The diet messages are everywhere: the National Cancer Institute has an "Eat 5 to 9 a Day for Better Health" program, the numbers referring to servings of fruits and vegetables, and the Prostate Cancer Foundation has a detailed anticancer diet.
Yet despite the often adamant advice, scientists say they really do not know whether dietary changes will make a difference. And there lies a quandary for today's medicine. It is turning out to be much more difficult than anyone expected to discover if diet affects cancer risk. Hypotheses abound, but convincing evidence remains elusive.
Most of the proposed dietary changes are unlikely to be harmful - less meat, more fish, more fruits and vegetables and less fat. And these changes in diet may help protect against heart disease, even if they have no effect on cancer.
So should people who are worried about cancer be told to follow these guidelines anyway, because they may work and will probably not hurt? Or should the people be told that the evidence just is not there, so they should not deceive themselves?
Dr. Barnett Kramer, deputy director in the office of disease prevention at the National Institutes of Health, said: "Over time, the messages on diet and cancer have been ratcheted up until they are almost co-equal with the smoking messages. I think a lot of the public is completely unaware that the strength of the message is not matched by the strength of the evidence."
But Dr. Arthur Schatzkin, chief of the nutritional epidemiology branch in the National Cancer Institute division of cancer epidemiology and genetics, said people wanted answers, even if they are not are not definitive.
"It is not enough to say that, well, this is complicated science and maybe in seven or eight years we will have new methods in place" that might resolve the issues, Dr. Schatzkin said. "We have a responsibility to give the best advice we can while pointing out where the evidence is uncertain and how we're working to improve the science."
That, however, is little consolation to cancer patients and family members who are terrified that cancer might strike them next. And there are more and more. As the population ages, the number of cancer patients is soaring. From 1997 to 2004, the number of Americans with cancer jumped, to 9.6 million from 9.4 million. Cancer strikes one in two men and one in three women in their lifetimes.
Most people want some sort of control, a way to prevent the disease from ever striking them or, if it does strike, to keep it from recurring. Many think of diet as a strategy.
Dr. Tim E. Byers, a professor of preventive medicine at the University of Colorado Health Sciences Center in Denver, was convinced that up to 20 percent of cancers were being caused by diet and he wanted to be part of the exciting new research that would prove it.
"I felt we were really on the cusp of important new discoveries about food and how the right choice of foods would improve cancer risk," Dr. Byers sad.
That was 25 years ago, when the evidence was pointing to diet. For example, cross-country comparisons of cancer rates suggested a dietary influence.
"For prostate cancer, if you look around the world, there might be 50-fold or greater differences in rates; they're huge," said Dr. Meir Stampfer, a professor of epidemiology and nutrition at the Harvard School of Public Health. "There are also big differences, many-fold differences, around the world for breast cancer and colon cancer."
And when people move from low risk countries to high risk countries, they or their children acquire the cancer rates of their new countries.
At the same time, some cancers were inexplicably becoming more common or, just as inexplicably, fading away in the United States.
In 1930, for instance, stomach cancer was the second leading cause of cancer death in women and the leading cause in men. Now, Dr. Stampfer says, stomach cancer is not even listed in the American Cancer Society's 10 leading cancers.
"So people think, 'What's happened in the past 70 years to make that change?' " he said. "Diet comes to mind." There were also differences in diets in countries where cancer rates were high and in those with low rates. With breast cancer, for example, researchers could draw a straight line directly relating the amount of fat in the diet to the rate of breast cancer in the population.
"People looked at it and said, 'Here it is - fat causes breast cancer,' " Dr. Stampfer said.
Next came studies that compared the diets of people who developed cancer to the diets of those who did not. Those studies, Dr. Schatzkin said, tended to show that dietary fiber protected against colon cancer, that fruits and vegetables protected against colon and other cancers and that a low-fat diet protected against breast cancer.
There were, of course, a few nagging questions. For example, people who had cancer might remember their diets differently.
"Whenever people get cancer, the first thing they ask is, 'Why me?' " Dr. Stampfer said. "And then they try to answer that question."
If colon cancer patients heard that fiber protected against colon cancer, for example, they might recall eating less fiber than people without cancer.
Dr. Stampfer said evidence from one of his studies indicated that was occurring, at least with fat and breast cancer. But, he said, when he published a paper saying so, "a lot of people didn't believe it."
The best studies are the hardest to conduct: prospective studies that that follow healthy people for years instead of looking backward and relying on memory. Even better - and harder and more expensive - are studies that randomly assign people to follow a particular diet or not.
But those more difficult studies were well worth doing, researchers said. And as more studies started, scientists hoped for definitive evidence that diet affected cancer.
The Fiber Theory
But as the results from those studies have begun to roll in, many researchers say they are taken aback. The findings, they say, are not what they expected.
Fat in the diet, the studies found, made no difference for breast cancer. "For fat and breast cancer, almost all of the prospective studies were null," Dr. Schatzkin said.
Fiber, in the form of fruits and vegetables, seemed to have a weak effect or no effect on colon cancer.
The more definitive randomized controlled trials were disappointing, too, with one exception. A study reported in May found that women with early stage breast cancer who followed a low-fat diet had a 20 percent lower risk of recurrence.
Even so, the effects were just marginally statistically significant. The study's principal investigator, Dr. Rowan Chlebowski of the Harbor-U.C.L.A. Medical Center, said it needed to be repeated before scientists would be convinced.
Nonetheless, the study contrasted sharply with those preceding it. Several involved beta carotene and antioxidant vitamins like C and E, substances that scientists thought were the protective agent in fruits and vegetables. The idea was that antioxidants could mop up free radicals in the body, which left unchecked could damage DNA, causing cancer.
Beta carotene was of special interest. People who ate lots of fruits and vegetables had more beta carotene in their blood, and the more beta carotene in the blood, the lower the cancer risk.
But a four-year study that asked whether beta carotene, with or without vitamins C and E, could protect against colon polyps, from which most colon cancers start, found no effect. People who took either beta carotene, vitamin C, vitamin E or all three had virtually identical rates of new polyps compared to participants taking dummy pills.
Another study, of 22,000 doctors randomly assigned to take beta carotene or a placebo, looked for an effect on any and all cancers. It found nothing. Two more, involving current and former smokers, found that those taking beta carotene actually had slightly higher lung cancer rates than those taking placebos.
Studies of fiber and colon cancer were similarly disappointing.
The fiber hypothesis had enormous appeal. Carcinogens from food can end up in stool. But when people eat a lot of fiber, their stool is bulkier and so carcinogens would be diluted. Bulkier stool is also excreted faster, reducing the time that the colon is in contact with cancer-causing substances. Fiber also binds bile acids in the bowel, substances that can damage the colon and, possibly, result in cancer. And the intestines metabolize fiber into short-chain fatty acids that seemed protective against cancer. Fiber also binds bile acids in the bowel, substances that can damage the colon and, possibly, result in cancer. And the intestines metabolize fiber into short-chain fatty acids that seemed protective against cancer.
Adding to the case for fiber was the fact that when researchers fed rodents carcinogens, the animals were protected against colon cancer if they also ate a lot of fiber.
Based on these indications, the cancer institute financed two studies on high-fiber diets and colon polyps. In one, 2,079 people were randomly assigned to eat low-fat high-fiber diets or to follow their usual diets. In the other, 1,429 people were assigned to eat high-fiber bran cereals or wheat bran fiber or to eat cereal and bars that looked and tasted the same but that were low on fiber. Fiber, the studies found, had no effect.
"We had high expectations and good rationale," Dr. Schatzkin said. But, he said, "we got absolutely null results."
Now, the largest randomized study ever of diet and cancer is nearing completion, involving 48,835 middle-age and elderly women. The women were randomly assigned to follow a low-fat diet with five servings a day of fruits and vegetables and two of grains or to follow their usual diet. The question was whether the experimental diet could prevent breast cancer.
The study is part of the Women's Health Initiative, a large federal project. When it began, the dietary fat hypothesis was ascendant. But after it was under way, other, less definitive studies failed to find any association between dietary fat and breast cancer.
The Women's Health Initiative diet study's results should be ready early next year, said its principle investigator, Ross L. Prentice, a biostatistics professor at Seattle's Fred Hutchinson Cancer Research Center.
And if it fails to find an effect?
Dr. Prentice said he would still wonder. Maybe what matters is diet earlier in life, he said, or maybe the women in the study did not stick to their diets.
Others say they suspect they were simply na´ve about the cross-country comparisons that persuaded them in the first place.
"People drew inferences that were in retrospect overenthusiastic," Dr. Stampfer said. "You could plot G.N.P. against cancer and get a very similar graph, or telephone poles. Any marker of Western civilization gives you the same relationship."
Because of the striking differences in daily life between people in countries with high cancer rates and those in countries with low rates, diet may have nothing to do with the incidence of the disease, Dr. Schatzkin said. Or diet may play a large role but the questionnaires used to measure what people were eating might have been inadequate to find it.
"That's the problem." Dr. Schatzkin said. "We just don't know."
As for Dr. Byers, who once had such high hopes for the diet and cancer hypotheses, he says he is sadder now, but wiser. "The progress has been different than I would have predicted," Dr. Byers said.
Specific food can affect general health, he added, but as for a major role in cancer, he doubts it. He now believes that it is the amount of food people eat, not specific foods or types of foods, that may make a difference. "I think the truth may be that particular food choices are not as important as I thought they were," Dr. Byers said.
Meanwhile, patients and those worried about cancer are adopting their own idiosyncratic dietary paths. Many know that the evidence is not solid, but they would rather take a chance that their diets will make a difference than wait helplessly for their fates to play out.
That is the view of John Napolitano, a New York graphic designer and marketer.
Three years ago, when he was 55, Mr. Napolitano found out that he had prostate cancer and that it had spread to his bones. Now, hoping to slow its progress, he avoids sugar and fat and almost never eats meat. He eats natural and organic foods. He drinks lots of water and green tea. He starts each day by whipping up a smoothie with a protein supplement and flaxseed.
"My diet is very different now than what it was three years ago," Mr. Napolitano said, adding that thinks that his new diet helped.
"Until recently, I was totally symptom free," he said. "I can't endorse anything I'm doing, but I've never had nausea, never had constipation" from his treatments.
Dr. Brad Efron, a professor of statistics at Stanford, has a different dietary approach. He does not have prostate cancer, but he had a couple of scares and he has friends who have it. So he is taking selenium, a trace mineral found in plants.
A study that randomly assigned people to take selenium or not to see whether it protected against skin cancer found that it had no effect on that cancer, but that the men taking it had only a third as many prostate cancers. Now, the National Cancer Institute is conducting a study on whether selenium protects against prostate cancer. Dr. Efron chose not to wait. He even published a statistical analysis concluding that the prostate effect was likely to be real.
"One of my colleagues said, 'Why do you think something that people thought would work on skin cancer has anything to do with you?' " he said. "There's always a leap of faith. But I'm scared of prostate cancer and I wanted psychological reassurance."