delayed non-IgE immune reactions to food are well-recognised clinically, and the immune biology governing oral tolerance is considerably more sophisticated than an antibody-as-exposure-marker model allows for.
The immune system’s relationship with food is not a binary of IgE-mediated allergy versus simple digestive intolerance. It is a dynamic, contextually governed, continuously calibrated process of tolerance and reactivity that involves innate and adaptive immune arms, mucosal and systemic compartments, regulatory and effector cell populations, and barrier systems whose integrity materially changes the nature of immune antigen encounters. That complexity does not make all food-specific IgG testing valid. It does mean that the dismissal of the entire category on immunologically thin grounds is not a service to the people the article is trying to protect.
