Collagen supplements: what the trials actually show

By the DieticiansWorldwide.com editorial team · 17 October 2026 · Supplements & industry

What collagen supplements are

Collagen is the main structural protein in skin, tendons, ligaments, cartilage and bone. Your body makes it from amino acids, especially glycine, proline and hydroxyproline, and needs vitamin C to do so. The US National Institutes of Health states that "vitamin C is required for the biosynthesis of collagen" (NIH Office of Dietary Supplements, vitamin C).

Supplements are made from the skin, bones or scales of cattle, pigs, chicken or fish. Most are hydrolysed collagen, also sold as collagen peptides, in which the protein has been broken into short chains that dissolve easily in drinks. A smaller category, undenatured type II collagen (often labelled UC-II), is taken in milligram doses and is not intended as a source of protein.

The core debate is simple. Collagen is digested like any other protein, so sceptics ask why swallowing it should build collagen in your face or knees any more than eating a steak would. Supporters point to studies showing that collagen-derived peptides can be detected in human blood about an hour after a dose and may reach the skin, as summarised in a 2023 review (Pu and colleagues, Nutrients, 2023). The trials are the only way to settle it, so this article goes through them by outcome.

Do collagen supplements actually work? The evidence by outcome

Collagen for skin

This is the most heavily marketed use, and the most contested.

The 2023 meta-analysis of 26 randomised controlled trials (RCTs) with 1,721 participants concluded that hydrolysed collagen improved skin hydration and elasticity, but it also "identified several biases in the included RCTs" and called for larger trials. The included studies lasted only 2 to 12 weeks.

A 2025 meta-analysis took a different approach: it split the trials by who paid for them and by quality. Across all 23 RCTs (1,474 participants), collagen improved hydration, elasticity and wrinkles. But "studies not receiving funding from pharmaceutical companies revealed no effect", and high-quality studies "revealed no significant effect in all categories". The authors concluded that "there is currently no clinical evidence to support the use of collagen supplements to prevent or treat skin aging" (Myung and Park, American Journal of Medicine, 2025).

That is a striking result, and it matches a pattern seen in other supplement research: small, short, sponsor-funded trials tend to show benefits that larger independent trials do not. The fair reading is that collagen for skin has not been shown to work once bias is accounted for. It is not proof that it never works, but it is a poor bet for the money.

Collagen for joints and osteoarthritis

The joint evidence is stronger. A 2024 meta-analysis in the journal Osteoarthritis and Cartilage pooled 35 RCTs with 3,165 patients and found that collagen derivatives produced "small-to-moderate effects on pain alleviation" (standardised mean difference −0.35, moderate certainty) and on function (−0.31, high certainty), without more side effects or withdrawals than placebo (Liang and colleagues, Osteoarthritis and Cartilage, 2024). A separate 2025 meta-analysis limited to knee osteoarthritis, with 11 RCTs and 870 participants, also found improvements in pain and function, but with high heterogeneity between studies, meaning the results varied a lot from trial to trial (Simental-Mendía and colleagues, Clinical and Experimental Rheumatology, 2025).

A standardised mean difference of about 0.35 is modest. In plain terms, some people with osteoarthritis may notice less pain, but collagen is not a replacement for exercise, weight management where relevant, and the treatments a GP or physiotherapist recommends. The evidence also mixes different products, including hydrolysed collagen and undenatured type II collagen, so a positive pooled result does not guarantee that the product on your shelf works.

For healthy, active people without arthritis, the picture is weaker. In 2011 the European Food Safety Authority assessed a claim that collagen hydrolysate helps maintain joint health in physically active people. The key trial, in 147 student athletes, found "no significant differences between groups for any endpoint" after correcting for multiple comparisons, and EFSA concluded that a cause and effect relationship "has not been established" (EFSA NDA Panel, 2011). The European Commission then formally refused the claim in Regulation (EU) No 379/2012 of 3 May 2012, which lists collagen hydrolysate and joint health among rejected health claims.

Collagen for bones

The most cited bone study gave 131 postmenopausal women with low bone density either 5 g of specific collagen peptides or a placebo every day for 12 months. Bone mineral density in the spine and the femoral neck (hip) rose slightly in the collagen group compared with placebo, and a marker of bone formation increased (König and colleagues, Nutrients, 2018). The changes in T-score were small (+0.1 at the spine versus −0.03 with placebo), and 29 of the 131 women did not complete the study.

A single trial is not enough to recommend collagen for osteoporosis. It is a reason for more research, not a substitute for calcium, vitamin D where needed, weight-bearing exercise and prescribed bone treatments.

Collagen for muscle

Collagen is sometimes sold for muscle building. In a 2015 trial, 53 men with sarcopenia (age-related muscle loss), average age 72, did 12 weeks of resistance training while taking 15 g a day of collagen peptides or a placebo. Both groups improved, but the collagen group gained more fat-free mass (4.2 kg versus 2.9 kg) and more quadriceps strength (Zdzieblik and colleagues, British Journal of Nutrition, 2015).

The catch is that the placebo was silica, not another protein. The trial cannot tell you whether collagen beats the same amount of whey, milk or soya protein. That matters, because collagen is a poor-quality protein on standard scoring.

Collagen as a protein source

By the standard PDCAAS scoring method, collagen "lacks one indispensable amino acid (tryptophan) and is therefore categorized as an incomplete protein source" (Paul and colleagues, Nutrients, 2019). The same study calculated that collagen peptides could make up as much as 36% of protein in a typical Western diet while still meeting essential amino acid needs, and noted that the doses used in research, 2.5 to 15 g a day, are well below that ceiling.

The practical message: a scoop of collagen counts towards your total protein, but it should not replace complete proteins such as dairy, eggs, fish, meat, soya or a mix of pulses and grains. If you are adding collagen specifically for protein, check your overall needs first with our protein calculator; our guide to how much protein you need explains the targets for different ages.

How long does collagen take to work?

The trials give a range rather than a single answer. Skin trials in the 2023 meta-analysis ran for 2 to 12 weeks, and longer use (more than 8 weeks) showed a larger pooled effect than shorter use (Pu and colleagues, 2023). Osteoarthritis trials typically measure pain over several months, and the bone trial ran for 12 months.

If you decide to try collagen for joint pain, a reasonable approach is to set a fixed trial period, such as three months, note your symptoms before and after, and stop if you notice no difference. Without that, it is easy to keep paying for something that is not helping.

Effective doses

Doses differ by type:

• Hydrolysed collagen or collagen peptides: the research literature uses roughly 2.5 to 15 g a day (Paul and colleagues, 2019). The bone trial used 5 g; the muscle trial used 15 g.
• Undenatured type II collagen: used in much smaller amounts, in milligrams, because it is not intended as a protein source.

More is not obviously better. Products with 20 g or more per serving are mainly delivering extra protein, which you could get from food.

Is marine collagen better than bovine?

There is no good evidence that it is. Marketing often claims marine (fish) collagen is "more bioavailable", but head-to-head trials are lacking. In the 2023 skin meta-analysis, 14 of the 26 trials used fish collagen, one used bovine, one chicken, two porcine and nine did not say. The authors found "no significant differences in the effects of different sources" on skin elasticity (Pu and colleagues, 2023). With only one bovine trial, a fair comparison was not really possible.

In practice the choice comes down to personal factors:

• Diet and belief: marine collagen suits pescatarians; people avoiding pork or beef for religious or cultural reasons should check the source.
• Allergies: marine collagen comes from fish. If you have a fish allergy, do not take it without advice from your doctor.
• Vegans: collagen is always animal-derived. Products sold as "vegan collagen" usually contain amino acids or vitamin C intended to support your own production; they are not collagen.
• Price: compare cost per gram of collagen; there is no proven extra benefit from paying more for a particular source.

Third-party testing: what labels do and do not guarantee

Because supplements are not approved before sale, independent testing is the main way to know a product contains what it says. In the US, the USP Verified Mark means a product "contains the ingredients listed on the label in the declared strength and amounts", does not contain harmful levels of specified contaminants "such as heavy metals and pesticides", follows FDA good manufacturing practices and will break down properly to release its ingredients (USP, Quality Supplements FAQs). USP also notes that "tests have shown that contents of some supplements do not match the label".

What a quality seal does not tell you is whether the product works. A collagen powder can be perfectly pure and still do nothing for your skin. Read seals as evidence of quality, and read the trials as evidence of benefit. Our guide to how supplements are regulated and sold across countries explains the rules in more detail, and the United States country guide covers the US regulatory system.

Other label points worth checking:

• Source animal (bovine, porcine, chicken, fish) and any allergen statement.
• Collagen per serving in grams, not just "proprietary blend".
• Added ingredients: many powders add sugar, sweeteners, caffeine, vitamins or "beauty blends", so check the full ingredient list and tell your doctor what you take.
• Claims: be wary of anti-ageing or disease claims. The EU joint-health claim for collagen hydrolysate was rejected in 2012.

Food-first versus supplement-first

Your body makes collagen from the protein you eat, with vitamin C as a necessary co-factor. So the food-first approach is to eat enough total protein and plenty of fruit and vegetables rich in vitamin C. Gelatine and slow-cooked cuts with connective tissue provide collagen-type amino acids too.

Bone broth is often promoted as natural collagen. An analysis of commercial, café and home-made broths found that amino acid levels in broth made to a standard recipe were significantly lower than those in a 20 g research dose of collagen, and that broth made to non-standard recipes varied widely; the authors concluded broth is "unlikely" to provide "a consistently reliable source" of the key amino acids (Alcock and colleagues, International Journal of Sport Nutrition and Exercise Metabolism, 2019). Broth can be a pleasant, nourishing food. It is not a dependable collagen supplement.

For skin, start with the basics, such as sun protection and a varied diet, before spending on supplements. You can check how your current eating pattern compares with dietary guidelines using our free healthy eating check. Fibre-rich plant foods, covered in our high-fibre foods list, also bring vitamin C and other nutrients that support general health.

Who should be careful

Collagen is generally well tolerated. The 2024 osteoarthritis meta-analysis found no increase in adverse events or withdrawals compared with placebo (Liang and colleagues, 2024). Some groups should still take advice first:

• People with fish, shellfish or egg allergies, depending on the collagen source and any added ingredients.
• People with kidney disease who have been advised to limit protein, because collagen adds to total protein intake.
• Pregnant or breastfeeding women, who should check any new supplement with their midwife or doctor.
• Anyone on a medically prescribed diet or taking several supplements; a dietitian can check for overlap between products and interactions with medicines.

When the money is wasted

Based on the trials, collagen is most likely to be wasted money when:

• you are buying it purely for wrinkles or skin "glow", where independent, high-quality trials show no benefit;
• you are healthy and active with no joint problems, where EFSA found no established effect;
• you are using it as your main protein, when complete proteins are cheaper and higher quality;
• the product is a "beauty blend" with doses that do not match any trial;
• you have been taking it for months without a clear before-and-after check.

It may be worth a time-limited trial if you have osteoarthritis pain and want to try something low-risk alongside standard care, ideally after discussing it with your GP, physiotherapist or dietitian.

How collagen fits with other supplements

Collagen is one of several supplements where marketing runs well ahead of evidence. The same critical approach applies to others on our site: omega-3 supplements, where benefits depend on dose and who you are; vitamin D supplements, where advice depends on season, age and where you live; vitamin B12 supplements, which are essential for vegans; and berberine, which is often marketed as a natural alternative to weight-loss drugs.

If you are taking several products, a single review of what you take, why, and what the evidence says can save money and reduce the risk of interactions.

Book a session with a dietitian

The practical takeaway

• Skin: independent, high-quality trials show no reliable benefit. Put sun protection and a good diet first.
• Osteoarthritis: small-to-moderate pain relief in pooled trials. A three-month trial alongside standard care is reasonable if you want to try it.
• Bone and muscle: promising single trials only. Not a replacement for proven treatment, exercise and adequate protein.
• Source: marine is not proven better than bovine; choose by diet, allergies and price.
• Quality: look for an independent seal such as USP Verified, and remember a seal proves purity, not benefit.
• Personal advice: for questions about your health, medicines or diet, book a session with a registered dietitian.