Chapter 30: Peptides in Dermatology and Cosmetics — What's in Your Skincare (and Does Any of It Work?)

"What is deepest in man is the skin." — Paul Valéry, L'Idée fixe (1932)

Overview

There is a good chance you own a bottle that brought you here.

It cost more than you would normally spend. The box said peptide, and possibly clinically proven, and almost certainly reduces the appearance of fine lines. Some days your skin looks better. You are not sure whether that is the serum, the sleep, the season, or the fact that you started paying attention to your face at the same time you started paying for something to put on it.

This chapter is not going to tell you that you were stupid. The underlying science is real. Skin is a signaling organ, fibroblasts in the dermis genuinely respond to short peptide fragments, broken collagen genuinely acts as a damage signal, and a tripeptide that binds copper genuinely does things in wound-healing research that have been reproduced for decades. Any account of cosmetic peptides that opens by calling the whole category a scam has already thrown away the part worth understanding.

The problem is not the biology. The problem is a layer of dead cells about as thick as a sheet of kitchen film whose entire evolved purpose is to stop molecules from getting through it.

That is the chapter, and everything else follows from it. When you finish, you will be able to look at any peptide on any ingredient list and ask the three questions that decide the matter: what is this molecule supposed to do, where would it have to be to do it, and is there evidence it gets there in a meaningful quantity? Those questions dispose of most claims in this space quickly and fairly — and not all of them, which is why this chapter contains a ✅, two ⚠️ ratings, and a ❌ rather than a single verdict. One full section is given to reading a cosmetic efficacy study, the most transferable skill here; once you know the genre's signature weaknesses you will never again be impressed by "clinically tested on 32 subjects over four weeks."

In this chapter, you will learn to:

• Explain what the stratum corneum is and why it is the organizing obstacle for every claim here
• Apply the 500-dalton rule of thumb, and say precisely what it is and is not
• Distinguish signal, carrier, and neurotransmitter-inhibiting peptides by their proposed mechanisms
• Explain why palmitoylation on a cosmetic peptide is a delivery modification, not an activity one
• State plainly what is genuine in the GHK-Cu literature and where the gap opens
• Explain why botulinum toxin works spectacularly and why the route is the whole answer
• Decode cosmetic marketing language as the legal artifact it is
• Identify the missing vehicle control, and ask the one question that settles most cosmetic studies
• Rate specific claims — not molecules — and say what would change each rating

Learning Paths

💄 Cosmetic — this is your chapter. Read every section, twice if you can. §30.2 is the physics, §30.8 is the skill, and §30.9 is the part you will use on a Tuesday in front of a shelf. Chapters 1 and 4 promised you this payoff; here it is. 🔬 Science — §30.2, §30.5, and §30.6 form a single argument about delivery that generalizes far beyond skin. §30.3's palmitoylation is Chapter 33's toolkit applied to a barrier problem instead of a half-life problem, which is a good test of whether Chapter 33 landed. 🏥 Clinical — §30.6 and §30.7 matter most. Patients ask about serums constantly, and §30.9 gives you a way to answer that is honest without being dismissive. Keep §30.6's safety callout. 💊 GLP-1 — mostly optional, but §30.7 (the structure/function line) explains the marketing language on compounded and "supplement" GLP-1 products too, and §30.8 transfers directly. 🏋️ Performance — §30.2 and §30.8 are the ones that pay. Topical and transdermal claims are common in the performance space, and the barrier argument here is the same argument.

30.1 A large ingredient category built on a real mechanism

Walk down any skincare aisle, at any price point, and you will find peptides. They appear in drugstore moisturizers and in serums costing more per milliliter than most prescription drugs. They appear as "palmitoyl tripeptide-1" and "acetyl hexapeptide-8" and "copper tripeptide-1" and dozens of other names built to the same pattern. They appear on the front of the box, which is the tell: an ingredient gets front-of-box treatment when it is doing marketing work, not merely formulation work.

I am not going to quote you a market size. Figures circulate, and when you chase them back they generally terminate in a paid industry report nobody outside the industry has read. A number with no traceable source is not evidence, even when it is probably roughly right.

What matters is that the category is large and rests on a premise that is scientifically respectable.

The premise, stated fairly

Skin ages in two ways dermatology distinguishes carefully. Intrinsic aging is chronological: the slow decline in cellular activity that happens on your inner forearm, where the sun rarely reaches. Extrinsic aging — overwhelmingly photoaging, from ultraviolet exposure — is responsible for most of what people actually dislike when they look in a mirror. Compare the back of an older person's hand with their upper inner arm and you are looking at the difference, on one body, with genetics controlled.

Underneath both processes is the dermis: a matrix of collagen (mostly types I and III), elastin, and glycosaminoglycans, produced and maintained by fibroblasts. With age and with ultraviolet exposure, collagen synthesis slows and degradation accelerates, partly through matrix metalloproteinases (MMPs), enzymes that chew up matrix proteins and that UV exposure upregulates. The visible result — thinning, laxity, lines that no longer smooth out — is the appearance of a matrix being disassembled faster than it is rebuilt.

So here is the premise: if fibroblasts respond to chemical signals, and if we know some of the signals, could we apply one to the skin and tell fibroblasts to rebuild?

That is a real question with real support. Fibroblasts in culture demonstrably change their behavior in response to short peptides, and wound healing genuinely involves peptide fragments acting as signals — a body that has been cut needs to know it has been cut, and fragments of its own broken proteins are one of the ways it finds out. Chapter 2's account of receptor signaling applies here without modification: a small molecule with the right shape, arriving at the right receptor, produces effects out of all proportion to its mass.

Everything in this chapter grants that premise. The disagreements start one step later.

The four families

The cosmetic literature sorts peptides into four groups. Notice, as you read them, that this is a taxonomy of proposed mechanisms — it classifies molecules by what they are supposed to do rather than by what has been shown. That is not disqualifying, but it is worth naming, because a classification built out of intentions will always look more organized than the evidence underneath.

The fourth family gets least attention here, not because it is less plausible but because its human evidence is thinnest of all and the arguments against it are the same arguments, only shorter.

WHERE A TOPICAL PEPTIDE HAS TO GET TO — and what is standing in the way

    product applied to the surface   ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼
   ┌──────────────────────────────────────────────────────────────┐
   │ STRATUM CORNEUM      dead, keratin-filled corneocytes packed  │ ← THE BARRIER
   │  ~10–20 µm           in a lipid matrix. "Bricks and mortar."  │   Its entire job
   │  ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ ▓▓▓▓ │   is to keep
   │                                                              │   things OUT.
   ├──────────────────────────────────────────────────────────────┤
   │ VIABLE EPIDERMIS     living keratinocytes; no blood vessels;  │
   │                      nourished by diffusion from below       │
   ├──────────────────────────────────────────────────────────────┤
   │ DERMIS               ★ FIBROBLASTS — signal and carrier       │
   │                        peptides are aimed HERE               │
   │                      collagen I & III, elastin, GAGs,        │
   │                      blood vessels, nerve endings            │
   ├──────────────────────────────────────────────────────────────┤
   │ SUBCUTIS / MUSCLE    facial expression muscles               │
   │                      ★ NEUROMUSCULAR JUNCTIONS — acetyl       │
   │                        hexapeptide-8 is aimed HERE           │
   │                      ★ and this is where botulinum toxin is   │
   │                        INJECTED, because injection is the    │
   │                        only reliable way to arrive           │
   └──────────────────────────────────────────────────────────────┘

Read that diagram as a map of distances. A signal peptide has to cross one barrier and traverse the epidermis. A neurotransmitter-inhibiting peptide has to do all of that and keep going, through the dermis and into muscle, and arrive in quantity. The targets are not equally hard to reach, which is why the ratings in this chapter are not all the same. Delivery is a spectrum, not a switch, and sorting claims by how far the molecule has to travel predicts the evidence surprisingly well.

🧬 The Molecule — what an INCI name tells you, and what it conceals

Cosmetic ingredients are named under the International Nomenclature of Cosmetic Ingredients (INCI), a system with entirely different rules from the drug-naming conventions of Chapter 1 §1.8.

Decode one: palmitoyl pentapeptide-4.

• palmitoyl — a 16-carbon fatty acid (palmitic acid) has been chemically attached
• pentapeptide — the peptide portion is five residues long
• -4 — a serial number distinguishing it from other registered pentapeptides

That is genuinely informative. You now know the molecule's size class and that someone deliberately made it greasier, which — as §30.3 explains — is a delivery move.

Now notice what the name does not contain. It does not give the sequence. Two pentapeptides with different sequences and completely different properties would be distinguished only by serial number. It does not give the concentration in the product, which cosmetic labels are generally not required to disclose. It says nothing about purity, or about stability in the finished formulation. And the serial number carries no ranking: acetyl hexapeptide-8 is not an improvement on acetyl hexapeptide-3. In that case they are the same molecule under two designations, because INCI naming was revised. The number changed; the chemistry did not.

Compare this with Chapter 1 §1.8. A generic drug name is assigned when a compound enters serious clinical development. An INCI name is assigned when an ingredient is registered for cosmetic use — a process requiring no efficacy evidence whatsoever. Both are real naming systems; only one is a signal about evidence.

30.2 The stratum corneum — the reason most of this is hard

Here is the fact that organizes the whole chapter.

The outermost layer of your skin, the stratum corneum, is a barrier of dead, keratin-filled cells called corneocytes, embedded in a matrix of lipids — principally ceramides, cholesterol, and free fatty acids. The usual metaphor is bricks and mortar, and it is a good one. The bricks are flattened dead cells cross-linked into tough envelopes; the mortar is a highly organized lipid phase in stacked bilayers. On most of the body it is only ten to twenty micrometers thick — thinner on the face and eyelids, dramatically thicker on palms and soles.

Its entire evolved function is to keep things out. And to keep water in: transepidermal water loss through an intact barrier is low, and the whole reason a land animal can exist without desiccating is that this layer works. It is not an incidental nuisance sitting between a serum and its target. It is one of the more impressive pieces of engineering in vertebrate biology, and it has been optimized over hundreds of millions of years against exactly what a cosmetic chemist is trying to do.

A molecule applied to skin has three possible routes inward. Intercellular — winding through the lipid mortar between corneocytes — is the dominant route for most substances, and it is a long, tortuous path strongly favoring small, moderately lipid-soluble molecules. Transcellular means going straight through the corneocytes, alternately crossing lipid and protein phases, which few molecules do well. Appendageal means down hair follicles and sweat ducts, bypassing the barrier; these openings are a small fraction of surface area, but the route is real.

The 500-dalton rule of thumb

Penetration falls sharply with molecular size. The commonly cited heuristic — proposed in the dermatology literature around the turn of the millennium and repeated ever since — is that molecules above roughly 500 daltons penetrate intact skin poorly.

Take that seriously, and also take it correctly. It is a rule of thumb, not a law. It was assembled from observations across several bodies of evidence: the molecular weights of common contact allergens, of topical drugs that work, and of compounds successfully formulated into transdermal patches. Nothing enforces a cutoff at 500. Lipophilicity matters, charge matters, the vehicle matters, and skin that is inflamed or broken behaves nothing like skin that is intact. What the rule captures is a real and steep trend, and its practical use is as a prior: if your molecule is well above 500 daltons and carries charge, the burden of proof for penetration sits squarely with you.

Now put the chapter's molecules on that scale.

Most cosmetic peptides exceed the threshold. GHK is a genuine exception, and that is worth remembering in §30.4 — it is one of the few molecules here with size on its side, which makes the rest of its evidence story more interesting rather than less.

So the central question for every claim in this chapter is Chapter 4's delivery question, transposed into a new setting: does the molecule reach living tissue in a meaningful quantity? In Chapter 4 the obstacle was your digestive tract and the first-pass metabolism behind it. Here it is a sheet of dead cells. The structure of the argument is identical, and so is the honest answer for most compounds: not established.

What "penetration" claims usually mean

Penetration is typically assessed with Franz diffusion cells (excised skin mounted between two chambers, measuring what crosses over hours), by tape stripping (removing successive layers of stratum corneum with adhesive), or by imaging labeled compound in a skin section. Each is useful, and each is easy to over-read in a consistent direction. Excised skin is dead skin: no circulation to carry a molecule away, no living metabolism to destroy it, usually frozen and thawed, which is not kind to a lipid barrier. Those results systematically overestimate what happens on a living face.

And tape stripping, which sounds like the most direct measurement of all, mostly answers the wrong question. If a study reports that a peptide was "detected in the skin," ask which layer. A molecule detected in the stratum corneum has been detected in the layer whose job is to hold things briefly and then shed them. Your outermost cells are on their way off your body. Arriving there is not arrival; it is the waiting room, and the waiting room is being demolished continuously.

The workarounds, and what each costs

Formulators know all of this perfectly well. The strategies are real, and each has a downside.

Penetration enhancers — ethanol, propylene glycol, certain surfactants and fatty acids — work substantially by disrupting the lipid mortar. That is the mechanism, stated honestly: they make the barrier work less well, and a barrier that works less well admits more of everything, including irritants and allergens. The tension between penetration and barrier integrity is unavoidable.

Occlusion — an ointment, film-forming polymer, or mask — hydrates the stratum corneum and substantially increases penetration of many compounds. This one genuinely works. It also, independently, makes skin look plumper and smoother for a while, which will matter enormously in §30.8.

Liposomal and lipid-conjugated carriers package the peptide in a lipid vesicle or hang a fatty acid off it to improve partitioning into the lipid phase. Palmitoylation is exactly this, and §30.3 is the case study. Evidence that liposomal delivery reaches the dermis specifically, as opposed to the stratum corneum, is much thinner than the marketing around the word suggests.

Microneedling punctures the barrier with fine needles. Unlike everything else here it unambiguously works as a delivery method, because it does not negotiate with the stratum corneum; it makes holes in it. That is also why it carries risks the others do not: introducing an ingredient formulated for intact skin into channels reaching living tissue is a different exposure entirely, and granulomatous and allergic reactions to substances applied during needling are documented. "It penetrates better" and "it is safer" pull in opposite directions, and a product boasting of both deserves a raised eyebrow.

🔍 Check Your Understanding

1. A serum's marketing states that its peptide "has been shown to penetrate the skin." What is the single follow-up question that determines whether this is meaningful, and why?
2. GHK is about 340 daltons and acetyl hexapeptide-8 is about 890. Does that settle which one works? What does it settle?
3. A formulator adds a penetration enhancer to a peptide serum. Name the benefit and the cost, and explain why they are two descriptions of the same physical event.

30.3 Signal peptides: palmitoyl pentapeptide-4 and the collagen-fragment idea

Of everything in this chapter, this is the most genuinely elegant mechanism.

Palmitoyl pentapeptide-4 — marketed under the ingredient trade name Matrixyl, and one of the most widely used peptides in skincare — is a five-residue fragment derived from type I procollagen with a palmitoyl group attached. Take those two halves separately, because they do completely different jobs.

The peptide half: a fragment as a message

Collagen is not secreted in its final form. Fibroblasts make procollagen, which carries extension peptides at both ends that are cleaved off as the molecule matures into the fibril-forming protein. The pentapeptide in question is a subfragment of one of those propeptide regions.

The proposed mechanism: when collagen in the dermis is damaged and broken down, the resulting fragments are in effect a chemical announcement that damage has occurred. Fibroblasts detect the announcement and respond by synthesizing new matrix. Apply the fragment topically and you counterfeit the announcement — the fibroblast believes there has been damage, and rebuilds.

The mechanism is genuinely plausible, and it was not invented to sell serum. Feedback regulation by breakdown products is a real and widespread motif in biology, and there is real supporting cell-culture work: the peptide has been shown in fibroblast culture to increase production of matrix components, including collagen. That work exists, it has been reproduced in various forms, and it is why this ingredient is taken seriously by people who are not selling anything.

It is also, precisely, cell-culture work. A fibroblast in a dish is bathed directly in the compound at a known concentration, with no barrier of any kind between them. Chapter 5's central discipline applies at full force: knowing how something would work is not evidence that it does. The dish shows the fibroblast can receive the message. It says nothing about whether the message arrives.

The palmitoyl half: a delivery modification

Now the fatty acid — and this is the part the 🔬 Science track should notice.

The pentapeptide by itself is small, water-loving, and charged: the worst possible properties for crossing a lipid barrier. Attaching a 16-carbon fatty acid makes it dramatically more lipid-soluble, so it partitions into the stratum corneum's lipid mortar instead of sitting on the surface.

That is a delivery modification, not an activity modification. Nobody added palmitate because it makes the peptide better at talking to fibroblasts. They added it because the unmodified peptide cannot get anywhere.

If that sounds familiar, it should. This is exactly the Chapter 33 logic, applied to skin. When a fatty acid chain is attached to semaglutide, receptor activity is not the target of the modification — half-life is, through albumin binding. Same chemical trick, same underlying insight (peptides are poorly suited to the environment we need them to survive, and lipid attachment changes how they partition), aimed at a completely different obstacle. The cosmetic and pharmaceutical industries arrived at the same tool from opposite ends of the problem.

There is a cost. Improved partitioning into a lipid phase is also a reason a molecule might get comfortable in the stratum corneum and stay there. Being good at entering a lipid barrier and being good at crossing it are not the same property, and a molecule can be optimized into a reservoir.

The human evidence

Here the honest account gets short.

Human evidence for topical palmitoyl pentapeptide-4 exists and is limited, typically small, short, industry-funded, and instrument-based. The studies tend to enroll a few dozen participants, run for several weeks to a few months, use silicone-replica profilometry or similar instrument endpoints, and be conducted or funded by parties with a commercial interest in the result. Some are vehicle- controlled and split-face, which is a genuine methodological credit and better than much of this literature. Effect sizes, where reported, are small.

I am not going to quote you a percentage improvement in wrinkle depth, and neither should any source you trust without first showing you the study design. A percentage from an unblinded, four-week, manufacturer-run study with an instrument endpoint is not a fact about your face — §30.8 explains why in detail.

What we can say fairly: a plausible mechanism, real supporting laboratory work, a rational delivery modification, and human data too thin to support the confident claims made for it.

📊 Evidence Rating

Claim: Topical palmitoyl pentapeptide-4 (Matrixyl), applied to the face of healthy adults with photoaged or age-related facial lines, produces a visible reduction in wrinkles. Rating: ⚠️ Promising but preliminary (as of this writing, 2026) Why: The mechanism is genuinely plausible and supported by real fibroblast culture work, and the palmitoylation is a rational answer to the penetration problem — but the human evidence is small, short, predominantly industry-funded, and reliant on instrument endpoints whose relationship to visible change is unestablished, and delivery to the dermis at an active concentration has not been convincingly demonstrated in living skin. What would change it: Adequately powered, independently funded trials of at least six months comparing the full formulation against the identical vehicle without the peptide, with blinded assessment of standardized photographs as the primary endpoint — plus a quantitative demonstration that the intact molecule reaches the dermis in living human skin. A ✅ needs both halves; delivery data alone would not do it, and neither would the clinical data.

A note on the "-3000" style combination products, since they are everywhere. Blends of two or more signal peptides are marketed as improvements, and may be. But a combination creates an attribution problem: if a blend shows a small effect, no component can claim it, and each component's own evidence is thinner than what was just described. Combinations expand the claim surface faster than they expand the evidence.

30.4 Carrier peptides: GHK-Cu, and a gap that is easy to miss

GHK is a naturally occurring tripeptide — glycyl-histidyl-lysine — first isolated from human plasma in the early 1970s. It binds copper with high affinity, forming the complex written GHK-Cu and listed on cosmetic labels as copper tripeptide-1.

Copper is a legitimate biological actor here, not a mystical trace element. It is the required cofactor for lysyl oxidase, the enzyme that cross-links collagen and elastin fibers into a mechanically competent matrix, and for one form of superoxide dismutase. A tissue rebuilding its matrix genuinely needs copper delivered to the right place, and a small peptide that binds copper and hands it off is a coherent physiological idea. That is what "carrier peptide" means: the peptide is proposed to be a chaperone for the metal.

What is genuinely established

Let me be unusually plain here, because this is a compound where boosters and debunkers both misrepresent the record.

GHK-Cu has genuine, decades-old activity in wound healing and in stimulating aspects of extracellular matrix production in laboratory systems. This is not fabricated science. There is a substantial body of work — cell culture, animal wound models, and some human wound contexts — showing effects on collagen synthesis, on the behavior of fibroblasts and other cell types, and on various markers of tissue repair. Papers on GHK and GHK-Cu have been appearing for roughly half a century. Whatever else is true, this molecule is not an invention of a marketing department, and anyone who tells you it is has not read the literature.

If your instinct at this point is "so it works," slow down. If your instinct is "so it's nonsense," slow down harder. The interesting thing is the specific shape of the gap.

The gap

The gap is between a wound — where the barrier is already breached and delivery is not the problem — and intact skin, where it is.

That sentence is the whole section. Every wound-healing result comes from a setting in which the stratum corneum has been cut, burned, abraded, or surgically removed. Applied to a wound, GHK-Cu is already past the barrier by definition; the molecule and the fibroblast are in the same room. The question this chapter exists to ask has been answered in advance by the injury.

Applied to intact facial skin, the barrier is present and functional, and the question reopens. GHK's small size — around 340 daltons for the free tripeptide, with the copper complex not much heavier — genuinely helps, and it is one of the few peptides in the category that clears the 500-dalton heuristic comfortably. That is a real point in its favor and I want it on the record. But clearing a size threshold establishes that penetration is not ruled out. It does not establish that a therapeutically relevant concentration reaches dermal fibroblasts from a cosmetic formulation applied to intact skin. Those are different claims, and only the first is settled.

A second complication is copper itself. Copper is redox-active: the same chemistry that makes it a useful enzyme cofactor makes free copper capable of generating reactive oxygen species, which is why copper homeostasis in tissue is tightly regulated. The complexed form is not free copper and cosmetic exposures are low — I am not raising an alarm. I am noting that "more copper delivered to the dermis" is not straightforwardly good in the way "more collagen" sounds, and that a mechanism story which is entirely upside is usually a story simplified for selling.

You may also have seen advice that copper peptide products should not be layered with vitamin C or acids. That is not marketing mysticism: ascorbic acid and low pH really can disrupt the copper complex. It is also a nice illustration of how much of a cosmetic peptide's fate is decided in the bottle rather than in the skin.

📊 Evidence Rating

Claim: Topical GHK-Cu (copper tripeptide-1), applied to intact facial skin in healthy adults, improves the appearance of aging skin. Rating: ⚠️ for modest instrument-measured effects; ❌ for the strong marketing version ("rebuilds collagen," "reverses skin aging," "resets skin to a younger state"). As of this writing, 2026. This reproduces the rating given in Chapter 6 §6.8 — same claim, same answer. Why: The wound-healing and cell-culture literature is real, substantial, and decades old, and the molecule is small enough that penetration is not ruled out on size grounds. But nearly all compelling data comes from settings where the barrier was already breached; on intact skin, human data is limited, small, short, largely industry-associated, and instrument-based. The strong claims extrapolate from a wound to a face. What would change it: For the ⚠️ to become ✅ — independently funded, vehicle-controlled trials on intact skin with blinded photographic endpoints over six months or more, plus demonstrated dermal delivery in living skin. For the ❌ to move at all — evidence that any measured change corresponds to a difference an observer notices, which is currently the missing link, not merely an unmeasured one.

Notice what just happened, because it is the rating system working as designed. One molecule, two ratings, because there are two claims (Chapter 5, rule 6). The modest claim and the strong claim are not the same assertion with different enthusiasm; they have different evidentiary requirements, and the evidence meets one and not the other. Any source giving GHK-Cu a single verdict has thrown away the only distinction that matters.

30.5 Neurotransmitter-inhibiting peptides: Argireline and the comparison it cannot support

Acetyl hexapeptide-8 — also seen as acetyl hexapeptide-3, the same molecule under an older designation, and under the ingredient trade name Argireline — is a six-residue peptide of roughly 890 daltons, derived from a portion of a protein called SNAP-25.

It is marketed, explicitly or by heavy implication, as a topical alternative to injected botulinum toxin. The rationale is mechanistically specific, which is part of why it persuades.

The mechanism, stated properly

When a motor nerve fires, it releases acetylcholine into the neuromuscular junction and the muscle contracts. Release requires vesicles inside the nerve terminal to fuse with the terminal membrane, and that fusion is executed by the SNARE complex — a bundle of proteins including SNAP-25, syntaxin, and synaptobrevin — which zippers together and pulls the two membranes into contact.

The proposal is that acetyl hexapeptide-8, being derived from a portion of SNAP-25, competes with the real SNAP-25 for a place in the assembling complex. A complex incorporating the decoy assembles less effectively, fewer vesicles fuse, less acetylcholine is released, the muscle contracts less forcefully, and the expression lines that repeated contraction carves into overlying skin become less pronounced.

That is a real mechanism operating on a correctly described piece of cell biology, and there is in-vitro work consistent with the peptide interfering with SNARE complex formation. I want to grant that clearly, because the failure here is not that someone made something up.

The problems, which compound

Now walk the molecule to its target.

WHAT ACETYL HEXAPEPTIDE-8 HAS TO ACCOMPLISH — five steps, all required

  1. Cross the stratum corneum       ~890 Da, charged residues. The 500 Da heuristic
                                      says: poorly. Not impossible; poorly.
        │  some fraction survives
        ▼
  2. Traverse the viable epidermis   diffusion through living tissue; peptidases present
        │  a fraction of that fraction
        ▼
  3. Cross the dermis                blood vessels here — anything reaching the dermal
                                      circulation is carried AWAY from the target and
                                      diluted into the whole body
        │  a fraction of that fraction
        ▼
  4. Reach the neuromuscular         these sit in muscle, below the dermis. This is
     junctions of facial muscle       farther than any other target in this chapter.
        │  a fraction of that fraction
        ▼
  5. Arrive at a concentration       it must OUT-COMPETE the cell's own SNAP-25 in a
     sufficient to compete            fast, efficient, redundant molecular machine —
                                      not merely be present, but be present in bulk

  Each step passes a fraction. Fractions multiply. Five modest fractions in series
  produce a very small number, and the last step needs a large number.

  ── Compare: botulinum toxin, 150,000 Da, skips steps 1–4 with a needle. ──

That diagram is the argument, and its most awkward feature deserves drawing out: step 3 works against step 4. The dermis is vascularized. A molecule that successfully reaches it has reached the bloodstream's collection point, which sweeps it away and dilutes it into several liters of circulation. To get past the dermis and into muscle in quantity, a molecule essentially has to beat the circulation, and the circulation is very good at its job. The better a topical agent is at reaching the dermis, the more of it is removed before it goes deeper.

Botulinum toxin is injected directly into the muscle precisely because that is the only way to get it there. Not because injection is traditional, or because the industry prefers procedures, or because nobody has tried hard enough to make a cream. Because a 150,000-dalton protein cannot cross intact skin, and even if it could it would not survive the trip to muscle in useful quantity. The needle is not a limitation of the technology. It is the technology.

Human evidence for topical acetyl hexapeptide-8 exists and is limited: small studies, short duration, instrument endpoints, small measured effects, and a strong association with commercial sponsorship.

⚠️ Hype Check — "Botox in a jar"

The claim, in its usual form:

"Argireline works like Botox without the needles — it relaxes the facial muscles that cause expression lines, so you get the same smoothing effect from a cream."

What's true in it. More than people usually concede. SNARE biology is exactly as described: SNAP-25 is a real component of a real fusion complex, and disrupting it really does prevent acetylcholine release — that is precisely how botulinum toxin works, so the mechanism being invoked is not merely plausible but demonstrably potent when it occurs. The peptide is a real, characterized molecule with in-vitro data behind the interference claim. And there is a genuine safety advantage: a topical peptide that does not reach a neuromuscular junction cannot cause the eyelid ptosis or asymmetry that an injection occasionally does. The product is nearly harmless.

Where it fails. The claim is a comparison, and the comparison is the part that breaks. The two agents differ in the one variable that determines whether a SNARE-interfering molecule does anything: whether it arrives at the SNARE complex. Botulinum toxin arrives because a clinician puts it there with a needle. Acetyl hexapeptide-8 must negotiate five sequential barriers, and the product of five fractions is small.

Note also the asymmetry in the evidence. Botulinum toxin's effect is large, obvious, and measurable by anyone — a treated forehead and an untreated one are distinguishable across a room. If topical acetyl hexapeptide-8 produced a comparable effect, demonstrating it would be trivially easy and spectacularly profitable. Decades in, the demonstrations remain small, short, instrument-dependent, and sponsor-associated. When an effect is claimed to be dramatic but can only be detected by delicate instruments, the instrument is measuring something other than the claim.

Verdict: the mechanism is sound and the comparison is not. This is a delivery failure wearing a mechanism's clothes — which, as Chapter 4 warned, is what most peptide delivery failures look like.

📊 Evidence Rating

Claim: Topical acetyl hexapeptide-8 (Argireline) reduces facial expression lines in healthy adults by an effect comparable to injected botulinum toxin — i.e., functions as an alternative to injection. Rating: ❌ Hype outpaces evidence (as of this writing, 2026) Why: The peptide is far above the 500-dalton heuristic and must reach neuromuscular junctions below the dermis, past a vascularized layer that clears molecules away from that target, at concentrations sufficient to compete with endogenous SNAP-25 in an efficient and redundant machine. Human evidence is limited, short, mostly sponsor-associated, and shows small instrument-measured effects; nothing supports the comparison being marketed. The reason for this rating is delivery, not mechanism — and ❌ describes the evidence, not the molecule (Chapter 5, rule 2). What would change it: A demonstration that intact acetyl hexapeptide-8 reaches facial muscle in living humans at a concentration capable of competing with SNAP-25, plus a randomized, vehicle-controlled trial with blinded photographic assessment showing an effect of a magnitude a person would notice. A delivery breakthrough — a validated carrier, or a needling protocol with its own safety data — would be the plausible route, and would create a new claim requiring its own rating.

30.6 Botulinum toxin: a protein that works spectacularly, by injection

If this chapter were only about failures, it would teach the wrong lesson. So here is the counterweight, and it is a large one.

Botulinum toxin is a protein produced by the bacterium Clostridium botulinum, and it is one of the most potent biological toxins known. Its natural role has nothing to do with faces; in humans it causes botulism, a descending flaccid paralysis that, untreated, kills by paralyzing the muscles of respiration.

The molecule is around 150,000 daltons — a heavy chain and a light chain joined by a disulfide bond, often found in larger complexes with accessory proteins. Several serotypes exist; type A is most used clinically. The heavy chain binds to and enters motor nerve terminals. Inside, the light chain acts as a zinc-dependent protease, and its substrate is the SNARE machinery: type A cleaves SNAP-25, the same protein acetyl hexapeptide-8 hopes to imitate a fragment of.

Sit with that. The mechanism Argireline proposes is a topical imitation of what botulinum toxin does enzymatically. The toxin does not compete for a spot in the complex; it cuts SNAP-25, catalytically, one toxin molecule destroying many substrate molecules. And it is delivered by needle to the tissue where the substrate lives. Same target, different chemistry, and — decisively — different route.

Cleaved SNAP-25 cannot support vesicle fusion. Acetylcholine is not released. The muscle does not contract. Injected in tiny, controlled quantities into a specific muscle, the effect is localized and temporary, wearing off over months as nerve terminals sprout and recover.

The evidence

It works spectacularly, and it is supported by extensive randomized evidence — for cosmetic use and for a long list of medical indications.

On the cosmetic side, the best-established indication is the glabellar lines between the eyebrows, for which randomized, placebo-controlled trials with blinded investigator and subject assessment have been conducted, replicated, and reviewed. Approvals followed — in the United States for glabellar lines in 2002, with lateral canthal ("crow's feet") and forehead lines added later, and comparable approvals in many other jurisdictions. Different products carry different approved indications and are not interchangeable with one another.

On the medical side the list is long and predates the cosmetic use. The first approvals, in 1989 in the United States, were for blepharospasm and strabismus — the cosmetic application was noticed in the course of treating eye conditions, a nice reminder of how often applications find their molecules rather than the reverse. Since then: cervical dystonia, spasticity in adults and children, severe primary axillary hyperhidrosis, overactive bladder and neurogenic detrusor overactivity, and chronic migraine (added in 2010, following randomized trials in that population).

📊 Evidence Rating

Claim: Botulinum toxin type A, injected into the target muscles of healthy adults, produces a temporary reduction in the appearance of glabellar lines and other approved cosmetic indications. Rating: ✅ Strong clinical evidence (as of this writing, 2026) Why: Multiple adequately powered randomized placebo-controlled trials with blinded assessment, consistent and large effects, decades of post-approval use, regulatory approval across many jurisdictions, and a characterized safety profile with known and manageable adverse effects. What would change it: Very little at the efficacy level; this is about as settled as cosmetic medicine gets. The live questions concern long-term effects of repeated treatment over decades, muscle atrophy with sustained use, immunogenicity and secondary non-response, and comparative effectiveness among products — none of which threaten the core claim.

📊 Evidence Rating

Claim: Botulinum toxin type A, injected per approved protocols, is effective for approved medical indications including chronic migraine, cervical dystonia, spasticity, and severe primary axillary hyperhidrosis. Rating: ✅ Strong clinical evidence (as of this writing, 2026) Why: Each indication is supported by its own randomized controlled evidence and its own regulatory review; these are separate claims that happen to share a molecule, and each was established independently. What would change it: Indication-specific findings would move indication-specific ratings. This rating does not extend to unapproved uses, which carry their own, weaker evidence and would require separate ratings — one molecule, many ratings (Chapter 5, rule 6).

Two ratings for one molecule, and neither transfers to the other. That is rule 6 doing real work rather than decorative work.

Why this section exists

Botulinum toxin demonstrates that the category's failures are about delivery and dose, not about proteins and peptides being inherently ineffective on skin.

Look at the numbers again. Acetyl hexapeptide-8 is about 890 daltons and does very little. Botulinum toxin is about 150,000 daltons — a hundred and seventy times larger, and utterly incapable of crossing intact skin — and produces one of the most reliable effects in all of cosmetic medicine. Size did not decide it. Route decided it. Put the molecule where its target is and it works; ask it to find its own way through a barrier evolved to stop it, and it does not.

This is why the honest position on cosmetic peptides is not cynicism. The biology is real, the molecules are real, and what is usually missing is a way to get them to the tissue — an engineering problem with a history of being solved, sometimes by chemistry and sometimes by a needle.

🩺 Safety and Risk — the most vivid possible illustration that "natural" means nothing

Botulinum toxin is produced by a living organism. It is as natural as anything in this book. It is also one of the most lethal substances known, and the entire clinical enterprise consists of using vanishingly small, precisely placed quantities of a poison. Chapter 1's point, in its most vivid available form: provenance predicts nothing about safety.

In competent hands, for approved indications, the safety record is good and common adverse effects are transient. But they are real:

• Local effects from spread to neighboring muscles — brow or eyelid ptosis, asymmetry, an altered smile, difficulty with certain expressions. These resolve as the effect wears off, which takes months, not days.
• Injection-site effects — bruising, pain, headache.
• Site-specific risks — dysphagia with neck injections; dry eye and altered blinking around the eyes; urinary retention with bladder treatment.
• A boxed warning about distant spread of toxin effect. Symptoms consistent with botulism — swallowing and breathing difficulty, generalized weakness — have been reported after injection, most often at the higher exposures used for spasticity, and can be life-threatening.
• Immunogenicity. It is a foreign protein; some people develop neutralizing antibodies and stop responding.

Two product facts, which are safety facts rather than dosing guidance. Units are product-specific and are not interchangeable between botulinum toxin products — a "unit" of one is not a unit of another, and treating them as equivalent has caused harm. And counterfeit or improperly handled product is a documented public-health problem: clusters of botulism-like illness following injections of unapproved or counterfeit product administered outside appropriate medical settings have been investigated by public health authorities, including a multi-state investigation in the United States in 2024.

This book gives no protocols and will not tell you whether or where to have anything injected. What it will say is that this is a prescription medicine with a boxed warning, that the difference between a good outcome and a bad one is substantially the difference between a qualified injector using authentic product and someone who is neither, and that this is a Chapter 39 conversation with a clinician rather than a Chapter 30 conversation with a book.

30.7 The drug/cosmetic line, and why every serum sounds the same

Read enough skincare copy and you notice a verbal tic. Nothing ever reduces wrinkles; things reduce the appearance of wrinkles. Nothing firms skin; products deliver a firmer look. Nothing increases collagen; products support the skin's natural renewal process. Nothing treats anything.

This is not timidity, and it is not writers running out of verbs. It is a legal artifact, and once you see it you cannot unsee it.

The two definitions

In United States law the categories are defined by intended use, and they are close to mutually exclusive in the claims they permit.

A cosmetic is a product intended to cleanse, beautify, promote attractiveness, or alter the appearance. Cosmetics do not require premarket approval. And crucially, a cosmetic may not claim to affect the structure or function of the body.

A drug is a product intended for the diagnosis, cure, mitigation, treatment, or prevention of disease — or intended to affect the structure or any function of the body. A drug may make those claims, and must prove them, through a process requiring evidence of both safety and efficacy before marketing.

Now the consequence, which people find genuinely startling: the category is determined by the claim, not by the chemistry. The same molecule, at the same concentration, in the same base, can be a cosmetic or a drug depending on what the label says.

THE SAME BOTTLE, TWO SENTENCES, TWO LEGAL CATEGORIES

  "Reduces the appearance of fine lines"        → COSMETIC
  "Skin looks firmer and more radiant"          → COSMETIC
  "Supports your skin's natural renewal"        → COSMETIC (and says nothing at all)
  "Visibly smooths and plumps"                  → COSMETIC
  ────────────────────────────────────────────────────────────────────────────
  "Increases collagen production in the dermis" → DRUG CLAIM  (structure/function)
  "Rebuilds the extracellular matrix"           → DRUG CLAIM  (structure/function)
  "Treats photodamage"                          → DRUG CLAIM  (treats a condition)
  "Blocks nerve signaling to facial muscles"    → DRUG CLAIM  (affects a function)

  Same molecules. Same jar. The sentence decides the category, and the category
  decides whether anyone has to prove anything.

That diagram explains almost everything about how these products are marketed. The vocabulary is not vague for its own sake; it is chosen to stay on the cosmetic side of a legal line, because crossing it converts a product that can be sold tomorrow into a drug requiring years and a great deal of money to approve. A company that believed its peptide genuinely rebuilt collagen and could prove it would have every incentive to say so — and would then be selling a drug. That essentially nobody in this space chooses that path is a data point about what they think they could prove.

Practical translation

"Cosmeceutical" has no legal standing in the United States. It is a useful conversational term for a cosmetic containing biologically active ingredients, and it is a marketing term. A product is a cosmetic or a drug; there is no third box.

"Clinically proven" is not the regulated term of art most people assume. On a cosmetic it frequently means an in-house consumer study — often small, often unpublished, often with self-report endpoints — and it does not indicate that any regulator reviewed anything.

"Dermatologist tested" means a dermatologist was involved in testing. It is not a claim about the result, though the natural reading is that it worked. "Data on file" means there is data and you cannot see it.

Jurisdictions differ, and the differences are instructive

This boundary and its wording vary by jurisdiction. The European Union requires a documented safety assessment and a product information file before marketing, and separately requires that claims meet common criteria including truthfulness and evidential support — on paper a stricter claims regime than the American one, though the practical gap is narrower than the texts suggest.

The clearest illustration of divergence is sunscreen. In the United States it is regulated as an over-the-counter drug, because preventing sunburn is a structure/function matter. In the European Union, Japan, and many other markets it is a cosmetic. The same bottle changes legal category when it crosses a border, and the permitted ingredients and claims change with it. Nothing about the chemistry moved.

In the United States the cosmetic framework was modernized by legislation enacted in 2022, adding facility registration, product listing, adverse-event reporting, and requirements to keep records substantiating safety. Note what that strengthens: it is a safety framework, not an efficacy one. Nothing in it requires anyone to demonstrate that a peptide does what the box implies.

30.8 Reading a cosmetic study: n, duration, self-report, sponsor

This is the most transferable section in the chapter. The skills work on any topical product, any supplement with a "clinical study," and most of the performance claims in Part III.

Cosmetic efficacy studies form a genre, and genres have conventions. Here are the seven that recur.

1. Small n. Twenty, thirty, forty participants. Not automatically fatal — a large, reliable effect can show up in a small sample — but it interacts badly with everything else. Small studies are underpowered for the modest effects claimed here, so a null result proves little and a positive one may be noise.

2. Short duration. Four weeks, eight weeks, twelve weeks. Consider what is being claimed: an effect on the architecture of the dermis, a process that took decades to produce the appearance being treated. Collagen turnover in adult human dermis is slow — years, not weeks. A four-week study of an anti-aging claim is measuring something, but it is very unlikely to be matrix remodeling. What it is most likely measuring is hydration, which brings us to the big one.

3. Self-report and unblinded assessment. Two of the commonest endpoints are the subject's own rating of their skin and an investigator's visual grading. Both are highly susceptible to expectation. A participant who knows they are using the active product, who was recruited on the promise of improvement, who has been applying something pleasant twice a day and looking closely at their own face in good lighting, will report improvement. So will an investigator who knows which side got the serum.

4. Instrument endpoints of unestablished relevance. The instruments are real and the measurements reproducible; that is not the problem. Profilometry captures surface topography, often from a silicone replica, quantifying line depth. Cutometry applies suction and measures deformation and recovery, reported as elasticity or firmness. Corneometry measures electrical capacitance of the stratum corneum, reported as hydration. Transepidermal water loss measures barrier function. Each measures something genuine. What is unestablished is the relationship between a change in any of them and anything a person notices in a mirror. A statistically significant change in a cutometer parameter is a fact about a cutometer. If the study does not also show that observers can tell the difference, nobody has demonstrated the thing the product is sold for. This is the surrogate endpoint problem from Chapter 5, wearing a lab coat.

5. No vehicle control, or an inadequate one. This is the big one, and it is close to disqualifying when it is missing.

A peptide is never applied alone. It is dissolved in a vehicle: water, humectants like glycerin and hyaluronic acid, emollients, occlusives, preservatives, sometimes silicones. That vehicle is a moisturizer. Not "similar to" one — it is one, and often a good one.

And here is the fact that should reorganize how you read every study in this genre: moisturized skin looks better and measures better. Hydrating the stratum corneum makes it swell. Swollen corneocytes push the walls of fine lines together, and fine lines become visibly shallower within hours. Hydrated skin reflects light more evenly, which reads as radiance. Corneometer readings rise, because that is precisely what a corneometer measures. Profilometry improves, because the surface really is smoother. Cutometer readings shift, because hydrated stratum corneum has different mechanical properties.

Every one of those changes is real. None of them requires a peptide, or any active ingredient whatsoever. Plain glycerin in water will do a substantial share of it.

Therefore a comparison against no treatment is close to uninformative. It cannot separate the peptide from the moisturizer, and the moisturizer alone is expected to produce most of the measurable result. A study whose control arm is an untreated cheek has established that applying a moisturizer to your face beats not applying one, which we already knew and which does not cost \$90.

The single most useful question you can ask about any cosmetic study: was it compared against the same formulation without the peptide? It is short, it is fair, it is answerable, and a large fraction of this literature cannot answer it yes.

6. Split-face designs — a real help, not a fix. Each participant gets active on one side and control on the other, so genetics, sun history, sleep, and season are matched perfectly. That is a genuine improvement. It does not fix blinding: if the two preparations differ at all in feel, scent, absorption, or finish, participants often work out which side is which — and even when they do not, they frequently believe they have, which has the same effect on their ratings. Split-face designs are also vulnerable to carryover, since people touch their faces, sleep on their sides, and use the same hands on both cheeks.

7. Sponsorship, and the studies you never see. Most cosmetic efficacy research is funded by parties selling the ingredient or the product. This does not make individual studies false; it shapes the population of studies you get to read. A negative in-house study is a business document, not a publication. Nobody is obliged to register a cosmetic trial in advance, so there is no public record of what was started and no way to notice what quietly stopped. Add the venues — journals with industry ties, sponsored supplements, conference posters, "white papers" never peer reviewed at all. A citation is not a study; it is sometimes a PDF.

THE COSMETIC STUDY CHECKLIST — seven questions, in order of decisiveness

  1. VEHICLE CONTROL?     Same formula minus the peptide?    ← if no, mostly stop here
  2. BLINDING?            Subjects blinded? Assessors blinded?
  3. ENDPOINT?            Blinded photo assessment > investigator grading
                          > instrument reading > subject self-report
  4. DURATION?            Weeks measure hydration. Months measure remodeling.
  5. SAMPLE SIZE?         Powered for the effect claimed, or for any effect at all?
  6. FUNDING?             Who paid, and who employed the authors?
  7. WHERE PUBLISHED?     Peer-reviewed journal, sponsored supplement, poster, or PDF?

  A study can fail 5, 6, and 7 and still be informative. A study that fails 1 and 2
  is measuring expectation and moisturizer, in unknown proportion.

🔬 Read the Study — a constructed abstract, and what is wrong with it

The following is a [constructed teaching example]. It is not a real study; it is assembled from the recurring features of the genre so you can practice before meeting the real thing.

"Thirty-two healthy female volunteers aged 35–60 with visible periorbital fine lines applied a serum containing 3% of a proprietary pentapeptide complex twice daily to the full face for 28 days. Wrinkle depth was assessed by silicone-replica profilometry at baseline and day 28. Investigators graded periorbital lines on a five-point scale. Subjects completed a satisfaction questionnaire. Mean wrinkle depth decreased significantly from baseline (p < 0.05). Investigator grading improved significantly. 87% of subjects agreed that their skin looked smoother. The serum was well tolerated."

Work through the checklist before reading on.

No control arm at all. Every comparison is against baseline. The design cannot distinguish the peptide from the vehicle, from the twice-daily massage, from seasonal change, or from the fact that participants' skin was at its worst on the day they were recruited into a study about fine lines — regression to the mean, which produces improvement on its own.

Twenty-eight days. Long enough to hydrate a stratum corneum thoroughly. Not long enough for dermal collagen remodeling to be a plausible explanation of anything.

No blinding of anyone. Investigators knew what they were grading. Subjects knew what they were applying and had been recruited on the promise of it working.

All three endpoints are soft, and they are not independent. Profilometry is directly and strongly affected by hydration; investigator grading is unblinded; the questionnaire measures satisfaction. Three "confirming" results are three views of the same expectation. "87% agreed" is an agreement rate on a positively worded item, from unblinded participants, in a study they were recruited into: rhetorical work, no evidentiary work.

"Significantly (p < 0.05)." Against baseline, in n = 32, with an unstated number of comparisons, and no effect size reported — so we do not know whether the change was large enough for a human being to perceive. Statistical significance is a claim about noise, not about magnitude. And "3% of a proprietary pentapeptide complex" conceals the sequence, the identity, and what else is in the complex; the figure may refer to a supplied solution rather than to the peptide.

The one thing this study does establish, and I want to give it its due: the serum was well tolerated in 32 people over four weeks. That is a genuine finding of a genuine kind. It is also not the finding on the box.

The fix, in one sentence: run the identical formulation without the peptide on the other side of each face, blind everyone, run it for six months, and have independent assessors who never met the participants rank standardized photographs.

30.9 What is actually worth your money, claim by claim

Now the practical section, and I want to do it without contempt, because contempt is both unkind and analytically lazy.

You bought the serum for reasons that were not stupid. The mechanism you were sold is a real mechanism. The molecule really is in the bottle. And the product probably does make your skin look somewhat better, because it is a well-made moisturizer and moisturizers work. The question is not whether you were fooled. It is which part of what you bought is doing the work.

The honest hierarchy

Ranked by strength of evidence for changing how skin ages and looks over time — strongest first:

Sun protection. Not a close call. Ultraviolet exposure is the dominant driver of the visible changes people call aging, and daily sunscreen use has been tested in a randomized controlled trial with skin-aging endpoints: the Australian community trial reported by Hughes and colleagues in Annals of Internal Medicine in 2013 found that participants randomized to daily sunscreen use showed less skin aging by microtopography measurement over four and a half years than those using it at their discretion. A randomized trial, in a community, with an aging endpoint, over years. Nothing else in this chapter has evidence of that class for prevention.

Prescription retinoids. Tretinoin is an approved drug for photoaging, which means someone had to satisfy a regulator with efficacy data — the §30.7 distinction becoming useful. It is supported by randomized evidence, produces changes visible to blinded observers, and has real side effects (irritation, peeling, photosensitivity) that are the price of the mechanism. Over-the-counter retinol and its relatives sit below on both effect and evidence, and the trade-off between potency and irritation is the whole story of the category.

Moisturizer. Does exactly what it claims, reliably, immediately, and cheaply. Improves the appearance of fine lines within hours by hydrating the stratum corneum, and supports barrier repair. It is somewhat absurd that this has become the boring answer.

Then, some distance down: peptides. The best of them are ⚠️. The strongest claims made for them are ❌. None has evidence approaching sunscreen or prescription retinoids.

So the honest answer, for a great many products: the vehicle is doing most of the work, and a good moisturizer with sunscreen is better supported than any peptide in this chapter. I mean that as useful information, not a put-down. It is a real answer to a real question, it saves real money, and it points at the things that work.

Claim by claim

What would make a specific product worth trying

If you want to use a peptide product — and there are decent reasons to, including that you like it — these questions separate the more serious from the less:

• Does the maker publish anything specific? Not "clinically proven" but a citation, a design, an n, a control arm. Companies that have done real work usually enjoy saying so in detail.
• Was the comparison against the same formula without the peptide? §30.8's question, applied to the product in your hand.
• How modest is the claim? A surprisingly good heuristic. Modest claims correlate with honest ones, because a company willing to say "may improve the appearance of fine lines with regular use" did not feel the need to promise more than it had.
• Is it a good moisturizer at a fair price? Because that part is definitely working.

And the risk side, plainly: the downside here is overwhelmingly financial, not medical. Cosmetic peptides are, as a class, well tolerated. Irritation and contact allergy are possible, as with any topical, and copper-containing products can stain fabric and occasionally tint skin. But the realistic worst case for most of these products is that you spent money on a nice moisturizer with an expensive story attached. That is a materially different risk profile from most of the compounds in Part III, and it is one reason this chapter's tone is gentler than Chapter 19's.

💊 In the Clinic — what a dermatologist can offer, and what they will probably say about your serum

A dermatologist evaluating the same question has a different toolkit, and the evidence is stratified sharply.

Well-supported: photoprotection; prescription retinoids for photoaging; treatment of the conditions that actually change how skin looks day to day, which are frequently rosacea, acne, seborrheic dermatitis, and eczema rather than aging at all; and procedural interventions with randomized evidence — botulinum toxin for dynamic lines, several device-based and chemical resurfacing modalities, and injectable fillers for volume, each with its own indications and risks.

Reasonable but variable: topical vitamin C formulations, niacinamide, alpha hydroxy acids — real evidence of varying quality, formulation-dependent, generally modest effects.

Adjunct at best: cosmetic peptides.

What most dermatologists will not do is tell you to throw the serum away. It is not harming you, you like using it, adherence to any routine has value, and the emollient base is doing something real. What they are likely to say is that it is not the lever — that if the goal is to change how your skin ages, the sunscreen you are inconsistent about matters more than the serum you are devoted to.

Hear that without embarrassment. Nearly everyone allocates effort this way: the expensive, novel, mechanistically interesting intervention gets the attention, and the cheap, boring, well-evidenced one gets skipped. That is not a skincare phenomenon. It is a human one, and Chapter 39 is about having the conversation that fixes it.

30.10 Ratings across the category

Everything above, collected. Every rating is dated, attached to a population and an endpoint, and paired with what would change it — Chapter 5's rules 1 and 5, the ones that keep a rating system from becoming an opinion column.

Why the last row is the weakest one

That final rating is the least useful line in the table, and explaining why is the chapter's methodological payoff.

A rating attaches to a claim, with a population and an endpoint (rule 1). "Topical peptides" is not a claim. It is an aisle. The category contains a tripeptide with fifty years of wound-healing literature and a genuine size advantage; a procollagen fragment with a plausible feedback mechanism and real cell-culture support; a SNAP-25 fragment sold on a comparison to an injection it cannot imitate; and an unknown number of ingredients with nothing behind them but an INCI registration. Those are not one thing, and averaging them produces a number that describes none of them.

The category-level rating exists only as a rough prior — the answer to "should I expect a randomly chosen peptide serum's marketing to be supported?" — and the answer is no, which is worth knowing and not worth much. The per-claim ratings above are where the information is. Any source telling you "peptides work" or "peptides don't work" has compressed away everything you came for. That is Chapter 1's error arriving from the other end: there, the word peptide was used to borrow insulin's credibility; here, it is used to spend it.

📊 Evidence Rating

Claim: Topical peptides, as an ingredient category, produce visible anti-aging benefit in healthy adults using cosmetic formulations on intact skin. Rating: ⚠️ → ❌ depending on the specific claim (as of this writing, 2026) Why: The category is not a claim, and this rating is therefore the least useful one in the chapter. Modest, hydration-adjacent, instrument-measured effects are ⚠️ for several ingredients; the structural claims implied by most marketing are ❌ for all of them, because delivery to living dermis at an active concentration is not established for any peptide here. What would change it: Nothing at the category level, and that is the point — categories do not accumulate evidence, claims do. Ask instead which molecule, in which formulation, for which population, measured how. Use the six ratings above; use this one only as a default expectation while you find out.

What would move the category

Delivery data. Quantitative demonstration that an intact peptide reaches the dermis, in living human skin, at a concentration comparable to what produces effects in culture. Not detection in the stratum corneum. Not excised skin in a diffusion cell. Living dermis, quantified.

Trials built the right way. Six months or longer. Vehicle-controlled with the identical formulation minus the peptide. Blinded assessment of standardized photographs by assessors who never met the participants, as the primary endpoint, with instrument measures in supporting roles.

Independence and pre-registration. Funding without a stake in the answer, and a public record of what was planned before it was run — which would let us see the negatives, currently invisible.

Notice a final asymmetry, and let it inform your priors. This industry has both the money and an overwhelming commercial motive to run exactly these trials. A vehicle-controlled six-month trial with blinded photographic endpoints showing a clear peptide effect would be worth a great deal to whoever owned it. These trials are largely not run, or not published. That absence is not proof of anything — but it is weak evidence, and it points in one direction rather than both.

📋 Your Evidence Dossier

Field 4 — Route and Delivery. The field where, for this chapter's molecules, route decides everything.

Field 4 asks how a compound is administered, whether that route delivers it to the tissue where its target lives, and what fraction arrives. In most chapters this is a technical detail sitting between more dramatic fields. Here it is the whole entry. The same molecular strategy — interfering with SNARE-mediated acetylcholine release — earns a ✅ by one route and a ❌ by another, and nothing else about the two entries differs in a way that matters.

FIELD 4 — BOTULINUM TOXIN TYPE A (cosmetic use)      [worked demonstration]
  Route              Intramuscular injection, by a qualified clinician, into
                     specifically identified facial muscles
  Target tissue      Motor nerve terminals at neuromuscular junctions in the
                     injected muscle
  Barriers crossed   NONE. The needle bypasses every barrier between the
                     surface and the target.
  Fraction arriving  Effectively the administered quantity, minus local
                     diffusion, delivered to the intended tissue
  Delivery evidence  Not in question. The clinical effect IS the delivery
                     evidence — the muscle demonstrably stops contracting
  Consequence        Large, reliable, blindable, replicable effect
  RATING             ✅ for approved cosmetic indications

FIELD 4 — ACETYL HEXAPEPTIDE-8 (topical)             [worked demonstration]
  Route              Topical application to intact facial skin
  Target tissue      Neuromuscular junctions in facial muscle — THE SAME
                     TARGET as the entry above
  Barriers crossed   (1) stratum corneum, ~890 Da vs. a 500 Da heuristic
                     (2) viable epidermis, with peptidases
                     (3) dermis, vascularized — clears molecules AWAY
                     (4) into muscle, at competitive concentration
  Fraction arriving  Unknown, and unmeasured in living human muscle.
                     Five sequential fractions multiply.
  Delivery evidence  Not established. In-vitro SNARE interference data does
                     not address arrival.
  Consequence        Small instrument-measured effects; no support for the
                     comparison being marketed
  RATING             ❌ as an alternative to injected botulinum toxin

Read the two entries side by side. Same target. Same biological mechanism, in kind. Opposite ratings. The only field that differs materially is Field 4 — and it is not a detail in the entry, it is the entry. Everything else follows.

Your turn

For each peptide in your dossier, complete Field 4:

FIELD 4 — ROUTE AND DELIVERY
  Route              how is it actually administered?
  Target tissue      where does its target live? (be specific — which organ,
                     which cell type, which side of which barrier)
  Barriers crossed   list them in order. For each: is there evidence this
                     one is crossed, or only an assumption?
  Fraction arriving  known? estimated? unmeasured? say which
  Delivery evidence  what is the actual evidence that the molecule arrives
                     where it needs to be, in the amount it needs to be there?
  Alternative routes does the SAME molecule have a different route with a
                     different rating? If so, they are different entries.

Two instructions specific to this chapter. First, if any of your peptides has both a topical and an injected form, give it two Field 4 entries and expect two ratings. That is not duplication; it is rule 6, and it will be one of the more clarifying half-hours of this project.

Second, for any topical entry, write one sentence answering: what is the evidence that this arrives, as distinct from the evidence that it would work if it did? If you cannot find any, write "no delivery evidence located" and date it. That is not a failure of your search. It is a finding, and it is the finding this chapter exists to teach you how to make.

Conclusion

Cosmetic peptides sit on a real mechanism and an unsolved delivery problem.

The mechanism is not in dispute. Fibroblasts respond to peptide signals. Collagen fragments really do function as damage announcements. Copper really is a cofactor for the enzymes that cross-link matrix. GHK-Cu really does have half a century of laboratory and wound-healing literature behind it, and palmitoyl pentapeptide-4 really does change what fibroblasts do in a dish. Anyone who tells you the whole category is fiction has skipped the reading.

The problem is the ten to twenty micrometers of dead cells on top, whose entire evolved function is to keep molecules out and which is very good at it. Most cosmetic peptides are above the 500-dalton rule of thumb. For most of them, delivery to living dermis at an active concentration is not established — not disproven, not established, which is a different and more honest thing to say. Formulators attempt to solve it with penetration enhancers that work by degrading the barrier, with occlusion that works and also independently makes skin look better, with lipid conjugation borrowed from pharmaceutical chemistry, and with needles that bypass the question entirely.

And then there is botulinum toxin: a hundred and fifty thousand daltons, produced by a bacterium, utterly incapable of crossing intact skin, and one of the most reliable interventions in cosmetic medicine — because a clinician puts it exactly where it needs to be. Route, not size. Delivery, not mechanism. If you take one thing from this chapter, take the pair: acetyl hexapeptide-8 and botulinum toxin, aimed at the same molecular machinery, separated by a needle, and rated ❌ and ✅.

So what do you do with the bottle in your bathroom? Use it if you like it. It is very unlikely to hurt you, it is probably a decent moisturizer, and the appearance of your fine lines really may be modestly better for it — mostly through hydration, which is a real effect and not a fake one. Just know what you bought, and know that the sunscreen you keep forgetting has better evidence behind it than anything else in this chapter apart from a needle.

Chapter 31 turns to a different kind of claim entirely, and you will bring the same three questions with you: what is this molecule supposed to do, where would it have to be to do it, and what is the evidence that it gets there.

Key Terms

Stratum corneum — the outermost layer of the epidermis: dead, keratin-filled corneocytes embedded in a lipid matrix, roughly 10–20 µm thick. Its evolved function is to keep substances out and water in, and it is the central obstacle for every topical peptide claim.

Corneocyte — a flattened, dead, keratin-filled cell; the "brick" of the stratum corneum.

Lipid matrix — the ceramide-, cholesterol-, and fatty-acid-rich "mortar" between corneocytes; the principal route by which most substances cross, and what penetration enhancers disrupt.

Transepidermal water loss (TEWL) — the rate at which water passes through skin and evaporates; a standard measure of barrier integrity.

500-dalton rule — a rule of thumb holding that molecules above roughly 500 daltons penetrate intact skin poorly. A heuristic and a strong prior, not a law; most cosmetic peptides exceed it.

Penetration enhancer — an ingredient that increases absorption, typically by disrupting the stratum corneum lipid matrix. Benefit and cost are the same physical event.

Occlusion — covering skin to reduce water loss, which hydrates the stratum corneum, increases penetration, and independently improves the appearance of fine lines.

Fibroblast — the dermal cell that produces and maintains collagen, elastin, and other matrix components. The intended target of signal and carrier peptides.

Extracellular matrix (ECM) — the structural network outside cells; in dermis, principally collagen (types I and III), elastin, and glycosaminoglycans.

Type I procollagen — the precursor of type I collagen, carrying propeptide extensions cleaved during maturation. Palmitoyl pentapeptide-4 derives from one of these regions.

Matrix metalloproteinase (MMP) — an enzyme that degrades matrix proteins; upregulated by UV.

Photoaging — the component of visible skin aging caused by ultraviolet exposure, as distinct from intrinsic chronological aging. The dominant driver of what people dislike about aging skin.

Signal peptide (cosmetic sense) — a peptide proposed to act as a message to fibroblasts prompting matrix synthesis, typically by mimicking a collagen breakdown fragment. Not the molecular-biology meaning of the term, which denotes a protein-trafficking sequence.

Carrier peptide — a peptide proposed to deliver a trace metal cofactor, usually copper, to enzymes requiring it. GHK-Cu is the archetype.

Neurotransmitter-inhibiting peptide — a peptide proposed to reduce acetylcholine release, and thereby muscle contraction, by interfering with SNARE complex assembly.

Palmitoylation — attachment of a 16-carbon fatty acid to a peptide, increasing lipid solubility. In cosmetics a delivery modification aimed at barrier crossing; related acylation in pharmaceuticals targets half-life (Chapter 33).

INCI name — a name assigned under the International Nomenclature of Cosmetic Ingredients. Encodes modification, length, and a serial number; encodes nothing about sequence, concentration, purity, or evidence.

SNARE complex — the protein assembly, including SNAP-25, that drives fusion of neurotransmitter vesicles with the nerve terminal membrane. Target of both botulinum toxin and acetyl hexapeptide-8, by different chemistry and very different routes.

SNAP-25 — a SNARE protein; cleaved enzymatically by botulinum toxin type A, and the protein from which acetyl hexapeptide-8 is derived.

Neuromuscular junction — the synapse between a motor nerve terminal and a muscle fiber. It lies below the dermis, which is why route matters.

Botulinum toxin — a ~150,000-dalton protein from Clostridium botulinum that cleaves SNARE proteins and blocks acetylcholine release. Injected in tiny controlled quantities it produces localized, temporary muscle relaxation, with extensive randomized evidence for cosmetic and medical indications — and a boxed warning.

Vehicle control — a control arm receiving the identical formulation minus the active ingredient. The single most important design feature of a cosmetic efficacy study, because the vehicle is itself a moisturizer.

Split-face design — a study in which each participant receives active on one side of the face and control on the other. Controls for between-person variation; does not by itself solve blinding.

Profilometry — measurement of skin surface topography, often from a silicone replica, quantifying line depth. Strongly affected by hydration.

Cutometry — measurement of skin deformation and recovery under suction, reported as elasticity or firmness.

Cosmeceutical — an informal term for a cosmetic containing biologically active ingredients. No legal standing in the United States; there is no third category between cosmetic and drug.

Structure/function claim — a claim that a product affects the structure or any function of the body. Making one converts a cosmetic into a drug, which is why cosmetic marketing says "the appearance of" instead.

Spaced Review

1. (Chapter 4 + 30.2) Chapter 4 introduced the delivery question for oral and injected peptides. Restate it in the form it takes for a topical cosmetic peptide, then explain why a study reporting that a peptide was "detected in the skin" may not answer it. Which layer would the detection have to be in for the answer to be yes?

2. (Chapter 5 + 30.4 + 30.6) GHK-Cu receives two different ratings in §30.4, and botulinum toxin receives two in §30.6. Explain which of Chapter 5's six rules each pair illustrates, and why collapsing either pair into a single verdict would destroy the information a reader needs.

3. (Chapter 2 + 30.3 + 30.5) Chapter 2 explained how receptor binding produces amplified downstream effects. Both palmitoyl pentapeptide-4 and acetyl hexapeptide-8 have mechanisms that are coherent in those terms, yet they receive different ratings. Identify the variable that separates them, and state why it is not mechanism.

4. (30.8) You are shown a study: 45 participants, 12 weeks, split-face, active serum on one side and no treatment on the other, investigator grading plus corneometry, funded by the ingredient supplier. List the design's genuine strengths first, then its weaknesses, then state the single change that would most improve it and why that change matters more than doubling the sample size.

5. (Chapter 1 + 30.6 + 30.7) Chapter 1 argued that "natural" predicts nothing about safety, and §30.7 argued that cosmetic marketing vocabulary is a legal artifact. Using botulinum toxin as your example, write three sentences for someone with no science background connecting those two points: why the most natural molecule in this chapter is the most dangerous one, and why the product legally permitted to say the least about changing your body is the one that changes it most.