Non-Celiac Gluten Sensitivity: What It Is, and What Nobody Can Test For Yet

Your celiac test came back negative, and bread still makes you feel terrible. That gap has a name in the medical literature: non-celiac gluten sensitivity. It is real enough to have a name, a consensus paper, and a research field. It is also unsettled enough that the scientists working on it are still arguing about whether "gluten" is even the right word in it. There is no blood test. It is diagnosed by ruling two other things out first.

Here is the message I get most often, and I have stopped being surprised by it.

"They tested me for celiac. Negative. My doctor told me it was probably stress. But I know what happens when I eat bread, and I am not imagining it."

I am not going to tell you what is happening in your body, because I do not know, and neither does anyone who tells you they do with confidence. What I can do is walk you through what researchers have actually found, what they have not, and where the honest gaps are. Then I will show you the one thing in all of this you can genuinely act on today, which has nothing to do with buying anything from me.

Table of contents

1. Three different conditions people call "gluten problems"
2. What non-celiac gluten sensitivity actually is
3. How it gets diagnosed, and why that order matters
4. Celiac disease is getting more common, and not only because we test better
5. Is it even gluten? The honest state of the argument
6. What the evidence does not show
7. Does the kind of wheat make a difference? What has actually been measured
8. If you are choosing flour and pasta anyway
9. Frequently asked questions
10. What I actually think

Three different conditions people call "gluten problems"

When someone says they cannot eat gluten, they could mean any of three separate conditions. They involve different parts of the immune system, they do different things to the gut, and only two of them can be confirmed by a test.

Sorting them out is the whole foundation of this article, so here they are side by side.

Frequencies as reported in Marco Pallotti, Grani Antichi, Figure 28, printed page 78, citing Volta 2011 and Catassi et al. 2013. The question mark on the last figure is the book's own, printed inside the figure. I have left it there because it is the most honest thing on the chart. These are population statistics, not personal risk.

That last column is the subject of this article, and the empty cell in the "confirming test" row is the reason it is difficult to write about.

What non-celiac gluten sensitivity actually is

Researchers describe it as a pattern rather than a finding. Symptoms appear within hours to a few days of eating gluten, clear when gluten is removed, and return when it comes back, in people who have already tested negative for celiac disease and wheat allergy.

The symptoms fall into two clusters. Inside the gut, researchers have catalogued abdominal pain, bloating, meteorism, altered bowel transit, diarrhea, constipation, or bowel habits that alternate between the two. Outside the gut, the literature records headache, what the Italian source calls mente annebbiata or brain fog, numbness in the limbs, joint and muscle pain, dermatitis and eczema and rash, depression, fatigue, and anemia. In children the picture leans toward the gut and looks more like gastroenteritis. In adults the symptoms outside the gut are more common.

I want to be careful about how you read that list, because a list like that is not a diagnostic checklist and I refuse to present it as one. It is a catalogue of what clinicians observed in patients who had already been through testing. Almost every symptom on it has dozens of possible causes, and several of them are things you would want a doctor to look at for entirely different reasons. Anemia and chronic fatigue are not things to self-manage with a diet change.

For a long time these patients had nowhere to go. The book I have been reading for this describes the condition as a terra di nessuno, a no man's land, and the reason is specific and a little bleak: the celiac specialists did not take these patients because their tests were negative, and the irritable bowel specialists did not take them either, even though the symptoms overlapped. Patients were sometimes referred to psychiatry, for depression that the situation itself was producing.

If you have been told your symptoms are stress, that history is worth knowing. It is not a diagnosis and it does not mean anything about your case. But you are not the first person to be handed that answer, and the field has moved since.

How it gets diagnosed, and why that order matters

There is a defined sequence, and it runs in one direction. This is a description of what clinicians do, not instructions for you.

Celiac disease is ruled out first. That means serology, and then, in most pathways, a duodenal biopsy. My source is explicit that biopsy "is still today considered fundamental" for a definitive diagnosis, treating the blood tests as a screening advance layered on top rather than a replacement. Some newer pediatric pathways allow a biopsy-sparing route in specific circumstances, which is a development after that book's citation window.

Wheat allergy is ruled out second. Standard allergy testing.

What is left is a diagnosis of exclusion. No antibody, no biopsy finding, no genetic marker confirms non-celiac gluten sensitivity. Both my source and current clinical consensus agree on this, and it is one of the places where the book is at its most reliable. What is described as a diagnosis is really a description that fits after other things have been ruled out.

Why you should not cut gluten before you are tested

This is the one genuinely actionable thing in this article, and my source does not say it, so I am adding it from standard clinical practice.

Celiac serology looks for antibodies your immune system produces in response to gluten. If you have already removed gluten from your diet, those antibodies fall, and the test can come back negative in someone who actually has celiac disease. The same applies to the biopsy. You can lose the answer by going looking for it.

So if you suspect something and you are planning to get tested, the order is: talk to your doctor first, get tested while you are still eating normally, and change the diet after. Not before. I have watched people spend two years in the dark because they did it the other way round, and it costs nothing to get right.

Celiac disease is getting more common, and not only because we test better

The intuitive explanation for rising celiac diagnoses is that we simply got better at finding it. Before the 1980s, celiac disease was thought to be rare, almost exclusively a childhood condition, and largely European. Diagnosis required gut symptoms plus a biopsy showing villous atrophy, which is a definition narrow enough to produce a low number mechanically. Then serological testing arrived and the number went up.

That explanation is widely believed, and it does not survive contact with two datasets.

The 1948 blood samples nobody expected to still exist

A 2009 analysis (Rubio-Tapia et al.) found more than 9,000 blood samples collected between 1940 and 1954 from recruits at Warren Air Force Base in Wyoming, still in storage. Mean age 37. Researchers ran modern celiac serology on them and compared the result against two present-day groups, one of 5,500 people born in the same era and one of 7,200 men matched to the same mean age.

The historical samples showed 0.2% seroprevalence. The two modern groups showed 0.8% and 0.9%. Same test, same laboratory methods, different decades. That is roughly a fourfold difference, and better detection cannot explain it, because the same test was used on all three groups.

Finland, twenty years apart

A 2007 Finnish study (Lohi et al.) did the same thing on a shorter timescale. About 7,200 adults sampled between 1978 and 1980, compared against 6,700 sampled in 2000 and 2001. Marker positivity went from 1.05% to 1.99%. Nearly doubled, in twenty years.

The iceberg

The other thing that changed is what celiac disease is understood to look like. There are five recognised clinical forms, and the classic one, the childhood case with obvious digestive symptoms, turns out to be the least frequent once the atypical presentations are counted. The rest sit below the waterline: silent cases with gut damage and no symptoms, latent cases with symptoms and positive blood work but a normal biopsy, and potential cases with positive serology and nothing else yet.

Diagnosis is also increasingly made in adulthood now, including in older patients, and the male-to-female ratio runs about 1 to 2.5.

I want to be careful here, because this is the section where an article like this usually starts pointing at a villain. The genetics are necessary but not sufficient: you are not born with celiac disease, you are born with a predisposition that may or may not be triggered, and the triggers named in my source are things like stress, intestinal infection, and pregnancy, plus what it honestly calls factors not yet known. That is where the evidence stops. It does not name a cause, and neither will I.

Is it even gluten? The honest state of the argument

This is the part of the topic that almost nobody writes about, and it is the most interesting thing in the entire field.

"Non-celiac gluten sensitivity" is a provisional label. The researchers who use it say so. The name was chosen before anyone had established what the trigger actually is, and there are currently three candidates.

Candidate one: gluten

The obvious one, the one the condition is named after, and the one with the least direct evidence behind it as a specific trigger. Which is not the same as saying it is wrong.

Candidate two: ATI, the wheat proteins that are not gluten

Amylase/trypsin inhibitors are a family of wheat proteins that resist being broken down in the intestine. Laboratory work (Junker et al. 2012; Bucci et al. 2013) has shown they can activate the innate immune system through monocytes, macrophages, and intestinal dendritic cells. That is a genuine and frequently cited immunology finding.

It is also in vitro. Cells in a dish, not people eating bread. Researchers propose it as a mechanism. Nobody has demonstrated it explains human symptoms, and treating a cell-culture result as a confirmed clinical explanation is exactly the leap this field keeps being asked not to make.

Candidate three: FODMAPs, and the trial that complicates everything

FODMAPs are a group of fermentable carbohydrates found in wheat and in plenty of other foods. Here the evidence gets genuinely contested, and I am going to report the disagreement rather than pick a side, because the disagreement is the honest answer.

My source cites a 2013 crossover trial (Biesiekierski et al.) to argue that FODMAPs are unlikely to be the main driver, on the grounds that symptoms resolved on a gluten-free diet that still contained FODMAPs from legumes. Much of the field reads that same trial the opposite way: once FODMAPs were controlled for, a gluten-specific challenge produced no more symptoms than placebo, which is usually taken as evidence against a gluten-specific mechanism and in favour of FODMAPs, expectation effects, or both.

I am not qualified to referee that, and I am not going to pretend the paper says what suits me. Two competent readings of the same trial disagree. That is a fair summary of where this field is, and you deserve to know it.

Why some researchers want to rename it

Because of all this, one group has proposed dropping "gluten" from the name entirely and calling it non-celiac wheat sensitivity. Their argument is that molecules other than gluten are probably involved, and that a name pointing at one molecule discourages people from looking at the others.

What the evidence does not show

Everything above is real, sourced, and page-cited. Now let me be equally clear about what none of it establishes, because articles on this topic have a habit of stacking true facts until the reader draws a conclusion that no sentence actually made.

• Nothing here identifies a cause of the rise in celiac diagnoses. The seroprevalence studies measure that it happened. They do not say why, and they do not point at any agricultural practice, farming method, or wheat variety. Anyone who tells you those studies name a culprit has not read them.
• No study in this material tested whether eating one kind of wheat rather than another changes anyone's symptoms. Not one. That trial does not exist in what I have read.
• Non-celiac gluten sensitivity has no biomarker, which means that even if someone wanted to run that trial cleanly, they would have to define the condition by symptom report first. That is a real and acknowledged obstacle in the field, not a technicality.
• The prevalence figure is not established. You will see "6% of the population" and similar numbers on the internet. My source prints "greater than 1%" with its own question mark attached. There is one study of 920 patients with irritable-bowel-type symptoms in which 30% turned out to have wheat sensitivity or multiple food hypersensitivity, and that number gets quoted constantly. It was a gastroenterology referral population, not the general public, and the book that cites it hedges it conditionally in the same sentence.
• The ATI mechanism is proposed, not demonstrated.

None of this means your symptoms are not real. It means the science has not caught up to them yet, and pretending otherwise would be a different kind of disrespect.

Does the kind of wheat make a difference? What has actually been measured

There is one piece of research that speaks directly to sensitivity and to specific wheats, and because I sell some of those wheats, I am going to describe it more carefully than I would otherwise.

The CXCL10 study: blood cells, not dinner

Valerii et al., 2015, as reported in Grani Antichi, printed page 101. A follow-up in children (Alvisi et al., 2017) found the same pattern and describes itself in its own title as a pilot study. These findings come from independent research and do not guarantee the same results with our product.

The most quoted line from that study is that khorasan produced less of the inflammatory signal. The more interesting line, and the one almost nobody repeats, is that the effect was independent of gluten content. That is why the same researchers proposed renaming the condition after wheat rather than after gluten. Read properly, this study is weaker evidence for ancient grain than it looks and stronger evidence that we have been blaming the wrong molecule.

The contradiction nobody in this niche mentions

There is a related question about whether older wheats carry fewer of the gluten fragments that celiac immune systems react to, and here the published evidence points in two directions at once.

A 2010 molecular analysis (van den Broeck et al.) compared 36 modern European varieties against 50 ancient and traditional ones grown until the early 1900s, and found one particular reactive fragment noticeably more common in the modern group. Its own title says breeding "may have contributed" to rising celiac prevalence.

Then a 2016 study (Ribeiro et al.) looked at 126 varieties and found that ancient soft-wheat landraces and spelt actually carried higher levels of the five toxic fragments it studied than the modern varieties did.

Both papers exist. Both are peer reviewed. Every page in my corner of the internet cites the first one and none of them cites the second, and I think that tells you something about how this subject usually gets written. It is unresolved. I would rather hand you both than sell you one.

If you are choosing flour and pasta anyway

Nothing above tells you what to buy. I want that stated plainly before I say a word about my own products, because the research in this article does not support a purchase decision and I am not going to pretend it does.

What I can talk about is labels, because that is a different kind of question and it has actual answers.

What "stone ground" on the front of a bag does not tell you

Here is something I had to learn properly myself, and it changed how I describe my own products.

Stone milling describes how the grain is ground. The whole kernel goes between the stones. But then comes the step that actually decides what ends up in your bag: if the meal is sifted afterward, the bran and the germ come out with the siftings. That is how semola has always been made, and it is a completely legitimate way to make pasta. It is also why "stone ground" and "whole grain" are two different claims that have to be checked separately. I went further into how grinding actually works in what stone ground really means.

So check the bag. Mine included.

• My Khorasan rigatoni, my Khorasan pastina and my Tumminìa pasta are whole grain. The proof is on the back of the bag, not the front: the ingredient line reads organic whole wheat semolina, one ingredient, nothing taken out and nothing added back.
• My einkorn flour is whole grain the same way. One ingredient on the label, whole grain einkorn.

I had that pasta tested at an independent laboratory in Minneapolis, and the numbers agree with the labels. Millers measure how much bran is left in a flour by burning a sample and weighing the mineral ash behind it: more bran, more ash. My Tumminìa came back at 1.95% ash on a dry-matter basis with 9.5 grams of fibre per 100 grams, and my Khorasan at 1.61% with 7.4 grams. Those sit where a wholemeal should. They are single lots, tested once, so read them as a snapshot of what I received rather than a specification.

What "ancient grain" on a label does and does not guarantee

Five things worth knowing, and none of them are about my products specifically.

1. It is not a regulated term. No agency defines it. Anyone can print it.
2. It never means gluten-free. Einkorn, emmer, spelt, khorasan and durum are all wheat, and all contain gluten.
3. It says nothing about milling. Whole or sifted is a separate question, and it is the one that changes what is actually in the bag.
4. It says nothing about where the grain was grown, or by whom.
5. It can be a modern registered variety. The Saragolle are centuries-old southern Italian landrace populations. Saragolla, singular, is a durum registered in 2004. Same word on a label, different thing in the bag. I wrote about that at more length in what modern wheat breeding actually changed.

Which brings me to the only shopping advice in this article, and it applies to every brand including mine: turn the bag over. "Stone ground" on the front tells you about a machine. The ingredient line tells you what is inside. If it says whole wheat or whole grain, the bran and germ are in there. If it says semolina with no "whole" in front of it, they were sifted out, no matter how the grain was ground.

The line that does not move

None of these grains are food for someone with diagnosed celiac disease. Not einkorn, not emmer, not spelt, not khorasan, not the heritage durums.

I want to be firmer about this than my own source is, and my source is already firm. It states it separately and explicitly for every single grain it discusses: einkorn cannot be eaten by people who have already developed celiac disease, emmer must not be included in their diet, spelt likewise, khorasan likewise. It even documents a case where a diagnosed celiac patient introduced spelt into a gluten-free diet and relapsed.

If you have celiac disease, the answer is a strict gluten-free diet, and nothing I sell belongs in it. I would rather lose the sale than be unclear about that.

Frequently asked questions

What is non-celiac gluten sensitivity?

A symptom pattern that appears within hours to a few days of eating gluten, clears when gluten is removed, and returns when it comes back, in people who have tested negative for celiac disease and wheat allergy. It has no confirming blood test.

Is non-celiac gluten sensitivity real?

It is a recognised clinical category with an active research literature and a consensus nomenclature paper behind it (Sapone et al., 2012). What remains genuinely unsettled is how common it is and what triggers it, not whether patients are experiencing something.

How is it different from celiac disease?

Celiac disease is autoimmune, involves both innate and adaptive immunity, damages the intestinal lining, and can be confirmed by serology plus biopsy. Sensitivity is believed to involve only innate immunity, produces no villous atrophy, and has no identified autoantibody. The literature calls them two sisters with real differences.

Is there a test for gluten sensitivity?

No. As of the research reviewed here, no antibody, biopsy finding, or genetic marker confirms it. It is diagnosed by excluding celiac disease and wheat allergy first. Anyone selling you a test for it is ahead of the science.

Can I have gluten sensitivity if my celiac test was negative?

That negative result is the starting point of the assessment, not the end of it, which is exactly why the sequence matters. This is a conversation for your doctor.

Should I go gluten-free to find out?

Not before testing. Removing gluten can make celiac serology come back falsely negative, which means you can lose the answer by looking for it. Talk to your doctor about the right order.

Do ancient grains help with gluten sensitivity?

No study has shown that. What exists is narrower: in one 2015 experiment, blood cells taken from 48 people with sensitivity released less of an inflammatory signal when exposed to proteins from khorasan and Senatore Cappelli than from modern wheats. That is cells in a laboratory, not people eating pasta.

Can people with celiac disease eat ancient grains?

No. Einkorn, emmer, spelt and khorasan are all wheat and all contain gluten. The source behind this article says so explicitly and separately for every one of them. I went through the einkorn case in detail in is einkorn flour gluten-free.

Why did I feel fine eating pasta in Italy?

This is the question I get more than any other, and it deserves its own answer rather than a guess tacked on here. I wrote it up separately in why Italy feels different for gluten-sensitive travellers. The short version is that several things change at once when you travel, and no single one of them has been isolated as the reason.

What I actually think

I think the people writing to me are describing something real, and I think the science has not caught up to it. Those two things are allowed to both be true.

What bothers me about how this subject gets written is that both of the available answers online are dishonest in opposite directions. One says gluten is fine and you are imagining it. The other says modern wheat is poisoning you, and here is the bag to buy. The actual literature says: something is going on, we have three candidate explanations, the best blinded trial keeps failing to pin it on gluten specifically, and there is no test yet. That is less satisfying and much more useful.

So here is what I would do in your position. Get tested before you change anything. Take the sequence seriously, because it only works in one direction. And when you do shop, turn the bag over and read the ingredient line, because that is the one thing on any package that has to tell you the truth.

I am not a doctor. This is my food and my story. Talk to your doctor, and read your labels.

Disclaimer: These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. Individual results may vary. Please consult a healthcare professional before making changes to your diet or wellness routine. These findings come from independent research and do not guarantee the same results with our product.

Disclaimer: the content in this blog is for informational purposes only and should not be taken as professional, medical advice.