Joint supplements for osteoarthritis: what the evidence shows

Osteoarthritis is the wear-and-repair joint disease that becomes common after 50, and the supplement aisle sells hope by the bottle. Glucosamine, chondroitin, turmeric, collagen, fish oil, and more all carry claims about cushioning cartilage or calming pain. This article does not tell you what to take. It reports what randomized trials, systematic reviews, and the major osteoarthritis guidelines have found, and where the evidence is thin it says so.

Glucosamine and chondroitin: the most-studied, most disappointing

These two are the giants of the joint-supplement market, and they have been tested harder than any of the others. The landmark test was the Glucosamine/chondroitin Arthritis Intervention Trial (GAIT), funded by the U.S. National Institutes of Health rather than by a manufacturer. It randomly assigned 1,583 people with symptomatic knee osteoarthritis to glucosamine, chondroitin, both, the prescription drug celecoxib, or placebo for 24 weeks (Clegg et al., NEJM, 2006).

The result was sobering. Overall, glucosamine and chondroitin, alone or together, were not significantly better than placebo at the primary goal of cutting knee pain by 20 percent. Part of the reason is that the placebo group did well on its own, with a 60.1 percent response rate. An exploratory look suggested people with moderate-to-severe pain might get some relief from the combination, but that finding came from a subgroup and was not a confirmed result (Clegg et al., NEJM, 2006). A follow-up phase tracked joint structure over two years and found that changes in the space between the knee bones, a marker of cartilage loss, did not differ between placebo and any supplement group (NCCIH summary of GAIT; Sawitzke et al., 2008).

When many trials are pooled, the picture stays flat. A network meta-analysis of 10 trials in 3,803 patients found the pain benefit of glucosamine was about -0.4 cm on a 10 cm scale, chondroitin about -0.3 cm, and the combination about -0.5 cm. The authors had prespecified -0.9 cm as the smallest difference a patient would notice, and none of the estimates reached it. Their conclusion was blunt: these supplements do not have a clinically relevant effect on joint pain or on joint-space narrowing (Wandel et al., BMJ, 2010).

That same analysis flagged a pattern worth remembering across this whole category. Industry-funded trials reported larger benefits than independent trials, and the difference was statistically significant (Wandel et al., BMJ, 2010). Guidelines have largely followed the strongest evidence. The 2019 OARSI guideline strongly recommends against both glucosamine and chondroitin for knee osteoarthritis, and the 2019 American College of Rheumatology and Arthritis Foundation guideline strongly recommends against glucosamine, alone or combined with chondroitin (OARSI, 2019; ACR/AF, 2020). Not everyone agrees: some European bodies and the American Academy of Orthopaedic Surgeons take a softer line, noting the evidence is inconsistent. On safety, large studies have not turned up major harms, though glucosamine can raise blood sugar in some people, and both may increase bleeding risk alongside the blood thinner warfarin (NCCIH).

Turmeric and curcumin: the most promising short-term signal, with caveats

Curcumin, the active compound in turmeric, is the one bright spot in newer randomized data. Several systematic reviews of knee osteoarthritis trials report that curcumin extracts reduce pain and improve function compared with placebo, with some analyses finding relief roughly comparable to nonsteroidal anti-inflammatory drugs (NSAIDs) over the short term, and no clear rise in side effects. One meta-analysis pooled 11 randomized trials with 1,258 participants (systematic review, PMC, 2022).

The caveats are large. The trials are mostly small and short, often run for weeks rather than years, and use many different formulations at different doses, which makes them hard to compare. A 2025 critical review of these meta-analyses warned that the underlying study quality is uneven and that industry involvement is common, so the enthusiastic pooled numbers should be read with caution (Frontiers in Pharmacology, 2025). Taken together, curcumin has the best short-term randomized signal in this group, and also the least mature evidence base. No major guideline endorses it as established therapy.

Collagen: modest meta-analysis effects, mostly from makers

Collagen peptides, sold as hydrolyzed collagen or as undenatured type II collagen, show small but statistically significant benefits in pooled analyses. A trial-sequential meta-analysis of 35 randomized trials in 3,165 patients reported small-to-moderate effects on pain and function versus control (Osteoarthritis and Cartilage, 2024). Individual trials, such as one testing 3,000 mg a day of low-molecular-weight collagen peptide over 180 days, have reported reduced knee pain with no notable adverse events.

The weakness is familiar. Much of this literature is funded by collagen manufacturers, the formulations vary widely, and the same funding-effect concern that dogs glucosamine applies here. The signal is real in the pooled numbers, but it rests on a foundation that independent, long-term trials have not yet firmed up.

Fish oil and omega-3: no clear win for joints

Fish oil is anti-inflammatory in the lab, which makes it a plausible candidate, but the joint trials have underwhelmed. The most-cited randomized test compared high-dose fish oil against a low-dose comparator in 202 people with knee osteoarthritis. Both groups improved, and the low-dose group did slightly better on symptom scores at two years, with no difference in cartilage loss (Hill et al., 2016). Because both arms received some omega-3, the trial could not show whether either beat true placebo. Broader reviews of omega-3 for chronic pain are mixed at best. There is no strong randomized case that fish oil relieves osteoarthritis pain.

Avocado-soybean unsaponifiables: a smaller, mixed evidence base

Avocado-soybean unsaponifiables (ASU) are an extract sold mainly in Europe. A systematic review of four rigorous placebo-controlled trials found that three suggested benefit for osteoarthritis symptoms, and later meta-analyses reported a significant reduction in pain, with better results for knee than hip osteoarthritis (systematic review, DARE/NCBI). The evidence base is smaller than for glucosamine, several trials are industry-linked, and guideline support is limited. It sits in the “some signal, not settled” bucket.

Vitamin D: tested and negative for joints

Low vitamin D has been linked to osteoarthritis in observational studies, but association is not proof, and the randomized test was clear. A two-year trial randomized 413 people with symptomatic knee osteoarthritis and low vitamin D to supplementation or placebo. The findings did not support vitamin D for preventing cartilage loss or for improving WOMAC knee pain (Jin et al., JAMA, 2016). Vitamin D matters for bone health, but the joint-pain claim did not hold up when tested properly.

Bottom line

Sorting the shelf by evidence quality, not by marketing: glucosamine and chondroitin have been tested the most and have largely failed the best tests, which is why two major 2019 guidelines advise against them. Curcumin has the most promising short-term randomized signal but an immature evidence base. Collagen and ASU show modest pooled benefits that lean on manufacturer-funded studies. Fish oil and vitamin D have not beaten placebo for joint pain in their strongest trials. Running through all of it is a funding effect: studies paid for by sellers tend to look better than independent ones.

This article is evidence reporting, not medical advice. Supplement responses vary, some products interact with medications such as blood thinners, and only a clinician who knows your history can help you decide what is right for you.

Common questions

Do glucosamine and chondroitin work for knee arthritis?

In the highest-quality trials they do not clearly beat placebo. The NIH-funded GAIT trial found no significant advantage over placebo for most people, and a large network meta-analysis found the pain benefit fell below the level a patient would notice. The 2019 OARSI and American College of Rheumatology guidelines recommend against them. Some people report feeling better, and both are generally safe, but the group data do not support a reliable effect.

Is turmeric or curcumin better than glucosamine for joint pain?

Curcumin has the more promising short-term randomized signal. Several systematic reviews find it reduces knee pain versus placebo, sometimes comparably to anti-inflammatory drugs over a few weeks. The catch is that these trials are mostly small, short, and often industry-linked, so the evidence is less settled than the number of positive studies suggests. No major guideline endorses it as established treatment yet.

Why do some studies show a benefit and others show none?

A big driver is who paid for the study. A 2010 BMJ analysis found that trials funded by supplement makers reported larger benefits than independent trials, and the difference was statistically significant. Small, short, or poorly blinded trials also tend to show bigger effects. When you weight the large, independent, long-term trials most heavily, the benefits shrink or disappear.

Does fish oil or vitamin D help osteoarthritis?

Not clearly for joint symptoms. The main fish oil trial could not show a benefit over a true placebo, and a low dose did as well as a high dose. A two-year randomized vitamin D trial found no reduction in cartilage loss or knee pain versus placebo. Both nutrients matter for general and bone health, but the osteoarthritis-specific claims did not hold up in testing.

Are these supplements safe to try?

Most have good safety records in trials, but safe is not the same as effective, and some carry real interactions. Glucosamine can raise blood sugar in some people, and glucosamine and chondroitin may increase bleeding risk with the blood thinner warfarin. This is general information, not medical advice. Check with a clinician or pharmacist before starting anything, especially if you take medication or manage diabetes.