Collagen: The Claim That's Illegal to Make

Walk into any Irish pharmacy and you will find a shelf of it. Sachets, liquid shots, powders to stir into coffee, gummies. Jennifer Aniston is the Chief Creative Officer of one of the largest brands — an actual executive role, not an endorsement. Kourtney Kardashian owns another. Estimates put the global collagen market somewhere between $4bn and $11.4bn depending on whose report you read and how generously they define it.

Here is the thing almost nobody mentions.

In the European Union, it is illegal to tell you that collagen supplements improve your skin. Not discouraged. Not frowned upon. Refused, in law, by name, twelve years ago — and the refusal is still on the statute book.

What actually happened in 2013

GELITA AG, the German company that makes VERISOL — the collagen peptide inside a great many of the products on that shelf — did the responsible thing. It applied to the European Food Safety Authority for permission to make a health claim. The proposed wording asked to say that its collagen peptide had:

a beneficial physiological effect on the maintenance of skin health, as indicated by an increased skin elasticity and a reduction of wrinkles volume

EFSA assessed the dossier and concluded:

A cause and effect relationship has not been established between the consumption of VeriSol®P and a change in skin elasticity leading to an improvement in skin function.

The panel's reasoning is worth knowing. It accepted that improving skin elasticity would be a genuine benefit. It just found that the wrinkle studies submitted "did not assess a function of the skin", that no other skin function had been measured, and that there were "no significant differences between the groups for transepidermal water loss or skin hydration".

The European Commission then wrote the refusal into law: Commission Regulation (EU) No 1154/2014, of 29 October 2014, refusing to authorise certain health claims. VERISOL is named in the annex. The claim was not added to the Union list.

A separate GELITA application for joints had already been rejected in 2011.

So there is no authorised health claim for collagen and skin anywhere in the EU. There is no authorised claim for collagen and joints either. The Food Safety Authority of Ireland's guidance is unambiguous that these rules cover not just packaging but "labelling, presentation, or advertising that the final consumer may see".

The board-certified dermatologist Dr Dray on whether oral collagen delivers what it is sold on.

The other side of the ledger: Vital Proteins announcing Jennifer Aniston as its Chief Creative Officer, from the brand's own channel.

The vitamin C manoeuvre

So how is any of this on shelves with skin claims attached?

Look at the back of the pack. Almost every collagen product sold in Ireland contains added vitamin C. There is a reason, and it is not nutritional generosity.

The only authorised health claims on the EU Register that mention collagen at all are vitamin C claims. One of them reads, in full:

Vitamin C contributes to normal collagen formation for the normal function of skin

That is an approved claim, and any product containing enough vitamin C may legally make it. Read the words carefully, though, because the wording was negotiated with precision. It is a claim about vitamin C, not about collagen. And it says "contributes to normal" formation — maintaining a baseline. It does not say increases, boosts, restores or replenishes.

Put that authorised sentence on a tub of collagen powder and the consumer joins the dots unaided. The brand has said something true and legal. You have heard something else entirely.

Even that has limits. In December 2023 the UK's Advertising Standards Authority ruled against Get A Drip, whose ad said "Vitamin C […] contributes to collagen formation". The company argued consumers would read it narrowly. The ASA disagreed: in context, people would take it to mean the product would "stimulate increased collagen formation". Upheld as misleading.

So the claim got outsourced

If a brand cannot say it, someone else can. This is the part that should interest anyone who works in media.

The rules are identical for influencer posts — the CAP Code applies wherever there is "a payment or any other incentive (requested or unsolicited)", affiliate links included. In practice, enforcement is retrospective and thin, and the ASA's collagen file shows what it catches when it looks:

• Revive Collagen, September 2021 — upheld. An Instagram post with before-and-after images: "lines and wrinkles have drastically reduced. Her skin has become more plump, hydrated and her skin tone has improved." The ASA ruled the hydration line was a health claim, because it implied "a beneficial physiological effect on or change to the structure of the skin" — and no authorisation for it existed. The appearance claims were permissible in principle but needed "scientifically robust studies". A customer testimonial was not enough.
• Kollo Health, November 2023 — upheld on both counts. Kollo submitted clinical studies on its ingredient. The ASA rejected them for small sample sizes of 20 to 52 people, narrow demographics — "primarily Japanese or French women aged 54–64" — a lack of peer review, and dosages that did not match the advertised product.
• Dermacoll, June 2019 — upheld on "reduce wrinkles with this anti-ageing drink" and similar.

Now the Irish part. I searched the Advertising Standards Authority for Ireland's database for collagen complaints. It returns nothing at all.

The UK regulator has ruled against collagen advertisers at least four times. Ireland's appears never to have looked. That is not evidence that Irish marketing is cleaner — Irish shelves carry the same brands making the same claims.

But does it actually work?

Regulation is not the same as reality, and the honest answer here is more interesting than either the hype or the debunk.

In September 2025 the American Journal of Medicine published a meta-analysis by Myung and Park: 23 randomised controlled trials, 1,474 participants. Pooled together, all 23 showed collagen significantly improved hydration, elasticity and wrinkles. Good news.

Then they did something almost nobody in this literature does. They split the studies by who paid for them.

In the trials not funded by supplement or pharmaceutical companies, there was no effect on hydration, no effect on elasticity, and no effect on wrinkles. In the high-quality subgroup, the same. Their conclusion: "There is currently no clinical evidence to support the use of collagen supplements to prevent or treat skin aging."

The industry pushed back hard, and one line in that pushback is extraordinary. Arguing the subgroup analysis was unreliable, the Collagen Stewardship Alliance's position was that the researchers had underestimated the commercial involvement — that 21 of the 23 trials had some form of industry backing, and so the "independent" comparison group was not really independent.

Read that again. The defence of collagen is that the evidence base is more compromised than the critics assumed.

Harvard's Nutrition Source puts it plainly: "most if not all of the research on collagen supplements are funded or partially funded by related industries", and "non-industry funded research on collagen supplements is lacking".

Four papers, one phrase

The disclosure norms in this field deserve their own paragraph.

Bolke et al. (2019) was funded by Quiris Healthcare, whose product it tested, and states "the authors declare no conflict of interest". Kim et al. (2018) was authored by staff of the marketing company and a collagen producer, and declares no conflict. Asai et al. (2020) was funded by commissioned research with Nitta Gelatin, one author works in Nitta Gelatin's peptide division, and declares no conflict.

And a 2019 paper in Nutrients arguing that collagen peptides are a sound dietary protein lists an author whose stated affiliation is GELITA AG, Uferstrasse 7, Eberbach — the same address as the company whose skin claim EFSA rejected — under a declaration of no conflict of interest.

None of this is fraud. It is something more ordinary and harder to fix: these studies exist at all because manufacturers commissioned them, run through a small circle of commercial testing houses and a manufacturer-adjacent research institute. As the nutrition scholar Marion Nestle has long argued, industry funding shapes research at the design stage, usually without anyone intending it.

One executive said the quiet part out loud. Douglas Jones of BioCell Technology, defending the sector: "Most research that is done… on ingredients and products are done by the companies that make them." He is right, and it is true across nutrition science. It is still the reason you cannot take the effect size at face value.

Dr Owen Thomas works through the clinical evidence as a value-for-money question.

Where the sceptics are wrong too

The standard takedown goes: collagen is a protein, your gut breaks it into amino acids, so eating it cannot put collagen in your skin. It sounds decisive. It is only half right.

Digestion is not complete. Certain di- and tripeptides containing hydroxyproline resist the brush-border enzymes and cross into the bloodstream intact. Pro-Hyp and Hyp-Gly have been repeatedly measured in human plasma after ingestion, rising within half an hour, peaking around two hours, gone by six. That is real and it is not seriously disputed.

What has not been shown is the next step. Every experiment demonstrating those peptides stimulating fibroblasts was done in a dish or in mice. The much-repeated line that collagen peptides "remain in the dermis for up to 14 days" comes from a study in rats — and gets laundered through review papers and marketing copy with the species quietly dropped.

The closest thing to human evidence is a single study using suction-blister biopsies, which found increased procollagen and elastin. It was run by the manufacturer's research institute, on the manufacturer's product, and has not been independently replicated.

The experiment nobody has run

Here is the gap that should end the argument, and it is remarkable that it is still open.

Collagen is an incomplete protein. It contains zero tryptophan — a figure confirmed in the industry's own paper on the subject. Roughly a third of it is glycine. The typical dose is 2.5 to 10 grams a day, which against a normal intake of 100 grams or more of protein is a rounding error — unless you are someone genuinely short on protein, which describes a real slice of the older women who make up almost every trial.

So: is it collagen doing something specific, or is it just protein?

Nobody knows, because no skin trial has ever compared collagen against an equivalent dose of a complete protein. Every one of them compares it to a placebo. Until someone runs collagen against whey, "collagen works" and "a bit more protein works" are the same result wearing different labels.

While we are here: the "type I and III for skin, type II for joints" distinction on the packaging has no basis in anything that survives your stomach. What is absorbed are short peptide fragments common to all fibrillar collagens. And topical collagen — collagen in a cream — is a molecule roughly 600 times larger than the accepted limit for getting through the skin barrier. It works as a moisturiser. So does glycerin, for a fraction of the price.

What is genuinely true

I do not want to leave you with a hatchet job, because parts of this hold up.

The joint evidence is better than the skin evidence — considerably. A 2024 meta-analysis in Osteoarthritis and Cartilage pooled 35 trials and 3,165 patients and found small but robust improvements in osteoarthritis pain and function, with the function finding rated high certainty and a trial sequential analysis confirming the sample was large enough to be conclusive. The effects are modest — roughly comparable to paracetamol for knee osteoarthritis — but they are real. If collagen deserves a defence, that is where it lives.

It is safe. Consistently well tolerated, no harm signal. The genuine cautions are heavy metals in some marine-sourced products and the general absence of third-party testing.

And a modest real effect on skin remains plausible. Independently funded meta-analyses — one from a Taipei hospital grant, another declaring no funding at all — still find positive pooled results for hydration and elasticity. The mechanism is not absurd. The honest position is that a small effect probably exists and the evidence base is too commercially captured to size it.

The Irish split

Ireland's own experts do not agree, which tells you something.

The Dublin dietitian Aveen Bannon, in the Irish Examiner, makes the mechanistic objection: "The body breaks it down into building blocks called amino acids, which don't automatically become collagen to be used in the skin."

In the same article, Professor Caitríona Ryan, consultant dermatologist, takes the supportive view — that hydrolysed peptides are absorbed and "have been shown to stimulate fibroblast activity in the dermis".

Both are serious people reading the same literature. That is what a contested evidence base looks like from the inside.

Worth noting too that one of Ireland's most prominent medical voices on supplements, the Cork GP Dr Doireann O'Leary, founded a company that sells collagen. And that in February this year Irish Country Magazine ran a first-person testimonial for an Irish collagen product — thicker hair, stronger nails, plumper skin — of precisely the kind the ASA ruled insufficient as evidence in the Revive Collagen case, carrying no visible partnership disclosure.

The bit that never appears in the marketing

One last thing, because it belongs in any honest account. Collagen is not synthesised in a laboratory. It is rendered from animal hide, bone and fish scale — a by-product of the meat and leather industries. In March 2023 a joint investigation by the Bureau of Investigative Journalism and ITV News traced collagen sold on British high streets back to Amazon deforestation and invasions of Indigenous land.

Read the Bureau of Investigative Journalism's March 2023 investigation →

So should you take it?

If your joints hurt, there is a defensible case, and the evidence there is the strongest in the category.

If you are taking it for your skin, the position is this. Regulators in your own jurisdiction assessed the claim and refused it, and that refusal has never been overturned. The best-designed analysis of the trials finds the benefit disappears once you remove the studies paid for by the people selling it. Nobody has yet tested it against the obvious alternative of simply eating more protein. And if it does work, it works modestly, on instrument readings, over eight to twelve weeks, in studies almost entirely of women aged 35 to 65.

None of that means it does nothing. It means the confidence with which it is sold to you is not earned by the evidence behind it — and that the reason a person on your phone is telling you rather than the company itself is a regulatory fact, not a coincidence.

The dermatologist Cory Dunnick of the University of Colorado was asked whether the data convinced her to start taking it. Her answer was no, followed by this:

If you're looking for the fountain of youth, it's sunscreen, exercise and avoiding smoking. — Dr Cory A. Dunnick

Not a product anyone can put an ambassador on.

A note on the sourcing

Every study, ruling and regulation named here was checked against its primary source: EFSA's published opinions, the text of Commission Regulation (EU) No 1154/2014 on EUR-Lex, the EU Register of authorised claims, and the ASA's own rulings database. Funding statements were read from the papers themselves.

Two limits worth stating. The full text of the 2025 meta-analysis sits behind a paywall, so its effect sizes are reported here only as the published abstract states them. And the industry's formal reply in the same journal was also inaccessible — its arguments are quoted here from trade-press coverage rather than the reply itself.

This is journalism, not medical advice. If you are considering supplements for a specific health problem, that is a conversation for your GP or a registered dietitian. Videos are shown via YouTube's official embed and remain the property of their creators.