SkincareProduct FormAug 2026

Collagen Peptides: Hydrolysed Powder, Absorption, and Whether Type Matters

The tub says collagen peptides, hydrolysed, type I and III, marine. Four claims, and almost nobody explains what any of them mean. This page is about the product form only: what hydrolysis does to the protein, how small the fragments are, whether the source animal changes anything you can measure, and what human blood tests find after a dose.

Pro-Hyp

Main fragment found in blood

1 to 2 h

Peak after a single dose

2.5 to 10 g

Hydrolysate dose in reviewed trials

Scope: this page answers form questions: hydrolysis, peptide size, source animal, and absorption. For what collagen is, what it does in skin, and whether swallowing it helps at all, the collagen guide covers that separately and is the better starting point if that is your question.

What Hydrolysed Means

Collagen, gelatin and collagen peptides are the same protein at three stages of being taken apart. Native collagen is a long triple helix, tough and insoluble in cold water, which is why you cannot stir a piece of hide into a glass of water. Processing with heat unwinds that helix into gelatin, which dissolves in hot water and sets into a jelly when it cools. Enzymes then cut the gelatin chains into much shorter pieces. That last product is hydrolysed collagen, and hydrolysed collagen is what the word peptides on a supplement label refers to.

The behaviour in water is the easiest tell. A hydrolysate stirs into a cold drink and does not set. Gelatin needs heat, then sets as it cools. If a powder you own turns your drink into jelly overnight, it is behaving like gelatin rather than like a hydrolysate, whatever the front of the pack says.

The one thing to take from this: peptides on a label is a claim about fragment size. It is not a different ingredient, not a special type, and not evidence of quality on its own. Every collagen powder that dissolves cold has been hydrolysed, including the cheapest one on the shelf.

What Reaches The Blood

This is the part people argue about, and it has actually been measured in humans. In a 2005 study, healthy volunteers fasted for 12 hours and then took 9.4 to 23 grams of gelatin hydrolysate made from porcine skin, chicken feet or cartilage. Before intake, the peptide form of hydroxyproline in blood was negligible. Afterwards it rose to a maximum of 20 to 60 nmol per mL of plasma at 1 to 2 hours, then fell to half that peak by 4 hours. The major constituent identified was Pro-Hyp, a two-unit fragment, alongside small but significant amounts of Ala-Hyp, Ala-Hyp-Gly, Pro-Hyp-Gly, Leu-Hyp, Ile-Hyp and Phe-Hyp.

So the flat claim that collagen is simply destroyed into single amino acids like any other protein is not quite right. Some of it circulates as intact di- and tripeptides. Hydroxyproline is the marker these studies track, because it is concentrated in collagen and uncommon in other food protein, which makes a rise in blood traceable to what was swallowed.

The size of that fraction is worth knowing before anyone oversells it. A 2007 study measured free hydroxyproline and peptide-bound hydroxyproline separately over 24 hours and found the peptide-bound share was roughly 30 percent of all hydroxyproline detected. The other 70 percent had been broken down completely. Collagen peptides do not travel through you untouched.

Where this evidence stops: these studies measured fragments in blood. Finding Pro-Hyp in plasma is not the same as showing it arrives in skin, and it is not the same as showing it changes anything once there. Those studies did not test that, and a supplement page that jumps from a blood measurement to a skin benefit has skipped the step that matters.

Does Peptide Size Matter

Premium products advertise low molecular weight, measured in daltons, or a high tripeptide content. The logic is easy to follow: since what circulates is mostly di- and tripeptides, a hydrolysate already cut down to that size has less work left to do. That is a reasonable argument. It is not yet a proven one, and the difference matters when you are paying two or three times more for it.

A 2019 systematic review of dermatological applications included 11 randomised placebo-controlled trials with 805 patients. Eight used ordinary collagen hydrolysate at 2.5 to 10 grams a day for 8 to 24 weeks, across pressure ulcers, dry skin, skin aging and cellulite. Two used a collagen tripeptide at 3 grams a day for 4 to 12 weeks and reported improvement in elasticity and hydration. One used a collagen dipeptide and suggested the effect tracked with dipeptide content. The same review noted that regulation is lacking on quality, absorption and efficacy, which is the honest frame for the whole category.

The most quoted size trial is a 2018 Korean study of a low molecular weight collagen peptide with a tripeptide content above 15 percent, including 3 percent Gly-Pro-Hyp. Sixty-four women aged 40 to 60 with photoaged skin took either placebo or 1000 milligrams once daily for 12 weeks. Hydration was significantly higher than placebo at both 6 and 12 weeks. At 12 weeks the visual wrinkle score and three wrinkle parameters were significantly better than placebo. Elasticity was more mixed and worth reading carefully: one of three elasticity parameters improved from baseline at 12 weeks, while two of three were higher than placebo.

One gram is a small fraction of the 2.5 to 10 grams used in the ordinary hydrolysate trials, and that gap is the strongest hint that fragment size does something real. Read the limits with it. It was a single trial of one proprietary ingredient, three of the five authors are listed with a commercial affiliation, two at Newtree Co., Ltd. and one at the research institute of Jellice, a gelatin and collagen manufacturer, and there was no arm giving a standard hydrolysate at the same 1000 milligrams, which is the comparison that would settle it.

What is missing: the searches behind this page did not turn up a human trial that randomised the same people to a high molecular weight and a low molecular weight hydrolysate at an identical dose. Until that exists, treat a low-dalton claim as a plausible engineering argument, not a demonstrated advantage, and price it accordingly.

Marine, Bovine, Porcine, Chicken

Marine collagen is sold at a premium almost everywhere, usually with a line about superior absorption. There is a real study behind that line, and it does not say what the marketing implies.

The 2007 study was a single-blind crossover in five healthy male volunteers, each taking type I gelatin hydrolysates from fish scale, fish skin and porcine skin on separate occasions after a 12 hour fast. The total area under the concentration-time curve for the fish scale group was significantly higher than for the porcine skin group. The peptide mix also differed: Ala-Hyp, Leu-Hyp, Ile-Hyp, Phe-Hyp and Pro-Hyp-Gly were detected only with the fish scale or fish skin hydrolysates, and Ala-Hyp-Gly and Ser-Hyp-Gly only with fish scale. The authors concluded that both the quantity and the structure of hydroxyproline-containing peptides in blood depend on the source of the gelatin.

Read that population line again: five healthy men, and no bovine arm. The study compared fish against pork. It is routinely cited as proof that marine beats bovine, and it did not test bovine at all. It is also a blood measurement, not a skin outcome.

The larger picture comes from a 2023 meta-analysis of 26 randomised controlled trials covering 1,721 participants. Hydrolysed collagen improved skin hydration and elasticity against placebo. In the subgroup analyses, the effect on hydration varied with the source of collagen and with how long people supplemented for. However, the same analysis found no significant difference between sources for elasticity, at p equal to 0.21. The authors also reported that they identified several biases in the included trials and called for larger randomised studies before treating any of it as settled.

Put plainly: source appears to change what shows up in blood, the pooled human evidence for source changing a skin outcome is partial and inconsistent between measures, and nobody has demonstrated that marine is worth a large price premium over bovine on evidence.

Source	Usual raw material	Non-efficacy reasons it matters
Marine	Fish skin and fish scale	A fish-derived food. Relevant to anyone with a fish allergy, and to pescatarian diets.
Bovine	Cattle hide and bone	Excluded by some religious dietary rules and by Hindu practice. Usually the cheapest per gram.
Porcine	Pig skin	Excluded by halal and kosher diets. Common in gelatin and in older research.
Chicken	Sternum cartilage, feet	The usual source when a product is sold as type II rather than type I.

Those last reasons are often the ones that actually decide the purchase. Heavy processing does not make an ingredient automatically allergen free, so anyone with a fish allergy has good reason to treat a marine collagen as a fish product and raise it with a doctor or pharmacist rather than working it out from the label. Religious dietary rules are a question for the buyer, not for a supplement page, but they need the source species stated to answer it, which is exactly what many labels leave out.

Whether The Type Number Matters

Type describes the tissue the collagen came from. Type I dominates skin, bone and tendon. Type II is the collagen of cartilage. Type III sits alongside type I in skin and blood vessels. On the front of a tub, type I and III is mainly telling you the raw material was skin or hide rather than cartilage.

Once a product is hydrolysed, that distinction gets blurry. In the 2005 blood study, the hydrolysates came from three different tissues, porcine skin, chicken feet and cartilage, and Pro-Hyp was identified as the major circulating fragment across the study. Different starting tissues, and the same principal fragment turning up in plasma. That does not prove type is irrelevant, but it does mean the type number describes the input, and the input is not what your blood sees.

Type does change the product in one place: when the collagen has deliberately not been broken down. Undenatured type II collagen from chicken sternum is a separate category, and the dose gives it away. In a 2013 randomised, double-blind, placebo-controlled trial, 55 healthy subjects with no history of arthritic disease, who reported knee pain after a standardised stepmill test, were randomised to 40 milligrams a day of undenatured type II collagen or to placebo for 120 days, 27 in the active group and 28 on placebo. Average knee extension improved against placebo, 81.0 against 74.0 degrees, at p equal to 0.011. The active group also exercised longer before initial joint discomfort at day 120, 2.8 minutes against its own baseline of 1.4 minutes, at p equal to 0.019. That second result is a comparison with baseline rather than with placebo, and the placebo group showed no significant change.

Why that number is the point: 40 milligrams is roughly one hundredth of a hydrolysate serving. A product working at that dose is not a smaller scoop of the same thing, it is a different product with a different rationale, tested in healthy exercisers with activity-related knee discomfort rather than in people diagnosed with osteoarthritis. The paper describes the ingredient as a patented form, and the corresponding author is listed at a commercial research company, both worth knowing when you read the result.

As for type III specifically, the searches behind this page did not turn up a human trial comparing a type I hydrolysate against a type III hydrolysate at the same dose. Type I and III on a label is a description of the source tissue. Nobody has shown it is a performance claim.

What The Trials Actually Used

These are trial figures, reported so you can compare a label against them. They are not a recommendation, and how much of anything belongs in your day is a question for a doctor or pharmacist who knows your situation.

The clearest look at amount comes from a 2013 double-blind, placebo-controlled trial in which 69 women aged 35 to 55 were randomised to 2.5 grams of collagen hydrolysate, 5.0 grams, or placebo once daily for 8 weeks, with 23 allocated to each group. By the end of the study, skin elasticity improved significantly against placebo in both dose groups. Skin moisture and skin evaporation showed a positive influence in a subgroup analysis, but that data failed to reach statistical significance, and the abstract does not report the 5 gram arm beating the 2.5 gram arm.

A companion 2014 trial tested a named commercial bioactive collagen peptide. A hundred and fourteen women aged 45 to 65 were randomised, 57 to 2.5 grams once daily and 57 to placebo, for 8 weeks. Eye wrinkle volume was significantly lower than placebo after 4 and 8 weeks, reported as 20 percent at 8 weeks, and the difference was still present 4 weeks after the last dose. In a subgroup that gave suction blister biopsies, procollagen type I was 65 percent higher and elastin 18 percent higher than placebo after 8 weeks. Fibrillin was 6 percent higher, and that difference failed to reach statistical significance.

Form tested	Daily amount in the study	Duration
Collagen hydrolysate	2.5 to 10 g	8 to 24 weeks
Collagen tripeptide	3 g	4 to 12 weeks
Low molecular weight peptide	1 g	12 weeks
Undenatured type II collagen	40 mg	120 days

Who these trials studied is part of the result. A 2021 systematic review and meta-analysis pooled 19 randomised, double-blind, controlled trials with 1,125 participants aged 20 to 70, of whom 95 percent were women, and concluded that 90 days of hydrolysed collagen reduced wrinkles and improved elasticity and hydration. Ninety-five percent women is not a footnote. The evidence base for skin outcomes is overwhelmingly female, which is worth remembering if you are not.

One more thing a large scoop will not do for you. Collagen protein lacks tryptophan, an indispensable amino acid, so under the standard PDCAAS scoring method it is classified as an incomplete protein source. Modelling published in 2019 on a Western pattern diet found that up to 36 percent of daily dietary protein could come from collagen peptides while still meeting indispensable amino acid requirements, and noted effective amounts in the literature of 2.5 to 15 grams a day. Those percentages were calculated against United States dietary patterns and do not transfer directly to a Vietnamese diet. One of the three authors discloses providing research services to a supplement manufacturer, and the other two are listed at GELITA AG, a collagen peptide manufacturer, and at the Collagen Research Institute in Kiel. The practical takeaway holds regardless: a collagen powder is not a substitute for a complete protein, which is a different product with a different job. If protein intake is the actual question, the protein shake guide is the right page.

Reading The Label And The Spec Sheet

Whether you are buying in a pharmacy in Vietnam, from an online marketplace, or from a Japanese or Korean import shop, the front of the pack is marketing and the specification sheet is the document. Three things a real spec sheet states and a marketing label often does not:

1Source species and tissue, not just marine or bovine. Fish skin and fish scale are different raw materials, and the 2007 blood study found they produced different peptide profiles.
2Average molecular weight in daltons, or the peptide profile. If a brand charges a premium for low molecular weight, that number should be on paper somewhere. A claim with no figure behind it is a slogan.
3Grams of collagen protein per serving, not grams of powder per scoop. A 10 gram scoop of a flavoured, sweetened blend is not 10 grams of collagen, and the difference is where the trial comparison lives.

What you cannot verify yourself is the source species. A finished hydrolysate is a white powder, and the animal it came from is not visible, smellable or tastable. Laboratories settle it with species-specific marker peptides read on a mass spectrometer. Published work has built exactly those methods for gelatins where substituting a cheaper species is a known problem: a 2023 paper validated seven marker peptides to distinguish soft-shelled turtle gelatin from other animal gelatins, and a 2022 paper did the same for donkey-hide gelatin and its adulterants. Both are traditional medicine gelatins studied by Chinese research groups, not skincare collagen powders, so this is not evidence that supplement collagen is widely mislabelled. It is evidence that if a source claim were false, you would have no way of knowing without a laboratory.

A listing is not an approval. A collagen powder is sold as a food supplement, not as an approved medicine, and no amount of clinical-looking packaging changes that.

Checked on 5 August 2026, the United States FDA National Drug Code directory returned 35 records for products listing collagen as an active ingredient. The marketing categories were 20 unapproved homeopathic, 8 over-the-counter monograph drug, 4 unapproved drug other, 2 over-the-counter monograph final, and 1 Biologics License Application. Only that last category is an approval pathway, and the single record in it is an intra-articular implant used in cartilage repair, not an oral supplement. Exactly one record in the 35 was an oral powder, and it sat under unapproved drug other. Dietary supplements do not appear in that directory at all. This is United States data and says nothing about how any product is registered in Vietnam.

Two questions cover most of it at the counter: what species and tissue is this made from, and how many grams of collagen protein are in a serving. A seller who can answer both is working from a real spec sheet. A seller who cannot is reading the same front-of-pack text you are.

The Short Version

Collagen, gelatin and collagen peptides are one protein at three stages of being cut apart. Peptides means hydrolysed, which means it dissolves cold and does not gel.
Human blood studies do find intact fragments after a dose. Pro-Hyp is the main one, peaking at 1 to 2 hours, but peptide-bound hydroxyproline was only about 30 percent of the total detected.
Measuring a fragment in plasma is not the same as showing it reaches skin. That step has not been shown in the studies cited here.
Low molecular weight is a plausible argument with one supportive proprietary trial at 1 gram, and no head-to-head trial against a standard hydrolysate at the same dose.
The marine versus bovine study most often quoted compared fish against pork in five healthy men. It had no bovine arm.
A 2023 meta-analysis found source affected hydration in subgroup analysis, however it found no significant source difference for elasticity, and it flagged bias in the pooled trials.
Type on a hydrolysate describes the source tissue. Type matters far more when a product is undenatured, such as type II collagen trialled at 40 mg a day.
Trials used 2.5 to 10 g of hydrolysate; a trial testing 2.5 g and 5.0 g found both beat placebo on elasticity. The skin evidence base is about 95 percent women.
Collagen protein lacks tryptophan, so it is scored as an incomplete protein and is not a protein powder substitute.

Frequently Asked Questions

What is the difference between collagen and collagen peptides?+

They are the same protein at different stages of being cut up. Native collagen is a long triple helix that will not dissolve in cold water. Heat and processing unwind it into gelatin, which sets into a jelly. Enzymes then cut gelatin into much shorter fragments, and that is what gets sold as hydrolysed collagen or collagen peptides. The practical tell is behaviour in water: a hydrolysate dissolves in a cold drink and does not gel, gelatin does the opposite. So the word peptides on a tub is a claim about fragment size, not about a different or better raw material.

Do collagen peptides actually survive digestion?+

Some of them do, in small amounts. In a 2005 study, healthy volunteers who fasted for 12 hours and then took 9.4 to 23 grams of gelatin hydrolysate showed almost no peptide-form hydroxyproline in blood beforehand, and a rise to a peak of 20 to 60 nmol per mL of plasma 1 to 2 hours after, falling to half the peak by 4 hours. The main fragment identified was the dipeptide Pro-Hyp. A later study measuring both forms found the peptide-bound share was about 30 percent of all detected hydroxyproline, so most of what turns up in blood is still free amino acid. Measuring a fragment in blood is not the same as showing it reaches skin or does anything there, and these studies did not test that.

Is marine collagen better than bovine collagen?+

The evidence usually cited for this does not answer the question. The blood study most often quoted was a single-blind crossover in five healthy men comparing type I hydrolysates from fish scale, fish skin and porcine skin. Fish scale produced a significantly higher total area under the concentration-time curve than porcine skin, and some peptides appeared only with the fish sources. There was no bovine arm and only five participants. A 2023 meta-analysis of 26 randomised trials found that skin hydration effects varied by collagen source, however it found no significant difference between sources for elasticity, and it reported several biases in the trials it pooled.

Does the molecular weight of a collagen peptide matter?+

It is a reasonable argument that is not yet settled. What circulates after a dose is mostly di- and tripeptides, so a hydrolysate already cut to that size is skipping a step in principle. One Korean trial of a low molecular weight collagen peptide, with a tripeptide content above 15 percent, reported improvements in hydration, wrinkles and some elasticity measures at 1000 mg a day, far below the 2.5 to 10 grams used in ordinary hydrolysate trials. That is suggestive, but it was one trial of one proprietary ingredient in 64 women aged 40 to 60 with photoaged skin, three of the five authors are listed with a commercial affiliation, two at Newtree Co., Ltd. and one at the research institute of Jellice, a gelatin and collagen manufacturer, and it had no arm testing a standard hydrolysate at the same 1000 mg. The searches behind this page did not turn up a trial that randomised the same people to a high and a low molecular weight version at an identical dose.

Does type I, II or III on the label matter?+

On a hydrolysed powder it mostly describes the tissue the raw material came from, not what you absorb. In the 2005 blood study the hydrolysates came from porcine skin, chicken feet and cartilage, and Pro-Hyp was identified as the major circulating fragment across the study. Type carries much more weight when a product is deliberately not hydrolysed: undenatured type II collagen from chicken sternum is trialled at 40 milligrams a day, roughly a hundredth of a hydrolysate serving, which tells you it is a different kind of product rather than a different flavour of the same one.

Research And Sources

Every figure on this page comes from one of the papers below, read as published abstracts, with the full paper and the author affiliations checked where a detail was not in the abstract. Study designs, populations, funding disclosures and author caveats are reported as the papers state them. Where a study hedged or a result failed to reach significance, that is carried through rather than dropped. The regulatory figures come from the United States FDA National Drug Code directory, queried through the openFDA interface on 5 August 2026 for products listing collagen as an active ingredient.

Identification of food-derived collagen peptides in human blood after oral ingestion of gelatin hydrolysates · Iwai K, Hasegawa T, Taguchi Y, Morimatsu F, Sato K, Nakamura Y, Higashi A, Kido Y, Nakabo Y, Ohtsuki K · Journal of Agricultural and Food Chemistry (2005) PMID:16076145
Comparison of quantity and structures of hydroxyproline-containing peptides in human blood after oral ingestion of gelatin hydrolysates from different sources · Ohara H, Matsumoto H, Ito K, Iwai K, Sato K · Journal of Agricultural and Food Chemistry (2007) PMID:17253720
Oral supplementation of specific collagen peptides has beneficial effects on human skin physiology: a double-blind, placebo-controlled study · Proksch E, Segger D, Degwert J, Schunck M, Zague V, Oesser S · Skin Pharmacology and Physiology (2014) PMID:23949208
Oral intake of specific bioactive collagen peptides reduces skin wrinkles and increases dermal matrix synthesis · Proksch E, Schunck M, Zague V, Segger D, Degwert J, Oesser S · Skin Pharmacology and Physiology (2014) PMID:24401291
Oral Intake of Low-Molecular-Weight Collagen Peptide Improves Hydration, Elasticity, and Wrinkling in Human Skin: A Randomized, Double-Blind, Placebo-Controlled Study · Kim DU, Chung HC, Choi J, Sakai Y, Lee BY · Nutrients (2018) PMID:29949889
Oral Collagen Supplementation: A Systematic Review of Dermatological Applications · Choi FD, Sung CT, Juhasz ML, Mesinkovsk NA · Journal of Drugs in Dermatology (2019) PMID:30681787
Significant Amounts of Functional Collagen Peptides Can Be Incorporated in the Diet While Maintaining Indispensable Amino Acid Balance · Paul C, Leser S, Oesser S · Nutrients (2019) PMID:31096622
Effects of hydrolyzed collagen supplementation on skin aging: a systematic review and meta-analysis · de Miranda RB, Weimer P, Rossi RC · International Journal of Dermatology (2021) PMID:33742704
Effects of Oral Collagen for Skin Anti-Aging: A Systematic Review and Meta-Analysis · Pu SY, Huang YL, Pu CM, Kang YN, Hoang KD, Chen KH, Chen C · Nutrients (2023) PMID:37432180
Undenatured type II collagen (UC-II) for joint support: a randomized, double-blind, placebo-controlled study in healthy volunteers · Lugo JP, Saiyed ZM, Lau FC, Molina JP, Pakdaman MN, Shamie AN, Udani JK · Journal of the International Society of Sports Nutrition (2013) PMID:24153020
Identification and application of species-specific peptide biomarkers from soft-shelled turtles (Pelodiscus sinensis) using post-translational modification detection-based peptidomics analysis · Xia H, Liu X, Cai S, Chen S, Li S, Zhao M, Duan JA, Liu R, Han S · Food Chemistry (2023) PMID:37011573
Species-specific identification of donkey-hide gelatin and its adulterants using marker peptides · Zhang J, Wu M, Ma Z, Zhang Y, Cao H · PLoS One (2022) PMID:35960756