Nutrition, Diet & Fasting - Diets
Low-carb vs low-fat: what the head-to-head trials found
By Ryan Lowe. Last reviewed 3 August 2026. How we write this.
The argument that consumed nutrition for thirty years, and the year-long randomised trial that put around 600 people on each side and found essentially no difference.
For most of living memory, dietary advice has been fought over one axis: is fat the problem or is carbohydrate the problem. Whole publishing careers, food industries and public health guidelines have been built on each answer. It is worth knowing what happened when someone tested it properly, because the result reframes almost every other argument in this classroom.
The trial
DIETFITS, published in JAMA in 2018, randomised around 600 adults with overweight or obesity to a healthy low-fat or a healthy low-carbohydrate diet for twelve months. Both groups got extensive dietitian support, and crucially both were told to prioritise food quality - vegetables, whole foods, minimal added sugar and refined grain - rather than simply to hit a macronutrient target. Neither group was given a calorie limit.
Average weight loss was around 5-6 kg in both groups, and the difference between them was not statistically significant. Within each group the spread was enormous: some people lost more than 20 kg and some gained, on both diets.
The study also tested the most popular explanation for that spread. The researchers had prespecified that they would check whether a genetic pattern associated with fat or carbohydrate metabolism, or a person's baseline insulin secretion, predicted which diet suited them. Neither did. The idea that a test can tell you which macronutrient split is right for you did not survive its own trial.
What that means, and what it does not
It does not mean the two diets are identical in every respect - they have different effects on blood lipids, on blood glucose, and on how easy different people find them. It means that for the outcome most people care about, over the timescale that matters, the macronutrient argument was much smaller than either camp believed.
The enormous individual variation is the more useful finding, and it is consistently misread. It is not evidence that people have secret metabolic types, because the trial looked for exactly that and did not find it. The better explanation is more boring: adherence. The people who did well were the people who kept doing it, on either arm.
This is why so much of this classroom is about mechanisms and honesty rather than picking a winner. If two opposite diets produce the same average result when both are done well, the interesting questions move elsewhere - to food quality, to ultra-processed food, to protein, and to whether a person can sustain the thing at all.
Where the evidence stops
Both DIETFITS arms were "healthy" versions with heavy dietitian support, which is not how either diet is usually done in the wild. A trial comparing a well-run low-carb diet with a well-run low-fat diet does not tell you much about a badly run one. Twelve months is also short relative to a lifetime, and weight regain after diet trials is common on every arm. And this compared diets on weight - a trial powered for heart attacks might have found differences that this one could not detect.
Educational content only and not medical advice. Individual circumstances, medication and existing conditions change what is appropriate; a qualified healthcare professional is the right person to advise on yours.
Related in this classroom
The studies behind this page
- Effect of low-fat vs low-carbohydrate diet on 12-month weight loss and the association with genotype pattern or insulin secretion (DIETFITS) - JAMA, 2018
Sources for this classroom
Educational content only. Not medical advice, diagnosis or treatment. Always consult a qualified healthcare professional before changing your health regimen.
