IF THERE ARE FADS in the search for causes of cancer -- many say there are -- the hottest new theory on the research circuit is that cancer is related to diet, mainly the amount of fat you eat.

The National Cancer Institute recently issued a report, recommending among other things a redcuction in fat consumption. Americans now eat more than 40 percent of their calories in fat. Dr. Arthur Upton, director of NCI, said it would be "prudent" to cut that to around one-third. It was the first time such an institute had told people what to eat for their health, and Sen. George McGovern (D-S.D.), chariman of a nutrition subcommittee, was ecstatic. He noted his committee's "dietary guidelines" had suggested the same thing nearly three years ago.

But there is hardly unanimity in the scientific community over the NCI action. One author of the report says it was "rewritten 14 times" to accommodate critics both inside and outside NCI.

"Let's say it was very controversial," says Dr. Neil M. Ellison, who drafted the first versions and has left NCI.

Some believe that NCI went out on a limb with skimpy, conflicting evidence. Dr. Gregory O'Connor, director of NCI's division of cancer cause and prevention, says the recommendations "can't do any harm," but he isn't sure they will prevent cancer.

Other scientists contend NCI is being pressured by Sen. McGovern and researchers like Dr. Ernst Wynder, head of the American Health Foundation, to shift its attention from environmental chemicals as a primary cause of cancer to "life style factors," including diet. Dr. Samuel Epstein, professor of occupational and environmental health at the University of Illinois, sees the whole thing as an "industry-inspired" tactic to divert attention from chemicals as carcinogens.

"It's 'Let's blame the victim and let industry off the hook,'" he says. He ridicules the evidence. "There's not a scrap of evidence that fat causes cancer. There's more evidence cancer is caused by the GNP."

Nevertheless, evidence that fat is associated with cancer comes tumbling in from many research centers and the government is buying it.

It started when Dr. Wynder, well-known for his anti-smoking crusades and a charismatic figure in scientific circles, presented a paper in 1965 showing a decided correlation worldwide between high colon cancer rates and the high consumption of animal fats.

For example, the colon cancer rate among Japanese, who then ate one-fourth as much fat as we do, was 2.5 per 100,000 compared with 14 per 100,000 in this country. The pattern, typing colon cancer to the so-called "high-fat western diet," was also clear in such countries as the Netherlands, Belgium, Denmark, West Germany, France, England and Canada, according to Wynder.

This "landmark" paper has been quoted, reprinted, cited and confirmed by several other studies -- and the correlations have been expanded to include other types of cancer, mainly of the breast but also of the prostate, pancreas, ovaries, kidney and bladder.

Wynder further cites figures showing that when the Japenese, or other people on low-fat diets, move to the United States, their rate of colon and breast cancer increases. The most logical reason, he says, is change of diet. Also, he notes, Japan's cancer rate has started to rise since the introduction of Big Macs and other high-fat western food.

Furthermore, numerous studies, dating to the 1940s, consistently find that laboratory animals exposed to cancer-causing chemicals are much more likely to develop cancer if they're on high-fat diets.

For the last 15 years Wynder has bombarded his colleagues with these findings at scientific meetings and in journals. And many others have joined the crusade. However, as the theory made its way through the labyrinth of scientific scrutiny, holes became apparent. Thus, the theory is ever-changing -- bending and twisting -- to accomodate emerging evidence.

Some populations with high animal fat diets have low rates of cancer: for example, Finland. Seventh Day Adventists, many of them vegetarians, have a low cancer death rate. But so do the Mormons, who eat meat. Some of the correlations break down.

To help explain it, Wynder and other scientists have come up with other theories: One is that some cultures also have dietary customs that counteract the onslaught of fat.

For example, Wynder and his colleagues at the American Health Foundation, Dr. John Weisburger, have studied a rural Finnish area where farmers gulp down high-fat milk and cheese and load their bread and butter. All told, they eat the same amount of fat as New Yorkers, but have much less colon cancer. One difference: Those farmers also eat much more grain, mainly in heavy, coarse bread. The theory is that fiber, or roughage, in the bread acts as an antidote against the cancerous effect of the fat.

Lest it seem a fanciful notion, the theory is backed up by tests on animals showing that fiber, especially bran and pectin (found in the white membranes of citrus fruits) does counteract the effects of fat and reduce the risk of colon cancer. Current thinking is that fat and fiber have an intricate, interdependent relationship to colon cancer. That's why NCI's Dr. Upton suggested eating more fiber as well as cutting fat intake.

Gradually, however, a more important flaw cropped up. At first, cancer was blamed entirely on saturated meat and dairy fats, namely beef, butters, milk and cheese. But recent overwhelming evidence in animals shows that not only saturated meat and dairy fats but also vegetable fats are incriminated. In fact, unsaturated fats, the kind recommended for heart disease victims, send cancer charts zooming.

Recent, well-accepted studies by Kenneth Carroll, a biochemist at the University of Ontario, show that when you feed carcinogen-exposed rats high amounts of unsaturated fats -- such as olive, corn and safflower oil -- they develop twice as many tumors as rats exclusively fed saturated fats such as butter and meat fat.

Additionally, when you mix the two types of fat -- as occurs in the human diet -- the effects are even worse. This throws a lot of confusion into the issue and destroys some old stereotypes. The scientific reaction, in the absence of more precise knowledge, is to advise cutting back on both saturated and unsaturated fats -- the total fat intake.

At this point it sounds like a bizarre Mother Nature joke on humanity. A built-in cancer agent shot through our entire diet. But it's not that simple. Scientists do not see fat as an "initiator" of cancer, like pesticides and other environmental chemicals that trigger basic genetic changes in cells, causing them to go awry. Instead, fat is seen as a cancer "promoter." Once cells are exposed to chemical carcinogens -- as everyone's are in a modern environment -- the excess of fat somehow reportedly feeds the cancer, causing it to grow.

How fat might do that is still a mystery, and the lack of knowledge about the mechanism bother scientists like the NCI's Dr. O'Connor. He believes to prove the theory researchers must find out what is happening biologically in people. Such work is going on, and one of the most prominent theories for colon cancer is that a high-fat diet causes an increase of bile acids and steroids in the intestines. Bile acids have been declared cancer promoters in animals. And studies show colon cancer victims do have higher concentrations of bile acids.

As for breast cancer, Wynder and colleagues say eating fat causes an upsurge in production of a hormone called prolactin that concentrates in the breast. When women go on vegetarian diets -- and eat less fat -- their prolactin levels go down.

Prolactin, Wynder says, encourages the growth of breast cancer in animals. He is so convinced of the rightness of this theory that he urges women with breast cancer to cut down on eating fat to slow the cancer growth.

One of the problems with such advice is that there's no proof of how much fat you have to cut out to protect yourself. If we're eating 40 percent of our calories in fat today, what is the safe level? Is it 33 percent, as NCI suggested? Or 20 percent, or even 15 or 10?

The one-third recommended by NCI actually is taken from the so-called "prudent" diet to prevent heart disease. But nobody knows whether that is low enough to prevent cancer. Tests in England found that once animals were eating 30 percent of their calories in fat, it didn't matter whether they ate more than that -- up to 60 percent. There was no difference in cancer rate.

Dr. Weisburger thinks to prevent cancer, especially of the breast, we might have to cut our fat intake in half -- to 20 percent of calories, or even as low as the pre-World War II Japanese diet of 10 percent, which is comparable to the Pritikin diet. Since "there's a big gap in the facts," he says we need definitive studies to find out precisely how much fat is too much.

Clearly, the fat-cancer theory is stirring up scientific passions, as viewpoints vie for public attention and researchers compete for newly available research funds.

Patricia Hausman, a nutritionist with the Center for Science in the Public Interest, argues persuasively that the anti-fat evidence is strong enough to merit action. She points out that similar gaps and inconsistencies in other cancer areas have not deterred action. Not everyone who smokes cigarettes gets lung cancer, but that doesn't negate the danger of cigarettes.

"It's the weight of the evidence against fat that's important, not the exceptions," she says. Dr. William Lijinsky, director of chemical carcinogenesis at the Frederick Cancer Research Center, sees no conflict between the fat and chemical theories. He speculates carcinogenic chemicals may be created when fat is heated. To find out, he is testing certain chemicals.

Dr. David Kritchevsky, well-respected researcher at Philadelphia's Wistar Institute, is cautious. He says: "Some scientists have a tendency to oversell their views. I don't think the fat-cancer relationship is as cut and dried as some would have you believe. I haven't changed my fat intake one bit."

Some of the harshest critics -- at least of Dr. Wynder -- are two biochemists at the University of Maryland, Mary G. Enig and Mark Keeney. They contend their analysis shows that the correlations in Dr. Wynder's original "landmark" paper are in fact inaccurate -- and do not prove a relationship between animal fat and colon cancer.

They say just the opposite: Animal fat consumption has dropped while the cancer rate has gone up. They think the fat culprit in cancer, if one exists, is processed vegetable oils that somehow change the membranes of cells, triggering cancer. They maintain it's more dangerous to eat margarine than butter.

So the scientific skirmishes go on and will undoubtedly escalate after NCI's pronouncements. The issue is terribly complicated and ill-understood, and the impact of fat on cancer is far from clear-cut.

Still, if officials at NCI turn out to be right, they will be praised for their farsightedness. If they are wrong, probably few will remember or hold it against them.

Many think saying something is better than saying nothing. They share Dr. Weisburger's view: "Certainly telling people to eat less fat isn't hazardous, and it could do a hell of a lot of good."