Market access & commercial
Health claim substantiation
Why most EU health claims fail: the characterisation requirement that defeats variable botanical extracts, the demand that an effect be a beneficial physiological function, and the population a study has to be run in.
Regulation (EC) No 1924/2006 turned health claims on food from an advertising question into an evidentiary one. The assessment carried out by EFSA’s NDA Panel reduces to three questions asked in a fixed order, and a dossier that fails the first is never assessed on the third. Is the food or constituent sufficiently characterised? Is the claimed effect a beneficial physiological effect that can be measured? And is a cause-and-effect relationship between the two established in humans, at the intake achievable through consumption, in the population the claim addresses?
Characterisation comes first, and it is where botanicals stop
A claim is made for a substance, so the substance has to be a defined thing. For a purified compound at a stated dose this is trivial. For a botanical extract it is often fatal. An extract is defined by plant species and part, growth and harvest conditions, extraction solvent and process, and the resulting profile of constituents; change any of them and the material is, for assessment purposes, a different substance. If the active constituents are unknown and the preparation is not standardised against a marker, then the batch used in the published trial cannot be shown to be the batch in the product, and the evidence does not attach to it. The same logic makes strain-level identification non-negotiable for micro-organisms: the evidence belongs to the strain, not the species.
The effect has to be a benefit, not a change
The second question is where the large-scale rejections happened. Of the roughly two thousand seven hundred general-function claims assessed under Article 13.1, the great majority received an unfavourable opinion, and the recurring reason was not weak trials — it was that the claimed effect was not accepted as a beneficial physiological effect in the first place.
Gut microbiome and immune claims illustrate it exactly. A measured shift in the composition of the gut microbiota is a change, and the Panel’s position has been that a change in composition is not in itself established as beneficial: a direction of benefit must be defined before an outcome can be evidence of it. Immune claims fail for the mirror reason — “supports the immune system” names no measurable function, and movement in cytokine concentrations or lymphocyte counts within the normal range is not, without a linked clinical outcome such as reduced incidence or duration of infection, evidence of a health benefit. Claims that succeeded tended to have a defined, measurable and accepted endpoint: plant sterols and stanols against LDL cholesterol at a stated daily intake, or sugar-free chewing gum against tooth demineralisation.
The population and the pathway
Two further constraints eliminate otherwise reasonable dossiers. Claims address the general healthy population or a defined subgroup within it, so evidence generated in patients being treated for a disease does not transfer to a claim aimed at healthy consumers. And the effect must be shown at an intake plausibly obtained from the food as consumed, not at a pharmacological dose.
Article 14 raises the bar further: reduction of a disease risk factor requires that the factor itself be an accepted risk marker for the disease, which is a question about the epidemiology and not about the product.