The Ancel Keys affair: how the fear of fat was manufactured

In 1953, the American physiologist Ancel Keys published a striking chart: plotting fat consumption against cardiovascular mortality across six countries, he got an almost perfect curve. The conclusion seemed to make itself: eating fat kills the heart.

The problem, pointed out as early as 1957 by the statisticians Jacob Yerushalmy and Herman Hilleboe, is that Keys had data on 22 countries, not 6. With all 22, the neat curve disappears. The correlation vanishes.

Keys was not a marginal figure. He was one of the most influential nutrition scientists of the 20th century: he ran the starvation studies during the Second World War, designed the K rations for American soldiers, and launched the Seven Countries Study in 1958, the first large multi-continent epidemiological study. His academic authority was immense, and his access to decision-making committees direct.

It is precisely because he was credible that his error, or his deliberate selection, had such consequences.

Facing Keys, another researcher argued the opposite: the Briton John Yudkin, professor at the University of London, showed from the 1960s that it was sugar, not fat, that correlated with cardiovascular disease. In 1972 he published Pure, White and Deadly, a book for a general readership, explaining how excess dietary sugar lay behind the epidemic of metabolic disease.

Keys then ran an active campaign to discredit Yudkin, publicly calling his work "quackery" and "junk science". Yudkin was gradually pushed out of influential academic circles. His work was forgotten for thirty years.

In 2016, an analysis published in JAMA Internal Medicine by Cristin Kearns, Laura Schmidt and Stanton Glantz (UC San Francisco) revealed what the internal archives of the Sugar Research Foundation (SRF, today the Sugar Association) contained.

The Keys-SRF thesis won out. In 1977, the first official American dietary guidelines put fat reduction at the heart of public health policy. Europe followed. Manufacturers reformulated their products as "low fat", making up for the lost taste with added sugar.

The outcome is documented: between 1970 and 2000, American fat consumption did indeed fall, in line with the guidelines. Over the same period, obesity doubled and type 2 diabetes exploded. The cardiovascular mortality curve did not improve in proportion to what the hypothesis predicted.

The reassessment has been under way since the 2010s, but it has accelerated and hardened recently. These studies are no longer dissenting opinions: they are meta-analyses published in the world's leading medical journals.

That last point is worth underlining: Yamada et al. 2025 covers randomized clinical trials specifically, the highest level of evidence in medicine. These are not observational correlations like Keys's in 1953. They are controlled experiments, across 13,000 participants. And the conclusion is the same: restricting saturated fat does not improve cardiovascular outcomes.

The "fat causes heart disease" hypothesis rests on an original study that was methodologically fragile (countries were selected), reinforced by an industry with direct financial interest in its winning, which actively funded research to steer that outcome.

Seventy years later, the most rigorous clinical trials available refute that hypothesis. It is not alternative voices or low-carb bloggers saying so: it is The Lancet, the European Journal of Preventive Cardiology and the JMA Journal.

The dogma lasted because it was institutional, not because it was right. And institutions are slow to change.