Collagen: From Boiled Bones to Hydrolyzed Peptides

Collagen started as a jelly from boiled bone and hide, and the capsule on the shelf is the same protein cut into fragments. This is what the sources show about that journey, what they cannot confirm, and what a label can and cannot tell you.

A collagen capsule is the newest step in an old kitchen process: boiling animal skin and bone until the liquid sets into a jelly. The change that matters most for a capsule happened after that jelly was broken into smaller pieces. According to PubMed, Iwai and colleagues (2005) gave human subjects 9.4 to 23 grams of collagen hydrolysate and reported that the main peptide found in their blood was Pro-Hyp, a two-amino-acid fragment called prolyl-hydroxyproline ([DOI](https://doi.org/10.1021/jf050206p)). The protein on the shelf is therefore a source of small pieces, and the history of how those pieces were made, tested and labeled explains most of what the label says. This page follows the whole arc: where collagen comes from, the earliest records, the Eastern texts, discovery, the market, the science, the countries that have studied it, the regulatory positions, the trials, and where a modern capsule fits. Where the record runs out, the page says so instead of filling the gap. **Jump to** - [Where collagen comes from](#where-collagen-comes-from) - [The earliest records and the Eastern texts](#the-earliest-records-and-the-eastern-texts) - [Ejiao and the donkey hide trade](#ejiao-and-the-donkey-hide-trade) - [Broths in the Western record](#broths-in-the-western-record) - [Discovery and the steam digester](#discovery-and-the-steam-digester) - [How gelatin went to market](#how-gelatin-went-to-market) - [How the science changed from gelatin to peptides](#how-the-science-changed-from-gelatin-to-peptides) - [Who has studied it and where](#who-has-studied-it-and-where) - [Regulatory positions side by side](#regulatory-positions-side-by-side) - [What the trials measured and where they disagree](#what-the-trials-measured-and-where-they-disagree) - [Why the range of collagen products matters](#why-the-range-of-collagen-products-matters) - [Where the record is thin](#where-the-record-is-thin) - [Where a modern capsule fits](#where-a-modern-capsule-fits) - [Frequently asked questions](#frequently-asked-questions) ## Where collagen comes from Collagen is the main structural protein of animal connective tissue. According to PubMed, Shoulders and Raines (2009) describe it as the most abundant protein in animals ([DOI](https://doi.org/10.1146/annurev.biochem.77.032207.120833)). It forms the framework of skin, tendons, ligaments and cartilage, and it builds the protein scaffold of bone, whose mineral side is covered in the [vitamin D guide](/health/vitamin-d-complete-guide). Its basic unit is a triple helix of three chains wound around one another, and the amino acid hydroxyproline helps hold that helix together. In a 2011 study of synthetic peptides, Shoulders and Raines reported that hydroxyproline in one particular position makes the triple helix more stable ([DOI](https://doi.org/10.1074/jbc.M110.199984)). The raw material for gelatin has always been animal connective tissue, including hooves, hides, bone and skin. Today's supplements use bovine, porcine, chicken and marine sources. A 2007 study of five men compared collagen peptides made from fish scale, fish skin and pig skin. According to PubMed, Ohara and colleagues found that the fish-scale peptides produced a higher area under the blood curve than the pig-skin peptides ([DOI](https://doi.org/10.1021/jf062834s)). Five men is a small group, so that result describes those five men and says little about the market. The earliest records in this page are not about supplements. They describe jelly, cookery and medicine, and the distance between those uses and a capsule is the subject of the rest of the page. ## The earliest records and the Eastern texts The earliest pharmacology texts on this page are Chinese. The University of California Press translation of the Bencao Gangmu, volume VI (excerpt), dates the Shennong Bencao Jing to the 1st century BCE to the 1st century CE. The same excerpt lists the work of Tao Hongjing, dated there to 500 CE, and the Xin xiu bencao of 659. Li Shizhen lived from 1518 to 1593. The Bencao Gangmu he compiled covers 1,892 substances, arranged under ten subheadings, one of them for pharmaceutical preparation, according to the excerpt (ISBN 9780520395176). The record does not show the Shennong text naming donkey-hide gelatin or any other animal skin or bone preparation. The excerpt covers chapters 26 to 33 and has no ejiao entry, so this page quotes no Chinese passage on gelatin or collagen. Ejiao gets its own section below, where the sources are modern and the classical uses are not settled by the record. Ayurveda has the largest gap. Neither the Charaka Samhita nor the Sushruta Samhita is quoted here for gelatin, bone broth or collagen. No date is given for such a passage either. This page therefore makes no claim about Ayurvedic use of collagen or gelatin. The Canon of Medicine by Ibn Sina is the text on this page tied to the Persian and Unani tradition. According to PubMed, a 2025 review lists the Canon among the core texts of Persian medicine ([DOI](https://doi.org/10.22038/ABJS.2024.83621.3805)). No passage about gelatin or broth in the Canon is claimed on this page. Korean and Japanese classics are also missing. No passage from Dongui Bogam, the Korean medical classic, or from the Kampo texts of Japan that names collagen or gelatin is quoted here. Readers who need those passages will have to consult the primary texts or a translation that quotes them. ## Ejiao and the donkey hide trade Ejiao is a gelatin made from donkey hide. A Hong Kong Baptist University project on donkey-hide gelatin describes ejiao as used for over 2,000 years. That figure is the project's own claim, and this page gives no independent date for it. The same project describes donkey skins from a particular breed, water from a specific well in Shandong, and donkey supply chains that run across several countries. The market has a quality problem. According to PubMed, Wu and colleagues (2022) tested 110 commercial ejiao batches and classed 57 as unqualified ([DOI](https://doi.org/10.3390/molecules27144643)). Three of the authors list their affiliation as Dong-E-E-Jiao Co., Ltd., a company in Liaocheng whose name includes Dong'e. Readers should weigh that affiliation when reading the study. According to PubMed, Liu and colleagues (2021) describe donkey hide as having low yield and high price, and they developed a real-time PCR test for pig hide ([DOI](https://doi.org/10.4315/JFP-20-140)). A 2020 study in the Journal of Food Science and Technology reported a loop-mediated isothermal amplification test that detects donkey gelatin at 0.1 percent, and described ejiao demand rising while donkey production falls ([DOI](https://doi.org/10.1007/s13197-020-04319-3)). What is in ejiao? According to PubMed, Huang and colleagues (2020) fractionated the glycosaminoglycans in ejiao and reported hyaluronic acid among them ([DOI](https://doi.org/10.1007/s10719-019-09904-z)). That is a composition finding, not a finding about effects in people. Modern trials cited here tested ejiao inside formulas. According to PubMed, a 2021 systematic review and meta-analysis pooled nine randomized trials with 844 patients. The trials tested traditional formulas containing ejiao or velvet antler for uterine fibroids ([DOI](https://doi.org/10.21037/apm-21-1755)). Because the trials tested whole formulas, they do not isolate what the gelatin did. A 2025 review proposes research directions for ejiao's pharmacology in chronic obstructive pulmonary disease ([DOI](https://doi.org/10.1002/ame2.70077)). A review of mechanisms is a proposal for study, not evidence of an effect. What did classical Chinese texts recommend ejiao for? No passage naming the indications is quoted here. Everything in this section concerns modern formulas, composition and markets, so none of it is a record of historical use. A reader who needs the classical indications will have to consult the Chinese texts themselves. ## Broths in the Western record The main source for the European and Islamic record here is Mikhailov's 2023 review in the International Journal of Molecular Sciences ([DOI](https://doi.org/10.3390/ijms24043583)). According to that review, the Viandier, a French cookery text traditionally attributed to Taillevent and dated there to about 1375, includes a recipe for a jellied meat broth. The review also records that 15th-century Britain used hooves and meat waste for jelly. The review cites Kitab al-Tabikh, an Arabic cookbook, for a fish jelly made by boiling fish heads. It dates the book to the tenth century BCE. That date cannot be reconciled with an Arabic text, so this page does not use it, and the fish jelly is recorded here without a date. Greek and Roman sources are where a reader would expect the earliest European broth records. No dated Greek or Roman source for bone or skin broth is on record here, so the page does not claim one. The gap stays open. ## Discovery and the steam digester Denis Papin demonstrated his digester, a sealed pressure vessel, to the Royal Society in 1679. In 1681 he published an English book, A new digester or engine for softening bones. The MacTutor biography of Papin gives both dates ([MacTutor](https://mathshistory.st-andrews.ac.uk/Biographies/Papin/)). The same biography cites a 1947 account by Robinson of extracting gelatin from bones with high-pressure steam. Mikhailov's review credits Papin's 1681 work with bone boiling and dates a hydrochloric acid method for bone gelatin to d'Arcet in 1812 ([DOI](https://doi.org/10.3390/ijms24043583)). Both methods were ways to get gelatin out of bone, one with heat and pressure and one with acid. The record does not settle who first identified collagen or when, so the page gives neither. A commonly repeated 1838 attribution is not supported by the record. What the sources do establish is a sequence. Jelly from boiled meat appears in the Viandier, dated to about 1375. Papin's digester was shown in 1679 and described in 1681. D'Arcet's acid method dates to 1812, and Mikhailov's review places most of the modern foundations in the 20th century. ## How gelatin went to market The review places the start of powdered gelatin in the United States in the mid-19th century, and it names Jell-O as a gelatin product ([DOI](https://doi.org/10.3390/ijms24043583)). It also links gelatin to silver halide photography, which it dates to 1871, and it lists gelatin capsules among pharmaceutical uses. These market dates come from that review, not from a separate record. No date is given here for when hydrolyzed collagen first went on sale, because the record does not show one. The rules that govern what a product may say are set out in the regulatory section below. ## How the science changed from gelatin to peptides The science moved in steps: the molecule, then absorption, then measurements in people. Shoulders and Raines's 2009 review and their 2011 synthetic-peptide study, both covered above, describe collagen's stability as a property of its amino acid sequence. Hydrolysis is the step that separates a dissolving supplement from a jelly. It cuts collagen's long chains into shorter pieces, using heat, acid or enzymes, and the label word hydrolyzed names that step. A hydrolysate dissolves in cold liquid and does not set, which is the practical difference from the gelatin of the kitchen. In a 1999 mouse study, Oesser and colleagues reported that about 95 percent of the dose was absorbed within 12 hours. The absorbed peptides fell between 2.5 and 15 kilodaltons, and they accumulated in cartilage (According to PubMed, [DOI](https://doi.org/10.1093/jn/129.10.1891)). A mouse study is an early step, and nothing in it tests a supplement in people. The human measurements came in 2005. Iwai and colleagues measured peptides in blood after people took 9.4 to 23 grams of hydrolysate, and the main peptide was Pro-Hyp ([DOI](https://doi.org/10.1021/jf050206p)). Those measurements were of peptides in the blood. Hydroxyproline links the two strands of work. It is the residue that stabilizes the helix in the 2011 study, and Pro-Hyp contains it. Source comparisons followed the same logic. Ohara and colleagues's 2007 comparison of fish scale, fish skin and pig skin measured how the peptides appeared in the blood. It did not test whether they changed skin or joints, which is the question the later trials had to answer. The label question follows from this. A kilodalton is a unit of molecular mass. A label that gives a size range describes the fragments a maker produced, not the dose, the amount absorbed or the effect. The 2.5 to 15 kilodalton range from the mouse study shows how such numbers are measured, and it came from mice. Mikhailov's review gives the longer arc. Most of the modern foundations were laid in the 20th century, and the human trials covered later on this page came after them. ## Who has studied it and where Two studies cited here, Iwai and colleagues (2005) and Ohara and colleagues (2007), are about peptides and fish collagen. This page does not assign their authors to a country. The 2015 paper by Asserin and colleagues does list affiliations with Unitec Foods in Tokyo and Rousselot in Ghent, and both are disclosed in the trials section. Japan runs on two tracks, FOSHU from 1991 and FFC from 2015, and the table below sets out what each requires. Primary pages from Japan's regulators are not cited here, so that description rests on an academic review of the system. Most of the ejiao papers cited here cover composition, adulteration, market quality and formulas. This page does not report Chinese Pharmacopoeia or National Health Commission positions. Europe's main record here is EFSA's 2011 opinion on collagen hydrolysate and the maintenance of joints. Gelita AG made the application, which went through a German competent authority. Regulation (EU) No 379/2012 is the EU's list of refused health claims, and this page does not state whether its collagen entry is included. This page does not report positions from the UK's SACN, the German BfArM and the historical Commission E monographs, or India's ICMR and Ayurvedic Pharmacopoeia. That silence is a gap, not a finding, and the table marks each one. In the US, FDA's public description of supplement regulation says manufacturers carry responsibility before sale, and FDA acts after products reach the market. No NIH Office of Dietary Supplements fact sheet on collagen is cited on this page. Among the reviews cited here, Mikhailov's 2023 history and the 2023 and 2026 skin meta-analyses by Pu and colleagues and by Nukaly and colleagues are scholarly syntheses. None of them is a regulator's position, and this page keeps them separate from the table. ## Regulatory positions side by side Each row gives what the body said or set up, and when. Rows marked "not covered" are gaps, not negative findings. | Body | What it said or set up | Year | Source | |---|---|---|---| | FDA (US) | Manufacturers are responsible for safety and labeling before sale; FDA acts after products reach the market | Page updated October 2024 | [FDA Dietary Supplements](https://www.fda.gov/food/dietary-supplements) | | EFSA (EU) | Opinion on collagen hydrolysate and joint maintenance: cause and effect not established; application by Gelita AG | 2011 | [EFSA opinion](https://efsa.europa.eu/en/efsajournal/pub/2291) | | EU law | Regulation (EU) No 379/2012 lists refused health claims; this page does not state whether its collagen entry is included | 2012 | [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2012/379/oj/eng) | | Japan, FOSHU | Requires trials and national review | Since 1991 | [Hattori and Nagamura, 2022](https://e-jabs.umin.jp/10/10.49.pdf) | | Japan, FFC | Manufacturer responsibility; systematic review accepted; trials not mandatory; over 5,000 notifications by year eight | Since 2015 | [Hattori and Nagamura, 2022](https://e-jabs.umin.jp/10/10.49.pdf) | | NIH ODS and NCCIH (US) | No NIH ODS collagen fact sheet is cited here, and no NCCIH position is stated | Not covered here | None cited | | China (NHC, Chinese Pharmacopoeia) | Not covered on this page | Not covered here | None cited | | Germany (BfArM, Commission E) | Not covered on this page | Not covered here | None cited | | UK (SACN), India (ICMR, Ayurvedic Pharmacopoeia) | Not covered on this page | Not covered here | None cited | | WHO, Health Canada, TGA | Not covered on this page | Not covered here | None cited | The sources agree on one point. FDA's description and Japan's FFC rules both put the first responsibility for a supplement's safety and label on the manufacturer. The bars for evidence differ. FFC accepts a systematic review and does not require trials, while EFSA examined the trial record and concluded that cause and effect had not been established. The table is a comparison, not a ranking. ## What the trials measured and where they disagree One joint trial cited here is Clark and colleagues's 2008 report. It describes a 24-week trial of 147 athletes with joint pain, 97 of whom were analyzed. Participants took a 25-milliliter liquid with 10 grams of collagen hydrolysate. Six joint-pain measures reached significance, but the analysis was not adjusted for multiple comparisons (According to PubMed, [DOI](https://doi.org/10.1185/030079908x291967)). EFSA's 2011 opinion describes a 147-athlete trial with no difference on the endpoints after adjustment for multiple comparisons. The participant count matches Clark's report, so the two appear to describe the same trial. The pair shows why adjustment matters: the same data can read as six signals or as none. Skin trials came later. According to PubMed, Proksch and colleagues (2014) ran a placebo-controlled trial of 114 women who took 2.5 grams of a collagen peptide product for eight weeks ([DOI](https://doi.org/10.1159/000355523)). The authors reported about a 20 percent drop in eye-wrinkle volume, with procollagen type I up about 65 percent and elastin up about 18 percent. Asserin and colleagues's 2015 paper reports two randomized trials. Hydration was measured by corneometry, dermal density by ultrasound and collagen fragmentation by confocal microscopy ([DOI](https://doi.org/10.1111/jocd.12174)). Several authors list affiliations with Unitec Foods in Tokyo and Rousselot in Ghent. That is a disclosure point, not a verdict on the data. Two newer meta-analyses pool the skin trials. According to PubMed, Pu and colleagues (2023) pooled 26 randomized trials with 1,721 participants and found improved hydration and elasticity ([DOI](https://doi.org/10.3390/nu15092080)). Nukaly and colleagues (2026) pooled 19 trials with 1,341 participants, covering oral and topical peptides. They found improved hydration and brightness, a modest effect on wrinkles, and inconsistent results for elasticity ([DOI](https://doi.org/10.3389/fmed.2026.1618306)). The joint evidence is weaker. According to PubMed, Liu and colleagues (2017) reviewed osteoarthritis supplements in the British Journal of Sports Medicine. Collagen hydrolysate was one of seven supplements with a large short-term effect, but the overall evidence was rated very low, and the authors found no clinically important effect at medium or long term ([DOI](https://doi.org/10.1136/bjsports-2016-097333)). For another joint supplement and how its evidence was judged, see [the glucosamine post](/blog/bone-joint-support/exploring-the-glucosamine-chondroitin-benefits-of-supplements). Grades, per outcome, on this site's scale: - Skin hydration: Moderate. The meta-analyses agree on direction, but products and doses vary and at least one trial has industry-affiliated authors. - Skin elasticity: Insufficient. The two meta-analyses disagree. - Wrinkles: Moderate for a modest effect, resting on one placebo-controlled trial and a meta-analysis. - Joint pain: Insufficient. The trials conflict, and the pooled evidence is rated very low. Most skin outcomes in these records are instrument readings, not what a person reports. The joint outcomes are closer to what a person feels, and that is where the evidence is weakest. ## Why the range of collagen products matters Collagen is interesting because the same protein runs through very different products. A Viandier recipe of about 1375, a 1681 book on a pressure vessel for softening bone, a Chinese donkey-hide gelatin and a modern capsule all sit on one line, and the evidence has moved from texture to fragments to trials. Source names matter in their own right. Bovine, porcine, chicken and marine labels are not interchangeable in the research. The one source comparison on this page is the fish-scale and pig-skin study in five men. That is a single small study, and it measured absorption only. The word collagen on a label can cover several different things. It can mean gelatin from hide or bone, a hydrolysate from bovine or marine sources, or ejiao, a donkey-hide gelatin used in Chinese medicine. These materials have different histories and different evidence, and a label should say which one it is. Authentication is not theoretical. The ejiao studies above describe detection tests for donkey and pig material, and one 2022 study classed 57 of 110 commercial batches as unqualified. The label has a second half that is easy to miss: a kilodalton figure describes fragment size, and it says nothing about dose or effect. ## Where the record is thin The record does not settle who first identified collagen or when, so the page gives neither. No dated Greek or Roman source for bone or skin broth is on record here, and no classical passage on gelatin or collagen from the Ayurvedic, Persian, Korean or Japanese texts is quoted. No Shennong or Bencao Gangmu entry on ejiao is quoted, and the only date offered for when ejiao production began in Dong'e is an "over 2,000 years" claim from a university project. Official positions from China, the UK, Germany, India, WHO, Health Canada, TGA, the NIH Office of Dietary Supplements and NCCIH are not covered on this page, and the primary MHLW pages of Japan are not cited. Whether the EU list of refused health claims names collagen is also not stated here. ## Where a modern capsule fits A modern capsule sits at the end of this history, not at its center. Its label tells you what the product contains, and the history tells you what that label can and cannot say. The product is sold as hair, skin and nail support. Its page lists 60 capsules, two capsules per serving and 30 servings, with a 750 mg blend of kosher gelatin, hyaluronic acid and chicken collagen. The page does not give the amount of each ingredient inside that blend, so the collagen dose cannot be read from the label. At $27.99, the product works out to about $0.93 per day. Hyaluronic acid is one of the blend's named ingredients. For hyaluronic acid on its own, see [the hyaluronic acid guide](/blog/bone-joint-support/uncovering-the-truth-about-hyaluronic-acid-benefits-side-effects-and-dosage). The trials above tested other products at other doses. The skin trial used 2.5 grams of a different collagen peptide product, and the joint trial used 10 grams in a liquid. Neither tested this capsule. No trial of nail outcomes turned up in the records reviewed, so this page makes no nail claim. For how the company sources and tests its raw materials, read [our raw material sourcing and testing post](/blog/new/our-approach-to-raw-material-sourcing-and-testing). You can also buy the [NutriSilk capsules](https://mtangelvitamins.com/products/nutrisilk-hair-skin-nail-support?utm_source=learn_hub&utm_medium=content&utm_campaign=blog_post&utm_content=history-of-collagen-from-boiled-bones-to-peptides) directly from the company. The oldest records on this page describe jelly, not capsules. The newest describe fragments measured in blood and in skin. So when a label gives a size range and no dose, what is it actually telling you? ### Next reads - [Vitamin D complete guide](/health/vitamin-d-complete-guide): the mineral and vitamin side of bone health, which this page does not cover. - [Glucosamine and chondroitin post](/blog/bone-joint-support/exploring-the-glucosamine-chondroitin-benefits-of-supplements): how another joint supplement's evidence has been judged. ## Frequently asked questions ### What is collagen, and how was it first identified? Collagen is the most abundant protein in animals, according to PubMed (Shoulders and Raines, 2009). It forms a triple helix in skin, tendon, cartilage and bone. The record does not settle who first identified collagen or when, so the page gives neither. ### How did gelatin from bones and hides become food and medicine? The dated record starts with the Viandier, traditionally attributed to Taillevent and dated to about 1375, which includes a jellied meat broth. Denis Papin demonstrated his digester in 1679 and described its bone-softening use in 1681. D'Arcet's acid method dates to 1812. Gelatin also appears in pharmaceutical capsules, and in China ejiao, a donkey-hide gelatin, is used in traditional medicine. ### What changed when collagen was broken into smaller peptides? Hydrolysis produces pieces that dissolve in cold liquid and do not set into a jelly. According to PubMed, Iwai and colleagues (2005) found Pro-Hyp to be the main peptide in blood after people took 9.4 to 23 grams of hydrolysate. Absorption studies measure fragments. What those fragments do in skin or joints is a separate question, and it has to be tested in trials. ### Where do collagen supplements come from today: bovine, marine, or chicken? Supplements use bovine, porcine, chicken and marine sources. The label or product page should name the source, and a buyer should read it. This page cites one source comparison, a five-man study in which fish-scale peptides produced a higher blood curve than pig-skin peptides. That study measured absorption, not skin or joint outcomes, and no head-to-head test across all the sources is cited. ### What do the collagen peptide size numbers on labels mean? A kilodalton (kD) is a unit of molecular mass. A size range describes the fragments a maker produced. It does not state the dose, the amount absorbed or the effect. The 2.5 to 15 kilodalton range in a 1999 mouse study shows how such numbers are measured in research, and that work was done in mice. ### What did the early collagen trials actually measure? Mostly instrument readings and joint-pain scores. Clark and colleagues's 2008 joint trial reported six pain measures as significant, but the analysis was not adjusted for multiple comparisons. Skin trials measured hydration, dermal density, collagen fragmentation and wrinkle volume. On this site's scale, skin hydration is graded Moderate and joint pain is graded Insufficient. ## Sources According to PubMed, the records below are listed with their PubMed entries. - Shoulders and Raines, 2009, Annual Review of Biochemistry. [DOI](https://doi.org/10.1146/annurev.biochem.77.032207.120833) - Shoulders and Raines, 2011, Journal of Biological Chemistry. [DOI](https://doi.org/10.1074/jbc.M110.199984) - Mikhailov et al., 2023, review, International Journal of Molecular Sciences. [DOI](https://doi.org/10.3390/ijms24043583) - Oesser et al., 1999, The Journal of Nutrition. [DOI](https://doi.org/10.1093/jn/129.10.1891) - Iwai et al., 2005, Journal of Agricultural and Food Chemistry. [DOI](https://doi.org/10.1021/jf050206p) - Ohara et al., 2007, Journal of Agricultural and Food Chemistry. [DOI](https://doi.org/10.1021/jf062834s) - Proksch et al., 2014, Skin Pharmacology and Physiology. [DOI](https://doi.org/10.1159/000355523) - Asserin et al., 2015, Journal of Cosmetic Dermatology. [DOI](https://doi.org/10.1111/jocd.12174) - Clark et al., 2008, Current Medical Research and Opinion. [DOI](https://doi.org/10.1185/030079908x291967) - Liu et al., 2017, British Journal of Sports Medicine. [DOI](https://doi.org/10.1136/bjsports-2016-097333) - Pu et al., 2023, Nutrients. [DOI](https://doi.org/10.3390/nu15092080) - Nukaly et al., 2026, Frontiers in Medicine. [DOI](https://doi.org/10.3389/fmed.2026.1618306) - Wu et al., 2022, Molecules. [DOI](https://doi.org/10.3390/molecules27144643) - Liu R et al., 2021, Journal of Food Protection. [DOI](https://doi.org/10.4315/JFP-20-140) - Sheu et al., 2020, Journal of Food Science and Technology. [DOI](https://doi.org/10.1007/s13197-020-04319-3) - Huang et al., 2020, Glycoconjugate Journal. [DOI](https://doi.org/10.1007/s10719-019-09904-z) - Zhang et al., 2025, Animal Models and Experimental Medicine. [DOI](https://doi.org/10.1002/ame2.70077) - Li Q et al., 2021, Annals of Palliative Medicine. [DOI](https://doi.org/10.21037/apm-21-1755) - Khosravi et al., 2025, Archives of Bone and Joint Surgery. [DOI](https://doi.org/10.22038/ABJS.2024.83621.3805) - EFSA, opinion on collagen hydrolysate and maintenance of joints, 2011, EFSA Journal. [EFSA](https://efsa.europa.eu/en/efsajournal/pub/2291) - Commission Regulation (EU) No 379/2012, EUR-Lex. [EUR-Lex](https://eur-lex.europa.eu/eli/reg/2012/379/oj/eng) - Hattori and Nagamura, 2022, Int J Anal Bio-Sci 10(3):49-59. [PDF](https://e-jabs.umin.jp/10/10.49.pdf) - US FDA, Dietary Supplements, page updated October 1, 2024. [FDA](https://www.fda.gov/food/dietary-supplements) - MacTutor, biography of Denis Papin. [MacTutor](https://mathshistory.st-andrews.ac.uk/Biographies/Papin/) - Hong Kong Baptist University, project on donkey-hide gelatin (Köhle). [HKBU](https://scholars.hkbu.edu.hk/en/projects/eating-tradition-towards-a-social-history-of-donkey-hide-gelatin-/) - Li Shizhen, Bencao Gangmu, volume VI excerpt, University of California Press, ISBN 9780520395176. [UC Press](https://content.ucpress.edu/title/9780520395176/bencaogangmuvolumeviexcerpt.pdf) - Product label facts, from the product's live page as of October 2026. The product link appears in its section. ### Related reading - [Vitamin C: from scurvy to ascorbic acid](/blog/history/history-of-vitamin-c-from-scurvy-to-ascorbic-acid) - [Calcium supplements: from chalk to bone trials](/blog/history/history-of-calcium-supplements-from-chalk-to-bone-trials) - [Glucosamine: from shellfish to joint supplement](/blog/history/history-of-glucosamine-from-shellfish-to-joint-supplement) - [Creatine: from meat extract to sports shelf](/blog/history/history-of-creatine-from-meat-extract-to-sports-shelf)