The evidence for medical cannabis is insufficient for most medical conditions.

The evidence for medical cannabis is insufficient for most medical conditions.

Medical cannabis: research results and risks

Medical cannabis does not have sufficient scientific support for the common uses for which it is used, such as chronic pain, anxiety and insomnia, according to a comprehensive review led by UCLA Health.

Overview of the study

The study, published inJAMA, reviewed more than 2,500 articles published from January 2010 to September 2025, including randomized clinical trials, meta-analyses, and clinical guidelines. Over 120 studies were prioritized based on large sample sizes, timeliness, topics covered, and relevance.

According to a 2018 survey, 27% of people in the United States and Canada reported using cannabis for various purposes such as pain relief, anxiety management, and sleep problems.

Public perception vs. scientific findings

Dr. Michael Hsu of UCLA Health, the lead author of the review, said there is a disconnect between the public's understanding and the latest scientific evidence regarding the purported medical benefits of cannabis.

"Although many people seek cannabis for relief, our review highlights significant gaps between public perception and scientific evidence regarding its effectiveness for most medical conditions. Clear guidance from clinics is critical to support safe, evidence-based decisions in the medical cannabis conversation."

Dr. Michael Hsu, clinical assistant professor of health sciences, UCLA Health Department of Psychiatry and Biobehavioral Sciences

Effectiveness of pharmaceutical cannabinoids

The review confirmed that the U.S. Food and Drug Administration (FDA)-approved pharmaceutical cannabinoids are only effective in a narrow range of conditions. These include medications for loss of appetite in HIV/AIDS, nausea and vomiting caused by chemotherapy, and certain severe childhood seizure disorders such as Dravet syndrome and Lennox-Gastaut syndrome.

For most other conditions, evidence remains either insufficient or missing, the review found. Over half of people who use medical cannabis report using it to treat chronic pain However, current clinical guidelines do not recommend cannabis medicine as a first-line treatment for chronic pain.

Health risks associated with cannabis

The review also highlighted potential health risks associated with cannabis use. Long-term data from adolescents suggested that high-potency cannabis may be associated with increased rates of psychotic symptoms (12.4% compared to 7.1% for low-potency cannabis) and generalized anxiety disorder (19.1% compared to 11.6%).

About 29% of people who use medical cannabis also met criteria for cannabis use disorder. Daily use of cannabis, particularly use of inhaled and/or high-potency products, may also be associated with cardiovascular risks, including higher rates of coronary heart disease, heart attack and stroke compared to less frequent use.

Important considerations for medical professionals

The review emphasizes that physicians should carefully screen their patients for cardiovascular disease and psychotic disorders and evaluate possible drug interactions before considering products containing THC for medical purposes.

“Patients deserve honest conversations about what science tells us and doesn’t tell us about medical cannabis,” Hsu said.

Limitations of review

The authors noted some limitations of the review. The work was not a systematic review and no formal assessment of the risk of bias of the included studies was performed. Many of the studies were observational and could be subject to bias. Additionally, recommendations from clinical trials may not apply to all patients due to differences in design, patient characteristics, and cannabis products tested.

"Further research is important to better understand the potential benefits and risks of medical cannabis. By supporting more rigorous studies, we can provide clearer guidance and improve clinical care for patients," said Dr. Hsu.

Researchers from Harvard, UC San Francisco, Washington University School of Medicine and New York University contributed to the study.

Sources: