Durum or Bread Wheat: What Is Actually in Your Pasta

There are two different species of wheat in the pasta aisle, and the box does not always tell you which one you are holding. In Italy the law settles it for you. In America it does not, and the ingredient line is the only place it is written down. Here is how to read it in four seconds, and the research behind why the difference might matter, including the parts that argue against my own pasta.

The question that sent me looking

Ask me what I get asked most and it is not about olive oil, and it is not about which shape goes with which sauce. It is this, and it arrives almost every time in the same words: I cannot eat wheat at home, I went to Italy, I ate pasta and bread every day for two weeks, and nothing happened. Then I came home, and it started again.

People want me to tell them why. I cannot, and that has bothered me for a long time.

It is not a rare question. It is one of the most common things people type into Google before they land on my website. They ask whether pasta in Italy is different. They ask whether gluten itself is different in Italy. They ask what Italy does to its wheat. They ask, in the plainest words they have, why they were fine there and are not fine here.

I sell pasta for a living, so you should assume I would love the answer to be "because of how the wheat is grown in Sicily." I have spent a long time now reading the actual research, and I cannot tell you that. What I can tell you is that there are six serious explanations, that scientists disagree about them, and that the honest state of this question is unsettled.

So I am going to walk you through all six. I am going to start with the two that would make my pasta no different from anybody else's, because those are the ones an article like this usually buries. Then I will show you the ones that are more flattering to me, along with the studies that argue against them.

Before I go one step further: this is wheat, and wheat has gluten

The rest of this page is about non-celiac gluten sensitivity and the broader thing researchers now call non-celiac wheat sensitivity: people who react to wheat, whose celiac tests come back negative, and for whom no diagnostic test exists. If that language is new to you, I wrote a longer piece on what non-celiac gluten sensitivity actually is and what nobody can test for.

Semolina or wheat flour: the four-second label test

Most of this page is honest about how much nobody knows. So let me put the part that is not in doubt right at the front, because it is written into law in two countries and you can verify it in an afternoon.

In Italy, dry pasta has to be made from durum wheat semolina. That is D.P.R. 9 febbraio 2001, n. 187, and it allows only a small tolerance for common wheat getting in by accident.

In America there is no such rule. The federal standard of identity, 21 CFR 139.110(a), defines macaroni products as made from "semolina, durum flour, farina, flour, or any combination of two or more of these, with water." Semolina and durum flour are durum, the pasta wheat. Farina and flour are common bread wheat, a different species entirely. The same body of US rule is also why American pasta is enriched in the first place.

So a box of pasta in Italy is durum because the law says so. A box in America is durum only if the company decided it should be. Which makes the ingredient line the only place you find out.

That is not a small difference and it is not a matter of opinion. It is two sentences of statute. If you do nothing else with this article, do that one thing the next time you are in front of a shelf.

Maybe it was never the gluten: wheat fructans

Here is the first one, and it is the one that would put me out of the argument entirely.

The point of this section in one sentence: the thing you think you react to may not be the thing you react to.

Wheat carries two suspects. Gluten is a protein. Fructans are a carbohydrate, a type of FODMAP, and plenty of people do not absorb them well. Fructans are in all wheat: ancient, modern, Sicilian, Canadian, mine, everyone's. When researchers separated the two and tested them blind, the carbohydrate produced the symptoms and the protein did not.

In 2013 a research group in Australia took thirty-seven people who were certain gluten was their problem, put everyone on a diet low in fermentable carbohydrates first, and then challenged them with gluten without telling them when. This was a double-blind, placebo-controlled crossover trial in humans, which is about as good as this kind of study gets. Once the fructans were controlled, gluten produced no specific and no dose-dependent worsening of symptoms. People also got worse on whey protein, which the authors could not explain (Biesiekierski et al., Gastroenterology, 2013;145(2):320-328).

Five years later a Norwegian group went further and tested gluten and fructan separately in the same fifty-nine blinded people. Fructan produced significantly worse symptoms than gluten. Gluten was indistinguishable from placebo. On the bloating score specifically: gluten 9.3, placebo 10.1, fructan 11.6 (Skodje et al., Gastroenterology, 2018;154(3):529-539).

Read that middle number again. On the exact symptom this whole conversation is built around, gluten scored slightly better than a dummy bar.

Now the part that is uncomfortable for me. If fructans are the answer, then the wholemeal pasta I make is on the wrong side of it. Fructans concentrate in the outer layers of the grain, so wholegrain flour carries more of them than white flour does. A 2025 laboratory analysis of thirteen commercial pastas found the refined durum samples among the lowest, and the wholegrain ancient-grain samples among the highest (Torbica et al., Foods, 2025;14(4):667). That paper also found that boiling washes a large share of those fructans into the water you pour down the sink, between roughly a third and four fifths depending on how much water you use and how long you cook. But its own headline is that almost all wheat pasta still classifies as high in FODMAPs.

These are measurements of food on a laboratory bench, and not one of them measured a person in Italy. But here is what this explanation would mean if it turns out to be yours: gluten was never the lever. Dose was, and cooking water, and how long a dough was left to rise. All three of those change on holiday without anyone deciding they should.

These findings come from independent research and do not guarantee the same results with our product.

Maybe it was never the food: the holiday effect

The second explanation is worse for me than the first, because it says the plate had nothing to do with it.

Think about what actually changed on that trip. You slept more. You walked ten miles a day instead of sitting. You ate at a table with people instead of standing at a counter. You drank more water and probably more wine. You were not braced for a reaction, because you were on holiday and had decided not to think about it. Every one of those things moves digestion, and they all moved at once.

And there is a measured effect underneath this. When researchers pooled the double-blind gluten challenges that had actually been run, only 38 of 231 patients showed symptoms that tracked gluten specifically, and roughly forty percent of those showed a nocebo response, meaning they felt the same or worse on the placebo (Molina-Infante and Carroccio, Clinical Gastroenterology and Hepatology, 2017;15(3):339-348). In the Norwegian trial above, twenty-two of the fifty-nine participants felt worst of all on the bar that contained neither gluten nor fructan.

Nearly four in ten people peaked on nothing.

I am not saying your symptoms are imaginary, and I want to be careful here, because that is what people with this condition get told constantly and it is both cruel and wrong. Expectation is not imagination. It is a real physiological input, it is measurable, and it works in both directions. In Italy, everything about the context was telling your body to relax. That is a genuine candidate explanation and it deserves to sit near the top of the list.

What nobody has done is the study that would settle it: follow gluten-sensitive travellers to Italy and measure something. There is no cohort, no diary study, no controlled re-exposure. The starting observation for this entire article, including every version of it I have been told in person, is anecdote. Anecdote is a good reason to ask a question. It is not an answer.

Maybe it is the same wheat: Italy imports durum too

This is the one that surprises people most, and I think it is the most useful thing on this page.

Italy does not grow enough durum to supply its own pasta industry. It is one of the largest wheat importers in the world, and a substantial share of the durum milled in Italy arrives from Canada, the United States, and elsewhere. The trade figures behind that come from re-tabulations of data from ISMEA, Italy's public agri-food market institute, rather than from a border record.

Italian law is strict about the species in your pasta and silent about the country. Under D.P.R. 187/2001, dry pasta produced and sold in Italy has to be made from durum wheat semolina, with only a small tolerance for common wheat as cross-contamination. There is a separate origin-labelling rule, and when one country supplies at least half the wheat it permits wording that names that country and leaves the rest as a catch-all phrase. Legal is not the same as informative.

So the romantic version of your holiday, where the wheat came from a field you drove past, is often not what happened. That plate in Rome may well have been milled from the same North American durum that goes into pasta on an American shelf. If that is true, then whatever changed for you was not the field.

I would rather you knew this. It is the sort of thing a pasta company is not supposed to tell you, and it is checkable.

Maybe it is the species: durum versus bread wheat

Now we get to the strongest grain-level fact I found. Before I show it to you, the important part goes first: durum pasta contains gluten, it is not safe in celiac disease, and what follows is a fact about a plant, not a claim about a food.

Bread wheat and pasta wheat are not the same species. Pasta wheat, durum, is tetraploid: it carries two genomes, called A and B. Bread wheat is hexaploid: it carries A and B plus a third, the D genome, which entered the family thousands of years ago from a wild goatgrass.

The single most studied gluten fragment in celiac research is called the 33-mer. In 2015 a Spanish group sequencing gliadin genes found that the full 33-mer fragment traces to gliadin subtypes carried on that D genome (Ozuna et al., The Plant Journal, 2015;82(5):794-805). Two years later a German group measured it directly by mass spectrometry across forty wheat flours. They found the 33-mer in all twenty-three modern bread wheat cultivars, in all fifteen old bread wheat cultivars, and in both spelts. They did not find it at all in durum, in emmer, or in einkorn (Schalk et al., Scientific Reports, 2017;7:45092).

Look at what that result actually says, because it is not the story the ancient-grain aisle tells. The old bread wheats carried the fragment exactly like the modern ones did. The modern durum did not carry it. The line falls between species, not between old and new. That does not make the old Sicilian wheats less interesting, and I have written about the ancient grains of Sicily and where they actually come from, but it does mean age is the wrong axis for this particular question.

A Canadian team found the same shape of answer from a different direction, comparing how strongly two antibody assays bound gluten fragments across wheats. Modern durum grouped with khorasan and emmer in the lower-reactivity tier; ancient spelt grouped with modern bread wheat in the higher one. The split was by species, not by age. Those authors are blunt about the limits of their own work, writing that it is not clear whether gluten components are even involved in wheat sensitivity and that clinical studies are still needed (Gélinas and McKinnon, International Journal of Food Science and Technology, 2016;51(8):1801-1810).

Here is the honest complication, in the same breath. If age were the axis, einkorn should be the gentlest wheat there is. It is the most ancient one. Yet a Dutch group found two celiac epitopes present in every single full-length einkorn gliadin gene they examined, and wrote that it may be premature to start breeding wheat on this basis at all (van Herpen et al., BMC Genomics, 2006;7:1).

That is the same split the law draws, which is where this page started: Italian pasta must be durum, American pasta need not be, and the ingredient line is where you find out which one you are holding.

It is still only a fact about flour. It measures one peptide out of dozens. It says nothing about cooking, digestion, or symptoms, and its authors draw no safety conclusion from it. If you want the longer version of how the species diverged, I wrote it up in modern wheat versus ancient wheat, and what plant breeding actually changed.

These findings come from independent research and do not guarantee the same results with our product.

Maybe it is the variety: where the research contradicts itself

This is the explanation the ancient-grain business is built on: that old varieties are gentler than new ones. It is also the explanation where the evidence is in the worst shape, and I am going to show you both sides of it in the same section rather than putting the awkward part at the bottom.

Start with the best human evidence that exists. A team at the Gemelli hospital in Rome ran a double-blind randomized crossover trial: forty-two people with non-celiac gluten sensitivity enrolled, thirty-four finished, each eating 100 g of pasta a day for two weeks, then switching, never knowing which box was which. One pasta was Senatore Cappelli, an old durum variety released in 1915. The other was standard commercial pasta. Bloating scored lower on the Cappelli.

Then the authors did the honest thing that most write-ups of this study leave out. They corrected for having tested many symptoms at once. After that correction, one result survived: a feeling of incomplete evacuation (p = 0.019). Abdominal distension became a trend (p = 0.076). Bloating did not survive (Ianiro et al., Nutrients, 2019;11(4):712). The authors call it a proof of concept with no sample size calculation.

There is a second problem with reading it as a variety result. It was not a variety experiment. It compared an organic, single-variety pasta against a standard commercial pasta made from mixed grains of Italian and non-European origin. Variety, provenance, organic status, milling, drying, and die all changed at the same time, and neither pasta was analysed for gluten, protein, or fructan content. Something moved. Nobody can say what.

Now the contradiction, and it is a direct one. An Italian group screened eighty-two durum varieties by simulated digestion and mass spectrometry, measuring how many celiac-relevant peptides each one generated. Senatore Cappelli, the 1915 variety, landed in the high cluster at 1,488 ppm. Saragolla, a modern cultivar released in 2004, landed in the low cluster at 513 ppm (Gaiani et al., Nutrients, 2020;12(11):3566). The variety that won the human trial lost the laboratory test to a modern one.

The molecular surveys contradict each other too. One Dutch study of eighty-six bread wheats found a key celiac epitope higher in modern varieties and concluded breeding may have contributed (van den Broeck et al., Theoretical and Applied Genetics, 2010). A Portuguese-led study of 126 varieties found the opposite: landraces carried more potential immunostimulatory epitopes than modern varieties, and modern durum and durum landraces tied in the same lowest tier (Ribeiro et al., Food Chemistry, 2016;213:8-18). Both are antibody measurements of flour by competent groups. They cannot both be the general truth, and no third study has settled it.

What about khorasan? There is a human trial: twenty people with irritable bowel syndrome, not gluten sensitivity, ate a whole basket of khorasan foods for six weeks and reported less bloating. The khorasan was grown in Saskatchewan and the comparison wheat in Italy, so the variety and the field changed together, and the trial was part-financed by the company that owns the KAMUT trademark (Sofi et al., British Journal of Nutrition, 2014;111(11):1992-1999). There is a rat study, in which rats fed pasta from KAMUT-brand khorasan showed different duodenal histology from rats fed modern durum pasta (Carnevali et al., 2014). Rats are rats. There is ex vivo work in which blood cells from forty-eight patients released less of an inflammatory signal called CXCL10 when exposed to protein extracts from older wheats, notably independent of how much gluten they contained; only about 45 percent of patients responded at all, and cells in a dish never chewed, cooked, or digested anything (Valerii et al., Journal of Cereal Science, 2015). A follow-up in eleven children calls itself a pilot study in its own title (Alvisi et al., 2017).

Two independent reviews look at this whole body of work and decline to endorse it. One notes that the ancient-wheat literature is concentrated on a single branded grain, that most of those publications acknowledge support from the companies behind it, and that genotype and growing environment have never been separated (Shewry, Journal of Cereal Science, 2018;79:469-476). The other concludes that no single wheat type can be recommended as better or safer, and that ancient wheats do not contain less gluten than modern durum (Brouns et al., Nutrition Bulletin, 2022).

And the study closest to my own grain came back quiet. An independently funded Italian trial, double-blind and crossover, fed twenty adults pasta made from an evolutionary durum population derived from ten Sicilian varieties against modern organic durum pasta for eight weeks per arm (Baldi et al., Frontiers in Nutrition, 2022;9:971666). Sicilian material, no trademark, no grain company paying for it. I am telling you about it because it is the fairest test near my own doorstep, not because it helps me.

These findings come from independent research and do not guarantee the same results with our product.

Maybe it is the making: what happens after the field

The last explanation is about the making rather than the growing, and it splits cleanly in two. This is where most articles get sloppy, so I want to be precise: what is true of bread and pizza in Italy is not true of dried pasta, and the difference is that pasta is not fermented.

Long fermentation genuinely changes wheat. Lactic acid bacteria break gluten proteins apart, and researchers have pushed this a long way. In one much-cited experiment, gliadins from durum semolina were almost entirely broken down. But look at what it took: ten selected strains of lactobacilli plus added fungal proteases, running for forty-eight hours at 37 degrees, in a laboratory (Rizzello et al., Applied and Environmental Microbiology, 2007). That is detoxification biotechnology. No bakery in Naples works that way.

The one time researchers built this into pasta, they had to add fermentation as a deliberate manufacturing step that does not otherwise exist: ferment the semolina in liquid culture, freeze-dry it, then blend it three parts to seven with buckwheat flour. The resulting product was mostly buckwheat, and its residual gluten was still about 1,045 ppm, roughly fifty times the legal threshold for calling something gluten free (Di Cagno et al., Journal of Agricultural and Food Chemistry, 2005;53(11):4393-4402).

Dry pasta, mine included, is semolina and water, extruded and dried. There is no fermentation step at all. So if fermentation is the answer to your Italy question, it is an answer about the pizza and the bread, and it does not reach the spaghetti.

Then the Dutch put the whole idea to the test. Two randomized, double-blind, crossover, multicentre trials asked exactly this: in people with non-celiac wheat sensitivity, do wheat species and fermentation type matter? Sourdough baking did measurably reduce fructans and other wheat proteins in the bread. Symptoms did not follow. Neither species nor fermentation changed how people felt (de Graaf et al., The American Journal of Clinical Nutrition, 2024;119(4):896-907). A smaller Finnish pilot found the same shape: the sourdough loaf really was lower in fructans, and the people eating it did not feel better.

I should say something here about the claims my own industry makes, including ones I have made. Slow drying and bronze dies are real things that I do and that I believe in for texture. Bronze extrusion produces a rougher, more porous noodle, which is why sauce clings to it instead of sliding off. But when I went looking for evidence that either one matters to a sensitive gut, there is none. The field's own twenty-year review of pasta processing does not discuss digestion at all (Bresciani et al., Foods, 2022;11(3):256). And the one study that tested drying temperature against simulated digestion found no meaningful difference across the range real factories use (De Zorzi et al., Food Chemistry, 2007;104(1):353-363). Bronze die is a texture claim. I am not going to dress it up as a digestion claim.

These findings come from independent research and do not guarantee the same results with our product.

What each of these studies can and cannot tell you

A study is not a fact until you know what kind of study it was. Here is the whole cast, in plain language. Note the last column.

Not one of these studies tested my pasta

That column is not a formatting accident, and it is the part articles like this one always leave out.

Not one study above tested my pasta. Not one tested my Tumminìa. Not one tested my Perciasacchi. The khorasan in the Sofi and Carnevali papers is KAMUT-branded grain grown in North America under a trademark programme. Same species as Perciasacchi, Triticum turgidum subspecies turanicum, and a different landrace, a different country, a different farm, a different year. Treating those as the same grain would be a sleight of hand.

As far as I can find, nobody has published a health, digestion, or immune study on either Sicilian landrace. The one peer-reviewed paper that names both is about telling varieties apart for traceability, and it has no health endpoint of any kind (Palombieri et al., Plants, 2024;13(5):741). I cite it for exactly one thing: it independently confirms that Perciasacchi is a khorasan wheat, which is why I use the more recognisable name on the bag.

If that ever changes, I will publish it here. Including if it goes against me.

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

These findings come from independent research and do not guarantee the same results with our product.

So which one is it? I am not going to pick for you

I said at the top I would not hand you an answer, and I am keeping to that, because the research has not picked one either.

What I think is likely, and this is a structural point rather than a claim, is that these are not six competing answers to one question. They are six different things that can be true for six different people, in different proportions, in the same week. Two people can describe this to me in identical words and have nothing in common except the sentence.

And your experience can be completely real while the mechanism stays unexplained. Those are two different sentences. People with this condition get told constantly that because nobody can measure it, it is not happening. That does not follow, and it never did. One customer wrote up her own version of it, twenty years gluten free before she tried an ancient grain pasta. That is her account of her own experience, not a study, and not a prediction of yours.

There is one more thing worth knowing. A properly blinded trial of close to this exact question is running right now at the University of Palermo, comparing an ancient wheat against modern wheat in people with non-celiac wheat sensitivity, and it has not reported yet. When it does, I will read it and tell you what it says.

Until then the honest answer is that we do not know, and I would rather say that than sell you a story.

What I can hand you is the part that was never in doubt. Turn the box over. Semolina and durum are pasta wheat; wheat flour and farina are bread wheat. In Italy the law makes that choice for you. Here it does not, so the ingredient line is where you find out.

If you are actually travelling, the practical side sits on a different page. I put the labelling rules, the celiac association, and how eating out works into eating gluten in Italy as a traveller.

If you have a version of this story, I would genuinely like to hear it in the comments, because right now the people living it know more about it than the literature does.

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

What I do know about my pasta

Everything above is a question I cannot answer. So let me end on the part I can, which is not about you at all. It is about the food.

Three pastas, one ingredient each. The durum semola and the Tumminìa come from Sicily, from the valley my mother's town and my father's town sit in. The third is the khorasan, which Sicily has always called Perciasacchi. All three are organic and certified by CCPB, stone ground, slow dried, and bronze die cut.

Every harvest goes to Medallion Labs in Minneapolis, accredited by A2LA to ISO/IEC 17025. The panel runs to more than thirty analytes per pasta.

Rather than repeat those numbers here, I put the whole panel on one page: our pasta lab results, published in full, with the good source and excellent source lines worked out per serving, and the tests I have not run yet named as not run.

On the durum I went further than the recipe. Enriched macaroni is required by law to carry 198 to 265 micrograms of folic acid per 100 grams (21 CFR 139.115), which is the whole of what separates enriched from non-enriched pasta. If the two words read as interchangeable to you, they are not: I wrote out the difference between folate and folic acid separately. Mine is not enriched, so I had it measured: folic acid came back below the limit of detection, under 10 micrograms, in the 2024 harvest and again in 2025. The full reports are published, unedited, on my pasta lab results page.

None of that is proof of anything about you. It is a description of a food: the grain, the mill, the ingredient line, the numbers. Whether your body notices any of it is exactly the question this whole page says nobody has answered yet. What I can promise is that the bag holds what I say it holds, and that I will show you the paperwork.

If you want the paperwork rather than my summary of it, the three reports are published unedited on our pasta lab results page, including the analytes that came back unflattering and the one test I have not run yet.

Questions people ask me about this

Is pasta in Italy gluten free?

No. Italian dry pasta is made from durum wheat semolina, and durum wheat contains gluten. The question gets asked constantly because of the experience described at the top of this page, but the pasta itself is not gluten free and is not safe in celiac disease.

Is durum wheat lower in gluten than bread wheat?

No. Durum contains plenty of gluten, and it is often higher in total protein than bread wheat. What the research shows is a difference in which gluten fragments are present, not in how much gluten there is. Those are different questions and they get mixed up constantly.

Do ancient grains have less gluten?

No. A multi-author review looking specifically at this concluded that ancient wheats do not contain less gluten than modern bread wheat, and are not lower than modern durum either (Brouns et al., 2022). Einkorn, the oldest wheat there is, carried two celiac epitopes in every full-length gene one study examined.

Can someone with celiac disease eat any of this?

No. Not durum, not einkorn, not khorasan, not Tumminìa, not sourdough, not pasta from Italy. There is no exception anywhere on this page, and no study cited here claims one.

Is khorasan the same thing as KAMUT?

Same species, different thing commercially. KAMUT is a registered trademark for khorasan wheat grown under a specific programme, mostly in North America. Khorasan is the species, Triticum turgidum subspecies turanicum. Sicily has its own landrace of it, called Perciasacchi. Nearly all the published khorasan research used the trademarked North American grain.

Why can I eat sourdough bread but not regular bread?

Some people report this and the biochemistry behind it is real: long fermentation measurably lowers fructans and breaks down gluten proteins. But when two blinded Dutch trials tested whether that translates into how people actually feel, it did not (de Graaf et al., 2024). It is an open question. And it does not transfer to pasta, because dry pasta is not fermented at all.

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