# Three signals, three stations, three very different evidence files

> Compare GHK-Cu, GLOW and BPC-157 — Recovery & Tissue Repair Research Peptides — renovapeptides — The three Recovery & Tissue Repair research peptides side by side: where each acts in the repair sequence, how fast each clears, how mature its evidence is, and how sharply their regulatory standing differs.

**SIDE BY SIDE**

GHK-Cu, GLOW and BPC-157 set against each other on mechanism, timescale, evidence maturity and regulatory standing — the four axes on which they actually diverge.

## Start here

The quickest way to keep these three apart is to stop ranking them and start placing them.

Tissue repair runs in a rough order: get blood to the injury, get repair cells to arrive and move, lay down structural material, then spend a long time rebuilding and cross-linking that material into something that holds. **BPC-157** is studied at the first step. **GHK-Cu** is studied at the last one. **GLOW** is a vial containing both of those plus a third peptide, sold on the idea of covering the whole run at once.

Placed that way, they are not competitors. They are different jobs, and the interesting comparisons are not about which is strongest but about how differently each is evidenced, how long each stays in the body, and how differently the law treats them. Those four comparisons are what this page does.

## The comparison at a glance

| | GHK-Cu | GLOW (research blend) | BPC-157 |
| --- | --- | --- | --- |
| **Station in repair** | Matrix synthesis and cross-linking — late | All three at once, in theory | Blood supply and cytoprotection — early |
| **What it is** | Tripeptide chelated to one copper(II) ion | GHK-Cu + BPC-157 + TB-500 in one vial | Fifteen-amino-acid gastric peptide |
| **Best-documented mechanism** | Fibroblast matrix synthesis; metalloproteinase/TIMP rebalancing; copper for lysyl oxidase cross-linking [4][9] | Its parts', assembled — none of its own [6] | VEGFR2 up-regulation with VEGFR2-Akt-eNOS signalling [8] |
| **Strongest human data** | Small topical trials; procollagen up in 70% of treated subjects [1]; one 45-man combination hair trial [3] | None as a blend [6] | Three pilot studies, one of them in two subjects [7][12] |
| **Strongest animal data** | Broad remodeling review across models [9] | Inherited from its parts [10][11] | Transected rat Achilles tendon [10]; rat gastric ulcer, 45.7-65.6% inhibition [14] |
| **Known pharmacokinetics** | No validated human PK systemically; topical dermal depot quantified [5] | None for the combination | Rats and dogs only; half-life under 30 min [13] |
| **Regulatory status** | Legal cosmetic ingredient topically; unapproved by any other route | Not approved; two constituents are unapproved research chemicals | Not approved; flagged by the FDA as not eligible for pharmacy compounding pending evaluation |
| **Anti-doping** | Not itemised on recent Prohibited Lists; the S0 catch-all may still apply | Prohibited — two constituents are banned | Prohibited at all times |
| **The honest one-liner** | Modest, real, and narrow | The best story and the least evidence | A deep rodent file and an almost empty human one |

## Mechanism: a relay, not a rivalry

The three mechanisms do not overlap much, which is why the blend exists at all.

BPC-157's documented activity is vascular. It up-regulates VEGFR2 and promotes its internalisation in endothelial cells, activating VEGFR2-Akt-eNOS signalling and raising vessel density in cell and animal models [8]. Additional reported routes — the FAK-paxillin complex, growth-hormone-receptor sensitisation in tendon fibroblasts, the nitric-oxide system — sit alongside that rather than replacing it.

TB-500, the third peptide inside GLOW, works on movement. Its parent protein sequesters G-actin and promotes cell migration, angiogenesis and reduced scarring; in a rat wound model, full-length thymosin beta-4 raised re-epithelialisation by 42% at day four and 61% at day seven, and as little as 10 pg stimulated migration two- to three-fold [11].

GHK-Cu works on the material itself. It stimulates fibroblasts to synthesise collagen, elastin, glycosaminoglycans and decorin, rebalances metalloproteinases against their TIMP inhibitors, and delivers the copper that lysyl oxidase needs to cross-link fibres [4][9].

Supply line, workforce, structure. As a division of labour it is coherent, and it is the one genuinely good argument for the blend. The argument has never been tested.

## Timescale: fast signals, slow structure

The three also differ on an axis the marketing never mentions: how long each is present, set against how long the process it is credited with actually takes.

BPC-157 is the extreme case. Formal characterisation in rats and dogs found linear pharmacokinetics, an elimination half-life under thirty minutes, intramuscular bioavailability of roughly 14-19% in rats and 45-51% in dogs, and rapid breakdown into small fragments entering ordinary amino-acid metabolism [13]. A molecule cleared that fast is not present across the weeks over which tendon or gut healing is reported to improve. Whatever it contributes has to be initiated quickly and then carried forward by the tissue.

Topical GHK-Cu behaves in the opposite way, at least in skin. Of the copper crossing dermatomed skin over 48 hours in a penetration study, 136.2 plus or minus 17.5 micrograms per square centimetre permeated while 97 plus or minus 6.6 micrograms per square centimetre was retained in the dermis as a depot [5]. A depot releases slowly, which suits a phase measured in weeks.

For GLOW the mismatch becomes the caution itself: three peptides with very different clearance rates — a small tripeptide, a short-lived pentadecapeptide, a thymosin fragment — are combined into a single injection whose joint kinetics have never been characterised.

And the process being targeted is slow throughout. Re-epithelialisation gains in the rat wound model were measured at day four and again at day seven [11]; the hair trial ran six months [3]; community accounts of skin change describe six to twelve weeks. Fast signals, slow structure. The two are easy to conflate and worth holding apart.

## Evidence maturity: the gap is the story

Ranked by controlled human data, the order is stark and does not match the order of enthusiasm.

GHK-Cu is highest, and still modest: small topical dermatology studies with placebo-controlled improvements in laxity, clarity, fine lines, wrinkle depth and density [4], a procollagen comparison favouring it over vitamin C and retinoic acid [1], one six-month randomised trial in 45 men of a combination formulation [3], and a quantified human skin-penetration study [5]. Real measurements in real people, in a narrow lane, with a mechanistic literature concentrated in a limited number of groups sitting above them.

BPC-157 is next, at a distance. Three pilot studies, no rigorous large-scale trial, and a formal recommendation that it be treated as investigational [7]. Its animal file is deep and its human file is nearly empty, and the size of that discrepancy is the single most useful thing to know about it.

GLOW is last, and not because its ingredients are unstudied. It is last because the vial is not what was studied. Every claim made for it is transferred from single-component research, and the one peer-reviewed review naming all three constituents together concluded that human safety data in this class are scarce and the potential for serious harm is real [6].

## Regulatory and anti-doping status diverges sharply

Legal standing separates these three more than any biological property does.

GHK-Cu is split by route. As topical Copper Tripeptide-1 it is a legal cosmetic ingredient in the US, EU and UK with a long safety record. Taken into the body by any other route it is an unapproved research compound with no validated human pharmacokinetics. It is not currently itemised on recent World Anti-Doping Agency Prohibited Lists, though the catch-all category for non-approved pharmacological substances can still apply, and the current list is the only reliable check.

BPC-157 has no approved therapeutic indication anywhere. In 2023 the FDA placed it in a category of bulk drug substances identified as not eligible for pharmacy compounding pending further evaluation, and it is prohibited in sport at all times under the non-approved-substances category.

GLOW takes the strictest status of its parts on every axis. Two of its three peptides are prohibited in sport at all times — BPC-157 under the non-approved-substances category and TB-500 as a fragment of thymosin beta-4, which is named among peptide hormones and growth factors — so the blend is off-limits to a tested athlete no matter how cosmetic the framing around it is [6]. It is also not a regulated product in any sense: ratios, purity and content vary by formulator and are unverified outside formal analysis.

## The caution that matters most for each

If only one caution per compound could be carried away, these are the three.

**For GHK-Cu, it is the route.** Topical cosmetic use has decades of consumer safety behind it and its worst common outcome is irritation. Injection is a completely different proposition with no validated human pharmacokinetic basis, and the community protocols that exist for it rest on nothing published. The evidence supporting the compound was collected almost entirely on skin.

**For BPC-157, it is the evidence base.** Not a specific hazard — the reported safety within the tiny human dataset and the animal work is reassuring as far as it goes. The problem is how far it goes: three pilot studies and no long-term human safety data mean the profile is genuinely unknown rather than established as benign [7][12].

**For GLOW, it is the untested combination.** Each caution belonging to a component still applies, and two more are added by the act of combining: joint pharmacokinetics that have never been characterised, and a blanket anti-doping prohibition inherited from two of the three peptides. A blend can be no better evidenced than its weakest-studied part, and its weakest-studied part is formally investigational [7].

Beneath all three sits the same structural fact from the 2026 musculoskeletal peptide review: favourable animal tissue-repair outcomes, scarce rigorous human safety data, and a category operating largely outside regulatory oversight [6].

---

renovapeptides is an independent reading desk for the tissue-repair literature: every figure here traces to a listed study, and none of it is a product, a protocol, or clinical advice.