Collagen is the most abundant protein in animals. In humans it makes up about a third of all protein and three quarters of the dry weight of skin[1]. Collagen peptides, also sold as hydrolysed collagen or collagen hydrolysate, are this protein cut into short pieces, and they are taken as a supplement in the hope of supporting skin, joints and muscle.
The idea sounds simple: eat collagen to replace collagen. But swallowed collagen is not carried to the skin and built into it. Like any other protein it is digested, and most of it ends up as free amino acids. This page follows what happens between the jar and the skin. It starts with the protein itself, then looks at how the supplement is made and at the few collagen fragments that do reach the blood intact, and ends with what clinical trials have found and why researchers read them so differently.
| Parent protein | Collagen: about a third of human protein, three quarters of the dry weight of skin |
|---|---|
| Repeating unit | Gly-Xaa-Yaa, with Xaa often proline and Yaa often hydroxyproline (Hyp) |
| The supplement | Peptides of about 3–6 kDa made by heating collagen and cutting it with enzymes |
| Main sources | Cattle and pig hide and bone, fish skin and scales |
| Found intact in blood | Mainly the dipeptide Pro-Hyp and the dipeptide Hyp-Gly |
| Clinical evidence for skin | Most consistent for skin hydration; disputed overall because of trial quality and industry funding |
Collagen itself
There is no single collagen. Vertebrates have 28 types, built from at least 46 different polypeptide chains[1]. Type I is the main collagen of skin, bone, teeth and tendons, type II of cartilage, type III of skin, muscle and blood vessels, and type IV forms the thin basement membranes that underlie tissues[2]. All of them share one structure. Three chains, each a left-handed helix of the polyproline II type, wind around one another into a right-handed triple helix[1]. A single chain of type I collagen has about 1,000 amino acids and a mass near 100 kDa. The assembled molecule weighs about 300 kDa and is a rod some 280 nm long and only 1.4 nm thick[2].
The triple helix explains collagen's unusual composition. The three chains are packed so tightly that only the smallest amino acid, glycine, fits at the centre, so every third residue has to be glycine. The chain therefore consists of repeats of Gly-Xaa-Yaa. The Xaa position is most often proline (28% of the time) and the Yaa position most often 4-hydroxyproline, Hyp (38%), and Pro-Hyp-Gly is the most common triplet in collagen, at 10.5%[1]. In total, glycine accounts for about a third of the residues, and proline with hydroxyproline for about a fifth[2]. The structure was worked out in the 1950s, first by Ramachandran and Kartha in 1954 and in its accepted form by Rich and Crick in 1955[1].
Hydroxyproline is not one of the 20 amino acids built into proteins by the genetic code. After the chain is made, the enzyme prolyl 4-hydroxylase adds a hydroxyl group to proline residues in the Yaa position[1]. The extra group makes the triple helix much more stable to heat, through an electronic effect on the shape of the proline ring rather than through extra hydrogen bonds with water[1]. Glycine is just as essential: mutations that replace a single glycine in the repeat are a cause of osteogenesis imperfecta, an inherited disease of brittle bones[1]. Because Hyp occurs almost only in collagen, it also serves as a marker: when hydroxyproline-containing peptides appear in the blood, they must come from collagen[3]. That fact becomes important below.
From hide to powder
Collagen for supplements comes from the parts of animals that are richest in it. The main commercial source is cattle, because it is plentiful; pig collagen is also widely used. Fish skin, scales and bones are a growing source, alongside smaller ones such as chicken and duck[2]. The choice is not only technical. Fears of bovine spongiform encephalopathy and swine influenza, and religious restrictions on pork and beef, have increased interest in marine collagen[2].
To make the supplement, the collagen is first heated. Above about 40 °C the triple helix comes apart into three separate chains with no fixed shape. Proteolytic enzymes such as alcalase, papain or pepsin then cut the chains into peptides, mostly of 3–6 kDa, roughly 30–60 amino acids[2]. Chemical hydrolysis with acid or alkali is also possible but is corrosive and leaves a lot of salt[2]. Besides supplements, hydrolysed collagen is used in foods, cosmetics and medicines[2].
One consequence is easy to miss. After this treatment the product no longer contains the triple helix at all, only its broken chains. When a label names a collagen "type", it describes the tissue the collagen came from.
What reaches the blood
Peptides of 3–6 kDa are still far too large to be absorbed. In the gut they are broken down further, mostly to free amino acids, and the level of amino acids in the blood rises after a dose of collagen hydrolysate[3]. This is the fate of most dietary protein. It is also why many claims about "bioactive" food peptides fail. Peptides that work in a test tube often reach the blood at only nanomolar levels, far below the micromolar to millimolar concentrations they need to act[3].
Collagen is an exception, and hydroxyproline is what showed it. Kenji Sato of Kyoto University and colleagues looked for peptides that actually appear in human blood after a meal of collagen hydrolysate, rather than for peptides that work in cell culture[3]. After 10 g of hydrolysate, peptide-bound hydroxyproline in plasma reached about 20 µM at 60 minutes and was still high at 180 minutes. That is more than 10,000 times the nanomolar levels reported for other food-derived peptides[3].
Two dipeptides account for most of this: Pro-Hyp, the most abundant, and Hyp-Gly. No significant amounts of peptides longer than three residues were found in the blood[3]. Both dipeptides resist the peptidases of the blood, and both come straight from the collagen repeat, where the motif Gly-Pro-Hyp-Gly is common[3]. Pro-Hyp appears to enter through the gut lining by the proton-coupled carrier PEPT1, the same transporter that takes up other dipeptides[3].
Pro-Hyp is not unique to supplements. The body makes it when it breaks down its own collagen, and it has been found at sites of inflammation and wound healing in mice that were given no collagen at all[3]. Experiments with skin cells suggest what it might do there. Fibroblasts, the cells that make collagen in skin, do not grow on a gel of collagen fibres even when serum is added. Pro-Hyp at 200 µM makes them grow, in a dose-dependent way, while free proline and hydroxyproline do not; Hyp-Gly has a similar effect. Pro-Hyp also raises the production of hyaluronic acid by fibroblasts, and Pro-Hyp applied to wounds in mice speeds their healing. Sato infers that Pro-Hyp, whether from food or from the body's own collagen, promotes wound healing by stimulating fibroblasts[3].
Sato concluded that these peptides can "partially explain" the effects reported for collagen hydrolysate[3]. Two cautions apply. The 200 µM used in the cell experiments is about ten times the peak plasma level after a 10 g dose, and that peak counts all hydroxyproline-containing peptides together, of which Pro-Hyp is only a part. And no study has yet shown that this mechanism works in human skin: a 2026 meta-analysis describes it as proposed but not fully established in living tissue[4].
Skin: the trials
Skin is where most of the clinical work has been done, and where the evidence is most disputed. Between 2025 and 2026 three meta-analyses of randomised controlled trials were published, drawing on largely overlapping trials and reaching different conclusions.
The first, by Myung and Park (2025), pooled 23 trials with 1,474 participants[5]. Taken together, the trials showed significant improvement in skin hydration, elasticity and wrinkles. The authors then divided them. Thirteen trials had a low risk of bias, and in these there was no significant effect on any outcome. Nor was there an effect in trials not funded by companies. Only the low-quality, industry-funded trials showed benefit. The authors concluded that "there is currently no clinical evidence to support the use of collagen supplements to prevent or treat skin aging"[5].
Manufacturers disputed the analysis. A supplier and two industry groups argued that the funding classification was wrong for some trials, with studies labelled independent despite links to companies, that it combined different collagen products, doses and mixtures with other ingredients, and that the pooled result showing benefit sat oddly with a conclusion of no evidence[6]. These objections come from an interested party, but some of them are questions about method that a reader can check.
The second, by Batool and colleagues (2026), was larger: 35 trials with 2,534 participants[4]. It found clear improvement in three measures of skin function, each given as a standardised mean difference (SMD), the gap between groups in units of standard deviation: hydration (SMD 0.44), elasticity (SMD 0.62) and water loss through the skin (SMD −0.39)[4]. Hydration improved from about four weeks, while elasticity and the skin barrier needed about twelve. For wrinkles and dermal density, the apparent benefit disappeared in sensitivity analyses; the authors attribute it to outlying trials and heterogeneity[4]. On the funding question they reached the opposite view to Myung and Park: independently conducted trials showed effects comparable to sponsored ones[4]. They also found widespread signs of publication bias, with small negative studies apparently missing, so the pooled effects are probably somewhat overstated[4].
The third, by Nukaly and colleagues (2026), looked at peptides for skin ageing more broadly; 17 of its 19 trials used peptides taken by mouth[7]. Peptides improved hydration and had a modest pooled effect on wrinkles, both driven mainly by the oral trials; effects on elasticity and density were inconsistent, and heterogeneity between trials was extreme[7]. The authors judged the data insufficient for definitive conclusions[7].
Read side by side, the three analyses agree more than their headlines suggest. All three find the clearest signal for skin hydration and the weakest for wrinkles and density. Where they differ is in how much weight to give to trial quality and funding, and that decides whether the signal counts as evidence. The trials themselves explain part of the uncertainty. They lasted 4 to 16 weeks, most of them 12; daily doses ranged twentyfold, from 0.5 g to 10 g; most were done in South Korea, Japan, China, Thailand and Taiwan, and most recruited women[4]. Whether there is a dose–response relationship, or a dose above which more collagen adds nothing, is not known[4].
Muscle recovery
Collagen is also sold for recovery after exercise. A 2024 review collected eight randomised trials with 286 participants[8]. Their results conflict. In one trial, twelve weeks of collagen peptides in untrained men reduced the rise in blood markers of muscle damage (creatine kinase, lactate dehydrogenase and myoglobin) after exercise; in two others there was no difference from placebo[8]. Of four trials that measured muscle soreness, only one found a significant reduction[8]. Some trials gave collagen together with vitamin C or whey protein, which makes the effect of collagen itself harder to isolate[8]. The review's authors acknowledge the conflict in their results section, yet their conclusion describes the literature as consistently supportive[8]. The trials themselves support a more cautious reading.
Safety and open questions
In the skin trials, collagen peptides were well tolerated. No serious adverse events were reported, and mild stomach discomfort occurred in isolated cases[7]. Most trials, however, lasted three months or less[4], so they say little about years of daily use. The animal source matters to some buyers for religious reasons, and fears about diseases carried by cattle and pigs are one reason producers have turned to fish[2].
The larger open questions are about effect, not safety. Products differ in source, peptide size, dose and added ingredients, and trials have rarely compared them directly[4]. The mechanism that links Pro-Hyp to skin cells has been shown in cell culture and in mice, not in human skin[3][4]. And the evidence that does exist is shaped by who pays for it and which results get published, which is exactly what the meta-analyses disagree about.
So the answer to the question this page started with is partly known. Most of the collagen in a supplement is digested like any protein. A small part reaches the blood as Pro-Hyp and Hyp-Gly, and in the laboratory these dipeptides stimulate skin cells, although at concentrations well above those measured in plasma. Whether that translates into visibly better skin is still contested, with the most consistent evidence for hydration and little for wrinkles.
In short
- Collagen is a triple helix built from Gly-Xaa-Yaa repeats rich in proline and hydroxyproline; it makes up about a third of human protein.
- Collagen peptides are made by heating collagen from cattle, pigs or fish and cutting it with enzymes into peptides of about 3–6 kDa; the triple helix is gone.
- Most of a dose is digested to amino acids, but the dipeptides Pro-Hyp and Hyp-Gly reach the blood at micromolar levels; at higher concentrations they stimulate skin fibroblasts in cell culture.
- Meta-analyses of skin trials agree best on hydration and find little for wrinkles; they disagree on whether the benefit survives once trial quality and industry funding are taken into account.
- Evidence for muscle recovery is conflicting; the trials are short, mostly 12 weeks, and products and doses vary widely.
References
- Shoulders MD, Raines RT (2009). Collagen structure and stability. Annual Review of Biochemistry 78:929–958. DOI: 10.1146/annurev.biochem.77.032207.120833
- León-López A, Morales-Peñaloza A, Martínez-Juárez VM, Vargas-Torres A, Zeugolis DI, Aguirre-Álvarez G (2019). Hydrolyzed collagen — sources and applications. Molecules 24:4031. DOI: 10.3390/molecules24224031
- Sato K (2017). The presence of food-derived collagen peptides in human body — structure and biological activity. Food & Function 8:4325–4330. DOI: 10.1039/c7fo01275f
- Batool A, Abbas MS, Farooq R, Akbar MA, Ahmed A, Kanwal W, Ali H, Ahmad A, Junaid M, Hanif MS, Nunes Junior EBD, Aminpoor H (2026). Oral collagen peptides and skin rejuvenation: a systematic review and an updated meta-analysis of randomized controlled trials. Journal of Cosmetic Dermatology 25:e71041. DOI: 10.1111/jocd.71041
- Myung S-K, Park Y (2025). Effects of collagen supplements on skin aging: a systematic review and meta-analysis of randomized controlled trials. The American Journal of Medicine. DOI: 10.1016/j.amjmed.2025.04.034
- NutraIngredients (26 August 2025). Industry pushes back on meta-analysis concluding collagen supplements show no benefit for skin aging. nutraingredients.com
- Nukaly HY, Halawani IR, Irtaza HM, Alturkistani T, Serafi MR, Alhawsawi W, Bogari HO, Ahmed FA, Alhaddad Y, Shadid A, Alharithy R, Jfri A (2026). Oral and topical peptides for skin aging: systematic review and meta-analysis of randomized controlled trials. Frontiers in Medicine 13:1618306. DOI: 10.3389/fmed.2026.1618306
- Inacio PAQ, Gomes YSM, de Aguiar AJN, Lopes-Martins PSL, Aimbire F, Leonardo PS, Sá Filho AS, Lopes-Martins RAB (2024). The effects of collagen peptides as a dietary supplement on muscle damage recovery and fatigue responses: an integrative review. Nutrients 16:3403. DOI: 10.3390/nu16193403