Collagen for joints: does it help and what does the research say?

Yes, there is evidence of collagen's benefits for joints, but it's important to understand it correctly. Modern meta-analyses of randomized trials show that in people with osteoarthritis, primarily of the knee, collagen supplements can, on average, slightly reduce pain and improve function. This isn't just the effect of individual small studies: a meta-analysis of 35 randomized trials with 3,165 participants found a small to moderate reduction in pain and improvement in function compared to controls. [1]

However, this does not mean that collagen "restores joints," regrows lost cartilage, or stops osteoarthritis. Most clinical studies have assessed pain, stiffness, and mobility, rather than conclusively proven preservation or restoration of articular cartilage. Therefore, collagen is more appropriately viewed as an optional treatment adjunct that may alleviate symptoms in some people, rather than as a chondroprotector with proven structural action. [2]

Moreover, the word "collagen" covers different products. The best-studied are hydrolyzed collagen peptides, which are typically taken in grams, and undenatured type II collagen, which is studied in significantly smaller doses. Comparing them solely by the milligram number on the packaging is not possible: they are different forms with different proposed mechanisms of action. [3]

What collagen research shows in osteoarthritis

The most comprehensive meta-analysis from 2024 included 35 randomized trials with 3,165 patients. Compared with control groups, collagen derivatives reduced pain with a standardized effect size of -0.35 and improved function with an effect size of -0.31. The authors assessed the certainty of the evidence as moderate for pain and high for function. In other words, there is a moderate effect, but its magnitude is closer to small than to radical symptom relief. [4]

Another updated meta-analysis, published in 2025, pooled 11 randomized trials with 870 participants specifically with knee osteoarthritis. In the collagen group, pain and functional impairment were reduced more significantly than with placebo. However, the range of results between studies was very wide: statistical heterogeneity reached 88% for pain and 75% for function. This is an important detail, as it demonstrates that the results of individual drugs and individual patient groups vary significantly. [5]

In 2026, an even broader umbrella review was published, combining data from 16 systematic reviews, 113 randomized trials, and 7983 participants on the various effects of collagen. For osteoarthritis, a favorable signal was again found: overall pain severity, the total osteoarthritis symptom score, and stiffness were reduced. However, individual components of some functional scales were not improved in all analyses. The authors disclosed no commercial funding for their review and declared no conflicts of interest. [6]

The new data, however, is not all positive. In an independent, double-blind study from 2025, 68 people with knee osteoarthritis received a combination of undenatured type II collagen, hydrolyzed collagen, or placebo for 12 weeks. Pain and function improved in both groups, but no significant advantage of collagen over placebo was found. The authors declared no conflicts of interest. This clearly illustrates why one cannot make promises about the effect of a specific individual based on a positive average result from a meta-analysis. [7]

Data layer	What was discovered	How to interpret this
Meta-analysis of 35 trials	Mild to moderate reduction in pain and improvement in function	The effect exists on average, but is usually not strong.
Meta-analysis of 11 trials in knee osteoarthritis	Improved pain and function	Results varied widely between studies.
Umbrella Review 2026	General favorable signal for osteoarthritis symptoms	The overall data has become more compelling, but heterogeneity among drugs remains
A separate independent trial in 2025	No benefit over placebo was found.	Collagen doesn't work for everyone, and different supplements aren't necessarily equivalent.

Sources: meta-analyses and a randomized trial from 2024–2026. [8]

What kind of collagen was studied for joints?

Hydrolyzed collagen

During hydrolysis, long protein molecules are first broken down into shorter peptides. After ingesting such a product, amino acids and small peptides containing hydroxyproline actually enter the bloodstream. In a 2024 randomized bioavailability study, this occurred after ingesting hydrolysates of fish, pork, and beef origin. No significant advantage was found for one source in terms of free hydroxyproline absorption. [9]

This last fact is particularly important when choosing a supplement. The advertising claim that "marine collagen is absorbed much better than beef" cannot be considered a proven clinical benefit for joints. A modern study demonstrated the formation of circulating metabolites after all studied hydrolysates, regardless of source and molecular weight; however, the authors emphasize that it is not yet known what concentration of these substances is required for specific clinical effects in tissues. [10]

Osteoarthritis studies have often used gram-scale amounts of hydrolyzed peptides. For example, a randomized 2024-2025 study in which patients with knee osteoarthritis received 10 grams of collagen peptides daily for six months found better pain and function scores than those with a placebo. Another 2025 study also examined 10 grams of hydrolyzed peptides daily. These results are specific to the study medications and do not mean that any powder labeled "collagen" will produce the same effect. [11]

Undenatured collagen type II

Undenatured type II collagen functions differently. Its purpose is not to provide the body with large quantities of amino acids for building new cartilage. It is believed that the preserved three-dimensional structure of the protein interacts with the intestinal immune tissue and promotes the development of so-called oral tolerance—a modified immune response to type II collagen. A biological explanation for this mechanism exists, although its clinical significance in osteoarthritis remains a subject of research. [12]

This is why the doses of undenatured type II collagen being studied are significantly lower. A well-known 180-day randomized trial of 191 people used 40 mg per day; this group showed better results in the overall symptom score compared to placebo. A review of studies published in 2025 also considered 40 mg per day as the most studied option, but clearly noted the need for larger and longer studies before such collagen can be recommended for routine use. [13]

Form	Typical quantities in studies	main idea	What is known
Hydrolyzed collagen peptides	Often several grams, in a number of modern trials 10 g/day	Amino acids and small peptides after digestion	There is evidence of a reduction in the symptoms of osteoarthritis
Undenatured collagen type II	In a number of studies 40 mg/day	Presumed immune oral tolerance	There are positive trials, but the base is smaller
Gelatin	There is no standardized treatment regimen.	Partially digested collagen	It is not possible to automatically transfer data from collagen preparations to regular food gelatin.
"Sea", "beef", "pork"	Depends on the specific hydrolyzate	Source of raw materials	Origin alone does not prove greater effectiveness

Sources: Clinical studies and bioavailability studies.[14]

Does collagen restore articular cartilage?

There's currently no convincing evidence that dietary collagen can restore damaged articular cartilage. This is one of the main barriers between what research shows and what advertising often promises.

There is preliminary structural data. For example, a small pilot randomized study of just 30 patients found changes in some cartilage magnetic resonance imaging parameters after using collagen hydrolysate. However, a pilot study with such a small number of participants does not prove that the drug is capable of restoring lost cartilage or altering the natural course of osteoarthritis. [15]

The mere fact that walking up and down stairs becomes less painful after a few months doesn't mean that cartilage thickness has increased. Pain in osteoarthritis depends not only on the cartilage, which contains virtually no pain receptors, but also on the subchondral bone, synovial membrane, ligaments, muscles, and other joint structures. Therefore, symptomatic improvement and structural restoration of the disease are different endpoints.

This is where a cautious formulation is particularly helpful: collagen may alleviate symptoms in some patients, but it cannot yet be considered a proven means of repairing cartilage or slowing the progression of osteoarthritis. Even a 2025 review that favorably evaluates undenatured type II collagen emphasizes the need for larger and longer trials before its routine clinical use. [16]

Why do the results of different studies differ?

The general term "collagen" encompasses products with vastly different compositions: hydrolysates of varying molecular weights, peptides from various sources, native type II collagen, and multicomponent mixtures. Additionally, dosages, treatment durations, osteoarthritis severity, and pain assessment criteria vary. Combining such diverse interventions inevitably creates statistical heterogeneity, as clearly demonstrated in a 2025 meta-analysis. [17]

Another issue is the scale of the studies. Many trials are significantly smaller than typical drug trials and last only a few months, whereas osteoarthritis develops over years. Therefore, the short-term symptomatic effect has been relatively well studied, but the long-term impact on pain medication requirements, disease progression, joint replacement surgery, and quality of life over several years is much less well understood. A 2026 umbrella review explicitly cites the lack of long-term clinical outcomes as one of the limitations of the collagen evidence base. [18]

Finally, some studies of specific collagen products are affiliated with their manufacturers. This doesn't automatically mean the results are invalid, but it does increase the value of independent replication. It's telling that an independent 2025 study, whose authors declared no competing interests, found no advantage of the combination of the two forms of collagen over placebo. [19]

Is collagen included in the treatment recommendations for osteoarthritis?

Despite positive meta-analyses, collagen is not yet considered a primary treatment for osteoarthritis in major clinical guidelines. This is not the same as concluding that "collagen is proven useless." Rather, the accumulated research has not yet led professional societies to rank it alongside exercise, weight management, and other standard treatments.

For example, the American Academy of Orthopedic Surgeons' guidelines for knee osteoarthritis consider dietary supplements, but among the listed supplements are turmeric, ginger extract, glucosamine, chondroitin, and vitamin D—collagen is not included. Moreover, even for the supplements listed, the evidence is considered limited and inconsistent. [20]

The American College of Rheumatology and the Arthritis Foundation also do not recommend collagen as a standard treatment. The document also thoroughly evaluates other supplements and, for example, advises against the use of glucosamine for osteoarthritis of the knee, hip, and hand. [21]

European guidelines for basic non-drug management of hip and knee osteoarthritis emphasize individualized interventions, patient education, exercise, and maintaining a healthy body weight or losing weight if overweight. These approaches remain the foundation of treatment, regardless of whether a person decides to additionally try collagen supplements. [22]

What is more important than collagen in joint arthrosis?

If a person has confirmed osteoarthritis, collagen should not replace methods with a more robust evidence base. Regularly selected physical activity and exercise improve pain and function, and in overweight individuals, weight loss reduces mechanical stress on the joint. These interventions form the core of current recommendations. [23]

For knee osteoarthritis, the American Academy of Orthopaedic Surgeons strongly recommends exercise. The same guidelines also provide significantly stronger evidence for topical nonsteroidal anti-inflammatory drugs than dietary supplements. This clearly demonstrates the role of collagen: as a possible supplement, not a mainstay of therapy. [24]

Therefore, the situation "I take collagen instead of exercise, weight loss, and prescribed treatment" is much less medically sound than the situation "the main treatment is carried out, and collagen is used as an additional attempt to reduce residual pain."

Is collagen worth trying for joints?

In cases of confirmed knee osteoarthritis, such a trial approach can be considered reasonable if the patient understands the limitations: moderate relief is possible, there is no guaranteed effect, and restoration of damaged cartilage is not to be expected. A body of modern meta-analyses makes this option more reasonable than the assertion that collagen has no effect at all. [25]

In this case, it makes sense to evaluate the specific product, not the word "collagen" on the label. If hydrolyzed peptides are chosen, the composition and daily dosage should at least roughly correspond to the forms that have been clinically studied. If undenatured type II collagen is being considered, gram doses of hydrolysate cannot be used as a guide: studies of this form used tens of milligrams. [26]

The effect should also not be assessed after just a few days. In studies, treatment duration ranged from several weeks to several months; positive trials of undenatured type II collagen lasted three to six months, while modern studies of hydrolyzed peptides lasted from eight weeks to six months. Therefore, collagen is fundamentally different from a fast-acting pain reliever. [27]

It's more practical to determine in advance what exactly constitutes success: for example, whether it's easier to walk, climb stairs, or perform a regular workout, and whether pain has subsided. If there's no noticeable change after an adequate period of use, there's no scientific basis for continuing the supplement indefinitely or constantly increasing the dosage.

Is collagen safe?

Based on available short- and medium-term studies, collagen is generally well-tolerated. In a meta-analysis of 35 randomized trials, the rate of adverse events and treatment discontinuations was no higher than in control groups. This is one reason why a potentially small symptomatic effect may be acceptable to someone wishing to try such a supplement. [28]

It is important, however, to consider the origin of the raw materials. Hydrolysates can be produced from fish, cattle, or pork, while many undenatured type II collagen products are made from chicken cartilage. In the case of food allergies, the source of the product is of direct importance. Current consumer guidelines also recommend checking the full composition of the supplement, not just the collagen content. [29]

The quality of supplements also cannot be considered uniform. The American Academy of Orthopedic Surgeons specifically notes that requirements for dietary supplements differ from those for prescription drugs, and product variability is possible between manufacturers. For international consumers, the practical conclusion is the same: transparent ingredients, source information, and independent quality control, if available, are preferable. [30]

What is often misunderstood

"Cartilage is made of type II collagen, so it's essential to take type II collagen." This sounds logical biologically, but tissue composition alone doesn't prove the effectiveness of a supplement. Hydrolyzed peptides and undenatured type II collagen are being studied as different interventions, and it's currently impossible to confidently say one form is universally best for all patients. [31]

"Marine collagen is better than beef for joints." There is no convincing clinical evidence to support this blanket statement. In a 2024 bioavailability study, hydrolysates of fish, pork, and beef origin delivered characteristic collagen metabolites into the bloodstream, and differences in source prevented the establishment of a universally superior option. [32]

"Collagen goes straight to the knee." This is an oversimplification. After consuming hydrolysate, amino acids and small peptides enter the bloodstream. Detection of these substances in the bloodstream confirms absorption, but it does not mean that the entire dose is sent directly to the joint and converted into new cartilage. [33]

"If the pain goes away, the cartilage is restored." No. Most positive studies show primarily a symptomatic result. To claim cartilage restoration, convincing long-term studies with structural endpoints would be needed, which are currently insufficient. [34]

How to make a decision in practice

If joints are healthy and collagen is only supposed to be taken "for wear and tear prevention," the evidence is significantly weaker than for symptomatic osteoarthritis. The presence of collagen in normal cartilage does not necessarily mean that supplementation prevents the development of osteoarthritis.

If knee osteoarthritis has been diagnosed and basic measures are already being taken, collagen can be considered as an additional trial with realistic expectations. It's best to choose one well-researched form, avoid switching to multiple supplements at once, and monitor specific clinical results. This will help you understand whether the supplement is truly beneficial for your specific individual.

If joint pain has recently developed, the joint is noticeably swollen, red, or hot, the pain occurred after an injury, or you are unable to bear weight properly, starting with collagen is incorrect. First, it's important to determine the cause of the symptoms: osteoarthritis is only one of many sources of joint pain.

Key points from experts

Mario Alberto Simental-Mendia, PhD, holds a PhD in molecular biology and genetic engineering and is a researcher at the Faculty of Medicine of the Autonomous University of Nuevo León. His research interests include critical evaluation of treatments for musculoskeletal disorders. A 2025 meta-analysis published by his group of 11 randomized trials demonstrated improvements in pain and function in knee osteoarthritis with collagen, although the results of individual trials varied significantly. [35]

Nicola Maffulli, Professor of Orthopaedics and Traumatology, is an orthopaedic surgeon specializing in musculoskeletal disorders. A review of undenatured type II collagen prepared by his research group found the current data promising, but the authors also emphasized the need for larger, multicenter, and long-term studies before routine clinical use is justified. This position reflects the current balance of evidence: there is a signal of efficacy, but scientific uncertainty remains. [36]

Frequently Asked Questions

Which collagen is better for joints - type I, II or III?

It's not possible to simply select a type by number. Both hydrolyzed collagen peptides containing various types of collagen and undenatured type II collagen are being studied for osteoarthritis. There is no convincing evidence yet of the universal superiority of one variant. [37]

Is type II collagen more effective than hydrolyzed?

The current evidence base does not support this conclusion. Studies of both forms yield positive results, but direct qualitative comparisons are lacking, and the mechanisms and dosages differ significantly.

How much collagen should I take for joints?

There is no single approved therapeutic dose. In modern studies of hydrolyzed peptides, approximately 10 g per day has often been used, while undenatured type II collagen has been studied, in particular, at a dose of 40 mg per day. These values cannot be transferred from one form to the other. [38]

How long does it take for collagen to start working?

Studies have assessed effects over weeks and months, not just days. Therefore, if improvement does occur, it is usually seen as a gradual result. [39]

Is it possible to take collagen on a regular basis?

The long-term safety of long-term, continuous use has been studied significantly less than that of a few months of use. Most clinical studies do not allow for a firm conclusion about the benefits of continuous, lifelong use. [40]

Should I take vitamin C at the same time?

Ascorbic acid is essential for the normal synthesis of collagen, but this doesn't mean everyone needs a separate vitamin C supplement alongside collagen. Many commercial supplements include it, but the effectiveness of collagen in osteoarthritis cannot be explained solely by the presence of vitamin C.

Is marine collagen better absorbed?

There is no convincing general evidence for this. A modern bioavailability study showed the release of characteristic metabolites into the blood after fish, pork, and beef hydrolysates. [41]

Does collagen help with rheumatoid arthritis?

Findings from osteoarthritis cannot be automatically applied to rheumatoid arthritis. This is a different disease with a different mechanism, requiring specific treatment targeting the immune-inflammatory process. Collagen should not replace standard therapy for rheumatoid arthritis.

Can collagen be replaced with gelatin or bone broth?

These are sources of collagen proteins, but their composition and quantity are not as standardized as in clinical trials of specific collagen preparations. Therefore, it cannot be claimed that a bowl of broth or regular gelatin will provide the same clinical effect.

Should a healthy person take collagen to prevent arthrosis?

There is insufficient convincing evidence that collagen supplementation prevents future osteoarthritis in healthy individuals. The primary clinical evidence base concerns individuals who already have symptoms or have been diagnosed with osteoarthritis.

Result

Collagen for joints is no longer considered a supplement for which there is "no evidence at all": several modern meta-analyses do show a small to moderate reduction in pain and improvement in function in osteoarthritis. Hydrolyzed collagen peptides and undenatured type II collagen have been the most studied. [42]

But the evidence is limited. Collagen has not yet been proven to restore damaged cartilage, prevent endoprosthetics, or halt the progression of osteoarthritis. It does not replace exercise, weight loss if you're overweight, or other methods with stronger clinical support. Therefore, the most accurate formulation to date is: collagen can be tried for osteoarthritis; the expectation is likely to be moderate symptom relief, not "joint regeneration." [43]