Peptides are where I find myself arguing in both directions in the same conversation. Against the people who dismiss them as expensive marketing, because there is a genuine research literature here. And against the brands quoting that literature, because the effect sizes in it are considerably smaller than the packaging implies.
Let us start with the thing that causes most of the confusion.
“Peptide” is a category, not an ingredient
A peptide is simply a short chain of amino acids. That is the entire definition. It says nothing about what the chain does, any more than “plant extract” tells you whether you are dealing with chamomile or foxglove.
The peptides used in cosmetics fall into roughly three groups, which Schagen's 2017 overview sets out clearly and which do completely unrelated things:
- Signal peptides — fragments that resemble the breakdown products of collagen. The proposed logic is elegant: skin interprets those fragments as evidence of damage and responds by making more collagen. This is the group with the most cosmetic research behind it.
- Carrier peptides — chains that transport a trace element, most commonly copper, into the skin, where it acts as a cofactor in wound-healing and matrix processes.
- Neurotransmitter-inhibiting peptides — the ones marketed with allusions to injectables, on the basis that they interfere with the signal that makes a muscle contract.
So when a product says “with peptides”, you have learned nothing yet. You need the actual name on the ingredient list.
The study everyone is quoting
Most signal-peptide claims trace back to Robinson's 2005 trial of palmitoyl pentapeptide — pal-KTTKS, sold as Matrixyl. Twelve weeks, vehicle-controlled, on photoaged facial skin. It reported statistically significant improvements in wrinkle appearance and skin roughness against the vehicle.
That is a real result from a properly controlled study, and I want to give it full credit before I qualify it.
Now the qualifications. The improvements were modest — the kind that show up reliably in instrumental measurement and standardised photography, and much less reliably in a mirror. The comparison was against vehicle, not against a retinoid, which is the comparison most people actually care about. And the participants were also using sunscreen, as they should have been, which is doing some of the work in any twelve-week photoaging study.
Gorouhi and Maibach's 2009 review is the more sobering read, and I would recommend it to anyone who wants the honest state of the field. Their assessment across peptide classes is that the mechanistic rationale is often stronger than the clinical evidence, that many studies are small, and that a substantial number are conducted or funded by the companies selling the ingredient. None of that makes the results wrong. It does mean the certainty being sold is ahead of the certainty available.
Copper peptides
Copper is genuinely a cofactor in the enzymes that cross-link collagen and elastin, and the wound-healing literature on copper complexes is real. The leap being made is from wound healing — a specific biological context involving broken skin — to daily cosmetic use on intact skin. That leap is plausible. It is not the same as demonstrated.
There is also a formulation caveat worth knowing: copper peptides do not sit comfortably alongside strong acids or high-dose vitamin C in the same routine. If you use both, separate them by time of day rather than layering them.
The injectable-adjacent ones
This is the group where I am most sceptical, and the scepticism is about a specific gap rather than the molecules themselves.
The mechanism concerns muscle contraction. Facial muscles sit well below the skin surface. For a topical peptide to affect them it has to cross the barrier, traverse the epidermis and dermis, and reach the neuromuscular junction in a meaningful concentration. The published evidence that this happens to a degree that produces a visible result is, in my reading, the thinnest in the whole category.
The studies that exist tend to measure wrinkle appearance rather than muscle activity, which is measuring the hoped-for outcome rather than the claimed mechanism. Those are different things, and the difference is exactly where I would want the evidence to be strongest.
What I would honestly expect
From a well-formulated signal peptide, used consistently for twelve weeks: a modest improvement in skin texture and in the appearance of fine lines. Real, measurable in a study, and probably at the edge of what you would notice unprompted.
Peptides are also generally well tolerated, which is a genuine advantage. If retinoids are too irritating for you, a peptide product is a reasonable thing to be using instead — with the understanding that you are choosing comfort over the stronger evidence base, not getting both.
How to read a peptide product
- Find the actual peptide name on the ingredient list. If the front says “peptide complex” and the list has one, far down, near the preservatives, you have a marketing ingredient.
- Ask which of the three groups it belongs to, then ask what that group is claimed to do.
- Treat “clinically proven” as a prompt to find the study, not as a conclusion. Check the duration, the number of participants, and who paid for it.
- Give it twelve weeks, alone, before deciding — which for this category is not a slogan, it is the length of the trials the claims come from.
My summary, which is unlikely to appear on any packaging: peptides do something. It is smaller than you have been led to expect, and it is better supported than the sceptics allow. Both of those are true at once, and that is a perfectly reasonable thing to buy — as long as you are buying it for what it is.
