Soy protein and soy isoflavone supplements are being heavily promoted by the powerful soy industry as 'miracle cures' for cancer. With cancer being the second leading cause of death in the US and in Europe, the idea that a simple food could save lives sounds like very good news indeed.
Unfortunately, the truth is another story. While a few studies do suggest that soy protein - or its isoflavones - might help prevent cancer, far more studies show it to be ineffective or inconsistent. Some studies even show that soy can contribute to, promote or even cause cancer.
Numerous experts - including scientists from the FDA's own National Laboratory for Toxicological Research - have warned of soy protein's cancer-causing potential and the health dangers that ensue from excess soy consumption.
Indeed, soy isoflavones - the plant oestrogens in soy most often credited with cancer prevention - are listed as 'carcinogens' in many toxicology textbooks, including the American Chemical Society's 1976 Chemical Carcinogens. Over the years, they have been proven to be mutagenic, clastogenic and teratogenic - they cause chromosomal damage and genetic abnormalities, the precursors to cancer (Mutat Res, 2003; 542: 43-8; J Chromatogr B Analyt Technol Biomed Life Sci, 2002; 777: 211-8; Int J Cancer, 2003; 105: 312-20). In addition, the modem industrial soy-processing techniques used to make soy-protein isolate (SPI), textured vegetable protein and other modern soy products create toxic and cancer-causing residues.
Nevertheless, the soy industry argues that Asians eat a lot of soy and have less cancer.
The death rates for breast and prostate cancers are four and 18 times lower, respectively, in China than in the US. Soy proponents often cite these and other similarly favourable cancer statistics as reasons to eat a lot of soy protein.
But if decreased rates of breast, prostate and colon cancer in Asia are to be attributed to soy consumption, then the same logic requires that soy take the blame for the higher rates of cancers of the oesophagus, stomach, thyroid, pancreas and liver found in those countries.
While soybeans in the diet might be a factor in either reduced or increased incidences of cancer in Asia, there is no direct evidence of cause and effect. Several dietary and lifestyle factors are undoubtedly involved. Epidemiological studies of prostate cancer, for example, have not only associated a reduced cancer incidence with soy, but also with rice, vegetables, fruits, nuts, grains, green tea, fish, other legumes and/or combinations of foods. None of the research supports crediting soy alone (Cancer Sci, 2004; 95: 238-42).
Oestrogen and anti-oestrogens
In the many laboratory studies, soy seems to prevent cancer, cause it or have no effect, depending on the study. Like most hormonally active agents, soy isoflavones can either stimulate or inhibit cell growth. Although soy proponents often laud the ability of soy oestrogens to act against human oestrogens, soy oestrogens are just as likely to act with them. The former might decrease cancer, while the latter increases the risk.
This means that increasing soy consumption to prevent cancer is unreliable, unpredictable and risky. Kenneth D.R. Setchell, PhD, warned of the potential harm decades ago when he wrote: "Oestrogens exert dose-dependent dual effects upon tumour induction and growth.
High doses inhibit tumour development and suppress growth while physiological doses [doses usually found in food] stimulate growth of human tumour cells" (Am J Clin Nutr, 1984; 40: 569-78).
This warning is supported by a number of recent studies. In 1997, researchers from the University of Minnesota at St Paul found that phytoestrogens inhibited breast-cancer cells at high concentrations, but stimulated growth at low concentrations (Nutr Cancer, 1997; 28: 236-47).
In 2001, the British Columbia Cancer Agency in Vancouver reported that the soy isoflavones genistein and daidzein at high concentrations inhibited tumour growth and enhanced the effect of tamoxifen in vitro (in the lab). However, at dietary levels, soy stimulated existing breast-tumour growth and antagonised the effect of tamoxifen both in vivo (in people) and in vitro. Accordingly, the Agency warned that "women with current or past cancer should be aware of the risks of potential tumour growth when taking soy products" (Ann Pharmacother, 2001; 35: 1118-21).
Likewise, Craig Dees, PhD, of the Oak Ridge National Laboratory in Tennessee, came to the conclusion that "dietary oestrogens at low concentrations do not act as anti-oestrogens, but act like DDT and oestradiol to stimulate breast cancer cells to enter the cell cycle" (Environ Health Perspect, 1997; 105 [Suppl 3]: 633-6) [our italics].
Women eating soy to avoid breast cancer can develop the very disease they are trying to prevent. Although some in vitro studies show that isoflavones can inhibit the proliferation of breast-cancer cells (Res Commun Chem Pathol Pharmacol, 1989; 64: 69-77), there are plenty of contradictory findings showing that soy can cause breast-cancer cells to proliferate (Endocrinology, 1978; 103: 1860-7).
Warning bells also sounded with a series of rodent studies run by William Helferich, MD, at the University of Illinois at Urbana-Champaign (J Nutr, 2001; 121: 2957-62). The team tested SPI-based feeds containing increasingly higher concentrations of genistein, and discovered that the more isoflavones the mice ate, the higher the incidence of breast-cell proliferation and cancer growth. The researchers also found that the tumours regressed when the mice were switched back to an isoflavone-free diet.
Dietary genistein notably stimulated the growth of mammary tumours in a low-oestrogen environment similar to that found in menopausal women.
In his most recent studies, Dr Helferich announced that soy products that contain isoflavones in purified forms provoked far greater tumour growth than minimally processed products such as soy flour.
What's the take-home message? Dr Helferich is clear: "Our preclinical laboratory animal data suggest that caution is warranted regarding the use of soy supplements high in isoflavones for women with breast cancer, particularly if they are menopausal." But animal data do not necessarily apply to humans.
However, in people, results have been equally sobering. Phytoestrogens in the diet can have oestrogenic effects on breast tissue (J Clin Endocrinol Metab, 1999; 84: 4017-24) and cause proliferation of breast cells (Am J Clin Nutr, 1998; 68 [6 Suppl]: 1431S-5S). Researchers at the University of California at San Francisco, who hoped to find that soy foods could protect against premenopausal breast cancer, instead found that six months of consuming soy protein containing genistein and daidzein had a "stimulatory effect on the premenopausal female breast, characterised by increased secretion of breast fluid, the appearance of hyperplastic epithelial cells and elevated levels of plasma oestradiol" (Cancer Epidemiol Biomarkers Prev, 1996; 5: 785-94). Breast-cancer risk is higher for women with abnormal cells in their breast fluid (J Natl Cancer Inst, 2001; 93: 1791-8).
Barry Goldin, PhD, of Tufts University in Boston, MA, warns that a high soy consumption might increase the oestrogenic activity in cells by as much as 25-30 per cent in postmenopausal women, who have low levels of natural oestrogen. Goldin believes that premenopausal women are less at risk because their own oestrogens 'overpower' the plant oestrogens in soy, but he warns that soy might pose "an added potential risk" for women who either have or are at risk of breast cancer.
Epidemiological studies show lower rates of prostate cancer in Asia, but provide little or no proof that soy should take the credit. A Japanese study of 122,261 men over 40 revealed that green and yellow vegetables appeared to be protective, and that soy miso - including the cheap, fast-fermented types that became popular after World War II - significantly increased the risk (Natl Cancer lnst Monogr, 1979; 53: 149-55). Over the years, researchers have concluded that anything and everything - from green tea, rice, nuts and fish to monogamy, fidelity and poverty - could be responsible for lower rates of prostate cancer.
Both SPI and conjugated linoleic acid (CLA), found in butter, have been associated with a reduced risk of prostate cancer, but researchers found that neither, used alone or in combination, inhibited prostate-cancer growth. Moreover, at the highest concentrations of SPI, there was a significant increase in tumour size in the test tube (Prostate, 2003; 54: 169-80).
Also, soy has either failed to improve PSA (prostate-specific antigen) test results or actually increased PSA levels in middle-aged and elderly men (J Urol, 2003; 169: 507-11). PSA is a marker of prostate tumour growth.
However, those few human studies suggesting that soy can reduce rates of prostate cancer show that it does so only in those who produce equol (Jpn J Clin Oncol, 2004; 34: 86-9; Jpn Clin Oncol, 2002; 32: 296-300). Equol is a metabolite of daidzein that some people produce in the gut, and drinking green tea may help to improve its production (Asian Pac J Cancer Prev, 2003; 4: 297-301).
Soy proponents rarely tell men that the reason why soy might protect them against prostate cancer is because it has a feminising effect. Levels of soy that might be useful to prevent or treat prostate cancer significantly decrease testosterone and androgen (high levels of which may promote prostate cancer) while increasing oestrogen (In Vivo, 2000; 14: 389-92).
The soy industry tells us that soy has a long track record in the prevention and treatment of gastrointestinal cancers. Yet, the most notable study to come along in recent years reveals that soy protein is associated with a lowered risk of stomach cancer, but a higher risk of death from colorectal cancer (Int J Epidemiol, 2000; 29: 832-6).
Certainly, soybeans have the potential to induce epithelial-cell damage and proliferation, markers of colon cancer risk (Carcinogenesis, 1999; 20: 927-31). However, a few in vitro studies suggest that high levels of genistein can inhibit the growth of human colon-cancer cells (Int J Oncol, 2001; 18: 997-1002).
Once again, these findings don't mean we should all eat more soybeans, but rather indicate that isoflavones might have promise in the short term. Indeed, in one case study, a 66-year-old man who took 160 mg/day of phytoestrogens for one week before a radical prostatectomy showed benefit from cancer-cell death and tumour regression (Med J Austr, 1997; 167: 138-40). This short-term course of genistein drug therapy offered soy benefits without much risk.
Recently, findings from the Singapore Chinese Health Study sent shock waves through the industry: soy was associated with a two- to threefold increase in bladder cancer risk. A follow-up study strengthened the connection. No other food in the diet was associated with the increase - just soy (Cancer Epidemiol Biomarkers Prev, 2002; 11: 1674-7; Int J Cancer, 2004; 112: 319-23). As for the rising rate of pancreatic cancer, scientists have known for half a century that trypsin inhibitors in soy protein put stress on the pancreas, contributing to and possibly causing pancreatic cancer (see box, page 8). Recently, four alarming studies linked soy oestrogens to infant leukaemia (Proc Natl Acad Sci USA, 2000; 97: 4790-5).
Another problem with soy is that we're bombarded with oestrogenic mimics in our environment. Increased levels of soy oestrogens in the diet have a cumulative or exponential effect when combined with other environmental oestrogens. As toxicologists at the Centre for Toxicology at the University of London note: "Oestrogenic agents are able to act together to produce significant effects when combined at concentrations below levels that would be toxic separately."
In the first half of the 20th century, John Harvey Kellogg, Henry Ford and others envisioned a grand future for soy. By the 1960s, vegetarians, hippies, environmentalists and other idealists took up the banner, recommending soy foods as the solution to world hunger, the path to good health, the key to healthy ageing and the way to preserve our environment. Sadly, big business and big government have usurped their dream. Old-fashioned whole-soy foods that contribute to health if eaten in moderation have given way to ersatz products that lead inevitably to malnutrition and disease.
As Regina G. Ziegler, PhD, a nutritional epidemiologist at the National Cancer Institute in Bethesda, MD, said in summarising the soy and breast cancer research to date: "it's complicated, inconsistent and inconclusive". When patients ask whether they should eat more soy foods, she says, "I think we have to be cautious" (Am J Clin Nutr, 2004; 79: 183-4).
Israel has taken it one step further. The 13-member committee of nutritionists, oncologists, paediatricians and other specialists, who took a year to examine the evidence, are distributing information on the dangers of soy foods and infant formula to public and healthcare workers, suggesting that all the population, but particularly infants and children, limit their consumption. Their advice is clear: soy is a condiment, not a superfood.
Kaayla T. Daniel, PhD, CCN
Dr Daniel is the author of The Whole Soy Story: The Dark Side of America's Favorite Health Food (Washington, DC: New Trends Publishing, 2005)