In the 12 months ending in October 1999, supermarket sales of soy foods were up 45 percent over the previous year, to nearly $419 million, according to Spins, a San Francisco market research company.

But is soy really as beneficial as people believe and as some ads say?

The most solid evidence of soy's benefit comes from studies of cholesterol. In October, the US Food and Drug Administration was convinced enough of its benefit to begin letting manufacturers put health claims on soy products indicating that soy may lower heart disease risk.

Some studies suggest soy can fight hot flashes and may lower the risk of breast and prostate cancer. But paradoxically, the very ingredients that make soy beneficial may endow it with risks.

Soybeans are legumes that are rich in plant estrogens, specifically, genistein and daidzein, which are substances known as isoflavones. Like the estrogens that humans make in their bodies or buy by prescription, phytoestrogens may drive cell proliferation - a red flag that means soy could theoretically spur cancer growth, particularly breast cancer, which is often driven by the hormone estrogen.

In general, because soy isoflavones are weak estrogens, they may be taken in by molecules known as estrogen receptors, possibly blocking the stronger estrogens made in the body.

In premenopausal women, this means that plant estrogens may protect against breast cancer by blocking a woman's own, stronger estrogens; in postmenopausal women, who have less estrogen than younger women, adding plant estrogens to the diet could yield an overall increase in estrogen levels, a possible concern for women who have or may develop breast cancer.

So how do the plusses and minuses of soy stack up in all areas of health? Much of it depends on individual risk factors for various diseases. Here's the latest data to go on:

Cholesterol. Overall, human and monkey studies suggest that soy reduces cholesterol 10 to 15 percent, a figure that reflects a grab-bag of studies on tofu, soy powder, extracts, and supplements. Specifically, soy lowers LDL or "bad" cholesterol, and countries where people eat a lot of soy tend to have lower heart disease rates, as well as lower cholesterol.

To get the beneficial effect of soy, a person has to eat at least four servings containing 6.25 milligrams of soy protein a day, as part of a diet low in saturated fat and cholesterol, the FDA noted.

In other words, soy provides "a definite, but modest reduction" in cholesterol, says Dr. Sherwood Gorbach, a professor of community health and medicine at Tufts University School of Medicine, who has also developed and patented a soy supplement called Healthy Woman.

Bone density. Here, the studies are mixed, with some showing a protective effect and others not. Some preliminary data suggest soy may restore bone loss, but this is not proved.

Hot flashes and other menopausal symptoms. Again, the data are mixed. Some data suggest soy reduces hot flashes by 40 to 50 percent, says Dr. Machelle Seibel, an endocrinologist at the Fertility Center of New England in Dedham and medical director of Inverness Medical, Inc., which makes SoyCare supplements.

But hot flashes often improve by 20 to 30 percent on placebo (or dummy drugs), too, he notes, adding that in some women, soy probably adds 20 to 30 percent to the benefit from any placebo.

On the other hand, Margo Woods, a Tufts nutritionist, has just completed a study of 85 women and found no difference between soy protein and placebo - both reduce hot flashes by about 25 percent, she says.

As for vaginal dryness, another common menopausal symptom, at least one study shows soy helps some women.

Prostate Cancer. Epidemiological studies from Asia suggest that men there are less likely than American men to get prostate cancer and die of it. One possible reason is that soy isoflavones may inhibit an enzyme that converts the hormone testosterone to its chemical cousin, dihydrotestosterone, which can spur prostate cell growth.

In a few case studies of men with prostate cancer, high doses of isoflavones seem to reduce the number of cancer cells, notes Seibel. So far, however, there's no solid evidence for using soy to treat prostate cancer.

In fact, despite encouraging data from animal studies that suggest soy isoflavones also act as angiogenesis blockers (which stop blood vessel growth around tumors), so far, no company appears to have asked the FDA to approve a health claim for soy on the grounds that it may fight prostate - or any other - cancer.

Breast Cancer. This is the diciest area of all because some research suggests that soy prevents breast cancer, while other indicates that, at least theoretically, it could increase it.

On the one hand, strong epidemiological evidence from Singapore and elsewhere shows that Asian women have a three- to five-fold lower risk of breast cancer than American women, though whether this is due to eating soy is unclear. At the very least, it would seem to suggest that high, lifelong soy consumption lowers the risk of breast cancer risk.

In fact, Japanese women with breast cancer typically continue to eat soy as part of their diet - and their survival rates are better than those of American women with breast cancer.

Furthermore, research suggests that in postmenopausal women, the levels of estradiol, a natural estrogen made in the body, decrease while women take soy, suggesting that soy may be protective against breast cancer, says Woods of Tufts.

The mere suggestion that soy might increase breast cancer risk rankles some soy researchers, including Gorbach of Tufts. "That's a ridiculous contention," he says. "Everywhere in the world where soy is consumed, the amount of breast cancer is remarkably decreased."

Still, in the interests of full disclosure, here's another side of the story.

When human breast cancers are transplanted into mice who are then given varying doses of isoflavones, strange things happen, says Tufts biochemist Barry Goldin. At high doses, the plant estrogens inhibit the growth of human breast cancer cells, while at lower doses, they may enhance it.

At the University of California in San Francisco, Dr. Nicholas Petrakis, an emeritus professor of preventive medicine and epidemiology, has also studied American women given soy and found some had an increase in the number of hyperplastic, or potentially precancerous, breast cells as well. This does not prove that soy raises the risk of breast cancer, he cautions, but it does suggest that soy probably has a "stimulatory estrogenic effect."

Also troubling is a 1998 study of nearly 50 premenopausal women published in the American Journal of Clinical Nutrition. The women were randomly assigned to continue their normal diets or to add 60 grams a day of soy supplements (containing 45 milligrams of isoflavones) for 14 days. Those who added soy had a greater increase in proliferation of breast cells.

Bobbie Hayes , a nurse in the menopause consultation service of Harvard Vanguard Medical Associates, puts it this way: "We don't feel [soy] is completely benign." It may be that soy is protective in Asian women who consume it all their lives, she says, but not for American women who start taking large doses at menopause to combat hot flashes.

At the National Cancer Institute, Dr. Peter Greenwald, director of the division of cancer prevention, adds that women who take soy are "taking an estrogenic compound. We know estrogens promote breast cancer and cell proliferation. So there is a theoretical risk."

Unfortunately, he adds, so far "we have no evidence" from clinical trials on either "the benefits or the harm of soy. If women are taking it to combat hot flashes, I would not see a problem with short-term use - a year or six months.. . .I think the longer you go, the less sure we are without studies."

Bill Helferich, a professor of nutrition at the University of Illinois, shares that concern. If given early in life, he says, isoflavones can cause breast tissue to differentiate, just as estrogen does - a good thing because the more differentiated breast cells are, the less likely they are to become cancerous.

On the other hand, if taken later in life, when tiny tumors may already be starting to grow, plant estrogens may cause those tumors to grow faster, he says.

Of particular concern, Helferich says, is women who take the prescription drug tamoxifen to prevent breast cancer but who want to switch to soy. That makes no sense, he says, because tamoxifen is a proven way to reduce cancer risk and soy is not.

So what's the bottom line in terms of breast cancer? Given that nobody has proved that soy reduces or raises breast cancer risk, much less that the supplements have the same risk-benefit profile as soy food, the prudent course might be for women to make decisions about soy just as carefully as they would about prescription estrogen. Among other things, that might mean talking to your doctor about taking soy for a limited period of time to combat hot flashes, then stopping.

And what about food versus supplements? Alas, once again, there are no good data, but Woods, the Tufts nutritionist, says that "people should increase their consumption of soy as a food," she says. "It's better than hamburger or fried chicken."

Most of the data showing a benefit to soy come from studies of food, not pills. "We have no data on what happens when you take phytoestrogens as supplements. . . I think taking supplements is too big a leap from the data on food."

If you want to add soy to your diet, try adding about three ounces of tofu or its equivalent a day, which will give you 45 milligrams of soy phytoestrogens, the amount that's believed to be beneficial to the heart.

If you prefer supplements instead, you should aim for about 45 to 55 mg a day. And it may be best to break this into two doses, one in the morning, one at night.

Judy Foreman is a member of the Globe staff. Her e-mail address is: foreman@globe.com.

Previous "Health Sense" columns are available through the Globe Online searchable archives at http://www.boston.com. Use the keyword columnists and then click on Judy Foreman's name.

All content herein is © Globe Newspaper Company and may not be republished without permission. If you have questions or comments about the archives, please contact us at any time.