Evidence library

Every number. Its source. Our grade.

What each ingredient has actually shown in people, with panel size and duration. Grade A: multiple peer-reviewed trials. Grade B: a manufacturer in-vivo study with a stated panel. Grade C: small, old or secondary-source evidence; we use the ingredient but do not lean on the number.

Matrixyl 3000Evidence B

Palmitoyl Tripeptide-1 + Palmitoyl Tetrapeptide-7 · Sederma (Croda)

~20%reduction in wrinkle depth and volume, 2 months, n=24 women aged 40 to 65

Signal peptides that encourage skin to make more of its own structural proteins. Men in the same programme (n=39) saw about 10% depth and 30% density reduction.

Tested level: ~3% of blend, twice daily · In: Verglas, Thaw

Source: chemistconfessions.com

ArgirelineEvidence C

Acetyl Hexapeptide-8 · Lipotec (Lubrizol)

-30%wrinkle depth around the eyes after 30 days, n=10 (Blanes-Mira 2002)

A neuro-peptide that softens the look of expression lines. The original study is small and a 2025 review notes inconsistent significance, so we keep the claim to "the appearance of expression lines" and nothing stronger.

Tested level: 5 to 10% solution, twice daily · In: Verglas, Halflight

Source: www.mdedge.com

EyeserylEvidence B

Acetyl Tetrapeptide-5 · Lipotec (Lubrizol)

70%of volunteers showed reduced eye-bag volume after 28 days at 1%, n=17

A tetrapeptide tested specifically for under-eye puffiness. At 10% over 60 days (n=20) the effect was significant from day 14 and elasticity improved 30% by day 30.

Tested level: 1 to 10% solution · In: Halflight

Source: alfaspa.ua

HaloxylEvidence B

Chrysin, N-Hydroxysuccinimide, Palmitoyl Tripeptide-1, Palmitoyl Tetrapeptide-7 · Sederma (Croda)

-12.5%red colour parameter under the eye after 56 days, n=22 (p<0.01); 72% of volunteers improved

Targets the blood-pigment build-up that gives vascular dark circles their colour. Does nothing for hollows or melanin-based circles, and we say so.

Tested level: 2% · In: Halflight

Source: www.cosmacon.de

SYN-COLLEvidence C

Palmitoyl Tripeptide-5 · dsm-firmenich

84 daysdose-dependent wrinkle reduction vs placebo and vs a Matrixyl comparator, n=60 (Schagen 2017)

A collagen-supporting signal peptide. The percentages quoted around the web are secondary-source only, so we report the study design, not a number.

Tested level: 1 to 3% (guideline 2.5%) · In: Snowline

Source: www.mdpi.com

Copper Tripeptide-1Evidence C

GHK-Cu · Generic (multiple suppliers)

n=7112-week facial study (Leyden 2002, poster): increased density, reduced fine lines and laxity

The most famous peptide in skincare and the least supplier-substantiated. The evidence is literature-based rather than a manufacturer datasheet, which is why Verglas does not rest its claim on it alone.

Tested level: 1% pure basis · In: Verglas

Source: pmc.ncbi.nlm.nih.gov

NiacinamideEvidence A

Niacinamide (vitamin B3) · Generic

4 to 8 wkmultiple peer-reviewed trials at 2 to 5% on tone, barrier lipids and sebum

One of the best-evidenced cosmetic actives there is. We dose it at 4% in Verglas and 2% in Snowline, the range the literature actually tested, not the 10% arms race.

Tested level: 2 to 5% · In: Verglas, Snowline

Source: pubmed.ncbi.nlm.nih.gov

Ceramides NP, AP, EOPEvidence A

Ceramide NP, Ceramide AP, Ceramide EOP · Generic (Evonik and others)

3:1:1ceramide to cholesterol to fatty acid ratio associated with fastest barrier recovery in TEWL studies

Skin-identical lipids. Decades of barrier science say the ratio matters more than the headline percentage.

Tested level: 0.5 to 1% total with cholesterol and fatty acids · In: Snowline, Halflight, Thaw

Source: pubmed.ncbi.nlm.nih.gov

Our own study. Supplier data describes ingredients. Before our second production run we intend to commission an independent instrumental study on Verglas (target n=30, 8 weeks, hydration, texture and wrinkle depth) and publish it here in full, whatever it says.

Every number. Its source. Our grade.

First access. Founding price. Lab notes.