Soybean peapod. Photos Courtesy of American Soybean Association.

Our medical minds like simple answers, a tendency sometimes at odds with the complexity of biological phenomena. Take the question of soy and women’s health, for example. There’s little doubt that overall, soy is a healthy food, and a great source of protein for women. But when one looks at specific tissues, and tries to determine whether soy isoflavones enhance or inhibit endogenous estrogenic effects, the simple answers quickly evaporate.

Two recent studies of the endometrial effects of soy isoflavones underscore the complexity of soy phytoestrogen biochemistry, leaving us with the need for far more information before we can posit definitive answers.

Investigators at the AGUNCO Obstetrics & Gynecology Centre, Rome, undertook a randomized, double-blind, placebo-controlled study to determine the endometrial effects of 5 years of treatment with soy phytoestrogens in postmenopausal women. Three hundred seventy-six women, all with an intact uterus, were randomized to 2 groups: one group (n=179) received soy tablets containing 150 mg of soy isoflavones per day; the second group (n=197) received identical placebo tablets. The researchers obtained endometrial biopsies at baseline, 30 months, and 5 years after treatment began.

A total of 298 participants completed the 5-year trial. Surprisingly, the incidence of endometrial hyperplasia, though low overall, was higher in the isoflavone group (3.3% vs 0%). This included 5 cases of simple hyperplasia and one of complex hyperplasia. Seventy percent of women on the 150 mg of soy isoflavones had atrophic tissue versus 81% receiving the placebo. No cases of atypical hyperplasia or endometrial cancer occurred during the 5 years (Unfer V, Casini M, Costabile L, et al. Fertil Steril 2004; 82:145–148).

Though there were no endometrial cancers in the soy-treated women, these findings are noteworthy because this is the first study to raise concern about the effects of long term, high dose isoflavone supplementation on the endometrium. Previous observational and clinical trials looking at this question showed no increased incidence of endometrial hyperplasia.

In fact, a recent study from China suggests that soy isoflavones have an anti-carcinogenic effect on the endometrium. This population-based, case-control study was based on detailed information from a food frequency questionnaire on soy food intake over five years. The participants were 832 women aged 30 to 69, who had had endometrial cancer from 1997 to 2001. Similar data were obtained from 846 matched control subjects selected from the Shanghai Residential Registry.

The average intake of isoflavones from soy food was 42.5 mg per day in both controls and cancer patients, but the data suggest that regular consumption of soy foods, as either soy protein or soy isoflavones, was inversely associated with endometrial cancer. The inverse association seemed to be greater among women with high body mass index and a higher waist-to-hip ratio (Xu W, Zheng W, Xiang Y, et al. Br Med J 2004; 328:1285–1288).

Epidemiologically, Asian countries have a low incidence of endometrial cancer and a high consumption of soy food. Endometrial cancer incidence in Asia is only one-fifth to one-third that of Western countries. We know that phytoestrogens have been shown to alter circulating concentrations of endogenous estrogen and they can also lower serum estradiol levels.

The Xu study is one of a number of studies from China showng an inverse association between phytoestrogen consumption and risk of endometrial cancer, though there is one earlier study suggesting soy phytoestrogens increase risk. What was striking about the current study is the finding of a more profound inverse association between soy food intake and risk of endometrial cancer (p<0.05), and further, that the effect was strongest in overweight women. This is consistent with the hypothesis that soy foods have an antagonistic “anti-estrogenic” effect on endometrial proliferation.

But it is difficult to draw a definitive conclusion because the Xu study included pre- and postmenopausal women, did not include a measurement of serum endogenous estrogen levels, and relied on patients recalling their habitual dietary intake of soy foods. In light of the Unfer study suggesting that soy isoflavones may have estrogenic, proliferative effects on endometrium, one cannot help but wonder if the net effect is dependent on isoflavone dose, menopausal status, baseline estrogen levels, and duration of isoflavone exposure.

Prior to the Unfer study, there were two shorter-term trials of isoflavones in non-Asian women. One study showed that 30 months of treatment with 150 mg of isoflavones did not result in an increased incidence of endometrial hyperplasia or neoplasia.

Soy phytoestrogens appear to have polyvalent effects on estrogen-sensitive tissues. They are able to act as a partial agonist, binding to the estrogen receptor. But because their action is weaker than that of endogenous estrogens, the final effect seems to be antagonistic and is able to counteract the proliferative effects of endogenous estrogens.

When isoflavone exposure is prolonged, and at higher doses, the agonist effects seem to be more evident, and the isoflavones exert a net estrogenic effect. It is important to note that in the Unfer study, at 30 months, there was no difference between the isoflavone group and the placebo group. It was only after 5 years that the isoflavone supplements had an estrogenic effect in a small number of women.

If soy isoflavones do, in fact, have a pro-estrogenic proliferative effect on endometrium, it is probably fairly small, and not likely to be cause for clinical concern unless a woman is already at increased risk for endometrial hyperplasia or endometrial cancer.

Thyroid Function and Soy

	Soy leaf.

Concerns about the possible goitrogenic effects of soybean isoflavones abound. Unfortunately, much of the concern is emotional, and not based on a full understanding of the facts. These concerns are based primarily on in vitro research, animal studies, and older reports of goiter occurring in infants who were fed soy formulas that did not contain added iodine.

The best available human data on this subject comes from a 2003 randomized, double-blind, placebo-controlled study, in which researchers investigated the effect of a daily soy supplement on thyroid function in 38 postmenopausal women who were not on hormone replacement therapy. The soy product contained 90 mg/day of isoflavones. Women were randomized to receive either the soy product or placebo. Customary thyroid tests, TSH (thyroid stimulating hormone), thyroxine (T4), and triiodothyronine (T3) were measured at baseline and after 3 months and 6 months.

Values for all three measures were statistically almost identical at six months, and levels were similar between the isoflavone group and the placebo groups at each of the measurements. The authors concluded that in healthy individuals, who did not have an iodine deficiency, soy isoflavones do not adversely affect thyroid function (Bruce B, et al. J Med Food 2003; 6:309–316).

Potential adverse effects of soy have been aggressively communicated by individuals and institutions that seem to have a very emotional, vehement, and at times hostile attitude about soy. I do not know or understand the source of this heated emotion, but I believe it has created unnecessary confusion amongst consumers and patients of natural medicine. I prefer a studious, academic approach to evaluating and understanding the benefits and risks of soy, based on reliable research, rather than perspectives dominated by opinion and heated emotion.

In vitro studies have shown that genistein and daidzein, the two main isoflavones in soy, interfere with thyroid enzymatic reactions that are critical to the production of thyroid hormone (Divi R, et al. Biochem Pharmacol 1997;54:1087–1096). Animal studies have demonstrated that soy isoflavones, in a dose dependent manner, have an inhibitory effect on thyroid peroxidase, and goiter development is more common in rats fed a soy diet, especially if they are deficient in iodine (Doerge D, Sheehan D. Environ Health Perspect 2002; 110 (Suppl 3): 349–353). Other aspects of thyroid function including thyroid hormone levels and thyroid gland weight, were not adversely affected in the soy-fed rats.

Years ago, there were some case reports of goiter development in infants who were exclusively fed soy formulas. Once iodine was added to the formulas, no further cases have been reported. A study of women with elevated cholesterol levels re-stimulated this concern. These women were fed 40 g of soy daily with either 90 mg of soy isoflavones or 56 mg of isoflavones per day for six months. Cholesterol figures, steroid hormone levels and thyroid hormones were measured at 90 days and 180 days. While there were small but measurable effects on thyroid hormone, this was not associated with any detectable clinical changes (Bruce B, et al. J Med Food 2003; 6:309–316).

The dose of soy isoflavones (90 mg/day) in the current study, is a higher daily dose of isoflavones than what is consumed in the typical Asian diet. I would assert that this study is very significant and clarifying regarding concerns about the effect of soy isoflavones on thyroid function in adult women. The potential anti-thyroid effects of soy, appears to be directly related to iodine. If there is an iodine deficiency, then soy may be able to interfere with optimal thyroid function. However, iodine deficiency is easily corrected with dietary or iodine supplementation. With normal iodine levels, women should feel reassured that there is no clinically significant effect of soy on thyroid hormone levels or thyroid function. A daily serving of wholesome soy foods can be eaten safely by the vast majority of women.