Cannabis and Cognition
It's 1978 and my friend Charles is rolling a joint while I ease my Toyota Corona onto I-45. We're planning to smoke up at Galveston Beach, then drive back home before our midnight curfew.
Clearly, Charles and I weren't very bright.
To be more specific, we exhibited poor judgment, since (a) we lived about 40 minutes north of Galveston, (b) It was already after 10 p.m., and (c) I was a new driver and had never before attempted to operate a vehicle while stoned.
You might ask why we had such poor judgment. Was it our age? Was it the fact that we'd each been smoking marijuana for over a year? Was it residual effects from having smoked the prior weekend?
Surely our age played a role. Studies show that the prefrontal cortex, which handles planning, decision-making, and impulse control, among other things, doesn't become fully mature until one's mid-20s. One predictor of risky behavior among young people is immaturity in this area of the brain (depicted in red below).
As for whether marijuana affected the way Charles and I made decisions, this newsletter was prompted by a new review suggesting that marijuana use does impair cognition, even after the effects wear off. ("Cognition" refers to attention, memory, learning, decision-making, etc.) According to this review, Charles and I exhibited poor judgment because we were teenagers, and because we were marijuana users.
Should you trust this review? The question is worth asking. Marijuana is the third most widely-used drug in the world (right behind alcohol and tobacco). Currently, recreational use of marijuana is legal in 24 U.S. states and the District of Columbia, and the laws in other states may be changing soon. It's critical to know whether marijuana truly impairs cognitive functioning.
The dose-response problem
As it turns out, the new review provides an excellent example of how health data can be distorted in subtle but important ways. Journalists, educators, politicians – and researchers themselves – sometimes make the mistake these authors made. This mistake can be described as insufficient attention to dose-response relationships.
What does that mean exactly? Well, if you say that marijuana use impairs cognition, but you don't say more, the implication is that the mere use of marijuana is harmful. This is what the authors say, and it allows them to draw a simple contrast between "users" and "nonusers". Lots of people make that kind of distinction, including journalists at CNN and elsewhere who reported on their review.
However, the actual data suggests that dose matters. By "dose" I mean the amount of marijuana consumed over time. Low levels of consumption do not appear to be harmful, cognitively speaking. Thus, contrasting users vs. non-users is too simplistic, and offers no guidance to those who choose, legally or illegally, to engage in low levels of use.
I'll explain later what "low levels of use" might mean. The point here is simply that instead of referring to one monolithic category of "users", the data suggests that we should be talking about things like frequency of use, amount consumed, and strength of product.
A terminological note
"Marijuana" refers to cannabis plants that are psychoactive (i.e., intoxicating).
The most well-known chemical compounds in cannabis plants are "tetrahydrocannabinol" (THC), which is psychoactive, and "cannabidiol" (CBD), which is not. Marijuana contains both compounds, and others, but it's only the THC that gets people high.
In this newsletter, I refer to "marijuana". Researchers themselves often use the term "cannabis", but I'll only be discussing studies in which marijuana was the form of cannabis investigated.
An ideal source?
If you want answers to a big question – e.g., how does marijuana use affect cognition? – but you're not an expert in the field, your best source of information may be a review of studies, published recently in a reputable, peer-reviewed journal.
This newsletter was prompted by just such a review, led by Dr. Marco Solmi at the University of Ottawa and published in The BMJ last week.
It's an exhaustive review. Exhaustingly exhaustive:
–Dr. Solmi was supported by an international team of 31 co-authors.
–The review systematically assessed the impact of cannabis on dozens of health problems and outcomes.
–The review covered 101 meta-analyses published between 2008 and 2022. (Since each meta-analysis is itself a review of multiple studies, Dr. Solmi and colleagues created an "umbrella review" – i.e., a review of reviews.)
What caught my eye, and prompted this newsletter, is the authors' assertion that marijuana use is linked to impairments in cognition, specifically memory and learning. Not just when you're under the influence, but all of the time.
The language of this assertion is consistent throughout the review, from the Abstract ("impaired cognition in healthy cannabis users") to the Conclusion ("Convincing or converging evidence supports that cannabis use is associated with poor mental health and cognition").
In other words, if you use marijuana, your ability to think will be impaired, even when you're sober.
The types of impairment are pretty specific. With respect to memory, for instance, your ability to recognize a photo of Barack Obama would not be affected, but you might be less articulate in describing what you recall about his presidency. And, if you made a mental note to Google something about President Obama later, your chances of remembering to do so would be diminished.
Scary stuff. But I was troubled by the lack of information about dose-response relationships. Solmi and colleagues didn't say whether the cognitive impairments depend on amount of consumption. They simply indicated that any amount is harmful.
Evaluation of review
Since Solmi et al.'s conclusion is drawn from mutiple reviews of published research, you can't argue with it, right?
Wrong. The cognitive data that Solmi and colleagues reviewed actually came from just one source: A large, carefully-done meta-analysis of 88 studies published in 2016 by Tabea Schoeler, now at University of Lausanne. This meta-analysis found no memory effects for "light" use of cannabis, which the researchers defined as "less than four times per month or less than 10 joints per month". Rather, the effects were driven by heavy users, defined as "more than 20 times per month or more than 30 joints per month."
If you're a light marijuana user, this may seems like good news, although I don't recommend taking the researchers' definition of "light" too literally, because joints vary in size, people don't always smoke alone, and the THC content of the marijuana in any product can vary widely. The key takeaway is that there is some low level of consumption that's not harmful, though it's difficult to precisely identify the cut-off.
In sum, Solmi and colleagues' umbrella review misrepresents the cognitive data. In their review they claim that marijuana use, in general, undermines certain aspects of memory and learning. But the basis for this claim, the 2016 meta-analysis, only links heavy use to impairments.
Solmi and colleagues were mainly concerned with marijuana products used for medicinal purposes, so it's understandable if they didn't do a great job with the literature on cognitive effects. Those effects, which they misrepresented, only constituted a small part of their review.
Better data
Last year, Dr. Laura Dellazizzo and colleagues at the University of Montreal published an umbrella review in the journal Addiction that offers a more comprehensive and careful examination of the cognitive effects of marijuana consumption.
Some of Dellazizzo and colleagues' findings were consistent with those of the 2016 meta-analysis (in part due to overlap in studies reviewed). Their newer review also added or more strongly emphasized three points:
(a) Many aspects of cognition can be impaired while a person is high on marijuana. No surprise there. Charles and I knew this from personal experience, from talking to friends, and from Cheech and Chong's Up in Smoke. (If you're too young to know the movie, don't worry – it wasn't all that funny. Unless you were stoned when you saw it.)
(b) There may be slight residual effects of marijuana use. Because THC is stored in body fat, the impact of being high may continue to affect cognition once a person is sober. The duration of these effects is unclear from the data, because time between last marijuana use and testing has not been systematically explored, but for a small number of skills, such as verbal memory and verbal learning, there may be very small effects lasting as much as a week.
(c) Light use of marijuana is not linked to cognitive impairment. The meta-analyses that Dellazizzo and colleagues reviewed contrasted "chronic", "heavy", and/or "regular" users of marijuana with non-users, or, in some cases, "minimal" users. Although minimal use was not defined consistently across studies, as a group, these people showed no significant impairments.
Is marijuana harmless, cognitively speaking?
I see no evidence that low levels of marijuana consumption have enduring cognitive effects.
A critic might say, wait, the fact that we haven't found harmful effects for light consumption doesn't mean they don't exist. Absence of evidence is not evidence of absence.
I've always disliked that saying. It's catchy, but easily overstated.
For instance, right now, sitting in my back yard, I see no evidence of ants in the grass. That doesn't prove there are no ants. Perhaps I just haven't seen them yet. I should admit that absence of evidence is not evidence of absence.
However, my observation does provide a little bit of evidence that my grass is free of ants. Seeing no ants is evidence of absence, even though it falls short of proof.
Similarily, for research on the other type of grass (marijuana), absence of evidence for harmful effects (at low levels of consumption) does constitute a little bit of evidence that those effects just don't exist.
However, before you fire up your bong or reach for that edible, here are three reasons for caution:
1. "Low" levels of consumption are not clearly defined.
As with many nutrition studies, low levels of marijuana consumption are inherently difficult to measure. No doubt 10 bowls or 150 milligrams of gummies per day would addle your brain. The evidence on that is clear. But we don't know the exact cutoffs for safe levels of consumption. As with alcohol, saturated fats, and lots of other things that are bad for you when consumed in excess, it's hard to make fine distinctions as to how much marijuana can be safely ingested. I believe the evidence suggests that some amount is safe, but it's hard to say exactly how much.
2. Residual effects may occur.
For a few kinds of cognitive functioning, small effects of marijuana consumption have been observed up to a week following use. Here again, there are lots of question marks, as the strength of the marijuana consumed and the time since consumption have not been systematically explored across studies.
3. All of the data are group data.
Studies that show no effects for low levels of marijuana consumption typically compare the mean performance of "light" users vs. non-users. There's variability around means, of course, and so low levels of consumption may be harmless in general, but not for certain individuals.
4. Marijuana has been understudied.
Bear with me for a moment, and I'll link this somewhat lengthy anecdote to my theme.
In 1970, thanks in part to strenuous efforts by Richard Nixon, the newly-passed Controlled Substances Act classified marijuana as a Schedule I drug, right up there with LSD, heroin, and, more recently, fentanyl.
This will probably change soon, because the DEA officially defines Schedule 1 drugs as those with "no currently accepted medical use and a high potential for abuse", but marijuana has medical uses that are now widely accepted among both researchers and clinicians. For instance, the new Solmi meta-analysis confirmed that marijuana reduces pain and nausea among people with a variety of illnesses.
Classification of marijuana as Schedule I has created barriers to studying its impact, leaving researchers with two options: (a) Ask people about their prior marijuana use (which is inherently messy, because memory is imperfect, people don't know how much THC they've consumed, etc), or (b) engage in a complex process of seeking permission from three different federal agencies in order to administer marijuana in controlled experiments. Once permission is granted, the only place where researchers can obtain marijuana for their experiments is from the University of Mississippi, where the National Institute on Drug Abuse (NIDA) maintains a supply.
Cannabis researchers complain that the THC content of NIDA's marijuana supply is low compared to commercially available products in states where it's legal. Thus, federal law impedes research on the impact of marijuana by making it difficult to obtain permission, and by limiting what can be learned about dose-response relationships.
Although the data suggests that low levels of consumption are not harmful, I recommend caution, because new studies are slowly cropping up, and we may see effects that were not previously explored. (For instance, a study published last week links marijuana consumption to higher concentrations of cadmium and lead in the body, although these effects may be driven by chronic, heavy users.)
Conclusion
Charles and I got lucky that night. Though the streetlights seemed oddly fuzzy to me on the drive home, we did manage to get back safely. We weren't stopped by the police. Our parents didn't seem to notice that we'd missed curfew and our clothes smelled funny.
We were idiots, truly, but I don't think marijuana was the culprit. We shared a couple of joints per week, though not every week. According to current data, our decision-making would not have been affected by such low levels of consumption.
More research is needed on this topic. Owing in part to state-level deregulation, marijuana use has steadily increased in recent years and has now reached an all-time high. For instance, NIDA reports that in 2021, 43% of adults between the ages of 19 and 30 used marijuana at least once during the prior year. (Rates had been increasing for over a decade, so this is not a pandemic effect.) People need to know more about the impact of consumption, particularly as the THC content of commercial products continues to rise. The DEA is currently reviewing a proposal to downgrade marijuana from Schedule I status. I believe this will happen soon, and when it does, experimental studies will provide a better understanding of dose-response relationships – and thus clearer guidance for users.
Oh, and in case you're wondering, I stopped using marijuana decades ago. If you see me on the road, don't worry – I'll be sober, and driving more skillfully than I did in high school.
Thanks for reading!