Increases in psychosis and schizophrenia attributable to cannabis use disorder in Ontario pre and post legalization.

An important new study was published on February 4th in JAMA Network Open. It is one of the largest population-based studies examining changes in psychosis and schizophrenia in relation to cannabis use disorder, covering the period before and after the legalization of medical cannabis (various measures in 2001, 2014, etc.) and recreational cannabis (late 2018) in Canada.

The authors found a marked increase in cannabis use disorder (CUD). While the overall incidence of schizophrenia did not change, other forms of psychosis (unspecified) increased dramatically. Schizophrenia incidence increased in younger men after 2015, following an earlier period of decline. The percentage of schizophrenia cases attributable to cannabis use disorder (population attributable risk fraction or PARF) rose dramatically, tripling from 3.7% pre-legalization to 10.3% post-legalization. This was highest in males aged 19-24, where one in five cases (18.9%) were associated with CUD, and much lower in females and older individuals.

These findings are similar to those from population-based data from Denmark (Hjorthoj and colleagues). Gilman has an excellent accompanying editorial, also attached. These findings likely represent conservative estimates of the cannabis impact because, first, they only detected CUD in people who were seen in emergency rooms or hospitalized. There is likely much more CUD that was not detected. Second, many cases of psychosis and schizophrenia associated with cannabis may not involve individuals with CUD, as we know from reports of naïve users whose psychosis or schizophrenia is triggered immediately.

This paper’s methods provide one of the most reliable epidemiological assessments of this important issue for cannabis policy.

The more than doubling of the population rate of unspecified forms of psychosis to the point where it surpassed the incidence of schizophrenia should be of great concern as well.

While it is still somewhat unclear whether schizophrenia is being triggered earlier in those who might eventually develop it or would not have occurred otherwise, either situation is deeply harmful.

Psychosis is one of the most devastating mental health symptoms, severely disrupting people’s lives and families, and one that our health system continues to fail at addressing. Now, we are allowing cannabis and hemp markets to aggressively market high-potency products that are aggravating this issue, which affects every city in America.

Our public policy in almost all states that have legalized cannabis still fails to address this issue.

Third, electronic medical records (EMRs) do not contain critical, fine-grained information on the THC potency of cannabis products, which is a key factor associated with the risk for psychosis. EMR studies that compare individuals who use cannabis with those who do not are often biased against finding a strong link, as individuals using cannabis infrequently or using low-potency products can dilute the overall signal in the data. The link between cannabis use and psychosis is considerably stronger for those using more potent products. Even these data may not include the variety of THC products on the market today in localities with thriving commercial markets.

The researchers focused specifically on individuals with CUD rather than examining all cannabis use, potentially allowing them to detect associations with schizophrenia among individuals who are most likely to experience adverse mental health outcomes. Even so, CUD diagnoses in medical records likely represent only a fraction of actual cases of CUD; individuals with heavy cannabis use who received outpatient care or did not seek care would be misclassified as having no CUD, again biasing results toward the null.

As legalization of cannabis becomes more widespread, along with a rapidly expanding commercial cannabis market, a natural experiment in population exposure to commercial cannabis markets is occurring. It is important not to misinterpret findings suggesting no significant acceleration in adverse outcomes after cannabis legalization as a statement that policy does not matter. Policy does matter; allowing unfettered commercial markets to exist exposes more individuals to a greater variety of readily available, high-potency cannabis products. Earlier use, more frequent use, and use of high-potency products all increase the risk of psychosis. Epidemiological studies should include long time intervals to account for processes of legalization and the emergence of mental illness, sufficient numbers to identify subpopulations at risk, such as young adults, and information concerning not only the quantity and frequency but also the potency of cannabis used. Without these critical factors, this research will be skewed toward nonsignificant findings, potentially obscuring important associations between cannabis policy and mental health outcomes.