A Keto Diet Accelerated Small-Intestinal Tumors in Mice — but Ketones Weren’t the Cause

The same diet produced opposite outcomes in two neighboring parts of the gut. In mice genetically predisposed to intestinal tumors, a ketogenic diet accelerated adenoma formation in the small intestine while suppressing tumor development in the colon.

The more surprising result was that ketone bodies — the molecules most closely associated with ketogenic diets — did not appear to drive either effect.

A ketogenic diet sharply restricts carbohydrates and shifts metabolism toward fatty-acid use. This produces ketone bodies, including β-hydroxybutyrate, which can serve both as fuel and as signaling molecules. Earlier research had suggested that β-hydroxybutyrate might help suppress colorectal tumors.

The researchers studied mice whose intestinal cells were genetically prone to losing the function of Apc. Disruption of this tumor-suppressor gene is common in intestinal tumors, while inherited *APC* mutations cause familial adenomatous polyposis in humans.

The animals received a control diet, a conventional high-fat diet, or an experimental ketogenic diet in which roughly 80% of calories came from fat. The keto-fed mice remained relatively lean, yet their small-intestinal tumor burden was comparable to, or greater than, that of mice eating the obesity-inducing diet. Their survival was also shorter. Most of the lesions in this model were adenomas — abnormal epithelial growths that can precede invasive cancer.

The team then manipulated ketone metabolism directly. They genetically reduced or increased ketone production and separately disrupted the ability of intestinal cells to use ketones as fuel. None of these interventions substantially changed tumor development.

Interfering with fatty-acid metabolism did.

The high influx of dietary fat activated PPARs, a family of lipid-responsive proteins that regulate gene activity. Intestinal stem cells became more proliferative and more capable of generating new cell colonies. That response can be useful when the intestinal lining needs repair, but in genetically susceptible tissue it also expands the pool of cells from which tumors can arise.

CPT1A proved to be a critical part of the pathway. This protein allows long-chain fatty acids to enter mitochondria, where they can be oxidized for energy. Removing CPT1A from the intestinal epithelium markedly limited the ketogenic diet’s ability to promote adenomas, while having little effect under the control diet. The tumor-promoting signal therefore came from the processing and oxidation of dietary lipids, not from ketosis itself.

The colon responded differently. The same ketogenic diet reduced both the number and area of colonic tumors, and altering ketone metabolism did not eliminate that protection. The researchers do not yet know why two adjacent intestinal tissues interpret the same diet in opposite ways.

These were genetically engineered mice, not people, and the experimental diet was extremely high in fat and largely lard-based. The findings therefore do not establish that ketogenic diets cause cancer in humans. They are most directly relevant to questions about inherited intestinal-cancer susceptibility, including *APC*-associated conditions, which will require human evidence.

A diet’s name describes a whole-body metabolic state. It does not tell us how every tissue will respond to the same nutrients.