After many years of telling Americans that if they eat less meat and eggs and lower their cholesterol levels they may help avoid heart attacks, the federal government has released a new study raising the possibility that men with very low cholesterol may suffer more cancers. the report, published in today's Journal of the American Medical Association, is based on the government's study of 5,200 residents of Framingham, Mass, who have been followed since 1948.

Men with very low cholesterol levels there had three times more colon cancer and 50 percent more cancer generally more colon cancer and 50 percent more cancer generally than men with high cholesterol. The starting finding seems to confirm similar ones that been appearing in some other studies.

Still other studies sharply disagree, however. The Japanese typically have both low cholesterol levels and a very low rate of colon cancer, the very kind most affected in Framingham.

It is possible, scientists said yesterday, that early, incipient cancers actually caused the low cholesterol levels in the Framingham men's bloodstreams rather than their low cholesterol levels being the cause and cancer the effect.

Still, the new data are certain to raise new questions about doctors' -- and recent federal government -- advice to eat less fatty food to lower cholesterol levels.

What the Framingham figures indicate is that men with cholesterol readings below 190 milligrams per deciliter (1/10th of liter) of blood had 4.6 colon cancers and 13.4 other cancers per 100 over a 24-year period.

Men with cholesterol levels between 180 and 280 (very high) had about the expected amount of cancer, 1 to 1 1/2 colon cancers and nine or so others. Those with levels above 280, very high, had still fewer.

The statistics were adjusted to rule out extraneous effects of such factors as age or simply living long enough to get cancer after avoiding attacks.

Framingham women showed no relationship between cholesterol levels and cancer. This is only one of many ham and other, similar results. Women, like men, are likelier to have heart attacks if they have high cholesterol readings.

The average American man's cholesterol level is 220, with most men ranging between 150 and 280. Many doctors have been saying that optimum levels to dodge heart disease should be 150 to 160, or at least below 200.

Heart doctors yesterday stuck to such advice. Officials of the National Heart, Lung and Blood Institute, which runs the Framingham study, said an expert advisory panel concluded last February that the evidence on low cholesterol and cancer was still uncertain. Most people, the panel said, should still consume a "prudent" low-fat, low-chesterol diet.

Dr. Robert Levy, Heart Institute director said,"We have only suggestion, not a true, proved association, about cancer and cholesterol. We have to investigate this very thoroughly. But against this possible association, which may have many other explanations, we have a known relationship between cholesterol levels and heart disease.

"These studies also indicate only what happened to persons with low cholesterol over many years. They do not tell us what happens when you lower cholesterol by diet or other means. There is absolutely no reason to change the present suggestion that those with elevated blood cholesterol prudently seek to lower it."

At the same time, said Dr. Manning Feinleib, associate director, "It well may be that on many measures -- cholesterol level, weight -- we have to try to find a golden mean. We know that it is not good to be either too overweight or too underweight. We may have to find a sensible cholesterol level that gives us the best chance for health."

Dr. William Kannel, co-director of the Framingham study, said, "It's still true that cardiovascular [heart and blood vessel] disease is the chief cause of mortality, and this is still the main problem to keep your eye on."

The Framingham study produced another surprise. Smokers of a pack or more of cigarettes daily had only a fourth as much colon cancer as non-smokers. But they had a 19 times greater risk of lung cancer.

Scientists now must learn far more. Among other things, they must study the role of specific blood lipids, or fats, not just blood cholesterol generally. They must re-examine the evidence that high fat consumption seems to increase the risk of both colon and breast cancer.

"In medicine we're always in a state of evolution," Kannel said.