Collagen Peptides: What the Trials Show for Skin, Joints, Bone and Muscle

Collagen peptides have more randomized trials behind them than most supplements, and the results depend heavily on who ran them. For skin, pooled trials show better hydration and elasticity, but a 2025 meta-analysis found the benefit disappeared when it looked only at trials not funded by companies, or only at high-quality ones.^[1] For knee osteoarthritis the pooled effect on pain is small to moderate.^[4] For muscle, collagen is a poor protein: in controlled feeding studies it did not raise muscle building where whey and pea did.^[10][11] If you are losing weight on a GLP-1, that last point matters most.

What collagen peptides are

Collagen is the main structural protein in skin, tendon, cartilage and bone. Supplement makers extract it from cow hide, pig skin or fish skin and break it into short chains with enzymes, which is what “hydrolyzed collagen” and “collagen peptides” mean on a label. Gelatin is the same protein in longer pieces, and gelatin and jello for weight loss covers how the two compare. Despite the shared word, these food peptides have nothing to do with the injectable research peptides sold by telehealth clinics, which are explained in what peptide therapy is.

Collagen is low in several essential amino acids and contains no tryptophan at all, so nutrition scientists class it as an incomplete protein.^[12] That review, co-written by a scientist at an industry-backed collagen institute, calculated that collagen can make up a large share of daily protein without unbalancing a normal Western diet, as long as the rest of the protein comes from other foods.^[12]

The evidence by claim

Skin: the clearest case of funding changing the answer

Skin is the most studied use and the most marketed one. A 2023 meta-analysis of 26 randomized trials with 1,721 people found that hydrolyzed collagen improved skin hydration and elasticity compared with placebo, while noting several biases in the trials it pooled.^[2] A 2026 review of 25 trials found collagen beat placebo on hydration in 10 of 15 trials that measured it, on elasticity in 10 of 13 and on wrinkles in 9 of 10, and rated most of those trials at high risk of bias.^[3]

The 2025 analysis in The American Journal of Medicine asked a question the earlier ones had not: does it matter who paid? Across all 23 trials it included, collagen improved hydration, elasticity and wrinkles. Split by funding, trials not funded by pharmaceutical companies showed no effect on any of the three, and trials rated high quality showed no significant effect either. The authors concluded that there is currently no clinical evidence to support collagen supplements for preventing or treating skin aging.^[1]

That does not prove collagen does nothing for skin. It means the positive results rest mostly on trials run or paid for by the companies selling the product, and the independent trials so far have not reproduced them. People who notice looser skin after large weight loss are dealing with a structural change that a supplement has not been shown to reverse; what changes in skin after major weight loss explains why.

Joints: a modest effect on osteoarthritis pain

Joints have the most consistent record. A 2024 meta-analysis of 35 randomized trials with 3,165 people with osteoarthritis found collagen products reduced pain by a small to moderate amount (a standardized difference of 0.35, rated moderate certainty) and improved function (0.31, rated high certainty), with no more side effects or dropouts than the comparison groups.^[4] A 2025 meta-analysis limited to knee osteoarthritis, with 11 trials and 870 people, also found better pain and function scores, though the results varied a great deal from trial to trial.^[5]

A standardized difference of about 0.3 is the size of effect that some people notice and many do not. The trials used different products, from hydrolyzed peptides to other collagen preparations, so the pooled number does not point to a particular powder.

Bone: one positive trial, with the maker in the room

The trial cited most for bone gave 131 postmenopausal women with low bone density either 5 g of collagen peptides or a placebo every day for a year. Spine and hip T-scores rose slightly with collagen and slipped slightly with placebo, a difference of about a tenth of a T-score point.^[6] The paper declares no conflicts of interest, but its acknowledgments say the collagen maker partly funded the study, which used that company’s product, and one author worked at an industry collagen research institute.^[6] A Greek trial of 51 women with osteopenia found better bone density when collagen was added to calcium and vitamin D than with calcium and vitamin D alone, but the comparison group is not described as taking a placebo in place of the collagen.^[7]

Neither trial measured fractures. Bone loss during rapid weight loss is a real concern, and the treatments with fracture evidence are covered in bone density and fracture risk on GLP-1s.

Muscle: why collagen is the wrong protein to lean on

The muscle studies split into two groups that seem to disagree. In a German trial, 53 older men with sarcopenia did 12 weeks of strength training and took either 15 g of collagen peptides or a non-protein placebo after each session. The collagen group gained more fat-free mass (4.2 kg against 2.9 kg), lost more fat and gained more strength.^[8] The collagen maker supplied the product and paid part of the costs.^[8] A 2024 meta-analysis of 19 such training studies found more fat-free mass with collagen, rating the certainty moderate for body composition and low for most other outcomes; its senior author also led the German trials.^[9]

The second group of studies compared collagen with other proteins instead of with nothing. In 22 older women, 30 g of whey twice a day raised muscle protein synthesis at rest and after exercise, while the same amount of collagen did not raise it above baseline.^[10] In 31 men aged about 72 eating a controlled diet, an extra 50 g a day of whey or pea protein raised muscle protein synthesis, and the same extra amount of collagen left it unchanged.^[11]

Read together, collagen does better than a placebo with no protein in it and worse than a complete protein. That distinction is the one that matters on a GLP-1, where appetite is low and every gram has to count. A 2025 joint advisory from four obesity and nutrition societies lists preserving muscle and bone through resistance training and diet as a priority during GLP-1 treatment, and notes that protein targets of 1.2 to 1.6 g per kg a day have been proposed during active weight loss.^[14] Collagen grams count toward that total on a label, but they do far less for muscle than the same grams of whey, eggs, dairy, meat or soy. Hair and muscle loss on GLP-1s covers how much lean mass the trials recorded, and the protein calculator works out a daily target.

Weight: small changes in short trials

Collagen is sometimes sold as an appetite or fat-loss aid. The trials are few, small and short. In one 12-week Korean trial in adults over 50, 15 g of collagen a day went with a fat-mass change of −0.49% against +2.23% on placebo.^[13] The appetite studies and the claim that collagen or gelatin raises GLP-1 are reviewed in the gelatin article; none shows weight loss anywhere near what the drugs produce.

Regulatory status

Collagen peptides are sold in the United States as dietary supplements or food. FDA does not have the authority to approve dietary supplements for safety or effectiveness, or to approve their labels, before they are sold; companies are responsible for making sure their products meet the safety standards.^[15] A collagen powder can say it “supports” skin or joints without any trial behind that brand.

If you are on a GLP-1 and thinking about collagen

• Count it, but do not build on it. Collagen adds protein grams, and it lacks tryptophan.^[12] Complete proteins did more for muscle in head-to-head trials.^[10][11] The trade-offs between protein sources are set out in plant-based protein on a GLP-1.
• Strength training does the heavy lifting. Every positive collagen muscle trial paired it with supervised training, and the training groups improved on placebo too.^[8][9]
• For joints, expect a modest effect at best. The pooled pain benefit in osteoarthritis is small to moderate.^[4]
• For skin, the independent evidence is negative so far.^[1]