Skip to main content

The diet and heart idea

After the Second World War, heart attacks were killing middle-aged men in rich countries at alarming rates. The American physiologist Ancel Keys argued that saturated fat in the diet raised blood cholesterol, which clogged the arteries. In 1953 he published a graph of six countries showing a close link between fat intake and heart disease deaths. In 1957 the statisticians Jacob Yerushalmy and Herman Hilleboe replied that data existed for 22 countries, and that with all 22 countries the link between fat and heart disease looked much weaker. Keys's defenders note that his 1953 paper was not meant as the final word.

To do better, Keys launched the Seven Countries Study in 1958, following about 12,000 men in the United States, Finland, the Netherlands, Italy, the former Yugoslavia, Greece and Japan. Men in eastern Finland, who ate a great deal of dairy fat, had very high rates of heart disease; men in Crete and Japan had very low rates. The study became a pillar of the diet and heart hypothesis, though critics point out that the countries were not chosen at random and that sugar intake tracked fat intake closely in the data.

A rival suspect

The British physiologist John Yudkin saw a different culprit. In his 1972 book, Yudkin argued that sugar was a major cause of heart disease, along with other ills. His evidence was also largely observational, and it did not persuade most researchers at the time. His reputation faded, then revived decades later when sugar came under suspicion again.

Advice becomes policy

In 1977 a United States Senate committee chaired by George McGovern issued the Dietary Goals for the United States. The goals urged Americans to eat less fat and more complex carbohydrates, as well as less sugar, salt and cholesterol. The first Dietary Guidelines for Americans followed in 1980, and similar advice spread across many countries. Critics, including some scientists at the time, argued that the evidence did not yet justify telling a whole nation what to eat.

The food industry responded with a flood of low-fat products. Many replaced fat with refined starch and sugar to keep the taste, and total calorie intake in the United States rose during the 1980s and 1990s. How much the advice itself contributed to rising obesity is still argued, since people did not follow it closely and many other things changed at once.

The trans fat reversal

For decades, stick margarine made from partially hydrogenated oil was promoted as a heart-friendly substitute for butter. In 1990 the Dutch researchers Ronald Mensink and Martijn Katan fed volunteers carefully controlled diets and found that trans fats raised LDL cholesterol and lowered HDL cholesterol, the worst of both directions. Large cohorts then linked trans fats to heart disease. Denmark capped industrial trans fat in foods in 2003, and later studies found heart disease deaths fell faster there than in similar countries. The United States removed partially hydrogenated oils from its list of safe ingredients in 2015, with a deadline in 2018.

Buried data and retracted trials

Between 1968 and 1973 the Minnesota Coronary Experiment fed thousands of residents of mental hospitals and a nursing home either a typical diet or one in which corn oil replaced much of the saturated fat. Much of the data went unpublished. When researchers recovered and analyzed it in 2016, the corn oil diet had lowered cholesterol without reducing deaths. Some read this as a blow to the diet and heart idea; others note the short follow-up, the trans fat in the margarine used and the frail population.

A Spanish trial called PREDIMED, published in 2013, found that a Mediterranean diet with extra olive oil or nuts cut heart attacks and strokes by about 30 percent. In 2018 it was retracted and republished, because some sites had not randomized participants properly, with whole households assigned together. A reanalysis reached similar conclusions, but with weaker claims to proof.

What has held steady

Despite the swings, some findings have been stable. Replacing saturated fat with unsaturated fat, especially polyunsaturated oils, is linked to less heart disease, while replacing it with refined starch and sugar shows little benefit. What a nutrient is replaced with often matters as much as the nutrient itself. Diets rich in vegetables, fruit, legumes and whole grains look good in almost every kind of study, and industrial trans fat is harmful.

Worked example: "Eat butter"

In 2014 a major American news magazine put a curl of butter on its cover with the words "Eat Butter," after a large pooled analysis found no clear link between saturated fat intake and heart disease. A reader might conclude that the old advice was simply wrong. A careful reader looks closer. The analysis compared people eating more saturated fat with people eating less, without asking what the low-fat eaters ate instead; if they ate white bread and sugar, a null result is what the stable finding above would predict. The cover was news; the evidence underneath was more modest and more interesting.

Quiz

5 questions · Practice: scored, not saved

1. What did Yerushalmy and Hilleboe point out in 1957?
2. Which study did Ancel Keys launch in 1958?
3. What did John Yudkin argue in his 1972 book?
4. What did the 1977 Dietary Goals for the United States urge?
5. What did Mensink and Katan report in 1990?
6. What did Denmark do in 2003?
7. What did recovered data from the Minnesota Coronary Experiment show?
8. Why was the PREDIMED trial retracted and republished in 2018?

CivilSimian.com created by AxiomaticPanic, CivilSimian, Kalokagathia