Collagen

Collagen is the protein that holds animals together. Skin, bone, tendon, cartilage: the structural scaffolding is collagen, and it is the most abundant protein in a mammal.

Which means the first thing worth saying is the thing the marketing never says. Collagen and Gelatin are the same material. Heat collagen and it becomes gelatin. Break gelatin down further into short peptides and it becomes the hydrolysed collagen sold in sachets. One protein, three processing stages, three price points.

What that means if you avoid animal products

The same thing gelatin means. A collagen supplement is made from cattle hide, pig skin or fish, and no amount of packaging it as a beauty product changes that.

Marine collagen is from fish and is used where pork must be avoided, which makes it relevant to reading labels when you keep halal or kosher, and it is also a fish product for anyone with that allergy.

There is no vegan collagen, and there cannot be, because the molecule is animal structural protein. Products sold as “vegan collagen” are either amino acids and vitamin C intended to support your own production, or they are marketing. Both may be reasonable purchases. Neither is collagen.

Does it do anything

Here the honest answer is more interesting than either of the usual ones.

The sceptical version says collagen is digested into amino acids like any protein, so eating it cannot deliver collagen to your skin. The first half is broadly right and the conclusion does not follow: some collagen-derived di- and tripeptides survive digestion intact and enter the bloodstream, which is the mechanism the research turns on.

The evidence itself is better than that dismissal. A 2023 meta-analysis pooling fourteen randomised controlled trials and 967 participants found significant improvements in skin hydration and elasticity against placebo, with smaller effects on wrinkles.

The same paper names the problem with reading too much into it: “significant heterogeneity, mainly due to differences in the composition of the supplements used”. Fourteen trials of fourteen different products at different doses for different durations. Something works. Which something, at what dose, in whom, is not settled.

Add the ordinary caution for this category, which is that a great deal of collagen research is funded by companies selling collagen, and you get a position that is neither dismissal nor endorsement: modest measurable effects, uncertain generalisation, and no reason to expect what an advertisement promises.

Reading a collagen label

The source is often stated and worth finding. Bovine, porcine, marine or chicken. It appears in the ingredient list rather than the Supplement Facts panel, a split covered in the guide to reading a supplement label.

“Type I” and “Type II” are real distinctions, not marketing. Type I dominates skin and bone, type II dominates cartilage, and products aimed at joints usually use type II.

A large number in grams is not a large amount of evidence. Trials cluster at modest daily doses, and a sachet containing far more has not been shown to do proportionally more.

Collagen in a cream is a different claim entirely. The molecule is far too large to cross the skin barrier, so topical collagen works as a humectant sitting on the surface. That is a real if unremarkable job, and it is not the job the packaging implies.

The one thing nobody disputes

It is protein. Ten grams of collagen is ten grams of protein, weighted oddly towards glycine and proline and short on some essential amino acids, which is why it is a poor choice as your only protein source and a perfectly ordinary addition alongside others.