Only a handful of places left for our September 2026 cohort. Applications close soonApply now →

MSTA
Your future
starts here.

Collagen

Also known as: collagen, type I collagen, type III collagen, collagen supplements

Collagen is the dominant structural protein of the dermis, providing tensile strength. Types I, III and V predominate in skin, and it turns over extremely slowly — a half-life estimated in years, not weeks. Most figures quoted about its loss and replacement do not survive checking.

Evidence status

Moderate

The structural biology and slow turnover are well established. The widely quoted rates of age-related loss are traceable to a hypothesis paper or to menopause-specific data. For oral supplements the only review that stratified by funding found benefit in industry-funded trials and none in independently funded ones.

What collagen is#

Collagen is the dominant structural protein of the dermis, giving skin its tensile strength. Types I, III and V predominate.

Almost everything else the industry says about it — how fast it is lost, how quickly it can be rebuilt, what percentage anything increases it by — turns out to be harder to source than it sounds.

How slowly it turns over#

Using aspartic acid racemisation as a marker of how long a protein has resided in tissue, human skin collagen has an estimated half-life of about 15 years.[1]

That single fact should reframe how a practitioner talks about results. This is not a rapidly replenishing tissue. Anything you do to it is a slow, cumulative intervention on a structure that has been there for years and will be there for years.

One caution on a number you will see quoted: the same method estimated articular cartilage collagen at about 117 years. That figure is routinely attributed to skin. It is not skin.

The type percentages#

Of types I, III and V combined, adult skin is roughly 85–90% type I, 8–11% type III and 2–4% type V. In fetal skin the balance sits differently: 70–75%, 18–21% and 6–8%.[2]

The phrase “of types I, III and V combined” is doing real work. Without it, these read as a share of allskin collagen, which they are not — skin contains other collagen types as well. You will see the figures quoted without that qualifier constantly.

The fetal comparison is where the familiar “fetal skin is type III-rich, which is why it heals without scarring” story comes from. That is a reasonable observation about composition; it is not by itself an explanation of scarless healing.

How it is lost — and the 1% myth#

Every course teaches that you lose about 1% of your collagen a year after some age in your twenties. Here is where that comes from.

What we can and cannot show

The figure is usually traced to Shuster’s sun-protected forearm work. That abstract establishes the direction — skin collagen falls with age and is lower in women than in men at every age — but we could not read the full article to check its complete quantitative content.[3, 4]

The number itself appears in a 2005 paper in Medical Hypotheses, where the same author restates his earlier measurements in support of a theory linking skin and bone collagen loss: women have less collagen than men, and it decreases by 1% a year in exposed and unexposed skin.

What we are not going to tell you is that the number was invented there.This page said that until August 2026, and it was a confident historical accusation built on an unverified trace — the project’s own verification record had already rejected the reconstruction. What we searched for and could not find is a primary passage establishing one universal 1%-per-year rule across ages, sites and measures.[no source found] Later papers repeat it. Repetition is not derivation, and neither is our failure to find it proof of fabrication.

The figure that is properly sourced

A decline of roughly 1–2% per yearis well documented — but it tracks years since the menopause, not chronological age, alongside falls in skin thickness and bone mineral content.[5]

That is a genuinely useful clinical fact, and a different claim from the one usually made.

And a straight line is a poor description

In 45 sun-protected abdominal samples taken at autopsy from the periumbilical region, spanning one month to 95 years, collagen measures varied with age in a pattern better fitted by quadratic and cubic models than by a straight line.[6]

Read what that is, though. It is cross-sectional — different people at different ages, sampled once — so it describes how tissue differs across a population, not how much collagen any one person loses in a year. Cross-sectional age differences and within-person change are not the same quantity, and the paper does not claim to measure the second. It makes a fixed annual percentage a poor description; it does not refute every use of a linear approximation.

What actually degrades it#

A single ultraviolet exposure induces matrix metalloproteinases in human skin and increases degradation of endogenous type I collagen.[10]

That is the most useful sentence on this page for a consultation. Not cumulative years of sunbathing — one exposure, measurable enzymatic degradation. It makes daily photoprotection an argument about collagen preservation rather than about burning.

What is shown to build it#

The strongest single demonstration that a topical agent restores collagen formation in human skin: in photodamaged skin on the extensor forearm, daily 0.1% tretinoin for 10 to 12 months increased a marker of newly synthesised collagen I by about 80%, against a 14% decrease with vehicle (p=0.006) — 15 patients treated, 14 on vehicle.[11]

Three scope notes, because this study is universally cited as facial evidence for total collagen. The site was the forearm. The measurement was collagen I formation, assessed immunohistologically — not total collagen, and not appearance. And buttock skin was not treated: it was a sun-protected comparator, used to show that collagen I formation was 56% lower in photodamaged skin than in protected skin. It is a real and important result. It is a biomarker, on an arm.

Oral collagen: the funding problem#

This is the question every client asks, and the honest answer is more interesting than either of the usual ones.

Several meta-analyses report positive pooled results. Hydration improved with a standardised mean difference of 0.63 (95% CI 0.38–0.88) — pooled from 18studies — and elasticity 0.72 (95% CI 0.40–1.03) from 19, within a review of 26 trials and 1,721 participants. A separate review of 14 trials found a small wrinkle effect of −0.21 (95% CI −0.40 to −0.01), pooled from just five studies.[8, 9]

Denominators travel with outcomes, and this page previously reported one headline count for all three. There is a second confound worth more than the effect sizes: many of the pooled trials tested multi-ingredient products— collagen with vitamin C, zinc, biotin, astaxanthin, coenzyme Q10, hyaluronic acid or glucosamine. One review says it outright: improvements in the skin can be caused by the influence of not only collagen alone.[8, 9] So these pooled effects do not isolate collagen, and they do not tell you that one supplement behaves like another.

Those same reviews report substantial internal bias — 13 of 26 trials at risk from missing outcome data in one, seven of 14 at unclear risk of bias in the other.[8, 9]

Then someone asked who paid

Of four reviews, only one stratified its analysis by funding source. It covered 23 randomised trials and 1,474 participants, and the result is the single most important thing on this page:

Industry-funded trials showed significant benefit for hydration, elasticity and wrinkles. Independently funded trials showed no effect on any of the three.[7]

Its conclusion was that there is currently no clinical evidence supporting collagen supplements for skin ageing.

Notice what that does to the other reviews. They are not wrong — they pooled real trials and computed real numbers. They simply answered a different question, because they used risk-of-bias tools that do not ask who funded the work. The positive pooled effect and the negative stratified effect can both be true at once, and only one of them tells you what to expect for a client buying a sachet.

What you can legally claim#

This matters commercially if you retail supplements.

EFSA assessed a specific porcine collagen hydrolysate, submitted by its own manufacturer, and concluded that a cause-and-effect relationship had not been established between consuming it and improved skin function.[12]

And on the Great Britain Nutrition and Health Claims Register, eight nutrients carry an authorised skin claim: biotin, copper, iodine, niacin, riboflavin, vitamin A, vitamin C and zinc. Collagen is not one of them.[13]

Products that appear to make a skin claim are usually leaning on an authorised claim for an added vitamin — most often vitamin C — rather than for the collagen itself.

Two distinctions the shorthand collapses, and this page collapsed them until August 2026. Great Britain and Northern Ireland run different registers and legal routes for nutrition and health claims, so nothing should be inferred for either without checking the wording that applies where the product is sold. And the absence of an authorised healthclaim is not a ban on every cosmetic-appearance claim: those remain lawful in principle, and remain subject to CAP and consumer-law substantiation — which is an obligation to hold evidence, not a permission to assert.[13]i

Topical collagen#

The “500 Dalton rule” is a practical heuristicproposed in a review for passive delivery through intact stratum corneum — not an experimentally validated universal cutoff, which is how this page previously used it. Native collagen is orders of magnitude larger than that threshold, so nothing establishes that it reaches dermal fibroblasts from a topical product.[14]i

What the heuristic cannot tell you is what a particular finished formulation does on the surface. Humectancy and film formation are properties of a product— its vehicle, its molecular weight distribution, everything else in it — and cannot be assigned to the whole ingredient class. A collagen cream may well be a good humectant. That is a claim about that cream.

In professional practice#

  • Lead with the timescale. A ~15-year half-life sets honest expectations better than any before-and-after.
  • Stop quoting 1% a year. Say collagen declines with age, faster after the menopause, non-linearly.
  • Use the single-UV-exposure finding. It is the most persuasive true thing you can tell a client about sunscreen.
  • On supplements, be straight. Positive pooled results exist; independently funded trials found nothing; and you cannot legally claim skin benefits for collagen in the UK.
  • Never promise a percentage increase. See the microneedling entry for what happened to the 400% figure.

What remains uncertain#

  • Whether the industry/independent split reflects publication bias, design differences, or something else.
  • Whether ingested collagen peptides reach skin in amounts that could matter.
  • How much of the menopausal decline is oestrogen-specific and how much is ageing.
  • Whether any aesthetic procedure produces durable collagen change beyond the remodelling window.

Common misconceptions#

“You lose 1% of your collagen every year from 25.”

That figure appears in a hypothesis paper, not a measurement. The sourced 1–2% relates to years since menopause, and loss is non-linear anyway.[3, 4]

“Collagen has a 117-year half-life.”

That is articular cartilage. Skin is about 15 years.[1]

“Collagen supplements are proven to work.”

In industry-funded trials. Independently funded trials showed no effect on hydration, elasticity or wrinkles.[7]

“Collagen cream replaces lost collagen.”

Intact collagen is far too large to penetrate. It works as a humectant.[14]i

“Tretinoin is proven to build facial collagen.”

The landmark trial names forearm and buttock, not face. The result is real; the site is not what people assume.[11]

Frequently asked questions#

Do collagen supplements work?

The most rigorous review — and the only one that checked who funded the trials — found benefit in industry-funded studies and none in independently funded ones. You also cannot lawfully claim skin benefits for collagen in the UK.[7]

How fast does collagen actually decline?

Not at a fixed annual rate. It declines with age, non-linearly, and faster in the years following the menopause.[6]

What is the strongest evidence for building collagen?

Topical tretinoin over 10–12 months, with an 80% increase in collagen I formation — measured on forearm and buttock skin.[11]

Can I sell a collagen drink and say it helps skin?

No. Collagen has no authorised skin claim in Great Britain. Eight nutrients do, and vitamin C is the one most often carrying the claim on a collagen product.[13]

Why does sunscreen matter for collagen specifically?

Because a single UV exposure induces the enzymes that degrade type I collagen. It is not about accumulated decades — it is about today.[10]

References#

Each source is graded by evidence tier. Tier 4 material (manufacturer documents, expert consensus, practitioner experience) is useful for protocol and context, and is never presented as equivalent to independent clinical evidence.

  1. Verzijl N, DeGroot J, Thorpe SR, et al. Effect of collagen turnover on the accumulation of advanced glycation end products. Journal of Biological Chemistry. 2000;275(50):39027–39031.Tier 3Supports: Using aspartic acid racemisation as a marker of how long a protein has resided in tissue, estimated the half-life of human SKIN collagen at about 15 years, and of ARTICULAR CARTILAGE collagen at about 117 years. The 117-year figure is routinely mis-attributed to skin.
  2. Smith LT, Holbrook KA, Madri JA. Collagen types I, III, and V in human embryonic and fetal skin. American Journal of Anatomy. 1986;175(4):507–521.Ex vivoTier 3Supports: Proportions OF TYPES I, III AND V COMBINED. Adult skin: type I 85–90%, type III 8–11%, type V 2–4%. Fetal skin (5–26 weeks' gestation): type I 70–75%, type III 18–21%, type V 6–8%. The 'of types I, III and V combined' qualifier is load-bearing — without it the figures read as a share of all skin collagen.
  3. Shuster S, Black MM, McVitie E. The influence of age and sex on skin thickness, skin collagen and density. British Journal of Dermatology. 1975;93(6):639–643.Tier 3Supports: Forearm measurements in a large group of normal subjects. Establishes the DIRECTION only: skin collagen falls with age and is lower in women than men at every age. It reports no annual rate.
  4. Shuster S. Osteoporosis, a unitary hypothesis of collagen loss in skin and bone. Medical Hypotheses. 2005;65(3):426–432.Tier 4Supports: This is where the '1% a year' figure actually appears: 'women have less collagen than men and it decreases by 1% a year in exposed and unexposed skin'. Published in Medical Hypotheses, a journal for hypothesis-led argument, restating the author's earlier measurements in support of a theory linking skin and bone collagen.
  5. Brincat M, Kabalan S, Studd JW, Moniz CF, de Trafford J, Montgomery J. A study of the decrease of skin collagen content, skin thickness, and bone mass in the postmenopausal woman. Obstetrics & Gynecology. 1987;70(6):840–845.Tier 3Supports: Skin collagen, skin thickness, metacarpal index and forearm bone mineral content all declined by roughly 1–2% per year in postmenopausal women — tracking YEARS SINCE MENOPAUSE rather than chronological age.
  6. Marcos-Garcés V, Molina Aguilar P, Bea Serrano C, et al. Age-related dermal collagen changes during development, maturation and ageing — a morphometric and comparative study. Journal of Anatomy. 2014;225(1):98–108.Ex vivoTier 3Supports: Morphometric measurement across the lifespan found dermal collagen loss to be NON-LINEAR, following quadratic and cubic trajectories rather than a fixed annual percentage.
  7. Myung SK, Park Y. Effects of Collagen Supplements on Skin Aging: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. American Journal of Medicine. 2025.Tier 1Supports: 23 RCTs, 1,474 participants — and the only one of four reviews to stratify by who funded the trials. Pooled across all trials, collagen favoured hydration, elasticity and wrinkles. Split by funding source: INDUSTRY-FUNDED trials showed significant benefit on all three; INDEPENDENTLY FUNDED trials showed no effect on any of the three. Concluded there is currently no clinical evidence supporting collagen supplements for skin ageing.
  8. Pu S-Y, Huang Y-L, Pu C-M, et al. Effects of Oral Collagen for Skin Anti-Aging: A Systematic Review and Meta-Analysis. Nutrients. 2023;15(9):2080.Tier 1Supports: 26 RCTs, 1,721 participants, 2–12 weeks. Hydration SMD 0.63 (95% CI 0.38–0.88) across 18 trials; elasticity SMD 0.72 (95% CI 0.40–1.03) across 19. Did NOT stratify by funding. 13 of 26 trials were judged at risk of bias from missing outcome data, seven for deviations from intended intervention and two for selective reporting.
  9. Dewi DAR, Arimuko A, Norawati L, et al. Exploring the Impact of Hydrolyzed Collagen Oral Supplementation on Skin Rejuvenation: A Systematic Review and Meta-Analysis. Cureus. 2023;15(12):e50231.Tier 1Supports: 14 RCTs, 967 participants (92% female), 4–12 weeks. Wrinkle effect size −0.21 (95% CI −0.40 to −0.01, p = 0.04) — an upper confidence limit of −0.01 sits on the edge of no effect. Seven of the 14 trials were at unclear risk of bias. Did not stratify by funding.
  10. Fisher GJ, Wang ZQ, Datta SC, Varani J, Kang S, Voorhees JJ. Pathophysiology of premature skin aging induced by ultraviolet light. New England Journal of Medicine. 1997;337(20):1419–1428.Tier 2Supports: A single UV exposure induces matrix metalloproteinases in human skin and increases degradation of endogenous type I collagen fibrils.
  11. Griffiths CEM, Russman AN, Majmudar G, Singer RS, Hamilton TA, Voorhees JJ. Restoration of collagen formation in photodamaged human skin by tretinoin. New England Journal of Medicine. 1993;329(8):530–535.Tier 2Supports: In 29 patients using 0.1% tretinoin daily for 10–12 months, collagen I formation increased 80% versus a 14% decrease with vehicle (P = 0.006). NOTE: the published abstract does not name the treated site, and the only sites named in the paper are forearm and buttock — so this should not be described as facial data.
  12. EFSA Panel on Dietetic Products, Nutrition and Allergies. Scientific Opinion on the substantiation of a health claim related to VeriSol®P and change in skin elasticity leading to an improvement in skin function. EFSA Journal. 2013;11(6):3257.Tier 1Supports: Assessed VeriSol®P, a porcine collagen hydrolysate submitted by Gelita AG under Article 13(5), and concluded that 'a cause and effect relationship has not been established between the consumption of VeriSol®P and a change in skin elasticity leading to an improvement in skin function'. This is a product-specific opinion, not a general assessment of collagen.
  13. Great Britain Nutrition and Health Claims Register. 19 May 2026 edition.Tier 1Supports: Eight nutrients carry an authorised skin claim in Great Britain: biotin, copper, iodine, niacin, riboflavin, vitamin A, vitamin C and zinc. Collagen is not among them, and collagen joint claims are listed as non-authorised.
  14. Bos JD, Meinardi MMHM. The 500 Dalton rule for the skin penetration of chemical compounds and drugs. Experimental Dermatology. 2000;9(3):165–169.Tier 4Supports: Argues that compounds above roughly 500 Da do not reach therapeutic concentrations through intact skin, resting on three observations: common contact allergens fall under 500 Da, routine topical pharmacological agents fall under 500 Da, and transdermal delivery drugs fall under 500 Da.

Continue learning with MSTA#

Skinipedia is written by the educators at MSTA, the medic-led skincare training academy in Liverpool.