Keeping abreast: the latest on diet and breast cancer.

An estimated 211,000 American women will be diagnosed with breast cancer this year. disease will kill 40,000. The only cancer that will kill more women (73,000) is lung.

Yet high breast cancer rates aren't universal, North America leads the pack, followed closely by Europe and other industrialized regions. Lower rates are found almost exclusively in developing countries. The exception: Japan.

Japanese women are only one-third as likely as American women to get breast cancer. genes don't explain the difference, because the risk rises when Japanese women emigrate to the U.S., and rates in Japan have been rising (they used to be one-fifth of ours). What is it about the U.S. diet, environment, or lifestyle that accounts for the discrepancy? That question has fueled much of the research on preventing breast cancer for the last 30 years. what researchers have found so far.

"Study of breast cancer patients finds benefit in low-fat diets," ran the headline in The New York Times in May. "Moderate walks help breast cancer survival," proclaimed USA Today.

Women have heard some good news about breast cancer lately. And it's not just for those who already have the disease. The risk of being diagnosed with breast cancer has declined slightly since 2001, after inching up from 1970 until about 2000. But over the long haul, the news has been either absent or discouraging.

"Breast cancer is always complicated," says Walter Willett, chair of the Nutrition Department at the Harvard School of Public Health. "Just when you think you've got it understood, something else emerges." Here's the latest on what you can do to lower your risk of the cancer that strikes one out of eight American women.

Fatty Foods

Could fatty foods promote breast cancer? Researchers seized on fat as a culprit in the 1970s, in trying to explain why Japanese women were much less likely to get breast cancer than American women. The Japanese were getting only 10 percent of their calories from fat, far less than the typical 40 percent that U.S. women were eating.

But the link between breast cancer and fatty foods didn't hold up as researchers gathered more precise data, especially when they pooled data in 2001 from studies that tracked more than 350,000 women. (1)

With blame shifting away from fatty foods, "we thought that Asian women had a lower risk of breast cancer because they consumed fewer calories, had a lower body weight, or were more physically active than American women," says cancer researcher Regina Ziegler of the National Cancer Institute in Bethesda, Maryland.

But the spotlight turned back to fat last May, when researchers reported that a low-fat diet (20 percent of calories) cut the risk of recurring breast cancers in a study of more than 2,400 postmenopausal women who had already undergone the usual medical treatment for early-stage breast cancer.

The low-fat eaters averaged only 33 grams of fat a day, considerably less than the 51 grams a day eaten by similar breast cancer patients who were instructed to follow their usual diet.

"The low-fat group decreased their portion sizes of dairy, meat, and baked goods like sweet rolls and drastically reduced their oils and spreads," says lead investigator Rowan Chlebowski of the Los Angeles Biomedical Research Institute in Torrance, California.

"This is the first randomized clinical trial showing that diet may have an impact on breast cancer outcome."

Even more surprising: the risk of recurring breast cancer was reduced--by 42 percent--only in low-fat dieters whose original tumors weren't sensitive to estrogen. Those "estrogen-negative" cancers--which account for about a quarter of all breast cancers in the U.S.--are hard to treat because they don't respond to estrogen-blocking drugs like tamoxifen.

But experts are cautious about the results.

"The press made it sound like a clear answer, but the study has to be regarded as preliminary," says Harvard's Willett. "It hasn't been published, and they have another eight months of follow-up that weren't included in the analysis," which was presented at a scientific meeting in May.

What's more, the results were statistically significant, but just barely--that is, they were almost chalked up to chance, largely because the number of estrogen-negative cancers was small.

"I don't want to oversell results with a borderline significance," says Chlebowski. "We need to wait for more follow-up and other studies. We can't tell women that they should change what they eat or their cancer is more likely to recur."

On the other hand, it's possible that earlier studies failed to find a link between fat and breast cancer because they didn't distinguish between estrogen-negative and estrogen-positive tumors. "Until recently, we didn't have data on estrogen-receptor status," says Willett. "It may turn out to be an important distinction."

Willett, his Harvard colleague David Hunter, and others recently found similar results in one of their studies. Women who reported eating a "prudent" diet pattern had a lower risk of breast cancer than those who are a "Western" pattern. Here again, the diet link appeared only in women who were diagnosed with estrogen-negative cancers. (2)

The prudent pattern (more fruits, vegetables, whole grains, low-fat dairy foods, fish, and poultry) is closer to the low-fat diet in Chlebowski's study than the Western pattern (more red and processed meats, refined grains, sweets, and high-fat dairy).

"It's possible that estrogen-positive tumors are so dominated by the effect of estrogen that it may mask other factors," suggests Willett. So diet may only affect estrogen-negative tumors.

And it may not be less fat, but more fruits and vegetables, that matters. "Almost all the benefit of the prudent pattern was coming from fruits and vegetables," Willett notes.

That seems to contradict the results of a study that pooled data on more than 350,000 women, which showed that fruits and vegetables had no impact on breast cancer risk. (3) But, like earlier studies on fat, it didn't separate estrogen-negative and estrogen-positive cancers.

Experts are eagerly awaiting results from the Women's Health Initiative to see if diet can reduce the risk of breast cancer and other diseases. "We have 48,000 otherwise-healthy postmenopausal women randomized to a typical diet or a low-fat diet with more fruits and vegetables," says Chlebowski, one of the trial's investigators. The results are due out in the first quarter of 2006.

Until researchers know more, says Hunter, "women who want to cut fat should do it by replacing much of their fat with fruits and vegetables, not by eating low-fat processed foods." Loading up on cakes, cookies, ice cream, white bread, and

other foods that are often labeled "low-fat" would be a mistake.

"We have good evidence that eating more sugar and refined flour will increase the risk of obesity, diabetes, heart disease, and quite probably cancer," says Hunter. "A low-fat diet could do harm if you replace fat with those carbohydrates."

Weight

For years, studies could not nail down a clear link between weight and breast cancer. Then, in the 1990s, scientists figured out why the link was so fuzzy. Among other things, it shows up:

* With modest weight gain. It's possible that weight seemed to have little impact on breast cancer in American women because the vast majority are considerably heavier than Japanese women.

But Ziegler and colleagues saw the impact of weight when they looked at Asian Americans (who weigh less than most other Americans). Asian American women who were heavier-than-average and had gained more than 10 pounds in the previous decade had three times the breast cancer risk of leaner Asian Americans. (4)

"We may have to get much slimmer to reach the low risk of Asians," she suggests.

* Only in postmenopausal women. Being overweight doesn't raise the risk of breast cancer in premenopausal women, so any studies that mixed them in with postmenopausal women might not have detected a link.

* Only in women who don't take hormones. For years, estrogen pills obscured the link between breast cancer and weight. "We found twice the risk of breast cancer in women who gained the most weight after age 20 compared to those who kept their weight constant," says Willett, "but only when we looked at women who had never used hormone replacement therapy."

Experts believe that extra fat cells raise the risk of breast cancer because they produce hormones that the body converts to estrogen.

"Estrogen levels in the blood are three times higher in overweight women than in lean women," says Willett. "That's a huge difference."

But extra estrogen from fat cells gets swamped by the estrogen from hormone replacement pills.

Recent studies suggest that it's not just estrogen, but also androgens, that raise the risk of breast cancer in overweight women. (5,6) "We think that higher circulating estrogens explains why postmenopausal women who are overweight have a higher breast cancer risk," says Ziegler, "but it's possible that other hormones are also important."

That's not to say that being overweight can make or break your chances of getting breast cancer. "Plenty of lean women get breast cancer," notes Harvard's Hunter. So, by staying trim, "you're reducing your risk, not abolishing it."

Nevertheless, the link between higher weight and the risk of postmenopausal breast cancer is "fairly strong and highly reproducible," says Willett. "It's a very important public health message."

And the risk rises long before you're shopping for plus-size clothes. "Gaining two or three pounds a year after age 18 or 20 adds up to quite a bit of excess risk by menopause," he explains. "Even a quarter of a pound a year--five to eight pounds by menopause--is a significant excess risk."

In fact, it's unusual for American women to weigh the same as they did at age 20. In a study that tracked women through the 1980s and early 1990s, "only 5 percent didn't take hormones or gain weight during adulthood," says Willett. "But in Japan, it was almost 100 percent of women, at least until recently."

On his visits to Japan, Willett has seen a different attitude toward eating than ours. "They don't have a culture of overindulging," he explains, especially among women.

As the Japanese diet becomes more Westernized, "the men have gained weight, but the women have remained remarkably lean," says Willett. "To a Japanese woman, one of the worst things that can happen is to become obese."

Exercise

Does a walk a day keep breast cancer away? For years, researchers weren't sure, because exercise seemed to protect women in some studies but not others.

"In the Nurses Health Study, we had to use a 16-year follow-up of more than 85,000 women, more than 3,000 cancer cases, and repeated measures of physical activity to see a relationship with breast cancer," says Willett. "Very few studies are that powerful."

The nurses study found an 18 percent lower risk in women who reported moderate or vigorous physical activity for at least seven hours a week compared to women who exercised for less than an hour a week. (7)

"But unless you have a very large, very long study, you could miss the association," Willett explains.

That's partly because exercise doesn't drive down risk dramatically. "Physical activity has a weaker relationship with breast cancer than with diabetes, heart disease, and colon cancer," he says.

Last June, researchers reported that exercise seems to improve survival in women who already have breast cancer. (8)

"We found the maximum benefit in women who walked the equivalent of three to five hours a week," says researcher Michelle Holmes of Harvard Medical School. They walked at an average pace--two to three miles an hour--with no benefit in walking faster or longer.

"The effect was most prominent in women whose tumors were hormone-positive," she adds. That comes as no surprise to Holmes.

"Women who are physically active have lower hormone levels, and hormones stimulate breast cancer growth," she notes. Other studies suggest that an exercise program--45 minutes a day for five days a week--can lower estrogen levels in sedentary, overweight postmenopausal women. (9)

"Physical activity may also improve the quality of life and self-image of women who've had breast cancer," adds Holmes. And it can help women avoid weight gain.

"Patients who gain weight have a poorer survival," she notes. Of course, physical activity can also help women fend off the extra padding that boosts the risk of getting breast cancer in the first place.

Staying active can prevent so many problems that it's tough to argue against it. "You don't have a lot to lose," says Holmes, "except maybe a higher risk of diabetes and heart disease."

Alcohol

"The evidence is solid that alcohol is a small, modest risk factor," says Willett. "It's been seen in 100 studies by now."

What's small? "One drink a day is associated with an 8 to 10 percent increase in risk, and two drinks a day are associated with a 25 percent increase," he says. "That's not huge, but given that the risk for non-drinkers is high, the increase is appreciable."

It's not clear how alcohol influences the breast. Some studies suggest that it may raise estrogen levels. Willett argues that the B-vitamin folate plays a role.

"Alcohol increases the requirement for folic acid, so when you have marginal folate intake and add alcohol, you really have inadequate folate," he says.

If he's right, his studies suggest that you can dodge that excess risk just by taking a multivitamin, which usually has 400 micrograms of folic acid. "It's one of the few easy fixes," says Willett.

Soy

Asian women eat more soy than their U.S. counterparts. Could that explain the lower breast cancer rates in Japanese women?

"We found a protective effect of soy in adults, but we couldn't rule out other aspects of an Asian lifestyle, like more physical activity, lower weight, and lower fat intake," says the National Cancer Institute's Regina Ziegler.

But overall, results are inconsistent. "If anything, the evidence for adults has weakened," she adds. "The protective effect could be stronger in children."

In theory, the plant estrogens in soy foods are ant/-estrogens--that is, they may block estrogen from reaching cells' estrogen receptors, like a natural version of the drug tamoxifen. But if women aren't making much estrogen, soy could become a pro-estrogen.

"When women are premenopausal and their natural estrogen levels are high, it may be helpful for soy to compete with estrogen by blocking estrogen at the receptor," says Willett.

But if natural estrogen levels are low, he adds, soy may simply add estrogen to the system, which increases risk.

"I do get worried about postmenopausal women who load up on soy and assume that more is better," cautions Willett. "I don't think we have an adequate basis to recommend or not recommend soy for breast cancer, and we don't have evidence on its long-term safety."

The safest approach: don't go overboard. "Soy is a good overall part of the mix as a replacement for red meat," says Willett. "But it may be possible to overdo it."

Estrogen

As researchers search for risk factors for breast cancer, they find a recurring theme. "One thing that's consistent is the association with estrogen," says Hunter.

For example, risk is lower in women who have at least two pregnancies, breastfeed for more than one year, or get their first menstrual period--menarche--late (at age 15 or older). (10) What ties those risk factors together? Fewer menstrual cycles and less estrogen.

"It's natural menstrual cycles--the cycling of estrogen in combination with progestin--that really kicks up breast cancer risk because it makes cells multiply," says Willett.

Breast cells are multiplying to get ready for breastfeeding should the woman become pregnant. But until recently, those cycles were few and far between.

"Women hardly ever had a menstrual cycle, because they were pregnant or lactating throughout their premenopausal years, and then menopause came along," explains Willett. "But we've created a non-natural state in human history, where women are ovulating and cycling every month for 30 years."

Menstrual periods don't explain all risk factors. For example, having a first baby at an early age (20 or younger) also lowers the risk of breast cancer, probably because the pregnancy makes breast cells differentiate, or specialize, earlier. Once they've differentiated, estrogen may cause less damage.

Of course, American teenagers aren't about to start having more babies to lower their breast cancer risk.

"It looks like breast cancer comes with the territory of an abundant food supply, early menarche, and late age at first birth," says Hunter. "Once you get beyond the reproductive factors, we haven't found ways to influence it, apart from avoiding alcohol, weight gain, and postmenopausal hormones."

But the picture may not be as bleak as it sounds. Researchers are trying to figure out exactly how hormones or growth factors promote cancer.

"If we can understand why they're so important at the cellular level," explains Ziegler, "we might find out how to lower the risk without asking women to bear children early or to choose early menopause."

(1) International Journal of Cancer 92 767, 2001.

(2) International Journal of Cancer 116:116, 2005.

(3) J. American Medical Assoc. 285: 769, 2001.

(4) J. National Cancer Institute 88: 650, 1996.

(5) J. National Cancer Institute 95:1218, 2003.

(6) J. National Cancer Institute 97: 755, 2005.

(7) Archives of Internal Medicine 159: 2290, 1999.

(8) J. American Medical Assoc. 293: 2479, 2005.

(9)Cancer Research 64: 2923, 2004.

(10) Cancer Causes & Control 11: 477, 2000.

Beyond Diet & Exercise

Here are some of the factors that raise your risk of breast cancer. To estimate your risk of cancer (or diabetes, heart disease, or stroke), go to www.yourdiseaserisk.harvard.edu.

* Age. The older you are, the higher your risk. The average age at which women are diagnosed is 62, but the highest rates are in women aged 70+.

* Family history. Risk is twice as high in women who have one first-degree relative (mother, sister, or daughter) with breast cancer than in women with none. Women who have two first-degree relatives with breast cancer have five times the risk. Having any relative with breast or ovarian cancer also raises the risk (especially if the cancers were diagnosed before age 50).

* Genes. An estimated 5 to 10 percent of breast cancer cases are a result of gene mutations. A woman who inherits the most common mutations--in the BRCA-1 and BRCA-2 genes--has up to an 80 percent chance of getting breast cancer in her lifetime.

* Jewish ethnicity. Jewish women--especially those of Ashkenazi (European) descent--have a higher risk of breast cancer, because they are more likely to have gene mutations linked to the disease.

* Menstrual periods. Women who started menstruating early (before age 12) or who went through menopause late (after age 55) have a higher risk.

* Births. Women who give birth to fewer than two children or who have their first child after age 30 have a higher risk. An earlier pregnancy may protect the breast by making breast cells specialize.

* Breastfeeding. Women who breastfeed children for at least a year (all babies combined) have a lower risk of breast cancer, at feast in some studies. Breastfeeding may help by cutting the number of menstrual periods.

* Estrogen pills. Women who take estrogen even for a year or two after menopause have a higher risk, but the risk seems to return to normal within five years of stopping.

* Benign breast disease. Hyperplasia raises the risk of breast cancer, possibly because extra cells in the breast can become abnormal.

* Height. Taller women have a higher risk of breast cancer, but researchers don't know why.

What doesn't raise your risk. Little or no evidence supports the claims that antiperspirants, underwire bras, abortion, or breast implants raise the risk of breast cancer. A few studies have found that antibiotics and working at night raise risk, but more research is needed to confirm those findings.

Sources: American Cancer Society and Cancer Causes & Control 11: 477, 2000.

Still Murky

It's too early to say whether these factors alter the risk of breast cancer:

* Carbohydrates. Insulin resistance promotes breast cancer, suggest some researchers. Because insulin resistance is hard to measure, scientists look at blood sugar instead. But so far, studies have found no consistent link between breast cancer risk and glycemic index (how often people eat carbs that raise blood sugar). (1,2)

"Insulin resistance could still be associated with breast cancer risk," says Harvard's Walter Willett. "But the evidence isn't very clear."

In part that's because "being overweight elevates insulin and estrogen levels in postmenopausal women," he says. "So unless you carefully factor out the impact of estrogen, you don't know if insulin is related to breast cancer risk."

* Flavonoids. Women who ate more flavonols such as kaempferol, quercetin, and myricetin--found in foods like tea, onions, apples, string beans, broccoli, green peppers, and blueberries--had no lower risk of breast cancer. (3)

* Insulin-like Growth Factors (IGF). Insulin-like growth factors--not to be confused with insulin--are hormones that may keep abnormal cells alive.

"The IGF theory relates to virtually every cancer," explains Willett. "But the evidence for colon and prostate cancers is a lot more consistent than it is for breast cancer."

In fact, most recent breast cancer studies have come up empty. (4) "The evidence is not looking as good as it did now that several larger studies have found no relationship between IGF and breast cancer," says Harvard's David Hunter.

It's possible that growth factors do matter in childhood and adolescence, though. Taller women have a higher risk of breast cancer, he points out, and the major determinants of height are the growth factor IGF-1 and genes.

"Data from Denmark suggest that rapid gain in height between 10 and 14 years of age is linked to a higher risk of breast cancer," Willett notes. (5)

But even if researchers were sure that less or later growth protects against breast cancer, there's not much they could recommend.

"Height may be important for breast cancer and may explain some of the international variation," says Hunter. "But no one's going to recommend that we stunt our daughters to lower their risk."

* Meat & Dairy. In a study of more than 350,000 women, Hunter and colleagues found no link between breast cancer and dairy or meat consumption in postmenopausal women. (6) However, in premenopausal women, eating high-fat meat and dairy is linked to a higher risk.

* Pesticides. So far, researchers haven't found good evidence that pesticides raise breast cancer risk. (7) A study Hunter and colleagues published in 2001 found no link with total PCBs and DDT residues. But, he acknowledges, "that doesn't absolve thousand of other chemicals for which the data are not adequate."

(1) International Journal of Cancer 114: 653, 2005.

(2) American Journal of Epidemiology 159. 732, 2004.

(3) International Journal of Cancer 114: 628, 2005.

(4) Cancer Epidemiol. Biomarkers Prev. 14: 699, 2005.

(5) New England Journal of Medicine 351: 1619, 2004.

(6) International Journal of Epidemiology 31: 78, 2002.

(7) International Journal of Cancer 91: 568, 2001.

The Bottom Line

* Screening. A doctor should examine your breasts every 1 to 3 years between ages 20 and 39 and every year starting at age 40. Women aged 40 or older should get a mammogram every year. All women should do a monthly breast self-exam.

* Weight. Avoid weight gain, especially after menopause.

* Alcohol, If you consume alcohol, don't exceed one drink per day.

* Exercise. Shoot for at least an hour a day of exercise like brisk walking.

* Fatty foods. It's too early to say whether a lower-fat diet, fruits, or vegetables can lower your risk. However, cutting back on saturated and trans fats lowers the risk of heart disease, and eating more fruits and vegetables may lower the risk of heart disease, stroke, and other cancers.

* Soy. Researchers don't know whether soy can lower the risk of breast cancer.

Breast Cancer Worldwide
(2002 estimates)

                                        NEW CASES     DEATHS

UNITED STATES                               101.1       19.0
FRANCE                                       91.9       21.5
UNITED KINGDOM                               87.2       24.3
CANADA                                       84.3       21.1
AUSTRALIA                                    83.2       18.4
GERMANY                                      79.8       21.6
ITALY                                        74.4       18.9
ARGENTINA                                    73.9       21.8
AVERAGE OF MORE-DEVELOPED COUNTRIES          67.8       18.1
GREECE                                       51.6       15.4
SPAIN                                        50.9       15.9
BRAZIL                                       46.0       14.1
RUSSIAN FEDERATION                           38.8       18.0
JAPAN                                        23.7        8.3
NIGERIA                                      31.2       21.9
MEXICO                                       26.4       10.5
AVERAGE OF LESS-DEVELOPED COUNTRIES          23.8       10.3
TANZANIA                                     21.1       15.5
INDIA                                        19.1       10.4
CHINA                                        18.7        5.5

per 100,000 people

Japanese women are diagnosed with--and die of--breast cancer at
rates that are closer to those of less-developed countries than
to other industrialized countries. When Japanese women move to
the U.S., their rates start to rise.

Source: International Agency for Research of Cancer, GLOBOCAN 2002
(www-dep.iarc.fr/).

Note: Table made from bar graph.