In recent years, medical cannabis has been hailed as a miracle drug that can reduce chronic pain, prevent seizures, and even cure cancer. Using cannabis to alleviate pain is one of the most frequent reasons it’s prescribed—but do the results from clinical trials live up to the hype?
Treating Seizures with Medical Cannabis
Let’s look at some of the specific uses promoted for medical cannabis products, and the state of the science supporting the uses compared to the state of media reporting on these same issues.
Seizures are dramatic and sometimes frightening events. For many years, people experiencing seizures were thought to suffer demonic possession, be dying, or both.
This is a transcript from the video series The Skeptic’s Guide to Health, Medicine, and the Media. Watch it now, Wondrium.
Many people with epilepsy—the medical term for a condition that causes repeated seizures—can have most or all of their seizures prevented with medications or other therapies. But some people have intractable or difficult-to-control seizures.
Epilepsy medications, like all medicines, can have side effects; because of this, many people with epilepsy look for a safer or more effective alternative.
Parents, especially, are looking for help, particularly for children whose seizures can’t be prevented with traditional therapy. CBD, the non-psychoactive compound in marijuana, has been shown in animal and other early studies from the 1970s and 1980s to have anti-seizure properties.
Stories like one from CNN in 2013, “Marijuana Stops Child’s Severe Seizures,” give parents hope, but the science has not confirmed the anecdotes.
Looking at studies of seizure therapy in children, overall, there’s about a 30% response rate, which sounds appealing.
But some of the studies that support FDA-approved seizure medications showed about a 30% response rate to placebos. A large review from the Cochrane collaboration concluded that the research on cannabis for epilepsy was inconclusive, finding no solid evidence to recommend it for treatment.
There is one bright spot, though. A New England Journal of Medicine study from 2017 used a randomized, controlled design to look at CBD for children with one specific, very difficult kind of epilepsy called Dravet syndrome. In this study, adding CBD to the usual medical regimen was significantly more effective than the placebo, providing the first really solid evidence for CBD in seizure disorders.
But, this was only used to treat one specific kind of epilepsy, a very serious but rare disease, and it’s not clear that these results apply to other kinds of seizure disorders in children or adults. Headlines, unfortunately, didn’t always make this clear.
NBC claimed “Cannabis Drug Reduced Seizures in Severe Epilepsy Cases,” without pointing out that it was only in one specific kind of severe epilepsy.
Rolling Stone magazine’s wording was better and at least made it clear that the study was more specific: “Marijuana Reduces Seizures in Kids with Rare Form of Epilepsy.” But Rolling Stone referred to the intervention as “marijuana,” when in fact the study used isolated, purified CBD alone—the children were not consuming or smoking marijuana itself.
Can Marijuana Effectively Ease Chronic Pain?
Let’s consider another potential use for medical marijuana: The treatment of pain, especially chronic pain.
Pain is not such a rare condition. An estimated 100 million Americans suffer from chronic pain, costing $600 billion a year in medical bills and lost productivity.
Many traditional pain-treating medications are narcotics, derivatives of opium or morphine that can be addictive and can have some pretty serious side effects. Better ways to treat pain are sorely needed.
Can cannabis be part of the solution?
As with seizure disorders, the answer is “maybe.” An in-depth, comprehensive review of all available literature through 2016 done by the National Academy of Sciences found 30 randomized controlled trials using cannabis products to treat pain.
Many of these used specific synthetics or other derivatives that aren’t available in the United States.
Looking at one summary of the eight studies of plant-derived cannabinoids, marijuana extracts increased the chance of improvement by about 40% more than a placebo. That’s not a huge difference, but there is some effectiveness there.
Learn more about how to read beneath the hyperbole-filled headlines
Only one of those eight studies looked at smoked cannabis, in 50 patients with HIV-associated sensory neuropathy. It did document somewhat better pain reduction than a placebo, but the confidence interval—that is, the statistical measurement of the range of likely effect sizes—was very large.
Why is the Confidence Interval Important?
A confidence interval indicates that, based on our study, we think the real result is between these two numbers. You’ve heard of confidence intervals before, but maybe not by that name.
Think about the last presidential election: You probably heard that so-and-so led the other person by 45% versus 40% in the polls. But after those specific numbers, they’ll usually add a margin of error of plus or minus five percentage points.
That “margin of error” is just another way of saying a confidence interval. So 45%, plus or minus 5%, is the same as saying the confidence interval is from 40 to 50%.
Why is this important?
This measure is important because in finding out the truth, realistically, we can’t ask everyone in the population. So we take a sample, and we’re assuming that our sample is representative of everyone, but it might not be.
To use another example, perhaps you wanted to know the percentage of people in the United States with red hair.
Learn more about the oversimplification of “clickbait” headlines
The most accurate way would be to ask every single person, all 320 million, and record the results, but this is impractical. Instead, using care and caution, choose a smaller number, then sample randomly all over the country, because different areas might have fewer or more redheads around. Most important, though, you’d have to ask a high-enough number of people to get a fair estimate.
Think of it this way: If you asked four random people, and one said she had red hair, would you feel comfortable saying 25% of the U.S. population has red hair?
What if you asked 100 people? That would probably be more representative, or even 1,000, or 10,000?
As you can imagine, the more people you ask, the more likely your study percentage is to equal the real, entire population percentage of redheads.
Expressing this as a confidence interval, if you asked 1,000 people, you might be able to say the result is 5%, plus or minus 3%. But if you asked 10,000 people, you might be able to say the result is 5% plus or minus 1%.
Therefore, the larger the number of people studied, the less uncertainty there will be about the result, and the smaller the confidence interval will be.
Let’s relate this back to the study of inhaled cannabis for pain. Fifty people completed the study, about half given real cannabis, and half a placebo to smoke.
The result is expressed as an odds ratio expressing how many people responded to the medicine versus the placebo. A ratio of 1 means the responses were equal.
In this study, the odds ratio was about 3.4, meaning people getting cannabis were 3.4 times as likely to report a benefit than those given placebo.
Initially, it looks good, but when you look at the confidence intervals, the real odds ratio was somewhere between 1.03 and 11.48. The study only had 50 subjects, and that was too small to get a narrow, more accurate estimate.
Studies like these—small studies with large confidence intervals—have their place, as pilot studies, or as a starting point for larger investigations. But it’s difficult to draw broad conclusions about using marijuana to treat pain from the limited evidence available so far.
Nonetheless, marijuana is being used to treat pain and not the kind of extracts that were looked at in most of these studies. Chronic pain is the most commonly cited condition for medical marijuana use.
Learn more about interpreting scientific studies about cannabis
Barriers to Medical Marijuana Research
In Colorado alone, about 600,000 servings of edible marijuana are being sold each month for a combination of medical and now recreational use. And orally, eaten cannabis has not been studied as a pain reliever.
That brings up a crucial point about studying marijuana. There’s one significant barrier to conducting clinical trials on cannabis: As of late 2018 in the United States, marijuana is illegal to grow or possess, under federal law.
As of late 2018, researchers have to jump through many expensive, time-consuming hoops to get a study funded and approved. But now the wide and easy availability of medical marijuana in many states may be further discouraging research.
Many patients who are in pain are reading the newspapers, seeing headlines like “New Jersey Patients Beg: Treat Pain with Medical Marijuana Instead of Opioids,” and “Chronic Pain Sufferers in New York Will Soon be Able to Seek Relief through Medical Marijuana.”
Do you think they’re likely to sign up for a clinical study where they’ll be randomized, without their knowledge, to possibly receive a placebo? The drug is already assumed to be effective, at least in the eyes of many people and in many media reports.
Therefore, sensationalist headlines can be damaging not only to the people who fall for them but also to those medical researchers who wish to maintain integrity and rigorous standards.
Common Questions About Cannabis and Pain Relief
Studies show that ingesting cannabis can alleviate symptoms of pain to some degree.
Scientists differ, but it generally considered that the non-psychoactive component CBD helps with pain relief. At the same time, it is the entire chain of cannabinoids that synergize to help with pain.
Yes. The endocannabinoid system is a lipid-based receptor system of neurotransmitters that accept and process cannabinoids, that is biologically hard wired in.
Yes. Many studies have shown that people with epilepsy and other seizure varieties are greatly helped by cannabis, and in many cases the seizures stopped completely.