In short. Hydration and elasticity readings are surrogate endpoints. A device number that moves slightly is not the same finding as skin that looks different.
Key points
- The honest sentence is that a course of collagen peptides was followed by a small change in an instrument reading over several weeks in a small group.
- The sold sentence is that collagen reduces wrinkles.
- Both trace back to the same paper, and the distance between them is the whole of this subject.
A trial that reports a change in the skin has to say what it measured, and in this literature the answer is usually a device reading rather than a visible outcome. The choice is not dishonest, and noticing it is the most useful thing a reader can do with the results.
Surrogate endpoints, and why they get used
Hydration and elasticity are measured with instruments: a probe is placed on the skin and returns a number. A visible change in wrinkles is measured by a trained assessor, by standardised photography, or by the participant’s own report, and all three are slower and harder to keep honest. So studies reach for the instrument, which is cheaper and produces a number that can be analysed.
What a moving number means
A surrogate is only as good as its link to the outcome it replaces, and that link has to be demonstrated rather than assumed. Here it has not been, for the readings in common use. A hydration reading can change for reasons unrelated to collagen in the dermis, including the water content of the outer layer and the temperature of the room. When a study reports a small change in a reading and the claim is about wrinkles, the study has not addressed the claim.
The honest sentence and the sold sentence
The honest sentence is that a course of collagen peptides was followed by a small change in an instrument reading over several weeks in a small group. The sold sentence is that collagen reduces wrinkles. Both trace back to the same paper, and the distance between them is the whole of this subject.
Design problems that recur
Four recur often enough to be worth naming. Samples are small, which makes a small effect hard to distinguish from chance. Follow-up is short, usually weeks rather than the months over which skin visibly changes. Placebo control is frequently absent, in a product where belief affects how skin feels and is described. And the outcome reported is often the one that moved rather than the one the study set out to test.
What a convincing trial would look like
It would be blinded and placebo-controlled, because nothing else separates the product from the belief. It would run long enough for a visible endpoint to have a chance of moving. It would state its primary outcome in advance and report it whether or not it moved. It would use a visible outcome, or a surrogate whose relationship to that outcome has already been established elsewhere.
Three questions that grade a trial in a minute
- What was measured: a visible outcome, or a reading from a device?
- Was there a control group, and did participants know which group they were in?
- How long did it run, and was the reported outcome the one the study was designed to test?
Asked of the studies usually cited for oral collagen, those questions place the results correctly: small, short, surrogate, and not about the outcome the claim names.
Frequently asked questions
What is a surrogate endpoint?
A measurement used in place of the outcome that actually matters, because it is quicker and easier to obtain. A hydration reading from a probe stands in for the appearance of skin. A surrogate is only useful if its link to the real outcome has been demonstrated, and in this field that link has not been established for the readings in common use.
Why do trials not simply measure wrinkles?
Because visible change is slow and hard to assess consistently. Grading wrinkles requires trained assessors, standardised photography or participant report, and all three are costly and open to bias. An instrument is cheaper and produces a number that can be analysed, which is why the literature is full of precise measurements of things that stand in for the outcome.
Do any of these trials show a visible improvement?
Not in a way that would carry the claim. Where visible outcomes are reported they are usually secondary, assessed by participants who know what they are taking, over weeks rather than months. A trial that wanted to establish a visible effect would have to be designed around that endpoint from the start, and very few are.
Risks & Limitations
- The claim and the evidence are about two different things: a fact about collagen, and a claim about a swallowed product.
- Swallowed collagen is digested into amino acids, and those are the ones ordinary protein already supplies.
- The trials measure hydration and elasticity readings over short periods, and most are funded by the companies selling the product.
- The products are regulated as food, so no approved indication for skin ageing exists and none is required for them to be sold.
- The harm worth naming is delay: a skin change answered with a supplement instead of an assessment.
When to Seek Care
- A mole or patch that is new, changing in size, shape or colour, or not healing — have it assessed by a clinician.
- Skin that is persistently red, painful or broken, or a rash that does not settle — assessed by a clinician.
- A reaction to a supplement, such as a rash, swelling or digestive symptoms that persist, is assessed by a clinician.
- A declared allergy, a condition requiring protein restriction, or a medication interaction makes this an individual clinical question.
References & Further Reading
- World Health Organization (WHO) — ultraviolet radiation and skin cancer prevention guidance
- U.S. Food and Drug Administration (FDA) — dietary supplement regulation and permitted claims
- National Medical Products Administration (NMPA, China) — health food registration and permitted claims
- Cochrane Library — systematic reviews of nutritional supplements for skin
- Chinese Medical Association Dermatology Branch — clinical guidance
This page is patient education, not medical advice, and it does not cite individual studies. It reviews evidence and regulation; it is not a booking page and does not recommend a treatment. Discuss your own case with a qualified clinician. To ask a question about this review, see Contact & Consultation.