CBD and osteoarthritis: two placebo trials, no difference

Osteoarthritis is wear of the cartilage and of the bone just beneath it. It causes pain, stiffness and loss of mobility. The Inserm file puts its frequency at 3% of people under 45, 65% of those over 65 and 80% of those over 80. It involves the spine in 45 to 50% of cases, the hands in 35 to 45%, the knees in 30% and the hips in 10%. Inserm adds that no treatment can yet halt its progression.

The knee has its own figure. According to France's national health insurance, knee osteoarthritis affects 30% of people aged 65 to 75. Factors that favour it include age, excess weight, a sedentary lifestyle, family history and menopause.

The disease is common and no treatment changes its course, so wanting to try CBD is understandable. What matters is what was measured when CBD was compared with a placebo.

Three published trials have tested CBD in people with osteoarthritis. Two are randomised, double-blind and placebo-controlled, meaning neither patient nor doctor knows who receives what. The third has no control group.

The Danish trial analysed 129 of its 136 enrolled patients. At 12 weeks, pain had fallen by more than 30 mm out of 100 in 22% of patients on CBD and in 21% on placebo. The authors saw no effect on sleep, anxiety or depression.

The Austrian trial used a very high dose of 600 mg a day. That is 300 times the provisional reference of 2 mg a day that EFSA published in 2026 (see our article on EFSA's reference dose). The two groups did not differ on knee pain, stiffness or function. Liver enzymes did rise in 15 patients on CBD, compared with 5 on placebo, and returned to normal once the trial ended. What these enzymes measure is explained in our article on CBD and the liver.

Patients in the Austrian trial who received only a placebo saw their pain fall by 2.4 points. Follow-up, the natural ups and downs of the disease and the expectation of an effect account for that improvement. With no placebo group, it would have been credited to the product.

That is where the Australian trial of a CBD gel falls short. Pain dropped by about 1.9 points on a 0 to 10 scale, then climbed again during the period without gel. Since there was no control group, the authors themselves write that proving efficacy requires a randomised placebo-controlled trial.

Ask the same thing of any testimonial: what was the improvement compared with? Our review of what clinical studies really say covers these biases. Epidyolex, the only authorised CBD-based medicine, is prescribed for rare epilepsies; osteoarthritis is not among its indications.

A CBD balm or cream comes under Regulation (EC) No 1223/2009 on cosmetics. Cosmetics exist to clean, perfume, protect or maintain the skin, and may not claim to treat a disease. Regulation (EU) No 655/2013 also requires their claims to be truthful and backed by evidence. A label that promises action on osteoarthritis breaks these rules. Our article on permitted claims gives examples.

The EU Scientific Committee on Consumer Safety (SCCS) adopted its final opinion on CBD in cosmetics on 26 March 2026. It considers CBD safe up to 0.19% in products applied to the skin, and THC present as an impurity up to 0.00025%. The committee concedes that its data were limited. Its opinion has no force of law: the European Commission uses it as the basis when it amends the annexes of the cosmetics regulation.

From that level, the arithmetic is quick. At 0.19%, a 50 ml jar, about 50 g, holds about 95 mg of CBD. A 50 ml balm labelled 500 mg of CBD is dosed at 1%, five times the level judged safe.

Check five points on the pack and the certificate before you buy a balm:

• the milligrams of CBD and the jar volume, so you can work out the percentage;
• the INCI list, where CBD is listed as CANNABIDIOL;
• the batch number, which should match on the jar and on the certificate of analysis;
• the measured Δ⁹-THC content, with no HHC or H4CBD;
• any warming or cooling ingredient such as menthol or camphor, to be tried first on a small patch of skin.

These criteria are covered in our guide to assessing a cream or balm, and each line of the certificate is explained in the one on the certificate of analysis.

The HAS published a summary on physical activity in osteoarthritis in July 2022. In its words, the disease "requires a global approach", resting on regular physical activity and on weight loss combined with a balanced diet. The first-line treatment it presents is an adapted physical activity programme that mixes muscle strengthening, mobility and endurance. It judges that programme's benefit/risk ratio better than that of medicines.

The advice is concrete:

• sit for less than 7 hours a day, and get up to move for at least one minute every hour;
• choose sports such as swimming or cycling, and do your first runs on soft ground;
• learn a few simple exercises, using heat or cold before and after exercise if needed;
• give the joint relative rest during a flare, when it is swollen and warm.

Like any comfort gesture, a massage with a balm can be part of this routine. It does not replace exercise or the physiotherapist's advice.

The GP makes the diagnosis and refers you to a rheumatologist or a physiotherapist. If pain becomes chronic, the GP can also refer you to a specialised chronic pain unit, certified by the regional health agency and open on medical referral. AFLAR, the French association against rheumatic diseases, provides information to patients and their families.

Swallowed CBD interacts with several medicines, since the liver clears it using the same enzymes. The drug families involved are listed in our article on the grapefruit effect. In any case, CBD products meant to be swallowed have been off sale in France since 15 May 2026, for lack of authorisation under the Novel Food regulation. Precautions for older readers are gathered on our page about CBD and older people. The one on CBD and pain describes what the legal framework lets sellers say.