Collagen Supplementation: Types, Doses, and What the Evidence Actually Supports

Collagen is the most abundant protein in your body. It forms the structural scaffolding for skin, tendons, cartilage, bones, and blood vessels. Production declines roughly 1% per year after age 20, which is why skin loses elasticity, joints stiffen, and recovery slows. Supplementing with collagen peptides has become one of the fastest-growing segments in the supplement market.

The question is whether swallowing collagen actually does anything useful, or whether your digestive system simply breaks it down into generic amino acids indistinguishable from those in a chicken breast. The answer is more mixed than either side of that debate usually admits.

Collagen types: what they are and where they work

Your body produces at least 28 types of collagen. Three of them account for the vast majority, and they are the ones relevant to supplementation.

Type I

Makes up roughly 90% of the collagen in your body. Found in skin, tendons, bones, teeth, and connective tissue. Most collagen supplements on the market contain Type I, typically sourced from bovine hides or marine (fish) sources. When people talk about collagen for skin health or general anti-ageing, Type I is what they mean.

Type II

Found primarily in cartilage. This is the relevant type for joint health, particularly osteoarthritis. Undenatured Type II collagen (UC-II) works through a different mechanism than hydrolyzed collagen peptides: it modulates the immune response to cartilage degradation rather than providing structural building blocks. The dose is correspondingly lower (40 mg vs several grams).

Type III

Found alongside Type I in skin, blood vessels, and organs. It contributes to skin elasticity and vascular integrity. Most bovine collagen supplements contain a mix of Types I and III. Marine collagen is almost exclusively Type I.

Hydrolyzed collagen peptides: what the digestion argument misses

The common objection: collagen is a protein, your stomach breaks it down into amino acids, and your body rebuilds whatever it needs from those amino acids regardless of the source. This is partly true and partly wrong.

Hydrolyzed collagen (collagen peptides) is pre-broken down into small peptide fragments, typically 2 to 5 amino acids long. Some of these specific peptides, particularly hydroxyproline-containing dipeptides and tripeptides, are absorbed intact and detectable in blood within hours. A 2005 study in the Journal of Agricultural and Food Chemistry showed that prolyl-hydroxyproline and hydroxyprolyl-glycine reached the bloodstream as intact peptides after oral ingestion.

These specific peptides appear to act as signaling molecules, stimulating fibroblasts to produce more collagen and hyaluronic acid. The mechanism is not 'eat collagen, become collagen.' It is closer to 'eat collagen peptides, signal your cells to ramp up their own collagen production.' This distinction matters because it explains why collagen supplements can show effects that an equivalent dose of generic protein does not.

Evidence for skin

Skin elasticity and hydration are the best-studied outcomes for collagen supplementation. A pattern of positive results runs through the research, though trial quality varies.

A 2019 systematic review in the Journal of Drugs in Dermatology analysed 11 studies with 805 participants. The pooled data showed improvements in skin elasticity, hydration, and dermal collagen density compared to placebo. Doses ranged from 2.5 to 10 grams daily, with most trials running 8 to 12 weeks. The effect sizes were modest but consistent.

A well-designed 2014 RCT by Proksch and colleagues found that 2.5 grams of specific collagen peptides (Verisol) daily for 8 weeks significantly improved skin elasticity in women aged 35 to 55, with effects persisting 4 weeks after stopping supplementation. A follow-up trial by the same group found improvements in skin roughness and collagen density at the same dose.

The caveats: many of these trials are funded by collagen manufacturers. The effect sizes, while statistically significant, are modest. And the outcomes (skin elasticity measured by cutometry, hydration measured by corneometry) are objective but may not translate to visible differences that people notice in the mirror. Setting expectations correctly matters here.

Evidence for joints

Joint health evidence splits along two lines: hydrolyzed collagen peptides for general joint comfort, and undenatured Type II collagen (UC-II) for osteoarthritis.

Hydrolyzed collagen for joint pain

A 2008 study in Current Medical Research and Opinion followed 147 athletes with activity-related joint pain. Those taking 10 grams of collagen hydrolysate daily for 24 weeks showed significant improvements in joint pain during walking, standing, at rest, and while carrying objects. The comparison was placebo, and the study ran long enough to be meaningful.

A 2017 systematic review found moderate evidence that collagen hydrolysate at 10 grams daily reduces joint pain in osteoarthritis, though the authors noted high heterogeneity across trials and called for larger, better-designed studies.

UC-II for osteoarthritis

Undenatured Type II collagen works differently. At 40 mg daily, it exposes the immune system to small amounts of Type II collagen, training it to stop attacking cartilage in the joints. This is oral tolerance, the same principle behind some allergy desensitisation protocols.

A 2016 RCT published in the International Journal of Medical Sciences compared UC-II (40 mg daily) against glucosamine plus chondroitin (1500 mg + 1200 mg daily) in 191 participants with knee osteoarthritis over 180 days. UC-II showed significantly greater improvement in WOMAC scores (the standard osteoarthritis assessment) than the glucosamine-chondroitin combination. That result got attention because glucosamine-chondroitin had been the standard joint supplement recommendation for decades.

Dose ranges

• Skin trials: the trials used 2.5 to 10 grams of hydrolyzed collagen peptides daily, most of them 2.5 to 5 grams, over 8 to 12 weeks. Higher doses did not produce better measurements. Collagen has no authorised wording on the GB nutrition and health claims register, so those are trial doses attached to trial outcomes, not a dose to take for your skin.
• Joint pain trials: 10 grams of hydrolyzed collagen daily was the dose in most of the trials that reported a change in joint pain scores. Lower doses have less data behind them either way.
• UC-II trials: 40 mg of undenatured Type II collagen daily. This is not a typo: the proposed mechanism is immune modulation rather than structural supplementation, so the dose is in milligrams. Osteoarthritis is diagnosed and managed by a clinician, and nothing here treats it.
• Bone trials: 5 grams daily, in a small number of trials whose results are too thin to conclude anything from. Bone density is a clinical measure your GP interprets, not a supplement target.

The vitamin C requirement

Your body cannot synthesize collagen without vitamin C. Specifically, vitamin C is a required cofactor for prolyl hydroxylase and lysyl hydroxylase, the enzymes that stabilize the collagen triple helix structure. Without adequate vitamin C, the collagen your fibroblasts produce is structurally weak and cannot form proper fibres. This is literally what scurvy is: collagen synthesis failure.

Most collagen supplement studies do not control for vitamin C intake. If participants had adequate vitamin C status (likely in well-nourished Western populations), the collagen peptides had the raw materials to work with. If your vitamin C intake is marginal, collagen supplementation may produce weaker results. Taking 100 to 200 mg of vitamin C alongside your collagen dose is a reasonable insurance policy.

Source matters less than you think

Marine collagen (from fish skin and scales) is marketed as superior to bovine collagen. The claim is that marine collagen has smaller peptide fragments and better bioavailability. The evidence for this is thin. A few in vitro studies suggest higher bioavailability for fish-derived peptides, but head-to-head human trials comparing marine vs bovine collagen for the same outcome are scarce.

The practical difference: marine collagen is almost exclusively Type I. Bovine collagen typically contains Types I and III. If your goal is skin health, either works. If you want Type III for vascular or gut health, bovine is the better source. Marine collagen is more expensive, tastes milder, and is suitable for pescatarians. Those are valid reasons to choose it. Superior efficacy is not, based on current evidence.

What collagen cannot do

Collagen supplements will not reverse deep wrinkles, rebuild severely damaged cartilage, or replace medical treatment for osteoarthritis. They will not make you look 20 years younger. The evidence supports modest, measurable improvements in skin elasticity, hydration, and joint comfort over 8 to 24 weeks. That is a useful outcome. It is not the transformation that some marketing implies.

How Stack Almanac handles this

Stack Almanac keeps the collagen type and the amount in your record, dated, so a switch between hydrolysed peptides and UC-II is visible later rather than remembered. The amounts each trial used sit on the ingredient page with their sources, for you to read against your own label. The app does not rule on whether your product is the right one, and a diagnosed joint condition is a conversation for your GP.

The timing information sits beside the ingredient, with its sources: collagen is absorbed better on a relatively empty stomach, so 30 minutes from meals and other protein supplements is the usual advice. Your record shows which of your supplements share a time block, so you can see where that advice collides with the day you actually have. Moving anything is your decision.