A team of Australian and Dutch researchers (Gillespie et al), writing in the June 2019 issue of the British medical journal The Lancet, criticized a March 2019 study by British researchers (Di Forti et al) in the same journal that suggested variations in cannabis use and potency may be responsible for variations in psychotic disorders. Opponents of cannabis legalization have used the Di Forti study as evidence to support their position. 

The criticism rests on the fact that the authors of the March study “assume that cannabis causes psychosis or psychotic symptoms without acknowledging compelling, alternative hypotheses.” 

Gillespie and colleagues point out that most studies looking at associations between cannabis and psychosis don’t adjust for “confounding” that arises from correlated genetic and environmental individual differences. They point to their own findings as well as those of other researchers showing cannabis use may be higher among individuals with a genetic liability that predisposes them to both cannabis use and the development of psychotic disorders. 

To address the shortcomings of the March study, they specifically point to the results of their recent meta-analysis of the largest genome-wide association study of lifetime cannabis use to date. The study indicated that genetic risk factors for cannabis use and schizophrenia are positively correlated. The meta-analysis applied bidirectional randomization and found a “consistent pattern of evidence supporting a causal effect of schizophrenia risk on lifetime cannabis use.” The study “found little evidence for any causal effect of cannabis use on schizophrenia.” While conceding their analyses were not based on cannabis use frequency or potency but rather genetic risk factors, they felt confident making the following statement:

Nevertheless, our findings strongly suggested that associations between measures of cannabis use and psychosis or psychotic disorders are far more nuanced than Di Forti and colleagues assume. In addition to correlated genetic liabilities, indirect and bidirectional processes are likely to affect the associations between cannabis use, misuse, and psychotic disorders. By not acknowledging the alternative, compelling and plausible mechanisms, Di Forti and colleagues' conclusion regarding the harmful effect of high-potency cannabis use on mental health is likely to be overestimated.

Opponents of cannabis legalization have seized on the warnings that Alex Berenson published in “Tell Your Children: The Truth About Marijuana, Mental Illness, and Violence” earlier this year in an effort to slow the momentum for legalization in the US. His book has come under criticism from many quarters for cherry-picking data, ignoring trends and confounders, and over-emphasizing correlations. Science journalist Maia Szalavitz, critiquing Berenson’s book shortly after its release, raised the same points as Gillespie and colleagues:

But we still don’t really know whether a third factor—such as a genetic predisposition that creates both an affinity for weed and an increased risk of psychosis—accounts for this link. According to Matthew Hill, an associate professor of neuroscience at the University of Calgary in Canada, candidate genes that do both have already been identified.

I have pointed out other criticisms here.

With Canada having legalized cannabis for recreational use last year, Mexico about to do the same, and 11 states plus the District of Columbia having legalized recreational cannabis, momentum continues to build for federal decriminalization or legalization. Opponents will grab on to any study that might generate enough fear to apply the brakes. The evidence and arguments put forth by Gillespie and colleagues should help to calm such fears.