Portal de Periódicos da CAPES

High-potency cannabis and incident psychosis: correcting the causal assumption – Authors' reply

2019; Elsevier BV; Volume: 6; Issue: 6 Linguagem: Inglês

10.1016/s2215-0366(19)30176-2

2215-0374

Marta Di Forti, Craig Morgan, Jean-Paul Selten, Michael T. Lynskey, Robin M. Murray,

Schizophrenia research and treatment

We are grateful for the opportunity to respond to the letters published in response to our Article.1Di Forti M Quattrone D Freeman TP et al.The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): a multicentre case-control study.Lancet Psychiatry. 2019; 6: 427-436Summary Full Text Full Text PDF PubMed Scopus (371) Google Scholar We do not believe—nor do our findings imply—that the cause of psychotic disorder is simple and attributable to one factor; rather, we acknowledge that multiple factors combine to cause it. A complex pathway underlying psychotic disorder might involve predisposing genes, exposure to prenatal complications, and use of high-potency cannabis. Removing any one of these factors would prevent all cases of the disorder attributable to this complex set of factors. In this way, population attributable fractions can add up to more than 1 (100%), because some individuals with more than one risk factor can have disease onset prevented in more than one way.2Rowe AK Powell KE Flanders WD Why population attributable fractions can sum to more than one.Am J Prev Med. 2004; 26: 243-249Summary Full Text Full Text PDF PubMed Scopus (91) Google Scholar Therefore, we respond to Iris Sommer and Wim van den Brink by highlighting that our paper shows, assuming complex causality, that removing high-potency cannabis use—as one component cause among others—could prevent up to 50% of new cases of psychotic disorder in Amsterdam, the Netherlands. Thus, our proposal that high-potency cannabis use accounts for a sizeable proportion of incident cases of psychosis is neither implausible nor inconsistent with what is known about the genetics of multifactorial disorders such as psychosis. Furthermore, in estimating heritability, gene–environment interactions are typically attributed to the genetic component, leading the unwary to underestimate the effect of the environment.3Mayhew AJ Meyre D Assessing the heritability of complex traits in humans: methodological challenges and opportunities.Curr Genomics. 2017; 18: 332-340Crossref PubMed Scopus (80) Google Scholar The comments of Sommer and Van den Brink on the characteristics of our control sample reflect low familiarity with epidemiological case-control designs, which, as we clearly explain in the paper, require controls representing the population at risk of developing the target disorder, rather than controls with other diseases, to produce robust findings. We agree with Clas Linnman that pollution and tobacco smoking might have a causal role in psychotic disorder, although the evidence is much less substantial than that for heavy cannabis use; nor did they explain the excess of psychotic disorder cases we found in northern European cities, compared with case numbers in Spain and Italy. Firstly, we controlled for cigarette smoking. Secondly, although, as Linnman points out, we found an effect of tobacco smoking on the odds ratio for psychotic disorder in the overall sample, this effect was mostly driven by cities in Spain and Italy, where the prevalences of tobacco smoking were among the highest in all of our European sites, both in cases and controls. In our study, the proportions of controls smoking more than 10 cigarettes per day were highest in Madrid (16%) in Spain and Bologna (14%) in Italy, which were around double those in Amsterdam (7%) and London, UK (5%; unpublished). Thirdly, in relation to pollution, the data from the European Environmental Agency referenced by Linnman indicate that Madrid was one of the cities with the highest NOx pollution in Europe during the years preceding illness onset in our patient group. Finally, we agree with Linnman that Δ9-tetrahydrocannabinol is not the only cannabinoid contained in cannabis; however, it is the one consistently linked to psychosis and the one used by international agencies to indicate cannabis potency. Nathan Gillespie and colleagues appear unaware of the large body of evidence from epidemiological, experimental, and neuroimaging studies supporting a causal link between cannabis use and psychosis. Nor do they refer to the study of Boydell and colleagues,4Boydell J van Os J Caspi A et al.Trends in cannabis use prior to first presentation with schizophrenia, in South-East London between 1965 and 1999.Psychol Med. 2006; 36: 1441-1446Crossref PubMed Scopus (68) Google Scholar which provides evidence of an increase in the incidence of psychosis alongside increases in the prevalence of cannabis use.4Boydell J van Os J Caspi A et al.Trends in cannabis use prior to first presentation with schizophrenia, in South-East London between 1965 and 1999.Psychol Med. 2006; 36: 1441-1446Crossref PubMed Scopus (68) Google Scholar Gillespie and colleagues quote findings from Degenhardt et al showing that psychosis incidence in Australia goes against patterns of cannabis use; however, these findings were generated by modelling trends on the basis of assumptions that even Degenhardt and colleagues accept might not be accurate. Although we controlled for the potential effect of other risk factors and used incidence values already adjusted for age and migration, Degenhardt and colleagues assumed that changes in prevalence of cannabis use alone might explain the incidence of psychosis; however, changes in other risk factors that they do not estimate could have also led to decreased cases of psychosis. The Mendelian randomisation studies published to date on the direction of causality between cannabis use and schizophrenia5Vaucher J Keating BJ Lasserre AM et al.Cannabis use and risk of schizophrenia: a Mendelian randomization study.Mol Psychiatry. 2018; 23: 1287-1292Crossref PubMed Scopus (104) Google Scholar do not come to the same conclusions as the Pasman et al study referenced by Gillespie and colleagues. Pasman and colleagues report a causal effect of schizophrenia risk genes on cannabis use, but their study only has data relating to what they define as cannabis initiation (ie, having ever used cannabis, yes or no). However, having used cannabis once is no more useful as a measure of psychosis risk than is a single event of alcohol consumption as a measure of liver disease risk. Thus, the phenotype they use is not comparable to the detailed measures of patterns of use, daily use, and use of high-potency cannabis that we found to affect the incidence of psychotic disorder. In response to Carey Clark, the types of high-potency cannabis available across our European sites were unlikely to be contaminated by the same pesticides or have the same concentrations of metals. We also did control in all our analyses for the potential confounding effect of all of the drugs to which Clark refers. To conclude, most importantly, we found that heavy cannabis use is a modifiable risk factor for psychotic disorders and, as such, a potential target of preventive efforts that might reduce the number of individuals who develop these devastating disorders. MDF reports personal fees from Janssen, outside of the submitted work. RMM reports personal fees from Janssen, Lundbeck, Sunovion, and Otsuka, outside of the submitted work. All other authors declare no competing interests. High-potency cannabis and incident psychosis: correcting the causal assumptionIn their recent paper, Marta Di Forti and colleagues1 conclude that removing one environmental factor—daily high-potency cannabis use—would reduce the incidence of all psychotic disorders in Amsterdam, the Netherlands, by 50%, from 37·9 to 18·8 cases per 100 000 person-years. We think that this is very unlikely given that Sullivan and colleagues2 confirmed the heritability of schizophrenia to be about 80%. Therefore, attributing this complex multifactorial brain disorder to one environmental factor such as high-potency cannabis use seems counterintuitive, especially given that 33·6% of the patients assessed by Di Forti and colleagues had never used cannabis. Full-Text PDF High-potency cannabis and incident psychosis: correcting the causal assumptionMarta Di Forti and colleagues1 claim that the frequency of cannabis use and cannabis potency are responsible for substantial variation in the incidence of psychotic disorders. The authors assume that cannabis causes psychosis or psychotic symptoms without acknowledging compelling, alternative hypotheses.2 Most reports examining associations between cannabis and psychosis have been unable to adjust for confounding that arises from correlated genetic and environmental individual differences. This oversight includes the common omission of appropriate methods for resolving causality (eg, random assignment to case and control conditions, discordant twin pairs, propensity score matching, or recently advanced genome-based restricted maximum likelihood methods). Full-Text PDF High-potency cannabis and incident psychosis: correcting the causal assumptionMarta Di Forti and colleagues1 found that daily cannabis consumers have increased odds of developing a psychotic disorder (odds ratio [OR] 3·2), with even higher risk for high-potency cannabis users (OR 4·8), suggesting that Δ9-tetrahydrocannabinol (THC) is a causal factor. Full-Text PDF The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): a multicentre case-control studyDifferences in frequency of daily cannabis use and in use of high-potency cannabis contributed to the striking variation in the incidence of psychotic disorder across the 11 studied sites. Given the increasing availability of high-potency cannabis, this has important implications for public health. Full-Text PDF Open AccessHigh-potency cannabis and incident psychosis: correcting the causal assumptionPsychosis has many different causes and I have concerns with the methods used in the Article by Marta Di Forti and colleagues,1 which aimed to correlate the use of cannabis of high Δ9-tetrahydrocannabinol percentage with the onset of psychotic disorders. Full-Text PDF