Soy Isoflavones: No Evidence of Endometrial Harm at 3 Years

Miriam E. Tucker

October 18, 2014

NATIONAL HARBOR, Maryland — Soy isoflavone supplements appear to be safe during a 3-year period, new data suggest.

D. Lee Alekel, PhD, from the National Center for Complementary and Alternative Medicine (NCCAM), Bethesda, Maryland, and colleague presented the latest findings from the randomized, placebo-controlled Soy Isoflavones for Reducing Bone Loss (SIRBL) study October 16 at the North American Menopause Society (NAMS) 2014 Annual Meeting.

Several previous short-term studies have also demonstrated no adverse effects of soy isoflavone supplements on endometrial thickness, but one 5-year study showed an association between their use and endometrial hyperplasia, although not cancer.

By contrast, in the SIRBL study, Dr Alekel and colleagues found no evidence of treatment effects on endometrial thickness, adverse events, or circulating hormone concentrations, including thyroid function, among 255 women randomly assigned to receive either 80 mg or 120 mg of soy isoflavones or placebo. Similarly, a subgroup analysis of women with greater than 80% compliance showed no evidence of harm.

The authors note that although the primary findings from the trial, first reported in 2009, were largely negative (no bone-sparing effect, except a modest one at the femoral neck), women still use the supplements.

"Women will take them because they have closely held beliefs.... The message [from our data] about [these] supplements is that they're probably not harmful," Dr Alekel told Medscape Medical News.

Session moderator Katherine M. Newton, PhD, director of research and external affairs and senior investigator at Group Health Research Institute, Seattle, Washington, agrees. "Many women continue to use these products even when randomized trials show that they really don't do what women are trying to have them do. I think it is reassuring that the endometrial safety appeared to be there over 3 years," she told Medscape Medical News.

"It was a nice, long-term study, and that's important. Most menopausal symptom studies are very short, around 8 to 12 weeks. Rarely are they done for a year. The fact that she had 3 years of data with really good adherence was nice," Dr Newton said.

Standardized Supplements

Women enrolled in the SIRBL trial were younger than 65 years at baseline, had undergone natural menopause within 1 to 8 years, did not have significant vasomotor symptoms, were nonsmokers, were not obese or extremely underweight, were not vegan (because vegans are likely to eat more soy), and did not drink more than 7 alcoholic drinks a week, because alcohol interferes with hormone metabolism. They also were not taking long-term medications, including hormones.

Unlike supplements sold in health food stores or on the Internet, the soy isoflavone supplements used in the trial were made under controlled conditions to ensure dose consistency. "We were pretty much assured that we had [the specified] amount in the tablets. You can't always be sure what you get off the shelf. That's true of any supplement study you do," Dr Alekel told Medscape Medical News.

The researchers included a total of 224 women in the intent-to-treat analysis, and 208 women actually completed the study. Compliance, defined as having taken at least 80% of all pills, as assessed by pill count, was above 95% for all three treatment groups and was confirmed by urinary daidzein levels. The investigators collected data at 6, 12, 24, and 36 months.

They found no differences between treatment groups in circulating hormone levels during the study period, although there were significant differences in estrone sulphate and thyroid stimulating hormone (TSH) levels between the two study sites.

Median endometrial thickness declined from baseline through 36 months in all three treatment groups, from 1.5 to 1.1 mm at one study site and from 2.6 to 1.9 mm at the other. Neither the 80-mg nor the 120-mg dose exerted an effect (P = .57 and P = 0.43, respectively).

Adverse event rates differed between the two study sites, but not between treatment groups. The most common of these were upper respiratory tract infection and nonspecific complaints. There were 18 serious events, including four in the control group, eight in the 80 mg/day group, and six in the 120 mg/day group. None were deemed related to treatment, Dr Alekel said.

Factors that did significantly predict less loss in endometrial thickness across time points were higher exposure to estrogen (P = .0013) and plasma 17ß-estradiol (P = 0.0086) and greater alcohol intake (P = .023). Serum TSH did not achieve statistical significance (P = .066).

In addition to presenting the trial results at the meeting, the study authors published the data recently in Menopause.

Dr Alekel is a federal employee and has disclosed no relevant financial relationships. Dr Newton, through her institution, once received funding from Otsuka Pharmaceutical Co, Ltd, which makes a soy-based product.

North American Menopause Society (NAMS) 2014 Annual Meeting. Presented October 16, 2014. Abstract S-8.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.