Red Clover for Menopausal Symptoms

February 2000; Volume 2: 9-12

By Adriane Fugh-Berman, MD and Fredi Kronenberg, PhD

Red clover (trifolium pratense) has been used in america for more than 100 years as a medicinal herb for skin problems such as psoriasis and eczema, to treat lung conditions such as asthma and whooping cough, as well as for gout pain and even cancer.1,2 Leaves, stems, flowers, and sprouts are occasionally eaten but, unlike soybeans, clover has never been a dietary staple of humans. Red clover dietary supplements have been on the market for quite some time, without mention of specific uses; however, recently red clover has been marketed specifically as a dietary supplement for peri- and post-menopausal women.

Despite a paucity of clinical data, red clover for the treatment of menopausal symptoms has become popular in the United States, and the worldwide market for red clover is now substantial. The most heavily promoted brand is Promensil™, a standardized extract manufactured by Novogen Ltd., an Australian company. According to the company’s web site, sales of its red clover products for the year ending June 30, 1999 were $19.69 million, including $8.8 million from Australia and New Zealand and $10.9 million from the United States and Europe. The package states that the product is "...for women concerned about maintaining healthy estrogen levels to preserve physical and emotional well-being during midlife."

Red clover contains the phytoestrogens formononetin, biochanin A, daidzein and genistein.1 This review of the available red clover literature assesses the current scientific data on the safety and efficacy of this supplement for menopausal symptoms. Although the data are by no means conclusive, the surging public interest and lay press coverage make it valuable to keep in mind the actual state of the science.

Hot Flashes and Other Menopausal Symptoms

There are only two clinical trials published in the English peer-reviewed literature on red clover extract for menopausal symptoms. One double-blind, placebo-controlled, crossover trial enrolled 51 women (whose last menses were at least six months past) having at least three hot flashes a day.3 They were randomized to one daily tablet of placebo or Promensil (containing 40 mg total isoflavones, including genistein 4 mg, daidzein 3.5 mg, biochanin 24.5 mg, and formononetin 8 mg.). The first phase of the trial lasted for three months, followed by a one-month washout period, after which they were crossed over to the other arm for 14 weeks (three months plus an additional two weeks to accommodate the possibility of change in reporting habits that might occur at the end of a trial). Symptom diaries utilizing the Greene Menopause Score (a validated symptom self-assessment scale) were kept by the subjects. At the beginning of the trial, and at the conclusion of each treatment arm, subjects underwent a medical examination and blood tests (CBC, liver function tests, and levels of follicle-stimulating hormone [FSH], estradiol, and sex hormone binding globulin [SHBG]; a 24-hour urine sample was also collected for analysis of genistein, daidzein, formononetin, and biochanin A). A vaginal smear for vaginal maturation index was collected and transvaginal ultrasound was performed to assess endometrial thickness.

Forty-three women completed the study. Hot flash frequency decreased in both placebo and active groups at 12 weeks (about 18% and 20% respectively); there were no statistically significant differences between groups in Greene Scores at any time point. No significant differences were seen between groups in SHBG levels, blood counts, serum electrolytes, urea, creatinine, or liver function tests. No change in body weight was observed in either group. There were no differences in endometrial thickness or in the vaginal maturation index between treatment groups at the beginning and end of the trial. Urinary isoflavones increased significantly in the red clover group but not in the placebo group (P < 0.001). There was no increased endometrial thickness by ultrasound; however, this study was too short to adequately assess estrogenic effects on endometrium. Information on adverse events was not reported. The number of subjects was small, and the efficacy data are not compelling, as no significant effects on hot flashes or Greene Score were noted. The treatment period was too short to determine conclusively the safety of the product in terms of endometrial stimulation.

In another trial of the effects of red clover extract on menopausal symptoms, 37 postmenopausal women having at least three hot flashes per day were randomized to placebo or one of two doses of red clover (Promensil 40 mg or 160 mg) for 12 weeks.4 Endpoints were similar to the previous trial. Hot flash frequency decreased in all groups over the 12 weeks (35%, 29%, 34%); there was no significant difference among the three groups. There were no significant differences from baseline in any group in FSH, SHBG, vaginal maturation scores, or vaginal pH (the latter two are biological indicators of estrogenic activity). No "gross abnormalities" in hematological profile or liver function were reported. Urinary isoflavone level increased from baseline to week 12 when women took Promensil. The authors state that "a smaller but notable increase [in urinary isoflavone levels] was also observed in the placebo group." Although the significance of this increase is not stated, it may be inferred from the table that the increase was insignificant. In the placebo group urinary isoflavones levels rose from 2.68 ng/ml +/-1.92 to 3.67 ng/ml +/-2.79; in the 40 mg Promensil group urinary isoflavones increased from 3.43 ng/ml +/-2.07 to 9.40 ng/ml +/-5.67; in the 160 mg dose group urinary isoflavones increased from 3.70 ng/ml +/-2.96 to 28.18 ng/ml +/-17.52. The authors report that one of the subjects consumed alfalfa (which contains phytoestrogens), and that the small increase in urinary isoflavones in the placebo group may have been caused by this. However, this cannot be determined from the information given. Careful monitoring of subjects’ food consumption is difficult but should be addressed in future studies.

Although the investigators reported that they did post hoc analyses of stored serum for blood lipid profiles, only HDL levels are reported. The investigators’ stated reason for not supplying the data on total cholesterol and triglycerides is that these levels could not be accurately evaluated as the samples were collected in a non-fasting state. Although it is true that triglyceride levels are greatly affected by fasting state (and total cholesterol is affected to a lesser extent), these numbers should have been reported if only to indicate whether or not there was a difference in non-fasting state from baseline to week 12. Serum HDL cholesterol levels increased significantly by 18 % (P = 0.038) in subjects taking the 40 mg dose; HDL levels for those taking 160 mg were not different from the placebo group. These data, although intriguing, are not possible to evaluate without the total cholesterol levels. The ratio of HDL to total cholesterol is important; if the increase in HDL is due to an increase in total cholesterol, this may not be a benefit. There is a comment that "Two patients were subsequently withdrawn from the 160-mg group because of intervention by their general practitioners." No information is offered as to whether this was due to adverse effects.

The abstract (often the only part of a publication to be read) editorializes in the conclusion. The results section of the abstract states the facts: No significant difference in hot flashes or Greene Score, or anything else was measured. It states, "A statistically significant increase in high-density lipoprotein...occurred in the 40 mg group" without noting that there was no increase in the higher-dose group, and that this was a post hoc analysis. However, the authors postulate in the conclusion that "...a large placebo response and inadvertent use of dietary isoflavones in the placebo group may have obscured a significant change in flushing frequency. Previous uncontrolled studies claiming a beneficial effect of foods with a high isoflavone content on menopausal symptoms may have been confounded by a large placebo response." But in other studies of hot flashes, large placebo effects of 30-60% are common and this study is thus consistent with others. Placebo-controlled trials are done for the express purpose of separating placebo effect from treatment effect. It also ignores the fact that there have been several controlled studies showing at least some beneficial effect of soy and other high-isoflavone foods on hot flashes.5-7

The number of subjects in the reported studies have been few, so we await results of ongoing or future trials with a larger number of subjects, particularly to follow up on the intriguing but inconclusive HDL data.

Possible Adverse Effects

There are several areas where caution and vigilance are advised until additional data are collected. It has been documented that while clovers may be eaten by grazing animals, they can be associated with significant toxicity. White clover (T. repens) contains a cyanogenic glycoside, and is the species most commonly associated with poisonings.8 Subterranean clover (T. subterraneum), like red clover, contains large amounts of phytoestrogens, and its consumption by sheep in Australia in the 1940s resulted in "clover disease": Infertility, abnormal lactation, dystocia, and prolapsed uterus were attributed to the estrogenic effects.8 Whether this effect is specific to the animal species, the clover species, or both remains to be determined. Thus, long-term studies should be done to ensure that endometrial stimulation does not occur.

Breast cancer cell stimulation is also a concern. Breast cancer cell line studies would also be helpful; in one bioactivity assay, red clover was one of the top 5% of 150 herbs tested in an assay for both estradiol and progesterone binding.9

Finally, caution is warranted, as none of the published papers mentions coumarins, and blood clotting tests are absent from all reported clinical trials. It is known that coumarins are present in clover. The action of microorganisms on sweet clover (Melilotus alba) produces dicoumarol, a potent anticoagulant from which warfarin anticoagulants were originally derived.10 It remains to be determined whether this would be a concern for those taking the specific clover products and dosages used in the research studies. It is, therefore, important that PT/PTT tests be performed in subjects enrolled in clinical trials of red clover extracts to assess if this herb increases the risk of bleeding.

Summary of Data to Date

The two studies on menopausal hot flashes published in peer-reviewed journals have not demonstrated a reduction in hot flashes. The negative results of the two small published studies do not support the claim that red clover is an effective treatment for menopausal symptoms, and important safety questions remain. Additional studies are necessary to further elucidate these issues.

Funding of Reviewed Studies

The two reported red clover studies were supported by grants from Novogen Ltd.

References

1. Foster S, Tyler VE. Tyler’s Honest Herbal. 4th ed. Binghamton, NY: The Haworth Herbal Press; 1999.

2. Holmes P. The Energetics of Western Herbs: Treatment Strategies Integrating Western and Oriental Herbal Medicine. Vol. 1. Boulder, CO: Snow Lotus; 1989.

3. Baber RJ, et al. Randomized placebo-controlled trial of an isoflavone supplement and menopausal symptoms in women. Climacteric 1999;2:85-92.

4. Knight DC, et al. The effect of Promensil™, an isoflavone extract, on menopausal symptoms. Climacteric 1999;2:79-84.

5. Albertazzi P, et al. The effect of dietary soy supplementation on hot flushes. Obstet Gynecol 1998;91:6-11.

6. Murkies AL, et al. Dietary flour supplementation decreases post-menopausal hot flushes: Effect of soy and wheat. Maturitas 1995;21:189-195.

7. Brzezinski A, et al. Short-term effects of phytoestrogen-rich diet on postmenopausal women. Menopause 1997;4:89-94.

8. Lewis RA. Lewis’ Dictionary of Toxicology. Boca Raton, FL: CRC Press; 1998.

9. Zava DT, et al. Estrogen and progestin bioactivity of foods, herbs, and spices. Proc Soc Exp Biol Med 1998;217:369-378.

10. Dewick PM. Medicinal Natural Products: A Biosynthetic Approach. West Sussex, England: John Wiley and Sons, Ltd.; 1997.