Cannabis and cannabinoids for the treatment of mental and substance use disorders and symptoms: A systematic review and meta-analysis of experimental and observational studies.

Background: Interest in cannabinoids for mental and substance use disorders is increasing. We examined experimental and observational evidence for treating these disorders and their symptoms.

Methods: Systematic review and meta-analysis (PROSPERO CRD42023467536). We searched CENTRAL, MEDLINE, Embase and PsycINFO to May 2025 for studies of cannabinoids in adults (≥18 years) with ADHD, anxiety, depression, PTSD, psychosis or Tourette syndrome, or alcohol, cannabis, opioid or tobacco use disorders. Two reviewers screened, extracted and assessed quality using a risk-of-bias tool and GRADE.

Results: We included 82 experimental and 118 observational studies. In RCTs, cannabinoids reduced anxiety symptoms (SMD=-0.40; 95% CI: -0.57, -0.23; I²=90%) and, in one small trial, PTSD symptoms (SMD=-2.60; 95% CI: -4.58, -0.62; n=20), with trivial-to-no effect on depression (SMD=-0.20; 95% CI: -0.43, 0.04; I²=91.6%), ADHD, psychosis and Tourette syndrome. Much anxiety and depression evidence came from symptoms measured as secondary outcomes in other primary conditions. Cannabinoids worsened cannabis use disorder severity in one RCT (SMD=2.35; 95% CI: 1.49, 3.21), with no effect on craving or withdrawal; evidence for alcohol, opioid and tobacco use disorders was very limited. Observational studies suggested improvements but had high risk of bias. The only significant safety finding was increased withdrawals due to adverse events with THC (OR=2.78; 95% CI: 1.66, 4.65). Certainty was predominantly very low.

Discussion: The evidence base shows very low certainty, high heterogeneity and methodological limitations, and is insufficient to support cannabinoids as first-line treatment. Signals for anxiety and PTSD are limited by indirectness and low certainty; no benefit was evident for depression; THC-related safety signals warrant careful consideration.