Research on medicinal uses of marijuana: Is the cart before the horse?

I describe two studies of medicinal uses of marijuana in Chapter 4 of Tools for Critical Thinking in Biology to illustrate why experiments are considered the gold standard for research. In June 2015, two months after the book appeared, the Journal of the American Medical Association (JAMA) published two reviews and an editorial about research on medicinal uses of marijuana. In their editorial, D’Souza and Ranganathan concluded that: “Since medical marijuana is not a life-saving intervention, it may be prudent to wait before widely adopting its use until high-quality evidence is available to guide the development of a rational approval process. Perhaps it is time to place the horse back in front of the cart.” Before explaining why D’Souza and Ranganathan think the cart is before the horse in marijuana research, I want to tell you what the reviews in JAMA have to say about the two studies I discussed in detail in my book.

Marijuana plant.

Marijuana plant growing in Sherburne National Wildlife Refuge, Minnesota
(US Fish and Wildlife Service).

In one of the studies I discussed, a group of researchers in San Francisco studied the effects of smoking marijuana on a type of pain called peripheral neuropathy experienced by HIV/AIDS patients; in the other, a group of researchers in San Diego studied effects of marijuana on experimentally induced pain in healthy volunteers. Both studies were randomized controlled trials: subjects were randomly assigned to treatments and one of the treatments was a placebo, or control, in which THC, one of the main active ingredients of marijuana, had been removed from cigarettes used by the volunteers. In addition, the experiments were double-blind trials: the subjects were ostensibly unaware of the treatment they received in each phase of the experiments, as were the researchers until the experiments were completed and the researchers started analyzing the data.


Chemical structure of THC, one active ingredient of marijuana, by Yikrazuul.

Neither set of reviewers for JAMA mentioned the San Diego study, presumably because these researchers worked with healthy volunteers rather than testing marijuana for treating a specific disease. I included the San Diego study in my book because it illustrated some important considerations in designing experiments. For example, the researchers asked whether there was a relationship between the dose of THC received by subjects and alleviation of pain. They found that an intermediate dose of THC was more effective than either a lower or higher dose; patients actually felt the most pain with the highest dose. These results are clearly relevant to using marijuana for treating pain caused by a disease such as HIV/AIDS, even though the JAMA reviewers didn’t see fit to mention this study in their papers.

One of the JAMA reviews was couched as a response to a patient who had been treated for chronic back pain for 18 years with only modest success. The patient began using marijuana shortly after it was legalized for medical use in his state. The second review was a quantitative analysis of evidence for the effectiveness of marijuana for treating several medical conditions. Both reviews focused on experimental studies and they reached similar conclusions about strengths and weaknesses of evidence for effectiveness of marijuana for treating various conditions.

Penny Whiting and her colleagues considered 79 randomized controlled experiments in their quantitative analysis. In 28 of these experiments, including the San Francisco study that I described in my book, marijuana was used in an attempt to alleviate chronic pain. Most of these experiments used an oral spray containing purified forms of two active ingredients extracted from marijuana, but the San Francisco volunteers smoked marijuana cigarettes. Subjects in the treated groups felt less pain than subjects in the control groups in most of these experiments, with the greatest reduction in pain for the San Francisco experiment. I used the San Francisco study in my book because it was a good example of how to design a randomized controlled experiment. With Whiting’s review, I can go further and say that the results of this experiment are consistent with those of other recent experiments addressing similar questions; if anything, the results of the California study are even more persuasive than those of similar experiments.

In addition to finding credible evidence that marijuana can benefit patients with chronic pain, the JAMA reviewers found moderate support from experimental studies that marijuana can reduce spasticity in patients with multiple sclerosis but only weak evidence that it can reduce nausea and vomiting associated with chemotherapy. The reviewers found essentially no evidence that it helps patients with depression, anxiety disorders, sleep disorders, psychoses, Tourette syndrome, Parkinson’s disease, inflammatory bowel syndrome, or glaucoma. Whiting’s group also found evidence for many adverse side effects, at least in the short term.

Why did the editorial accompanying these reviews in Journal of the American Medical Association conclude we should put the horse back in front of the cart in marijuana research? About half of the US states allow residents to use marijuana for medical purposes, but each state lists a specific set of allowable conditions and these lists are very diverse. In most cases, support for including a medical condition on a state list is not based on randomized, controlled experiments – the gold standard for research – but on weak evidence at best: “anecdotal reports, individual testimonials, legislative initiatives, and public opinion.” Marijuana contains at least 400 secondary compounds, including 70 related to THC, the best known active ingredient. There is great variation in the chemical composition of different samples of marijuana, leading to unpredictable effects on the body. Interactions of marijuana with other drugs that may be taken concurrently are uncertain. There are numerous short-term side effects and potential long-term risks, especially for children and young adults whose brains are still developing, since brain development depends on a natural compound in the brain that is similar to THC and binds to the same receptor molecule on the membranes of brain cells. For these and other reasons, D’Souza and Ranganathan argue in JAMA that we need to gather more and better evidence to justify or discredit use of marijuana for specific medical conditions.

This entry was posted in Experimentation, Medicine. Bookmark the permalink.

One Response to Research on medicinal uses of marijuana: Is the cart before the horse?

  1. Pingback: Medical use of marijuana revisited | Critical Thinking in Biology

Comments are closed.