The Journal of Neuroscience recently published a study linking recreational marijuana use to subtle changes in brain structure. The researchers, led by Jodi Gilman of Massachusetts General Hospital, identified increased gray matter density in the left nucleus accumbens and some bordering areas. The study was fine, but the media coverage was abysmal. Reporters overstated the findings, mischaracterized the study, and failed to mention previous research done on pot smoking and health. Goldfish may not have a three-second memory, but some journalists seem to. When a new paper comes out, it’s often treated as the first ever and final word on the topic. There is a significant body of literature on the neurological and wider health effects of marijuana, and to ignore it when covering new studies seems to me a form of journalistic malpractice.
A press release from the Society for Neuroscience trumpeted the Gilman study’s importance because it looked at casual users rather than regular pot smokers, who form the basis of most marijuana studies. That claim is dubious in the extreme. The subjects averaged 3.83 days of smoking and 11.2 total joints per week. Characterizing these people as casual pot smokers was a great media hook, but it defied common sense. Occasional users wondered if they’d done permanent damage, and parents were concerned that their teenagers might face profound neurological changes from experimenting with pot. Any reporter who read the study, however, should have known not to take that bait.
Even by the standards of past medical studies, it’s a stretch to call these subjects casual pot smokers. Just two years ago, for example, Janna Cousijn and colleagues published a study on a group that she called “heavy” marijuana users. In the average week, they smoked 3 grams of cannabis—approximately 2 grams less than Gilman’s casual smokers. (A joint has about 0.5 grams of cannabis.) The justification for calling Gilman’s subjects casual smokers is that they didn’t meet the criteria for dependence, but when you count up the joints, the study doesn’t look so revolutionary.
Many stories also claimed that the Gilman study showed direct causation between pot smoking and brain abnormalities. That’s wrong. The study looked at differences between pot smokers and abstainers at a single moment. Only a longitudinal study, examining brain changes over time, could have suggested causation. As a letter writer to the Journal of Neuroscience noted, it’s possible that pre-existing brain differences cause some people to seek out marijuana. Gilman’s pot smokers also drank more and smoked more cigarettes than the control group, which supports this interpretation and also raises the possibility that other factors led to brain structure differences.
The biggest problem with the coverage of the marijuana study was that it failed to put the new research into context. Valentina Lorenzetti of the University of Melbourne recently published a widely cited review paper synthesizing dozens of studies on marijuana and the brain. Taking the literature as a whole, there is evidence suggesting that marijuana use causes structural changes in three parts of the brain: the frontal lobes, temporal lobes, and the cerebellum. The data also reinforces the idea that long-term, heavy smokers experience greater changes than casual users. The studies, however, have serious flaws. They are typically small and have been unable to show that the structural changes cause cognitive impairment. Gilman’s study of 20 smokers is a good contribution to the literature, but it doesn’t resolve those problems.
If you are considering smoking pot—or quitting—here is what you need to know.
Smoking marijuana once is very unlikely to harm you. It takes at least 15 grams of cannabis to kill a person, and probably much more than that. A healthy person would have to smoke dozens of joints in a single session to risk death from overdose. People who do die from the acute effects of marijuana die in accidents: A recent study suggested that more than 10 percent of drivers killed in car accidents test positive for cannabis.*
The more likely risk from trying marijuana is dependence. There is a debate over whether marijuana is addictive, but you needn’t bother with it—it’s mainly about semantics. The fact is, approximately 9 percent of people who start smoking pot become dependent by ordinary medical standards. That’s low compared with dependence rates for other drugs: More than 15 percent of people who drink become alcoholics, and 32 percent of people who try cigarettes get hooked. Still, you should think seriously about a 1-in-11 chance of addiction, especially if you have a personal or family history of substance abuse.
What sorts of health risks are these regular cannabis users taking? It’s extremely challenging to study the long-term health effects of marijuana in humans. You can’t legally ask 1,000 people to smoke three joints a week for 40 years just to see what happens, so researchers can only compare health data from people who admit to smoking pot with data from people who don’t admit to it. Retrospective correlational studies like these raise all kinds of problems, such as matching the groups for confounding variables. (Do they smoke cigarettes? Do they have a family history of cancer? What do they eat? Do they exercise?) Even if you assume that everyone is telling the truth, there are also bound to be wide variations in how much pot the subjects used. Most studies suggest that any potential health risks of cannabis are dose-dependent—people who smoke only a little face very few health risks, while people who smoke a lot are more likely to get sick—but this is still largely a matter of conjecture.
With that caveat out of the way, here are some findings. Studies consistently show that frequent marijuana smoking is associated with some forms of respiratory dysfunction. Smokers report problems with coughing, wheezing, and phlegm. Lung cancer is a murkier issue. Cannabis smoke contains higher concentrations of some carcinogens than cigarette smoke does. Some large studies show increased prevalence of respiratory tract cancers in cannabis users, while others find no correlation.
With the legalization of recreational marijuana in some states, many people have asked whether they can minimize cancer risk by ingesting rather than smoking cannabis. It’s a reasonable suggestion. At this point, however, the question is unanswerable. There simply aren’t enough people with a long history of eating marijuana, but not smoking it, to put together a study. Ingestion may be risky, because it seems easier to overindulge in food products than in smoking. Colorado is currently reviewing its regulations after accidental deaths involving ingested cannabis. Keeping cannabis brownies is especially risky if you ever have children in your home. A study released last year suggested that an increasing number of children in Colorado are accidentally eating marijuana-laced food products.
The cognitive effects of chronic marijuana use are uncertain. If you’re an adult who smokes occasionally, there appears to be little or no reason to believe your mental performance will suffer. Several studies also show that those who experience impairments may recover if they stop smoking. Heavy, long-term smokers may experience memory and attention loss. There is also some indication that heavy marijuana users are more likely to be diagnosed with schizophrenia, but it’s not clear which is the cause and which the effect, if indeed there is such a relationship; it may be the case that people with schizophrenia are self-medicating with marijuana.
As with alcohol and tobacco, it’s fairly clear that minors should not use marijuana. Many studies show that kids who smoke pot do poorly in school, and some studies suggest that they commit suicide at higher rates. Although the causal relationship isn’t clear, the risks are too great.
You probably have plenty of other questions. For example, is marijuana less bad for you than alcohol or tobacco? The comparison is basically impossible to make. Mountains of data link cigarette smoking to a staggering collection of adverse events. It’s difficult to know whether the same goes for marijuana, because fewer people smoke it, and those who do typically smoke less pot than cigarette smokers do tobacco. Comparing alcohol with marijuana—aside from differences in acute toxicity and driving competence—is also impractical.
Seriously, though, if you’re trying to decide among smoking pot, taking up cigarettes, and drinking alcohol based on health risks, I suggest finding a different hobby.
*Correction, May 1, 2014: Due to an editing error, this piece misstated that almost 25 percent of drivers killed in car accidents test positive for cannabis. Almost 25 percent test positive for non-alcohol drugs; of those, about 12 percent test positive for cannabis.
Brian Palmer is Slate’s chief explainer. He also writes How and Why and Ecologic for the Washington Post. Follow him on Twitter.