Ranked against randomized trials for osteoarthritis: curcumin (turmeric extract) leads with repeated small-trial wins at 500–1,500 mg/day; boswellia shows similar signals in fewer studies; glucosamine sulfate is a contested maybe (the big independent trial was negative); chondroitin, collagen and MSM sit between small and unreliable effects; omega-3's arthritis evidence belongs mainly to rheumatoid, not osteo; and "joint complexes" combine sub-doses of everything, testable as nothing. OA's placebo response — around 60% improvement rates in trial control arms — explains most testimonials. Strengthening exercise and weight loss beat every row of the table, and no supplement regrows cartilage.
The scorecard
Osteoarthritis's fluctuating pain and outsized placebo response (FAQ) make this the aisle where controlled trials matter most. Ranked:
| Ingredient | Verdict | What trials show |
|---|---|---|
| Curcumin / turmeric extract | Limited — best in aisle | Repeated small RCT wins for knee-OA pain at 500–1,500 mg/day; a few NSAID-comparable results; small, short, formula-fragmented trials; liver and blood-thinner cautions. |
| Boswellia serrata | Limited | Several small positive knee trials (100–250 mg standardized extracts); same size caveats, cleaner safety so far; often paired with curcumin in trials and blends alike. |
| Glucosamine sulfate | Mixed | The 30-year argument: big independent GAIT negative; European prescription-sulfate trials modestly positive; guidelines split. Hydrochloride form leans null. |
| Chondroitin | Mixed | Sponsored meta-analyses flatter, independent ones shrink toward zero; chronic product-quality problems; one decent hand-OA trial. |
| Collagen (hydrolyzed / UC-II) | Insufficient | Small, short, manufacturer-heavy positives; mechanism-as-marketed is biochemically naive (FAQ). |
| MSM | Insufficient | A few small trials, borderline effects at 3–6 g/day; benign safety. |
| Omega-3 / fish oil | Insufficient for OA | Real evidence lives in rheumatoid arthritis (reduced tender joints, NSAID-sparing); OA trials, including a high-vs-low-dose comparison, show little specific effect. |
| Avocado-soybean unsaponifiables (ASU) | Limited | Modest French-trial record for hip/knee symptoms; little recent replication; niche availability. |
| Vitamin D | Deficiency-only | Correcting deficiency serves bone and muscle; supplementing replete OA patients did not reduce pain or cartilage loss in trials. |
| SAM-e | Mixed | Older trials comparable to NSAIDs; expensive, quality-variable, interaction-prone (serotonergic); rarely a first pick. |
| "Joint complex" blends | Insufficient | Sub-trial doses of many rows above, untestable and untested as sold. |
The comparison that reframes the table
Strengthening-exercise programs for knee OA produce pain improvements in the same range as oral NSAIDs in trials; 5–10% weight loss produces measurable relief with side effects that are all bonuses. Nothing in the table approaches either — and every honest ingredient page above ends by saying so — which sets the only sensible ordering: the program first, a disciplined single-ingredient experiment second for those who want one, and the blend aisle never, since it cannot even be tested.
Cartilage regrowth, finally, is claimed by no trial for any row — "rebuilds joints" on a label is fiction, and a useful seller-honesty screen.
The honest playbook
- Confirm it's OA — the flagship guide's red-flag list (hot joints, prolonged symmetric morning stiffness, fever) routes elsewhere first.
- Run the real treatment: strengthening 2–3×/week, weight where relevant, pacing, heat/ice, topical NSAIDs — six weeks honestly.
- If experimenting: curcumin or boswellia first by evidence (safety lists permitting), glucosamine sulfate the traditionalist's pick; one at a time, 12 weeks, scored (FAQ protocol); quit nulls without sentiment.
- Skip: blends, "cartilage rebuilders", and anything whose evidence section is a testimonial reel — the product-by-product findings live in joint supplement research.
- Escalating disease belongs with a clinician — injections, surgery timing and inflammatory-arthritis rule-outs are medical calls no shelf answers.
Frequently asked questions
Why is the placebo response so big in arthritis trials?
OA pain waxes and wanes on its own, people enroll (and buy supplements) during bad stretches that were due to ease anyway, and pain is attention-sensitive — expecting relief genuinely delivers some. GAIT's placebo arm hit ~60% response. None of this means pain isn't real; it means uncontrolled "it worked for me" cannot distinguish any capsule from a sugar pill, which is the entire case for the table.
What about collagen — it's literally what cartilage is made of?
So is steak, by that logic — swallowed collagen is digested to amino acids like any protein. Hydrolyzed-collagen and undenatured type-II (UC-II) trials exist and skew positive but are small, short and heavily manufacturer-funded; the "building blocks" mechanism is marketing, and any real effect likely runs through immune signaling (UC-II) at best. A protein-adequate diet already supplies the amino acids.
Is there anything for flare-ups specifically?
Supplements are the wrong tool for flares — trialled effects, where they exist, build over 8–12 weeks. Flares respond to short relative rest, ice, topical NSAIDs, and pacing the return; the plan lives in the osteoarthritis guide. A capsule bought during a flare is how the placebo story in question one gets written.
How should I run a fair trial on myself if I want to try one?
One ingredient at a time, at the studied dose (the ingredient pages list them), for 12 weeks; score pain 0–10 and a marker activity (stairs, opening jars) before and after; keep the rest of your routine steady; and obey the verdict. Responders can confirm with a two-week break. Multi-ingredient blends make this impossible by design, which tells you who they're designed for.
References
- NCCIH. Glucosamine and Chondroitin for Osteoarthritis (2023). https://www.nccih.nih.gov/health/glucosamine-and-chondroitin-for-osteoarthritis
- NCCIH. Turmeric (2020). https://www.nccih.nih.gov/health/turmeric
- National Institute of Arthritis and Musculoskeletal and Skin Diseases. Osteoarthritis (2023). https://www.niams.nih.gov/health-topics/osteoarthritis
- U.S. Food and Drug Administration. Dietary Supplements (2024). https://www.fda.gov/food/dietary-supplements
This page discusses supplement ingredients or categories. It may mention a commercial product, with disclosure. It is general health information, not personal medical advice. Read our editorial policy and medical review policy.