Glucosamine ± chondroitin

Glucosamine is an amino sugar used to build cartilage glycosaminoglycans; chondroitin sulfate is itself a major cartilage glycosaminoglycan. Substrate provision is the pitch. Whether oral dosing raises joint concentrations enough to matter is contested — which is why this entry lives and dies on the clinical trials. And the trials disagree in an unusually clean pattern.

The positive trials — and who funded them

Two 3-year placebo-controlled RCTs of crystalline glucosamine sulfate 1500 mg once daily found essentially no joint-space narrowing on treatment versus roughly 0.3 mm of loss on placebo, plus modest symptom benefit: Reginster 2001 (n=212) and its replication Pavelka 2002 (n=202). These are the only supplement trials anywhere claiming structural protection of cartilage.

Both were funded by Rottapharm, the manufacturer of the prescription-grade product tested, with company-affiliated authors. And radiographic joint-space width — the entire basis of the structure claim — is sensitive to knee positioning, with debated per-millimeter clinical meaning.

The null trials — and who funded those

The NIH-funded GAIT trial (n=1583, 24 weeks) found glucosamine HCl, chondroitin, and their combination all indistinguishable from placebo on the primary endpoint: 60.1% of placebo patients responded versus 64–66.6% on the supplements. The frequently quoted combination benefit in moderate-to-severe pain was a post-hoc subgroup, never confirmed prospectively — and the 2-year GAIT continuation went back and tested it as a treatment interaction, where it came to nothing. That continuation stayed null throughout: no arm, including celecoxib, beat placebo. An investigator-initiated Spanish trial of the combination was stopped for futility with placebo numerically ahead on pain.

The meta-analytic view matches. The BMJ network meta-analysis of large trials (n=3803) put pooled pain effects at 0.3–0.5 cm on a 10 cm scale — below its 0.9 cm clinical-relevance threshold — with minimal joint-space effects, and found industry-funded trials produced systematically larger effects. The OA Trial Bank individual-patient-data analysis (n=1625) found no effect and no responsive subgroup — though manufacturer-held datasets from the positive sulfate trials were not contributed.

The industry-funded MOVES trial found the combination non-inferior to celecoxib at 6 months, but with no placebo arm both results could be regression to the mean.

Formulation: sulfate against hydrochloride

The mixed results split cleanly by salt. The positive trials used European crystalline glucosamine sulfate, 1500 mg once daily (not divided), for 6 months to 3 years. US supplements are typically glucosamine HCl — the salt that failed in GAIT — and supplement-grade chondroitin has documented content-quality variability. Most retail SKUs bundle HCl salt, low-grade chondroitin, and MSM in a combination matching no trialed regimen. If you use this at all, the only formulation with positive (if conflicted) data is crystalline glucosamine sulfate, and nothing is expected before 4–12 weeks.

There is a mechanistic reading of that salt split worth knowing, because it reframes what the supplement might be doing. Oral glucosamine, at any dose, does not measurably raise blood glucosamine — but glucosamine sulfate does raise serum sulfate, and synovial fluid sulfate tracks serum sulfate almost exactly. Sulfate is required for proteoglycan sulfation. On that account the active moiety is the sulfate, not the glucosamine, which predicts precisely what the trials found: non-sulfate salts should fail. It remains a hypothesis — no trial of sulfate itself with a cartilage endpoint exists — and it is set out under sulfur & sulfate.

Chondroitin on its own

Chondroitin is usually discussed as glucosamine's sidekick and has a better independent evidence base than that implies. The Cochrane review pooled 43 trials and 9,110 participants. Its single highest-quality node — high level of evidence, low risk of bias — found 53 of 100 people on chondroitin achieved a 20 percent pain reduction against 47 of 100 on placebo. A six percentage point difference. Loss of minimum joint space width was 4.7 percent less than placebo, also high quality and low risk of bias, and serious adverse events were lower than placebo.

That is a small, replicated, structurally-relevant effect — better than glucosamine's record. The reviewers' own qualifier is the one that matters most, and it echoes the pattern across this whole section: benefits "were uncertain when we limited data to studies with appropriate allocation concealment, or large samples, or studies without pharmaceutical funding."

There is an unresolved question underneath all of it: whether swallowed chondroitin reaches the circulation at all. The evidence that it does is one group of twenty healthy men, dosed twice: a 4 g bovine dose raised plasma chondroitin more than 200 percent above baseline, peaking at two hours, and a shark-derived dose in the same twenty volunteers raised it 120 percent but took nearly nine hours to peak. The baseline those rises are measured against comes from six of the twenty. An independent group measuring the same thing found endogenous levels of about 20 µg/mL unchanged by ingestion, and concluded that any effect would have to come from breakdown products. Separating a swallowed dose from a large endogenous pool is the hard part of the measurement, so the null is a failure to detect rather than a refutation — but two decades on, the field has not settled it. Preparation matters too: a biofermentative product gave plasma levels about 44 percent higher than an animal-derived one at the same 2,400 mg dose.

One preclinical pairing is worth recording, because it is the only one anyone has run with this material. In a systematic review of 33 cell and animal studies of light therapy, adding chondroitin sulfate and glucosamine sulfate to the light enhanced its effect on articular cartilage in a rat model, where adding exercise or a topical anti-inflammatory added nothing over light alone. It is a rat result about a combination, and it says nothing about either substance taken on its own.

One consequence for anyone hoping to get chondroitin from food: every trial above used pharmaceutical-grade material at 800–1200 mg/day, and commercial chondroitin varies enough in composition that some samples contain negligible active content. Nothing in this literature transfers to chondroitin delivered in a bowl of stock — the quantities are covered under glycosaminoglycans from food.

Where the guidelines landed

OARSI's 2019 guideline grades glucosamine "strongly recommended against" for knee, hip and polyarticular OA, with "no efficacy" as the panel's stated rationale, and chondroitin "conditionally recommended against" for knee OA — grades that sit in the guideline's supplementary tables rather than its main text. AAOS and ACR guidance is similarly negative. European bodies (ESCEO) endorse prescription-grade crystalline glucosamine sulfate specifically — via a committee with documented manufacturer ties, which readers deserve to know.

What I take from it

The contradiction is the story: every large trial free of manufacturer involvement is null, and every positive structure trial was funded by the product's manufacturer. After more than two decades, no independent group has tried to replicate the crystalline-sulfate findings — itself informative. Safety, at least, is genuinely strong: adverse-event rates matched placebo across 3-year exposures, with the usual cautions for shellfish allergy, glucose monitoring in diabetes, and case reports of chondroitin interacting with warfarin.