 When the Women's Health Initiative study found that menopausal women taking hormone replacement therapy suffered higher rates of breast cancer, cardiovascular disease, and overall harm, a call was made for safer alternatives. Yes, estrogen has positive effects, the Women's Health Initiative found, such as reducing menopausal symptoms and improving bone health, reducing hip fracture risk, but also negative effects, increasing the risk of blood clots in the heart, brain, and lungs, as well as breast cancer. So ideally, to get the best of both worlds, we'd need what's called a selective estrogen receptor modulator, something that has proestrogenic effects in some tissues, like bone, but antiestrogenic effects in other tissues, like the breast. Drug companies are trying to make them, but phytoestrogens, natural compounds in plants like genestine and soybeans that are structurally similar to estrogen, appear to function as natural selective estrogen receptor modulators. How could something that looks like estrogen act as an anti-estrogen? The original theory for how soy phytoestrogens control breast cancer growth is that they compete with our own estrogens for binding to the estrogen receptor. As you drip more and more soy compounds on breast cancer cells in the Petri disk, less and less actual estrogen is able to bind to them. So the estrogen blocking ability of phytoestrogens can help explain their anti-estrogenic effects. But how do we then explain their proestrogenic effects on other tissues, like bone? How can soy have it both ways? The mystery was solved when we discovered there are two types of estrogen receptors in the body. And so how a target cell responds depends on which type of estrogen receptors they have. This may be the key to understanding the health protective potential of soy phytoestrogens, the existence of this newly discovered estrogen receptor called estrogen receptor beta, to distinguish it from the classic estrogen receptor alpha. And unlike our body's own estrogen, soy phytoestrogens preferentially bind to the beta receptors. If you have people eat about a cup of cooked whole soybeans, with about 8 hours' genistin levels in the blood, reach about 20 to 15 nanomoles. That's how much is circulating throughout our body bathing ourselves. About half is bound up to proteins in the blood, so the effective concentration is about half that. So let's see what that means for estrogen receptor activation. This is the graph that explains the mysterious health benefits of soy foods. Down around the effective levels, you'd get eating a cup of soybeans. There's very little alpha activation, but lots of beta activation. So now let's look at where each of these receptors are located in the human body. The way estrogen pills increase the risk of fatal blood clots is by causing the liver to dump out all these extra clotting factors. But guess what? The human liver only contains alpha estrogen receptors, not beta receptors. And so maybe if we ate like 30 cups of soybeans a day, that could be a problem. But at the kinds of concentrations one would get with just normal soy consumption, no wonder this is a problem with drug estrogens, but not soy phytoestrogens. The effects on the uterus appear also to be mediated solely by alpha receptors, which is presumably why no negative impact has been seen with soy. So while estrogen-containing drugs may increase the risk of endometrial cancer up to a 10-fold, phytoestrogen-containing foods are associated with significantly less endometrial cancer. In fact, protective effects for these types of gynecological cancers in general. Women who ate the most soy had 30% less endometrial cancer and appeared to cut their ovarian cancer risk nearly in half. Soy phytoestrogens don't appear to have any effect on the lining of the uterus, but still can dramatically improve menopausal symptoms. The cuppermen index is like a compilation of all 11 of the most common menopausal symptoms. In terms of bone health, human bone cells carry beta estrogen receptor, so we might expect soy phytoestrogens to be protective, and indeed they do seem to significantly increase bone mineral density, consistent with population data suggesting high consumption of soy products is associated with increased bone mass. But can they prevent bone loss over time? Soy milk was compared to a transdermal progesterone cream. The control group lost significant bone mineral density in their spine over the two-year study period, but the progesterone group lost significantly less, and the two glasses of soy milk a day group ended up actually better than when they started. This is probably the most robust study to date, comparing the soy phytoestrogen genestine to a more traditional hormone replacement drug regimen in the spine over a year. The placebo group lost bone density, but gained in the phytoestrogen and estrogen groups, and the same with the hip bones. The study clearly shows that the soy phytoestrogen prevents bone loss and enhances new bone formation, in turn producing a net gain of bone mass. But the only reason we care about bone mass is we want to prevent fractures. Is soy food consumption associated with lower fracture risk? Yes, a significantly lower risk of bone fracture associated with just a single serving of soy a day. The equivalent of 5 to 7 grams of soy protein, or 20 to 30 milligrams of phytoestrogens, so that's just like one cup of soy milk, or even better, a serving of a whole soy food like tempeh, or edamame, or the beans themselves. We don't have fractured data on soy supplements, though, so if we seek the types of health benefits we presume Asian populations get from eating whole and traditional soy foods, maybe we should look to eating those rather than taking unproven protein powders or pills. Is there anyone who should avoid soy? Well, some people have soy allergies. But a national survey found that only about 1 in 2,000 people report a soy allergy. That's 40 times less than the most common allergy in dairy milk, and about 10 times less than all the other common allergies, like fish, eggs, shellfish, nuts, wheat, or peanuts.