Collagen for Knee Osteoarthritis: What the Trials Show

    Collagen for Knee Osteoarthritis: What the Trials Show

    Collagen may ease knee osteoarthritis slightly, but evidence is weak and industry-funded. Trials, effect sizes, UC-II vs peptides and guideline advice.

    Written by Dimitrun Irene. Updated November 19, 2026.

    Collagen may ease knee osteoarthritis symptoms slightly, but the evidence is weak, the benefit is modest, and almost every trial was funded by a company that sells collagen. Osteoarthritis (OA) is the common "wear and tear" form of arthritis, in which the cartilage that cushions the ends of the bones slowly breaks down, causing pain and stiffness. Several meta-analyses, which pool the results of many trials, report small improvements in pain and stiffness scores with collagen supplements. But the trials are small, many are judged at high risk of bias, and the improvements are often below what patients would notice. Major guidelines do not recommend collagen: the Australian RACGP guideline says it cannot recommend for or against it, Arthritis UK says the NHS does not recommend it, and the American College of Rheumatology (ACR) guideline summary does not mention it. The ACR guideline puts exercise first, and for people with overweight, even modest weight loss can ease knee pain. This guide sets out what the trials found, how big the effects are, who paid for the research, and what to do instead or alongside.

    Collagen for Knee Osteoarthritis: The Short Answer

    1. 1Meta-analyses report small improvements in pain and stiffness, but most trials are small and at high risk of bias.
    2. 2All five trials in a widely cited 2019 meta-analysis were funded by, or had authors paid by, collagen companies.
    3. 3Two different products are studied: collagen peptides at gram doses and undenatured type II collagen (UC-II) at 40 mg a day.
    4. 4No major guideline recommends collagen for OA; the RACGP is neutral and the NHS does not recommend it.
    5. 5There is no good evidence that collagen can rebuild cartilage that is already damaged.
    6. 6Exercise is a first-line treatment in the ACR guideline, and weight loss can ease knee pain for people with overweight.

    Does Collagen Help Knee Osteoarthritis?

    It might help a little, but the evidence is not strong enough to say for sure. The best-known meta-analysis, by García-Coronado and colleagues in International Orthopaedics in 2019, pooled five randomized trials. It found that collagen improved the total WOMAC score (a 0 to 100 questionnaire of pain, stiffness and physical function, where lower is better) by 8 points, and pain on a visual analogue scale (VAS, a 0 to 100 line where people mark their pain) by about 17 points. The authors concluded that collagen is effective in improving OA symptoms.

    Looked at more closely, the picture is less convincing. In the same analysis, the WOMAC pain and function subscores did not improve significantly; only stiffness and the total score did. Later reviews have found similar small benefits with wide variation between trials. For a broader overview of joint research, see our guide to collagen peptides for joints and exercise.

    What Is the Difference Between Collagen Peptides and UC-II?

    Two quite different products are both called "collagen" in arthritis research, and they should not be confused. Collagen peptides (also called hydrolysed collagen) are collagen broken down into small fragments, usually from cow, pig or fish. Trials use them at doses of several grams a day; one 2025 trial used 3,000 mg a day of low-molecular-weight fish collagen peptides. The idea is that the fragments supply building blocks and may signal cartilage cells. You can read about fragment size in collagen molecular weight.

    Undenatured type II collagen (UC-II) is made from chicken cartilage without breaking down the collagen's structure. It is used at a tiny dose, 40 mg a day in studies, which contains about 1.2 mg of undenatured type II collagen. It is thought to work through "oral tolerance", a process in which small amounts of a protein swallowed by mouth may calm the immune response against the same protein in the joints. Our guide to undenatured type II collagen explains this in more detail. The same oral tolerance idea was tested in rheumatoid arthritis with disappointing results, as covered in collagen for rheumatoid arthritis.

    Collagen peptides vs UC-II in knee OA research

    Feature
    Collagen peptides (hydrolysed)
    Undenatured type II collagen (UC-II)

    Dose used in studies

    Gram doses, e.g. 3,000 mg a day (Park 2025)

    40 mg a day (about 1.2 mg undenatured type II collagen)

    Proposed mechanism

    Supplies collagen fragments that may signal cartilage cells

    Oral tolerance: may calm immune activity in the joint

    Key trial

    Park 2025: 80 randomized, 180 days, early knee OA

    Lugo 2016: 191 randomized, 180 days, vs glucosamine-chondroitin and placebo

    Key trial result

    Better WOMAC pain and total scores; no change in VAS, joint space or CRP

    Lower total WOMAC than placebo and glucosamine-chondroitin at day 180

    Key trial funding

    NEWTREE Co. (product maker); one author an employee

    InterHealth Nutraceuticals; two authors employees

    These are doses used in studies, not dosing recommendations. Neither product is an approved drug for osteoarthritis.

    What Do the Meta-Analyses Actually Find?

    Several reviews have pooled the trials, and they agree that collagen shows small average benefits. They also agree that the evidence is low quality. The main ones are summarized below.

    Main reviews of collagen for osteoarthritis

    Review
    Collagen type
    Trials and participants
    Main result
    Caveats

    García-Coronado 2019, Int Orthop

    Mixed

    5 RCTs (total not reported in abstract)

    Total WOMAC -8.0; VAS -16.6; WOMAC pain and function not significant

    All 5 trials had industry ties; only 1 preregistered

    Lin 2023, J Orthop Surg Res

    Collagen peptides

    4 RCTs, 507 patients

    VAS pain SMD -0.58 (moderate quality)

    All 4 trials at high risk of bias; I² 68%

    Kumar 2023, Am J Transl Res

    UC-II only

    8 RCTs, 243 on UC-II

    WOMAC -8.9; VAS -1.65; Lequesne not significant

    Small numbers; 3 to 6 months follow-up

    Simental-Mendía 2025, Clin Exp Rheumatol

    Mixed

    11 RCTs, 870 participants

    Pain -13.6; function -6.5

    Very high heterogeneity (I² 75% to 88%); scales not named in abstract

    Ravindran 2026, Aesthet Surg J Open Forum (umbrella review)

    Mixed

    16 reviews, 113 RCTs, about 8,000 people

    Longer use linked to less OA pain and stiffness

    Reported via press release; no certainty rating or funding given there

    Heterogeneity, shown as I², measures how much the trials disagree with each other; values above about 75% mean the results vary a lot, which makes a single pooled number less trustworthy. A 2025 review by Gupta and Maffulli in Annals of Medicine looked at 12 studies without pooling them and concluded that larger, longer controlled trials are needed before collagen can be used routinely.

    How Big Is the Effect, and Is It Clinically Meaningful?

    Probably small. An 8-point drop on a 100-point WOMAC scale sounds useful, but it is an average that is pulled up by the smallest trials. In a critique published by STAT News, statistician von Hippel noted that the largest trials in the 2019 meta-analysis, with 54 to 111 people taking collagen, showed improvements of less than 5 points on a 0 to 100 scale. Smaller trials tend to report bigger effects, a pattern often linked to publication bias, where positive studies are more likely to be published than negative ones.

    Individual trials show the same mixed picture. In the 2025 Park trial in Frontiers in Nutrition, people with early knee OA (Kellgren-Lawrence grade I to II on X-ray) who took fish collagen peptides improved more on WOMAC pain and total scores than those on placebo. But there was no difference in VAS pain, in the joint space seen on X-ray, or in CRP, a blood marker of inflammation. The results were also based on the 60 people who completed the study as planned, not all 80 who were randomized.

    Who Funded the Collagen Trials, and Why Does It Matter?

    Funding matters because company-funded trials are more likely to report results that favour the product. According to von Hippel's analysis, the 2019 meta-analysis started with 114 articles and kept five randomized trials. All five were funded by a collagen company or had authors paid by one, and only one was preregistered, meaning its methods and outcomes were declared in advance. The trial with the largest effect had four of its five authors employed by Nitta Gelatin, a collagen maker.

    The pattern continues in newer trials. The Lugo 2016 UC-II trial was funded by InterHealth Nutraceuticals, and two authors were its employees. The Park 2025 trial was funded by NEWTREE Co., which made the product, and one author was an employee. Industry funding does not prove a result is wrong, but it is a reason to wait for independent trials. The ACR guideline already applies this logic to a similar supplement: it recommends against glucosamine, alone or with chondroitin, because it does not improve knee and hip OA in studies without industry funding.

    What Do Arthritis Guidelines Say About Collagen?

    No major guideline recommends collagen for knee osteoarthritis, as of November 2026.

    1. 1RACGP (Australia, 2018): "We are unable to recommend either for or against the use of collagen for people with knee and/or hip OA." This is a conditional statement based on low-quality evidence for the knee and very low-quality evidence for the hip.
    2. 2ACR and Arthritis Foundation (US, 2019): the AAFP summary of the guideline does not mention collagen. It recommends against glucosamine and chondroitin.
    3. 3Arthritis UK (Versus Arthritis, reviewed September 2026): there is not enough evidence to recommend collagen as a routine arthritis treatment, and the NHS does not recommend it.
    4. 4Arthritis Foundation consumer article: describes most findings as positive and mentions about 40 mg a day of UC-II, but it does not discuss industry funding or the ACR guideline.

    Can Collagen Rebuild Knee Cartilage?

    No good evidence shows that it can. Arthritis UK states that there is currently no good evidence that collagen supplements can rebuild cartilage already damaged by arthritis. In the Park 2025 trial, the one measure of joint structure, joint space width on X-ray, did not differ between collagen and placebo after 180 days. The benefits reported in trials are changes in symptom questionnaires, not changes in the joint itself. Claims that collagen "regrows" cartilage go beyond what the research shows.

    How Long Did Trial Participants Take Collagen?

    Most trials lasted three to six months. The UC-II trials in the Kumar review followed people for 3 to 6 months, and both the Lugo and Park trials ran for 180 days. The 2026 umbrella review found that longer supplementation was associated with less pain and stiffness. That means anyone trying collagen would need several months to judge it, and a few weeks without change does not tell you much either way. More on typical timelines is in how long collagen peptides take to work.

    Is Collagen Safe to Take With Arthritis Medicines?

    Collagen is generally well tolerated, but the safety data from OA trials are thin. In the Lin 2023 meta-analysis, adverse events were somewhat more common with collagen than placebo (odds ratio 1.66), but the result was not statistically significant and the evidence was rated very low quality. The reviews in this article do not report specific interactions with arthritis medicines. Because supplements can matter for people on several drugs, tell your doctor or pharmacist about any collagen product, especially combination products that add other ingredients. Common complaints, such as stomach upset, are covered in collagen peptides side effects.

    What Helps Knee Osteoarthritis Most?

    According to the AAFP summary of the ACR guideline, exercise, tai chi and self-management programs are first-line treatments for osteoarthritis. The same summary notes that OA pain is related to weight and that even a 5% weight loss provides notable improvement in knee and hip pain. These steps cost little and benefit overall health. Collagen, if you choose to try it, is at best an add-on to these steps, not a replacement.

    Knee OA self-care: what the evidence supports

    Approach
    Guideline stance
    What it involves

    Exercise

    First-line treatment (ACR, per AAFP summary)

    Regular walking, strengthening and other activity, ideally with a physical therapist's plan

    Weight loss (if overweight)

    Even 5% loss notably improves knee pain (ACR, per AAFP summary)

    Gradual weight loss through diet and activity

    Glucosamine and chondroitin

    Recommended against (ACR)

    No benefit in studies without industry funding

    Collagen peptides or UC-II

    Not mentioned (ACR); neutral (RACGP); not recommended (NHS)

    Low-quality, mostly industry-funded evidence of small symptom benefits

    Strengthening the muscles around the knee also supports tendons and ligaments; see collagen for tendons and ligaments. If weight is part of the picture, our article on whether collagen helps with weight loss explains why collagen is not a weight-loss tool on its own.

    When to Talk to Your Doctor

    1. 1Knee pain that is new, severe or getting worse
    2. 2A hot, red or very swollen joint, or fever with joint pain
    3. 3Morning stiffness lasting more than an hour or pain in many joints, which may point to inflammatory arthritis
    4. 4Pain that stops you walking, sleeping or working
    5. 5Before adding supplements if you take blood thinners or several medicines, or have kidney disease

    Should You Try Collagen for Knee Pain? What to Do Next

    If you decide to try collagen, treat it as a personal experiment and keep the core treatments in place.

    1. 1Start with the basics: an exercise plan and, if relevant, weight loss.
    2. 2Note your pain and stiffness on a 0 to 10 scale before starting, then monthly.
    3. 3Give it about three to six months, the length of most trials, then stop if you notice no change.
    4. 4Know which product type you are taking: gram-dose peptides and 40 mg UC-II are different.
    5. 5Ask your doctor: Which exercises suit my knee? Would physical therapy help? Is my weight affecting my knee? Is there any reason not to take this supplement with my medicines?

    The Bottom Line

    Collagen for knee osteoarthritis has some supportive trials, but the benefits are small, the trials are small and mostly industry-funded, and no major guideline recommends it. It is unlikely to rebuild cartilage. If you want to try it, give it a few months and judge it honestly, while keeping exercise and weight management at the centre of your knee care.

    FAQ

    There is no recommended dose. Studies used gram doses of collagen peptides, such as 3,000 mg a day, or 40 mg a day of UC-II. These are trial doses, not medical advice.

    One trial, funded by the UC-II maker, found UC-II lowered WOMAC scores more than glucosamine with chondroitin over 180 days. Independent trials are lacking.

    There is no good evidence that it can. Arthritis UK says there is no good evidence that collagen supplements can rebuild cartilage damaged by arthritis.

    Trials lasted three to six months, and longer use was linked to more benefit. Few people should expect change within weeks.

    Its consumer article describes the research as mostly positive, but the ACR and Arthritis Foundation guideline summary does not mention collagen.

    They have not been compared head to head. Peptides are used at gram doses; UC-II at 40 mg a day. Evidence for both is low quality.

    Sources

    This article summarizes published research for general information and is not medical advice. Most collagen trials for osteoarthritis were funded by collagen manufacturers.

    Dimitrun Irene

    Author

    Published November 19, 2026
    Share:TwitterFacebook

    Comments (0)

    Be the first to comment on this article.

    More articles

    How Long Do Collagen Peptides Take to Work