What the IgG test measures — and the argument about it

The test measures antibodies called IgG against a large panel of food and drink proteins, and reports how much of each one you have.

Here is what most clinics selling it will not tell you. The major allergy societies do not accept food-specific IgG as a diagnostic test. The European Academy of Allergy and Clinical Immunology concluded that food-specific IgG4 “does not indicate (imminent) food allergy or intolerance, but rather a physiological response of the immune system after exposition to food components”. The Canadian society called it “a marker of exposure and tolerance to food”. The British position is blunter: IgG antibodies to food are formed in all healthy asymptomatic people.

The laboratory that makes the test I use prints the same caveat on the report itself: the link between food, raised IgG and chronic symptoms “is still debated in the scientific community and a consensus has not been reached thus far”.

In plain terms: a raised IgG to bread may mean nothing more than that you eat bread.

So why use it at all?

Because the trials of what happens when you act on the result have been more encouraging than the theory predicts. In 2025, a double-blind, sham-controlled randomised trial across eight centres tested an IgG-guided elimination diet in 223 patients with irritable bowel syndrome. Patients guided by their real results did better than those given a convincing fake diet — 59.6% responded against 42.1%. The benefit was concentrated in constipation-predominant and mixed-pattern IBS, and was absent in diarrhoea-predominant IBS.

An earlier randomised trial pointed the same way, with symptoms returning when the eliminated foods were reintroduced. A 2026 systematic review of thirteen studies and 935 patients found consistent improvement in pain, distension, bowel habit and quality of life — while noting, fairly, substantial variation in study quality and a real risk of malnutrition when multiple foods are removed without supervision.

What about inflammation?

The mechanism people describe is real biology. When a food protein crosses the gut wall intact and meets an antibody, the two bind and form an immune complex, and immune complexes are a recognised route to low-grade inflammation.

What has not been established is that this is what is happening in any individual with a raised result — because those antibodies also appear in people with no symptoms whatsoever.

So I will not tell you your result proves you are inflamed. What I will tell you is which foods are worth removing, and that if inflammation was food-driven, it is one of the things that should change when we remove them.

Then we measure it. Where it is relevant, I take baseline inflammatory markers and a structured symptom score before the elimination and repeat them after, so the question is answered with numbers rather than impressions. Measure first, so you can tell whether anything actually changed.

What is on the panel

Panel 1:The panel covers 283 foods and drinks — milk and egg, meat, fish and seafood, cereals and seeds, nuts, legumes, fruit, vegetables, spices, edible mushrooms, coffee and tea, and a group of newer foods including algae and edible insects.

• Camel milk is on the panel, alongside cow, goat, sheep and buffalo. Very few panels test it, and in this region that matters.
• Four proteins are tested separately from the foods they come in — the three main cow’s milk proteins and wheat gliadin. This is useful: it can show that the problem is casein rather than dairy as a category, which changes what you actually have to avoid.
• Results are reported as an actual concentration for each food, not a score out of ten, in three bands — low, intermediate and highly elevated.

Panel 2: The panel covers over 200 foods and drinks across dairy and egg, gluten-containing and gluten-free grains, fruit, vegetables, fish and seafood, meat, herbs and spices, nuts and seeds.

• Gliadin is tested separately from the gluten-containing grains, and the three main cow’s milk proteins separately from milk.
• Results come back two ways — grouped by food type, and ranked in order of reactivity — in three bands: normal, borderline and elevated.
• The report is available in Arabic as well as English.

You can either do both panels or based on your condition and what is more common in your diet we help you coose one.

• The sample is a standard blood draw.
• You come in to go through the results. I do not email you the report with a leaflet attached. These reports run to dozens of pages and the software-generated commentary in them is generic — the same list of symptoms appears beside every food on the list. Reading the numbers against your history is the part that makes the test worth doing, and it is not something a PDF can do for you.

Before you test: what should be ruled out first

A food intolerance panel belongs after the conditions with definitive tests have been excluded, not before.

• Coeliac disease. Not an intolerance — an autoimmune condition causing real intestinal damage. It must be excluded with tTG-IgA and total IgA before anyone removes gluten, because the test only works while you are still eating it. Cutting gluten out first makes coeliac disease hard to diagnose for months.
• Lactose intolerance, where dairy is the suspect — hydrogen breath test.
• Thyroid function, ferritin, B12, vitamin D, full blood count and an inflammatory marker. Fatigue and brain fog attributed to food are very often anaemia, thyroid disease or vitamin D deficiency — all common here, all treatable.
• Helicobacter pylori where the picture is upper gastrointestinal.

When this is the wrong test

• Suspected allergy — see above.
• Any red flag: unintentional weight loss, blood in the stool, iron-deficiency anaemia, fever, vomiting, symptoms that wake you at night, a change in bowel habit after 50, or a family history of bowel cancer or inflammatory bowel disease. These need investigating properly, and a food panel delays that.
• An existing eating disorder, or a restrictive and anxious relationship with food. A list of forbidden foods is actively harmful here, and I will say so rather than take the booking.
• Pregnancy, and children — neither should undertake multi-food elimination without paediatric or obstetric dietetic supervision.
• Diarrhoea-predominant IBS, where the trial evidence showed no advantage over placebo. A low-FODMAP approach has better support.

What happens after the result

The test is the beginning of the process, not the end of it. The result on its own changes nothing.

1. We go through the panel together, set it against your symptom history and your actual diet, and decide which reactions are worth acting on. A raised result against a food you eat twice a year is not where we start.
2. Structured elimination, around three months. Not everything that came back raised — the shortlist we agreed. With a written plan naming what each food hides in, because the difficult part is rarely the obvious form: milk protein is in processed meat and margarine, gluten is in sauces and beer, and if you do not know that, the elimination is not real.
3. Replacements, not just removals. Every food that comes out is replaced by something specific. This is where most elimination diets fail.
4. Nutritional cover. The genuine risk of multi-food elimination is an inadequate diet, and it is the risk the research literature raises most consistently — removing wheat, for instance, takes out a major source of B vitamins unless something replaces it. Planned for, not left to chance.
5. Systematic reintroduction. One food at a time, every three to four days, with a symptom diary. This is the part people skip and the part that matters most.
6. The verdict. It comes from the reintroduction, not from the printout. A food that causes nothing on reintroduction goes back into your diet, whatever it scored.

What this test is not

• It is not an allergy test.
• It is not a diagnosis.
• It is not a permanent list — antibody levels follow what you eat.
• It is not a reason to spend your life eating twelve foods. If the plan is making your diet smaller rather than clearer, the plan is wrong.