Red Clover ( Trifolium pratense ) Isoflavones and Serum Homocysteine in Premenopausal Women: A Pilot Study

Introduction

Evidence is suggestive of a link between elevated homocysteine (Hcy) concentrations and neural tube defects, ¹ cardiovascular disease (CVD), ² osteoporosis, ³ and cognitive decline. ⁴ Although the etiology of hyperhomocysteinemia is not fully understood, enzymes and intermediates of the methylation pathway are likely to play an important role. Increasing the intake of folate and vitamin B₁₂ has been shown to decrease elevated plasma Hcy concentrations. ⁵ The effects of macronutrients and phytochemicals on plasma Hcy are less well understood.

In a cross-sectional study of premenopausal women, the intake of soy products was shown to be inversely associated with plasma Hcy concentrations, ⁶ whereas trials with soy protein have yielded mixed results, with a Hcy-lowering effect supported by some studies ^7,8 but not others. ^9,10 Similarly, supplementation with the isoflavone constituents of soy has produced conflicting results. ^11,12 There is little information on the homocysteine effect of isoflavones from sources other than soy. In the present study, we used purified isoflavones that were derived from red clover (Trifolium pratense), which, in contrast to those in soy, contain a mixture of methoxylated (biochanin A, formononetin) and unmethoxylated (genistein, daidzein) isoflavones.

Intervention studies that use red clover-derived isoflavone supplements would be informative in young healthy women to determine if any benefits can be found on risk factors that are associated with neural tube defects, breast cancer, and CVD. As part of a larger study aimed at exploring the effects of isoflavone supplements on women's health, ^{13 –16} the effect of red clover isoflavones on serum Hcy and folate concentrations was investigated in a group of healthy premenopausal women.

Subjects and Methods

Women aged 18–45 years were invited to participate in the study by advertisements in local print media. Subjects were included in the study if they reported having regular menses; no history of chronic illness or a family history of cancer; no diseases of the liver, gallbladder, bowel, or kidney; not taking lipid-lowering medication, antibiotics, oral contraceptive agents (OCA), steroids, or isoflavone supplements; stable dietary intake; and consuming less than one serving per week of foods rich in isoflavones. The protocol was approved by The University of Sydney Human Ethics Review Committee, and all volunteers provided informed consent. Details of the selection criteria and study protocol are reported elsewhere. ¹⁴

A randomized, double-blind, placebo-controlled, parallel trial was conducted over four menstrual cycles. Each subject commenced the study on the first day of menstruation, and subjects were requested to maintain their habitual dietary intake and exercise patterns. All subjects consumed two placebo tablets daily during the first menstrual cycle. For the ensuing three cycles, the placebo group (PL) continued consumption of the placebo tablets, whereas the intervention group (IF) was switched to isoflavone-containing tablets (Novogen Pty Ltd., North Ryde, Australia). The isoflavones were derived from subterranean red clover (T. pratense), and each tablet contained formononetin (9.3 mg), biochanin A (25.7 mg), genistein (4.3 mg), and daidzein (3.7 mg), providing a total of 86 mg isoflavones per day.

During menstrual cycles 1, 3, and 4, weekly blood samples were taken from subjects in the fasted state, between 7.30 and 10.30 am. Blood samples were collected into vacutainers (Becton Dickinson, Franklin Lakes, NJ), allowed to clot, and then centrifuged at 1500 g for 15 minutes at 4°C. Serum was removed and stored at −80°C until analysis. Serum folate concentrations were determined by electrochemiluminescence (Elecsys Folate Reagent Kit, Roche Diagnostics, Sydney, Australia), and serum homocysteine concentrations were measured by fluorescence polarization immunoassay (IMX system, Abbott Laboratories, North Ryde, Australia). First morning urine samples were collected in each cycle during the 2-week interval around ovulation (days 8–21, where day 1 was the first day of menstruation) for analysis of luteinizing hormone (LH) to determine the day of ovulation.

During week 3 of each month, subjects kept a 3-day weighed food record, and nutrient intake was analyzed using Foodworks 2005 (version 4.0, Xyris Software Pty Ltd., Highgate Hill, Australia), with the Australian Food and Nutrient database (AUSNUT), including brand-named foods. Folate values used in AUSNUT were largely based on the U.K. tables of food composition, with the exception of folate-fortified foods. Attention was paid to identifying only foods fortified at the time of data collection, based on information about the timing of fortification from Food Standards Australia New Zealand.

Mean monthly values of folate and Hcy were calculated by averaging the data from 4–6 weekly blood samples from each subject. For comparison between the intervention group (IF) and placebo group (PL) and phases of the menstrual cycle (follicular/luteal), a repeated-measures general linear ANOVA model was used, with between-subjects factor group and within-subject factor menstrual phase. Group and phase were used as independent factors, and mean baseline folate and Hcy values were included as covariates. Statistical analysis was performed using SPSS, version 14, (SPSS Inc., Chicago, IL), and p < 0.05 was considered statistically significant.

Results

The subjects' age, weight, and body mass index (BMI) were 32.8 ± 9.5 years, 61.9 ± 12.6 kg, and 23.0 ± 4.1 kg/m², respectively (mean ± SD, n = 23). The average menstrual cycle length during the first month of the trial was 29.7 ± 6.5 days, and the follicular and luteal phase lengths were 15.7 ± 4.6 and 13.2 ± 1.9 days, respectively. Baseline characteristics were similar in those allocated to IF (n = 11) or PL (n = 12). The baseline concentrations of folate and Hcy in serum were within the reference ranges and did not change significantly in either group. There were no significant differences observed in either folate or Hcy between the follicular and luteal phases of the menstrual cycle (Table 1).

Compliance with the study protocol was confirmed by analysis of urinary isoflavone concentrations. Supplementation with isoflavones resulted in a 15-fold increase in urinary isoflavone excretion (p < 0.0001), and tablet counts in returned containers indicated >95% compliance in both the PL and IF groups. ^14,15 There were no temporal or group differences in energy intake, percent of energy derived from macronutrients, dietary fiber, or folate, indicating that nutrient intake was constant (Table 2).

Discussion

The findings of the present supplementation trial suggest that methoxylated isoflavones have no effect on serum Hcy or folate concentrations. The strengths of the study were that purified isoflavones were used in the absence of any dietary manipulation, and the supplement was free of any matrix effects, particularly methionine ¹⁷ and phytate, ¹⁸ which have been reported as significant determinants of plasma Hcy. Stages of the menstrual cycle were also monitored, and up to six blood samples were collected from each participant during each menstrual cycle. The lack of effect is not a result of poor compliance, as the mean urinary isoflavone concentrations increased significantly in the IF group and remained low in the PL group. ¹³ Urinary isoflavone metabolites were significantly higher in the IF group compared with the PL group: biochanin A, daidzein, and genistein increased 6–10-fold, formononetin and dihydrodaidzein increased 13-fold, equol increased 27-fold, and O-desmethylangolensin increased 228-fold. ¹⁵

A limitation of this study was the nonsignificant trend toward a higher intake of alcohol and lower intakes of energy and folate in the PL group compared with the IF group. A low folate intake can cause elevated Hcy levels; however, the subjects' dietary intake included foods that were fortified with folate, and none were vegetarian, which suggests an adequate intake of folate and vitamin B₁₂, the main determinants of Hcy. ⁵ Body weight was monitored weekly throughout the study, and all subjects remained at a stable weight. Although mean energy intake was rather lower in the placebo group at the end of the study compared with the beginning, this change was not statistically significant nor accompanied by any change in body weight. Thus, the consistency in plasma Hcy may be related to the small sample size of the study, the subjects' adequate intake of folate, or the low concentrations of Hcy at baseline.

Changes in plasma Hcy concentrations have been reported during different stages of the menstrual cycle. In a study in 15 premenopausal women, Tallova et al. ¹⁹ showed a higher concentration of Hcy in the follicular phase compared with the luteal phase and a weak negative correlation between Hcy and estrogen concentrations in plasma. Folate concentrations were not reported, and it is unclear if folate fortification of the food supply was an important factor in that study. In an earlier report ²⁰ where folate intake was similar to that reported in the present study, serum folate concentrations were unaffected by stages of the menstrual cycle but were lower in OCA users. In the present study, OCA users were excluded, but it is possible that the effect of isoflavones is different in OCA users and nonusers.

Isoflavones possess two key structural features in common with estradiol: a rigid planar structure and the presence of two hydroxyl groups in the A and B rings, which enable isoflavones to activate the estrogen receptor. The relative binding affinities of isoflavones are lower than that of estradiol, indicating a lower estrogenic potency. ²¹ A lower estrogenic potency is more likely to occur in the presence of high baseline estrogen, as is the case in premenopausal women. Previous studies using soy isoflavones in young women over a 12-month period showed no effect on biomarkers of bone metabolism. ²² Similarly, previous studies with red clover isoflavones over an intervention period of 4 months have shown no effect on plasma lipoproteins or markers of insulin sensitivity. ¹⁴ The lack of effect in the present study may be related to the apparent lower estrogenic potency of isoflavones in premenopausal women, although this was not measured in the present study.

Studies in isolated human liver microsomes have shown that O-demethylase activity is associated with cytochrome P-450. ²³ O-Demethylase acts on biochanin A and formononetin to produce genistein and daidzein, respectively. We hypothesized that demethylation of red clover-derived methoxylated isoflavones in the liver will impact Hcy metabolism by providing methyl substrates for the one-carbon cycle. The lack of effect on plasma Hcy in the present study supports the observation that demethylation by the gut microflora ²⁴ may have a prominent role that limits the availability of methoxylated isoflavones in the liver. That the effects of isolated isoflavones are the same as those of the intact plant remains to be tested. In conclusion, the results of the present study in healthy young women with normal menses suggest that red clover isoflavones in isolation are unlikely to affect serum Hcy or folate metabolism.