Glycosaminoglycans from food

Cartilage is not mostly collagen by function — it is collagen fibres holding a water-swollen gel of glycosaminoglycans, and it is the gel that resists compression. So the second question about any collagen-rich food is what glycosaminoglycans it delivers: chondroitin sulfate, hyaluronic acid, keratan sulfate, dermatan sulfate.

The answers range from "surprisingly good" to "nobody has ever looked."

Chondroitin sulfate survives cooking

This is the genuinely encouraging finding. Chondroitin was extracted from ordinary chicken leg bone soup and compared against chondroitin obtained enzymatically from the same bones. The soup-derived molecule averaged 35.81 kDa, the enzymatic reference 37.18 kDa, and infrared spectroscopy and HPLC found no significant structural difference.

That is a roughly 4 percent difference in molecular weight. Chondroitin comes through a simmer essentially intact, which is not obvious in advance — plenty of food polysaccharides do not.

It also does something. In the one bone-broth animal study with a skeletal endpoint, chicken-vegetable broth was fractionated by ion-exchange chromatography, and the fraction containing hyaluronan and chondroitin sulfate was the one that most inhibited osteoclast differentiation. Both the whole broth and that fraction raised bone mineral density and bone volume in ovariectomised rats. Bone, not cartilage — but the constituents were identified rather than assumed.

Then the arithmetic

A carcass survey measured how much chondroitin a broiler chicken actually contains. The answer: about 1.9 g in the entire 1.66 kg bird. Keel cartilage, the part usually named as the good source, held under 10 percent of it; most is distributed through the limb bones and joints.

The trial dose of chondroitin is 800–1200 mg per day. So a whole chicken contains one to two days' worth — if you extracted all of it, which cooking does not. The one measurement of transfer into soup reports a 0.14 percent process yield.

There is no plausible reading in which a bowl of stock delivers a studied chondroitin dose.

What the chondroitin dose buys

Worth knowing before chasing it. The Cochrane review pooled 43 trials and 9,110 participants. Its highest-quality result — high level of evidence, low risk of bias — is that 53 of 100 people on chondroitin achieved a 20 percent pain reduction against 47 of 100 on placebo. An absolute difference of six percentage points.

Joint space width loss was a relative 4.7 percent less than placebo, also high quality. Serious adverse events were lower than placebo.

That is a real, replicated, small effect at a pharmaceutical dose. The reviewers add the sentence that matters most: the benefits "were uncertain when we limited data to studies with appropriate allocation concealment or a large study sample (> 200) or to studies without pharmaceutical funding."

Whether any of it is absorbed

This matters more for the food question than the efficacy numbers do, and the literature does not agree with itself.

One group of twenty healthy men, dosed twice, reports that it is. Four grams of bovine chondroitin sulfate raised plasma levels more than 200 percent above a 0.3–5.3 µg/mL baseline — itself measured on six of the twenty — peaking at two hours, with the author describing the absorbed material as a high-molecular-mass polysaccharide alongside partially depolymerised and desulfated derivatives. Four grams of shark-derived chondroitin, given to the same twenty volunteers, behaved differently again — a 120 percent rise to a Cmax of 4.87 µg/mL, but taking 8.7 hours to peak, with an altered disaccharide composition. Source changes the kinetics substantially.

An independent group reached the opposite conclusion. Measuring circulating chondroitin sulfate at around 20 µg/mL, they found those endogenous levels not detectably altered by ingestion at all, and concluded that any effect would have to come from breakdown products rather than the polysaccharide.

The disagreement is methodological — separating a swallowed dose from a large endogenous pool is the hard part of this measurement, and a null is a failure to detect rather than proof of non-absorption. It has not been resolved. A separate finding sharpens the practical point: a biofermentative preparation produced plasma concentrations about 44 percent higher than an animal-derived one at the same 2,400 mg dose, so even among pharmaceutical products, exposure is not determined by the number on the label.

For the food question this cuts one way. If absorption from a standardised gram-scale dose is still contested after twenty years, the milligram quantities a stock could plausibly deliver are not where the answer is going to come from.

Hyaluronic acid: present, and fragile

Hyaluronan was the co-constituent in the active bone-broth fraction, so it is demonstrably in there. Two problems follow.

It is heat-labile — hyaluronic acid depolymerises readily in hot water, and long simmering should be expected to fragment it substantially. Nobody has measured how much survives.

That may matter less than it appears. When 300 kDa hyaluronan was given orally to rats, intestinal bacteria degraded it to oligosaccharides, which were then absorbed in the large intestine and distributed to skin as di-, tetra- and polysaccharides. Nothing crossed as the intact polymer. So the gut fragments hyaluronan regardless of what the pot did first, and a simmer destroying high-molecular-weight structure is destroying something that was not going to be absorbed in that form anyway. What this does not tell you is whether any of it reaches cartilage, or does anything once there — the study followed it to skin, in rats.

And for chicken feet specifically, the only published hyaluronic acid figure comes from a paper that has since been retracted. This guide does not cite it, and the hyaluronic acid content of chicken feet is unknown.

Keratan and dermatan sulfate: nobody has looked

Keratan sulfate is a major glycosaminoglycan of aggrecan — it is genuinely part of the cartilage people are trying to eat. Dermatan sulfate decorates decorin and biglycan in tendon and skin.

Neither has ever been measured in any broth or stock. Not measured and found low — never looked for. There is likewise no oral keratan sulfate trial in any species, and no oral dermatan sulfate trial with a joint endpoint.

The distinction matters: unmeasured is not the same as absent. Claims about "the full spectrum of glycosaminoglycans in bone broth" are extrapolations from chondroitin, the one glycosaminoglycan that has actually been measured.

The glucosamine mistake

One widespread claim is worth correcting explicitly, because it fails at the first step.

Chondroitin sulfate contains no glucosamine. Its repeating disaccharide is glucuronic acid plus N-acetyl-galactosamine. Glucosamine occurs in hyaluronan, keratan sulfate and heparan sulfate — not chondroitin. So "long simmering breaks chondroitin down and releases glucosamine" cannot be right at any dose: chondroitin has no glucosamine to release.

Separately, liberating free glucosamine from the glycosaminoglycans that do contain it requires strong acid at length — the industrial process uses molar hydrochloric acid near boiling for hours. A tablespoon of vinegar in a stockpot is two to three orders of magnitude away, and a weak acid rather than a strong one.

There is no measurement of free glucosamine in bone stock, chicken feet, or any cooked food. Not low — unmeasured, and mechanistically implausible.

The glycosaminoglycans, side by side

Everything above as a ledger. The two right-hand columns are independent questions: whether a molecule survives into your pot, and what the evidence says about that molecule as a supplement, are unrelated facts, and neither answers the other.

Two rows carry no evidence chip, and that is the point. Nobody has run an oral keratan or dermatan sulfate trial with a joint endpoint in any species, so there is no tier to give them — an em-dash rather than a rating that would imply somebody looked and found nothing. They are unmeasured in the pot and untrialled in the body, which is a different and more complete kind of silence than any other row here.