"We know two things about how to prevent death in middle age: smoking and cholesterol. Each of those two things is responsible for about one third of all deaths in middle age. Nothing else is up there with them."

With that crisp verdict, Oxford University scientist Richard Peto boiled down for listeners at last week's American Heart Association meeting his overview of research on the question "Will lowering the cholesterol in an individual's blood help protect against heart disease?"

Peto's answer, after analyzing studies of a total of 36,000 patients, is an emphatic "Yes."

A high blood cholesterol level is known to raise the risk of heart disease. Evidence comes from studies that compare cholesterol levels in countries where heart disease rates are high -- such as the United States and Great Britain -- to those in countries where heart disease is much rarer, such as Japan and rural China. Such studies show a consistent correlation -- the lower a population's average cholesterol level, the less heart disease occurs.

Within a single country, such as the United States, the same pattern holds: People with the highest cholesterol levels are most prone to heart attack and "hardening of the arteries," or atherosclerosis. Peto said animal studies in some 36 species confirm that raising blood cholesterol levels creates damage to arteries, and lowering cholesterol produces partial healing of the damaged areas.

Proof that lowering cholesterol with drugs or diet prevents heart disease has been harder to come by. One reason is that, with current treatments, it is difficult to reduce blood cholesterol levels in a large group of people by more than about 10 percent, and even harder to maintain the improvement and monitor the group's health for long enough to detect a difference in the figures.

To overcome the inconclusive results of individual studies on the subject, Peto and Dr. Salim Yusuf, a cardiologist at the National Heart, Lung and Blood Institute, pooled data from randomized controlled trials that monitored the effects of lowering cholesterol in 36,000 individuals, for periods ranging from one to seven years.

They found that reducing cholesterol levels by 10 percent decreased heart disease by about one sixth, even within the relatively short period covered by the studies.

"It confirms that there is a short time scale" -- that lowering cholesterol improves the health of the heart and arteries within a few years, Peto said.

They saw no evidence that lowering cholesterol increased chances of dying from other disorders. The data showed small but statistically insignificant increases in death rates from cancer, accidents and all other causes. "It represents the play of chance," Peto said.

To Peto, the findings are decisive enough to justify vigorous efforts to persuade people in the western countries to change the way they eat. "A substantial shift in our diet away from the saturated fat in dairy and other land animal products, toward the complex carbohydrates in fiber-rich cereals and fresh fruits and vegetables would substantially reduce heart disease and, hence, total death rates," he said.

Together, smoking and high cholesterol cause half of all deaths in people between ages 40 and 69, both in this country and in Great Britain, he said. "Smoking causes about one third of all cancer deaths in all age groups ," he said. "That is 10 times as many as the next most important, reliably known effect on cancer."

High cholesterol -- specifically high levels of low-density lipoprotein (LDL), the cholesterol-carrying substance most implicated in damaging the walls of arteries -- "is a cause of the large majority of heart disease in the United States," he said. Smoking is another cause, and in many people the two risk factors operate together.

To lower the risk, "the first thing anybody should do . . . is to quit smoking," Peto said. The second thing is to cut down on saturated fats in the diet.

This is because foods that contain saturated fat -- not foods rich in cholesterol -- are the main influence on cholesterol levels in the blood, Peto said. Each person's cholesterol level is influenced both by external factors (diet) and internal factors (metabolism).

"Internal factors are much more important, because you make more cholesterol than you eat, and if you eat more, you make less," Peto said. "A chief determinant of how much you make is the amount of saturated fat from dairy and land animal products" in your diet.

Thus, the healthiest diet for the heart is one that favors fish over meat, contains low-fat rather than cream-rich dairy products, and provides "a general sort of emphasis on vegetables rather than animals," Peto said.

He applauded the dietary recommendations in the 1985 version of "The American Heart Association Diet," a new pamphlet available from local heart association offices, which gives specific instructions on which foods to emphasize and which to avoid.

The beauty of the diet, according to Peto, is that available evidence suggests it should protect not only against heart disease, but against cancer as well. "The most currently promising theories about what nutritional changes might protect against cancer would actually be satisfied" by the heart association's recommendations, he said.

Peto, a lean, engaging statistician who directs Oxford University's clinical trials service unit, said his research is geared to finding ways to reduce premature deaths among middle-aged adults in Great Britain and other countries. In the 19th century, one-quarter of Britons died in infancy, another quarter in young adulthood, and another quarter between the ages of 40 and 69. Only 25 percent of the population lived to what we consider old age.

Today, he said, only 1 percent of Britons die in infancy and another 2 percent before age 40, but 28 percent die between the ages of 40 and 70. The death rate among the middle-aged is also disproportionately high in the United States.

In both countries, disease of the heart and blood vessels is the number one killer. In the United States it causes an estimated 985,000 deaths a year, according to the American Heart Association.

Peto said there is already a suggestion that changes in national habits are affecting the figures. In Britain, butter consumption has risen since the 1950s, and so have heart disease statistics. In the United States, butter has lost ground to margarine, and heart disease death rates have dropped. The figures provide further evidence that dietary changes can produce results relatively quickly.

"If you drop the national cholesterol level by 10 percent, it will save not only my children," he said, "but also my own generation."