Framework comparisons are abstract. Following one ingredient through three regulators is not.
Soy isoflavone is the ideal test case. All three jurisdictions looked at it. Two reached opposite conclusions on the same endpoint, and the third didn’t rule at all.
What the EU concluded
EFSA assessed soy isoflavones under Article 13(1) of the EU health claims regulation, and the verdicts were comprehensive.
Menopausal symptoms — rejected. The Panel concluded that a cause-and-effect relationship had not been established between soy isoflavones and reduction of vasomotor symptoms associated with menopause[3].
Bone mineral density — rejected. Same conclusion: evidence insufficient to establish cause and effect[4].
The reasoning is worth seeing, because it explains why “some studies show benefit” wasn’t enough.
For menopausal symptoms, 13 randomised controlled trials were submitted. Six showed statistically significant reductions in hot flushes. Six did not. One didn’t report hot flush results[1].
And of the six positive trials, the Panel found that most were at high risk of bias due to major methodological weaknesses in the statistical analyses — inadequate handling of missing data among them[1].
For bone, 14 long-term studies of more than 12 months were evaluated. Only two showed an effect, at doses of 54 mg per day[2]. The Panel acknowledged that other studies provided some evidence for attenuating bone density loss, but judged the sum of evidence weak[1].
Germany and Austria subsequently requested further assessment. EFSA reviewed the additional information and reached the same conclusion[5].
What Korea concluded
Here is the part that stopped me.
Soy isoflavone is a recognised functional ingredient in Korea — classified under joint and bone health.
Not menopause. Bone.
Which means Korea recognised the ingredient for precisely the endpoint the EU examined and rejected. Fourteen studies, two positive, verdict insufficient in Europe. Recognised in Korea.
Two regulators, one ingredient, one endpoint, opposite conclusions.
I want to be careful here. This doesn’t establish that either is wrong. Reviews were conducted at different times, against different evidence standards, with different bodies of submitted data. Regulatory judgement involves deciding how much evidence is enough, and reasonable panels can set that threshold differently.
But it does dispose of the idea that approval status tracks the underlying science in any simple way.
What the US does
The US doesn’t appear in this comparison the way the other two do, because it didn’t reach a conclusion.
Under DSHEA, structure/function claims require no pre-market review. The manufacturer holds substantiation; notification goes to the FDA after marketing begins. So a US label making a menopause-related structure/function claim reflects a company’s own assessment, accompanied by the mandatory disclaimer that the FDA has not evaluated the statement.
There’s no US ruling to compare against EFSA’s, because no ruling was required.
The cholesterol precedent
One more data point shows this isn’t unique to menopause.
EFSA also rejected soy protein and isoflavones for maintenance of healthy cholesterol levels — despite similar claims being approved in more than ten countries, including the UK, Japan and the US[1].
Ten-plus jurisdictions said yes. EFSA said no. Same ingredient, same endpoint, same broad literature.
That’s the clearest available demonstration that an approved claim tells you what a regulator decided, not what the evidence unambiguously shows.
| Menopausal symptoms | Bone density | Review type | |
| EU (EFSA) | Rejected | Rejected | Pre-market scientific assessment |
| Korea (MFDS) | Not the recognised function | Recognised | Pre-market ingredient recognition |
| US (FDA) | No ruling | No ruling | No pre-market review |
[Table 1] Soy isoflavone across three jurisdictions · Source: Determination by each regulatory authority[1][3][4]
Why they diverge
Three mechanisms account for most of it.
Different questions. EFSA assesses whether a specific claim wording is substantiated. Korea recognises an ingredient for a defined function. The US asks only whether a manufacturer holds substantiation it never submits.
Different thresholds. How many positive trials, of what quality, constitute sufficient evidence? EFSA’s threshold in the Article 13.1 process proved high enough that most submitted claims failed — as we saw in an earlier post, the large majority of the 2,758 claims assessed did not survive.
Different timing. Reviews conducted years apart draw on different literature. A claim assessed in 2011 and one assessed later aren’t looking at the same evidence base.
Note also what EFSA’s process wasn’t asking. Rejection under Article 13.1 doesn’t mean an ingredient was found harmful, or that no effect exists. It means the submitted dossier didn’t establish cause and effect to that Panel’s standard.
One safety note, separately
Distinct from claims, EFSA’s ANS Panel was asked to assess whether isoflavone supplements pose risks to the mammary gland, uterus and thyroid in peri- and post-menopausal women[6]. Isoflavones possess oestrogenic properties, and may interact with thyroid hormone synthesis[6].
That assessment is a separate question from efficacy, and one worth raising with a clinician if you have thyroid or hormone-sensitive conditions — regardless of what any jurisdiction concluded about claims.
What to do with this
Practical conclusions for anyone buying across borders.
Check the claim against the market it came from. An approved claim carries authority only within the system that issued it. A Korean recognition means nothing in the EU, and an EU authorisation means nothing here.
A rejection isn’t a safety finding. EFSA rejecting a claim means the evidence didn’t meet a threshold, not that the ingredient is dangerous or useless.
An approval isn’t proof either. As the cholesterol case shows, ten jurisdictions can approve what one rejects.
The absence of a US ruling is itself information. No pre-market review means no verdict exists — the label reflects a company’s position.
Look up the wording where you are. For Korea, the recognised function and its exact wording are searchable on the official food safety portal[7].
Closing
I came into this expecting the three systems to differ mainly in strictness — a spectrum from permissive to demanding.
The bone finding doesn’t fit that shape. Korea recognises soy isoflavone for bone health; the EU examined bone density specifically and rejected it. Not a difference in how strict, but a difference in the answer.
And the cholesterol precedent removes any comfort that consensus resolves it. More than ten countries approved what EFSA turned down.
So the honest reading of an approved claim is narrower than it looks. It tells you a particular body, at a particular time, judged a particular dossier sufficient. That’s real information. It isn’t the same as knowing what the ingredient does.
At a Glance
- EFSA rejected soy isoflavones for both menopausal symptoms and bone mineral density
- Menopause dossier: 13 RCTs — 6 positive, 6 null, 1 not reported; most positive trials at high risk of bias
- Bone dossier: 14 long-term studies, only 2 positive at 54 mg/day
- Germany and Austria requested further assessment; the conclusion held
- Korea recognises soy isoflavone under joint and bone health — the endpoint the EU rejected
- The US issues no ruling — structure/function claims require no pre-market review
- EFSA also rejected soy for cholesterol despite approvals in 10+ countries including the UK, Japan and US
- A rejection is not a safety finding, and an approval is not proof
※ This article compares regulatory decisions and is for general information only. It does not recommend or discourage any ingredient or product, and does not replace medical advice. Isoflavones have oestrogenic properties and may interact with thyroid function — if you have a hormone-sensitive condition, thyroid disease, or take related medication, consult a clinician before use. Approved claims and ingredient lists are revised periodically; confirm current wording on official sources.
References
- 「Soy sector “surprised” by EFSA menopause and bone health rejection」, NutraIngredients (trial breakdown, risk of bias, cholesterol precedent), https://www.nutraingredients.com/Article/2012/08/10/Soy-sector-surprised-by-EFSA-menopause-and-bone-health-rejection/
- 「The potential health effects of dietary phytoestrogens」, PMC (EFSA evaluation of 14 long-term bone studies), https://pmc.ncbi.nlm.nih.gov/articles/PMC5429336/
- EFSA Panel on Dietetic Products, Nutrition and Allergies, Scientific Opinion on health claims related to soy isoflavones (vasomotor symptoms, LDL cholesterol, oxidative damage and others), EFSA Journal, https://www.efsa.europa.eu/en/efsajournal/doc/2264.pdf
- 「EFSA Publishes Final Batch of Health Claims, Soy Isoflavones Take a Beating」, Nutritional Outlook, https://www.nutritionaloutlook.com/view/efsa-publishes-final-batch-health-claims-soy-isoflavones-take-beating
- EFSA NDA Panel, Scientific Opinion on soy isoflavones and maintenance of bone mineral density (ID 1655) and reduction of vasomotor symptoms (further assessment), EFSA Journal 2012;10(8):2847, https://orbit.dtu.dk/en/publications/efsa-panel-on-dietetic-products-nutrition-and-allergies-nda-scien-42/
- 「Risk assessment for peri- and post-menopausal women taking food supplements containing isolated isoflavones」, EFSA Journal 2015, https://efsa.onlinelibrary.wiley.com/doi/10.2903/j.efsa.2015.4246
- Food Safety Korea (MFDS), recognised functional ingredient and approved wording lookup, https://www.foodsafetykorea.go.kr/