Understanding collagen: terms, sources and context

Collagen is not a single substance but a whole protein family. Twenty-eight types are described in humans, of which a few dominate by quantity. They share a very characteristic build: three protein chains wind into a triple helix. That structure is why collagen withstands tensile force, and it is also why the amino acid composition is unusual: glycine occupies every third position, alongside a great deal of proline and hydroxyproline.

Type I is the largest share by quantity in the body. It occurs in skin, tendons, ligaments and bone, forming fibres organised into bundles. It appears most often on product labels, frequently combined with type III.

Type II is the main component of hyaline cartilage, the tissue at joint surfaces. It forms a finer network than type I. Products declaring type II often source it from chicken cartilage.

Type III occurs together with type I, among other places in skin, blood vessels and internal organs. It forms finer fibres and makes up a larger proportion in young tissue than in older tissue.

What that means for a label. When a product declares “type I and III”, that describes the origin of the raw material and its tissue share, not a destination in the body. A label tells you what the powder consists of. It does not tell you where that protein ends up after you eat it: collagen is digested like any other dietary protein, that is, broken down into amino acids and smaller peptides. No targeted routing to a particular tissue follows from a type declaration.

Evidenced basis: the structural and type descriptions draw on the reviews by Shoulders and Raines on collagen structure and by Ricard-Blum on the collagen family. Both are linked in the sources section.

Direct source: Shoulders MD, Raines RT: Collagen structure and stability. Annual Review of Biochemistry (2009). · Ricard-Blum S: The collagen family. Cold Spring Harbor Perspectives in Biology (2011).

A label is the only source of information legally tied to the specific product. Advertising copy is not. So it is worth reading in this order:

1. The ingredient list, top to bottom. Ingredients appear in descending order of weight. If collagen hydrolysate comes first, it is the main component. If fillers, carriers or sugar come before it, it is not.

2. The reference quantity. Figures apply per serving, per scoop, per capsule or per 100 grams. Two products are only comparable once you convert the figures to the same reference. This is the most common source of mistaken comparisons between two tubs on a shelf.

3. The animal species and source tissue. “Bovine collagen” is a different statement from “marine collagen from fish skin”. Both are permitted, but they are different raw materials.

4. Additives. Flavourings, sweeteners, anti-caking agents, acidity regulators. They are not a problem in themselves, but they often explain why one product weighs more or tastes different from another.

5. The address of the food business operator. Under the EU Food Information Regulation (Regulation (EU) No 1169/2011), the packaging must carry, among other things, the name of the food, the ingredient list, allergen labelling, the net quantity, the date of minimum durability and the address of the responsible business. Anything missing there is a factual signal to look closer.

6. What does not belong on a label. Statements attributing to a food the property of eliminating or averting a disease are not permitted for foods. If you read something like that on a pack or in a product description, that is a reason to look more carefully.

What a label cannot do. It says nothing about what the product does for you. It describes composition and quantity. That is precisely why it is useful: it is the part of the packaging that can be verified.

Direct source: Regulation (EU) No 1169/2011 on the provision of food information to consumers (2011).

Collagen hydrolysate is sold in three forms. They differ in handling, not in raw material. Which suits you depends on your routine, not on a ranking.

Powder. The most common form. It is stirred into water, coffee or yoghurt and usually has little noticeable taste. Practical: the serving size is freely chosen with the scoop, and the ratio of content to packaging is favourable. Impractical: it needs a drink, a scale or a spoon, and travels badly.

Capsules. Practical: precisely portioned, taste-neutral, portable, no stirring. Note: a capsule holds considerably less powder than a scoop, so a powder serving corresponds to a good many capsules. Do that arithmetic when comparing two products; it is exactly the reference-quantity trap from the label section.

Ready-to-drink liquids. Ampoules or shots. Practical: no preparation, fixed portion, good for travel. Note: they often also contain water, flavourings and sweeteners, and the packaging per serving is larger. Read the ingredient list here especially carefully.

What applies to all three. The form changes nothing about the composition of the collagen fraction. We recommend no form as the most suitable and deliberately state no quantities: both would be an intake recommendation, and we have no robust basis for one. The manufacturer’s information on the specific product is the relevant source.

Evidence limit: This reading aid does not derive any further product, selection or efficacy claim here.

When collagen comes up, two vocabularies almost always arrive together: the anatomical (“connective tissue”, “cartilage”, “tendon”) and the commercial (“for your joints”). This page keeps them apart, because only the first can be described.

What connective tissue is anatomically. Connective tissue links, shapes and carries other structures. It consists of cells and a surrounding extracellular matrix, and that matrix contains, besides water and proteoglycans, mainly collagen fibres. In skin those fibres lie as a network; in tendons and ligaments they run in parallel, because there they take tension in one direction.

What cartilage is. The hyaline cartilage at joint surfaces is a particularly water-rich tissue. Its resistance to pressure arises from the interplay of proteoglycans, which bind water, and a network of type II collagen holding that swollen tissue together. Cartilage has no blood supply of its own; it is served by the joint fluid.

Why no product statement follows from that. The fact that a tissue contains collagen says nothing about what an ingested collagen product does at that tissue. Ingested protein is digested and used in the body as a pool of building blocks, without any destination following from the origin of a raw material. Presenting it otherwise turns an anatomical description into a product claim. That boundary is exactly why no sentence on this page connects a collagen product to a joint.

What you can do with this distinction. It is a filter for advertising copy. If a product description jumps straight from anatomy to product benefit without naming a study with a population and an endpoint in between, the gap is precisely at that point.

Direct source: Ricard-Blum S: The collagen family. Cold Spring Harbor Perspectives in Biology (2011).

Studies on collagen peptides exist. Citing them is easy; placing them is the actual work. These four questions do most of it:

1. Which study type? A cell culture study observes cells in a dish. An animal study observes a different organism with a different metabolism. A controlled human study with blinding and a control group is fundamentally different and can support conclusions only within its study population and design. When a product description names “studies” without naming the type, that is the first thing to check.

2. Which population? Who was measured? Number of participants, age, sex, starting point. A study in thirty people of a particular age group describes that group. It does not automatically describe you.

3. Which endpoint? What exactly was measured, and with what? An instrument reading, a questionnaire and a subjective self-assessment are three different things. Endpoints defined after looking at the data are weaker than those defined in advance.

4. Who funded it? Manufacturer funding does not make a study wrong, but it belongs to the appraisal, as does the question of whether results have been independently repeated.

What one result is worth. One study alone is not enough for a general conclusion. A finding becomes more reliable when several independent groups measure in the same direction with comparable methods. Until then it points to an open question rather than delivering an everyday conclusion.

Why there is no results list on this page. Because such a list inevitably reads as a promise, even when every individual item is cited correctly. Instead we give you the tool to check any study claim yourself, including the one on a pack.

Direct source: Schulz KF, Altman DG, Moher D: CONSORT 2010 Statement for reporting parallel-group randomised trials. BMJ (2010).