
There has been a lot of bad press about soy lately, and I wanted to take the opportunity to discuss some of the recent controversy surrounding this humble legume. It can be hard to find a balanced, evidence-based review of the issues surrounding soy, since most people seem to either love it or passionately hate it. I hope to consider both sides, and examine the research to give a relatively unbiased viewpoint. Apologies for the length of this post, but it's a very complex topic, and it still barely scratches the surface! If you are not interested in all the details, just skip to the end of the post for the bottom line.
Soy has been consumed for thousands of years in Eastern Asia, and is a mainstay of modern vegetarian diets. Unlike many other legumes, soy is a "complete protein," meaning that it contains all the amino acids that our bodies cannot make on their own. (It should be noted that although other vegetarian sources of protein may not contain all of the essential amino acids, it is very easy to get all of these amino acids by eating a diet with plenty of variety.) Soy is incredibly versatile, and can be consumed in its unprocessed form as
edamame; fermented into soy sauce,
miso,
natto, or
tempeh; or processed into soy milk and tofu. More "modern" products such as soy flour, textured vegetable protein (
TVP), and isolated soy protein have also become parts of our diet as ingredients in
faux meat products and soy protein powders.
Like all plants, soybeans contain a variety of
phytochemicals. Soy is unique from other food plants in that it contains
isoflavones, otherwise known by the less accurate (but more emotionally charged) term
phytoestrogens. These compounds are complex--there are 12 different types of soy
isoflavones, and their level of activity depends on many factors, including how the soybeans were processed. However, their
estrogenic activity is weaker than natural
estrogens, and they can also exert anti-
estrogenic effects, in addition to acting as antioxidants.
In fact, most soybeans (at least in North America) are not grown for humans to consume directly--they have become a major commodity crop. Soybeans are principally grown for the fats they contain; these are extracted and processed into soy lecithin, liquid soybean or partially hydrogenated soybean oil. (Though I expect that the manufacturing of the latter has decreased as a result of the rallying cry against trans fats.) The resulting
defatted soy flour is often used for animal feed. As such, huge
monocrops of soybeans are grown across the Midwest (supported by federal subsidies), and are targets for some of Monsanto's genetic engineering efforts.
You can imagine the North American soybean industry's delight when research in the 1980's and 90's started showing that consumption of soy could lead to decreased risk of heart disease, and may reduce menopausal symptoms--thus opening up an entirely new market. Tofu wasn't just for hippies anymore, and soy protein was added to many products, backed by a health claim from the Food and Drug Administration stating that increased soy protein intake could lower risk for heart disease. As a result, consumption of soy increased dramatically.
As is often the case in nutritional science, the pendulum soon swung back the other way. (An aside: It is true that nutrition advice often seems to be contradictory; one week it seems that chocolate is good for you; another week it should be avoided at all costs. This unfortunate situation arises from the fact that media outlets often seize on the results from single studies, regardless of their scientific merits, that have results that people want to hear, such as "chocolate is good for you," or, in this case, "all that soy that you grudgingly ate because you thought it was good for you is actually harming your health." For a discussion on different types of studies and why some are considered more credible than others, take a look at
this NYT article. And, remember that results from a single study never prove anything; multiple studies must be done to confirm findings.) The more research that was done on soy, the more the potential health effects seemed to go away; a few studies even suggested some adverse outcomes. These results were also publicized by several crusaders against soy, who began to declare that we are all poisoning ourselves by eating tofu. Their strong language and seemingly scientific reviews of the topic left some of us vegans wondering if, indeed, they were on to something. (Incidentally, you can read an excellent refutation of many popular anti-soy claims by John Robbins
here, or see below for my take on a few.)
So, what does the current research say about the safety of soy? First, a quick caveat: the body of studies is hard to interpret because not only do the experimental designs vary, but many studies were done in animals rather than humans, and different forms of soy (from isolated protein or
isoflavones to whole soy foods such as tofu and
tempeh) were studied. Below is what I have gleaned from my readings; you can link to many of the original studies if you are interested.
Soy "toxins":Uncooked soybeans contain
trypsin inhibitors that hinder digestion, causing severe gastrointestinal distress. Cooking or fermenting the soybeans deactivates most of these inhibitors, but the anti-soy community uses the presence of these compounds as evidence that soy is loaded with toxins. They also point to
phytates, compounds which bind minerals such as calcium and iron, making them less
absorbable by the body.
Phytates too are somewhat deactivated by cooking, and occur in many plant foods besides soybeans. While these may contribute to malnutrition in developing countries where food sources are limited, this slight decrease in mineral absorption is not of concern in individuals who eat a sufficiently varied diet.
Soy and heart disease, menopausal symptoms, and cancer:Although previous evidence showed that soy protein can help reduce cholesterol levels, resulting in the
FDA's heart-healthy claim for soy foods, a
recent review by an expert panel of the American Heart Association found that soy has very little effect on cholesterol. However, many of these studies were done with isolated soy protein or
isoflavones, and it should be noted that the researchers concluded that whole soy products have a favorable nutrient profile that should help reduce the risk of cardiovascular disease. This same review also found that soy products have not been shown to reduce symptoms of menopause; results were mixed regarding soy's effect on postmenopausal bone loss. Some studies suggest that soy may help prevent cancer, but that is not effective for cancer treatment.
One study in mice suggests that soy may even exacerbate certain types of hormone-sensitive breast cancer. These results have not been confirmed in humans, but many clinicians recommended that women with this particular type of cancer should not consume soy. Thus, the bulk of studies point to soy having little to no effect on these health outcomes, but again, many of these did not use whole soy foods.
Soy and Thyroid Function: Soy also contains compounds called
goitrogens, which affect thyroid function and could cause goiters. This seems to be of concern only if iodine levels are low (after all, how many of us tofu-eating vegans have goiters?). Sources of iodine include iodized salt (sea salt, however, does not contain high concentrations of iodine) and seaweed.
Soy and maternal and infant health: "Soy formula is birth control for babies," claim some anti-soy groups. Yes, it is true that soy formula contains
isoflavones, but again, these compounds have only weak
estrogenic effect and exert multiple functions in the body--drinking soy formula is not at all similar to taking birth control pills. A thorough
review in 2004 concluded that "available evidence from adult human and infant populations indicates that dietary
isoflavones in soy infant formulas do not adversely affect human growth, development, or reproduction." Although the American Academy of Pediatrics does not recommend routine use of soy formulas for infants (one exception being for "situations in which a vegetarian diet is preferred,") their 2008
review of the literature found "no conclusive evidence from animal, adult
human, or infant populations that dietary soy
isoflavones may
adversely affect human development, reproduction, or endocrine
function." Of course, breast milk is the best source of nutrition, but for vegans who cannot breastfeed, soy formula is a viable option that does not seem to cause harm. Many, many people have been raised on soy formula with no apparent ill effects. It is likewise safe to consume soy products while breastfeeding (only a small amount of
isoflavones are excreted in breast milk), and for infants to consume soy once solid foods are started (though, again, it is best to stick to traditional foods such as tofu and
tempeh). Further, consumption of soy products during pregnancy does not seem to result in adverse outcomes for mother or fetus.
Soy and Cognitive Function: "Soy products may increase dementia risk" read the headlines this summer, as the results of an Indonesian study were publicized. While the study did find that increased tofu consumption was associated with worse memory, the same study also found that increased
tempeh consumption was associated with better memory. These results suggest that it might be something other than soy itself that was related to cognitive decline (and, it must be noted, this was not a randomized trial, so it cannot establish cause and effect). The authors note that because
tempeh is fermented, it has higher levels of
folate, which is protective against cognitive decline. (You can read the abstract
here.) However, they also noted that the results could have been due to toxic compounds such as formaldehyde, which is apparently sometimes used as a preservative in Indonesia. This possibility was not controlled for in the study, so it certainly seems possible that this could account for the results. And what about the Honolulu Asian Aging Study, which is often cited as showing that soy products increase risk of dementia? This study was quite comprehensive and well done, and does seem to suggest an association between tofu intake in midlife and accelerated brain aging. The only major limitation of this study is that it was done on a very specific population--Japanese-American men living in Hawaii--and so the results can't be generalized to other populations. Still, this result should give us pause. If you want to wade through the results of this study, it can be found
here.
Soy and Male Infertility: For a great review of the highly-publicized study released this summer, check out
this post. Needless to say, many vegan families the world over have proven that soy does not cause infertility.
ConclusionsIn conclusion, while soy may not be the miracle food it was once touted to be, it is not the toxic poison that some groups would have you believe. Many of the studies showing adverse health effects were done in animals consuming large amounts of isolated
isoflavones or soy protein--conditions far removed from what we encounter in our everyday lives. (And, as John Robbins wisely says, "If you made it your policy to eat no food that contained substances which can in large enough concentrations cause damage, there would be literally nothing left for you to eat.") That said, it would be prudent to avoid eating soy in excess.
While researching this post, I marvelled at the complexity of this topic. The thousands of
phytochemicals present in soybeans, the
isoflavones themselves, the studies examining potential health benefits and risks--all are incredibly complicated, making it very difficult to disentangle all the variables to try and get to some sort of "truth." Please keep this in mind whenever you are reading any mainstream news article or sensationalist diatribe that oversimplifies the issues and leaves the impression that it is all black and white, good or bad, with no in between. Once again, as usual, it comes down to moderation. A diet that puts too much emphasis on any one food, be it soy, red meat, refined grains, etc., is not going to be health promoting. A diet that includes a variety of fruits and vegetables, whole grains, and lean protein sources has been shown time and again to reduce the risk of chronic disease. Thus, it is important to put this discussion in the context of the whole diet.
While I believe that the long-standing use of soy products in Asian cultures is further proof of their safety, it is true that vegans have the potential to have a much higher intake of soy than has been seen in traditional cultures. For that reason, we should be careful not to rely too heavily on soy as a protein source--there are so many other nuts, seeds, legumes, and whole grains that can fill this role.
The Bottom Line:Soy is a great source of complete protein, and likely poses no health risk within the context of a well-rounded diet containing a variety of fruits, vegetables, whole grains, nuts, seeds, and legumes. However, with all the
faux meat products and soy-enriched products that abound these days, it is very easy to eat soy in amounts that may be detrimental to long-term health. My recommendations are:
- Enjoy soy in moderation, in the forms that have been consumed by traditional cultures for thousands of years (tofu, tempeh, miso, natto, soy milk, and soy yogurt). Aim for no more than 3 servings of these foods per day. One serving of tempeh or tofu is 4 ounces (about 1/2 cup); a serving of soy milk is 1 cup.
- Limit consumption of products containing soy protein concentrate or soy protein isolate (many faux meats and protein powders) to one or less per day.
- Don't feel that you have to eat soy to be healthy; only eat it if you enjoy it! Many other great sources of vegetarian protein exist. Lentils, beans, quinoa, almonds, sunflower seeds... the list goes on!
I hope that this post cleared up some of the controversy; feel free to leave any comments you have about soy. Obviously, this is a complicated topic and the research is still ongoing.