Collagen is the most abundant protein in the human body, accounting for roughly 30% of total protein mass and forming the structural scaffold of skin, bone, tendon, cartilage, and vasculature. Twenty-eight genetically distinct collagen types have been catalogued, but three — Type I, Type II, and Type III — account for more than 90% of the collagen in the body and dominate both clinical research and the supplement market.
In the last decade, collagen has moved from an obscure connective-tissue protein into one of the best-selling supplement categories, with the global collagen peptide market estimated at several billion dollars in 2025. Research has clarified two distinct supplementation paradigms: hydrolyzed collagen peptides (gelatin-like fragments dosed at 2.5–10 g/day that deliver absorbable dipeptides such as Pro-Hyp and Hyp-Gly into circulation) and undenatured collagen Type II (UC-II), a low-dose oral-tolerance approach dosed at 40 mg/day for osteoarthritis.
These two approaches are mechanistically different, not interchangeable, and clinicians and consumers frequently confuse them. A hydrolyzed marine or bovine collagen powder is a source of bioactive peptide fragments and precursor amino acids; UC-II is an immune-modulating therapy that works through a completely different pathway involving gut-associated lymphoid tissue. The collagen type on the label matters, the form (hydrolyzed vs undenatured) matters, and the dose ranges are not comparable.
This guide walks through the three most clinically relevant collagen types, the absorption biology of collagen-derived dipeptides, and the practical distinctions between hydrolyzed and undenatured collagen products.
Collagen Types Primer: 28 Types Known, 3 That Matter Most
The collagen superfamily comprises 28 distinct types (designated Types I through XXVIII) encoded by more than 40 genes. Each type shares the characteristic triple-helix structure built from three polypeptide alpha chains rich in glycine, proline, and hydroxyproline, but they differ in which tissues they populate, how they assemble (fibrillar, network-forming, fibril-associated, or transmembrane), and what biomechanical role they play.
Three collagen types dominate human biology. Type I is the most abundant overall — the principal structural collagen of skin, bone, tendon, ligament, and cornea, accounting for roughly 90% of body collagen. Type II is effectively restricted to hyaline and elastic cartilage, where it forms the fibrillar meshwork that traps proteoglycans and gives articular cartilage its compressive resilience. Type III is the second most abundant fibrillar collagen, concentrated in vasculature, gut wall, reticular fibers of lymphoid tissue, and early wound-repair matrix where it lays down a provisional scaffold that is later remodeled into Type I.
The remaining 25 collagen types (Types IV–XXVIII) serve specialized functions — Type IV forms basement membranes, Type V co-polymerizes with Type I fibrils to regulate their diameter, Type VII anchors epithelium to basement membrane — but they are not commercially supplemented and are of limited relevance outside specialized research settings. Supplement labels that advertise 'multi-collagen' products typically combine Type I, II, and III from bovine, marine, chicken, and eggshell membrane sources.
Type I — Skin, Bone, Tendon: The Structural Collagen
Type I collagen is the workhorse fibrillar collagen of mammalian physiology. Its triple helix consists of two alpha-1(I) chains and one alpha-2(I) chain that assemble into thick, densely packed fibrils, giving tensile strength to tendon, the organic matrix of bone, the reticular dermis, cornea, sclera, and fibrocartilage.
Most hydrolyzed collagen supplements derived from bovine hide, fish skin, or fish scales are predominantly Type I (typically 85–95%), with a small Type III fraction. When hydrolyzed to average molecular weights of roughly 2,000–5,000 Da, these peptides produce the dipeptide markers Pro-Hyp and Hyp-Gly in plasma after oral dosing.
Randomized trials of 2.5–10 g/day of hydrolyzed Type I collagen have reported modest improvements in skin elasticity, dermal density, and wrinkle depth over 8–12 weeks, with several meta-analyses supporting the skin endpoints. Outcomes in bone mineral density (postmenopausal women, 5 g/day specific bioactive collagen peptides over 12 months) and tendinopathy (15 g collagen hydrolysate with vitamin C alongside rehabilitation) have been reported in smaller trials but remain less firmly established. Marine Type I is often marketed as superior for skin due to its lower molecular weight distribution and faster absorption, though head-to-head data against bovine Type I remains limited.
Type II — Articular Cartilage and the Undenatured UC-II Mechanism
Type II collagen is the defining fibrillar collagen of hyaline cartilage, including articular surfaces of synovial joints, intervertebral discs, tracheal rings, and the nucleus pulposus. It is a homotrimer of three alpha-1(II) chains and interweaves with aggrecan-rich proteoglycans to give cartilage its shock-absorbing properties.
Type II collagen supplementation splits into two distinct products. Hydrolyzed Type II collagen (typically chicken sternum-derived) follows the same logic as hydrolyzed Type I: dose 2.5–10 g/day, rely on dipeptide absorption, modest joint-symptom outcomes in some trials. Undenatured Type II collagen (UC-II or native Type II) is a different product entirely. UC-II retains the intact triple-helix epitope and is dosed at just 40 mg once daily — roughly 1/250th the dose of hydrolyzed collagen.
The proposed mechanism for UC-II is oral tolerance: intact Type II epitopes interact with gut-associated lymphoid tissue in the Peyer's patches, inducing regulatory T-cell responses that dampen the Type II collagen-directed autoimmunity thought to contribute to osteoarthritis and rheumatoid arthritis pathology. Several randomized trials (including the Lugo et al. 2016 osteoarthritis trial) reported UC-II at 40 mg/day produced larger improvements in WOMAC pain scores than glucosamine-chondroitin at standard doses over 180 days, although head-to-head data remains limited and the oral-tolerance model is not universally accepted. UC-II and hydrolyzed collagen are not interchangeable — they work through different pathways at incompatible doses.
Type III — Vascular, Gut, and Early Wound Matrix
Type III collagen is a homotrimer of alpha-1(III) chains that co-distributes with Type I in most soft tissues but predominates in environments where elasticity and rapid turnover matter: large artery walls, lung parenchyma, the gut submucosa, reticular fibers of lymph nodes and spleen, and the provisional granulation-tissue matrix of early wound healing.
In dermal physiology, Type III forms a larger proportion of neonatal and young-adult skin and declines with aging, while Type I remains dominant throughout life. The Type I-to-Type III ratio in dermis shifts from roughly 4:1 in young skin to higher Type I dominance in aged skin, and this shift is hypothesized to contribute to the loss of skin resilience and wrinkle formation.
Type III is rarely sold as a standalone supplement; it typically accompanies Type I in bovine-hide hydrolysates (Type I and III together) or eggshell-membrane preparations (which contain Types I, V, and X alongside glycosaminoglycans). Commercial 'Type I and III' marketing usually reflects the natural co-distribution of these collagens in bovine skin rather than a specifically engineered ratio. Clinical data specifically isolating Type III effects from Type I effects is essentially absent — supplementation studies have not meaningfully separated the two.
Collagen Dipeptides (Pro-Hyp, Hyp-Gly): Absorption Pharmacokinetics
The pharmacokinetic case for oral collagen rests on a small number of dipeptides that survive gastrointestinal digestion and appear intact in systemic circulation. The two best-characterized are Pro-Hyp (prolyl-hydroxyproline) and Hyp-Gly (hydroxyprolyl-glycine), with lesser contributions from Gly-Pro-Hyp, Ala-Hyp, and Ser-Hyp.
After oral ingestion of 5–10 g of hydrolyzed collagen, plasma Pro-Hyp concentrations rise to roughly 20–80 nmol/mL and peak at 1–2 hours, with half-life in the range of 4–6 hours. Hyp-Gly appears at lower concentrations but with similar kinetics. These dipeptides resist further hydrolysis by intestinal and plasma peptidases because the imide bond adjacent to hydroxyproline is poorly recognized by common dipeptidyl peptidases.
In vitro and ex vivo work has shown Pro-Hyp stimulates fibroblast proliferation and chemotaxis, upregulates hyaluronic acid synthase expression in dermal fibroblasts, and chondrocyte activity in cartilage explants at concentrations plausibly achievable in tissue after oral dosing. Whether these effects drive the clinical outcomes observed in skin elasticity, wrinkle depth, and WOMAC pain trials is the active mechanistic question — the dipeptide-signaling hypothesis is the leading explanation but has not been definitively confirmed in humans.
Hydrolyzed (Gelatin-Like) vs Undenatured UC-II Dosing Contrast
The most common source of consumer confusion in the collagen category is the dose-range mismatch between hydrolyzed and undenatured products. A typical hydrolyzed collagen regimen is 5–10 g/day of powder stirred into coffee, water, or a smoothie; a typical UC-II regimen is a single 40 mg capsule per day. These are not competing doses of the same thing — they are different products with different mechanisms.
Hydrolyzed collagen is a source of absorbable peptide fragments and amino acids. Its intended mechanism is systemic delivery of Pro-Hyp and Hyp-Gly plus a supply of glycine, proline, and hydroxyproline precursors for endogenous collagen synthesis. Higher doses (up to 15 g/day for tendon studies) are used in some research protocols.
Undenatured Type II collagen (UC-II) is an immunomodulatory product that works at the level of intestinal lymphoid tissue. Its intended mechanism is induction of oral tolerance to Type II collagen epitopes. Higher doses of UC-II have been hypothesized to overwhelm the tolerance-induction window and reduce efficacy, consistent with oral-tolerance models from autoimmune research. Mixing a scoop of hydrolyzed Type II powder is not an equivalent to 40 mg of UC-II — the triple helix is destroyed by hydrolysis and the tolerance mechanism is lost.
Practical consequence: consumers buying collagen for joint symptoms should clarify which product they want. UC-II is the evidence-backed option for osteoarthritis at 40 mg/day; hydrolyzed Type II at 5–10 g/day is a reasonable but less strongly supported alternative.
Practical Takeaways: Food vs Supplement vs Injection
Dietary collagen (bone broth, skin-on fish, chicken skin, gelatin desserts) supplies modest amounts of collagen-derived amino acids and small peptide fragments. The predictability of dipeptide delivery is low because the degree of hydrolysis varies, and the total collagen yield per serving (roughly 5–10 g from a concentrated bone broth) is at the low end of research-dose ranges.
Supplement-grade hydrolyzed collagen is standardized to a lower molecular-weight distribution (typically average 2,000–5,000 Da) and delivers consistent Pro-Hyp and Hyp-Gly exposure. Bovine and marine sources produce broadly similar results in skin endpoints in available trials, with marine collagen marketed for faster absorption based on lower mean molecular weight. Specific bioactive collagen peptide preparations (Verisol, Fortigel, Tendoforte) are standardized products used in many of the pivotal trials.
Injectable collagen is a separate category entirely. Historical Zyderm-type injectable bovine collagen has been largely replaced by cross-linked hyaluronic-acid fillers and poly-L-lactic acid, and is not a collagen-peptide therapy. The injectable research-peptide space occasionally includes collagen-derived dipeptides and copper peptides (GHK-Cu) that interact with collagen turnover indirectly, but injected GHK-Cu is a signaling peptide, not a collagen substrate.
Stacking considerations: hydrolyzed collagen is frequently co-dosed with vitamin C (a cofactor for prolyl and lysyl hydroxylase) and with GHK-Cu (a copper-peptide signaling molecule that upregulates collagen synthesis). BPC-157 is sometimes discussed in connection with collagen for tendon repair, though direct evidence of BPC-157 augmenting oral-collagen effects is limited to mechanistic plausibility rather than confirmatory trials.
Frequently asked questions
Randomized trials support modest benefits of 2.5–10 g/day hydrolyzed collagen for skin elasticity, hydration, and wrinkle depth over 8–12 weeks, and 40 mg/day UC-II for osteoarthritis symptom scores. Effect sizes are moderate, not transformative, and responses vary by individual. The evidence is strongest for dermatologic endpoints and joint pain; bone and tendon data are preliminary.
UC-II (undenatured Type II collagen) retains the intact triple helix and works at 40 mg/day through oral tolerance in gut-associated lymphoid tissue. Hydrolyzed collagen is broken into small peptide fragments dosed at 2.5–10 g/day and works by delivering absorbable dipeptides like Pro-Hyp. They are mechanistically different products, not interchangeable.
Yes. Plasma pharmacokinetic studies show Pro-Hyp and Hyp-Gly appear in circulation within 1–2 hours after oral hydrolyzed collagen, with half-lives of roughly 4–6 hours. These dipeptides resist peptidase hydrolysis because of the imide bond adjacent to hydroxyproline. They are considered the main bioactive markers of oral collagen absorption in humans.
In practice, less than is often claimed. Hydrolyzed Type I, II, or III all deliver Pro-Hyp and Hyp-Gly upon absorption. For joint-specific oral tolerance, UC-II (undenatured Type II) at 40 mg/day is mechanistically distinct and does require Type II. For skin, available trials with Type I and Type I/III blends show broadly comparable effects.
Marine collagen (fish skin, scales) is predominantly Type I with a lower average molecular weight and slightly faster absorption than bovine. Bovine collagen (hide) is Type I with a Type III fraction. Head-to-head clinical comparisons are limited, and most trials show broadly comparable skin-endpoint effects. Marine is often chosen by pescatarians or those avoiding bovine sources.
Mechanistically plausible. GHK-Cu is a signaling tripeptide that upregulates collagen synthesis and remodeling pathways; oral collagen supplies precursor substrate. BPC-157 has animal data for tendon and ligament repair. Direct human trial evidence of oral collagen combined with these peptides is limited, so any stacking is extrapolation rather than confirmed protocol.