Dr. Peter Greenwald, 49, has been director of the Division of Cancer Prevention and Control at the National Cancer Institute for the past four years. In that position, he has provided the scientific backup to the institute for its dietary recommendations to reduce cancer risk. The division he heads is responsible for much of the research on diet and cancer being sponsored by NCI.

A medical doctor with a PhD in epidemiology from the Harvard School of Public Health, Greenwald is a member of the National Academy of Sciences' Food and Nutrition Board and the author of more than 100 published papers.

Greenwald was interviewed by Patricia Picone Mitchell, a free-lance writer who specializes in food and health topics.

Q. The National Cancer Institute says in its booklet "Diet, Nutrition and Cancer Prevention" that 35 percent of all cancer deaths in the U.S. may be related to the way we eat. How did you come up with that kind of estimate?

A. First of all, that's a very rough estimate. We don't have the information on diet to make as precise an estimate as we can for smoking. For smoking, it's also a third or so, or more than 30 percent, and we can be very precise about it. With diet, that figure actually came from two English epidemiologists, but there have been several other people that made similar estimates. The main basis for it is looking at the international differences in cancer, the fact that cancer rates vary markedly from country to country, and looking at what happens when people move from one place to another. When they move, their risk of cancer changes. And people tend to take on the risks of the place to which they moved. And for certain cancers -- those in the gastrointestinal tract, and certain endocrine-related cancers like breast and prostate -- the leading hypothesis for what affects these changes is diet. And when you add up the magnitude of the differences of those cancers where the problem is thought to be at least in part related to diet, you can come up with that figure. If you live here in the Northeast, your risk of colon cancer is higher than in some parts of Florida where 40 percent of the population moved from the Northeast. There are population differences. If you live in Connecticut, it's 70 percent higher than in Utah.

Q. What do people eat in Connecticut that's so different from what they eat in Utah?

A. There are some good studies of the Mormon population in Utah. As I understand it, they do have more fiber in their food. That's probably the major difference.

Q. But wouldn't a greater cancer rate in the Northeast suggest some environmental factors?

A. There's more than one factor. You have to know about smoking rates for some cancers, although they tend to be different types of cancers. There are certainly more smokers in Connecticut than Utah. I don't think we can attribute much of today's cancer to synthetic chemicals in the environment, if that's what you mean, except for certain groups of workers that have been exposed to things like asbestos or some other industrial chemicals. But you can look back before many of those chemicals were even developed and find cancer rates and not tremendous changes in time since those chemicals were introduced. It's only in selected groups with specific exposures that there's really a major concern.

Q. NCI has been telling consumers that diets high in fiber and low in fat, with plenty of fresh fruits and vegetables and whole-grain breads and cereals, may reduce their risk of cancer. But some experts believe the data linking diet and cancer is too preliminary to be making such recommendations. What made NCI decide to publish the recommendations now instead of waiting for better evidence?

A. We have a great deal of evidence now, and over the past five or 10 years there's been a growing consistency of evidence coming from different research disciplines. If we look at the human epidemiological data -- that is, what are the differences in cancer rates in different parts of this country and in different parts of the world -- people don't really realize what the statistics show. In the United States, deaths from cancer of the colon are more than three times the rate in Finland, and there has been careful study of the diet here and in Finland. The biggest difference is fiber. The Finns eat about twice the amount of fiber as we do. There was a study in northern India amongst railroad workers that showed that the railroad workers who ate a lot of vegetable fiber had much lower rates of colon cancer than people in Southern India, where they have a much more refined high rice diet.

Now there are some studies that don't agree. About three-quarters show the relation with fibers. None of the studies to my knowledge go in the other direction, but there are a few that show no difference. We think we can explain that apparent discrepancy. We have examined all those studies and they tend to be done in populations that don't eat much fiber, so it's hard to expect a benefit among any of the groups. Or they use methods for measuring diet that are fairly crude, and we do have a research problem in that it's hard for people to remember what they ate and there's some inacurracy in finding what they ate. Well, the effect of that is to weaken the research ability to detect a difference, and so you get some negative studies.

Now beyond that we have now a lot of laboratory research studies, studies in animals, for example, that show that certain fibers, some of the insoluble fibers, tend to prevent cancers in animals, particularly the chemically induced cancers in animals, and there is a consistency between the animal studies and the human. Again, I'm not saying that it's total agreement or total evidence, but there is quite a lot of consistency.

Q. Didn't the (U.S. Departments of Agriculture and Health and Human Services) dietary guidelines advisory committee that just revised the guidelines state that there was no firm evidence linking diet and cancer?

A. There might be some slight differences, I'm not sure. I think we're quite consistent. When people ask us, they want to know, really, how can they make a judgment about what they're doing. So if you say, "Double your fiber," it's more understandable. They want somebody to get a handle on it, just what it is you're recommending. We try to give that. But I think that what we're saying is basically the same, very consistent with the others.

Q. As far as that recommendation to double your fiber, you're really the only expert body in or out of the government that has actually come out with a number. Did you stick your neck out to do that?

A. I don't think so. We have people here that are nutritional scientists that have been reviewng all the information there is on fiber, and have been calculating what the number of grams of fiber is in the U.S. diet. And it comes out quite low. There was an international meeting about a year ago, at which they had many experts on fiber from around the world, and at that meeting they were generally recommending about 30 grams of fiber a day.

Q. Do you feel it's appropriate for companies to sell products containing fiber -- as Kellogg has done in its All-Bran campaign -- based on the available evidence?

A. I think Kellogg has been very responsible, as have been a number of other companies in that they're very accurately portraying what we think are reasonable recommendations. They're not saying that one product prevents cancer. The information is that a variety of foods that are high in fiber and low in fat may reduce your risk for some cancers. And I think that's been accurately portrayed. I think there is a concern, a concern that certainly the Food and Drug Administration has and that we understand, that there might be some exaggeration or distortion about what's known. So I would say it's something the Food and Drug Administration and the Federal Trade Commission have to be aware of and assure that the public is accurately informed.

Q. The U.S. has among the highest incidence of breast cancer in the world. Is it because the U.S. diet is so high in fat?

A. We think the major risk factor that can be modified is fat in the diet. Now there may be some role or interaction with other things like obesity or total calories. But fat first of all is the most concentrated form of calories, and you're most able to convert it into fat because it doesn't have to be modified. It uses less energy to convert than carbohydrate or protein. So we think the best thing you can do is to cut down fat in your diet if you want to limit breast cancer risk, and keep trim. The basis for that, again, comes from a number of epidemiological studies much like the ones I described with fiber. The countries that have about half of the fat intake as the United States have slightly below half of the breast cancer mortality rates. So it appears to be a fairly big effect.

Q. One of the criticisms of these recommendations is that a lot of the evidence is based on population studies, and population studies don't really show cause and effect.

A. I think that you can show it, that what you have to do is really look at what are the criteria for evidence, how rigorous is the research, and for the people that say the evidence isn't there, I think they, too, have a responsibility to define just what the studies are that they think are missing and how would you go about them. It's quite hard to do. We've spent a lot of time and thought thinking about just what is the research that would be convincing, just how can you interpret the fairly large body of evidence that exists now, and what would be added to it with more information. Would what would be added be likely to change something like the fact that fat is a risk factor for cancer? I think that's extremely unlikely. There's just too strong and diverse a body of evidence. It's more likely that it would only refine it.

Q. NCI has recommended that people eat foods high in Vitamins A and C. What's the evidence that these vitamins have anticancer effects?

A. There's somewhere over 20 studies that suggest that those eating diets higher in green and yellow vegetables have somewhat lower cancer rates. And those studies pertain to cancers of the gastointestinal tract, lung and several other sites, bladder, breast. Again, we have additional evidence from the laboratory with specific substances. Now some of these are not naturally occurring and not what you can buy in health food stores. They're synthetic analogs of Vitamin A, for example. Chemical cousins. They're called the retinoids. And it's possible to reduce the frequency of some experimental cancers in animals with these substances. And you can get fairly marked effects. For example, there's an experimental bladder tumor in rats that's caused by a carcinogen. And the frequency can be reduced from 40 percent down close to 5 percent with some of these inhibitors, retinoids. There's an experimental mammary tumor in rats that can be reduced from 100 percent down as far as 20 percent with some of the best of the inhibitors in certain experimental situations. And we're finding enough examples of that with the same cancer inhibitors working in different species and different cell systems that we think there must be something useful in it for people.

And therefore, the National Cancer Institute has begun two dozen cancer prevention trials. These generally take people at high risk for cancer, they're randomized into two groups. One group gets a pill containing either a nutrient, sometimes Vitamins A, C, E, selenium, betacarotene, or they get a placebo that looks and tastes the same. And these people are now being followed to see whether and to what extent we can lower their cancer rate. The first results will be in in 1989 to 1992. It takes some time for cancer to develop, so the studies take a number of years.

Q. Why can't we get the effects that we may be seeing from foods from these nutrients just by taking a vitamin supplement?

A. We don't know. It's possible that we might, but we just don't know. And we don't feel that we have enough research information to be making any recommendation along those lines. And I think also for the very large unbalanced dosages that some people tend to take, we don't really have adequate information on safety. Now I'm not saying that about the average multivitamins.

Q. But are you giving people megadoses in these trials?

A. Some of them have fairly large doses. It's worked out in each study. A lot of them have dosages at the upper end of the usual intake for large populations. But it depends on the study.

qa What about other nutrients and their possible anticancer effects?

There have been a number of possibilities about which there is some speculation or evidence. We think we need more studies. The endols, substances that are present in the cabbage family of vegetables, are ones that have been tested in the lab a bit, and we're bringing them along to see whether they should be studied. The question of calcium in relation to colon cancer has been raised. Some of the B vitamins, like folate, are in several of the studies. But again, I want to be very clear that we do not feel that we have any specific information on the vitamins to be recommending vitamin (supplements). We're recommending foods high in these substances. I don't know where we'll be in a few years. We think there's enough to do the studies, obviously.

Q. There is a group of scientists who have just been really resistant to these types of recommendations. And they still are, even though the evidence keeps coming in. Are they just sticking their heads in the sand, or do they have some legitimate concerns that we might be telling people to change and then find out later that they shouldn't have changed?

A. There are always differences of opinion, of course. I think a lot of the leads in this case came from epidemiology, from carcinogenesis studies, from molecular biochemistry and such things, and there are only some of the more traditional nutrition departments that are now getting involved in the chronic disease area. A few of them have been for a long time, of course, and we're glad to see their involvement. I think we need people from different research areas working together to develop these fields, and there has been some debate, some of it probably more emotional than is useful. But I think that it's healthy.