The evidence is thin — and the conditions people use it for most are the ones it helps least
STUDY: Wilson J et al, Lancet Psychiatry 2026;13:304–315, PMID: 41856154
STUDY TYPE: Systematic review and meta-analysis
FUNDING: National Health and Medical Research Council (Australia)
Background
Around 1 in 3 US adults have tried medical cannabis, although the medical problems were its use is justified are few, as reported in JAMA:
- Chemotherapy nausea (a small benefit)
- Weight restoration in HIV+
Outside of that, the evidence is either non-existent or not outweighed by its risks:
- Heart and lung disease
- Addiction: 1 in 3 medical users progress to cannabis use disorder
- Car accidents
- Psychosis and anxiety
But what about its use in psychiatric disorders? Today’s paper is the largest analysis to examine that.
The Study
The authors pooled 54 trials and 2,477 participants across 14 conditions, including anxiety, PTSD, psychosis, insomnia, autism, Tourette’s, and several substance use disorders.
Cannabinoids showed no benefit for anxiety, PTSD, psychosis, OCD, anorexia, ADHD, bipolar disorder, opioid, tobacco, cannabis, and cocaine use disorders. The conditions with positive findings were:
- Cannabis use disorder (reduced withdrawal symptoms and weekly use)
- Insomnia (modestly more sleep time on device measurement, moderate certainty)
- Tic or Tourette’s syndrome (reduced tic severity, but only with combined THC/CBD)
- Autism spectrum disorder (reduced autistic traits, but very low certainty and high risk of bias in both studies)
In cocaine use disorder, cannabinoids worsened cravings for cocaine.
There were no randomized trials in depression, even though this is one of the most common reasons people seek medical cannabis.
Safety-wise, cannabinoids produced more adverse events for 1 in 7 people across all studies (number needed to harm = 1 in 7), but didn’t increase serious adverse events or dropouts. Most common adverse events were dry mouth, nausea, diarrhea, and dizziness.
Nearly half of the studies had a high risk of bias, and most findings were rated low to very low certainty.
Practice Implications
Many patients use cannabis for anxiety, depression, or PTSD, and these are where the evidence is weakest
The modestly positive findings — Tourette’s, insomnia, cannabis use disorder — are worth knowing, especially since effective pharmacotherapy is scarce for all three. But even these rest on small trials that don’t outweigh the better established risks.
One exception is cannabis use disorder. Here CBD is reasonable to try as a harm reduction strategy. It may work, and it is much safer than full cannabis. It is THC that causes most of the problems with addiction, anxiety, and psychosis.
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
What’s Your Take? Share in Comments
- Are your patients using medical cannabis? What benefits and harms do you see?
