Collagen Types and Forms: What the Distinctions Actually Mean
Type labels on supplements matter less than whether the product is hydrolysed and at a trial dose, with one genuine exception that works by a different mechanism entirely.
The Short Answer
Collagen supplements are marketed by type, with type I and III for skin and type II for joints. That framing implies the ingested type is routed to the corresponding tissue, which is not how digestion works. What matters is whether the product is hydrolysed into absorbable peptides and whether the dose matches the trials. There is one genuine exception, undenatured type II collagen, which works by an entirely different mechanism and at a tiny dose.
What Happens to Ingested Collagen
Collagen is a protein. Digestion breaks proteins into amino acids and small peptides, and those enter circulation as a pool rather than as a labelled cargo destined for a particular tissue.
The intuitive objection is therefore fair: eating type I collagen does not deliver type I collagen to skin, and the body synthesises whichever collagen type a tissue requires from available amino acids.
The counterargument that makes hydrolysed collagen interesting: specific di- and tripeptides, particularly prolyl-hydroxyproline, survive digestion, appear measurably in plasma, and may act as signals to fibroblasts rather than only as substrate. That is a mechanistically distinct claim from simple amino acid provision, and it has some support.
What follows practically: the type label on a hydrolysed product is largely marketing, since the peptide profile matters more than the source type. Hydrolysis, peptide profile and dose are the variables, and the type is not.
The amino acid point. Collagen is not a complete protein, being low in tryptophan, so it should not count toward a protein target as though it were whey or meat. It is a supplement rather than a protein source.
The Forms That Actually Differ
| Form | What it is | Typical dose |
|---|---|---|
| Hydrolysed collagen peptides | Enzymatically broken into absorbable peptides | 2.5 to 15 g daily |
| Gelatin | Partially hydrolysed; gels when cooled | Similar amounts; less convenient |
| Undenatured type II collagen | Native structure retained; acts through oral tolerance | 40 mg daily |
| Bone broth | Variable and generally low collagen content | Not a reliable dose |
| Marine, bovine or porcine source | Different source animals | Peptide profile matters more than species |
The undenatured type II row is the genuine exception and it is worth understanding, because the dose difference is the giveaway. At 40 mg daily it cannot be working as substrate. The proposed mechanism is immunological, involving oral tolerance and modulation of the immune response to joint collagen, which is a completely different pathway from hydrolysed peptides at 10 g.
Bone broth is worth naming because it is widely recommended as a collagen source. Collagen content varies enormously with preparation and is generally low relative to a supplemental dose, which makes it a food rather than a reliable intervention.
What the Evidence Supports
Skin. The best-evidenced area. Multiple small randomised trials report improvements in hydration and elasticity using instrumental measures such as corneometry and cutometry over 8 to 12 weeks, with some reporting reduced wrinkle depth. Meta-analyses of these trials report positive pooled effects. Doses of 2.5 to 10 g.
Joint symptoms. Small trials in athletes and in people with activity-related knee pain report symptomatic improvement with hydrolysed peptides, and separate trials of undenatured type II collagen at 40 mg report benefit in osteoarthritis.
Tendon. Trials report improved tendon properties when collagen is combined with a loading protocol, which is the important qualifier: the loading is doing substantial work, and collagen without loading is unlikely to reproduce the result.
Bone. Limited trial data in postmenopausal women on bone density markers.
The caveats that travel with all of it: small samples, frequent industry funding, short durations, modest effect sizes, and product-specific peptide profiles that may not transfer between brands.
Where the evidence is weakest: hair and nails, gut healing, and general anti-ageing claims.
How to Use It, If You Do
Dose. 2.5 to 10 g of hydrolysed peptides for skin outcomes, and around 15 g for tendon work in the trials that used a loading protocol. For undenatured type II, 40 mg, which is not interchangeable.
With vitamin C. Ascorbate is a required cofactor for the hydroxylation steps in collagen synthesis, so adequate vitamin C is a sensible pairing rather than an upsell.
Timing for tendon: trials took it around an hour before a loading session, on the reasoning that peptide availability coincides with the loading stimulus.
Expect 8 to 12 weeks before assessing, which is the trial window and the timescale on which skin and tendon remodel.
Loading is not optional for the tendon and joint use case. The compound is an adjunct to progressive loading, which has the strongest evidence in tendinopathy management by a wide margin.
Judge it on what you added it for, which for skin means standardised photographs at three-month intervals and for tendon means symptom and function change.
Consider the alternative use of the money. For skin, sunscreen and a retinoid have substantially better evidence. For tendon, the loading protocol is the intervention.
Quality and Practical Notes
Hydrolysed or not. The single most important label check, since unhydrolysed collagen is a poorly absorbed protein.
Peptide profile, where specified. Different processing produces different peptide distributions, and trial results attach to specific preparations. This is a reason a named trial-matched product has a legitimate premium here.
Source and allergy. Marine collagen is derived from fish, which matters for fish allergy. Bovine and porcine sources matter for dietary restrictions. There is no vegan collagen, since collagen is an animal protein; products marketed as vegan collagen supply cofactors and amino acids rather than collagen.
Contaminants. Heavy metals have been found in some collagen products, and third-party testing is worth having.
Flavour and tolerance. Generally well tolerated, with mild gastrointestinal effects in some people.
Not a protein source. Low in tryptophan and incomplete, so it should not displace complete protein in a protein target.
Interactions: minimal, which makes it a low-risk addition compared with several botanicals.
The honest summary: hydrolysed peptides at a trial dose with vitamin C, judged over 8 to 12 weeks, is a reasonable purchase with modest expectations. The type labels are marketing, undenatured type II is a genuinely different product, and the topicals still have better skin evidence.
The AEONNN Perspective
AEONNN reads collagen by form and dose rather than by type, because the type labels imply a routing that digestion does not perform. What makes hydrolysed collagen mechanistically interesting is narrower and more specific: prolyl-hydroxyproline and similar peptides survive digestion, appear in plasma, and may act as fibroblast signals rather than only as substrate.
Undenatured type II collagen is the one genuine exception in this category, and the dose gives it away. At 40 mg daily it cannot be acting as substrate, and the proposed mechanism is immunological through oral tolerance, which makes it a different product rather than a different type.
The qualifier the platform attaches to the tendon evidence is the one the marketing omits: several positive trials combined collagen with a progressive loading protocol, and the loading is doing substantial work. AEONNN recommends it as an adjunct to loading with adequate vitamin C as a cofactor, never as a substitute. And for skin the platform makes the comparison anyway, since sunscreen and a topical retinoid have substantially better evidence than any oral compound here.
Pillar Matrix mapping
Skin and Extracellular Matrix, Structural and Musculoskeletal Support
Database Matrix layers
- Pharmacokinetics Layer (HMDB, PubChem)
- Evidence Layer (PubMed, Cochrane, ClinicalTrials.gov)
- Quality / Formulation Layer (ConsumerLab, Labdoor)
- Mechanistic Layer (KEGG, Reactome, UniProt)
Frequently Asked
Do collagen types matter in supplements?
Largely no. Digestion breaks collagen into amino acids and peptides that enter a general pool, so ingested type I is not routed to skin. Hydrolysis, peptide profile and dose are the variables.
What makes hydrolysed collagen different from other protein?
Specific di- and tripeptides, particularly prolyl-hydroxyproline, survive digestion and appear in plasma, and may act as signals to fibroblasts rather than only as substrate.
What is undenatured type II collagen?
A genuinely different product working through oral tolerance at around 40 mg daily, which at that dose cannot be acting as substrate. It is not interchangeable with hydrolysed peptides.
How much should I take?
2.5 to 10 g of hydrolysed peptides for skin outcomes, around 15 g for tendon work in the trials that combined it with loading, and 40 mg for undenatured type II.
Does collagen work for tendons?
Trials report improved tendon properties when collagen is combined with a progressive loading protocol. The loading is doing substantial work, and collagen without loading is unlikely to reproduce it.
Is bone broth a good collagen source?
Collagen content varies enormously with preparation and is generally low relative to a supplemental dose, which makes it a food rather than a reliable intervention.
Is there a vegan collagen?
No. Collagen is an animal protein, and products marketed as vegan collagen supply cofactors and amino acids rather than collagen itself.
Evidence and review
Any dosage ranges cited here reflect the ranges used in published human trials, not personal recommendations. Evidence in this field moves, so this article is reviewed quarterly and carries its last-updated date above. Nothing here is intended as medical advice, and supplementation should be discussed with a qualified clinician, particularly alongside prescribed medication or an existing condition.