On the occasion of the 2nd International Workshop „Nutrients & Food Supplements in Europe – Regulatory Issues“ we shortly presented the history of the risk-benefit debate of isoflavones. Paolo Vergano (FratiniVergano European Lawyers, Brussels) then presented the available legal options in the context of international trade.

The benefits and safety aspects of soy products containing isoflavones are currently discussed on the level of the EFSA. Despite good clinical data demonstrating the hormonal safety and clinical effects of isoflavones a negative opinion may be anticipated. The opinions of the EFSA and the composition of the working group can hardly be challenged at the European Court of Justice, as EFSA is officially not taking decisions, although the negative opinions do have an immediate economic impact in the form of de facto trade bans. Challenging the decision of the EU Commission is possible, but rather lengthy and with questionable chances of success. In cases where international trading is compromised by a negative EFSA opinion the WTO dispute settlement system may provide a more promising way forward within a reasonable period of time.


Isoflavones, secondary plant metabolites naturally occurring in food sources such as soy or red clover, are currently under discussion in two working groups of the EFSA: whereas the NDA panel is examining the wording of submitted health claims, a working group of the AMU panel is looking into potential risks related to the intake of isoflavones.
Nutritional intake of isoflavones through soy food is related to a large number of health benefits: high consumption of soy is associated with less menopausal symptoms such as hot flushes, less osteoporosis, cardiovascular disease or decline of cognitive functions in higher age. Soy and the isoflavones contained therein are also associated with a prevention of hormone-dependent cancer such as uterus or breast cancer (McCarty 2006). Isoflavones are generally referred to as “phyto-estrogens”, as their mechanism of action is binding to and activation of the estrogen-beta receptor (ER-β). However, the denomination is misleading, as isoflavones do not act like estrogen. Under the conditions encountered in the human organism they lack the capability of activating the estrogen receptor alpha (ER-α). Whereas the hormone itself triggers important hormonal signalling such as the proliferation of hormone-dependent breast or uterine tissues via activation of ER-α, the function of the beta-receptor can be interpreted as that of a guardian against overshooting estrogenic effects (Heldring et al. 2007).

The activation of ER-β through isoflavones is probably also the most important mechanism of action for the application of soy extracts for the alleviation of menopausal symptoms as an alternative to or nutritional support of hormone replacement therapy (HRT). However, with the results of the Women’s Health Initiative (WHI study) doubts were cast on the safety of hormone substitution in menopause. In hormone-substituted women the study authors reported an increase of the relative risk of stroke by 41 %, of myocardial infarction by 29 %, of thrombosis by 50 % and of breast cancer by 26 % (Rossouw et al. 2002). Consequently, all kinds of estrogenic compounds came under a general suspicion, including the isoflavones, although it should have been clear that the reported adverse effects would be ER-α and not ER-β mediated. The suspicion of risks related to isoflavone intake were seemingly confirmed when soy preparations and isolated isoflavones were submitted to testing in models of cancer promoting hormonal effects (Helferich et al. 2008).

However, the corresponding models are considered irrelevant for the situation in the human body, as the test conditions are highly artificial and involve, among other factors, the complete absence of a working immune system, or circulating estrogen and of estrogen-beta receptors in the transplanted human cancer cells (Messina and Wu 2009). In addition, the model in rodents allows for blood levels of free isoflavones by far exceeding the levels possible achievable in humans even after extremely elevated doses of isoflavones (Setchell 2009). In fact, already small changes in the experimental conditions could be shown to change the outcome from cancer-promoting to cancer-preventive effects (Zhou et al. 2004).

Clinical safety of isoflavones

Hints to potential adverse effects must be taken seriously. However, rules of risk assessment also require a correlation of effects derived from experiments in animals or in vitro studies with clinical observations. This may not be possible with newly developed drugs, but with the very high number of publications on human application of isoflavones this group of natural compounds cannot be considered new. Correspondingly, hypothetical adverse effects cannot simply be extrapolated from animal models – the risk should at least be reflected in the documented use in hundreds of thousands of women exposed to soy, soy extracts and isolated isoflavones. To date the clinical trials and the safety parameters assessed therein do not confirm a hormonal risk of isoflavones, rather the contrary: the studies point to significant protective effects against the formation of hormone-dependent cancer (Marini et al. 2008; Maskarinec et al. 2009; Schmidt 2009; Wu et al. 2008). The only study which was interpreted as giving some evidence of endometrial hyperplasia (Unfer et al. 2004) was criticized for various reasons. One of the major flaws was the observation of some cases of endometrial hyperplasia in the isoflavone group, but none in the placebo group. With the known spontaneous rates of endometrial hyperplasia the complete absence of such events in menopausal women after 5 years of placebo treatment must be considered a medical anomaly (Legault et al. 1998; Messina 2009). In fact, the data published in at least 24 studies with the application of daily doses of 35 – 132 mg of isoflavones for a duration of up to 3 years (Schmidt 2009) points to protective effects of isoflavones with respect to endometrial cancer (Messina 2008).

Official safety assessments of isoflavones

Immediately after the publication of the results of the WHI study related to hormone therapy, the safety of isoflavones as “phyto-estrogens” was officially assessed by the UK Committee of Toxicology. In 2003, the COT published a thorough analysis named “Phytoestrogens and health”, where it was pointed out that the animal models show conflicting results, whereas the human data points to cancer-protective effects. The COT called for more and better human data, which is meanwhile available.

The COT assessment was followed by an evaluation of the French AFSSAPS (French Federal Agency for the Safety of Health Products) in 2005, published under the title “Security and benefits of dietary phytoestrogens: Recommendations”. This very detailed assessment found isoflavones safe up to a dose of at least 60 mg/day, whereas for higher doses the agency said it had insufficient data for conclusions.

In November 2006, the German senate commission for food (SKLM) analyzed “Isoflavones as phytoestrogens in food supplements and dietary foods for special medical purposes”. With this analysis the basic question was reformulated. Whereas the previous assessments focussed on the question whether there are indications that isoflavones might bear a risk, the question was now whether there is sufficient evidence to exclude a (hypothetical) risk. From the scientific point of view this is an important difference: whereas science can prove the existence of an effect, it can never definitively demonstrate its absence.

The SKLM stated that the potential for negative effects has not been sufficiently examined, since isoflavones might theoretically share the cancer-promoting risk of estrogen. This hypothesis was immediately picked up by the German Federal Institute for Risk Assessment (BfR), which in April 2007 published that “Isolated isoflavones are not free of risk”. This verdict was mainly based on the assumption of an effect of isoflavones at the ER-α. The BfR pointed out that based on extrapolations from experimental data breast-altering effects of isoflavones cannot be excluded. BfR did, however, not give clinical evidence to the allegation of cancer-promoting effects – with the exception of the already mention endometrial study (Unfer et al. 2004).

When confronted with the numerous factual errors in the assessment, the BfR corrected the statement of ER-α mediated isoflavone effects to ER-β in a new version of November 2007. However, the BfR failed to draw the logical conclusions from this correction: the arguments defending the hypothesis of a risk are still based on the assumption of ER-α-mediated effects, and on the extrapolation from animal models based on ER-α-induced cancer growth. Many of the errors in the text are still uncorrected.

The debates of the BfR statement finally led to a “scientific discussion” on invitation of the BfR – a remarkably single-sided discussion which did not follow the usual scientific standards. The selection of invited speakers was made by the BfR, and practically only adversaries of soy (mainly those who created the animal data) were given the right to speak. Since there was no real discussion, the BfR obtained from this event “confirmation” of its own point of view, and has since then not stopped informing the media of the dangers of soy consumption. Still the scientific discussion is not that easy to quench. Thus, several symposia on safety and benefits of soy with real scientific position statements and discussions have been held since then, with the most important symposium taking place in Milan in May 2009, organized by the Council for Responsible Nutrition (CRN) and held in the presence of representatives of the EFSA (Messina et al. 2009; Schmidt 2009). In these symposia, benefits and potential risks were likewise discussed in the context of experimental and clinical research. It was concluded that based on the available data isoflavones in quantities corresponding to the usual daily intake in Asian countries must be considered safe and beneficial for human health.

Assessment of safety and health claims by the EFSA

In the meantime the question of soy and isoflavone safety was passed on to the European level by giving the EFSA the mandate to assess the safety of isoflavones on request of “some EU Member States” – most likely under the leadership of the German BfR, which had announced such a step in its public statements. A working group was formed with the task of evaluating the available scientific data. The decision of the working group was originally announced for December 2009, but was recently postponed to December 2010. One of the group members (who according to the minutes has meanwhile no longer an active part) was Klaus Richter from the German BfR, and thus one of the persons responsible for the incriminating BfR statement. Under such conditions it is difficult to see how an impartial assessment could be reached.

Can EFSA opinions be challenged?

A negative opinion of the EFSA will have an immediate impact on the market of soy and soy derived products. The question therefore arises whether an EFSA opinion on safety and benefits of isoflavones could be challenged. However, an immediate legal action at the EC courts to challenge a negative EFSA opinion on isoflavones would not seem to have many chances of success. There have already been three court cases against EFSA opinions at the EU General Court, two of them were claims for damages for the loss allegedly sustained as a result of the adoption of the respective EFSA opinion. The applications were declared inadmissible by the General Court (Orders of 17 June 2008, case numbers T-312/06, T-397/06 and T-311/06), as the EFSA does only provide opinions, but not legally binding decisions. Although the General Court admitted that the EU commission usually adopts its decision after obtaining the opinion of the EFSA, there is nothing in the law to suggest that the EU commission is obliged to comply with EFSA opinions in substantive terms. As the EU Commission has in theory the discretion to disregard EFSA opinions, these can hardly be challenged. Only the EU Commission decision may be regarded as the final stage of the procedure.

Decisions of the EU (i.e. Commission Regulations on the authorisation or rejection of health claims made on food) may be instituted within two months of the publication of the measure at the General Court, with possible action for interim measures and action for damages. The way through the instances will take many years and will therefore probably be unaffordable for small and medium size enterprises.

Can EFSA expert appointments be challenged?

In relation to the safety of isoflavones, EFSA has decided that preparatory work is needed before the task can be assigned to the competent EFSA panel – hence the formation of a working group, which is requested to provide a report characterizing the potential hazards and health benefits associated with isoflavone consumption to the Executive Director by the end of 2010. However, it seems that the working group features members which have already decided that isoflavones are a major risk factor. It is unclear how these experts may come to an independent and politically unbiased conclusion.

Challenging the composition of the working group exclusively based on previous statements in the topic has, however, not many chances of success. The declaration of interests, which has to be signed by the experts, states that “high quality of scientific expertise is by nature based on prior experience and that, therefore, having an interest does not necessarily mean having a conflict of interest”. In the past, EFSA’s GMO panel and its members have been subject to criticism because of alleged conflicts of interests in the opposite sense: they were accused of “rubber-stamping anything the agro-biotech industry puts forward, with the blessing of the European Commission”. Correspondingly, there is a justified interest in not having only positive voices in the panel and the working groups. Even though the composition of the working group cannot be directly challenged, a petition to the EU Parliament can be made pointing out that the selection of experts may be biased.

The WTO dispute settlement procedure

A potential negative decision of the EFSA and the EU Commission regarding isoflavone safety would lead to a de facto trade ban of isoflavone-containing preparations. Such a decision would, however, most likely collide with WTO trade rules, especially since the major producers of soy extracts are seated outside the EU. Generally, trade barriers can only be established when a clear consumer risk from the use of the corresponding product is demonstrated. Producers from non EU member states economically suffering from such trade restrictions can, in cooperation with their governments, claim a violation of WTO trade rules. In the case of soy extracts this would essentially be the case for extract producers from, for example, the United States of Israel. WTO disputed may be brought only by WTO member states and not by private parties.

In comparison to an appeal to the EU Court of Justice, the WTO dispute settlement system has the advantage of being based on a set of procedures that require decisions to be taken within a predetermined period of time. As a general rule, no more than 9 months should elapse between the establishment of the panel and the adoption of a panel report. If the report is appealed, the total timeframe indicated by WTO rules is 12 months. The actual average time frame of a WTO dispute at the panel stage is around 14 months.

Using WTO dispute settlement is not unusual and stands good chances of success within a reasonable time frame. The system has already successfully been used in similar scientific matters related to questions of safety. With the opinions of the EFSA having immediate economic effects without a realistic possibility of being challenged based on EU legislation, the WTO dispute settlement might in fact be the most promising way forward.

Download: Presentation slides: Soy Isoflavones - Current status, pittfalls and opportunities. 2nd International Workshop Nutrients & Food Supplements in Europe – Regulatory Issues, Brussels, 3rd of December 2009.


Heldring, N., Pike, A., Andersson, S., Matthews, J., Cheng, G., Hartman, J., Tujague, M., Strom, A., Treuter, E., Warner, M., and Gustafsson, J. A. (2007). Estrogen receptors: how do they signal and what are their targets. Physiol Rev. 87 (3): 905-931.

Helferich, W. G., Andrade, J. E., and Hoagland, M. S. (2008). Phytoestrogens and breast cancer: a complex story. Inflammopharmacology. 16 (5): 219-226.

Legault, C., Espeland, M. A., Wasilauskas, C. H., Bush, T. L., Trabal, J., Judd, H. L., Johnson, S. R., and Greendale, G. A. (1998). Agreement in assessing endometrial pathology: the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. J. Womens Health 7 (4): 435-442.

Marini, H., Bitto, A., Altavilla, D., Burnett, B. P., Polito, F., Di Stefano, V, Minutoli, L., Atteritano, M., Levy, R. M., D'Anna, R., Frisina, N., Mazzaferro, S., Cancellieri, F., Cannata, M. L., Corrado, F., Frisina, A., Adamo, V., Lubrano, C., Sansotta, C., Marini, R., Adamo, E. B., and Squadrito, F. (2008). Breast Safety and efficacy of genistein aglycone for post-menopausal bone loss: A follow-up study. J. Clin Endocrinol. Metab 93 (12): 7487-7496.

Maskarinec, G., Verheus, M., Steinberg, F. M., Amato, P., Cramer, M. K., Lewis, R. D., Murray, M. J., Young, R. L., and Wong, W. W. (2009). Various doses of soy isoflavones do not modify mammographic density in postmenopausal women. J. Nutr. 139: 981-986.

McCarty, M. F. (2006). Isoflavones made simple - Genistein's agonist activity for the beta-type estrogen receptor mediates their health benefits. Med. Hypotheses 66 (6): 1093-1114.

Messina, M. (2008). The endometrial effects of isoflavones: a discussion paper. Complement Ther. Clin Pract. 14 (3): 212-214.

Messina, M. (2009). Effects of isoflavones on endometrial tissue and endometrial cancer risk. Symposium on Evaluating the Efficacy and Safety of Isoflavones for Postmenopausal Women, 13-14 May. Milan (Italy): Council for Responsible Nutrition.

Messina, M., Watanabe, S., and Setchell, K. D. (2009). Report on the 8th International Symposium on the Role of Soy in Health Promotion and Chronic Disease Prevention and Treatment. J. Nutr. 139 (4): 796S-802S.

Messina, M. and Wu, A. H. (2009). Perspectives on the soy-breast cancer relation. Am. J. Clin. Nutr. 89 (5): 1673S-1679S.

Rossouw, J. E., Anderson, G. L., Prentice, R. L., LaCroix, A. Z., Kooperberg, C., Stefanick, M. L., Jackson, R. D., Beresford, S. A., Howard, B. V., Johnson, K. C., Kotchen, J. M., and Ockene, J. (2002). Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women's Health Initiative randomized controlled trial. JAMA 288 (3): 321-333.

Schmidt, M. (2009). Soja-Isoflavone bei Wechseljahresbeschwerden. Welchen Nutzen, welche Risiken bergen Pflanzenestrogene? Gyne 30 (12): 25-34.

Setchell, K.D. (2009). Overview of isoflavone physiology and pharmacokinetics. Symposium on Evaluating the Efficacy and Safety of Isoflavones for Postmenopausal Women, 13-14 May. Milan (Italy): Council for Responsible Nutrition.

Unfer, V., Casini, M. L., Costabile, L., Mignosa, M., Gerli, S., and Di Renzo, G. C. (2004). Endometrial effects of long-term treatment with phytoestrogens: a randomized, double-blind, placebo-controlled study. Fertil. Steril. 82 (1): 145-148.

Wu, A. H., Yu, M. C., Tseng, C. C., and Pike, M. C. (2008). Epidemiology of soy exposures and breast cancer risk. Br. J. Cancer 98 (1): 9-14.

Zhou, J. R., Yu, L., Mai, Z., and Blackburn, G. L. (2004). Combined inhibition of estrogen-dependent human breast carcinoma by soy and tea bioactive components in mice. Int. J. Cancer 108 (1): 8-14.