The ketogenic diet has spent a century changing jobs. It began in the 1920s as a medical treatment for epilepsy, then became one of the most popular weight-loss plans in the world. Now a study from the Massachusetts Institute of Technology has given it a more complicated reputation: in the same animals, on the same diet, the keto diet appeared to raise the risk of tumours in one part of the gut while lowering it in another.
That split result is what makes the research worth a careful read. It is not a verdict on the keto diet as a whole, and it is not a study in people at all — it is a study in mice, and its most important message may be about how carefully diet research has to be read.
The short version
Researchers at MIT fed mice that were genetically prone to intestinal cancer one of three diets: a regular diet, a high-fat diet designed to promote obesity, or a ketogenic diet. The mice on the ketogenic diet developed more tumours in the small intestine than the mice on the regular diet, at rates similar to or higher than the obesity-promoting diet group — even though the keto-fed mice did not become obese. In the colon, the same diet suppressed tumour formation. The researchers traced both effects to the way intestinal stem cells burn dietary fat, not to the ketone bodies the diet is named for. Published in Nature in July 2026, the study has drawn fresh attention in November 2026 as new coverage brings it to a wider audience.
Two intestines, two opposite results
The gut is one continuous tube, but the study shows that its two main sections responded to the keto diet in opposite directions. The contrast is the heart of the finding.
| Small intestine | Colon | |
|---|---|---|
| Tumour outcome on a keto diet | More tumours than on a regular diet, at rates matching or topping the high-fat diet group | Fewer tumours than on a regular diet |
| Weight of the animals | Stayed lean; the keto mice did not become obese | Same animals, same diet |
| Likely driver, according to the researchers | Dietary fat burned by intestinal stem cells | Also linked to fat metabolism, not ketones |
| Earlier evidence | No earlier sign of protection here | A 2022 study had already suggested the keto diet protected the colon |
The table makes the puzzle plain. If the keto diet were simply good or bad for the gut, both sections would move in the same direction. They did not — and the researchers say the effects of a diet can depend on the tissue, not only on the diet itself.
The study, in its own timeline
The new findings did not arrive out of nowhere. They sit at the end of a chain of research that stretches back a hundred years, and the chain helps explain why the result surprised even the people who found it.
In the 1920s, the ketogenic diet was developed as a treatment for epilepsy, a medical diet built around very low carbohydrate intake and very high fat intake. It stayed a specialist treatment for decades before moving into the mainstream as a weight-loss plan, where its appeal was straightforward: by cutting carbohydrates sharply, the body turns to fat for fuel, and many people find it helps them eat less without feeling as hungry.
In 2022, a study published in Nature reported that a ketogenic diet could protect against colon cancer. That work proposed that a ketone body called BHB, the most abundant ketone produced during ketosis, played a central role in the benefit, and it set the question the MIT team would later take up.
In July 2026, the MIT team published its study in Nature, titled “Ketogenic diet mediates intestinal tumorigenesis through lipids not ketones”. The senior author was Omer Yilmaz, an associate professor of biology at MIT and director of the MIT Stem Cell Initiative, with postdoctoral researchers Jessica Shay and Fangtao Chi as lead authors. The team set out to test whether the protective effect reported for the colon in 2022 would also appear in the small intestine. It did not — and the explanation the team found challenged the earlier idea about ketones as well.
In November 2026, the study reached a wider audience through new health coverage that presents the research as a fresh warning about the keto diet. That coverage is why the study is being discussed again now, months after publication — the research itself is from July, while the current wave of attention is new.
How the researchers tested it
The MIT team worked with mice genetically engineered to be prone to intestinal cancer, animals that develop tumours readily and therefore make dietary effects easier to detect. The mice were divided into three feeding groups, and each group ate its assigned diet while the researchers watched what happened in the gut.
The regular diet group
The first group ate a standard control diet, the ordinary laboratory diet against which everything else was measured. These mice provided the baseline: how many tumours develop in cancer-prone mice when nothing about the diet changes.
The high-fat diet group
The second group ate a high-fat, high-calorie diet designed to promote obesity. This group separated two possible explanations: if the keto diet raised tumours only by making animals overweight, the high-fat group should have shown the same pattern for the same reason. The comparison let the researchers ask whether weight gain or the diet itself was doing the work.
The ketogenic diet group
The third group ate a ketogenic diet, very low in carbohydrate and very high in fat, the pattern that pushes the body to produce ketone bodies. These mice stayed lean, unlike the high-fat group, yet they developed small-intestine tumours at rates similar to or higher than the animals on the obesity-promoting diet. That pairing — lean animals with a heavy tumour burden — rules out body weight as the explanation.
When the researchers examined the colon in the same animals, the keto diet suppressed tumours, matching the direction of the 2022 finding. The same diet, the same animals, two opposite outcomes a short distance apart.
Fat, not ketones: the mechanism
The most surprising part of the study is not the split result but the mechanism behind it. The keto diet is named for ketones, the molecules the body produces when it burns fat instead of carbohydrate, and the 2022 work had proposed that a ketone body called BHB was central to the colon benefit. The MIT team tested that idea directly, and it did not hold.
The researchers altered the pathways that produce and break down ketones, raising ketone production, lowering it, and blocking it altogether. Whether ketone levels went up, down, or were switched off, tumour development stayed largely unchanged. The ketone bodies, in the researchers’ reading, were essentially bystanders.
What did change the outcome was the sheer amount of dietary fat and the way intestinal stem cells handled it. The large fat intake activated a metabolic pathway called fatty acid oxidation, the process cells use to burn dietary fat for energy. Increased activity in this pathway switched on a family of proteins called PPARs, which then encouraged intestinal stem cells to divide more rapidly. Faster division means more opportunities for a cell to turn cancerous, and in animals already prone to tumours, that shift was enough to raise the tumour count.
The same fat-driven pathway appears to sit behind the colon result as well, which challenges the earlier explanation there too. The protective effect in the colon remained, but the new experiments suggest it was not the ketones doing the protecting. Readers who want the full account of the study and its coverage can read the Blavity report on the keto diet and cancer risk, which sets out the findings and the experts’ cautions. That distinction carries a practical consequence the researchers themselves noted. Commercial ketone drinks and supplements, which raise ketone levels without the high fat intake of a ketogenic diet, would not be expected to reproduce either the risk in the small intestine or the benefit in the colon, because the effects appear to come from fat metabolism rather than from ketones.
What the researchers said
Yilmaz, the study’s senior author, put the lesson in plain terms when the research was released: ketogenic diets have distinct effects on different tissues even within the gastrointestinal tract, and the message is to be very careful in generalizing the effects these diets can have, because what might be beneficial for one tissue may be detrimental for another tissue.
The caution is aimed at a habit diet research invites: taking a result from one setting and stretching it into a general rule. A benefit in the colon does not prove a benefit everywhere, and a risk in the small intestine of cancer-prone mice does not prove a risk in people. Both stretches go beyond the evidence, and the researchers have been clear about the boundary.
Independent commentators have made the same point. The work was done in mice, not humans, and the researchers and commentators alike stress that it is not a basis for dietary recommendations. The study opens a question about how dietary fat affects intestinal stem cells; it does not answer what any individual person should eat.
What this does not show
Every study has limits, and this one has limits that matter more than most, because the subject is cancer and the diet is one millions of people follow.
First, it is a mouse study. The animals were genetically prone to intestinal cancer in a way most people are not, and mice are not small humans. A result in cancer-prone mice can point researchers toward a mechanism worth investigating, but it cannot tell anyone what a ketogenic diet does to cancer risk in healthy people.
Second, there is no human data in this study at all. No people were followed, no human tumours were counted, and no human trial has tested whether a ketogenic diet changes intestinal cancer risk. Until such evidence exists, the risk in people is unknown.
Third, the caution the experts raise is aimed at a specific group: people with inherited conditions that raise the risk of intestinal cancer, such as familial adenomatous polyposis, known as FAP. For those people, the researchers suggest extra care and a conversation with a clinician before starting a ketogenic diet. That advice is narrow, and it should not be read as a warning to everyone who eats a low-carbohydrate diet.
Fourth, the study does not overturn the reasons people choose a keto diet, and it does not speak to them. Weight loss, blood sugar control, and appetite are separate questions from tumour biology in mice, and this study measured none of them.
Fifth, the result is about the diet as a whole pattern, not about any single food. The keto diet is very high in fat by design, and the mechanism the researchers found runs through that total fat load. The study says nothing about individual foods, moderate low-carbohydrate eating, or ketone supplements taken on their own.
Frequently asked questions
Does the keto diet cause cancer?
No human study shows that. The MIT study found that a ketogenic diet increased small-intestine tumours in mice that were genetically prone to intestinal cancer, and decreased colon tumours in the same animals. It was an animal study, and the authors and commentators stress that it does not establish cancer risk for people.
What is the keto diet cancer risk the study actually found?
The risk appeared in the small intestine of cancer-prone mice, where keto-fed animals developed tumours at rates similar to or higher than animals on an obesity-promoting high-fat diet, even though the keto mice stayed lean. In the colon, the same diet suppressed tumours. Both effects were traced to how intestinal stem cells burn dietary fat, not to ketones.
If ketones are not the cause, what is?
According to the researchers, the driver is fatty acid oxidation, the process intestinal stem cells use to burn dietary fat. A heavy fat intake increased activity in this pathway, which switched on proteins called PPARs and pushed stem cells to divide faster, creating more chances for a cell to become cancerous. Changing ketone levels up or down did not change the outcome.
Should people with a family history of intestinal cancer avoid keto?
The researchers advise particular caution for people with inherited conditions such as familial adenomatous polyposis, which greatly increases the lifetime risk of intestinal tumours. Anyone in that position should discuss a ketogenic diet with a clinician before starting one.
Do ketone supplements carry the same risk?
The researchers say they would not be expected to. Because the effects in the study were driven by dietary fat metabolism rather than by ketones themselves, commercial ketone drinks or supplements are unlikely to reproduce either the small-intestine risk or the colon benefit seen in the mice.
What this means for weight loss and daily calories
The study leaves the keto diet’s weight-loss story where it was. Many people lose weight on a keto diet, and the usual explanation is straightforward: cutting carbohydrates sharply tends to reduce appetite and total calorie intake, and weight change follows energy balance over time. None of that is challenged by a mouse study about intestinal stem cells, and none of it is confirmed by it either. The two questions — does the diet help with weight, and does it affect tumour biology — are answered by different evidence and should be kept apart.
That separation is also what makes the study useful for anyone thinking about food in terms of energy. A diet is not only a list of foods; it is a daily energy budget, and the same diet can look very different depending on how much energy a person’s day actually requires. Readers who want that background can start with a plain explanation of what TDEE is, which sets out total daily energy expenditure in everyday terms, and a guide to TDEE versus BMR explains how the daily total differs from the smaller basal figure.
The MIT study is a reminder that diet research rarely delivers a simple verdict, and that a single diet can do opposite things in neighbouring parts of the body. For anyone considering a ketogenic diet, the sensible next step is the one the researchers suggest: treat the finding as a question to raise with a clinician, not as a decision made by a headline. And before changing how they eat, readers can work out their own daily energy needs with the TDEE calculator, and take that number into a conversation with a clinician or dietitian.
