Fax: (813) 632-1334
The specific role of isoflavones on estrogen metabolism in premenopausal women
Version of Record online: 20 FEB 2002
Copyright © 2002 American Cancer Society
Volume 94, Issue 4, pages 1166–1174, 15 February 2002
How to Cite
Kumar, N. B., Cantor, A., Allen, K., Riccardi, D. and Cox, C. E. (2002), The specific role of isoflavones on estrogen metabolism in premenopausal women. Cancer, 94: 1166–1174. doi: 10.1002/cncr.10320
- Issue online: 20 FEB 2002
- Version of Record online: 20 FEB 2002
- Manuscript Accepted: 15 OCT 2001
- Manuscript Revised: 11 OCT 2001
- Manuscript Received: 19 JUL 2001
- National Institutes of Health, National Cancer Institute. Grant Number: RO3 CA72588-01A1
- Department of Defense. Grant Number: DAMD17-98-1-8659
- breast carcinoma;
- steroid hormones;
- estrogen metabolism;
- menstrual cycle length
There is increasing evidence that dietary factors may play a role in the production, metabolism, and bioavailability of sex hormones and their impact on target tissues. The specific objective of this study was to evaluate the effectiveness of supplementing a group of premenopausal women who were free of breast carcinoma with a dietary supplement of isoflavones (40 mg per day) in producing a change in steroid hormones and menstrual cycle length.
Sixty-eight consecutively recruited, premenopausal, omnivorous women of all races and ethnicities between the ages of 25 years and 55 years were admitted to the study and randomized to an experimental group supplemented with soy (40 mg genistein per day) or to a control group that consumed a placebo for a 12-week period. Changes in their anthropometric, nutritional, and hormonal biomarkers from early follicular phase were analyzed at baseline and postintervention.
Serum-free estradiol and estrone levels decreased moderately in the experimental group. Serum hormone-binding globulin levels increased in 41.4% of women in the experimental group compared with 37.5% of women in the placebo group. Free estradiol decreased in 53.85% of women in the experimental group compared with 37.5% of women in the placebo group. Estrone decreased in 55.56% of women in the experimental group compared with 42.86% in the placebo group. Those women in the experimental group who were consuming soy had their mean menstrual cycle length increased by 3.52 days compared with a mean decrease of 0.06 days for women in the placebo group (P = 0.04) from baseline to the third menstrual cycle. In addition, women who were taking soy had their mean follicular phase increase by 1.46 days compared with a mean increase of 0.14 days for women who were taking the placebo (P = 0.08).
These data suggest that increased isoflavone intake affects estrogen metabolism by altering the steroid hormone concentrations and menstrual cycle length, thereby demonstrating a potential to reduce the risk for breast carcinoma. Cancer 2002;94:1166–74. © 2002 American Cancer Society.
Breast carcinoma is the most common cancer and the second leading cause of cancer death in women after lung carcinoma. The American Cancer Society predicts that there will be about 182,800 new incidents of invasive breast carcinoma in the year 2000 among women in the United States and about 40,800 deaths from the disease. It has been suggested that 30–60% of all malignancies in the developed world may be attributed, at least in part, to dietary habits.1 There is increasing evidence that dietary factors may play a role in the production, metabolism, and bioavailability of sex hormones and their impact on target tissues.2, 3 Although the most highly scrutinized dietary difference between the Asian countries, such as China and Japan, and the Westernized countries is the low-fat, higher fiber intake, it has been suggested that consumption of soy may contribute to the lower rates of breast, prostate, and colon carcinoma in Asian countries.4–8 Epidemiologic studies have demonstrated that populations consuming soy products have the lowest incidence of breast carcinoma.5–7, 9–12 Several natural anticarcinogens have now been identified in soybeans, such as protease inhibitors, phytates, phytosterols, saponins, lignans, and isoflavones.10, 13–17 Diadzein and genistein are the major forms of isoflavones, also referred to as phytoestrogens, present in soybeans and, after structural modifications by intestinal bacteria, convert to compounds that are known to possess weak estrogenic and antiestrogenic properties.9, 10, 18–20 Case–control studies have shown that urinary levels of phytoestrogens were lower in patients with breast carcinoma compared with control participants.11, 12, 18, 21 In addition to these substances influencing hormonal metabolism, isoflavones also affect intracellular enzymes, protein synthesis, growth factor action, cell proliferation, and angiogenesis.4, 13, 22, 23
There are several mechanisms by which soy isoflavones specifically may modulate the risk of breast carcinoma. It has been observed that these soy isoflavones, due to their similarity to estrogen both structurally and functionally, influence breast carcinoma risk by modulating the steroid hormones. Isoflavonoid phytoestrogens have been shown to increase serum levels of sex hormone-binding globulin (SHBG), which then decreases the bioavailability of estrogen, because higher levels of SHBG result in the lowering of free estradiol levels.18, 24–28 This also may be due to the weak estrogenic effect of phytoestrogens, which, like tamoxifen, stimulate the synthesis of SHBG in the liver.18 More recently, clinical studies examining the effects of soy isoflavones on serum estrogen levels have demonstrated moderate reductions in ovarian steroid levels.29–31 Other clinical studies have shown that increased intake of isoflavones converts the endogenous estrogens to the protective 2-hydroxylated estrogens in women and may play a critical role in lowering levels of 17-α hydroxyestrone,30, 32–35 a known stimulant of breast proliferation, thereby reducing the long-term risk of breast carcinoma. Several studies also indicate that phytoestrogens reduce the bioavailability of estrogens by actually occupying estrogen-binding sites, exerting a weak estrogenic effect (antiestrogenic), and decreasing the availability of estrogen receptors to endogenous, biologically active estrogen.18, 24, 26, 36, 37
Studies examining the effects of a soy protein diet on the menstrual cycle have demonstrated a significant increase in follicular phase length and delay in menstruation,29, 38 including the suppression of midcycle surges of luteinizing hormone (LH) and follicle-stimulating hormone (FSH), which potentially may reduce the risk of breast carcinoma.29–31 A change in the menstrual cycle length alters the relative duration of mammary epithelial cells in the luteal phase of a cycle, during which time the breast cells are more proliferative.32, 39, 40 Thus, the levels of these steroid hormones that affect proliferation and menstrual cycle length are recognized risk factors for breast carcinoma. Isoflavones, specifically genistein, have received a great deal of attention due to their interesting antiproliferative properties10, 15, 41–43 and have been shown to inhibit breast proliferation in vitro.27, 44 However, there are contradictory reports suggesting that isoflavones increase rodent mammary gland proliferation45 and proliferation in breast tumor cell lines,46 which may be a dose-dependant phenomena.47 Investigators recently demonstrated that supplementation with isoflavones for 14 days affected histologically normal breasts in premenopausal women, as indicated by an increase in the number of cells in the S-phase of the cell cycle and a small increase in progesterone receptor expression.48, 49 In a more recent report on one of the same studies with double the number of participants, Hargreaves et al. failed to see any difference in breast proliferation, apoptosis, or progesterone receptor expression, as observed in their preliminary report of 1998. In addition, these clinical trials, which report higher proliferation with short-term supplementation, at most, were pilot trials with small sample sizes and short duration of interventions that did not exceed 1 month.49, 50
Thus, the mechanism by which soy isoflavones influence the risk of breast carcinoma remains unclear. Although there are some epidemiologic studies and several animal studies supporting the cancer-preventative qualities of soy products, there have been few definitive, prospective clinical studies30–32, 51–53 that tested the effects of specific isoflavones on female steroid hormones and menstrual cycle length with substantial numbers of participants and duration of intervention. Although isoflavones have several interesting properties, our specific objective was to evaluate further the individual effectiveness of supplementing a group of premenopausal, breast carcinoma free women with a dietary supplement of isoflavones (40 mg per day) in producing a change in female steroid hormones that are implicated in the initiation and promotion of breast carcinoma. In addition, our goal was to observe the possible effects of isoflavone supplementation on menstrual cycle lengths, an increase of which has been implicated in the reduction of breast proliferation and, thus, breast carcinoma risk.
MATERIALS AND METHODS
Ninety-seven healthy, premenopausal, omnivorous (including nonvegan vegetarians) women, of all races and ethnicities between the ages 25 years and 55 years, inclusive, at first screening were admitted consecutively to the study. The women were volunteers who were recruited from the Tampa Bay area and from the Lifetime Cancer Screening Program at the H. Lee Moffitt Cancer Center and Research Institute and were solicited though the Tampa Tribune and the Cancer Center publication, Lifetime Choices. Interested participants were screened by a research dietitian using an initial screening questionnaire that included information pertaining to the inclusion and exclusion criterion. This study was approved by the local Institutional Review Board, and all participants provided written, informed consent. Sixty-six of 97 participants completed the study. On recruitment, participants were randomized to the experimental group or the control group using the procedure for telephone randomization,54 as previously reported. The study was double blinded. Participants in the experimental group consumed a soy protein supplement (Protein Technologies International, St. Louis, MO) containing 40 mg of genistein (20 mg per dose), and participants in the control group received an isocaloric placebo supplement of identical appearance using milk protein (Protein Technologies International). The dose of isoflavones was selected based on data reflecting that the Asian population, which has the lowest rates of breast carcinoma and prostate carcinoma, consumes an estimated 40–80 mg of isoflavones per day as part of their habitual diets compared with 1 mg per day consumed by Western populations. Most of the clinical studies, including our previous studies, using soy isoflavones have used 1–2 mg isoflavones per kilogram of body weight. At that dose, serum concentrations of isoflavones increased, and hormonal and proliferative modulations have been observed.7, 11, 53 The established dose of 40 mg was based on the data from those studies.
Criteria for exclusion included history of cancer of any type; pregnancy or breastfeeding within 12 months preceding the study; current use of hormones, including birth control pills or hormone replacement therapies; perimenopausal status; irregular menstrual cycles; body mass index > 38 kg/m2; antibiotic use within a 3-month period prior to recruitment; and vegans. In addition, participants, including nonvegan vegetarians, who already were consuming > 20 g of fiber per day or who were consuming oral fiber supplements, participants who were allergic to soy or casein, and participants who already were consuming soy products (e.g., tofu, miso, soy flour, and soy milk) were excluded from the study.
On consenting to participate in the study, eligible participants were interviewed in our clinic by our trained research clinical dietitian, initially confirming the accuracy of eligibility information by using a screening checklist for eligibility. Because compliance to study protocol was critical, we screened all eligible participants' potential for compliance in the initial screening phase, prior to randomization and entry into study. Potential participants performed tasks similar to those required of them during the initial recruitment period, such as completing daily monitors. In addition, using a structured questionnaire, demographic information, personal and medical history, hormonal and reproductive history, exercise, smoking, and alcohol use history were obtained. Once they were found eligible and were recruited, participants were randomized and required to wait to start the intervention phase until cycle Days 4–7 (midfollicular phase). The duration of the study, when patients received the study agent or placebo, extended from cycle Days 4–7 after the first day of menses in the first cycle to the same period (cycle Days 4–7) after 3 menstrual cycles were completed. Anthropometric measurements, such as height, weight, skin-fold, and circumference measurements using standards, as described previously,28 were obtained at baseline and postintervention. Blood samples (30 mL) were drawn into heparinized tubes in a nonfasting state at the same time of day, between 7:00 AM and 12:00 P.M., for each individual participant. Blood samples were centrifuged at ×1300 g for 10 minutes at room temperature within 3 hours of sample collection, and the serum was separated and stored at −80 °C until assayed. Serum levels of estrone, estradiol, and SHBG were determined at baseline and postintervention by Quest Diagnostics (San Diego, CA). Blood samples were collected at baseline and at the end of study (12 weeks or after the completion of 3 full menstrual cycles). Both blood draws were completed on cycle Days 4–7 between 7:00 AM and 12:.00 P.M. Patients received their first dose of study agents and started their intervention phase after a baseline draw of blood for hormonal analysis. Each participant was provided a 1-month supply of the product at each visit.
Participants were instructed and required to monitor the length of their menstrual cycle and the luteal hormone surge using a commercial kit (First Response Ovulation Predictor; Carter-Wallace, Cranbury, NJ), which provided qualitative information regarding when ovulation occurred and length of follicular phase. Menstrual cycle history was maintained by all participants for the entire study period. In addition, participants completed a daily Symptoms Monitor, which was to be used by participants to document 1) days of menstrual cycle and 2) specific nutritional symptoms. Participants were instructed to document the first day of period, days of ovulation, days of menses, and absence of menses or ovulation. The precise phases of their cycle—follicular, ovulatory, menstrual, and luteal—were calculated from these daily monitors. Nutritional symptoms, such as bloating, constipation, diarrhea, fullness, or other, were required to be monitored daily on the same form.
The participant was provided with a 4-day diet record and was instructed on reporting food intake, including weights/measures and methods of preparation of foods consumed using standard food models. At entry to the study, patients were instructed specifically on the study agents, including mixing instructions; ideas for improving taste of the product without increasing other nutrients; importance of compliance to quantity of the study agent provided; completion of 4-day food records throughout the period of the study; compliance and completion of the daily Symptoms Scale, including nutritional symptoms and menstrual cycle documentation; attendance to clinic to obtain monthly supply of the study agent; self-monitoring tools and the research staff-completed self-monitoring tools and leftover supplements, if any; and the importance of compliance to the diet without altering or reducing other foods. The research staff used a Participant Tracking Form to monitor all activities and variables observed during the study period. Both groups were instructed to continue their usual lifestyle, including eating and physical activity habits. Research staff verified incomplete data during clinic visits or on the telephone. Participants completed the last dose of the study agent on the day of blood draw, on completion of 3 full menstrual cycles on cycle Days 4–7 between 7:00 AM and 12:00 noon, the same time when the baseline blood draw was performed.
The sample size in each group was chosen so that the comparisons described above for each time would have a power of 0.80 if the true difference between mean hormone levels of the two groups was 0.7 standard deviations. At the conclusion of the intervention, a pooled t test was used to compare the mean values for hormonal levels, nutrient intake, anthropometric measures, and the lengths of menstrual cycle between the two groups. The proportion of women who reported increased stool frequency and gastrointestinal distress in the two groups was compared by chi-square tests. Although compliance was monitored, we employed the intent-to-treat principle in all group comparisons. Participants were analyzed according to the group into which they were randomized without regard to compliance or actual diet.
Ninety-seven participants were recruited in the study, and 66 women completed the intervention. Thirty-one participants dropped out of the study, including 15 women from the placebo group and 16 women from the isoflavone group. The reasons for attrition included 1) unable to comply (record keeping, etc.) in eight women (six women in the isoflavone group and two women in the placebo group), 2) intolerance to product (bloating, nausea, feeling sick) in four women (two women in the isoflavone group and two women in the placebo group), 3) constipation in two women (both in the isoflavone group), 4) taste intolerance in four women (all in the placebo group), 5) pregnancy or birth control pill use after starting the study in three women (two women in the isoflavone group and one woman in the placebo group), 6) no reason in seven women (four women in the isoflavone group and three women in the placebo), and 7) weight gain in three women (all in the placebo group). An initial comparison of baseline demographic variables, such as age, anthropometrics, smoking history, parity, age at menarche, family history of breast carcinoma, and personal history of benign breast disease, is displayed in Table 1. Of the participants who completed the study, excellent compliance to study agents and monitors were observed, including use of ovulation predictor kits, 4-day diet records, and documentation of daily gastrointestinal symptoms and menstrual cycle days. Participants in both groups had similar intake of macronutrients and micronutrients, including soy isoflavones, at baseline. Participants in both groups were instructed clearly not to consume soy or soy products during the study period. Energy, protein, fat, and cholesterol intake increased significantly in the isoflavone-supplemented group from baseline to the end of the study, and protein intake increased significantly in the placebo group (Table 2). However, on examination of food records during the entire intervention period, no other consistent dietary modification was observed in either group. Therefore, this was more likely due to normal variation in diet, as reflected by the 4-day diet record at baseline and post-treatment. The liquid (milk products, fruit juices, etc.) used to mix the supplement was remarkably consistent in the two groups. The intake of soy isoflavones in the experimental group, as predicted, was significantly higher from baseline to postintervention compared with the placebo group.
|Variable||Isoflavone supplemented (n = 33 women)||Placebo (n = 33 women)||P valuea|
|Baseline weight (pounds)||143.80||5.14||147.05||4.48||0.32|
|Age at menarche||12.53||0.20||12.97||0.27||0.10|
|No. of full-term pregnancies||1.41||0.19||1.45||0.21||0.44|
|Birth control pill use (months)||63.00||12.87||51.38||10.53||0.76|
|Current smoker (%)||3.1||—||9.14||—||0.31|
|Family history of breast carcinoma (%)||56.7||—||41.9||—||0.25|
|History of benign breast disease (%)||46.9||—||39.4||—||0.54|
|Variable||Isoflavone supplemented (n = 33 women)||Placebo (n = 33 women)||P valueb|
|Baseline||Final week||P valuea||Baseline||Final week||P valuea|
|Protein (gm)||72.4||110.3||< 0.001||72.8||97.2||< 0.001||0.18|
|Vitamin A (RE)||1151.1||1131.3||0.82||1275.9||1030.4||0.35||0.56|
|Vitamin C (mg)||130.1||110.1||0.11||118.2||140.9||0.64||0.36|
|Vitamin E (mg)||8.5||6.1||0.46||5.2||4.6||0.47||0.59|
|Thiamin B1 (mg)||1.53||1.31||0.36||1.3||1.2||0.77||0.49|
|Riboflavin B2 (mg)||1.70||1.53||0.45||1.5||1.5||0.96||0.54|
|Niacin B3 (mg)||19.67||18.21||0.58||16.4||16.0||0.81||0.71|
|Pyridoxine B6 (mg)||1.47||1.38||0.38||1.3||1.3||0.66||0.35|
Participants in the experimental group who consumed supplements of isoflavones gained an average of 1.7 pounds during the study period, whereas the placebo group gained an average of 1.1 pounds during the same time. No significant changes in anthropometric variables, such as skin-fold and circumference measurements, were observed during this period.
The baseline and final concentrations of serum steroid hormones are displayed in Table 3. Serum free estradiol and estrone levels decreased in the experimental group, although the changes in hormonal levels between the two intervention groups were not statistically significant. Changes in menstrual cycle length from baseline to postintervention between the two groups were statistically significant (Table 4). Those participants in the experimental group who consumed soy had their mean menstrual cycle length increase by 3.52 days compared with a mean decrease of 0.06 days in the placebo group (P = 0.04) from baseline to the third menstrual cycle. In addition, participants on soy had their mean follicular phase increase by 1.46 days compared with a mean increase of 0.14 days for participants on placebo (P = 0.08) (Table 5).
|Variable||Isoflavone supplemented (n = 33 women)||Placebo (n = 33 women)||P valueb|
|Initial||Final||P valuea||Initial||Final||P valuea|
|Free estradiol (pg/mL)||1.37||0.21||1.28||0.19||0.47||1.54||0.18||1.59||0.25||0.19||0.48|
|Total estradiol (pg/mL)||44.58||4.31||47.26||6.66||.08||47.39||4.34||52.03||14.4||0.76||0.67|
|Cycle||Isoflavone supplemented (n = 33 women)||Placebo (n = 33 women)||P valuea|
|Cycle||Isoflavone supplemented (n = 33 women)||Placebo (n = 33 women)||P valuea|
There is a general agreement that steroid hormones are involved in the development of breast carcinoma. In addition to genetic factors, number of pregnancies, use of oral contraceptives, lifestyle factors that influence age at menarche and menopause, the levels of these steroid hormones and menstrual cycle length are recognized risk factors for breast carcinoma. A change in the menstrual cycle length over a woman's lifetime alters the relative duration of exposure of mammary epithelial cells to steroid hormone surges in the luteal phase of the menstrual cycle, during which time the breast cells are more proliferative.32, 39, 55 In general, when cycle length increases, the length of the follicular phase increases more than the luteal phase.56 Because breast cells proliferate two to three times more rapidly during the luteal phase than during the follicular phase,40 an increase in the cycle length and an increase in follicular cycle length, as observed in our study in participants who consumed soy supplements, theoretically may shorten the exposure of the breast epithelia to progesterone in the luteal phase. If this occurs over a long period of time with consistent soy consumption, then the relative amount of time during which the breast epithelia is stimulated to proliferate may decrease accordingly, and this may decrease the overall breast carcinoma risk.32 Differences in cycle lengths between breast carcinoma patients and control participants57 and between populations of women in countries with different breast carcinoma risk58 were consistent with the theory that menstrual cycle length may moderate breast carcinoma risk.59 The results of the current study support the hypothesis that soy consumption may alter circulating ovarian steroid hormone concentrations in premenopausal women and increase menstrual cycle length. The menstrual cycle of women from Western populations ranges from 26 days to 29 days, whereas the average cycle length for Japanese women and other Asian women is longer, which may account, in part, for the lower risk of breast carcinoma in populations59–61 that consume soy in their daily diet. The current study was able to demonstrate that daily soy consumption can lead to statistically significant changes in menstrual cycles, even in Western women.
Although we observed only a modest change in steroid hormonal concentrations with soy supplementation for a 12-week period, we observed a trend that demonstrated an increase in SHBG and a decrease in free estradiol and estrone levels in a relatively larger percentage of participants who consumed the soy supplement compared with the placebo group. Previous studies examining the effect of soy isoflavones on steroid hormones in premenopausal women similarly have shown only modest changes. In a pilot study, Lu et al.56 examined the effects of soy in six premenopausal women for 1 month in a metabolic unit and observed reduction in serum 17 β-estradiol, progesterone, and dehydroepiandrosterone sulfate levels and an increase in menstrual cycle length. Similarly, Nagata et al.31 demonstrated a moderate but not statistically significant decrease in estrone and estradiol levels in an experimental study with premenopausal Japanese women supplemented with soy milk. However, participants in the experimental group increased their menstrual cycles by nearly 2 days. Previous studies also have shown that treating premenopausal women with soy supplements, in addition to increasing the length of menstrual cycle length, resulted in lower concentrations of LH and FSH.49 However, in two other studies,52, 53 no steroid hormonal or menstrual cycle effects were observed. The sample size in those studies was small, and/or the duration of intervention ranged from 1 month to 2 months.
In summary, we demonstrated that supplementing with 40 mg of soy isoflavones per day in a premenopausal group of Western women can increase menstrual cycle and follicular cycle length and can moderately increase serum SHBG levels and lower estrone and free estradiol levels in a relatively greater percentage of participants in the experimental group compared with a placebo group. This is an important finding, because an increase in menstrual cycle length would reduce the number of menstrual cycles during a lifetime, thereby reducing the total number of times the breast is exposed to estrogen. In addition, women will spend more days in the increased follicular cycle, when proliferation is at its lowest. These effects are mediated by the pituitary gland, and long-term effects on the pituitary gland may result in an overall antiestrogenic effect49 and, thus, in a lower risk for breast carcinoma.
The most promising approach to carcinoma control is a national commitment to prevention. Basic research has identified nutrients as agents that inhibit mutagenesis and hyperproliferation, as well as those that induce apoptosis or differentiation, as critical characteristics for chemoprevention regardless of their specific molecular targets. There is increasing evidence that soy isoflavones are promising nutrients, as we and others have demonstrated, that may alter the plasma concentration, production, metabolism, and excretion of estrogens and their impact on target tissues and that also possess antiproliferative properties. However, it is critical to continue to identify the cellular and molecular mechanisms that are operational in breast carcinoma and to determine how these pathways may be modified by these agents. It is imperative for research to demonstrate that the chemopreventive agents can modulate the biomarkers chosen as surrogate end points. Phase II and III trials are critical for evaluating chemoprevention efficacy in suitable cohorts, as defined by risk factors.62 Future studies must examine the individual effectiveness of supplementing patients and high-risk populations with varying doses of the isoflavone supplements in modulation of intermediate end point biomarkers, such as indices of cell proliferation, apoptosis, tyrosine kinase inhibition, DNA ploidy, and DNA repair enzyme expression. Most importantly, the nature of trying to prevent the occurrence of malignancy requires the use of chemopreventive agents with little or no toxicity. It is critical to evaluate the chronic administration of these agents and establish dosage regimens for chemoprevention. While these studies are in progress, it is extremely critical for practitioners to await the results of empiric research that clearly demonstrates the efficacy of the agents in chemoprevention prior to recommending supplemental doses for cancer prevention or treatment.
The authors thank Paul Woodward for his assistance with data management.
- 1Cancer facts and figures. Atlanta: American Cancer Society, 1999.
- 13Diet and urinary estrogen profile in various populations: a preliminary report. Proceedings of the 14th International Meeting on Polycyclic Aromatic Hydrocarbons, 1993., , , , , , et al.
- 16Isoflavanoids and 2-methyoxyyestradiol: inhibitors of tumor cell growth and angiogenesis. Proc AACR 1995; 35: 693–4., et al.
- 18Genestein, a dietary ingested isoflavanoid, inhibits cell proliferation and in vitro angiogenesis. J Nutr 1995; 125: 790–7., , , , , .
- 19In vitro hormonal effects of soybean isoflavones. J Nutr 1995; 125L: 751–6., , .
- 24Differential effects of genestein on p52 expression and proliferation in MCF-7 cells are concentration-dependent. Pro AICR 1994; 35: A503., , .
- 27Effect of genestein on in vitro and in vivo models of cancer. J Nutr 1995; 125: s777–83..
- 36Growth inhibitors of human prostatic cell lines by phytoestrogens. Proceedings of the Second International Symposium on the Role of Soy in Preventing and Treating Chronic Disease, Belgium, 1996., , .
- 37Genestein inhibition in prostate cancer cell growth and metastasis in vivo. Proceedings of the Second International Symposium on the Role of Soy in Preventing and Treating Chronic Disease, Belgium, 1996., , , .
- 40A biometric study of basal metabolism in man [no. 279]. Washington, DC: JB Lippincott, The Carnegie Institution of Washington Publication, 1919., .
- 43The free/total prostate-specific antigen test: how should it be used in 1997? Urol Int 1997; 4(3): 18–9..
- 49Effects of soy-protein supplementation on epithelial proliferation in the histologically normal human breast. Am J Clin Nutr 1998; 68(6 Suppl ): S1431–5., , , , , , et al.
- 53The specific role of genistein in reducing hormonal and proliferative risk parameters in prostate cancer. Proceedings of the Fourth Annual Symposium on Predictive Oncology and Therapy. International Society for Preventive Oncology, 2000., , , , , .
- 61Studies of cancer in migrant populations. IARC Sci (Lyon) 1993; 123: 1–10..