COMPARISON
Four Origins, Measured by the Distance Travelled
NAD+, KPV, PT-141 and GHK-Cu lined up on the dimensions that matter when the selling point is that the body already makes it.
Abstract, in plain English
All four compounds catalogued here can honestly claim a connection to something the body produces. That is where the similarity ends.
NAD+ is chemically identical to the endogenous molecule — no modification at all. GHK-Cu is an exact human sequence lifted out of collagen, but delivered on a schedule and by a route the body never uses for it. KPV is a three-amino-acid piece of a hormone, unmodified but incomplete, and it behaves differently from the hormone it came from. PT-141 is that same hormone's core deliberately rebuilt so it would survive long enough to work as a medicine.
Ranked by human evidence, the order comes out almost backwards from the order by naturalness. The most modified compound in the set has two Phase 3 trials and an approved label. The completely unmodified one has trials that move a blood marker and not much else. The unmodified fragment has no human trial at all. Origin, in other words, predicts neither how well a compound has been tested nor how well it works.
The comparison matrix
| Dimension | NAD+ | KPV | PT-141 | GHK-Cu |
|---|---|---|---|---|
| Endogenous parent | None needed — the molecule itself is endogenous | Alpha-MSH, residues 11-13 (C-terminus) | Alpha-MSH core, residues 4-10, rebuilt | GHK sequence inside type I collagen and SPARC |
| Native role of the parent | Redox carrier plus consumed substrate for sirtuins, PARPs and CD38 [4] | Pigment signalling and anti-inflammatory signalling in one hormone | Same hormone as KPV, different region | Released locally when collagen is degraded at injury |
| What the manufactured version changes | Nothing structural; only route, amount and timing | Subtraction only — ten residues removed, nothing added | Cyclisation, lactam bridge, Nle and D-Phe substitutions [17] | Copper chelation plus topical application to intact skin |
| Why the change was made | To raise a level that falls with age | To keep anti-inflammatory action without pigmentary action [10] | To defeat rapid enzymatic breakdown of the native hormone | To stabilise the peptide and enable delivery [18] |
| Route with the best evidence | Oral precursors, NMN and NR [2][5] | None validated in humans; carriers under study [7][6] | Subcutaneous injection [15] | Topical, and poorly at that [18][22] |
| Human evidence ceiling | RCTs moving blood NAD+ and one insulin-sensitivity endpoint [5][3] | Zero published human trials | Two Phase 3 RCTs plus a 52-week extension [15][16] | Small topical dermatology trials plus one 45-man hair study [20][21] |
| Best-supported quantitative result | Whole-blood NAD+ up 142% at the highest NR dose [5] | 8 of 8 rabbit corneas re-epithelialised by 60 hours versus none on placebo [11] | FSFI-desire +0.35 versus placebo, P<.001, over 24 weeks [15] | Hair count +71.5 versus +9.6 for placebo over 6 months [20] |
| Sharpest caution | Blood level rises without established clinical benefit [1] | No human safety data of any kind exists | Nausea in about 40% over long-term use [16] | Barely crosses intact skin, clogP −2.24 [18] |
Origin: four different things the word endogenous can mean
The word is doing four different jobs across this set, and collapsing them is the first mistake.
For NAD+ it means the identical molecule. Nothing is being imitated; a cell that takes up a precursor makes the same coenzyme it would have made anyway.
For KPV it means a subsequence of a hormone. The three residues exist inside alpha-MSH. Whether that tripeptide circulates freely after natural proteolysis is a separate and much less settled question than whether the sequence exists inside a human protein.
For GHK-Cu it means a fragment liberated by tissue damage. The sequence sits inside type I collagen and appears when collagen is broken down — a local, event-driven release rather than a secreted signal.
For PT-141 it means descended from, and no more than that. Bremelanotide is not a piece of alpha-MSH; it is a redesign of its core with a lactam bridge and two unnatural amino acids [17]. The lineage is real and the molecule is new.
Only the first of those four supports the intuitive claim that the body is receiving something it already recognises in the form it already recognises it.
Evidence maturity runs against the naturalness ranking
Sorting the four by how little they were modified gives: NAD+, then KPV, then GHK-Cu, then PT-141. Sorting them by the strength of human evidence gives almost the reverse.
PT-141, the most heavily engineered, has the deepest human record — two identical Phase 3 randomised trials in 1,267 women meeting their co-primary endpoints [15], a 52-week open-label extension in 684 women [16], a mechanistic fMRI crossover study [14] and a regulatory label with published pharmacokinetics [17].
GHK-Cu has small controlled topical work, including one 45-participant six-month hair-count study of a combination formulation [20] and comparative procollagen data [21][18].
NAD+ has good randomised evidence for the pharmacodynamic marker and thin evidence past it — precursors reliably raise blood NAD+ [5][2], one trial reached a clamp-measured insulin-sensitivity endpoint [3], and the 2025 synthesis still described human efficacy in ageing as limited [1].
KPV has no human trial at all. Its literature is cell culture and mouse colitis models [8][9][6][7], with one rabbit corneal study [11].
The correlation is not merely absent, it is close to inverted — and the reason is unglamorous. Modification is what makes a molecule stable enough, patentable enough and commercially defined enough to be run through an expensive trial programme. Being natural does not attract that investment. It attracts marketing instead.
Where the fallacy actually bites
The failure mode is not that people believe something false about chemistry. It is that a true statement about origin gets used to answer a question it cannot answer.
The body makes it is true of all four here and settles none of the following: whether the compound reaches the tissue where it acts, at what concentration, how often, for how long, what else it binds on the way, and whether any of that changes an outcome a person would notice. Every documented problem in this catalogue lives in that list, not in the molecule's pedigree. Pigmentary change with frequent bremelanotide dosing is the endogenous receptor family behaving normally at an unphysiological frequency [17]. GHK-Cu's central obstacle is a permeability coefficient [18][22]. KPV's is that a free tripeptide does not survive long enough to be studied in a person. NAD+'s is that a blood level is not a tissue level [1].
The compensating insight is worth stating too, because an overcorrection would be its own error. Endogenous origin is genuinely useful information: it identifies a receptor system that already exists, a physiological role that has been characterised, and a plausible reason to look. That is a good reason to run an experiment. It is not a result.