Key takeaways
- In Miller’s 2006 randomized skin-care trial, 13 people completed follow-up after laser resurfacing, and objective measurements did not distinguish the two groups.
- Bo’s 2026 randomized study measured eyebrow hair count and diameter in 18 healthy adults, not scalp hair.
- Collagen measured in cultured cells is not the same outcome as skin appearance measured in people.
- The cited studies’ follow-up dates do not establish whether a visible change would occur, when it would begin, or whether it would last.
01Start with the randomized human trials
If you searched “what does ghk-cu do”, start with the comparison, not the claim. Random assignment helps make groups comparable. It does not make a small study conclusive or turn one study setting into evidence for another.
Skin appearance, Miller et al., 2006 (T1). After carbon dioxide laser resurfacing around the mouth, researchers randomly assigned people to skin-care regimens with or without GHK-Cu. Thirteen people completed the study. There was no statistically significant difference between groups on objective measures of how soon redness resolved, or on evaluators’ ratings of wrinkles and overall skin quality. That finding is about the objective measures, not every measure in the trial. It is also a comparison within post-laser care, not a test of ordinary skin care. Not medical advice.
Clinical wound measurements, Mulder et al., 1994 (T1). A separate human trial at multiple centers randomized participants to a GHK-Cu formulation or placebo alongside standard wound care. Evaluators did not know who was in each group. The study measured wound closure and infections in a specific clinical setting; the accessible abstract does not state the sample size. This is a different question from whether skin looks different after laser resurfacing or in ordinary skin care.
02GHK-Cu evidence by claim area
Each row pairs a claim area with a study that measured something relevant. The tier labels describe the design and source, not how well the claim holds up. See how 6AM Research grades evidence.
| Claim area | Study and year | Model, sample, tier | What was measured | What the comparison cannot show |
|---|---|---|---|---|
| Skin appearance | Miller et al., 2006 | Human randomized trial; 13 completers (T1) | Redness, wrinkles, and skin quality after laser resurfacing | Whether the same findings apply outside post-laser care |
| Hair | Bo et al., 2026 | Human randomized split-face study; 18 healthy adults (T1, small) | Eyebrow hair count and diameter | What happens on the scalp |
| Collagen | Maquart et al., 1988 | Cultured fibroblasts; sample size not stated in the abstract (T4) | Collagen production by connective-tissue cells | Whether collagen or visible skin appearance changes in people |
| Wound tissue | Maquart et al., 1993 | Rat wound-chamber model; 10 rats per group in the collagen comparison (T3) | Collagen content in tissue collected from the chambers | What that tissue measurement means for wound closure in people |
| Lung tissue | Park et al., 2016 | Mouse lung model; 35 mice, 7 per group (T3) | Microscope-based scores of lung tissue | Whether the finding applies to people |
03What the follow-up dates mean
A follow-up visit tells you when the researchers checked. It does not tell you whether a change occurred, when it began, whether an individual would notice it, or whether it would persist.
Badenhorst et al. (2016) enrolled 40 women and analyzed 39 (T2 for this comparison). They measured wrinkle depth and volume. The control serum lacked both the peptide and its carrier, so the comparison cannot separate the contribution of one from the other.
Miller et al. (2006) assessed wrinkles and overall skin appearance at 12 weeks, with 13 completers (T1). The visit was an assessment point, not an estimate of when a difference would become visible.
Bo et al. (2026) used a double-blind, split-face design: each of the 18 adults served as their own comparison (T1, small). The conference proceedings describe eyebrow measurements, not a timetable for scalp hair.
A separate 2016 randomized scalp study reported 45 men (T1, small), but it tested copper-free GHK chemically joined to 5-aminolevulinic acid. It did not test GHK-Cu.
04What GHK-Cu stands for
GHK is glycyl-L-histidyl-L-lysine, a chain of three amino acids: glycine, histidine, and lysine. GHK-Cu is that peptide bound to copper(II). The research history goes back to papers on GHK in 1973; a 1982 solution study described how copper binds to the peptide. The date of the early GHK work is not the date of a human trial of the copper complex.
Binding copper is chemistry, not evidence of a visible effect. If you want to check which compound a paper names, start with the GHK-Cu guide. If the question is what a batch report actually tested, see how to read a certificate of analysis. A batch report and a clinical trial answer different questions.
05What we don’t know
- Scalp hair. The eyebrow trial does not show what GHK-Cu alone does on the scalp, and the scalp study cited here tested a different compound.
- Timing and durability. The studies cited here do not establish whether a visible change would occur, when it would start, or whether it would last beyond follow-up.
- Effects beyond the model tested. Collagen in cultured cells, collagen in rat wound chambers, and lung tissue scores in mice do not establish a corresponding effect in people.
- Long-term safety. The short cosmetic studies summarized here cannot establish long-term safety.
This is a plain-English summary of the cited studies, checked on 21 November 2026. Not medical advice.