What does the research say?
In a 12-week study of 136 people with hand osteoarthritis or joint inflammation linked to psoriasis (psoriatic arthritis), there was almost no difference in pain between those who took CBD and those who took placebo; no difference was seen in sleep, anxiety or depression either.
In 86 people with knee osteoarthritis who took acetaminophen (paracetamol) regularly, high-dose CBD was compared with placebo for eight weeks. The reduction in pain was almost the same in both groups; however, side effects and raised liver enzymes were more common in the CBD group. A third study, using a CBD-rich oil in knee osteoarthritis, also found no difference after 60 days.
In a small 18-person study of CBD applied to the skin in arthritis at the base of the thumb, improvements in pain and hand function were reported. In a study of CBD gel in hand osteoarthritis with no comparison group, pain also decreased, but it returned to its earlier level when the gel was stopped. There are few controlled studies of products containing THC. In 58 people with rheumatoid arthritis, a mouth spray containing THC and CBD was compared with placebo for five weeks; a small improvement was seen in pain on movement and at rest, in sleep and in disease activity. It is unclear how noticeable the difference of about 1 point (out of 10) in pain is in daily life; side effects such as dizziness and dry mouth were more common. Cochrane and other reviews found the evidence, which rests on this single study, weak. In two laboratory studies of 37 and 21 people with knee osteoarthritis, a single dose of a THC medicine (dronabinol) did not reduce joint pain more than placebo; giving it together with an opioid pain reliever did not add any benefit either. Studies testing THC for weeks in osteoarthritis are ongoing.
Hundreds of patients in the UK and Australia with osteoarthritis or inflammatory joint disease (rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis) who used a prescription cannabis product were followed; improvements from baseline in pain, sleep and quality of life were reported. However, these studies had no comparison group; some of the registries belonged to private clinics that prescribed the product, and in the inflammatory joint disease registry more than half of the patients were already using cannabis before treatment. Because 35–65% of chronic pain patients given placebo in controlled studies also report clear improvement, these results count as weak evidence of effectiveness. In a California study without placebo of 164 people (100 with rheumatoid arthritis or osteoarthritis) comparing three different cannabinoid capsules, symptoms also improved from baseline; but it is not known how much of the improvement came from the product.
Strength of evidence: ●●●○ Moderate evidence. Finding: No effect found. What does this mean?
Ongoing research on this topic
Trials testing CBD for this topic on ClinicalTrials.gov: 9 under way (7 recruiting); results are awaited from 1 that ended in the last two years. A trial being under way does not show that CBD works.
- NCT06787118: Phase 3, not yet recruiting, target 130 participants
- NCT05020028: Phase 2/3, recruiting, target 100 participants
- NCT06234631: Phase 2, recruiting, target 380 participants
Registries in other countries also list 3 recruiting trials on this topic: other registries.
Possible side effects and drug interactions
In the knee osteoarthritis study, side effects were more common in those taking high-dose CBD, and raised liver enzymes were seen in 15 people (5 people in the placebo group). These patients were also taking acetaminophen (paracetamol) regularly; for this reason, monitoring liver enzymes is recommended when acetaminophen and CBD are used together.
CBD can slow the breakdown of some medicines in the body. A case has been reported in which the blood-clotting value (INR) rose with the blood thinner warfarin; the effect of clopidogrel may be reduced. No significant documented interaction with anti-inflammatory pain relievers has been reported. Taken together with medicines that cause sleepiness, it can increase drowsiness.
With products containing THC: in rheumatoid arthritis, side effects such as dizziness, dry mouth and lightheadedness were more common than with placebo in people using a mouth spray containing THC and CBD. In the single-dose studies in knee osteoarthritis, the THC medicine dronabinol increased the feeling of being high; when it was given together with an opioid pain reliever, moderate side effects in one study, and nausea in the other, were more common than with the opioid alone.
Interactions with medicines used for this condition
Interactions reported for these medicines in peer-reviewed publications and official prescribing information. A medicine that is not listed can still interact with CBD; tell your doctor about every medicine you take.
| Medicine | Importance | Warning | Source |
|---|---|---|---|
| Other opioids (morphine, oxycodone, fentanyl, tramadol, etc.), Methadone | Caution | Methadone and other opioids: In a child, a high methadone level was found and drowsiness developed while CBD was being used; two weeks after CBD was stopped, the level fell to less than half and the symptoms resolved. In small controlled studies with morphine and buprenorphine, no clear change in drug levels was seen; with a single intravenous dose of fentanyl, there was no respiratory depression. Drowsiness may increase together with opioids. Tell the doctor who manages your pain or addiction treatment. | PMID 32444381, PMID 41622700, PMID 41467610, PMID 25748562, FDA Epidiolex label §7.4 (05/2026) |
| Gabapentin (Gabateva, Neruda), Pregabalin (Lyrica) | Caution | Benzodiazepines, sleep medicines and sedatives: Drowsiness and sedation may increase together with CBD. Talk to your doctor before using them together. | DailyMed Epidiolex label 7.4, EMA Epidyolex SmPC 4.4 |
| Acetaminophen (paracetamol) | Caution | Acetaminophen and the liver: In a knee osteoarthritis study, liver enzyme elevations were more common in people taking 600 mg of CBD together with 3 g a day of acetaminophen (paracetamol) than with placebo (15 vs. 5 patients). Enzyme elevations were also seen in healthy adults at a lower CBD dose (5 mg/kg a day). If you take acetaminophen regularly, talk with your doctor about liver tests. | PMID 38033459, PMID 40622698 |
| Tizanidine | Caution | Theophylline, tizanidine: CBD can raise the blood levels of these medicines; the prescribing information for approved CBD recommends considering a dose reduction. The tizanidine prescribing information itself asks that it not be combined with weak CYP1A2 inhibitors such as CBD; low blood pressure, a slowed heart rate or excessive drowsiness may occur. Tell your doctor. | FDA Epidiolex label §7.2 (05/2026), EMA Epidyolex SmPC 4.5, PMID 33951339, Tizanidine prescribing information §7.2, dailymed.nlm.nih.gov |
| NSAIDs (ibuprofen, naproxen, celecoxib, etc.) | Information | NSAID pain relievers: There is no documented meaningful interaction between CBD and these pain relievers; there are no direct studies either. The bleeding and kidney risks come from the NSAID itself. | FDA Epidiolex label §12.3 (05/2026), Celecoxib prescribing information §7, dailymed.nlm.nih.gov |
To check interactions with the medicines you take: CBD drug interaction checker.
How was CBD used in studies?
Studies of CBD by mouth used 20–30 mg a day (study dose; not a personal recommendation), 45 mg a day (study dose; not a personal recommendation) and 600 mg a day (study dose; not a personal recommendation); in none of them did pain decrease more than with placebo.
Studies of products applied to the skin used 1 mL twice a day of an oil containing 6.2 mg of CBD per milliliter (study dose; not a personal recommendation) and a 4% CBD gel three times a day (study dose; not a personal recommendation).
In the THC studies, a mouth spray was used in the evenings in rheumatoid arthritis; in the last week, the average was about 14.6 mg of THC and 13.5 mg of CBD a day (study dose; not a personal recommendation). In the laboratory studies in knee osteoarthritis, a single dose of 10 mg of dronabinol was given (study dose; not a personal recommendation).
Study details (for health professionals)
Three RCTs found no superiority of oral CBD over placebo: hand osteoarthritis/psoriatic arthritis (20–30 mg, 12 weeks), knee osteoarthritis (600 mg/day, 8 weeks) and knee osteoarthritis (45 mg/day CBD-rich oil, 60 days). In the 600 mg/day study, side effects and raised liver enzymes were more common in the CBD group. Small studies of topical products show conflicting results (see topical CBD). With products containing THC: in rheumatoid arthritis, a single RCT of a THC:CBD mouth spray (58 people, 5 weeks) showed a small improvement in pain and DAS28, with more frequent nervous system side effects (Cochrane: weak evidence, potential harms outweigh the small benefit; GRADE: low certainty; the reviews’ risk-of-bias assessments disagree); in knee osteoarthritis, a single 10 mg dose of dronabinol did not reduce clinical pain more than placebo in two crossover laboratory RCTs (37 and 21 people). A randomized study without placebo comparing three cannabinoid capsules (Kruger 2026) and patient registries (UK, Australia) reported improvement from baseline; with no comparison group, they do not count as evidence of effectiveness. October 2026 search (PubMed, Cochrane, ClinicalTrials.gov, Europe PMC): 8 randomized studies of cannabinoids in this area (7 placebo-controlled), 1 open-label study, 5 reviews (one Cochrane) and 3 observational studies were found; all are in the table below. Not included in the table: a skin gel study announced only as a conference abstract, a non-peer-reviewed 12-person preprint (oral CBD in knee osteoarthritis), completed studies with unpublished results (see unverifiable items) and studies with hemp seed oil (seed oil is not a CBD or THC product; these studies did not report cannabinoid content).
Evidence assessment: Moderate (pointing to no benefit for oral CBD)
CBD alone
Oral pure CBD did not reduce pain more than placebo in osteoarthritis and hand arthritis; at high doses, raised liver enzymes were more common. There is no published study comparing pure CBD with placebo in rheumatoid arthritis or ankylosing spondylitis (October 2026); the results of a completed study in rheumatoid arthritis (NCT04911127) have not yet been published.
Products containing THC
There is a single controlled study lasting several weeks of products containing THC in this area: in rheumatoid arthritis, a mouth spray containing THC and CBD showed slightly more improvement in pain and disease activity than placebo over 5 weeks, but nervous system side effects such as dizziness were also more common (58 people; low-certainty evidence). The Cochrane review wrote that the potential harms seem to outweigh this small benefit. In knee osteoarthritis, a single dose of dronabinol (synthetic THC) did not reduce joint pain more than placebo in two laboratory studies. In patient registries, people using THC-containing products reported improvement in pain, but there was no comparison group. Controlled studies testing THC for weeks in osteoarthritis are ongoing (NCT06878417, NCT04992624).
Evidence: Weak (a single small RCT in rheumatoid arthritis; in knee osteoarthritis, only single-dose laboratory studies)
Dose range in studies
Oral CBD 20–30 mg/day, 45 mg/day and 600 mg/day; topical 6.2 mg/mL (1 mL twice a day) and 4% gel; with THC: in RA, THC:CBD mouth spray averaging ~14.6 mg THC + 13.5 mg CBD/day; in knee OA, a single 10 mg dose of dronabinol
Starting dose and titration
Vela 2022: 10 → 20 → 30 mg/day; Pramhas 2023: 600 mg/day in three equal steps during the first week; Mojoli 2025: fixed 45 mg/day; Blake 2006: started with 1 spray in the evening and increased every 2 days (up to 6)
These are doses researchers gave under medical supervision; they are not a recommended dose for you. "mg/kg/day" means milligrams per kilogram of body weight per day.
| Study | Participants | Product | Dose and titration | Duration | Result | Source |
|---|---|---|---|---|---|---|
| Vela 2022 Double-blind, placebo-controlled RCT | 136 (129 analyzed) Hand osteoarthritis and psoriatic arthritis | Synthetic CBD, oral; no THC | Started at 10 mg/day; 20 mg/day at week 2; 30 mg/day at week 4 if no response Stepwise: 10 → 20 → 30 mg/day | 12 weeks | Difference in pain 0.23 mm (0–100 scale; 95% CI −9.4 to 9.9); no difference in sleep, anxiety, depression or pain catastrophizing either Side effects: Adverse events in 33 patients in the CBD group and 26 with placebo (119 events in total); serious events 2 vs. 2, none attributed to the drug | PMID 34510141, doi:10.1097/j.pain.0000000000002466, NCT03693833 |
| Pramhas 2023 Double-blind, placebo-controlled, parallel-group RCT | 86 randomized (43/43); 84 received the drug, 58 completed; analysis with all randomized patients (missing data imputed) Painful knee osteoarthritis (in addition to 3 g of acetaminophen [paracetamol] a day) | Hemp-derived CBD capsules (purity >99.8%), 200 mg/capsule | 600 mg/day (200 mg 3 times a day, with food) Increased to 600 mg/day in three equal steps during the first week | 8 weeks | Reduction in WOMAC pain (0–10) 2.5 vs. 2.4; difference 0.1 (95% CI −0.8 to 1.0; p=0.80). The study was funded by the company that makes the CBD capsules Side effects: Adverse events more frequent with CBD (135 vs. 105, p=0.008); transaminase/GGT elevation 15 vs. 5 (p=0.02) | PMID 38033459, doi:10.1016/j.lanepe.2023.100777, NCT04607603, PMID 38322713 |
| Mojoli 2025 (CANOA) Double-blind, placebo-controlled RCT | 45 randomized (43 analyzed) Knee osteoarthritis | Full-spectrum cannabis oil (in MCT), CBD 22.5 mg/mL; THC below the limit of detection (<0.075 mg/mL) | 45 mg CBD/day (morning and evening) None (fixed dose) | 60 days | No difference in WOMAC or VAS Side effects: No serious adverse events; no significant change in serum biomarkers | PMID 41487518, doi:10.3389/fphar.2025.1657065, NCT06588972 |
| Heineman 2022 Crossover, double-blind RCT | 18 Arthritis of the thumb base joint | Topical CBD 6.2 mg/mL in shea butter (control: shea butter alone) | 1 mL twice a day | 2 weeks + 1-week washout + 2 weeks crossover | During the CBD period, greater improvement in VAS pain, DASH and SANE scores than in the control period; grip and pinch strength and range of motion similar. Numbers not in the abstract (full text could not be read); each period 2 weeks, 18 people Side effects: No adverse events | PMID 35637038, doi:10.1016/j.jhsa.2022.03.002, NCT04611347, 2025, the primary measures in the registry are grip strength, PROMIS upper extremity, Kapandji and range of motion, not the pain scales highlighted in the article) |
| Bawa 2024 Open-label, single-arm, feasibility | 18 (15 completed) Hand osteoarthritis | Transdermal CBD gel 4% | 0.25 mL 3 times a day (~30 mg CBD/day) | 4 weeks | About −1.9 NRS in pain (no control group); returned to baseline during the washout period. The study was co-funded by the company that developed the gel, which also donated the product; four of the authors are employees of this company Side effects: 31 reports of mild adverse events and 1 moderate; no THC detected in urine | PMID 38783008, doi:10.1038/s41598-024-62428-x, ANZCTR ACTRN12621001512819 |
| Blake 2006 Double-blind, placebo-controlled, parallel-group RCT | 58 randomized (31 active drug, 27 placebo); 75 people screened Active rheumatoid arthritis; disease-modifying antirheumatic drugs (DMARDs) continued at a stable dose (United Kingdom) | Mouth spray containing THC and CBD (nabiximols; Sativex); 2.7 mg THC + 2.5 mg CBD per spray | Evenings only; mean 5.4 sprays a day in the last week (about 14.6 mg THC + 13.5 mg CBD/day) Started with 1 spray in the evening and increased every 2 days (up to 6); then 3 weeks at a fixed dose | 5 weeks | Compared with placebo, pain on movement (0–10) improved by −0.95 more (p = 0.044), pain at rest by −1.04 (p = 0.018), sleep quality by −1.17 (p = 0.027) and disease activity (DAS28) by −0.76 (p = 0.002); no difference in morning stiffness (baseline scores were low). The authors wrote that the differences were small and varied from person to person; the most recent review (Zintziovas 2026) found the clinical importance of the pain difference of about 1 point unclear. A single, small, short study in rheumatoid arthritis; Cochrane (Richards 2012) found the evidence weak, and the GRADE summary (Schulze-Schiappacasse 2022) rated it low certainty for pain and DAS28. Risk-of-bias assessments conflict: Cochrane, Fitzcharles 2016 and Johal 2020 found it high, the GRADE summary unclear for blinding, and Zintziovas 2026 (RoB 2) low; in Johal 2020’s recalculation, the pain difference was not significant Side effects: Mostly mild to moderate; no serious side effects or discontinuations due to side effects in the active drug group. By Cochrane’s calculation, the risk of side effects was higher than with placebo (RR 1.82; 95% CI 1.10–3.00); most often dizziness 26%, dry mouth 13%, lightheadedness 10% | PMID 16282192, doi:10.1093/rheumatology/kei183, Letter: Wright 2006, PMID 16621924 |
| Campbell 2023 Double-blind, placebo-controlled, within-subject crossover laboratory RCT (a single dose in each session) | 37 Knee osteoarthritis, mean pain at least 3/10 (USA; mean age 62, 65% women) | Dronabinol (synthetic THC) and hydromorphone (an opioid pain reliever), oral capsules; alone, together and placebo | Single dose: dronabinol 10 mg; hydromorphone 4 mg None (the first session was always hydromorphone for safety; the order of the next three sessions was random) | 4 sessions at least 7 days apart; measurements for several hours after the dose in each session | No difference in knee pain intensity (0–100) among the four conditions (peak value: placebo 12.6, dronabinol 13.5, hydromorphone 9.3, both together 11.0; p = 0.302); no difference in physical function tests either. In experimental pain tests, dronabinol did not clearly enhance the effect of hydromorphone. Single-dose laboratory study; does not reflect long-term use Side effects: Drug-related side effects in 52 of 148 sessions (35.1%) and in 26 of 37 people; no serious side effects. Hydromorphone + dronabinol increased moderate side effects compared with placebo and with hydromorphone alone; dronabinol raised 'bad drug effect' and 'high' scores above placebo; only hydromorphone impaired cognitive function | PMID 37202479, doi:10.1038/s41386-023-01597-1, NCT03098563, 2024, 58 people together with the Hamilton 2026 cohort, in the combined registry result, peak clinical pain (0–100) was 12.8 with dronabinol and 12.8 with placebo) |
| Hamilton 2026 Double-blind, placebo-controlled, within-subject crossover laboratory RCT (a single dose in each session) | 21 (different people from Campbell 2023) Knee osteoarthritis (USA, 2021–2023; mean age 63, 57% women) | Dronabinol (synthetic THC) and low-dose hydromorphone (an opioid pain reliever), oral capsules; alone, together and placebo | Single dose: dronabinol 10 mg; hydromorphone 2 mg None (the first session was always hydromorphone for safety; the order of the next three sessions was random) | 4 sessions at least 7 days apart; 4 hours of measurement after the dose | No difference in knee pain intensity (0–100) (placebo 13.1, dronabinol 12.8, hydromorphone 9.5, both together 15.0; p = 0.280); no difference in walking, timed up-and-go or stair tests. In experimental pressure pain, hydromorphone was superior to dronabinol; no added benefit of giving them together was seen. Single-dose laboratory study; does not reflect long-term use Side effects: Drug-related side effects in 24 of 84 sessions (28.6%); no difference between conditions (p = 0.082); no serious side effects. Dronabinol increased the 'high' score more than hydromorphone; when given together, working memory reaction time lengthened and nausea increased compared with hydromorphone | PMID 41481837, doi:10.1097/ALN.0000000000005925, NCT03098563, the results in the registry combine both cohorts) |
| Kruger 2026 Randomized study without placebo comparing three cannabinoid formulations (a real-world data study, according to the authors); blinding not stated in the abstract | 276 enrolled; 164 people who completed the questionnaires and continued the product were analyzed (25 rheumatoid arthritis, 75 knee/hip osteoarthritis, 64 fibromyalgia) Adults living in California diagnosed with rheumatoid arthritis, knee/hip osteoarthritis or fibromyalgia (USA) | Oral capsules, three formulations: 1) 12.5 mg CBD + 12.5 mg THC; 2) 10 mg THCA + 10 mg CBDA (the unheated acid forms of THC and CBD) + 5 mg CBG + 3 mg CBC; 3) 10 mg CBD + 10 mg CBDA | The amounts above per capsule; number of capsules per day not reported (in the abstract) Not reported (in the abstract) | 12 weeks | In the per-protocol analysis, improvement from baseline in all symptoms except cognitive function (small to large effect); most effects did not differ between products or pain types (rheumatoid arthritis, osteoarthritis, fibromyalgia). Because there was no placebo group, it is not known how much of the improvement came from the products; 39% of those enrolled did not complete the questionnaires; arthritis-specific numbers are not in the abstract; four of the six authors are affiliated with a company (MoreBetter), and the authors declared no conflicts of interest Side effects: Not reported (in the abstract) | PMID 42140793, doi:10.1016/j.clinthera.2026.04.018 |
| Richards 2012 (Cochrane) Cochrane systematic review and meta-analysis (neuromodulators in rheumatoid arthritis; RCTs; search through November 2010) | 4 RCTs; 1 RCT with a cannabinoid (58 people) Adults with rheumatoid arthritis | In the cannabinoid arm, only a mouth spray containing THC and CBD (Blake 2006); the other drugs were nefopam and capsaicin | As in Blake 2006 (mouth spray in the evening) | Over 5 weeks, the mouth spray was superior to placebo on the 'current pain' score by a small but significant margin (mean difference −0.72; 95% CI −1.31 to −0.13). The authors found the evidence weak (a single, small, low-quality study) and wrote that the potential harms seem to outweigh the small benefit. As of October 2026, there is no update on PubMed Side effects: Risk of side effects higher than with placebo: RR 1.82 (95% CI 1.10–3.00); number needed to harm 3; most often dizziness 26%, dry mouth 13%, lightheadedness 10% | PMID 22258992, doi:10.1002/14651858.CD008921.pub2, The same team’s summary on inflammatory joint diseases (J Rheumatol Suppl 2012, PMID 22942326 | |
| Schulze-Schiappacasse 2022 Structured evidence summary of systematic reviews (Epistemonikos, GRADE) | 26 systematic reviews; all rely on the same single RCT (58 people) Rheumatoid arthritis | Cannabis, cannabis products and synthetic cannabinoids; the only eligible study was a mouth spray containing THC and CBD | Mean 14.6 mg THC + 13.5 mg CBD/day (Blake 2006) | Disease activity (DAS28) 0.9 points lower than placebo (low certainty); current pain score (0–5) 0.7 lower (low certainty); the authors interpreted this as 'little or no difference' in pain. Certainty was downgraded because the risk of bias from blinding was unclear and the confidence interval was wide. It was emphasized that the evidence rests on a single 5-week study Side effects: Nervous system side effects (dizziness, lightheadedness, sleepiness, headache) 419 vs. 148 per 1,000 people (RR 2.83; 95% CI 1.05–7.65; low certainty); the evidence for serious side effects is very uncertain | PMID 33859125, doi:10.1097/RHU.0000000000001745 | |
| Zintziovas 2026 Scoping review (RCTs and observational studies; assessed with RoB 2 and ROBINS-I; search through September 10, 2026) | 593 records; 3 studies (1 RCT, 2 observational) Adults with rheumatoid arthritis | Cannabinoid products (in the RCT, a mouth spray containing THC and CBD; in the observational studies, CBD or cannabis that people used on their own) | Variable | The only RCT of a cannabinoid in rheumatoid arthritis is still Blake 2006; the authors rated it at low risk of bias with RoB 2 (Cochrane and other reviews had found high risk), but found the clinical importance of the pain difference of about 1 point unclear. The risk of bias of the two observational studies is serious or critical. According to the authors, current evidence does not support routine use of cannabinoids for rheumatoid arthritis pain; completed but unpublished studies (e.g., NCT04911127) could not contribute to the evidence Side effects: No long-term safety data | PMID 42831900, doi:10.1007/s00296-026-06314-x | |
| Dubois 2024 Systematic review (2012–2022; GRADE and ROBINS-I; narrative synthesis) | 7 studies (2 RCTs, 5 observational) Adults with osteoarthritis using medical cannabis | Medical cannabis and cannabinoid products | Variable | Of the two RCTs, only one reported an improvement in pain; all 5 observational studies reported improvements in pain, opioid use or function. Risk of bias high, study quality low; the authors found the evidence supporting the use of medical cannabis in place of first-line drug treatment for osteoarthritis to be of low quality Side effects: Not reported (in the abstract) | PMID 38832841, doi:10.1080/03007995.2024.2363945, PROSPERO CRD42022354026 | |
| Abafita 2026 Systematic review and meta-analysis (animal and human studies; RoB 2, GRADE; search through April 2024) | 6 human studies (5 RCTs, 1 open-label), 606 patients; pain meta-analysis 4 RCTs Knee and hand osteoarthritis (Denmark, Austria, USA, Australia) | CBD (oral and on the skin) and dronabinol (synthetic THC) | Variable | In the pooled analysis of four RCTs, no significant improvement in pain compared with placebo (standardized mean difference −1.52; 95% CI −4.63 to 1.59). The authors considered the two small hand studies reporting positive results limited because of small samples, lack of a comparison group and high risk of bias. Certainty of evidence low to moderate; positive findings in animal studies were not confirmed in humans Side effects: Side effects generally mild to moderate (nausea, dry mouth, dizziness, sleepiness); insufficient data on long-term safety | PMID 42515994, doi:10.1155/prm/3998239, PROSPERO CRD42024539941 | |
| Sajeev 2026 (UK Medical Cannabis Registry) Prospective patient registry case series (no comparison group) | 192 (patients in the registry for at least 24 months) Chronic pain from inflammatory joint disease (an umbrella term covering rheumatoid arthritis, psoriatic arthritis and ankylosing spondylitis; distribution by disease not reported) (United Kingdom; mean age 47; 54% were already using cannabis at baseline) | Prescription cannabis products (oil, dried flower or both; containing THC and CBD); at month 24, the daily THC dose was about seven times that of CBD | At month 24, median 146 mg THC and 21 mg CBD/day Median 20 mg THC/day at baseline; increased from the first month on | 24 months | Pain intensity (BPI, 0–10) fell from 5.95 to 3.98 at month 24; pain, sleep and quality of life were better than baseline at all follow-up points (p < 0.01); anxiety improved only in the first 3 months. At month 24, 62.5% of patients reported a clinically meaningful improvement in pain intensity (in the RCTs the authors cite, 35–65% of chronic pain patients given placebo also report such an improvement). No comparison group and no blinding, and missing questionnaires were filled in statistically; the registry belongs to the private clinic that prescribes the product, and seven of the nine authors are affiliated with this clinic (one is its research director) Side effects: 296 side effects in 27 patients (14.1%): 42.6% mild, 44.6% moderate, 12.8% severe; none life-threatening | PMID 42724108, doi:10.1177/11795441261480709, Earlier analysis of the same registry: Francis 2025, PMID 38976497, 12 months, the patients most likely overlap) |
| Francis 2024 (UK Medical Cannabis Registry) Prospective patient registry case series (no comparison group) | 77 Chronic pain from osteoarthritis (United Kingdom) | Prescription cannabis products (containing THC and/or CBD; details not in the abstract) | Not reported (in the abstract) Not reported (in the abstract) | 12 months | Pain intensity and the impact of pain on daily life (BPI) and the McGill pain score were better than baseline at all follow-up points (months 1, 3, 6 and 12) (p = 0.004, 0.005 and 0.017); improvements in quality of life, sleep and anxiety were limited to the first 3–6 months. No comparison group; the authors wrote that causality cannot be inferred; the registry belongs to the private clinic that prescribes the product, and seven of the eight authors are affiliated with this clinic Side effects: 218 side effects in 17 patients (22.1%): 76 mild, 104 moderate, 38 severe | PMID 38669060, doi:10.1080/15360288.2024.2340076 |
| Schubert 2023 Interim analysis of a prospective observational study (no comparison group; arthritis subgroup) | Arthritis subgroup 199 (part of a refractory chronic pain cohort) Patients prescribed medical cannabis for refractory chronic pain; arthritis subgroup (Australia) | Prescription medical cannabis: CBD-only, balanced CBD:THC and THC-rich products | Variable (grouped by THC:CBD ratio) Not reported (in the abstract) | Not reported (in the abstract) | In the arthritis subgroup, pain intensity decreased from baseline (p = 0.005); it also decreased in those using CBD-only (p = 0.018) and balanced CBD:THC products (p = 0.005). In the whole cohort, 22% of patients reported a clinically meaningful reduction in pain. No comparison group; some of the authors are from a private cannabis research company Side effects: 51% of the whole chronic pain cohort (364 people) reported at least one side effect: most often dry mouth 24%, sleepiness 19%, tiredness 12%; similar in the arthritis subgroup | PMID 37057257, doi:10.1177/20494637221147115 |
Reported side effects
- In knee OA with 600 mg/day, adverse events more frequent (135 vs. 105) and transaminase/GGT elevation 15 vs. 5 (Pramhas 2023)
- THC:CBD mouth spray in RA (5 weeks): side effects RR 1.82 (95% CI 1.10–3.00), NNTH 3; dizziness 26%, dry mouth 13%, lightheadedness 10%; no serious side effects or discontinuations due to side effects in the active group (Blake 2006; Richards 2012 Cochrane)
Monitoring done in studies
- Epidiolex label: at 10–20 mg/kg/day, ALT >3×ULN in 13% (placebo 1%); risk rises markedly with valproate; monitor at baseline, at months 1, 3 and 6, and after dose changes
- In healthy adults, 5 mg/kg/day led to ALT/AST >3×ULN in 5.6% (placebo 0%) and eosinophilia (Florian 2025, PMID 40622698); 1,500 mg/day led to >5×ULN in 31% (Watkins 2021, PMID 33022751)
- Liver enzymes should be monitored with acetaminophen (paracetamol) and other medicines that strain the liver (Pramhas 2023)
Who should be careful
- Liver enzymes should be monitored with high-dose CBD taken together with acetaminophen (paracetamol) (Pramhas 2023)
- CBD is not an approved medicine for any type of pain
Drug interactions
- Opioids: no increase in side effects with fentanyl + 400–800 mg CBD; 'bad drug effect' increased with hydromorphone + 200 mg CBD; pain threshold decreased with methadone + 1,200 mg CBD. In knee OA, hydromorphone + a single 10 mg dose of dronabinol: no additional analgesia; the feeling of being high increased, and moderate side effects (Campbell 2023) and nausea (Hamilton 2026; compared with hydromorphone alone) were more frequent. An opioid-sparing effect has not been proven; monitor for sedation
- CNS depressants (gabapentin, pregabalin, benzodiazepines, alcohol): additive sedation; somnolence with Epidiolex 32% vs. 11%. No direct clinical study of CBD + gabapentinoids has been reported
- NSAIDs: the label shows no significant change in tolbutamide, a CYP2C9 substrate; no direct study of CBD + NSAIDs has been reported; no documented significant interaction
- Acetaminophen (paracetamol): in knee OA, enzyme elevations were more frequent with acetaminophen 3 g + CBD 600 mg; monitor liver enzymes
- Warfarin: in a case report, INR rose and the warfarin dose was reduced; monitor INR when starting or stopping CBD
- Clopidogrel: antiplatelet effect may be reduced via CYP2C19 (label); consult cardiology
- DOACs (apixaban, rivaroxaban): no direct data reported; uncertain, should be disclosed and monitored
- Valproate (liver, high risk) and clobazam (sedation 46% vs. 16%)
- Tizanidine, theophylline (CYP1A2): the label recommends considering a dose reduction
- Tacrolimus, sirolimus, everolimus, digoxin: monitoring or dose reduction (label)
- Strong inducers such as rifampin: CBD levels fall by ~32%
- Note: label data come from high-dose pharmaceutical CBD; with nonprescription products the size of an interaction may differ, but its direction should be assumed to be the same
Abbreviations: RCT, randomized controlled trial (a study in which participants are assigned to groups by chance); placebo, an inactive substance; SMD, OR and RR, effect measures used in analyses that pool studies; CYP, liver enzymes that break down medicines.
If you are considering CBD
- You can talk with your doctor about the treatment options that suit your joint pain.
- If you take acetaminophen (paracetamol) regularly or other medicines that affect the liver, talk with your doctor about the fact that CBD can affect liver enzymes.
- If you take a blood thinner, be sure to mention it.
- If you are considering an oil, cream or gel, check the product’s contents and its independent certificate of analysis.
Choose your age group, medicines and situation to prepare a short summary you can take to your doctor. Nothing you enter is sent anywhere.
Bottom line
Moderate-strength evidence shows that CBD taken by mouth does not reduce pain more than an inactive substance in osteoarthritis and hand arthritis; at high doses, side effects increased. There are small positive signals for products applied to the skin, but it is still too early to draw a firm conclusion. The evidence for products containing THC is weak: in rheumatoid arthritis, a single 58-person study showed a small benefit and more frequent side effects; in knee osteoarthritis, pain did not decrease in two single-dose laboratory studies.
Frequently asked questions
Does CBD oil work for arthritis and osteoarthritis?
CBD taken by mouth was tested in three well-designed studies, and in none of them did it reduce pain more than an inactive substance (placebo). The evidence on this is moderate. At high doses, side effects and raised liver enzymes were more common. A mouth spray containing THC and CBD reduced pain slightly in a single small study in rheumatoid arthritis; but this evidence is weak and side effects were more common.
Does CBD cream help joint pain?
A small 18-person study in arthritis at the base of the thumb reported a positive result; in an uncontrolled study of CBD gel in hand osteoarthritis, pain also decreased. However, these studies are very small, and the results need to be confirmed by larger studies.
Can I take CBD while using acetaminophen (paracetamol)?
In the knee osteoarthritis study, raised liver enzymes were more common in people taking high-dose CBD together with regular acetaminophen (paracetamol). For this reason, if using them together is being considered, monitoring liver enzymes is recommended; it’s important to talk with your doctor.
Is a higher dose of CBD more effective?
No study shows this. In knee osteoarthritis, 600 mg of CBD a day (study dose; not a personal recommendation) did not reduce pain more than placebo; side effects, however, were more common.
Can CBD oil be obtained with a prescription for arthritis and osteoarthritis in Turkey?
In Turkey, cannabis-derived products are governed by the Kenevirden Elde Edilen Ürünlere Dair Yönetmelik (Regulation on Products Derived from Hemp) dated January 31, 2026. These products, including CBD, can be sold only in pharmacies. Those classed as medicines and health products are dispensed with a prescription; for supplement products and cosmetics with a THC content of no more than 0.3%, the regulation does not state a prescription requirement. As of October 2026, there is no approved (licensed) CBD medicine in Turkey. If you are thinking of using it, talk with your doctor first.
References
Show 17 references
- Vela 2022. Double-blind, placebo-controlled RCT. PMID 34510141, doi:10.1097/j.pain.0000000000002466, NCT03693833
- Pramhas 2023. Double-blind, placebo-controlled, parallel-group RCT. PMID 38033459, doi:10.1016/j.lanepe.2023.100777, NCT04607603, PMID 38322713
- Mojoli 2025 (CANOA). Double-blind, placebo-controlled RCT. PMID 41487518, doi:10.3389/fphar.2025.1657065, NCT06588972
- Heineman 2022. Crossover, double-blind RCT. PMID 35637038, doi:10.1016/j.jhsa.2022.03.002, NCT04611347, 2025, the primary measures in the registry are grip strength, PROMIS upper extremity, Kapandji and range of motion, not the pain scales highlighted in the article)
- Bawa 2024. Open-label, single-arm, feasibility. PMID 38783008, doi:10.1038/s41598-024-62428-x, ANZCTR ACTRN12621001512819
- Blake 2006. Double-blind, placebo-controlled, parallel-group RCT. PMID 16282192, doi:10.1093/rheumatology/kei183, Letter: Wright 2006, PMID 16621924
- Campbell 2023. Double-blind, placebo-controlled, within-subject crossover laboratory RCT (a single dose in each session). PMID 37202479, doi:10.1038/s41386-023-01597-1, NCT03098563, 2024, 58 people together with the Hamilton 2026 cohort, in the combined registry result, peak clinical pain (0–100) was 12.8 with dronabinol and 12.8 with placebo)
- Hamilton 2026. Double-blind, placebo-controlled, within-subject crossover laboratory RCT (a single dose in each session). PMID 41481837, doi:10.1097/ALN.0000000000005925, NCT03098563, the results in the registry combine both cohorts)
- Kruger 2026. Randomized study without placebo comparing three cannabinoid formulations (a real-world data study, according to the authors); blinding not stated in the abstract. PMID 42140793, doi:10.1016/j.clinthera.2026.04.018
- Richards 2012 (Cochrane). Cochrane systematic review and meta-analysis (neuromodulators in rheumatoid arthritis; RCTs; search through November 2010). PMID 22258992, doi:10.1002/14651858.CD008921.pub2, The same team’s summary on inflammatory joint diseases (J Rheumatol Suppl 2012, PMID 22942326
- Schulze-Schiappacasse 2022. Structured evidence summary of systematic reviews (Epistemonikos, GRADE). PMID 33859125, doi:10.1097/RHU.0000000000001745
- Zintziovas 2026. Scoping review (RCTs and observational studies; assessed with RoB 2 and ROBINS-I; search through September 10, 2026). PMID 42831900, doi:10.1007/s00296-026-06314-x
- Dubois 2024. Systematic review (2012–2022; GRADE and ROBINS-I; narrative synthesis). PMID 38832841, doi:10.1080/03007995.2024.2363945, PROSPERO CRD42022354026
- Abafita 2026. Systematic review and meta-analysis (animal and human studies; RoB 2, GRADE; search through April 2024). PMID 42515994, doi:10.1155/prm/3998239, PROSPERO CRD42024539941
- Sajeev 2026 (UK Medical Cannabis Registry). Prospective patient registry case series (no comparison group). PMID 42724108, doi:10.1177/11795441261480709, Earlier analysis of the same registry: Francis 2025, PMID 38976497, 12 months, the patients most likely overlap)
- Francis 2024 (UK Medical Cannabis Registry). Prospective patient registry case series (no comparison group). PMID 38669060, doi:10.1080/15360288.2024.2340076
- Schubert 2023. Interim analysis of a prospective observational study (no comparison group; arthritis subgroup). PMID 37057257, doi:10.1177/20494637221147115
Points that could not be verified from the full text
- Heineman 2022 was read only from the abstract; for Vela 2022, the accepted author manuscript was read instead of the publisher’s version. Of the studies added in October 2026, Blake 2006, Kruger 2026, Richards 2012 (Cochrane), Dubois 2024, Francis 2024 and Schubert 2023 were read only from the abstract; the numbers for Blake 2006 were taken from reviews, and its funding could not be verified
- Ongoing studies (enrolling patients as of October 2026, no results): NCT06878417 (CBD and THC in hip/knee OA, pilot), NCT04992624 (CBD, THC and both together in knee OA), NCT05020028 (orally disintegrating CBD tablet in knee OA), NCT07638189 (CBD on the skin in knee OA; not yet started), NCT06414473 (oral CBD in knee OA; still appears to be enrolling in the registry, its 12-person results were shared in October 2026 as a non-peer-reviewed preprint, not included in the table). Completed but with results unpublished or not found: NCT04911127 (CBD in rheumatoid arthritis, 67 people, December 2023), NCT07445568 (CBD on the skin plus exercise in knee OA, November 2025), ISRCTN29199098 (cannabis cream on the finger joints in rheumatoid arthritis, Canada; end date December 2014), a CBD trial in rheumatoid arthritis and ankylosing spondylitis in Denmark (EudraCT 2017-004226-15; only the protocol was published); NCT04992962 (CBD/CBN and CBD/THC sublingual tablets in knee OA) status unknown
- A study of synthetic CBD gel applied to the skin in knee osteoarthritis, announced only as a conference abstract in 2018 (Hunter 2018), could not be found as a full peer-reviewed publication; because the abstract could not be accessed, it was not included in the table
Research history
Here we record what changed on this page, when and why, with dates. The newest change is at the top.
In October 2026, PubMed, Cochrane, ClinicalTrials.gov and Europe PMC were searched systematically; 12 studies were added: 3 randomized studies of products containing THC (a mouth spray containing THC and CBD in rheumatoid arthritis, two single-dose dronabinol studies in knee osteoarthritis), a study without placebo comparing three cannabinoid capsules, 5 reviews including Cochrane, and 3 observational studies. A non-peer-reviewed preprint and hemp seed oil studies were not included in the table; the reason is given in the evidence summary. Registration numbers were added to the five existing rows; the number of participants and the result of the knee osteoarthritis study (Pramhas 2023) were given in more detail, a funding note was added to two rows, and a strength-of-evidence note to the thumb-base study (Heineman 2022). The conclusion for products containing THC was rewritten, and the THC strength of evidence, which had been blank, was written for the first time ('Weak'). The conclusion and strength of evidence for CBD did not change.
Related studies: Blake 2006, published January 2006 (PMID 16282192, doi:10.1093/rheumatology/kei183, Letter: Wright 2006, PMID 16621924); Campbell 2023, published October 2023 (PMID 37202479, doi:10.1038/s41386-023-01597-1, NCT03098563, 2024, 58 people together with the Hamilton 2026 cohort, in the combined registry result, peak clinical pain (0–100) was 12.8 with dronabinol and 12.8 with placebo)); Hamilton 2026, published May 2026 (PMID 41481837, doi:10.1097/ALN.0000000000005925, NCT03098563, the results in the registry combine both cohorts)); Kruger 2026, published September 2026 (PMID 42140793, doi:10.1016/j.clinthera.2026.04.018); Richards 2012 (Cochrane), published January 18, 2012 (PMID 22258992, doi:10.1002/14651858.CD008921.pub2, The same team’s summary on inflammatory joint diseases (J Rheumatol Suppl 2012, PMID 22942326); Schulze-Schiappacasse 2022, published March 2022 (PMID 33859125, doi:10.1097/RHU.0000000000001745); Zintziovas 2026, published October 5, 2026 (PMID 42831900, doi:10.1007/s00296-026-06314-x); Dubois 2024, published July 2024 (PMID 38832841, doi:10.1080/03007995.2024.2363945, PROSPERO CRD42022354026); Abafita 2026, published July 27, 2026 (PMID 42515994, doi:10.1155/prm/3998239, PROSPERO CRD42024539941); Sajeev 2026 (UK Medical Cannabis Registry), published September 9, 2026 (PMID 42724108, doi:10.1177/11795441261480709, Earlier analysis of the same registry: Francis 2025, PMID 38976497, 12 months, the patients most likely overlap)); Francis 2024 (UK Medical Cannabis Registry), published June 2024 (PMID 38669060, doi:10.1080/15360288.2024.2340076); Schubert 2023, published April 2023 (PMID 37057257, doi:10.1177/20494637221147115)
The names of the conclusion labels were simplified: the "No benefit seen" label on this page became "No effect found." The assessment did not change.
The full texts of the studies were reread and the information on the page was checked against its sources. Corrected: dose, dose titration, side-effect and product information in the study table. Points that could not be verified from the full text went down from 3 items to 2. The conclusion and the strength of evidence did not change.
Older changes (1)
Page published. This first review looked at 5 studies.