EDITOR'S NOTE: Dr. McCullough added the following statement 4/8/14 in response to questions related to sources of isoflavones:
Research on soy and cancer is highly complex, controversial, and evolving.
When concerns about soy are raised, they generally focus on findings from rodent models of cancer which tend to use isolated soy compounds like soy protein isolate or high doses of isoflavones (compounds found in soy). However, soy is metabolized differently in humans than it is in mice and rats, so findings in rodents may not apply to people. (See: http://www.ncbi.nlm.nih.gov/pubmed/16614407 for more on this.)(Setchell, AJCN, 2011). There is no evidence in the medical literature that soy protein isolate is bad for humans, compared to other forms of soy. Soy protein isolate is often used as a supplement in randomized studies of the effects of soy on health and none of these studies have shown harm.
Most of the studies suggesting benefits of soy consumption in people have measured how much soy foods people are eating, including tofu, soybeans, and soy milk. These foods are more commonly eaten in Asian countries. In the U.S., purified forms of soy are used in the food supply, including in energy bars and soy hot dogs. The few US studies that have measured these forms of soy do not suggest harm.
More research is needed to understand the relationship between specific forms of soy and doses of isoflavones on cancer risk and recurrence. We also need to learn more about childhood exposure to isoflavones and risk of cancer. Until more is known, if you enjoy eating soy foods, the evidence indicates that this is safe, and may be beneficial (but note that miso, a fermented soy product, is high in sodium.) It is prudent to avoid high doses of isolated soy compounds found specifically in supplements, as less is known about their health effects. As for other "hidden" sources of soy proteins, the evidence to date does not suggest harm or benefit. However, if you are concerned about these products, you can choose to avoid them.
Before writing a blog about soy and breast cancer, I took an informal poll of a few friends to get a sense of what women believe about soy. I asked them, "What do you know about eating soy food? Is it good for you? Not good for you?" (I didn't even mention breast cancer.) The responses I got were, "I think it acts like estrogen in the body"; "Consuming any soy products increases the risk of breast cancer"; "I don't eat it a lot because I heard something negative but I can't remember what it was;" and "I've heard you should only have it in moderation." Apparently, people are hearing that soy may not be good. But what's the truth? In this blog I'll walk you through what we know and what we don't know about soy and breast cancer, and give you some practical tips on eating soy.
Soy can act like estrogen
Soy foods (such as tofu, tempeh, edamame, miso, many veggie  burgers, and other products made with soy flour) contain isoflavones,  which are chemically similar to estrogens. Two major types, genistein  and daidzein, can act like estrogen in the body, although at a very  small fraction of the potency of circulating free estrogen in women.   These effects can be good or bad.  Let me explain.    
It is well established that estrogen is linked to hormonally-sensitive cancers in women, such as breast and endometrial cancer.   Breast cells contain estrogen receptors, and when the "key" (estrogen)  joins with the "lock" (the estrogen receptor), a series of signals are  sent which can spur on estrogen-receptor (ER) positive breast tumor growth. Common risk factors for breast cancer include conditions that involve longer exposure of breast tissue to estrogen:
In women who have been diagnosed with ER-positive breast tumors, tamoxifen and raloxifene are often used to block estrogen receptors so that estrogen can't reach them and stimulate breast tumor growth. 
Animal, human studies show different results
Ok. Now back to soy. Most of the concern about soy has come from  studies in laboratory animals.  Rats injected with ER-positive tumor  cells were given varying doses of genestein or daidzen.  Those given  more of the isoflavones had a greater growth of the breast tumors  compared to rats given little or no isoflavones.  But not all animal  studies have shown harmful effects.
Furthermore, while isoflavones may act like estrogen, they also have anti-estrogen  properties. That is, they can block the more potent natural estrogens  from binding to the estrogen receptor.  In addition, they stop the  formation of estrogens in fat tissue and stimulate production of a  protein that binds estrogen in the blood (to make it less able to bind  to the receptor).  They also have anti-oxidant and anti-inflammatory properties and work in other ways to reduce cancer growth.
Epidemiological studies, in which large populations of healthy  women have reported details about their usual diet and were followed for  many years, have either shown no association between soy and breast cancer, or a protective association, meaning that people who ate more soy had less breast cancer. 
In general, studies in Asian women have found a lower risk of  breast cancer with eating more soy, whereas studies in the U.S. have  tended to not find any association between how much soy a woman consumes  and her risk of breast cancer.  Indeed, a recent study combined data  from 14 epidemiologic studies on this topic and found that in Asian  countries, women who ate the most (compared to the least) soy  isoflavones had a 24% lower risk of developing breast cancer, while  there was no association in Western countries such as the U.S. 
Part of the challenge in studying this topic in the U.S. and  other Western countries is that women eat much less soy food than in  Asia, so evaluating the risk of breast cancer with high levels of soy  consumption is difficult.  For example, women who ate and drank the most  soy in a study in Shanghai consumed about 40 mg isoflavones/day  (roughly equal to 4 servings daily) and those who ate and drank the  least ate 5 mg/day (or half a serving). But in the U.S., the highest  category of consumption was 1-2 mg/day (less than half a serving per  day), and the lowest category was 'none.' In other words, women in the  highest categories in the U.S. would still fall into the lowest  categories in China. 
Keep in mind also that women in Asia likely have other cultural  dietary patterns that may not be assumed in the U.S., such as lifelong  consumption of soy and other foods that may have influenced breast  development during adolescence in ways not yet fully understood.  A  study in Shanghai found that women who ate a high amount of soy protein  consistently during adolescence and adulthood had a significantly  reduced risk of breast cancer before menopause, but not after menopause.
Soy for cancer survivors
What about women who have had breast cancer, especially  ER-positive breast cancer? A recent study looked at soy consumption in  the diets of more than 9,000 breast cancer survivors who were  participating in 3 studies of eating habits and other lifestyle factors  after breast cancer. Two of the studies were from the U.S. and 1 was  from China. Women from both the U.S. and China who consumed 10 mg/day or  more of soy had a 25% lower risk of breast cancer recurrence. These  protective associations were slightly stronger in women with ER-negative  tumors. In women with ER-positive tumors, the associations also seemed  protective (though not strongly so) for women regardless of whether they  were taking tamoxifen or not.    
But to find out for sure whether we should recommend soy foods to  women, researchers would need to repeat these findings, ideally through  a controlled study (considered the gold standard in research). At the  very least, the evidence from the studies in women reassures scientists  that moderate consumption is likely to be safe. 
The 2012 American Cancer Society Guidelines on Nutrition and Physical Activity for Cancer Survivors,  written by a panel of experts, including researchers with expertise in  this area, concluded that current research finds no harmful effects to  breast cancer survivors from eating soy.  Both the guidelines for cancer  prevention and the guidelines for survivors recommend against taking  soy supplements because they contain much higher isoflavone  concentrations than what you would normally find in the foods you eat,  haven't been as rigorously tested, and may have other potent effects on  body tissues.
When you make the decision about consuming soy, it's also  important to remember that breast cancer survivors (and the rest of the  population) are also at risk for other cancers and cardiovascular  disease. Tofu and other soy foods are linked to lower rates of heart  disease because they are excellent sources of protein, may replace other  less healthy foods in the diet (e.g. animal fats, red and processed  meats), and may help lower cholesterol, and so can be a good meal choice  for anyone. 
Bottom line: Even though animal studies have shown  mixed effects on breast cancer with soy supplements, studies in humans  have not shown harm from eating soy foods. Moderate consumption of soy foods appears safe for both breast cancer survivors and the general population, and may even lower breast cancer risk. Avoid soy supplements  until more research is done. So, enjoy your occasional tofu stir-fry or  tofu burger - they are unlikely to increase your risk of breast cancer  and, on balance, are some of the healthier foods you can eat!
http://www.cancer.org/cancer/news/expertvoices/post/2012/08/02/the-bottom-line-on-soy-and-breast-cancer-risk.aspx