Soft coral-like structures under diffused light

Food reactions: what the tests mean and what they do not

It starts with two foods. Eighteen months on, the list runs to nineteen items and the reactions have not stopped. That spiral is rarely caused by nineteen foods.

Written by Dr Mitra Basu Chhillar, M.D. Published 26 August 2026 Updated 2 September 2026 Reviewed by Team SOMA 6 min read
Listen to this article

It starts with two foods. Then a third goes, then dairy, then onions, then anything with wheat in it. Eighteen months on, the list runs to nineteen items and the reactions have not stopped.

That spiral is rarely caused by nineteen foods. It is usually caused by one category error, made early and never corrected. By the end of this article you will be able to place your own experience, and know which tests mean something.

How many different problems hide behind the word allergy?

At least six. They differ in how fast they come on, in whether portion size matters, and in whether any validated test exists.

True food allergy is the fast one: minutes, reproducible on every exposure, sometimes triggered by traces, occasionally life-threatening. It runs on an antibody called IgE, the immune system’s rapid alarm. Diagnosis needs a convincing history supported by skin prick or blood IgE testing, and that order matters. A positive test with no matching story means the immune system has merely noticed the food. Doctors call that sensitisation, not allergy.

Coeliac disease is immune too, but it is autoimmune rather than allergic. Gluten drives an attack on the lining of the small intestine. The damage reaches past digestion, into iron stores and bone density. Antibody testing comes first, usually confirmed by a small tissue sample.

The other four are not immune events at all. Lactose intolerance is an enzyme shortfall. Undigested milk sugar reaches the colon, draws water in, and is fermented into gas. FODMAP sensitivity works the same way with a wider cast of poorly absorbed sugars. Non-coeliac gluten or wheat sensitivity is real as a reported experience and unresolved as to cause; fructans, a sugar in wheat, are a serious candidate. Histamine reactions are the most confidently discussed of the six and the least established.

Dose is the most useful clue you can give yourself. A problem that scales with portion size is a capacity problem, not an immune attack. Half a serving fine and a full serving not is the signature of fermentation. An allergy does not negotiate over quantity.

The reaction Mechanism Onset Dose matters? Validated test? Food allergy (IgE) IgE antibodies Minutes No: traces count Yes, with history Coeliac disease Autoimmune, to gluten Weeks to years No Yes: blood test, then biopsy Lactose intolerance Missing lactase 30 min to hours Yes Yes: breath test FODMAP sensitivity Osmotic load, then fermentation Hours Yes, strongly No: trial only Non-coeliac wheat sensitivity Not established Hours to days Unclear No: diagnosis by exclusion Histamine reactions Proposed, unclear Variable Unclear NoSpeed and dose sort the categories; the last column decides what can be acted on.
Six reactions described in the same words, sorted by what actually separates them. Immune mechanisms arrive fast and ignore portion size; capacity mechanisms scale with the serving. The last column shows where a test can settle it.

One of these rows carries a rule that catches people out constantly, and it costs diagnoses.

Why must coeliac testing happen before gluten goes?

Because the blood tests look for the immune response to gluten that is currently being eaten. Take gluten out, and the antibodies fall. The lining recovers, and the tests drift negative whether or not the condition is there.

Removing gluten before testing does not make the answer clearer. It makes the answer unavailable. Getting it back means eating gluten daily again for several weeks before retesting. Many people decline. They then spend decades strictly gluten-free without knowing whether they need to be. They also miss the follow-up a diagnosis brings: relatives screened, and iron, B12 and bone density watched.

Feeling better without gluten does not separate coeliac disease from fructan sensitivity, or from simply eating less bread. That is an honest test doing its job within its limits. Another test, one many people have already paid for, works very differently.

Why does an IgG panel flag the foods you eat most?

IgG is the immune system’s everyday memory antibody. Making IgG to food proteins is what a working immune system does when it meets the same food repeatedly. These antibodies appear in healthy people with no symptoms at all. One form, IgG4, actually rises as tolerance develops during allergy treatment: the opposite of a reaction marker.

IgG to a food records exposure, not reaction. That is why the major allergy and immunology bodies advise against these panels for diagnosis.

If you have paid for one, the result is not nonsense. It is an accurate description of your diet, and that is exactly the problem. It flags whatever you eat often, and it arrives looking like a verdict. The real cost is not the fee. It is what usually follows.

Why does the shrinking food list feed itself?

Twenty flagged foods do not produce twenty careful experiments. They produce one long subtraction. Variety falls, and the range of plants falls with it. The microbial community that ferments what reaches your colon narrows too.

Restriction itself can narrow the microbiome and lower tolerance over time. Each reintroduction then goes badly, and the shrinking list appears to prove itself right. Lactase, the milk-sugar enzyme, behaves the same way: the body makes it in response to use. Expectation adds its own layer, and that effect is measurable physiology, not imagination.

When trouble spreads across foods with nothing chemically in common, the shared factor is usually the system meeting them, not the foods. The question stops being which food to remove. It becomes why ordinary food is being flagged at all.

Clinical pearl

A list that keeps growing is itself the finding, not a failure of the last elimination. Six safe foods and falling is the signal to stop subtracting. No single food explains a spreading pattern.

So the way out is not another elimination. It is one designed around its own ending.

When is an elimination diet worth doing?

Only when the reintroduction is already written. The method with real support is structured and time-limited. Defined foods come out for a defined window, commonly two to six weeks, with symptoms recorded throughout. The low FODMAP diet is the best studied example, and it was never meant to be permanent.

The part that usually gets abandoned carries all the information. Reintroduction is the experiment. An elimination without a planned reintroduction is not a test, it is just a smaller life. Foods return one group at a time, in a rising dose across a few days. Tolerance usually proves to be a threshold rather than a wall. Write that schedule before day one, ideally with a dietitian who knows the protocol.

Two limits matter more than any detail. None of this suits anyone for whom eating has ever been a difficult or dangerous subject, because restriction can reopen something far more serious than bloating. And a suspected allergy or coeliac disease belongs in a diagnostic pathway with your own doctor before anything is removed.

Evidence check

Proven: IgE testing read alongside a matching history, and coeliac antibody testing followed by biopsy, are established diagnostic standards. Hydrogen breath testing for lactose malabsorption is validated, with limitations. The low FODMAP diet with structured reintroduction has randomised trial support in irritable bowel syndrome, though not as a permanent diet. Being studied: non-coeliac wheat sensitivity, which has no established mechanism or test yet, and histamine intolerance, which rests on mechanism and small uncontrolled studies. IgG food panels are advised against by the major allergy societies, because the result marks exposure rather than symptom-causing foods.

What to hold on to

  • Six mechanisms hide behind one word, and only two are the immune system reacting to food.
  • Speed and dose sort them faster than any test. Dose-blind is immune, dose-dependent is capacity.
  • Coeliac testing works only while gluten is still eaten. Remove it first and the diagnosis goes.
  • IgG panels mark exposure, not reaction, and flag whatever is eaten most.
  • Restriction narrows the microbiome and can lower tolerance, so the list confirms itself.
  • Plan the reintroduction before the elimination begins, and skip both if eating has ever been a difficult subject.

Sorting a reaction into the right row is what this field keeps asking of you. The Gut and Immunity knowledge check will show you how fast you can do it.

Dr Mitra Basu Chhillar, M.D. Dr Mitra Basu Chhillar, M.D. Founder and Medical Director, SOMA Longevity Sciences. Over thirty years of clinical practice in preventive, functional and regenerative medicine.

Deep DiveAn elimination that actually ends: structuring a food trial so it answers somethingLearn how a food trial is built so that it produces an answer: why the reintroduction is the experiment, how the question and its threshold are fixed before day one, the three phases of low FODMAP and the one most people never leave, and the confounders that manufacture a false yes.

This article leads somewhere

Scroll to Top