The Wrong Molecule

Sometime in the last fifteen years, bread became suspect. A food that has anchored the human diet for ten millennia — that built the granaries of Mesopotamia, financed Rome, fed the cathedrals of medieval Europe — was suddenly accused of making people sick in numbers no one could quite explain. Patients arrived in gastroenterology clinics with the same complaint: bloating after pasta, fog after a sandwich, fatigue after pizza. They were tested for celiac disease and came back negative. They were tested for wheat allergy and came back negative. And then they were sent home with a shrug.

Many of them, scrolling the internet on the train ride back, found a confident answer. It’s the gluten. Cut it out, the forums said, and you will feel like yourself again. So they did. And, often, they felt better.

Which, on the face of it, would seem to close the case. Symptoms disappear when the suspect is removed; the suspect is guilty. But in the careful machinery of clinical science, that kind of reasoning is what blind trials are designed to interrogate. And when researchers ran those trials, beginning around 2011 and intensifying through the decade that followed, something disorienting happened. The people who reported feeling sick from gluten kept feeling sick. The diet that helped them kept helping them. But gluten itself, isolated and tested under controlled conditions, kept failing to behave like a culprit. The relief was real. The molecule was wrong.

A new diagnosis, almost as soon as it was named

True gluten disease — celiac — is not new. Physicians have described its symptoms since antiquity; Aretaeus of Cappadocia wrote about a wasting illness he called koiliakos in the second century, and the modern understanding of celiac as an autoimmune reaction to gluten was firmly established by the late twentieth century. It is a serious, lifelong condition in which the small intestine attacks itself in the presence of even trace amounts of wheat, rye, or barley protein. It affects roughly one percent of the global population.¹

What changed in the 2000s was the appearance, in unusually large numbers, of patients who did not have celiac and did not have IgE-mediated wheat allergy, but who nonetheless reacted badly to wheat. Their endoscopies looked normal. Their antibody panels were clean. And yet their lives had been narrowed by an unexplained intolerance. By 2011, an international panel of experts had given the condition a working name: Non-Celiac Gluten Sensitivity, or NCGS.² The label was deliberately provisional. No one had identified the molecule responsible. No one had a biomarker. Doctors could only listen, exclude celiac, and recommend an experimental gluten-free trial.

Meanwhile, the marketplace did what the marketplace does. Gluten-free aisles expanded from a corner of the health-food store into a category of their own. By 2015, the global gluten-free market had grown to roughly fifteen billion dollars, and surveys suggested that as many as one in five American adults was actively avoiding gluten — a population vastly larger than the one percent who actually had celiac disease.³ A diagnosis that had barely existed a decade earlier was now reshaping menus, supply chains, and family dinners.

It also produced something rarer in modern nutrition science: a researcher willing to look hard at evidence he had himself produced, and to wonder if he had gotten it wrong.

The Australian who doubted his own study

Peter Gibson is a gastroenterologist at Monash University in Melbourne, and the laboratory he runs with Jessica Biesiekierski had, in 2011, published one of the first rigorous trials of NCGS. The design was elegant in its simplicity. They recruited adults who said gluten made them sick, confirmed they did not have celiac, and then fed some of them gluten and others a placebo, in a double-blind, randomized fashion. The participants who got gluten reported significantly more symptoms — pain, bloating, fatigue — than those who got placebo. The finding seemed to anchor the new diagnosis in real biology.⁴

It also helped catalyze the gluten-free boom. But Gibson, by his own later account, was uncomfortable. The effect sizes were modest. The mechanism was murky. And in clinical practice, his patients were eating not isolated gluten but bread, pasta, breakfast cereal — foods that contained dozens of other potentially disruptive compounds. He suspected his first study had not accounted for everything that comes with wheat.

So in 2013, he ran the trial again, this time with what he called a much tighter protocol.⁵ Before any gluten was reintroduced, every participant was placed on a low-FODMAP diet — an acronym, coined at Monash, that stands for fermentable oligosaccharides, disaccharides, monosaccharides, and polyols. FODMAPs are a family of short-chain carbohydrates found in onions, garlic, beans, certain fruits, dairy, and, importantly, wheat. They are poorly absorbed in the small intestine and travel onward to the colon, where resident bacteria ferment them, producing gas, distension, and in some people the kind of cramping and bloating that feels indistinguishable from a food intolerance.

With the background diet stripped of FODMAPs, Gibson’s team then secretly gave their volunteers either gluten, whey protein, or a true placebo. The results were not what he had expected. Symptoms no longer tracked with gluten. Participants reported feeling sick across all three arms of the study, including when they were eating nothing offensive at all. The signal that had looked so clear two years earlier had vanished into the noise.

The most plausible interpretation was that the original study had measured something real — but had named it incorrectly. The real trigger had been hiding in the background diet all along: the fructans in wheat, a kind of FODMAP, fermenting in the gut. When the team removed fructans, gluten alone could not reproduce the misery.

What the fructans were doing

A fructan is, in essence, a short chain of fructose molecules with a glucose cap. Wheat contains relatively modest amounts by weight, but it is consumed in such volume in Western diets — in bread, pasta, pizza, pastries, cereal — that it becomes a leading dietary source. Once a fructan reaches the colon, the bacteria that live there fall on it eagerly. The fermentation produces short-chain fatty acids, which are largely beneficial, and gases, which are not always.⁶ In a person with a normally elastic and unbothered gut, the gas dissipates. In a person with a sensitized gut — perhaps from a past infection, perhaps from an altered microbial population, perhaps from causes that remain unclear — the same volume of gas registers as pain.

The fructan hypothesis was not Gibson’s alone. In 2018, a team led by the Norwegian clinical nutritionist Gry Skodje at the University of Oslo ran what may be the cleanest test of the question to date. They recruited fifty-nine adults with self-diagnosed gluten sensitivity, all of whom had been thriving on gluten-free diets, and challenged them with three identical-looking muesli bars: one containing gluten, one containing fructans, and one containing only a placebo. Neither the participants nor the researchers knew which was which. Each volunteer rotated through all three.⁷

The fructan bars produced significantly more gastrointestinal symptoms than the gluten bars did. The gluten bars, in fact, were statistically indistinguishable from placebo. A condition that had been named after gluten was, in its most carefully designed test, not behaving like a gluten condition at all.

For the patients, this was good news in a counterintuitive sense. Their symptoms were not imaginary; the trial reproduced them. But the molecule they had been told to fear was probably not the one harming them. A different and more tractable explanation was beginning to take its place.

A second suspect, hiding in the protein

If fructans were one part of the story, immunologists were quietly building the case for another. Wheat is more than starch and gluten. Like most cultivated grains, it produces a small family of defensive proteins meant to fend off insects and fungi. Among the most studied are the amylase-trypsin inhibitors, or ATIs — proteins that interfere with the digestive enzymes of would-be pests. They make up a small fraction of wheat protein, perhaps two to four percent, but they are remarkably bioactive.

Detlef Schuppan, an immunologist at the Mainz University Medical Center in Germany and a longtime researcher of celiac disease, became interested in ATIs in the early 2010s. His lab showed that when ATIs reached the intestines of mice, they activated the innate immune system through a receptor called TLR4, triggering a low-grade inflammatory response that did not look like celiac but did not look like nothing, either.⁸ In follow-up work, Schuppan suggested that ATIs might not only contribute to NCGS-like symptoms but could also worsen inflammation in unrelated autoimmune conditions elsewhere in the body.⁹

There is a second wrinkle: modern bread wheat, bred over the last century for higher yields and stronger gluten networks suitable for industrial baking, appears to contain measurably more ATIs than the older landrace varieties it replaced. The difference is not enormous, and the field is careful not to romanticize ancient grains, which still contain plenty of fructans and gluten of their own. But it is at least suggestive that a slow shift in the chemistry of the loaf has accompanied the rise of a diagnosis that did not exist when our grandparents were eating bread.

So the picture that has emerged is messier than the gluten-free aisle implies. Some people may react to fructans; some to ATIs; some, almost certainly, to both, or to other wheat components not yet characterized. Pure gluten, in trial after trial, fails to behave like the master villain it has been cast as. The diet works because it removes the actual triggers along with the imagined one.

The cost of solving the wrong problem

If gluten-free living were costless, the misdiagnosis would be of academic interest only. It is not. A strict gluten-free diet, especially one built around packaged substitutes, tends to be lower in fiber, iron, folate, and several B vitamins than the diet it replaces.¹⁰ Gluten-free breads are often more refined, with shorter ingredient lists dominated by rice flour and starch. Long-term gluten-free eaters who do not have celiac disease have shown, in some studies, modest shifts in cardiovascular risk markers and changes in gut microbial composition whose significance is still being worked out.

There are social costs too. Restaurants are navigated warily. Dinner parties become negotiations. Children’s birthdays carry a low-grade tension. A person who has been told that a ubiquitous food is poison to them lives differently from a person who has been told they react to a specific carbohydrate that can be managed with more flexibility. The first framing offers identity. The second offers options.

And the alternative, if fructans are the real issue, is genuinely different. A low-FODMAP diet, used clinically for irritable bowel syndrome, is typically run as an elimination-and-reintroduction protocol over a few months, with the goal of identifying which specific fermentable carbohydrates a given person tolerates and which they do not.¹¹ Many people who believed they could never eat bread again discover, under that protocol, that they tolerate sourdough — whose long fermentation depletes fructans — perfectly well. Others find their thresholds: a few crackers, fine; a bowl of pasta, not. The diagnosis becomes a dial, not a wall.

Renaming what we don’t yet understand

The field has, in its quiet way, been trying to correct itself. In recent consensus statements, many researchers have moved away from Non-Celiac Gluten Sensitivity toward the more honest Non-Celiac Wheat Sensitivity, or NCWS.¹² The new name concedes what the old one denied: that the offending molecule has not yet been identified, that wheat is a chemical orchestra rather than a single instrument, and that calling the condition by the name of a protein we cannot pin to it was always more confident than the data warranted.

None of this means that the patients were confused, or that the relief they found was placebo. The puzzle of NCWS is not that people feel better off wheat; that part is well-attested. The puzzle is that the body’s protest has been hard to map onto any single molecule, and that the simplest story — gluten is bad — has held the field hostage longer than the evidence supported. Science is often slower than diet trends. It is also, eventually, more interesting.

There is one final thing worth saying about the people in those clinics, the ones who described their bread-induced fog and were sent home with a shrug. The trials of the last decade have, in a strange way, vindicated them. Something in wheat is making them ill. Their gut is reporting a real signal. The signal has been misread, by them and by the doctors who first named the condition, but it was never imaginary. The investigation is not over. It is just being conducted, finally, with the right suspects in the room.

Which is a useful kind of humility to bring to a plate of food. The body knows when something is wrong. It is not always good at telling us what. Between the sandwich and the symptom lies a far longer chain of biology than any label on a loaf can capture, and the honest answer, for now, is that we are still learning to listen to what the gut is actually saying.

Sources

1. Singh P. et al., Global Prevalence of Celiac Disease: Systematic Review and Meta-Analysis, Clinical Gastroenterology and Hepatology, 2018. — https://www.cghjournal.org/article/S1542-3565(17)30783-8/fulltext
2. Sapone A. et al., Spectrum of gluten-related disorders: consensus on new nomenclature and classification, BMC Medicine, 2012. — https://bmcmedicine.biomedcentral.com/articles/10.1186/1741-7015-10-13
3. Reilly N. R., The Gluten-Free Diet: Recognizing Fact, Fiction, and Fad, Journal of Pediatrics, 2016. — https://www.jpeds.com/article/S0022-3476(16)00088-X/fulltext
4. Biesiekierski J. R. et al., Gluten Causes Gastrointestinal Symptoms in Subjects Without Celiac Disease, American Journal of Gastroenterology, 2011. — https://journals.lww.com/ajg/Abstract/2011/03000/Gluten_Causes_Gastrointestinal_Symptoms_in.41.aspx
5. Biesiekierski J. R., Peters S. L., Newnham E. D., Rosella O., Muir J. G., Gibson P. R., No effects of gluten in patients with self-reported non-celiac gluten sensitivity after dietary reduction of fermentable, poorly absorbed, short-chain carbohydrates, Gastroenterology, 2013. — https://www.gastrojournal.org/article/S0016-5085(13)00702-6/fulltext
6. Gibson P. R., Shepherd S. J., Evidence-based dietary management of functional gastrointestinal symptoms: The FODMAP approach, Journal of Gastroenterology and Hepatology, 2010. — https://onlinelibrary.wiley.com/doi/10.1111/j.1440-1746.2009.06149.x
7. Skodje G. I. et al., Fructan, Rather Than Gluten, Induces Symptoms in Patients With Self-Reported Non-Celiac Gluten Sensitivity, Gastroenterology, 2018. — https://www.gastrojournal.org/article/S0016-5085(17)36302-3/fulltext
8. Junker Y. et al., Wheat amylase trypsin inhibitors drive intestinal inflammation via activation of toll-like receptor 4, Journal of Experimental Medicine, 2012. — https://rupress.org/jem/article/209/13/2395/41540/Wheat-amylase-trypsin-inhibitors-drive-intestinal
9. Zevallos V. F. et al., Nutritional Wheat Amylase-Trypsin Inhibitors Promote Intestinal Inflammation via Activation of Myeloid Cells, Gastroenterology, 2017. — https://www.gastrojournal.org/article/S0016-5085(16)35452-8/fulltext
10. Vici G. et al., Gluten free diet and nutrient deficiencies: A review, Clinical Nutrition, 2016. — https://www.clinicalnutritionjournal.com/article/S0261-5614(16)30062-5/fulltext
11. Halmos E. P. et al., A diet low in FODMAPs reduces symptoms of irritable bowel syndrome, Gastroenterology, 2014. — https://www.gastrojournal.org/article/S0016-5085(13)01407-8/fulltext
12. Catassi C. et al., The Overlapping Area of Non-Celiac Gluten Sensitivity (NCGS) and Wheat-Sensitive Irritable Bowel Syndrome: An Update, Nutrients, 2017. — https://www.mdpi.com/2072-6643/9/11/1268