That’s the point at which I discovered that there are actually compounds in plants that are hormonal, and studying their effects in humans seemed like a fascinating way to combine my training and interest in nutrition and endocrinology. I applied for the grant and ended up getting it, and most of my research program since then has been on phytoestrogens.

I was invited to write it. Although I don’t often write review papers, this one seemed particularly timely. My collaborator on this paper was a post-doctoral fellow in the lab, Xia Xu, and I asked him if he would write it with me. He said yes, and he was heavily involved in the writing. The editors of Annual Reviews had decided that it was a good time for a summary review on the subject of dietary phytoestrogens and I agreed. It was just at the beginning of public awareness about phytoestrogens, the foods that contain them such as soy and flaxseed, and their biological effects in humans. I decided that the time was right to get the word out to a broader audience.

We began by defining what phytoestrogens are and discussing their dietary sources and chemistry. We then discussed how they’re metabolized in the body. It’s particularly interesting that they’re metabolized in the gut by bacteria, and so gut flora exert an extremely important influence on absorption and exposure. The implication of this is that intestinal metabolism determines the level of phytoestrogens that a person is actually exposed to. So if you imagine a group of people all consuming the same quantity of phytoestrogens, each person will be exposed to different levels because of different profiles of gut flora.

We then discussed the ranges of typical physiological levels in humans, and biological effects, which included hormonal effects. The primary interest in these compounds at the time was in their relationship to cancer, to tumor cell growth, anti-oxidation, etc.

In the end we talked about some of the newer emerging research. Although there had been a lot of research done on phytoestrogens at the time, most of the research was in vitro or using animal models—few human clinical studies had been performed. At that time, for instance, there was very little known about the effects on menopausal symptoms, bone mass, osteoporosis, heart health, etc. There were very few human studies looking at cancer. So the last section of the paper focused on human health.

It’s very interesting to me in retrospect that our review paper was so brief. If we’d written a comprehensive review now, it could be ten times longer because so much more is known.

They’re mainly found in soybeans and flaxseeds. Soy contains phytoestrogens called isoflavones; flaxseed contains a related group of compounds called lignans. Isoflavones are highly concentrated in soybeans; they’re found in much lower levels in other legumes, and not found in significant levels in any other foods. On the other hand, lignans are widespread in foods—in whole grains and a number of berries and fruits—but the most concentrated source used in research studies is flaxseed. Flaxseed is not a traditional food. Another name for it is linseed, which is used to make textiles. Flaxseed is currently used in health foods mainly as a concentrated plant source of omega-3 fatty acids. We were interested in flaxseed because of the concentration of lignans, and my research group performed one of the early studies looking at the effects of flaxseed lignans on hormones in humans, which may be what prompted the Annual Review editors to ask me to write the review.

I think it’s mostly a sign of how big the field has become, and that this was one of the early reviews. And, of course, the Annual Reviews are considered an excellent source of this kind of information. There have been many other reviews written on the subject, in all kinds of different journals, but I think the fact that it was in Annual Reviews gives it more credibility.

It’s just exploded, particularly in the area of soy effects. There’s an interesting history here. The soy industry is actually largely responsible for the enormous amount of research that’s been done. I give them a lot of credit for that. They supported the early research and were extremely smart in understanding the importance of research in evaluating their products and determining whether or not there were health effects. They believed that there were, and they understood that research would help them.

So a number of companies were supporting this research, and a number of states with big soybean crops have funds through which the growers devote a percentage of their profits to research. By the 1990s, there was enough research, primarily in animals, suggesting that soy is beneficial and has interesting properties, and that isoflavones could potentially have physiological properties in humans, that the federal government started looking at this as a real opportunity in public health and an area to be funded. So at that point the NIH started funding soy research, as did the USDA.

Since the 1990s, an enormous amount of research has been done on soy, and a lot on flaxseed. Most of the federal funding comes from the National Cancer Institute, because of the interest in cancer prevention. There have also been studies looking at the effect of soy and isoflavones on risk factors for heart disease, on hot flashes, and on bone health. All of these research areas have really taken off. There’s also been funding from The Institute of Environmental Health Sciences. They’ve been concerned about the negative potential of phytoestrogens acting as what they call endocrine disruptors.

I would say that the enthusiasm for soy has definitely diminished over the past couple of years, primarily because of a few publications that have recently come out suggesting from animal research that soy might actually promote breast cancer. Those papers sparked some immediate concern. We’ve also confirmed what many of us suspected all along, which is that even though it may exert some health benefits, soy is no magic bullet.

Soy consumption certainly has some very specific health benefits, aside from the fact that it’s a good source of protein. I would say it definitely eases hot flashes, but nowhere near as effectively as estrogen replacement. It certainly lowers cholesterol, but nowhere near as effectively as the pharmaceutical statins. Those are the two areas in which the data are strongest.

When it comes to bone health, the animal data and human data are fairly suggestive of a benefit. But the studies that need to be done to prove it are two-year and three-year studies, and those are just now being done. My suspicion is that even if soy is protective of bone health, it won’t be as effective as the pharmaceuticals used to reduce bone loss.

When it comes to breast cancer, the data suggest, both from animals and humans, that soy consumption during the years of breast development, which means childhood and pre-puberty, is cancer preventive, because it actually alters the structure of the breast making it less vulnerable to cancer. There are also intriguing data that soy may help prevent prostate cancer, but more studies in humans need to be done before we can make conclusive statements about this.

I have continued to do a lot of research with soy. I’ve done studies in pre- and post-menopausal women looking at hormonal effects. I’ve looked at the effects on blood lipids and bone metabolism, and I’m just finishing a study in men at high risk of prostate cancer. We’re looking at the effect of soy on intermediary biomarkers of prostate cancer.

My research program has stayed in the area of diet and cancer prevention, although I’ve broadened my focus to expand beyond phytoestrogens. I’ve recently been involved in a study of cancer-preventive effects of omega-3 fatty acids from fish in post-menopausal women, and we’re just beginning a study looking at exercise in young women to see if it has any cancer preventive effects.

That we don’t have any really good biomarkers for cancer and for breast cancer in particular. For example, with cervical cancer, we can look at Pap smears; we can do an intervention and use a Pap test as a marker of what’s happening in the cervix. With prostate cancer, we thought for many years that PSA was a reasonable biomarker. We could do a study and look at PSA as an outcome variable and we wouldn’t have to wait decades to get a meaningful result. But now PSA is no longer viewed as a good biomarker, because there are too many false positives.

We don’t have good biomarkers for breast cancer risk, which makes cancer prevention research in humans extremely difficult. We look at a number of different biomarkers that may be connected with breast cancer, but we really don’t know what they mean. If we take estrogen, for example, epidemiological studies suggest that circulating estrogens are associated with breast cancer risk. Does that mean if we do an intervention that causes women to have somewhat higher estrogen levels, they will get more breast cancer? We don’t really know.

So there are missing links, serious gaps in our knowledge when it comes to cancer prevention research in humans. We try to put as many pieces of the puzzle together as we can, but the limitations of the studies restrict severely what we can recommend to people who would like to lower their risk of cancer through dietary changes. Obviously when you’re doing cancer research, the very best thing would be to do enormous studies and actually look at cancer as the endpoint. But that would take many, many years, tens of thousands of subjects, and far too much money.

Before I answer that, let me go back to something I was saying earlier, which is that the public has been extremely interested in soy, soy supplements, and isoflavone supplements, and people have wondered if they consume soy, will it do all kinds of great things for them? Now that the data are showing only modest benefits and not enormous benefits, and perhaps even some negative effects, there’s a level of disappointment that has the same energy as the previous level of enthusiasm. People are just enormously disappointed. And the problem, I think, is that too many people want the health benefits of these foods to be the equivalent of the benefits they might get from taking a drug. But this is a food, it’s not a drug. You’ll never get the same kind of effects, and this is actually a good thing. If foods acted like drugs they would have to be regulated like drugs, and there would be concerns about toxicity, interactions, dosage, etc., that we would not want to worry about.

So I think it’s important that the public moderate their expectations of the kinds of benefits they might get from a specific food. And I would say when it comes to assessing the health benefits of specific foods or food components, we should all be careful about what sources we go to for information. I teach a freshman-level course on general nutrition in which we spend time teaching the students how to assess the sources of nutrition information.

The public needs to become more aware of how to assess this information as well, so they can learn to moderate their expectations. An advertisement in a magazine, for instance, is probably not the right place to get nutrition information. The public is rightfully confused by all the nutrition information in the media and it’s a general problem that we in the nutrition and medical communities have to deal with. In the meantime, we should try to teach the public to take a lot of the latest nutrition news with a grain of salt, and learn not to expect a food to act like a drug or be a magic bullet when it comes to our health.

When it comes to soy, I would say that soy is a great source of protein, and in addition has some interesting effects that are likely beneficial. Soy modestly lowers cholesterol and reduces hot flashes; although small, the benefits may be meaningful when soy is consumed over a lifetime and in combination with other beneficial phytochemicals. Soy probably reduces bone loss with menopause, but we’ll get more definitive information on this within the next two years or so.

As far as breast cancer goes, soy consumption appears to be beneficial when consumed before puberty, and the effects of consuming it postmenopausally are controversial. There is a small theoretical risk that it could stimulate breast cancer cell growth in women who already have cancer, but this is not serious enough for me to dissuade women from consuming soy, even high-risk women, unless they are not regular soy consumers already. There are interesting data suggesting that soy may help prevent prostate cancer, but again, this is an emerging area. If you remember that soy is a food and don’t expect drug-like effects, the benefits are actually quite remarkable.