The Largest-Ever Study Found No Link Between Cannabis and Mental Health Improvement

Medicinal cannabis has been sold to millions of people as a treatment for anxiety, depression, and PTSD. A major new meta-analysis says there is no evidence it works — and some reason to think it might be making things worse.

The study, published in The Lancet Psychiatry00015-5), reviewed 54 randomized controlled trials conducted worldwide over 45 years. It is the largest systematic review ever conducted on the safety and effectiveness of cannabinoids across mental health conditions. The conclusion is blunt: no evidence that medicinal cannabis is effective for anxiety, depression, or PTSD.

The timing matters. Medicinal cannabis has become a significant commercial industry. In Australia, where the study was led by researchers at the University of Sydney Matilda Centre, more than one million prescription approvals have been issued over the past four years, with sales of cannabinoid medications tripling. Roughly 700,000 Australians have used medicinal cannabis to treat over 250 different health conditions. In the U.S. and Canada, about 27% of adults aged 16-65 have used cannabis medically — and roughly half of those were using it for mental health symptoms.

The study lead author Dr. Jack Wilson said the results call into question why these approvals are being granted in the first place. The routine use of medicinal cannabis could be doing more harm than good, he said, citing risks of psychotic symptoms, cannabis use disorder, and the danger of delaying more effective treatments.

That is a significant claim given the marketing push behind these products. The study found some weak signals that cannabis might help with certain conditions — cannabis use disorder, autism symptoms, insomnia, and tics or Tourette syndrome — but Wilson was careful to note the overall evidence quality was low. For most of those conditions, the data barely clears the bar.

The picture gets more complicated with substance use disorders. Cannabis may help treat dependence on cannabis itself — similar to how methadone is used for opioid use disorder — but it appears to make cocaine dependence worse. Among people with cocaine-use disorder, cannabis use actually increased cravings, suggesting it should not be considered as a treatment option in that context.

The broader regulatory question is whether the approval frameworks for medicinal cannabis are ahead of the evidence. Major medical organizations including the Australian Medical Association, the Pharmacy Guild of Australia, and the American Medical Association have all raised concerns about the pace of approvals relative to what the science actually shows. The Therapeutic Goods Administration in Australia has initiated a review of its regulatory oversight, with more than 500 responses published in February.

The study does not argue that cannabis has no medical uses — the evidence for epilepsy, multiple sclerosis spasticity, and certain pain conditions is more established. But the mental health claims, which have driven much of the commercial expansion, do not survive scrutiny. When a product is approved and sold for conditions it probably does not help, and might harm, the burden of proof should run the other way.