Woman holding a jar of food supplements and reading an information leaflet

Research

EU health claims: how to tell whether an effect has been assessed

18 min read

The market for food supplements is astonishingly free. Anyone who wants to sell a product needs no authorisation, no clinical trial and no expert opinion. A notification to the responsible ministry is enough.

At exactly one point this freedom ends: at the statement about the effect. What a manufacturer may write on the pack is regulated EU-wide, and has been since 2006 in a regulation of its own.1 For the first list, the member states had notified around 44,000 statements. After merging and assessment by EFSA, the Commission approved a list of 222 claims in 2012.2

That list is therefore something unusual: a filter set up not by the manufacturer but by an authority. Anyone who knows it can assess any product on the shelf in 2 minutes.

What a health claim actually is

A health claim is not an advertising statement that was permitted. It is a statement that had to be examined beforehand.

Under the regulation, a health claim is any statement that establishes a relationship between a food and health.1 That applies not only to explicit sentences but also to implications and even to pictures. A heart on the pack is a claim.

The regulation reverses the usual relationship. It is not the authority that has to prove a statement false. The manufacturer has to prove that it is true - and before using it. What is not on the list is forbidden.1

The assessment is made by the European Food Safety Authority, EFSA. It does not examine whether a product is useful, but a far narrower question: can a causal relationship be established for this nutrient and this bodily function.

How 44,000 proposals became 222 claims

The process is well documented, because the chair of the responsible EFSA panel described it himself.3

Of more than 44,000 statements, 222 remained

Notified by the member statesin use, with duplicatesmore than 44,000
After merginglist without duplicates4,637
Assessed by EFSAin 341 opinions2,758
Authorisedlist in force since 2012222
Sources: EFSA, European Commission 2012, Regulation (EU) No 432/2012. Bars to scale. 1,548 statements on plant substances from the list remain on hold to this day, neither authorised nor rejected.

These numbers invite a wrong calculation: assessed minus authorised equals rejected. It does not add up. One authorised claim often covers several entries on the list. The 222 claims stand for nearly 500 entries, and more than 1,600 were not authorised.2 And the Commission put 1,548 statements on plant substances on hold. To this day they have been neither authorised nor rejected.3,4

What is telling is what applications failed on. In the area of appetite and weight management, every application for a new claim received a negative EFSA opinion. The main reason was not a refuted effect but insufficiently substantiated effects. An analysis of these applications therefore examined which outcome variables and measurement methods are suitable for demonstrating an effect.5 Such a level of proof had until then been familiar from medicines development.6

Which studies count as evidence

The question is not whether a relationship was observed. The question is whether it is causal.

As evidence, EFSA's approach counts in the first place studies in humans of sufficient quality; everything else can be drawn on only in support.3 The gold standard for that is the randomised, controlled, double-blind trial.

The 3 terms mean:

  • randomised - chance decides who receives the substance and who does not, not the investigators and not the participants
  • controlled - there is a comparison group that receives a dummy preparation
  • double-blind - neither participants nor investigators know who is in which group

The effort sounds excessive until you see what happens without it.

Why observation alone is not enough

For decades, observational studies showed a clear relationship: people with a high intake of beta-carotene and vitamin E developed lung cancer and heart disease less often. It was natural to conclude that the nutrients protect.

Then it was tested in randomised form. In the Finnish ATBC trial, 29,133 men who smoked received alpha-tocopherol, beta-carotene, both or a dummy preparation over 5 to 8 years. The result was the opposite of what was expected: in the beta-carotene group the lung cancer rate was 18 per cent higher, all-cause mortality 8 per cent higher.7 In the American CARET trial with 18,314 participants, the lung cancer rate under beta-carotene and vitamin A was 28 per cent higher. The trial was stopped 21 months earlier than planned.8

The reason for the discrepancy is well studied: people who take supplements differ in education, income, exercise and diet from those who do not. The observational study measures these differences too and attributes them to the product.9 A review of 68 randomised trials with 232,606 participants later took stock. In the 47 methodologically best trials, mortality under antioxidants was 5 per cent higher.10

Protection was expected, more risk was measured

Risk with the supplement compared with placebo. 1 means no difference, above 1 means more often

Lung cancer

ATBC 1.1829,133 men who smoked, beta-carotene, 5 to 8 years
CARET 1.2818,314 participants, beta-carotene and vitamin A, about 4 years

Mortality

ATBC 1.08beta-carotene
CARET 1.17beta-carotene and vitamin A
47 trials 1.05antioxidants, 180,938 participants
0.811.21.41.6

estimate95% confidence interval

Sources: ATBC Study Group 1994, Omenn 1996 (CARET), Bjelakovic 2007 (47 randomised trials at low risk of bias). ATBC as change in incidence, otherwise relative risk. If the whole interval lies above 1, the higher risk is statistically established.

That is exactly why EFSA demands the more demanding procedure. It is not bureaucracy but the lesson from an expensive error.

How much of a nutrient a claim requires

An authorised sentence alone is not enough. There also has to be enough of it in the product.

The same condition hangs on 121 of the 122 claims considered here: it may be used only for foods that meet the minimum requirements for a source of a nutrient.11 That requirement leads, through 2 regulations, to a concrete figure: 15 per cent of the nutrient reference value.12 For food supplements it relates to the recommended daily portion.13

The minimum amount is 15% of the reference value

Amount per daily portion from which a food supplement counts as a source of this nutrient

Vitamin Creference value 80 mg12 mg
Vitamin Dreference value 5 µg0.75 µg
Vitamin Ereference value 12 mg1.8 mg
Vitamin B6reference value 1.4 mg0.21 mg
Folatereference value 200 µg30 µg
Vitamin B12reference value 2.5 µg0.375 µg
Magnesiumreference value 375 mg56.25 mg
Zincreference value 10 mg1.5 mg
Seleniumreference value 55 µg8.25 µg
Sources: Regulation (EU) No 1169/2011, Annex XIII Part A (reference values, 15% as a significant amount); "source" condition attached to the claims in Regulation (EU) No 432/2012; amount per daily portion under § 4 of the German Food Supplements Ordinance. The only exception among the 122 claims considered here: vitamin C during intense physical exercise, 200 mg in addition to the daily dose.

From this follow 2 test questions anyone can answer for themselves:

  • Is the wording there exactly as authorised? Anyone who writes "strengthens the immune defences" instead of "vitamin C contributes to the normal function of the immune system" is not using an authorised claim. A general wording is permitted only if an authorised specific claim accompanies it.1
  • Does the daily portion reach 15 per cent of the reference value? If the nutrient appears only in the ingredient list, without an amount and without a percentage, that cannot be checked - and then the claim may not be used either.
Five jars of capsules and tablets in front of flowers
For food supplements, the 15 percent threshold refers to the recommended daily portion.

And a third observation that sharpens the eye: the authorised wording almost always says "normal". Not "improves", not "increases". A claim describes the maintenance of a normal bodily function, not its optimisation. Anyone who promises more has left the list.

Requirements, food sources and forms of the individual nutrients are covered in separate articles, for example on zinc, selenium, folic acid, vitamin B6, vitamin E and vitamin D3 and K2.

The test: 18 nutrients, 122 claims

What does that look like with a concrete product? ATLAS All-in-One contains 18 vitamins and minerals. For each one of them there are authorised claims, and each one lies above the 15 per cent threshold. The amounts relate to the daily portion of 4 capsules.

  • Thiamine (B1), 2.6 mg, 236 % - 4 authorised claims, among them: Thiamine contributes to normal energy-yielding metabolism.
  • Riboflavin (B2), 2.6 mg, 186 % - 9 claims, among them: Riboflavin contributes to the reduction of tiredness and fatigue.
  • Niacin, 16 mg NE, 100 % - 6 claims, among them: Niacin contributes to the normal functioning of the nervous system.
  • Pantothenic acid, 12 mg, 200 % - 4 claims, among them: Pantothenic acid contributes to normal mental performance.
  • Vitamin B6, 2.3 mg, 164 % - 10 claims, among them: Vitamin B6 contributes to normal homocysteine metabolism.
  • Biotin, 104 µg, 208 % - 7 claims, among them: Biotin contributes to normal psychological function.
  • Folate, 400 µg, 200 % - 8 claims, among them: Folate contributes to normal blood formation.
  • Vitamin B12, 75 µg, 3000 % - 8 claims, among them: Vitamin B12 has a role in the process of cell division.
  • Vitamin C, 161 mg, 201 % - 15 claims, among them: Vitamin C increases iron absorption.
  • Vitamin D, 50 µg, 1000 % - 7 claims, among them: Vitamin D contributes to the maintenance of normal muscle function.
  • Vitamin E, 15.7 mg α-TE, 131 % - 1 claim: Vitamin E contributes to the protection of cells from oxidative stress.
  • Vitamin K, 150 µg, 200 % - 2 claims, among them: Vitamin K contributes to normal blood clotting.
  • Magnesium, 176 mg, 47 % - 10 claims, among them: Magnesium contributes to normal functioning of the nervous system.
  • Zinc, 10 mg, 100 % - 18 claims, among them: Zinc contributes to normal cognitive function.
  • Manganese, 2 mg, 100 % - 4 claims, among them: Manganese contributes to the normal formation of connective tissue.
  • Selenium, 100 µg, 182 % - 6 claims, among them: Selenium contributes to the normal thyroid function.
  • Chromium, 50 µg, 125 % - 2 claims, among them: Chromium contributes to the maintenance of normal blood glucose levels.
  • Molybdenum, 101 µg, 202 % - 1 claim: Molybdenum contributes to normal sulphur amino acid metabolism.

Together that makes 122 authorised claims. Every single one was checked for this article against the official consolidated text of the regulation.11

One claim falls out of that, and it belongs said: vitamin C also carries a claim on immune function during and after intense physical exercise. It requires, however, 200 mg in addition to the recommended daily intake.11 With 161 mg in total, that condition is not met, so the claim cannot be used.

Notes that belong with every health claim: a varied and balanced diet and a healthy lifestyle are the foundation; a food supplement does not replace them. The effects named relate to the daily intake of 4 capsules. The stated recommended daily dose must not be exceeded. Keep out of the reach of children.

And the 17 substances for which there is no claim

For 18 substances the regulation permits a statement. For 17 it does not. Both are in this article.

The same composition contains 17 further substances. For none of them is there an authorised health claim. In a text like this, nothing may therefore be asserted about their effect, and nothing is asserted here.

That does not mean nothing is known about them. It means that the evidence is either not yet sufficient, that no application was made, or that an application did not pass the assessment. Anyone who wants to form their own picture will find the evidence in the article on each:

What that means for the opening question

The list does not answer the question of whether anyone needs a food supplement. It answers a smaller but more reliable question: which statements about a nutrient have withstood an official assessment.

On the larger question the evidence is inconsistent, and that cannot be written away. The COSMOS trial with 21,442 older adults found no effect of a multivitamin over 3.6 years on cancer overall, on cardiovascular events or on mortality.14 A review for the US Preventive Services Task Force covering 84 studies with 739,803 people found little or no difference in cancer, cardiovascular disease and mortality. The only exception was a small benefit for cancer incidence with multivitamins.15 In the same COSMOS trial, however, the sub-studies on cognition (thinking, remembering and concentration) showed a small, statistically significant benefit for memory.16,17

And more is not better. For vitamin C, an in-hospital study in 7 healthy volunteers showed that immune cells are saturated from 100 mg daily. From 500 mg as a single dose, the share the body absorbs (bioavailability) falls.18 A randomised trial in 2,256 older women found more falls and fractures after an annual high dose of vitamin D than under a dummy preparation.19 For selenium the relationship is described as U-shaped: those already well supplied can be harmed by additional intake.20

Anyone wanting to go deeper will find the evidence on the underlying question in 2 articles of their own: a healthy diet or supplements and multivitamins: benefits and limits. How a composition is justified where a substance has no authorised wording at all is set out in How the ATLAS composition came about.

Frequently asked questions

Is a health claim proof of efficacy as with a medicine?

No. An authorised claim describes the contribution of a nutrient to a normal bodily function. It says nothing about the treatment, alleviation or prevention of a disease - such statements are not permitted for foods.12 Under a separate procedure, only claims that a food reduces a risk factor for a disease are allowed.1,3

Why do so many plant substances carry no claim?

For plant substances, 1,548 statements were put on hold and to this day have been neither authorised nor rejected.4 The absence of a claim in these cases does not mean an effect has been refuted, but that no final assessment exists.

From what age does supplementation make sense?

That question cannot be answered from the claims list; it is not a recommendation but a language filter. Surveys from Europe and Germany offer pointers on the supply situation in different phases of life.21-23 They do not replace an individual assessment.

How can you tell on a pack that a claim is permitted?

By 3 features: the sentence is attached to a nutrient, not to the product. It contains the word "normal". And the nutrition table shows at least 15 per cent of the reference value per daily portion for that nutrient.

Why do some products say "supports the immune system" without naming a nutrient?

Such general wordings are permitted only if an authorised specific claim accompanies them.1 Standing alone, the statement is not covered.

Where can you check whether a claim is authorised?

In the EU Register of nutrition and health claims kept by the European Commission. It contains the authorised claims with their conditions and a list of rejected claims with the reasons.1

Key takeaways

  • What is not on the list is forbidden. The regulation reverses the burden of proof: it is not the authority that has to refute a statement, the manufacturer has to substantiate it - and before using it.
  • Around 44,000 notified statements became a first list of 222 authorised claims in 2012. The calculation "assessed minus authorised equals rejected" still does not add up. One claim often covers several entries, and 1,548 statements on plant substances are on hold to this day, not rejected.
  • The question is the cause, not the observation. For decades, observational studies saw protection from beta-carotene; the randomised ATBC trial in 29,133 men who smoked then found 18 per cent more lung cancer, and CARET was stopped early. The demanding procedure is the lesson from an expensive error.
  • An authorised sentence alone is not enough. Almost every claim carries a condition on amount: 15 per cent of the nutrient reference value per daily portion. Anyone who does not state the amount may not carry the claim.
  • The authorised wording almost always says "normal" - not "improves", not "increases". A claim describes the maintenance of a normal bodily function, not its optimisation. Anyone who promises more has left the list.
  • A missing claim is no counter-proof. It can mean that the evidence is insufficient, that no application was made, or that an application did not pass the assessment.
  • The list does not answer whether anyone needs a product. It is a language filter, not a recommendation - it says only which statements have withstood an official assessment.

Sources

  1. European Parliament and Council. Regulation (EC) No 1924/2006 on nutrition and health claims made on foods. Official Journal of the EU, 2006. In particular Art. 2(2)(1), (5) and (6), Art. 5, Art. 6, Art. 10, Art. 14, Art. 20.
  2. European Commission. Food: Commission adopts landmark list of permitted health claims. Press release IP/12/479, 16 May 2012.
  3. Martin A. Assessment and Management of Health Claims in Europe 9 Years after the Adoption of the Regulation. Journal of Nutritional Science and Vitaminology, 2015.
  4. European Food Safety Authority. Health claims: general function claims (Article 13). EFSA, accessed 2026.
  5. Martini D, Biasini B, Rossi S, et al. Claimed effects, outcome variables and methods of measurement for health claims on foods proposed under European Community Regulation 1924/2006 in the area of appetite ratings and weight management. International Journal of Food Sciences and Nutrition, 2018.
  6. Vero V, Gasbarrini A. The EFSA health claims "learning experience". International Journal of Food Sciences and Nutrition, 2012.
  7. The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group. The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. New England Journal of Medicine, 1994.
  8. Omenn GS, Goodman GE, Thornquist MD, et al. Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease. New England Journal of Medicine, 1996.
  9. Lawlor DA, Davey Smith G, Kundu D, et al. Those confounded vitamins: what can we learn from the differences between observational versus randomised trial evidence? The Lancet, 2004.
  10. Bjelakovic G, Nikolova D, Gluud LL, et al. Mortality in randomized trials of antioxidant supplements for primary and secondary prevention: systematic review and meta-analysis. JAMA, 2007.
  11. European Commission. Regulation (EU) No 432/2012 establishing a list of permitted health claims made on foods, other than those referring to the reduction of disease risk and to children's development and health. Consolidated text, as at 20 August 2025.
  12. European Parliament and Council. Regulation (EU) No 1169/2011 on the provision of food information to consumers. Consolidated text, as at 1 April 2025. Art. 7(3), Annex XIII Part A No 1 and 2.
  13. German Federal Ministry of Justice. Food Supplements Ordinance § 4 - labelling. Germany, paragraphs 2 and 3.
  14. Sesso HD, Rist PM, Aragaki AK, et al. Multivitamins in the prevention of cancer and cardiovascular disease: the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomized clinical trial. American Journal of Clinical Nutrition, 2022.
  15. O'Connor EA, Evans CV, Ivlev I, et al. Vitamin and Mineral Supplements for the Primary Prevention of Cardiovascular Disease and Cancer: Updated Evidence Report and Systematic Review for the US Preventive Services Task Force. JAMA, 2022.
  16. Yeung LK, Alschuler DM, Wall M, et al. Multivitamin Supplementation Improves Memory in Older Adults: A Randomized Clinical Trial. American Journal of Clinical Nutrition, 2023.
  17. Vyas CM, Manson JE, Sesso HD, et al. Effect of multivitamin-mineral supplementation versus placebo on cognitive function: results from the clinic subcohort of the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomized clinical trial and meta-analysis of 3 cognitive studies within COSMOS. American Journal of Clinical Nutrition, 2024.
  18. Levine M, Conry-Cantilena C, Wang Y, et al. Vitamin C pharmacokinetics in healthy volunteers: evidence for a recommended dietary allowance. Proceedings of the National Academy of Sciences, 1996.
  19. Sanders KM, Stuart AL, Williamson EJ, et al. Annual high-dose oral vitamin D and falls and fractures in older women: a randomized controlled trial. JAMA, 2010.
  20. Rayman MP. Selenium and human health. The Lancet, 2012.
  21. Cashman KD, Dowling KG, Škrabáková Z, et al. Vitamin D deficiency in Europe: pandemic? American Journal of Clinical Nutrition, 2016.
  22. Rabenberg M, Scheidt-Nave C, Busch MA, et al. Vitamin D status among adults in Germany - results from the German Health Interview and Examination Survey for Adults (DEGS1). BMC Public Health, 2015.
  23. Mensink GBM, Fletcher R, Gurinovic M, et al. Mapping low intake of micronutrients across Europe. British Journal of Nutrition, 2013.

ATLAS All-in-One - 35 ingredients, 4 capsules, once a day.