A marijuana-derived compound may help reduce the frequency of seizures for epilepsy sufferers, according to a small study published today. More than half of the children and adults treated with cannabidiol (CBD) experienced fewer seizures over the course of a 12-week trial.

The medicinal benefits of marijuana are a topic of great debate in the United States at the moment, where a number of states have recently legalized its use. Anecdotal reports and small-scale studies have suggested that cannabis could help sufferers of treatment-resistant epilepsy by reducing the number of seizures they suffer. The debate arises from the lack of properly rigorous research exploring the hypothesis, and it won't be settled by today's findings, which have important limitations that prevent them from being conclusive. Still, they do offer some of the strongest support yet for the basic premise that cannabis extracts can help.

CBD is a non-psychoactive compound derived from cannabis, meaning it can be administered to patients without getting them high. Nonetheless, side effects reported in the study included drowsiness for a fifth of the participants, as well as diarrhea, fatigue, and loss of appetite. So far, 213 patients between the ages of 2 and 42 have taken part, of whom 137 have completed the full 12-week course. Fifty-four percent of them have seen a reduction in epileptic seizures.

Dr. Orrin Devinsky of the NYU Langone Comprehensive Epilepsy Center is the lead researcher for this study, which he will present to the American Academy of Neurology’s Annual Meeting later this month. He cautions that it's still only an early trial that could be subject to a placebo effect: knowing you're taking something that's supposed to make you better could be helping by simply fostering a more positive mindset. Devinsky plans to soon begin a randomized, placebo-controlled trial, where neither the doctors nor the patients know whether they're receiving CBD or a placebo. That should preclude any potential bias in the findings and help establish a scientific basis for the enthusiasm (or otherwise) for using marijuana as a healing agent.

The study was supported by GW Pharmaceuticals, a UK company that provided the cannabidiol used for testing. Devinsky and GW presented similar findings at the end of last year, though the number of patients has now grown and progress is being made toward more rigorous trials. One of the main stumbling blocks, notes Devinsky, is that at least three US federal agencies have to provide clearance before a cannabis trial on human subjects can proceed: the Food and Drug Administration, the Drug Enforcement Agency, and the Bureau of Narcotic Enforcement. So even though everyone agrees that more research is required, getting there looks likely to be a long and drawn-out process.