Food Sensitivity, Food Allergy, Food Intolerance: Three Different Things

Food allergy — IgE

This is the one everybody means when they say allergy, and it's the only one of the three that can kill you.

The mechanism. The immune system produces IgE antibodies against a food protein. Those antibodies sit on mast cells. Next exposure, the protein cross-links the antibodies, the mast cell degranulates, and histamine and other mediators flood out.

The timing is the giveaway. Minutes. Occasionally up to two hours, rarely longer. If you ate something and your lips tingled or your throat tightened within fifteen minutes, that's an immediate-type reaction and it behaves quite unlike anything else on this page.

What it looks like. Hives, swelling — particularly lips, tongue, face — vomiting, wheeze, and at the severe end anaphylaxis: airway compromise, blood pressure collapse, a medical emergency.

Who diagnoses it. An allergist. Skin prick testing, specific IgE blood testing, and where necessary a supervised oral food challenge, which remains the reference standard. Not me, and not any functional panel.

This is the important paragraph on this page. If you have had an immediate reaction to a food — any swelling, any breathing difficulty, any sense of your throat closing — that is an NHS allergy clinic referral, not a functional medicine question. You may need adrenaline available. Do not investigate that with a food sensitivity panel and do not let anyone tell you it's a sensitivity. The two are not on a spectrum. They are different immunology.

Food intolerance — enzymatic and chemical

No immune involvement at all. This is chemistry.

Lactose intolerance is the clearest example. Lactase, the enzyme that breaks down lactose, declines after weaning in most of the world's population. Undigested lactose reaches the colon, bacteria ferment it, and you get gas, bloating and loose stools. Nothing immunological is happening. It's dose-dependent — a splash in coffee is fine, a pint of milk isn't — which is a useful diagnostic feature in itself.

Persistence of lactase into adulthood is, incidentally, the best-established gene–diet interaction there is. A genuine Mendelian trait, not a wellness claim.

Histamine intolerance works differently. Histamine in food is normally broken down by diamine oxidase in the gut wall. When that capacity is exceeded — high-histamine foods, impaired DAO, or medications that inhibit it — histamine gets absorbed. The result mimics allergy: flushing, headache, nasal congestion, loose stools, sometimes hives. But there's no antibody, and it's dose- and threshold-dependent rather than all-or-nothing. Aged cheese, fermented foods, cured meat, leftovers, and alcohol are the usual suspects.

Wine is worth mentioning because it gets blamed on sulphites so often. Placebo-controlled challenge studies using low-sulphite wine have generally failed to reproduce the reaction, and the current view is that wine reactions are multifactorial — histamine and alcohol itself are likelier culprits than sulphites in most people.

FODMAPs are fermentable carbohydrates that draw water into the small intestine and get fermented in the colon. Very common in IBS. Again, dose-dependent, and again, no immune mechanism.

Chemical sensitivities — salicylates, benzoates, tartrazine, glutamate — sit here too. Real in some people, over-diagnosed generally.

Food sensitivity — IgG

This is the contested one, and I'm going to be straight about where the argument sits rather than pretend it's settled.

The proposed mechanism. IgG antibodies form against food proteins. Unlike IgE, they don't trigger mast cell degranulation. The theory is that they drive a delayed, low-grade reaction, hours to days later. That part is mechanistic and untested: whether food-specific IgG causes symptoms hasn't been shown, and IgG to foods is common in people with no symptoms at all.

Why a delayed reaction would be hard to spot. If a food did cause symptoms a day or two later, you'd struggle to work it out by observation. You ate thirty things in that window. That's the case for a structured trial rather than guesswork.

The honest position on the evidence. Elevated IgG to a food indicates exposure and immune recognition. That much is uncontroversial — it's what IgG does. What's contested is whether it indicates a clinically meaningful problem. Some allergy bodies take the view that IgG to food is a normal marker of exposure and tolerance rather than pathology, and they have a point: you would expect IgG to the foods you eat most.

What I'd say from clinical practice is narrower than what's usually claimed for these panels. The results are a hypothesis generator, not a diagnosis. The panel gives you a shortlist. Structured elimination and reintroduction tells you whether the shortlist means anything. If a food comes up strongly, you leave it out for the 12-week trial, nothing changes, and it reintroduces cleanly — then whatever the antibody was doing, it wasn't causing your symptoms. That's a real answer too.

The one randomised trial is worth knowing about. In 150 people with IBS, leaving out IgG-flagged foods for 12 weeks reduced symptom scores by 10% more than a sham diet, and by 26% in people who stuck to it fully (Atkinson 2004, Gut, PMID 15361495, doi:10.1136/gut.2003.037697). Grade: mixed. One trial, in IBS only, a modest effect, not repeated at scale. How I use the panel, and the test itself, are on the food sensitivity testing page.

Anyone selling IgG testing as a definitive diagnosis of food allergy is overselling it. Anyone dismissing the pattern entirely is ignoring what happens when people actually run the elimination properly.

The comparison

Why the same food can appear in all three columns

Wheat is the example that shows why this gets confusing.

Someone can have a genuine IgE wheat allergy — immediate, dangerous. Someone else can have coeliac disease, which is autoimmune and needs a gastroenterologist, formal serology, and biopsy while still eating gluten. Someone else reacts to the fructans in wheat, which is a FODMAP intolerance and nothing to do with gluten at all. And someone else shows IgG reactivity to wheat proteins with delayed symptoms.

Four people, one food, four mechanisms, four different correct responses.

Coeliac deserves its own warning. If coeliac disease is a possibility, get tested (tTG-IgA with total IgA) before removing gluten. The test needs gluten in the diet: NICE's coeliac guideline (NG20) asks for gluten in more than one meal a day for at least 6 weeks beforehand. Serology becomes unreliable once you've been gluten-free, and people who self-exclude first often end up unable to get a clean answer without a formal gluten challenge they'd rather not do.

What actually helps

Start with what's dangerous. Immediate reactions go to an allergist. Coeliac serology before removing gluten. Blood in stool, weight loss, difficulty swallowing, night-time symptoms — those go to a doctor, not into an elimination diet.

Then look at the terrain. This is the part that gets skipped. Food reactivity doesn't happen in a vacuum. If the gut barrier is compromised, larger protein fragments cross into circulation where the immune system encounters them. If secretory IgA — the mucosal immune system's first line — is low, sensitisation happens more readily. If stomach acid or pancreatic enzyme output is inadequate, proteins arrive incompletely broken down.

Which means the reactive food list can be a consequence of gut dysfunction rather than the cause of the symptoms. I've seen people react to twenty-five foods, repair the barrier, and reintroduce most of them without incident. Eliminating twenty-five foods forever would have been the wrong answer — and a miserable one.

Then eliminate properly if you're going to. Leave out the panel-flagged foods for a 12-week trial (the length used in the one randomised trial, Atkinson 2004, Gut, PMID 15361495). Then reintroduce one food every 3 days: a small portion on day 1, a normal portion on day 2, none on day 3, with a daily symptom diary. If symptoms return, stop that food and wait until they settle before testing the next; retry it once after 3 months. A food that causes no reaction goes back in for good. Evidence: one RCT in IBS (mixed); the reintroduction steps are practice, modelled on staged FODMAP reintroduction, not trial-tested for IgG. Removing eight things at once and feeling better tells you nothing about which of the eight mattered.

And keep the endpoint in view. The aim is the widest tolerable diet, not the narrowest. Long-term restriction narrows the microbiome, risks nutritional gaps, and does something to a person's relationship with eating that's hard to undo. Elimination is a diagnostic tool with a defined duration. It is not a lifestyle.

Educational content, not medical advice. If you have experienced any immediate reaction to food involving swelling or breathing difficulty, seek medical assessment — that requires an allergist, not a nutrition practitioner.