Condition deep-dive · 8 min read

Knee osteoarthritis — what to add, what to avoid

Updated 2026-05-14 · Reviewed by SupplementScore editors · No sponsorships

For knee osteoarthritis pain and function, the best-evidenced supplements are curcumin (bioavailable forms) and Boswellia serrata, with collagen peptides and undenatured type-II collagen as second-tier options. The glucosamine + chondroitin evidence has weakened since the GAIT trial — modest effect at best, and only in certain subgroups. None replace the largest interventions: weight loss in the overweight, quadriceps strengthening, and (when appropriate) intra-articular injections or surgery.

Read this first. Knee OA pain that is acute, severe, accompanied by mechanical locking, redness/warmth, or significant night pain warrants clinical evaluation rather than supplement self-management. Septic joint, gout, and meniscal injury are treatable differential diagnoses that supplements do nothing for.

The supplement layer with credible evidence

Tier 2 evidence · Pain and function · Strongest supplement signal

Curcumin (bioavailable form)

500–1,500 mg/day of a bioavailable formulation (Meriva, Theracurmin, BCM-95, or similar)

Meta-analyses of curcumin in knee OA (Daily 2016, Bannuru 2018) show pain and function improvements comparable to NSAIDs in some trials, with better tolerability for the GI tract. The catch is bioavailability — plain curcumin powder has very poor oral absorption. Formulations using phosphatidylcholine (Meriva), nanoparticles (Theracurmin), or essential-oil-blended curcuminoids (BCM-95) achieve 5–30× the systemic exposure. The clinical trials use these bioavailable forms; do not expect equivalent effect from generic "turmeric" capsules.

Tier 2 evidence · Pain and function

Boswellia serrata (5-LOX inhibitor)

100–250 mg/day standardised to AKBA (e.g., 5-Loxin, Aflapin)

Boswellia inhibits 5-lipoxygenase, reducing leukotriene-driven joint inflammation. Multiple RCTs at standardised doses (5-Loxin 100 mg/day, Aflapin 100 mg/day) show pain reduction and function improvement on WOMAC scores over 30–90 days. Boswellia + curcumin combinations have additive evidence in some trials. The standardised products are markedly more effective than generic Boswellia powders.

Tier 2 evidence · Function · Slower onset

Collagen peptides (hydrolyzed type II or generic peptides)

5–10 g/day hydrolyzed collagen, or 40 mg/day undenatured type-II collagen (UC-II)

Two distinct preparations with different mechanisms: hydrolyzed collagen peptides (5–10 g/day) appear to act as substrate for cartilage matrix; undenatured type-II collagen (UC-II, 40 mg/day) is hypothesized to act via oral tolerance and downregulation of immune-mediated cartilage attack. Both have small RCTs showing functional improvement in knee OA at 90+ days. Effect is slower than NSAIDs but generally well-tolerated.

Tier 3 evidence · Older but still relevant for some subgroups

Glucosamine sulfate (1,500 mg) + Chondroitin sulfate (1,200 mg)

Glucosamine sulfate 1,500 mg/day + Chondroitin sulfate 1,200 mg/day for 6+ months

The GAIT trial (2006) was negative for the combination versus celecoxib overall, but a moderate-to-severe knee OA subgroup showed modest benefit. Subsequent European RCTs of pharmaceutical-grade glucosamine sulfate (rather than glucosamine hydrochloride) have been more consistently positive. Effect is slow (3+ months). Reasonable to trial if other options are limited; do not expect dramatic relief. Use glucosamine sulfate, not hydrochloride, and pair with chondroitin in the doses above.

Tier 2 evidence · If deficient

Vitamin D3 (if 25-OH-D is low)

1,000–2,000 IU/day to maintain normal 25-OH-D

Low vitamin D correlates with worse OA outcomes observationally; trial evidence for OA-specific endpoints is modest. Worth testing and supplementing if deficient — most adults benefit anyway.

What to skip

MSM as monotherapy at low doses — modest signal exists for 3 g/day or higher, but the typical 500–1,000 mg/day on combination labels is subtherapeutic. If using MSM, use 3+ g/day.
SAMe as primary OA treatment — has small trial signal but is expensive, marginal, and the mood and methylation indications it's better-evidenced for are different conditions.
Hyaluronic acid oral supplements — intra-articular injection is the trial-supported delivery; oral hyaluronate has thin and inconsistent evidence.
"Joint complex" products bundling 12 ingredients at subtherapeutic doses — pay for individual ingredients at trial-tested doses.
Sea cucumber, shark cartilage, "deer antler velvet" — historically marketed for joint health with minimal clinical evidence; sourcing concerns and contamination.
Topical capsaicin without expectations — capsaicin does have signal for OA pain but is uncomfortable to use (initial burning), needs consistent multi-week application, and is not what most users mean by "supplement."

The non-supplement layer that matters more

Weight loss is the largest single lever for knee OA — every 1 kg lost reduces effective knee compressive load by ~4 kg during normal walking. Quadriceps strengthening and aerobic conditioning produce real pain and function improvements. Bracing, off-loader insoles, and (for some patients) intra-articular corticosteroid or hyaluronic acid injections sit alongside supplements. Total knee replacement is the definitive intervention for end-stage OA with quality-of-life impact.

Practical quick-start. Bioavailable curcumin (Meriva 500 mg twice daily or Theracurmin 90 mg three times daily) PLUS Boswellia serrata standardised extract (5-Loxin 100 mg/day or Aflapin 100 mg/day). Trial for 8–12 weeks. Add hydrolyzed collagen peptides 10 g/day if function-limited. Skip the multi-ingredient "joint complex" products. Address weight and quadriceps strength in parallel — those produce larger improvements than any supplement.