For years I carried around a tidy little theory about keto and cancer, and I never once checked its wiring. You cut the carbs, starve the tumor of sugar, let the body run on ketones, and somewhere in there the ketones themselves do something protective. So when a lab at MIT reported that a ketogenic diet gave cancer-prone mice more tumors, my first instinct was to assume somebody had fed them the wrong chow. Then I read how they took the ketone half of my theory apart, one experiment at a time, and I’ve had to rebuild what I thought I knew.
The deeper question is why the same diet has opposite consequences in two adjacent parts of the gut.
The theory had excellent company. In 2022 a Penn team reported in Nature that ketogenic diets made mice strikingly resistant to colorectal tumors, and they pinned the effect on beta-hydroxybutyrate, or BHB, the main ketone your liver pours out when carbohydrates run dry. Giving BHB alone, without the diet, cut tumor burden in their mice. Penn’s own announcement noted that BHB is “widely available as a dieting supplement,” quoted a co-author predicting the molecule “could someday become a standard part of colorectal cancer care and prevention,” and said a clinical trial of BHB was being set up. If any result could make a ketone in a bottle sound like medicine, it was that one.
Ömer Yilmaz’s lab at MIT looked one stretch of gut higher and found something I did not see coming. Writing in Nature, Jessica Shay, Fangtao Chi, Yilmaz and colleagues put mice genetically predisposed to intestinal cancer on a ketogenic diet that drew about 80 percent of its energy from fat and kept them on it for roughly 150 days, according to Earth.com. In the small intestine, the keto mice developed more tumors than mice on a control diet, at rates similar to, and in some cases higher than those in mice fed an obesity-inducing, high-calorie diet. They died sooner, too. And they never got fat, so the extra tumors showed up without the weight gain we usually blame for that kind of damage.
| Measure | Share |
|---|---|
| of the mouse keto diet's calories | 80 percent |
Slide a few inches down into the colon and everything flips. In the same animals, on the same food, the diet suppressed colon tumors, just as the Penn team had found. One diet, two neighboring stretches of intestine, opposite outcomes.
The obvious suspect was ketones, and Yilmaz says the team expected exactly that: “We expected ketone bodies like BHB to be the direct drivers.” So they went after them. They shut down ketone production in stages, first in the gut lining and then in the liver, until the supply was a trickle, and tumors kept forming. They raised ketone levels in mice eating an ordinary diet, and tumor development didn’t budge. When they disabled the machinery cells use to burn ketones for fuel, tumor counts stayed flat (Earth.com). “Surprisingly, neither increasing nor eliminating ketone production altered intestinal tumor growth,” Chi said, as ScienceAlert reported. Yilmaz called the molecules “essentially metabolic bystanders.” Even the colon protection didn’t depend on them, and the receptor that the competing ketone theory used to explain that benefit was missing from colon tumors and stem cells altogether.
I’ll admit that one stopped me cold, because it puts the new paper squarely against the old one’s central claim. The mouse models and gut regions weren’t identical, so this isn’t a clean head-to-head. Still, one Nature paper says BHB on its own cut tumor burden, and the other says pushing ketones up or down did nothing to tumor growth. I know which of the two results the supplement aisle was rooting for.
Wait, so if the ketones aren’t doing it, what is? The answer is the fat itself, and what the gut’s stem cells do with it, and the mechanism is almost physical enough to picture. On keto, fat floods the intestine, and the stem cells lining it ramp up fatty acid oxidation, burning that fat for energy. The burning switches on a family of proteins called PPARs, which push the stem cells to divide faster (MIT News). Every extra division is another chance for a cell to pick up the wrong mutation, and in a mouse already carrying a cancer-primed gene, that’s a foot pressed on the accelerator. The team then tested the idea bluntly: when they deleted CPT1a, the enzyme that lets fat into the cell’s power plants, from the gut lining, the diet’s tumor-promoting effect dropped sharply (Earth.com). In Yilmaz’s words, “tumor acceleration is driven entirely by how stem cells process and burn” dietary fat.
Nobody can explain yet why the colon reads the same fat-burning signal and goes the other way. “We don’t know why they’re responding differently,” Yilmaz said, and Chi framed it as the next job: “The deeper question is why the same diet has opposite consequences in two adjacent parts of the gut.” I like a lab that leaves the hole where everyone can see it. A supplement label never does.
Which brings me to the ketone drinks, sold on the idea that you can skip the brutal diet and sip the benefit straight from a can. The MIT researchers say their work points the other way: because both the small-intestine risk and the colon protection trace back to how cells metabolize dietary fat rather than to ketones circulating in the blood, ketone supplements or drinks would not be expected to reproduce either effect (MIT News). You can’t buy the colon protection in a sachet. You aren’t buying the small-intestine risk either, which is cold comfort when what you’ve paid for is the bystander.
The people I’d worry about aren’t the weekend keto dabblers. Small-intestinal tumors have been rising in incidence in recent decades, and the risk falls hardest on people with inherited conditions like familial adenomatous polyposis, or FAP (MIT News). FAP carpets the gut in polyps, and many patients have their colon removed young. After that, the small intestine becomes the battleground: benign growths turn up in about 9 in 10 patients and cancer in roughly 1 in 10, and some of these patients get steered toward low-carb eating (Earth.com). Put that next to this study. The colon, the one place keto looked protective, is already gone, and what’s left is the stretch of gut where keto added tumors.
| Measure | Part | Whole |
|---|---|---|
| FAP patients who develop benign small-intestine growths | 9 | 10 |
These are mice bred to get intestinal cancer, and Yilmaz himself told Earth.com that for a healthy person, “this isn’t a reason to change course.” I’ll take him at his word on that, since nobody has run this in humans. But for someone with FAP, a blanket “keto fights cancer” is exactly the kind of advice that could quietly work against them, in the one organ the headlines skip.
Mostly, this paper broke a shortcut I’d been leaning on without noticing. Keto-against-cancer was appealing because it promised one lever that moved everything in the same direction, and the biology won’t cooperate: the same fat, burned by the same kind of cell, adds tumors in one stretch of gut and subtracts them in the next. If I had FAP, I would not start a ketogenic diet without a gastroenterologist keeping watch on my small intestine, and whatever I end up eating, I’m not spending a dollar on ketone drinks.
Sources
- Nature: Shay, Chi, Yilmaz et al., “Ketogenic diet mediates intestinal tumorigenesis through lipids not ketones” (2026)
- MIT News: Ketogenic diets may increase cancer risk in the small intestine (2026)
- Earth.com: Ketogenic diet may protect one part of the gut while harming another (2026)
- ScienceDaily: MIT scientists find a hidden cancer risk of the keto diet (2026)
- ScienceAlert: coverage of the MIT keto study (2026)
- Nature (via PMC): Dmitrieva-Posocco et al., “β-Hydroxybutyrate suppresses colorectal cancer” (2022)
- Penn Today: ‘Keto’ molecule may be useful in preventing and treating colorectal cancer (2022)