Do ketogenic diets cause cancer?

In the world of nutritional science, the precision of language and methodology matters. A lot. Yet somehow, we continue to see peer-reviewed studies published in reputable journals that use fundamentally flawed definitions, especially when it comes to hot-button topics like the ketogenic diet.

A recent example, published in the Nutrition and Cancer journal (July 2025), made headlines for its bold claim: that ketogenic diets are associated with an elevated risk for all cancers. Sounds alarming, right? The problem is: it doesn’t hold up under scrutiny. In fact, it’s not even studying ketogenic diets.

This study used data from NHANES (the U.S. National Health and Nutrition Examination Survey) between 2001 and 2018 to analyse self-reported dietary intake and health outcomes. The researchers applied a metric called the Dietary Ketogenic Ratio (DKR), based on the ratio of fat to carbohydrate and protein intake.

They then claimed that diets with higher DKR values (interpreted by the authors as “more ketogenic”) were associated with an increased risk of cancer.

Here’s the problem: It’s not a ketogenic diet Leaving aside the fact that this study was designed with a dietary recall (normally another sign to take the results with a pinch of salt), the main issue here is that the claim in the title is a blatant lie.

A true ketogenic diet is defined by very low carbohydrate intake, typically below 50 grams per day, and often closer to 20–30 grams. This is required to shift the body into a state of nutritional ketosis, where it begins producing ketones as an alternative fuel source.

However, the study’s “high DKR” group had a median carbohydrate intake of 181 grams per day. That’s not just non-ketogenic, it’s not even low-carb by most standards. This means the study does not include actual ketogenic diets at all.

Calling these high-carb, high-fat dietary patterns “ketogenic” is scientifically inaccurate. Why does it matter? Studies like this do more than just misinform, they actively distort public understanding of nutrition. Headlines circulate. Fear spreads. People living with cancer, or trying to prevent it, may be dissuaded from dietary approaches that could offer therapeutic benefit in certain contexts, all because of a study that never actually measured the thing it claimed to evaluate.

Even worse, clinicians, dietitians, and public health policymakers (many of whom rely on abstracts and summaries rather than digging into full methodology) may take the conclusions at face value.

This speaks to a broader issue in nutrition science: a worrying tolerance for imprecise terminology and weak methodological standards, especially in large observational studies. NHANES data is widely used, but it relies on short-term dietary recall and is inherently limited in establishing long-term health effects. When those limitations are combined with faulty definitions, the result is not just low-quality evidence, it’s misleading evidence.