Verified by 
Since the 1970s, research has repeatedly found evidence of an association between the use of cannabis and risk of developing psychosis, a condition that may manifest as delusions, hallucinations, or disorganized speech and behavior. As the drug becomes decriminalized in places around the world, some researchers have voiced concerns that an increase in usage could drive up the prevalence or severity of psychosis.
A new study may fuel these worries, finding that daily use of cannabis—particularly high-potency cannabis, or cannabis high in tetrahydrocannabinol (THC)—was associated with increased odds of developing psychosis. The researchers also found that in locations where high-potency cannabis is more widely used, such as Amsterdam, Paris, and London, a strikingly higher incidence of psychosis was observed than that seen in the other sites.
Other experts caution that the psychosis-cannabis relationship is complicated, and proving that one leads to the other is difficult, if not impossible, with current evidence.
The study, published last month in The Lancet, used data from 901 patients with first-episode psychosis and 1,237 controls, recruited from 10 European sites and one location in Brazil between 2010 and 2015. Participants were asked about their past drug use and, if they reported using cannabis, how often they used it and what type they used. Cannabis types were cross-referenced with government data to measure potency; strains of cannabis with more than 10 percent THC were classified as high-potency. The measure of consumption is imperfect, as it was based on participants’ recall and excludes other possibly relevant measures, like amounts of cannabidiol (CBD). Still, it allowed researchers “to assess patterns of lifetime THC exposure, whereas a laboratory test of THC concentration could only estimate a single point in time,” notes Tom Freeman, lead director of the Addiction and Mental Health Group at the University of Bath and a co-author on the study.
Compared to those who had never tried cannabis, those who used the drug on a daily basis had a three-times greater risk of having psychosis; those who used high-potency cannabis daily had five times the risk. In Paris, London, and Amsterdam—the three sites with the highest reported consumption of high-potency cannabis—odds of developing psychosis were four times, five times, and nine times greater, respectively, than for controls in those locations who did not use cannabis.
But assuming causality, experts say, may not be possible with the current evidence, and past research on the relationship has yielded mixed results. One 2017 genome-wide association study, for instance, looked at genetic data for more than 180,000 individuals and found that genetic risk factors for cannabis use were strongly correlated with those for schizophrenia, a condition that commonly causes psychosis. But when assessing the relationship between the two sets of risk factors, using a technique called Mendelian randomization, the authors found only weak evidence for a causal pathway from cannabis use to schizophrenia. The evidence was much stronger for a pathway from schizophrenia to cannabis use.
In her own work, Suzanne Gage, a psychologist at the University of Liverpool who was not involved in the research, has found evidence of a bidirectional causal pathway—the possibility that heightened psychosis risk leads to cannabis use and vice versa. She points out that those at higher risk for psychosis could be using cannabis to self-medicate, which could be confounding the association. The study published in The Lancet, she says, “is really well designed, but it still has an inherent problem.” It was observational; randomized, controlled trials, in contrast, could provide clearer evidence of cause-and-effect—but in this case, they are both impractical and unethical, she notes. “Because you don’t get that element of randomization, there are [possible] differences between people who choose to use cannabis and people who choose not to use cannabis.”
Many past studies have followed a specific cohort for several years to measure their cannabis use over time and assess their odds of developing psychosis, Gage says, which can provide a more accurate measure of cannabis use. But since psychosis is overall quite rare, it can be difficult to observe enough cases to to give the studies enough statistical power to draw clear conclusions. The case-control design used in the latest study—where a population with preexisting psychosis was matched with a population without psychosis—better addressed that problem by ensuring there were adequate sample sizes in each condition. “But the problem is it then also runs the risk of recall bias,” she says. “Since there have been media reports that heavy use of cannabis is associated with risk of psychosis,” those who have psychosis may assume that their cannabis use was a contributing factor to their psychosis, she adds, which may make it easier for them to recall times they used cannabis or inflate the drug’s potency in their self-reports.
Freeman is also cautious about causality. But the risk of assuming the relationship isn’t causal, he adds, could be high. “From a public health perspective, the safest option may be to assume a possible causal relationship between cannabis and psychosis.” In places where cannabis use is regulated, he says, setting a tax based on THC content could discourage both daily use and the use of higher-potency strains.
Rather than being seen as a gateway to increased incidence of psychosis, Gage argues, legalization ought to be viewed as a way to reduce the potential risk. “This is actually a great opportunity to do things like limit the levels of THC that can be in cannabis,” she says. The ratio of THC to CBD could also be controlled, she adds, as preliminary evidence suggests that higher levels of CBD may counteract the more negative properties of THC, such as psychotic symptoms, memory disruption, and anxiety. “It’s impossible to do this when the markets are controlled by criminals. But if the markets are regulated, it would be very easy to do.”
Regardless of psychosis risk, she notes, reducing the use of cannabis would likely confer benefits to heavy users. “If someone is using high-potency cannabis every day, that could be a marker for all sorts of things,” she says. “That’s worrying in and of itself. Outside of any potential risk of psychosis, it’s a maladaptive behavior pattern.”
