Cannabinoids are increasingly promoted and used for the treatment of mental disorders and substance use disorders (SUDs), despite limited clarity regarding their efficacy and safety. These compounds, including delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD), interact with neurobiological systems implicated in psychiatric conditions. However, prior evidence has been inconsistent, often based on small or low-quality studies, and there is concern that expanding clinical use may outpace the supporting evidence. This study aimed to systematically evaluate randomized controlled trial (RCT) evidence on cannabinoids as primary treatments for mental disorders and SUDs.

This systematic review and meta-analysis included RCTs published between 1980 and May 2025 from major databases (e.g., MEDLINE, Embase, Cochrane). Eligible studies evaluated plant-based or pharmaceutical cannabinoids used as primary treatments for mental disorders or SUDs. A total of 54 trials (2477 participants) were included. Primary outcomes were remission or symptom reduction, while safety outcomes included all-cause and serious adverse events. Data were pooled using random-effects meta-analysis, with results expressed as standardized mean differences (SMDs) or odds ratios (ORs). Risk of bias was assessed using Cochrane tools, and evidence quality was graded using GRADE.

Overall, evidence quality was low, with 44% of studies at high risk of bias. Cannabinoids demonstrated modest benefits in select conditions, including decreased tic severity in Tourette’s syndrome (particularly with combined THC and CBD), increased total sleep time in insomnia (without improvements in sleep quality or latency), and small reductions in autistic traits in autism spectrum disorder. However, no significant effects were observed for anxiety disorders, psychotic disorders, PTSD, opioid use disorder, anorexia nervosa, and OCD. Unfortunately, there was insufficient data to get an accurate assessment of how cannabinoids affect ADHD, bipolar disorder, or tobacco use disorder. Cannabinoids were associated with increased cocaine craving in cocaine use disorder. Safety analyses showed higher odds of all-cause adverse events with cannabinoids (OR 1.75; number needed to harm ≈7), but no differences in serious adverse events or treatment discontinuation compared with controls.

This comprehensive meta-analysis demonstrates that cannabinoids provide limited and condition-specific benefits, with generally low-certainty evidence. While some symptom improvements were observed (notably in cannabis use disorder, insomnia, and tic disorders), cannabinoids were not effective for most psychiatric conditions and were associated with increased non-serious adverse events. Given the mismatch between widespread clinical use and limited supporting evidence, routine use of cannabinoids for mental disorders and SUDs is rarely justified at present. Higher-quality, larger-scale RCTs are needed to clarify therapeutic roles and safety, and clinicians should exercise caution when considering cannabinoid-based treatments.

Reference:

Wilson J, et al. Lancet Psychiatry 2026;13:304-15. Abstract