Undenatured Type II Collagen (UC-II): What the Trials Show

This content is for informational purposes only and is not medical advice. Collagenium is published by a research-driven editorial team, not licensed medical professionals. See our editorial standards and methodology. Consult a registered dietitian, physician, or pharmacist before starting any supplement, especially if you have existing health conditions, are pregnant or breastfeeding, or take medications.

Undenatured type II collagen, sold mostly under the UC-II trademark, is intact collagen from chicken breastbone cartilage taken at just 40 mg a day. It is not a protein dose. It is thought to work by calming the immune reaction against cartilage, and ten controlled trials mostly report modest gains in knee pain, stiffness and range of motion over 90 to 180 days. Most of those trials were paid for by the companies that sell it.

That last point matters, and it is the reason this guide exists. The top results for this topic are research papers written for scientists and veterinarians, product pages, and an Amazon listing. Below we explain the mechanism in plain English, put every human trial we could verify in one table with its size, duration, comparator and funding, show why 40 mg of UC-II and 10 g of hydrolyzed collagen are not the same thing, and state what the evidence does not show.

What is undenatured type II collagen?

Undenatured type II collagen is collagen taken from chicken sternum (breastbone) cartilage and processed at low temperature, without the heat, acids or enzymes that normally break collagen apart, so the protein keeps its natural folded shape (Arthritis Foundation, 2022). “Native” type II collagen means the same thing. Type II makes up about 90% of the collagen in joint cartilage (Arthritis Foundation, 2022), which is why it is the type aimed at joints rather than skin.

Three details clear up most of the confusion around this ingredient:

• UC-II is a brand, not a generic term. It is the registered trademark of the most studied ingredient of this kind, developed by InterHealth Nutraceuticals and now owned by Lonza, as the NOW Foods label states. Other branded undenatured ingredients exist, including NEXT-II from Japan and a native type II collagen from the Spanish company Bioiberica, and each has its own, smaller set of trials.
• 40 mg is not 40 mg of collagen. A 40 mg dose of UC-II is 40 mg of standardized chicken cartilage powder, which contains about 10 mg of total collagen, part of it undenatured type II. The NOW Foods supplement facts panel says exactly this (“40 mg, providing 10 mg total collagen”), and a 2020 review puts the figure at 10.4 mg of native type II collagen per 40 mg dose (Gencoglu et al., 2020).
• It is roughly one thousandth of a peptide dose. A typical scoop of hydrolyzed collagen delivers 10 g. Ten milligrams cannot matter as nutrition. The whole argument for undenatured collagen rests on its shape, not its amount.

How does undenatured type II collagen work?

The proposed mechanism is called oral tolerance, an immune process that happens in the gut. Most of the detailed evidence for it comes from animals, so we grade the mechanism in humans as preliminary: plausible and consistent in rodents, but barely measured in people.

Here is the idea in plain English:

1. Your gut tolerates food proteins every day. Patches of immune tissue in the wall of the small intestine, called Peyer’s patches, sample what passes through and decide what to ignore and what to attack.
2. Intact type II collagen looks like your own cartilage. When undenatured collagen reaches those patches with its structure intact, it carries the same recognition sites, called epitopes, as the type II collagen in your knees. Repeated small exposure is thought to train regulatory T cells to treat type II collagen as harmless (Bagchi et al., 2002).
3. Those cells travel to the joint. When they meet type II collagen in cartilage, they release calming signals, including interleukin-10 and TGF-beta, that dampen the inflammatory attack thought to contribute to cartilage wear (Gencoglu et al., 2020).

Two consequences follow. First, shape is everything: hydrolyzed collagen is chopped into short peptides and loses its epitopes, and laboratory assays can detect those epitopes in undenatured but not in hydrolyzed collagen (Bagchi et al., 2002; Harris et al., 2021). Second, small doses are deliberate. In oral tolerance, low doses favor the regulatory response. In a 1998 multicentre trial of oral type II collagen in 274 people with rheumatoid arthritis, only the lowest dose tested, 20 micrograms a day, reached statistical significance, and only on one of three response criteria (Barnett et al., 1998).

The approach began with rheumatoid arthritis. A 1993 trial published in Science gave 60 people with severe rheumatoid arthritis chicken type II collagen for three months and reported fewer swollen and tender joints than placebo (Trentham et al., 1993). The larger 1998 follow-up was largely negative, and no modern UC-II trial has been run in rheumatoid arthritis, so that early promise remains preliminary.

What has been shown directly in people? Symptom scores and knee angles, not immune cells. The regulatory T cell and cytokine findings come from rodents, including a 2024 mouse study, run through a Lonza-partnered PhD programme, in which oral undenatured type II collagen roughly halved the incidence of experimental arthritis and reshaped gut immune signalling and the microbiome (Communications Biology, 2024). In humans, one trial measured urinary markers of cartilage breakdown and found no change after three months (Bakilan et al., 2016); another found a difference in a cartilage-degradation marker only in a subgroup (Möller et al., 2026).

How is it different from hydrolyzed collagen?

They are different molecules, different doses and different mechanisms, and they are not interchangeable. A product that contains both is two separate ingredients in one capsule, not a stronger version of either.

	Undenatured type II collagen (UC-II)	Hydrolyzed collagen (collagen peptides)
What it is	Intact, folded collagen protein from chicken sternum cartilage	Collagen broken into short peptide fragments by enzymes
Usual source	Chicken sternum cartilage only	Bovine hide, fish skin, chicken, eggshell membrane
Daily dose in trials	40 mg of UC-II (about 10 mg collagen)	2.5 to 15 g
Proposed mechanism	Immune signalling in the gut (oral tolerance)	Absorbed peptides supply amino acids and may signal cells to make collagen
What trials measured	Knee pain, stiffness and function (WOMAC), knee range of motion, time to discomfort during exercise	Skin hydration and elasticity, activity-related joint pain, bone density, nail growth
Typical trial length	90 to 180 days	8 to 24 weeks
Format	One small capsule	Powder scoop or several large capsules
Does more help?	Not shown; the mechanism predicts low doses work best	Dose ranges are gram-level by design

Sources for the hydrolyzed column: the 2.5 to 15 g daily range is the Arthritis Foundation’s summary of hydrolyzed collagen trials (Arthritis Foundation, 2022); the trial lengths and outcomes are those of the skin, joint, bone and nail trials summarised, with their sources, in our hydrolyzed collagen guide. The undenatured column is drawn from the trial table below.

The practical trap is the word “type II” on multi-collagen powders. Those products contain hydrolyzed chicken cartilage, not a standardized 40 mg undenatured dose, and they are not substitutes for a UC-II product. Our multi-collagen guide and best collagen for joints roundup cover that distinction product by product. The reverse mistake is just as common: 40 mg of a hydrolyzed powder is a pinch of protein and has no trial behind it, and nobody has studied 10 g of UC-II, because the mechanism gives no reason to.

What do the human trials show?

We found ten controlled human trials of undenatured collagen that we could open and verify, published between 2009 and 2026, with 39 to 191 participants and 12 to 26 weeks of supplementation where the abstracts state them. Most report modest improvements over placebo or a comparator. Most were funded or staffed by the companies that make the ingredient. The one fully independent placebo-controlled trial found no benefit.

Trial	Who (n)	Dose and duration	Compared with	Main result	Funding
Crowley et al., 2009	52 adults with knee osteoarthritis	UC-II 40 mg, 90 days	Glucosamine + chondroitin (no placebo arm)	WOMAC score down 33% vs 14%; pain score down 40% vs 15%	InterHealth (UC-II maker)
Lugo et al., 2013	55 healthy adults with knee discomfort after a step test	UC-II 40 mg, 120 days	Placebo	Knee extension 81 vs 74 degrees; exercised longer before discomfort	InterHealth; three authors employees
Lugo et al., 2016	191 adults with knee osteoarthritis	UC-II 40 mg, 180 days	Placebo and glucosamine + chondroitin	Total WOMAC lower than placebo (p = 0.002) and than G+C (p = 0.04)	InterHealth sponsor; two authors employees
Bakilan et al., 2016	39 adults with knee osteoarthritis	Native type II collagen 10 mg plus acetaminophen, 3 months	Acetaminophen alone	Walking pain better; urinary cartilage-breakdown markers unchanged	University (Turkey); no industry link stated
Schön et al., 2022	96 healthy adults aged 20 to 55 with activity-related knee discomfort	UC-II 40 mg, 24 weeks	Placebo	Knee flexion gained 3.2 vs 0.2 degrees	Lonza (UC-II owner); three authors employees
Sadigursky et al., 2022	105 women aged 60 to 80 with knee osteoarthritis	UC-II, 90 days	No supplement (open label)	Pain score change minus 3.8 vs minus 1.3; WOMAC minus 9.5 vs minus 1.3	None declared
Santana et al., 2022	Women with knee osteoarthritis (group sizes not in abstract)	UC-II one capsule daily (amount not in abstract)	Twelve-session exercise programme, and no intervention	Walking tests improved vs control, as did exercise; only exercise improved total WOMAC	University (Brazil, Portugal); no industry link stated
Shiojima et al., 2022	64 healthy Japanese adults	NEXT-II 3.2 mg undenatured type II, 12 weeks	Placebo	Knee flexion range, walking and stair tests improved	Manufacturer (Ryusendo) authors
Yuenyongviwat et al., 2025	68 adults with knee osteoarthritis	UC-II plus hydrolysed collagen tablets, 12 weeks	Placebo	No difference in pain, function, rescue medication or satisfaction	Thai medical council foundation; no industry ties
Möller et al., 2026	74 healthy adults with joint discomfort during activity	Native type II collagen 40 mg, 180 days	Placebo	Both groups improved; earlier improvement and subgroup differences only	Bioiberica (maker) employees and consultant

A few things to notice when you read that table. Only two trials had more than 100 participants. Four had no placebo arm, comparing UC-II with glucosamine and chondroitin, with acetaminophen alone, with exercise, or with nothing. The knee range-of-motion gains in healthy adults are a few degrees, and no trial has established what a 3-degree change means for daily life. And the placebo groups improved too, in almost every study, which is normal in joint research and is why the between-group difference is the number that counts.

For knee osteoarthritis symptoms

Some evidence. Three controlled trials favour UC-II on the WOMAC pain, stiffness and function index over 90 to 180 days (Crowley et al., 2009; Lugo et al., 2016; Sadigursky et al., 2022), and the 2016 trial is the strongest because it was placebo-controlled, multicentre and the largest. Against that, the 2025 Thai trial of a combined UC-II and hydrolysed collagen tablet found nothing beyond placebo at 12 weeks (Yuenyongviwat et al., 2025). It was shorter, used a combination product and had 68 participants, so it does not cancel the earlier trials, but it is the only one with no manufacturer involvement, and it was negative.

For active adults with knee discomfort

Some evidence, and small. Three placebo-controlled trials in healthy people who get knee discomfort with exercise report better knee extension or flexion and longer time to discomfort (Lugo et al., 2013; Schön et al., 2022; Möller et al., 2026). The 2026 trial is the most honest about its limits: both groups improved on the standard knee questionnaire, and the collagen group mainly improved sooner, with between-group differences appearing in subgroups rather than overall.

Compared with exercise

Preliminary. A second 2022 Brazilian trial, separate from the Sadigursky study and run by a different university group, compared twelve exercise sessions with daily UC-II and with no intervention in women with knee osteoarthritis. Both active groups improved walking tests over the control, but only exercise improved the overall WOMAC score and stiffness (Santana et al., 2022). It was open label and the abstract does not report group sizes, so it is one trial, not a verdict, but it points the same way as every joint guideline: a capsule is not a replacement for movement.

What do the systematic reviews conclude?

The independent reviews are more cautious than the trial abstracts. The most rigorous one, a 2018 meta-analysis of 69 trials and 20 supplements in the British Journal of Sports Medicine, found that undenatured collagen produced a statistically significant short-term pain improvement of unclear clinical importance, that it was one of only two supplements with a clinically important pain effect at medium term, and that no supplement had a clinically important effect long term (Liu et al., 2018). An earlier 2012 systematic review re-examined the Crowley trial and concluded that undenatured collagen showed no significant between-group difference against glucosamine and chondroitin, and called for more independent studies (Van Vijven et al., 2012).

Two newer reviews are warmer but say the same about quality. A 2023 review pooled eight trials with 243 UC-II recipients, reported better WOMAC and pain scores than placebo, and described the literature as limited (Kumar et al., 2023). A 2025 network meta-analysis ranked it third of 21 supplements for WOMAC pain, with low-quality evidence overall (Chen et al., 2025).

Putting the trials and reviews together, here is how we grade the main claims:

Claim	Our grade	Why
Less knee pain and stiffness in knee osteoarthritis over 3 to 6 months	Some evidence	Three positive controlled trials, one independent negative trial, meta-analyses call the effect real but modest
Better knee range of motion in active adults	Some evidence	Three small placebo-controlled trials, gains of a few degrees
Works through oral tolerance at 40 mg	Preliminary in humans	Mechanism shown in animals and lab assays, not measured in trial participants
Better than glucosamine and chondroitin	Preliminary	One trial for (Lugo 2016), one reanalysis against (Van Vijven 2012)
As useful as an exercise programme	Preliminary, leaning no	One open-label trial; exercise matched UC-II on walking tests and beat it on WOMAC (Santana 2022)
Helps rheumatoid arthritis	Preliminary	One small positive 1993 trial, one large largely negative 1998 trial, no modern UC-II trial
Rebuilds or regrows cartilage in humans	Marketing claim	No human imaging study; the one trial that measured cartilage-breakdown markers found no change

What the evidence does not show

• Independence. The Crowley, both Lugo and the Schön trials were funded by InterHealth or Lonza, the owners of UC-II, and list company employees as authors. The Möller trial authors work for or consulted for Bioiberica; the Shiojima authors work for the NEXT-II manufacturer. Only the Bakilan, Sadigursky, Santana and Yuenyongviwat trials have no manufacturer involvement, and the only one of those that was placebo-controlled was negative.
• Size and duration. No trial ran longer than six months, and no trial has reported what happens when people stop. Long-term benefit and long-term safety are both unknown.
• Structure. Every human outcome is a questionnaire score or a goniometer angle. No human trial has imaged cartilage before and after UC-II, and urinary cartilage-breakdown markers did not improve in the one trial that measured them (Bakilan et al., 2016).
• Interchangeability. A 2021 laboratory analysis of commercial powders labelled undenatured type II collagen found marked differences in collagen fibre size and in antigenic configuration between products, and concluded they should not be assumed interchangeable (Harris et al., 2021). The trials used specific branded ingredients. A generic capsule may not contain the same structure.
• Effect size. Even the positive trials describe improvements, not resolution. People in placebo groups improved as well, and the independent meta-analysis called the short-term pain effect of unclear clinical importance (Liu et al., 2018).

How much did the trials use, and for how long?

Almost every UC-II trial used 40 mg once a day, and that is the amount on most labels. Trials of other undenatured ingredients used different amounts, 10 mg of native type II collagen in the Turkish trial and 3.2 mg of NEXT-II in the Japanese one, so the right dose depends on the branded ingredient, not on the words “type II collagen”. Trials ran 90 to 180 days. In the 2013 trial, knee extension first separated from baseline at day 90 (Lugo et al., 2013); the 2016 osteoarthritis trial reported its primary result at day 180 (Lugo et al., 2016). Anyone trying it should expect to wait three months before judging.

On timing, the Crowley trial gave UC-II in the evening, the NOW Foods label suggests bedtime on an empty stomach, and Examine suggests before breakfast. No trial has compared times of day, so timing advice is label practice, not evidence. We do not give personalised dosing; a pharmacist or physician can tell you whether 40 mg a day fits alongside what you already take.

Is it safe, and who should be cautious?

Short-term tolerability is well-supported for up to six months. The 2013 trial reported no product-related adverse events (Lugo et al., 2013), the 2016 trial found no safety differences between UC-II, glucosamine-chondroitin and placebo (Lugo et al., 2016), the 2025 trial reported no adverse effects in either group (Yuenyongviwat et al., 2025), and the 2018 meta-analysis found no safety difference between any supplement and placebo except diacerein (Liu et al., 2018). A 2010 toxicology programme in rats and rabbits found no toxicity at an oral dose above 5,000 mg per kilogram, no mutagenicity, and no organ or blood chemistry changes over 90 days (Marone et al., 2010).

Where data are missing, caution replaces evidence:

• Chicken allergy. The ingredient is chicken cartilage.
• Pregnancy and breastfeeding. Excluded from every trial; no data.
• Autoimmune disease or immune-suppressing medication. The mechanism is immune modulation. The 1990s rheumatoid arthritis trials reported no side effects, but no modern UC-II trial has included people taking biologics or immunosuppressants. Ask a rheumatologist or pharmacist first, and never substitute a supplement for prescribed therapy.
• Children. Adult-only trials.

Our collagen side effects guide covers the general collagen picture, which is mostly about hydrolyzed products.

Can you get it from food?

Not in any realistic way. Type II collagen does sit in cartilage, which is why chicken feet, chicken breastbone and trachea are the raw material, but undenatured collagen is by definition collagen that has not been exposed to heat, acids or enzymes, and cooking turns collagen into gelatin (Arthritis Foundation, 2022). Denatured collagen lacks the intact epitopes that the oral tolerance mechanism depends on (Bagchi et al., 2002; Harris et al., 2021). So bone broth and simmered chicken feet deliver gelatin and amino acids, which is the hydrolyzed collagen story, not the undenatured one. Eating raw cartilage is not a sensible alternative, and the trials used a low-temperature manufacturing process to preserve structure, not kitchen methods (Gencoglu et al., 2020).

How do you read a label, and what does our catalogue have?

Because branded ingredients differ (Harris et al., 2021), the label matters more here than for most supplements. A checklist:

• Look for “UC-II” with the registered mark, or “undenatured” or “native” type II collagen with a stated milligram amount, usually 40 mg.
• Expect a line such as “providing 10 mg total collagen”. That is normal and correct, not a shortfall.
• One capsule a day means a 60-count bottle is 60 servings; compare prices per capsule.
• “Type II collagen” listed inside a multi-collagen powder measured in grams is hydrolyzed, not undenatured.
• Check the other ingredients. The NOW Foods capsule, for example, also contains 80 mg of calcium from seaweed minerals.

We do not currently review a standalone 40 mg UC-II capsule, so we can only be precise about what our catalogue does contain. AlgaeCal’s collagen powder lists UC-II among four branded ingredients alongside hydrolyzed peptides for skin and bone. At 55.80 dollars for 30 servings it works out to about 1.86 dollars a serving, and our listing does not record the UC-II milligrams, so check the supplement facts panel for the 40 mg figure before buying it for that reason.

NOW Foods BioCell Collagen is the product people most often mistake for UC-II. It is hydrolyzed type II collagen from chicken sternum with naturally occurring chondroitin and hyaluronic acid. It has its own trials, but it is a different molecule and does not deliver the undenatured mechanism described above.

The bottom line

Undenatured type II collagen is the rare supplement whose tiny dose is a feature of its mechanism rather than a sign of under-dosing. For adults with knee osteoarthritis symptoms, or active adults with knee discomfort, 40 mg a day for three to six months has some evidence of modest benefit, most of it produced by the companies that sell it and one independent trial finding none. For anyone who wants the broader hydrolyzed collagen effects on skin, bone and athletes’ joints, it is the wrong tool, and our collagen for joint health guide compares the two bodies of evidence side by side.

Taking both is reasonable in principle, because they act differently, but the only trial of a combined tablet found no benefit over placebo at 12 weeks. If you try UC-II, pick a product that names the branded ingredient and its milligram amount, give it 90 days, keep moving, and judge it on your own pain and stiffness rather than on the marketing.