Collagen Peptides: Questions to Ask Before Comparing Products

Collagen powders make similar promises, and their labels look alike. The research behind them is narrower than the marketing: a typical trial tests one particular collagen ingredient, measures a handful of outcomes, and runs for weeks or months. The four questions below help you hold a product page up against that research before you compare prices.

What collagen peptides are

Collagen is the main structural protein in skin, bone and tendon. Supplements use collagen that has been broken down (hydrolyzed) into short peptides, extracted mainly from cattle and pigs, and also from fish skin, scales and bones or from chicken ^{Source [1]}. Labels call it "collagen peptides", "hydrolyzed collagen" or "collagen hydrolysate", often followed by the animal source.

Two products can share that name and still differ in source animal, processing and peptide size. The studies below do not treat them as interchangeable, and a fair comparison should not either.

Question 1: What outcome was measured?

"Supports skin" is not an outcome. Trials measure specific things: skin hydration and elasticity with instruments, wrinkle depth from images, or symptom questionnaires in people with knee osteoarthritis. A result on one outcome does not carry over to another.

Skin has the largest body of research. A 2021 meta-analysis pooled 19 randomized, double-blind trials with 1,125 participants, 95% of them women, and found better hydration, elasticity and wrinkle scores with hydrolyzed collagen than with placebo ^{Source [2]}. A 2023 meta-analysis of 26 trials with 1,721 participants found improved hydration and elasticity, and flagged several sources of bias in the included trials ^{Source [3]}.

Joint research measures symptoms, not skin. A 2019 meta-analysis of placebo-controlled trials in osteoarthritis found lower total questionnaire scores and lower visual pain-scale scores, although the questionnaire's separate pain and function subscores did not differ from placebo ^{Source [4]}. An updated 2025 analysis of 11 trials with 870 people with knee osteoarthritis reported better pain and function scores, with high variation between trials ^{Source [5]}. Both analyses pool "collagen-based supplements", a broad label, so it is worth checking which product each trial used.

Question 2: Which product was studied?

Many trials test a branded ingredient made by a collagen supplier, not the powder on a store shelf. A 2014 trial randomized 69 women aged 35 to 55 to one of two amounts of a hydrolysate described as "specific collagen peptides", or to placebo, for eight weeks. Skin elasticity improved in both collagen groups, while changes in skin moisture did not reach statistical significance ^{Source [6]}.

A 12-month trial in 131 postmenopausal women with reduced bone density tested a named product, FORTIBONE from GELITA AG, and reported higher bone mineral density at the spine and femoral neck than with placebo ^{Source [7]}. That is one trial of one ingredient, not evidence about collagen powders in general.

When a product page mentions research, look for the name of the ingredient that was tested. If the page does not name it, you cannot tell whether the study applies to that product.

Question 3: How much, for how long, and compared with what?

Read the amount and duration as facts about the study, not as instructions. The 2014 skin trial used 2.5 g or 5 g a day for eight weeks; the bone trial used 5 g a day for a year. The three labels in our comparison list 12 g to 20 g of collagen per serving. A larger serving of a different ingredient is not the same thing as the studied product. We do not give dosing guidance.

Then check the comparison group. A 2017 study gave 25 people with brittle nails a branded collagen peptide product (VERISOL) for 24 weeks and reported faster nail growth and fewer broken nails. It was open-label with no placebo group ^{Source [8]}, so it cannot separate the product's effect from time, the season, or participants' expectations. Our guide to reading a peptide study explains why the control group matters.

Question 4: Who paid for the research?

Funding does not make a result wrong, but it is a known source of bias, and in collagen research it appears to matter. A 2025 meta-analysis of 23 trials with 1,474 participants found that collagen improved hydration, elasticity and wrinkles when all trials were pooled. When the authors split the trials by funding, those without pharmaceutical-company funding showed no effect on any of the three, and neither did the higher-quality trials. The authors concluded that, for now, clinical evidence does not support taking collagen supplements to prevent or treat skin aging ^{Source [9]}.

Funding is usually stated in the full paper, not the abstract. The bone trial above, for example, states that GELITA AG provided part of the funding ^{Source [7]}.

Applying the questions to three labels

We read the published labels of three unflavored collagen powders sold in the United States. Each Supplement Facts panel lists collagen peptides with no added vitamins or botanicals: two from bovine sources at 20 g per serving, and one from Alaskan pollock at 12 g per serving ^{Source [10] Source [11] Source [12]}.

None of the three product pages names a published study of that exact product as of Oct 18, 2026. One footnote says "various studies have suggested benefits" over several months without naming them ^{Source [10]}. Another page says its marine collagen is "more bioavailable" than bovine collagen and points to molecular weight in a footnote rather than to a study ^{Source [12]}.

The full side-by-side is in our comparison of three collagen powders, and our Vital Proteins documentation review walks through one label in detail.

What these questions cannot tell you

A well-documented label does not show that a product will work for you, and a study of one ingredient does not show that another works. FDA does not approve dietary supplements before they are marketed, and claims such as "supports healthy skin" are structure/function claims that FDA has not evaluated ^{Source [13]}. For how we weigh different kinds of evidence, see our methodology.