Risk assessment for peri- and post-menopausal women taking food supplements containing isolated isoflavones

Tabs

Article
Panel on Food Additives and Nutrient Sources Added to Food
EFSA Journal
EFSA Journal 2015;13(10):4246 [342 pp.].
doi
10.2903/j.efsa.2015.4246
Panel members at the time of adoption
Fernando Aguilar, Riccardo Crebelli, Alessandro Di Domenico, Birgit Dusemund, Maria Jose Frutos, Pierre Galtier, David Gott, Ursula Gundert-Remy, Claude Lambré, Jean-Charles Leblanc, Oliver Lindtner, Peter Moldeus, Alicja Mortensen, Pasquale Mosesso, Dominique Parent-Massin, Agneta Oskarsson, Ivan Stankovic, Ine Waalkens-Berendsen, Rudolf Antonius Woutersen, Matthew Wright and Maged Younes.
Acknowledgements

The Panel wishes to thank the members of the Working Group on Isoflavones: Catherine Bennetau-Pellissero (since December 2014), Cristina Bosetti, Daniel Doerge, David Gott, Julie Glanville, Ursula Gundert-Remy, Sabine Kulling, Gunter Georg Kuhnle, Claude Lambré (until December 2014), Alicja Mortensen, Pasquale Mosesso, Agneta Oskarsson, Ferruccio Santini and Rudolf Antonius Woutersen for the preparatory work on this scientific output and EFSA staff members: Elisa Aiassa, Davide Arcella, Fulvio Barizzone, Annette-Cecilia Forss and Camilla Smeraldi for the support provided to this scientific output.

Type
Opinion of the Scientific Committee/Scientific Panel
On request from
Bundesinstitut für Risikobewertung (BfR)
Question Number
EFSA-Q-2013-00916
Adopted
8 September 2015
Published
21 October 2015
Affiliation
European Food Safety Authority (EFSA), Parma, Italy
Note
Download Article (5.45 MB) Download Annex A (1019.85 KB) Download Annex B (2.77 MB)
Abstract

The EFSA ANS Panel was asked to deliver a scientific opinion on the possible association between the intake of isoflavones from food supplements and harmful effects on mammary gland, uterus and thyroid in peri- and post-menopausal women. Isoflavones are naturally occurring substances which can be found in, among other sources, soy, red clover and kudzu root. The main isoflavones are genistein, daidzein, glycitein, formononetin, biochanin A and puerarin. Their chemical structures are related to 17β-oestradiol and they possess oestrogenic properties. Furthermore, isoflavones may interact with the synthesis of thyroid hormone. Food supplements targeted at peri- and post-menopausal women typically provide a daily dose of isoflavones in the range of 35–150 mg/day. A systematic review was performed to investigate whether an association could be found between intake of isoflavones from food supplements and adverse effects on the three target organs in peri- and post-menopausal women. The human data did not support the hypothesis of an increased risk of breast cancer from observational studies nor of an effect on mammographic density nor on proliferation marker Ki-67 expression in interventional studies. No effect was found on endometrial thickness and histopathological changes in the uterus up to 30 months of supplementation with 150 mg/day of soy isoflavones. After 60 months some non-malignant histopathological changes were reported. Thyroid hormones levels were not changed following intake of isoflavones from food supplements. The background exposure from the diet in the general European population was estimated to be lower than 1 mg/day, whereas in consumers of soy-based foods it could be higher. The Panel concluded that it was not possible to derive a single health-based guidance value for the different preparations in post-menopausal women. However the doses used in the intervention studies and their duration could serve as guidance for the intake of food supplements.

Summary

In accordance with Article 29 (1) of Regulation (EC) No 178/2002, the Bundesinstitut für Risikobewertung (BfR) asks the European Food Safety Authority to provide a scientific opinion on the possible health risks associated with the intake of isolated isoflavones in food supplements by peri- and post-menopausal women.

Isoflavones are a class of naturally occurring substances, present in a number of plants, especially in soybeans, red clover and kudzu root. Isolated isoflavones used in dietary supplements are defined as extracts from soybeans containing a mixture of predominantly glycosylated genistein, daidzein and glycitein and isolated forms of these soy-based compounds and extracts from red clover containing a mixture of predominantly glycosylated formononetin and biochanin A. Kudzu essentially contains the oestrogenic glycosides of daidzein and genistein in a ratio of two-thirds to one-third, usually as well as puerarin.

The scientific opinion requested by the BfR should address the relevant available scientific evidence on the potential adverse effects associated with intake of isolated isoflavones in food supplements by peri- and post-menopausal women, including data from both human and animal studies. The scientific opinion should focus on possible harmful effects on mammary gland, uterus and thyroid. If adverse effects are identified, it should provide indication of the underlying potential modes of action. The scientific opinion should also provide an estimate of exposure of the target populations to the isolated isoflavones from food supplements and, if possible, give advice on a safe intake level of isolated isoflavones from dietary supplements.

Thus, the assessment is a focused assessment concerning the population of interest (peri- and post-menopausal women) as well as the target organs of interest (breast, uterus, thyroid).

The isoflavones (and some of their unconjugated metabolites) considered relevant for this risk assessment were daidzein, genistein, glycitein, biochanin A and formononetin, and their glycosides daidzin, genistin, glycitin and puerarin. The potential hazards of concern for this assessment are those expected from the interaction of isoflavones with endocrine pathways. Interactions are known with the oestrogen receptors ERα and ERβ. Apart from the inhibition of thyroid peroxidase, several other mechanisms have been described from in vitro and animal experiments, which could indicate an effect of isoflavones on thyroid function.

A review of the association between isoflavones intake from food supplements and possible effects on mammary gland, uterus and thyroid was performed in accordance with the principles of systematic review, based on an a priori established protocol.The criteria defined for the selection of the studies limited inclusion to studies conducted in the relevant population of interest and to ovariectomised animals. All the studies included in the systematic review were appraised for validity and risk of bias using pre-defined criteria agreed before the start of the assessment. The literature on kinetics and metabolism of isoflavones was reviewed by a narrative approach. Published literature on genotoxicity, retrieved by a focused literature search, was conducted using the search strings detailed in the protocol.

For the systematic review, a total of 7 841 hits were retrieved. After screening for relevance, 43 human studies and 62 animal studies were retained for this opinion.

For the exposure, background dietary isoflavone exposure was estimated using levels of isoflavones in soy and soy-based products reported in the literature in combination with food consumption data from the EFSA Comprehensive European Food Consumption Database for the group of women older than 40 years of age. Exposure to isoflavones from food supplements in the population of peri- and post-menopausal women was taken from the range of doses used in the intervention studies in the target population included in the systematic review. In addition, information from the relevant industry associations was obtained on the labelled amount of isoflavones in the supplements and the recommended range of daily doses. This information was compared with published data reporting on the measured isoflavone content in the food supplements.

For humans, no data on absolute bioavailability can be given. In mice, absolute bioavailability amounted to 9–14 % for genistein and 29–34 % for daidzein. In rats, absolute bioavailability of genistein was 7 % in males and 15 % in females. In humans, the maximum concentration (Cmax) of genistein aglucone and that of daidzein aglucone varied between 0.4 % and 3.9 % and between 1.4 % and 4.2%, respectively, expressed as percentage of the total isoflavone concentration. Hence, it can be deduced that the absolute bioavailability of the two main isoflavones must be low in humans. At least some data on the metabolism of the soy isoflavones daidzein and genistein are available for humans, monkeys, rats, and mice. The main metabolites are the sulphate and glucuronide conjugates of the substances. Monkeys, rats, and mice are described as 100 % equol producers, meaning that the microbiotas of these animals are uniformly able to transform daidzein to a considerable extent to S-equol. In humans only a part of the population is able to produce S-equol. Oxidative phase I metabolites of daidzein and genistein, mainly 6-hydroxy-, 8-hydroxy- and 3′-hydroxy-daidzein as well as 6-hydroxy-, and 3′-hydroxy-genistein, are found in humans, rats and mice, although as minor metabolites.

For the endpoint breast, four epidemiological studies investigating breast cancer incidence (involving, in total, 2 216 isoflavone users), eight interventional controlled studies, measuring mammographic density (741 participants), and two interventional controlled studies, investigating histopathological changes (75 participants), did not suggest an association between exposure to isoflavones-containing food supplements and adverse effects in the mammary gland. Ten studies in ovariectomised animals were found, investigating breast cell proliferation, and 11 animal studies were identified, investigating histopathological changes in the mammary gland of animals treated with isoflavones. Although in the majority of the studies no effect was noted, a stimulating effect on the mammary gland was observed in two rat studies (genistein 5.4 and 54 mg/kg/day and 221 mg/kg body weight (bw)/day, both studies carried out for 90 days). The findings are consistent with the results from the US National Toxicology Program study conducted in non-ovariectomised animals administered genistein at doses ranging 0.3–44 mg/kg bw/day, in which there was some evidence of carcinogenic activity of genistein in female Sprague–Dawley rats based on an increased incidence of mammary gland adenoma or adenocarcinoma (combined).

For the endpoint uterus, no study was found that investigated the association between intake of isoflavones from food supplements and risk of uterine cancer in the target population. Endometrial thickness was measured in 25 interventional controlled studies (1 484 participants) and histopathological investigations of endometrium were carried out in nine interventional controlled studies (677 participants). None of the studies reported statistically significant changes in endometrial thickness compared with control. In only two studies were some histopathological effects noted. One study was not properly controlled and had further methodological flaws. In the other study, there were no findings after 2.5 years of intervention, whereas after 5 years of intervention only five cases of simple hyperplasia and one case of complex hyperplasia of the endometrium were observed, but no cases of endometrial carcinoma. The findings could indicate an oestrogenic effect. Thirteen studies in animals investigated uterus cell proliferation, among them three in monkeys, and 22 animal studies studied uterus histopathological changes, among them four studies in monkeys. An effect of isoflavones was not seen in most of the studies. However, a daidzein-rich soy extract containing daidzein at doses above 40 mg/kg bw/day, caused an increase in cell proliferation of the epithelium and the stroma of the uterus as well as the vaginal epithelium of rats. Racemic equol (36 mg/kg/day), administered to rats induced an increase in proliferation of stromal cells of the uterus, whereas an equivocal effect was observed on proliferation of the epithelial cells of the uterus. In monkeys and rats, isoflavones obtained from soy extracts or soy protein isolates did not result in significant changes in endometrial thickness, endometrial hyperplasia, epithelial area or endometrial gland area. Daidzein-rich soy extract and red clover extracts at high doses (≥ 125 mg/kg bw/day) caused a significant increase in endometrial area, endometrial thickness, number of glands and myometrial area in rats. Genistein and daidzein did not induce histopathological changes such as hypertrophy, hyperplasia or squamous metaplasia in the uterus. Racemic equol at a dose of 10 mg/kg bw/day or higher resulted in a significant increase in uterine wall thickness when administered for 90 days, but no such changes were observed after 35 days’ administration of the same dose.

Eleven human controlled randomised studies that reported effects of isoflavones administration on some thyroid-related endpoints were identified. In total, 925 subjects were allocated to isoflavones. In none of the studies was a clinically relevant effect on the thyroid detected. Although the studies have some flaws (thyroid function not the primary endpoint, sample size calculation not given, low power to detect changes) the Panel’s conclusion is that administration of food supplements containing isoflavones is not associated with clinically relevant changes in thyroid function (hypo- or hyperthyroidism) in the population of interest.

As regards the genotoxicity of genistein and the two catecholic oxidative metabolites of daidzein, 3’,4’,7-trihydroxyisoflavone and 4’,6,7-trihydroxyisoflavone, genotoxic effects found in vitro in mammalian cells, for which a thresholded mechanism of action has been demonstrated, have not been reproduced in valid in vivo micronucleus tests in rats and mice or in the comet assay and micronucleus test in human studies. Based on these findings, it can be concluded that isoflavones are not genotoxic.

In the human intervention studies included in this assessment, doses of isoflavones ranging 30–900 mg/day were used. On the basis of the published studies included in this assessment and measuring the actual content of isoflavones in a number of food supplements available on the market, it can be estimated that intake of isoflavones from food supplements is extremely variable, ranging approximately 0.1–100 mg/day for soy isoflavones, 30–160 mg/day for isoflavones from red clover and 20–50 mg/day for kudzu root isoflavones (all the values above are expressed as aglycones). The values above are in line with information provided by the relevant food sector business operators, with respect to the recommended daily doses of isoflavones in products mainly targeted at menopausal women. The overall intake of isoflavones from the diet in women (0.27–1.43 mg/day) is lower than the lowest recommended daily intake of isoflavones from food supplements (20–35 mg/day), although in women with special dietary habits (e.g. soy consumers and vegetarians) the intake of isoflavones from the diet could be within the range of exposure from food supplements. Consumption of certain soy-based food products (e.g. soy drink, soy yoghurt, tofu) alone in a single day might contribute to the intake of isoflavones of the same order of magnitude as estimated for food supplements.

In this opinion, data from human and animal studies have been assessed; however, the Panel identified a number of limitations which could introduce uncertainties in the evaluation. Owing to the nature of the uncertainties identified, it was not always possible to state in which direction they might have influenced the conclusions.

The Panel noted that the conclusions drawn in this opinion are based on the assumption that intake of isoflavones from the use of food supplements represents the major contribution to the intake of isoflavones. In line with the Terms of Reference, this risk assessment was focused on three target organs, mammary gland, uterus and thyroid, in a sub-group of the general European population: peri- and post-menopausal women. The assessment was limited to isoflavones ingested as food supplements at doses used in the human studies available in the published scientific literature and following a request for information from relevant interested parties.

The Panel considered that results obtained from the human studies and the studies in ovariectomised animals were most relevant for the target population. Furthermore, the Panel noted that, in addition to all the limitations described for the relevant studies, studies in other human populations (e.g. males) or animal models (e.g. juvenile animals, transgenic models) would not provide additional relevant information on the specific risks being assessed in peri- and post-menopausal women. Therefore, the result from this assessment cannot be extrapolated to other groups and other situations in the general population.

Based on the data reviewed and presented in the current opinion, and taking into account the uncertainties described above, the Panel reached the following conclusions:

Overall, the Panel concluded that it was not possible to derive a single health-based guidance value or a safe intake level for food supplements containing isoflavones. The doses and duration of treatment with the individual preparations used in the interventional studies may serve as guidance for a dose and duration of use at which no effect has been observed in all three target organs in the evidence considered for this opinion. The Panel noted that recommended daily doses of the marketed food supplements, with the exception of one product containing soy isoflavones and one product containing red clover, seem to fall within these ranges, albeit the food supplements do not bear any clear indication with respect to the recommended duration of use. For products containing kudzu root, no data were available which could guide their use in post-menopausal women.

The proposed values are applicable only to post-menopausal women without a current diagnosis or history of oestrogen-dependent breast or uterine cancer.

The Panel noted that the conclusions drawn in this opinion are based on the assumption that intake of isoflavones from the use of food supplements represents the major contribution to the intake of isoflavones, as has been the situation in the human interventional studies.

The Panel considered that more data on the doses and duration of consumption should be generated as this would improve the available database on the safety of prolonged use of food supplements containing isoflavones.

This assessment identified the need for a harmonised way of reporting the isoflavone content of food supplements.

Future studies should use a standardised description of the isoflavones and should explicitly state whether or not the content of isoflavones is expressed as aglycones and should report the ratio of the individual isoflavones.

  1. An assessment could be provided only for human studies in the relevant population of peri- and post-menopausal women or from animal studies in OVX animals investigating the relevant pre-defined endpoints in mammary gland, uterus and thyroid. Details of the isoflavone composition of the preparations tested in the human and animal studies are given in the opinion.
  2. In assessing the effects of isoflavones on the three target organs, the Panel decided that differences in functions, receptor density, proportions of ERα and ERβ and effects of receptor activation meant that it was not possible to directly extrapolate observations from any one organ to the others. The Panel noted differences in biological effects and activity between isoflavones from different sources and in different organs and, therefore, concluded that currently it is not generally feasible to apply a read-across approach either between different preparations or between similar preparations in different organs. Hence, a full evaluation is possible only if study results are available for all three target organs.
  3. There is overlap between peri- and post-menopause. The World Health Organization defines ‘perimenopause’ as the period immediately prior to the menopause (when the endocrinological, biological and clinical features of approaching menopause commence) and the first year after menopause and ‘postmenopause’ as the period dating from the final menstrual period. Although women progress through peri-menopause to post-menopause it may not be possible to definitively categorise them as peri- or post-menopausal. Only a small proportion of the participants included in the interventional studies would be classified as peri-menopausal women according to the definition above. Despite the uncertainties and limitations described, and given the overlap in the definitions, the Panel considered that the data on mammary gland and thyroid allow conclusions that are applicable to post- and peri-menopausal women.
    With respect to the data on uterus, the Panel considered that the database is not sufficient to draw conclusions on peri-menopausal women.
    Because not all three target organs were covered by the intervention studies in the peri-menopausal population, the overall conclusions of this opinion apply only to post-menopausal women.
  4. For the target organ mammary gland, three case–control studies and one prospective cohort study did not support the hypothesis of an increased risk of breast cancer associated with the intake of isoflavones from food supplements. The Panel acknowledged that the central tendency was around 1, consistently across all the studies included in the review, and the upper limit of the confidence interval for the estimated odds ratio was always below 1.67.
    Based on interventional trials encompassing 816 women, the Panel concluded that neither enhanced breast density (741 women) nor histopathological changes (75 women) were observed for soy isoflavones/soy extracts, soy protein, daidzein-rich isoflavones, genistein and red clover extract. The Panel concluded that, on the basis of the evidence reviewed, there is no indication for adverse effects on the mammary gland in post-menopausal women from isoflavones when taken in doses and for durations as described above.
    The information on women with breast cancer obtained from this systematic review is limited; therefore, the opinion cannot conclude on the risk of oestrogenic isoflavones-based food-supplements in postmenopausal women with a current diagnosis or history of oestrogen-dependent cancer.
  5. For the target organ uterus, neither changes in endometrial thickness (studies involving, in total, 1 484 participants) nor remarkable histo(patho)logical findings (studies encompassing 677 participants) were observed in any of the human interventional studies, with the highest isoflavone dose being 150 mg/day administered for a period of 2.5 years. An effect on uterine weight was found in rats with doses of various isoflavones of between 10 mg/kg body weight (bw)/day and 100 mg/kg bw/day. The Panel considered that this was not an adverse effect.
    On the basis of the evidence from human studies and the considerations on the findings from the animal studies, the Panel concluded that no adverse effects on the uterus were noted for soy isoflavones/soy extract, soy protein, daidzein-rich isoflavones, glycitein-rich isoflavones, genistein and red clover extract in post-menopausal women when taken in doses and for durations as described above.
    No information on women with uterine cancer was obtained from this systematic review; therefore, the Panel cannot conclude on the risk of oestrogenic isoflavones-based food-supplements in post-menopausal women with current diagnosis or a history of oestrogen-dependent cancer.
  6. The assessment of effects on the thyroid function was exclusively based on the results from human interventional studies. Based on these studies (involving 925 participants taking isoflavones and 576 serving as controls), the Panel concluded that there are no statistically significant changes to indicate that food supplements containing soy isoflavones/soy extract, soy protein, daidzein-rich isoflavones, genistein or red clover extract exert a hypothyroid effect in post-menopausal women with normal thyroid function.
  7. Human studies investigating effects on the three target organs of intake of food supplements containing extracts from kudzu root and the isoflavones genistin, daidzin, daidzein, glycitin, glycitein, puerarin, biochanin A and formononetin were not retrieved. No animal studies other than those evaluating the effect of isoflavones on uterine weight were available. Taken together, these findings preclude an assessment of these substances and mixtures.
  8. For genotoxicity, positive findings expressed in vitro in mammalian cells by the two catecholic oxidative metabolites of daidzein 3’,4’,7-trihydroxyisoflavone and 4’,6,7-trihydroxyisoflavone and by genistein through the stabilisation of the ‘cleavable complex’ and generation of DNA double-strand breaks at topoisomerase II–DNA binding sites, for which a thresholded mechanism of action has been demonstrated, have not been reproduced in valid in vivo micronucleus tests in rats and mice and in comet assay and micronucleus test in human studies. On these bases, the use of isoflavones in food supplements is not of genotoxic concern.
  9. A comparison of the estimated intake of isoflavones from food supplements with estimates of intake based on food consumption data showed that the levels of daily intake of soy isoflavones from food supplements may be achieved by consumers of specific soy foods, such as tofu, soy yoghurt, soy milk and drinks.

Overall, the Panel concluded that it was not possible to derive a single health-based guidance value or a safe intake level for food supplements containing isoflavones. The doses and duration of treatment with the individual preparations used in the interventional studies may serve as guidance for a dose and duration of use at which no effect has been observed in all three target organs in the evidence considered for this opinion. The Panel noted that recommended daily doses of the marketed food supplements, with the exception of one product containing soy isoflavones and one product containing red clover, seem to fall within these ranges, albeit the food supplements do not bear any clear indication with respect to the recommended duration of use. For products containing kudzu root, no data were available which could guide their use in post-menopausal women.

The proposed values are applicable only to post-menopausal women without a current diagnosis or history of oestrogen-dependent breast or uterine cancer.

The Panel noted that the conclusions drawn in this opinion are based on the assumption that intake of isoflavones from the use of food supplements represents the major contribution to the intake of isoflavones, as has been the situation in the human interventional studies.

The Panel considered that more data on the doses and duration of consumption should be generated as this would improve the available database on the safety of prolonged use of food supplements containing isoflavones.

This assessment identified the need for a harmonised way of reporting the isoflavone content of food supplements.

Future studies should use a standardised description of the isoflavones and should explicitly state whether or not the content of isoflavones is expressed as aglycones and should report the ratio of the individual isoflavones.

Keywords
isoflavones, phytoestrogens, food supplements, mammary gland, uterus, thyroid, systematic review
Print on demand
Number of Pages
342