rethinkTHC Search
Menu
Study breakdown

Large Meta-Analysis Found Cannabis Provides Only Modest Pain Relief With Significant Side Effects

Meta AnalysisStrong evidence
The takeaway

Across 104 studies with nearly 10,000 participants, cannabinoids provided a statistically significant but clinically small reduction in chronic pain (equivalent to 3 mm on a 100 mm scale), with a high rate of adverse events.

Pain management providers, chronic pain patients considering cannabis, medical cannabis policy makers.

NNT of 24 for 30% pain reduction and NNH of 6 for adverse events suggest cannabinoids are unlikely to be highly effective pain medicines for most patients.

What the researchers found

For 30% pain reduction: 29% on cannabinoids vs 25.9% on placebo (NNT=24). For 50% pain reduction: no significant difference. Overall pain intensity reduction was equivalent to just 3 mm on a 100 mm visual analogue scale. Adverse events were common: NNH was only 6 (81.2% vs 66.2%). No significant impacts on physical or emotional functioning.

Why it matters

This is one of the largest and most rigorous meta-analyses of cannabinoids for chronic pain. Its sobering conclusion challenges the widespread belief that cannabis is highly effective for pain, while noting that some patients do benefit.

The numbers in context

104 studies, 9,958 participants. NNT for 30% pain reduction: 24 (CI 15-61). Pain intensity reduction: 3 mm on 100 mm scale. NNH for adverse events: 6 (CI 5-8). 48 studies on neuropathic pain, 7 on fibromyalgia.

How the study worked

Systematic review and meta-analysis of 104 studies (47 RCTs, 57 observational) with 9,958 participants, searched across MEDLINE, Embase, PsycINFO, CENTRAL, and clinicaltrials.gov. IMMPACT guidelines followed.

What this study cannot tell us

Heterogeneity across studies in cannabinoid type, dose, formulation, and pain condition. Most RCTs were relatively short-term. Publication bias possible. Average effects may mask meaningful responses in subgroups.

How to read the evidence

Strong - large, rigorous meta-analysis following IMMPACT guidelines with nearly 10,000 participants across 104 studies.

When this study was published

Published in 2018. Literature searched through July 2017.

The bigger picture

While cannabis is widely promoted for pain, this meta-analysis suggests the average benefit is small and comes with frequent side effects. However, the NNT of 24 is not zero - some patients do experience meaningful relief, and cannabis may still be preferable to more dangerous alternatives like opioids.

Questions still open

  • Are there identifiable subgroups who respond better to cannabinoids for pain? Would different cannabinoid formulations or doses produce better results? How does cannabinoid pain relief compare to other analgesics on the same scale?

Common questions

Does cannabis actually help with chronic pain?
This meta-analysis of 104 studies found that cannabis provides a real but very small average pain reduction - equivalent to 3 mm on a 100 mm scale. About 29% of cannabis users achieved 30% pain reduction vs 26% on placebo. Side effects were common.
How effective is cannabis compared to other pain medications?
Based on this analysis, the number needed to treat for cannabis to achieve 30% pain reduction was 24, which is high compared to many conventional analgesics. However, cannabis may be considered when other options carry greater risks, such as opioids.

Read the original research

Cannabis and cannabinoids for the treatment of people with chronic noncancer pain conditions: a systematic review and meta-analysis of controlled and observational studies.

Pain, 159(10), 1932-1954

Citation

Stockings, Emily; Campbell, Gabrielle; Hall, Wayne D; Nielsen, Suzanne; Zagic, Dino; Rahman, Rakin; Murnion, Bridin; Farrell, Michael; Weier, Megan; Degenhardt, Louisa. (2018). Cannabis and cannabinoids for the treatment of people with chronic noncancer pain conditions: a systematic review and meta-analysis of controlled and observational studies.. Pain, 159(10), 1932-1954. https://doi.org/10.1097/j.pain.0000000000001293

Explore the wider topic