What Is Gluten and Why Do Some People React to It?

Gluten has become one of those words that carries more weight in a grocery aisle than it does in a biology textbook. Shelves are stacked with “gluten-free” labels on products that never contained gluten to begin with, and somewhere along the way the actual protein got lost behind the marketing. So let’s back up and look at what gluten really is, and why it causes real trouble for some people and, for most others, essentially none at all.

What gluten actually is

Gluten is the main storage protein in wheat. When wheat dough is washed with water to remove the starch and soluble components, what’s left behind is a rubbery, elastic mass, and that’s gluten [2]. It’s not one protein but a mixture of hundreds of related ones, split mainly into two families: gliadins and glutenins, which together make up roughly 85 to 90 percent of the protein in a wheat kernel [2]. Similar storage proteins show up in other grains under different names: secalin in rye, hordein in barley, and avenins in oats, and all of these get grouped under the same “gluten” umbrella [2].

The two families do different jobs in dough. Gliadins are more responsible for viscosity and extensibility, while glutenins give dough its strength and elasticity [2]. That combination is why gluten is so useful in baking, and also why it shows up as a hidden additive in things that have no obvious business containing wheat, like processed meat, ice cream, and even some medications, where it’s used as a binder or filler [2].

The reason gluten causes problems for anyone at all comes down to its amino acid makeup. Gliadin is unusually rich in proline and glutamine, and that combination creates tight, compact peptide sequences that resist being broken down by gastric, pancreatic, and intestinal enzymes [2]. Most of gluten gets digested just fine, but these resistant fragments survive intact into the small intestine, where in certain people they trigger a response that has nothing to do with normal digestion.

Three different reactions, one shared culprit

This is where a lot of the confusion sets in, because “reacting to gluten” can mean three distinct things, and they don’t share a mechanism, a treatment threshold, or even a diagnostic test.

Coeliac disease is an autoimmune condition, not a food intolerance. In people carrying specific genetic markers (HLA-DQ2 or HLA-DQ8), an enzyme called tissue transglutaminase modifies gliadin fragments in a way that makes them bind more tightly to immune cells, which then mount a full T-cell response against the lining of the small intestine [5]. This causes measurable damage: flattened villi, crypt hyperplasia, and a gut that struggles to absorb iron, folate, and fat-soluble vitamins. It affects around 1 percent of the population globally, though the range across studies runs from about 0.7 to 1.4 percent depending on how it was diagnosed [5]. Coeliac disease is confirmed through blood antibody tests (IgA tissue transglutaminase being the standard first screen) and usually a small intestine biopsy, and the only treatment is a strict, lifelong gluten-free diet, since even trace amounts can keep driving the immune response [5].

Wheat allergy is a separate mechanism entirely. It’s IgE-mediated, meaning the immune system treats wheat proteins the way it would treat pollen or peanuts, and it produces classic allergic symptoms: hives, swelling, respiratory reactions, and in rare cases anaphylaxis, usually within minutes to hours of exposure [3]. It’s uncommon in adults (under 1 percent) and most children outgrow it [2].

Non-coeliac gluten sensitivity, usually abbreviated NCGS, is the one without a clean answer. It describes people who get gastrointestinal and other symptoms after eating gluten or wheat, but who test negative for both coeliac disease and wheat allergy [1]. And this is where the story gets more complicated than the name suggests.

The problem with non-coeliac gluten sensitivity

Around 10 percent of people worldwide self-report a sensitivity to gluten or wheat, but that number badly overstates how many people actually react to gluten specifically when tested under controlled conditions [1]. When researchers run double-blind, placebo-controlled gluten challenges, where neither the participant nor the person assessing them knows whether gluten or a placebo was given, only about 16 to 30 percent of self-reported cases show a response that’s actually tied to gluten [1]. In the largest meta-analysis on this, a gluten-specific effect only showed up in three studies that met strict methodological criteria, and across all the studies pooled together, the effect didn’t differ from placebo [1].

So if it’s not gluten, what’s causing symptoms in people who actually feel better on a gluten-free diet? Three things keep showing up in the research.

Fructans, a type of fermentable carbohydrate that falls under the FODMAP umbrella, are found alongside gluten in wheat and are fermented by gut bacteria in a way that produces gas, bloating, and abdominal discomfort in sensitive individuals [1]. A well-designed trial by Skodje and colleagues gave people gluten, fructans, or placebo in a crossover design and found that fructans, not gluten, were the ones triggering symptoms [1]. An earlier study by Biesiekierski found something similar: when participants were kept on a low-FODMAP diet throughout, gluten added on top of that produced no extra symptoms at all [1].

That’s the second piece: a documented nocebo effect. In that same trial, the combination of expecting gluten and actually receiving it produced the worst symptoms, but expecting gluten while receiving a placebo bread produced nearly as strong a reaction [1]. Across multiple studies, nocebo responses in placebo groups average around 40 percent and have reached as high as 56 percent [1]. This doesn’t mean people are imagining their symptoms. It means the gut and the brain are more tightly linked than a simple food-trigger model accounts for, and expectation itself is doing measurable physiological work.

The third candidate is a wheat protein most people have never heard of: amylase-trypsin inhibitors, or ATIs. These are non-gluten proteins that make up 2 to 4 percent of wheat’s total protein and act as natural pest defenses in the plant [2]. In lab and animal studies, ATIs activate toll-like receptor 4, a pathway that triggers the release of inflammatory molecules from immune cells [2]. The catch is that no controlled human trial has actually confirmed ATIs cause NCGS symptoms in real patients, so their role stays plausible but unproven [1].

Put together, wheat contains at least three separate irritants bundled into the same food (gluten, fructans, and ATIs), plus a nervous system that’s demonstrably responsive to what people believe they’re eating. That combination is a big part of why “gluten sensitivity” has been so hard to pin down as one clean diagnosis, and why some researchers now argue for the broader term non-coeliac wheat sensitivity instead, since gluten may not even be the main offender in a lot of cases [1].

Why this matters beyond the label

None of this is an argument against gluten-free diets working for the people who need them, whether that’s coeliac disease, wheat allergy, or a subset of NCGS patients who do show a real, reproducible response to gluten in controlled testing [1]. What it does mean is that self-diagnosis is unreliable in both directions. Studies find that only about a third of self-reported food sensitivities hold up under objective testing [1], and separately, over two-thirds of people avoiding gluten haven’t actually been diagnosed with anything that requires it [3].

There’s also a cost to getting this wrong. Gluten-free products often run over a third more expensive than their wheat-based equivalents, and they’re frequently lower in fibre and several B vitamins unless deliberately fortified [1][3]. For someone with coeliac disease, that trade-off is non-negotiable. For someone whose actual trigger is fructans, a full gluten-free diet might solve the problem by accident, since it usually cuts fructan intake too, but a targeted low-FODMAP approach could get there with a lot less restriction.

Where the research stands now

The current state of the field is this: coeliac disease and wheat allergy are well-defined, with clear diagnostic pathways and a clear mechanism. NCGS is real as a symptom cluster, meaning people do experience gastrointestinal and extraintestinal symptoms tied to wheat consumption, but it’s not yet clear that gluten is the consistent cause. The current best-practice approach for anyone with suspected NCGS is a structured elimination and reintroduction process, ideally with a low-FODMAP phase included, rather than jumping straight to permanent gluten avoidance based on a hunch [1].

What’s most interesting to me is how much of this traces back to one shared property: gliadin’s resistance to digestion. That single biochemical quirk (a peptide sequence too proline-rich for human enzymes to fully break down) is the root of coeliac disease’s immune attack, a contributing factor in wheat allergy, and part of the reason wheat’s other components get blamed for symptoms they may not be causing. Same molecule, three very different stories depending on who’s eating it.

Common Questions

Is gluten bad for everyone?

No. For most people, gluten is digested without incident. It only becomes harmful in coeliac disease, where the immune system attacks the small intestine in response to it, and in wheat allergy, which is a separate IgE-mediated reaction.

What's the difference between coeliac disease and gluten sensitivity?

Coeliac disease is an autoimmune condition confirmed by blood antibodies and intestinal biopsy, and it causes measurable damage to the gut lining. Non-coeliac gluten sensitivity has no confirmed biomarker or tissue damage, and controlled trials suggest gluten itself is often not the actual trigger.

If gluten isn't always the culprit in gluten sensitivity, what is?

The leading candidates are fructans, a type of FODMAP carbohydrate found in wheat, along with amylase-trypsin inhibitors and a documented nocebo effect, where the expectation of eating gluten produces symptoms on its own.

Can gluten sensitivity be confirmed with a blood test?

Not yet. Unlike coeliac disease, there is no validated biomarker for non-coeliac gluten sensitivity. Diagnosis still relies on a double-blind, placebo-controlled gluten challenge, which is difficult to run outside of research settings.

References