In the town of Framingham, Massa-chusetts, the number of fatal heart at-tacks went down during the decline of animal fat consumption, but the number of non-fatal heart attacks increased during the same period (N Engl J Med, 1990; 322: 1635-41).

In England, the intake of animal fat has been relatively stable since at least 1910, whereas the number of heart at-tacks increased 10 times between 1930 and 1970 (Lancet, 1957; ii: 155-62).

The most we can say for the many studies that have been performed to test the diet-heart hypothesis is that there is a weak association between

the CHD mortality in various countries and the amount of fat available for them to eat, but no difference between the amount of fat eaten by coronary patients and by healthy individuals.

Such discrepancies clearly indicate that fat is not a causal factor. Usually, the common denominator in countries where people eat lots of high-fat food

is prosperity. In prosperous countries, high-fat foods are abundant, but so

are stress-provoking factors. Also, more people smoke, fewer people perform manual labour, industrial pollution of the environment is often worse and the ability to diagnose CHD is

better.

In Framingham, a small town near Boston, a large number of citizens have taken part in a study since the early 1950s surveying all the factors that may play a role in the development of atherosclerosis and heart disease. Among other things, their blood cholesterol levels have been measured frequently.

After five years, the researchers observ-ed what has become one of the cornerstones of the diet-heart connection. When they divided the participants into three groups of low, medium and high cholesterol values, they observed that, in the

lattermost group, more had died from a heart attack than in the two other groups. This must mean, they reasoned, that high cholesterol is a risk factor for CHD.

In the study, they showed that, on average, one per cent of all men with high cholesterol died each year during the 30 years of follow-up. During the first 10 years, about a quarter of one per cent of

the total died each year. Women with low cholesterol died equally as often as did women with high cholesterol. Among those with the lowest cholesterol values, only half as many died.

But these figures covered all causes of death. The researchers said nothing about death from heart disease. And heart mortality was the main issue of the project.

Furthermore, for men above age 47, cho-lesterol levels made no difference. Those who had low cholesterol at the age of 48 died just as often as those with high cho-lesterol. Blood cholesterol is usually at its highest at about age 50. It is after this age that heart attacks usually appear, increasing in frequency year by year. But few die from a heart attack before age 48, and most of those who do are diabetics or have a rare genetic component. More than 95 per cent of all heart attacks occur in people over 48. If cholesterol has no correlation with heart disease after 48, it cannot be associated with heart attacks in the majority.

Even more damning, the study showed that low cholesterol levels were associated with greater mortality and that people with low cholesterol levels were dying of other diseases.

The Framingham fingings are not

a rare exception. High cholesterol has no importance in Australian men over 74, according to an Australian study (Ath-erosclerosis, 1995; 117: 107-18). A New York study also found that neither total