How Ancel Keys Created The Diet-Heart Hypothesis

The idea that saturated fat is unhealthy is assumed to be common sense, but this theory has only been in existence since the 1950s, when biologist Ancel Keys created it.

Built on fear

Americans were terrified of heart disease, a scary new killer suddenly taking men in their prime. The theory at the time was that arteries narrowed as a result of aging, but Keys thought the disease could be avoided. He sought to offer a proactive solution, which was in stark contrast to the defeatist attitude of the time.

A powerful man

Ancel Keys with his cherry-picked data

Keys was fiercely independent, with an indomitable will, and would argue a point to the death. This meant he was often labelled “arrogant and ruthless”. He earned a PhD in biology at Berkeley, and a second in physiology at Kings College London. His passion for human physiology was sparked by studying the affects of altitude on his own body, and ventured into nutrition when he conducted research on starvation during WW2, which resulted in Keys developing the ‘K Ration’ food bar. He then turned to the study of heart disease. The main focus at the time was cholesterol, which is a vital component of every cell membrane in the body, and highly concentrated in the brain. It also comprises atherosclerotic plaque, which was thought to cause narrowing of the arteries, which would then lead to a heart attack. Although modern science has shown this outcome is more complex, the early observation was compelling, and created the relatively simple theory that cholesterol was bad for you.

Children dying from heart disease

Scientists found some children with a rare genetic disease that caused elevated serum cholesterol levels (the amount of cholesterol in the blood) and fatty lumps under the skin. The children also died young from heart attacks. Researchers observed that older people with similar lumps and elevated serum cholesterol also had heart attacks, and thus concluded that elevated serum cholesterol lead to heart attacks. They assumed that the lumps under the skin were the same “build-up” of cholesterol found in the arteries. Although other scientists suggested that this may not be a sound reasoning, but the theory prevailed.

Russian animal testing

A Russian researcher found he could induce arterial lesions in rabbits if they were fed massive amounts of saturated fat, and it was replicated in cows, sheep, and horses. Other scientists noted that this research used herbivorous animals, and therefore was probably not a good model for humans, and when the study was replicated in carnivorous dogs, the outcome was different. But the theory had already been embraced by heart disease researchers, and thus the recommendation to limit saturated fat in order to prevent build-up in the body began. Ancel Keys initially refuted this idea, because his experiments showed that massive amounts of dietary cholesterol did not raise serum cholesterol levels in his human test subjects. A Swedish researcher concluded similarly, finding that egg consumption actually reduced cholesterol. So what does raise serum cholesterol?

New tech redirection

Photo by ThisIsEngineering on

Heart disease research had been focused on carbohydrate and protein intake up until the invention of the gas liquid chromatography. This new technique allowed different types of fats to be analysed, and research grants flooded in. They discovered saturated fats, mono-unsaturated fats and polyunsaturated fats. Unsaturated fats are found in seed oils, and saturated fat is found in animal fats. Researchers found that replacing animal fats with seed oils dramatically lowered serum cholesterol, and that vegetarians also had lower levels. Studies showed that butter and lard increased serum cholesterol, while seed oils dramatically lowered it. This convinced researchers that saturated fat was causing heart disease through increasing serum cholesterol level.

Simple is powerful

Ancel Keys did not agree with this idea, favouring his idea that total fat was more important than the type of fat. He tested his diet theory on patients in a mental hospital, and found that a low-fat diet led to a small decrease in serum cholesterol. Despite the ethics of the study, small sample size, short study length and minimal effect, Keys presented his study as definitive. He concluded there was a direct relationship between dietary fat, serum cholesterol and heart disease. Keys then published a paper presenting his ‘diet heart hypothesis’ which charted a clear correlation between dietary fat consumption and heart disease in 6 countries. All of our modern ideas about the correlation between fat and disease stem from this idea originally put forth, and vigorously promoted, by Ancel Keys. The global impact that he achieved with this theory is unparalleled. Keys also believed that fat made people fat, due to its caloric density. Modern science has shown that this is not true, quite the opposite in fact, but the impact of Keys’ theory still remains today.

Fight fire with fire

Photo by Alexy Almond on

Keys recognised data about the Masai, Inuit, and Native Americans went against his theory, so he found his own international populations to support his theory. He found people in the Mediterranean and Japan had lower rates of heart disease and fat consumption, while some Finnish loggers that ate butter, cheese and beer that did have heart disease. While Keys tested only a few subjects in each location, and used no methodology for assessing their diet, he still wrote with certainty that dietary fat caused heart disease. He also used his journey to make alliances with international researchers who confirmed the theory using minuscule sample sizes. Keys used reduced heart disease deaths during war time as support for his theory, postulating that it was a result of meat rations. Other researchers reasoned that sugar and flour were also scarce during this time, along with an increase in exercise and reduction in car pollution, but Keys dismissed these ideas outright.

A quiet turnaround

By the 1950s, Keys started to change his writings about fat, turning away from total fat and towards the type of fats consumed. By testing on a few mental hospital patients again, he found that saturated fat raised serum cholesterol, and seed oils lowered it. His papers pronounced this as fact in several high-profile journals, even including a formula that supposedly calculated exactly what health outcomes the whole population would experience as a result of eating different amounts of fats. He confidently asserted that if saturated fat was cut from the diet, heart disease would become very rare.

Public victory

Photo by brotiN biswaS on

When President Eisenhower had a heart attack, his personal doctor Paul Dudley White (who was a Harvard professor, American Heart Association Founder and National Heart Institute creator) turned to Ancel Keys’ research as the answer. Keys had already cultivated powerful alliances in government, and taken White on international research trips. The day after Eisenhower’s heart attack, White held a press conference announcing to the public that if they want to prevent a heart attack, they need to cut out saturated fat and cholesterol from their diet. This was followed by prominent front page articles, which focused solely on Keys ‘diet-heart hypothesis’.

Your version of the data is wrong

Keys had many fans of his work, but Jacob Yerushalmy (founder of the biostatistics department at Berkeley) was not impressed with the data presented by Keys. He noted there are populations eating large amounts of animal fat that don’t get heart disease, as seen in the so-called ‘French Paradox’. It was clear by looking at the 22 countries for which data was available, there were “paradoxes” in several countries. Yerushalmy suggested that Keys had cherry-picked his data, and that alternative explanations were available such as cars, cigarettes, margarine and sugar consumption. He plotted the data for all 22 countries, and found no correlation between saturated fat and heart disease. George Mann was hopeful that this would stop the progress of Keys’ theory, but that was not to be. Instead, Keys was angry about Yerushalmy’s paper, and responded by writing a scathing rebuttal, claiming the data was unreliable. Even though it was a double-standard to make such a claim, no one questioned Keys’ statement.

Choose wisely to succeed

Photo by Ylanite Koppens on

Keys was not going to accept defeat, so he set about designing his “Seven Countries Study”, the first multi-country epidemiological study in history. He was awarded $200,000 to undertake the study, an enormous amount in the 1950s. Using existing data to make his selection, Keys chose to study rural middle-age male labourers in a variety of countries that had already been shown to confirm his hypothesis; USA, Finland, Netherlands, Italy, Greece, Yugoslavia and Japan. He avoided studying any countries that disprove his hypothesis, such as France, Switzerland, Germany, Norway or Sweden. The study took place during a time when the population of Greece, Italy and Yugoslavia were still recovering from WW2, affecting the childhood development of these men and current food availability.

Keys only selected 3.9% of the men in his study for food composition analysis, meaning that 96.1% of the men’s diets was unknown. The data was also collected erratically, with only one day recorded for the USA, and up to seven days for other countries. Some foods analysed before cooking, others after, and others a mixture of both. Different techniques of analysis were also used, with results significantly varying for the same foods. Keys even took measurements during the Greek period of Lent, which was a time of fasting from all animal foods, and these findings were used as the basis for the so-called “Mediterranean Diet”.

Shout success, hide facts

Photo by Lukas on

Keys published the results of his “Seven Countries Study” in a 211 page monograph, a book, and over 600 articles from supporting researchers. Interestingly, Keys chose to quietly publish all the original data in an obscure Dutch journal, rather than the prestigious American or British journals his other articles were published in. It is arguable that, if all the flaws of the study had been known at the time, it is unlikely that any would have been published.

Bad science is big news

It is estimated that by 2004, over 1 million references had been made to the “Seven Countries Study”, due to its seemingly clear correlation between the rate of heart attacks and saturated fat consumption. In fact, Keys’ study could at best only show an association between diet and heart attacks, not causation. Even though the data seemed elegantly clear, there were even differences within the countries he studied, with neighbouring areas demonstrating contradictory outcomes. Then when Keys followed up with his original populations 15 years later he discovered “paradoxical” outcomes. To justify this, he sought explanations beyond diet, refusing to accept the idea that his ‘diet-heart hypothesis’ could be flawed.

Independent review points to sugar

Photo by Suzy Hazelwood on

In 1999, the study’s lead Italian research reviewed the original data and found that the best association between a food and coronary mortality was actually sweets. He found a correlation coefficient of 0.821 (a perfect correlation is 1.0). The analysis even excluded chocolate, ice cream and soft drinks, due to it being “too troublesome” to re-categorise the data. He found that the animal foods had a correlation coefficient 0.798, but this also included margarine, a food known to have negative health effects. Keys knew all along that sugar was a possible alternative explanation, due to an ongoing debate with Dr John Yudkin, professor of physiology at Queen Elizabeth College London. However, Keys was not open to any alternative to his own theory, claiming that Yudkin’s theory was “a mountain of nonsense”.

The method is bunk

Epidemiological methods are not suitable for studying chronic diseases of dietary origin. Smoking was the most successful use of epidemiology, as it demonstrated a clear dose-dependent outcome. Smoking increases disease 30-fold, whereas Keys found only a 2-fold increase with no dose-dependent effect. A better methodology for the studying diet would be controlled trials, but these are expensive and time-consuming, and so conducted less often. Regardless, Keys used his epidemiological evidence to spearhead the first dietary guidelines.

Logical but ethically questionable

Photo by Ketut Subiyanto on

As a famous researcher in the field, with many powerful allies and years of study under his belt, it was understandable that Keys went to any lengths to publicise his findings, as it led to greater professional acclaim and further funding. This powerful influence wielded by Keys silenced not only his contemporaries, but also modern scientists, who still find themselves ridiculed, slandered, and silenced by their peers who see the issue of saturated fat as settled. Those who question the theory continue to lose jobs, funding, career advancement, conference invitations and the ability to publish of their work, so is it any wonder there is an appearance of consensus? The reality of nutrition science is not one of careful study, analysis and cautious interpretation, but rather one of strong personalities steering events with their charisma, wits and intelligence.


This article was based on the incredible book ‘The Big Fat Surprise’ by Nina Teicholz. Her deeply researched book is filled with data and stories detailing the evolution of the ‘diet-heart hypothesis’, which is the basis for modern food guidelines and medical treatments. Please consider purchasing a copy from her website, a local book store, or Audible.

2 thoughts on “How Ancel Keys Created The Diet-Heart Hypothesis

  1. The modern world has come to realize with more and more research to back it up that carbs, especially the highly processed ones are the real culprit to all of our health and weight related issues. All carbs are not built the same, and as time goes on, more and more data will keep on poking holes into Mr key’s “groundbreaking discovery”. Great post!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: