Just when everyone was convinced that the white, spongy blandness of tofu was the food medicine of the millennium, out comes a study likely to scare the soy right off your dinner plate.
A group of Hawaiian researchers who aired their study in the latest Journal Of The American College Of Nutrition have stunned the scientific community with their claim that the protein-rich substance can actually rot your brain as you grow older.
Their findings, which show a significant statistical relationship between two or more servings of tofu a week and "accelerated brain aging" -- even an association with Alzheimer's disease -- is hardly reassuring to bean curd aficionados.
Veggie types For years, veggie types have been regularly throwing the Oriental staple into their woks, encouraged by the literature touting the stuff as an elixir against heart disease, cancer and hot flashes.
I myself indulge heartily in this magic food and I'm not sure, based on this study alone, that I'll stop anytime soon.
The offending research comes from the Hawaii Center for Health Research in Honolulu, where a team collected data from 3,734 Japanese-American men enrolled in the Honolulu Heart Program.
The men, born between 1900 and 1919, were interviewed twice -- first in the late 1960s, and then in the early 90s -- about their consumption of 27 different foods and drinks. Their cognitive functioning was assessed, as well as that of 502 wives who lived with the subjects.
The data showed that the men who ate more tofu were more apt to have impaired mental ability, with tofu being the only consistent link.
"The test results were about equivalent to what they would have been if they had been five years older," says chief investigator Lon White. "Guys who ate none, their test scores were as though they were five years younger."
The brains of 300 men who died were examined in a unique autopsy study conducted as part of the Honolulu aging project. "What I saw was (that) the simple weight of the brain was lower," White says. Shrinkage occurs naturally with age, he says, but atrophy progressed more rapidly in those who had consumed more tofu.
"We can't be sure that the causal factor was really the tofu, but no alternative is apparent at this time."
Now, of course, one of the first principles of epidemiology is not to confuse a correlation with a cause-and-effect relationship. The question here is whether tofu consumption simply represents a marker for some other cause of brain damage, perhaps a lifestyle factor or food contaminant.
Susan Love, a women's health guru and adjunct professor of surgery at UCLA, is skeptical about the study. "It's like saying everybody who has lung cancer has a cigarette lighter in their pocket, so lighter fluid causes lung cancer."
She reminds me that the anecdotal data, at least in women, is that "tofu improves cognitive function."
Causal relationships, Love says, are always elusive. Consider the case of the thesis that hormone replacement therapy prevents heart disease. "Guess what? When they did the randomized controlled study, it showed no benefit.'
Noting that all the participants lived in one place, Love muses, "Maybe the tofu in Hawaii is fermented in some brain-numbing stuff, so it's not that soy is bad but tofu made that way is bad."
Teasing out the real variables to determine what causes an illness is a tricky business. Can data obtained from two interviews two decades apart be trusted to produce reliable and persuasive results?
One researcher who has his doubts is Claire Hasler, executive director of the Functional Foods for Health Program at the University of Illinois in Urbana-Champaign.
An expert on soy and heart disease, Hasler stresses that the Hawaii research is "only an associative study. In fact, East Asian countries have a lower incidence of dementia and Alzheimer's than European countries."
Hasler notes, "This may be an indicator, not a cause, of other risk factors for dementia. There are a lot of environmental factors involved in this. We've never noticed this before, and I would never make public policy as a consequence of this."
For the many boosters of soy as a superfood, of course, the study is a real downer.
Over the last decade or so, more and more evidence has been accumulating that the benefits of soy have to do less with its nutritional content than with some of its other molecules, which essentially act as drugs.
Chief among these are natural estrogens, or "phytoestrogens." Paradoxically, these exert both estrogen-like and anti-estrogen effects in the body, and may actually lower natural estrogen levels and alter cell metabolism.
A lot of this data is more suggestive than scientifically conclusive. Asian women, great soy eaters, report less frequent hot flashes and other menopausal symptoms than their North American counterparts. A study in the American Journal Of Epidemiology found that women who ate the most phytoestrogen-rich foods had a 54-per-cent reduction in their risk of developing endometrial cancer. In addition, a study in the journal the Lancet found that Australian women with high concentrations of phytoestrogens in their urine had lower rates of breast cancer.
Indeed, the incidence of breast, ovarian and endometrial cancer is significantly lower in China, Japan and other Asian countries than in North America and Europe.
The same is true of heart disease, and a study published in the Journal Of The American Menopause Society concluded that adding 20 grams of soy protein -- containing 34 milligrams of phytoestrogens -- to one's daily diet significantly lowers cholesterol and blood pressure.
More seriously This finding, among others, led the U.S. Food and Drug Administration to issue a rare health claim last October, giving American food manufacturers permission to label products high in soy protein -- everything from tofu, soy-based beverages, soy meat and soy protein bars and powders -- as possibly helping to lower the risk of heart disease.
In the context of all this positive soy news, it's difficult for many researchers to get their heads around the meaning of the new study. Still, some think there may be something to it.
Tom Clarkson, a professor of comparative medicine at Wake Forest University School of Medicine, who studies post-menopausal health in monkeys for a large U.S. National Institutes of Health-supported study on soy and health, notes, "It has a plausible mechanism" -- for men, at least-- "and that's why people are taking it more seriously than they might otherwise."
The cognitive portion of the male brain, Clarkson explains, is driven by a hormone called estradiol. Its production requires a certain protein -- aromitase -- which the phytoestrogens found in soy products have been shown to inhibit, although it's not clear whether levels of this enzyme in the brain are affected.
"There's a great level of uncertainty," Clarkson admits. "If (tofu) is an aromitase inhibitor in the brain as it is in other tissues, then you can imagine some mechanism. If, on the other hand, it is not present in the brain, then there's probably not a reasonable biological mechanism by which tofu consumption could adversely affect cognitive function."
But even if this mechanism holds true, this still leaves open the question of why this effect should be seen in women -- if indeed it is.
It's also unclear why White found no association between consumption of miso -- which contains many of the same phytoestrogens as tofu -- and cognitive function or brain aging. Perhaps something else is intervening.
That's the suggestion of Abby Lippman, director of graduate studies in the department of epidemiology and biostatistics at McGill University. A vegetarian, Lippman says she consumes tofu almost every day.
She says she buys her stash fresh from a Japanese store, and notes that "even though I change the water every day, soft tofu may in fact be sitting there collecting more crappy bacteria than the stuff that's packaged -- which, then again, may have preservatives in it that could be potentially harmful."
Pretty strong According to Bhagu Bhavnani, who studies menopause, hormone replacement and brain function at the department of obstetrics and gynecology at U of T's Institute of Medical Science, the data in the new study "looks pretty strong."
"The only thing I am concerned about is the interpretation regarding the effect in women, because that was essentially done by proxy rather than by direct questioning of the individuals."
He also notes that the study subjects were first interviewed in 1966, and only the men who survived have been followed. "What happened to the ones who didn't survive? What did they die of? No one knows that. That introduces what we call selection bias."
Further, Bhavnani says many misunderstand the source of soy's benefits -- it may not be the phytoestrogens at all.
He notes that while monkeys fed high-cholesterol soya diets had lower arterial plaque formation, when they were subsequently fed the pure phytoestrogen components, there was no benefit.
Other protein "So there could be something other than the phytoestrogens, most likely some other protein that is there that may be having some beneficial effect."
And he says soy is being overhyped in general, particularly for menopause. "Most of these studies are very poorly designed," he says.
In his estimation, soy doesn't take care of hot flashes or vaginal dryness and certainly doesn't prevent osteoporosis. "It's a very common disease in the Chinese population, who consume a lot of tofu and soya products," he says.
No one advises throwing out your bean curd just yet, but there is a lot of anxiety out there, and the next study is eagerly awaited.
"What you take from a study like this is not, 'Oh, my god, I should never eat tofu again,"' says Love, "but rather, 'Isn't that interesting?' Now we need to do a controlled study of randomized people to eat X amount of tofu and then test them in a year."
The Joy of Soy
- is a complete protein
- lowers "bad' cholesterol
- lowers blood pressure
- may reduce risk of endometrial, breast and ovarian cancer
- may help to regulate glucose in diabetes
- may lessen menopausal symptoms
- may not prevent osteoporosis as touted
- may diminish cognitive functions