Collagen for Skin, Joints and Bone: Evidence by Use

Evidence labels used on these pages

Promising but limited means some positive trials with weaknesses. Mixed means reviews disagree. Insufficient evidence means too few or too weak studies. Proper evidence not found means we searched and found nothing we could responsibly cite. A marketing claim is a claim made by sellers without trials behind it.

What reviews found, use by use

Why the same supplement gets different verdicts

Three things explain most of the disagreement.

• Who paid and how good the trial was. The 2025 skin meta-analysis split trials by funding source and by quality. Trials not funded by pharmaceutical companies and the high-quality trials showed no significant effect. [3] Most trials in a 2026 skin review were rated at high risk of bias. [9]
• What was measured. Many skin and bone studies report instrument readings or blood markers. A lab marker changing is not the same as a person feeling or seeing a difference, or avoiding a fracture.
• Which product. Trials differ in product type, amount, duration and who took part. A 2026 review noted multi-ingredient formulas and industry funding limit confidence. [10]

Joints, in a few lines

A 2024 meta-analysis of 35 trials (3,165 people with osteoarthritis; the main pain analysis used 25 trials) found small-to-moderate effects on pain and function compared with control, and no extra adverse events. [4] A 2018 review of 20 supplements rated the overall quality of evidence as very low. [11] The NHS lists exercise and a healthy weight among the main osteoarthritis treatments. [12] The NHS pages we read do not mention collagen, and it does not replace medical care.

Bone, in a few lines

In one 12-month trial of 131 postmenopausal women with reduced bone density, those given specific collagen peptides had small gains in spine and femoral neck T-scores compared with placebo. [13] Reviews say the bone literature has limitations that prevent firm conclusions [14] and that collagen should be seen as an add-on, not a replacement for osteoporosis treatment. [7]

Hair and nails

The NHS says most hair loss does not need treatment, and advises seeing a GP to learn the cause. [15] A systematic review of nutritional supplements for hair loss did not name collagen peptides among the products it highlighted. [16] Hair claims for collagen are mostly marketing claims at present.

Amounts used in studies

Study amounts are not advice. A 2026 tendon review described 15 to 30 g a day of collagen with vitamin C alongside heavy training. [8] A 2019 paper gave 2.5 to 15 g a day as the range seen in the literature. [19] See the reference table of amounts.