Peptides and Collagen in Ageing Skin: What the Trials Actually Measured

August 2, 2026

The short version

Collagen in the dermis fragments with age rather than simply draining away, and that damages the mechanical properties of skin and the cells living in it Rittié, 2015.
UV switches on the enzymes that degrade collagen, and that has been measured in living skin Quan, 2009. No sunscreen was tested in that study, so limiting UV is reasoning from a mechanism, not a trial result.
The best pooled evidence covers nineteen randomised trials and 1,341 people. Hydration and brightness improved, the wrinkle effect was modest, and most of it came from oral formulations rather than serums Nukaly, 2026.
The topical peptide literature is small and often uncontrolled. Only six of fifteen studies in one review used a placebo control Michalek, 2019.
We sell the two Medik8 peptide products below, and neither has a published trial of its own that we can find. Weigh what we say accordingly.

What happens to collagen as skin ages

Skin does not simply run out of collagen. Rittié and Fisher describe the dermal collagen matrix as undergoing gradual fragmentation, damaging the mechanical properties of skin and the cells living in the dermis Rittié, 2015. Long, orderly bundles break into shorter, disorganised pieces.

The next step is often told as settled fact: fibroblasts lose support and make less collagen. That review measured neither fibroblast shape nor collagen output. It is a reasonable explanation, not a number a cream can move.

UV and the enzymes it switches on

Quan and colleagues report that UV irradiation induces expression of certain members of the matrix metalloproteinase family, enzymes that degrade collagen and other matrix proteins in the dermis of living people Quan, 2009.

Now the inference, labelled as one. That study tested no sunscreen and no hat. It measured enzyme expression after UV, not what happens to people who cover up.

Limiting UV is still the step with the clearest documented mechanism behind it, but that is reasoning from a mechanism rather than a trial result.

What the peptide trials actually measured

Peptides are short amino acid chains, and the theory is that certain sequences signal the cells that build the dermal matrix.

The most useful recent answer is a 2026 meta-analysis of nineteen randomised trials covering 1,341 participants, run from university and hospital departments rather than by an ingredient maker. It found:

Peptides, oral formulations in particular, significantly improved hydration and brightness.
The pooled effect on wrinkles was modest: a mean difference of 0.27, p = 0.04.
Subgroup analysis showed even that was largely driven by oral polypeptides, at a mean difference of 1.5.
Effects on elasticity and skin density were inconsistent, and peptides were well tolerated with minimal adverse events reported Nukaly, 2026.

Read that again if you are buying a serum. The hydration, brightness and most of the wrinkle effect came from things people swallowed, not things they applied.

The two topical trials, and their catches

A 56-day randomised, double-blinded, vehicle-controlled trial put a cyclised hexapeptide-9 serum at 0.002 per cent against a retinol serum at the same strength, and reported reductions in crow's feet and forehead wrinkles in the peptide group Chang, 2025. Every author works for the companies behind the ingredient, and 0.002 per cent retinol sits far below retail strengths.
A randomised, blinded, split-face study in twenty people tested a tripeptide and hexapeptide serum against a bland moisturiser, with everyone also having two passes of a 1927 nm non-ablative fractional laser. The peptide side improved more on all measures at thirty and sixty days, except radiance at sixty days Wang, 2020. That is an add-on to laser, in twenty people.

Where the evidence is thinner than the packaging

A 2019 systematic review found fifteen independent studies on matrix-stimulating cosmetic peptides. Only six used a placebo control and five were double-blinded. The authors concluded the literature is sparse and better controlled trials are needed Michalek, 2019.

A wider review of over-the-counter anti-ageing ingredients says it plainly: evidence and good clinical practice of the data supporting their use are oftentimes lacking Imhof, 2021.

Where the two Medik8 products sit

Liquid Peptides is a 30ml serum aimed at the appearance of expression lines. We will not call it high-concentration, because we have no total peptide percentage to check. Advanced Pro-Collagen+ Peptide Cream is a moisturiser, so it combines the peptide and moisturising steps.

Clients tell us both feel comfortable and we see very little reaction in the treatment room. That is shop floor impression, not trial data, and it does not inherit the tolerability figure above, which came from a pooled set dominated by oral supplements.

Medik8 publishes its own performance figures for both. We have not found them in a peer-reviewed journal where the methods can be checked, so we will not repeat them here as findings. That is what our searching turned up, not an accusation about the testing.

What to actually do

Sort your sun exposure first. It is the dull step with the best documented mechanism behind it.
Treat a peptide serum or cream as a comfortable optional layer. No trial we can find establishes a measured result for either of these two products.
If your skin is reactive, patch test on your inner arm first.
Do not buy a seventh bottle on our say-so. Bring your current routine into Parnell and we will look at it.

Frequently asked questions

Do topical peptides actually build collagen in the skin?

That is the theory, and the published human evidence does not yet show it. A 2026 systematic review and meta-analysis of 19 randomised trials covering 1,341 participants found peptides improved hydration and brightness, with a modest pooled wrinkle effect (mean difference 0.27, p = 0.04) largely driven by oral polypeptides. In a 2019 review, 10 of the 15 studies judged the product with image-based methods, so collagen synthesis inside the dermis is inferred from surface measurements rather than demonstrated.

Are peptides better than retinol?

One trial tested that comparison and we will not stretch it further than it goes. A 2025 randomised, double-blinded trial ran a cyclised hexapeptide-9 serum against a retinol serum, both at 0.002 per cent, for 56 days and reported larger wrinkle reductions for the peptide. Every author is employed by the companies behind the ingredient, and 0.002 per cent retinol is far below retail strengths, so this covers one matched low dose rather than the retinoid category.

Why does UV exposure matter so much for collagen?

Because that step has been measured in living skin: UV irradiation induces expression of certain matrix metalloproteinases, the enzymes that degrade collagen and other matrix proteins in the dermis. That study tested no sunscreen, hat or other protective measure, so it cannot tell you what reducing UV exposure does to your collagen. It is still the step with the clearest documented mechanism behind it, which is why we mention it first.

What about the performance figures on the Medik8 packaging?

Medik8 publishes its own performance figures for both Liquid Peptides and Advanced Pro-Collagen+ Peptide Cream. We have not found those figures in a peer-reviewed journal where the methods can be independently checked, so we do not repeat them as established findings. That describes what our searching turned up and is not an accusation about the testing. If a published trial appears, we will link to it.

Which of the two Medik8 peptide products should I choose?

Liquid Peptides is a 30ml serum that layers under a moisturiser, so it suits someone adding a peptide step to a routine they already run. Advanced Pro-Collagen+ Peptide Cream is a moisturiser, so it puts the peptide and moisturising steps together and suits drier skin or a shorter routine. Clients tell us both are comfortable, but that is our impression from the shop rather than trial data, so patch test first if your skin is reactive.

Back to blog