In the United States during the first quarter of 2016, Professor’s Charles Ksir (University of Wyoming, Psychology) and Carl L. Hart (Columbia University, New York, Psychology) read with intense interest the meta-analysis by Tabea Schoeler (Kings College, London, United Kingdom) and colleagues, on continued Cannabis use in patients with psychosis. They applauded the authors for covering this timely, important issue, commending them for attempting to provide empirical evidence to inform public policy. However, their enthusiasm was dampened because the interpretation extended beyond the available data.
They noted that it is of utmost importance to remember the meta-analysis was based on correlational studies. Each study pointed out causation has not been shown; however, a strong tendency exists to accept Cannabis use as a so-called component cause of psychosis, which then leads to the conclusion it is imperative to reduce Cannabis use in patients with, or at risk, for psychosis. Although the Professor’s understood this impulse is motivated by a concern for public health, they agreed it should not allow the consistency of these correlational findings to substitute for actual evidence of causality. In 2016, Professor’s Ksir and Hart published a critical review of the scientific literature on Cannabis and psychosis and concluded that the literature supports the hypothesis that both psychosis and Cannabis use are more likely in individuals with a shared vulnerability to misuse of various substances and increased risk for various mental disorders. In other words, the correlation between Cannabis use and psychosis is not specific, either with regard to the chemicals found in Cannabis or to psychosis as opposed to other disorders.
Schoeler and colleagues stated that rates of Cannabis use in patients with psychosis are “higher than … those of people with other psychiatric diagnoses”. To support this statement the authors cited an article by Agosti and colleagues, even though Agosti and colleagues clearly concluded, “Alcohol dependence, antisocial personality disorder and conduct disorder had the strongest associations with Cannabis ‘dependence’, followed by anxiety and mood disorders”. They did not report any association between Cannabis and psychosis, presumably because of the low frequency of psychosis in the participants studied. In their own review, Professor’s Ksir and Hart included seven studies published between 2013 and 2016 that provided information on the issue of specificity. After reviewing the scientific literature they found evidence that bipolar disorder, anxiety disorder and mood disorder have all been correlated with Cannabis use and reported that psychosis has been correlated with heavy tobacco smoking, heavy alcohol use, stimulant misuse and sedative misuse. They found no clear evidence for a causal relation between Cannabis and psychosis.
The Professors’ reviewed research reports on Cannabis and psychosis, giving particular attention to how each report provided evidence relating to two hypotheses:
1) Cannabis as a contributing cause, and
2) shared vulnerability.
Two primary kinds of data are brought to bear on this issue: studies done with schizophrenic patients and studies of first-episode psychosis. Evidence reviewed suggests that Cannabis does not in itself cause a psychosis disorder. Rather, the evidence leads to the conclusion that both early use and heavy use of Cannabis are more likely in individuals with a vulnerability to psychosis. The role of early and heavy Cannabis use as a prodromal sign (early symptom that may mark onset of a disease) merits further examination, along with a variety of other problem behaviours (e.g., early or heavy use of cigarettes or alcohol and poor school performance).
According to their shared vulnerability hypothesis, in a given group of Cannabis users who have had psychotic episodes, the individuals with the greatest degree of the shared vulnerability would be the most likely to continue Cannabis use rather than to discontinue and they would be the most likely to have recurring episodes of psychosis and require more hospital treatment. As such, these two outcomes should be correlated, even if neither is a cause of the other. As to whether a public health benefit can be obtained from efforts to reduce Cannabis use in patients with psychosis, two randomised controlled trials, published in 2013, comparing treatment as usual with treatment as usual plus motivational interviewing and cognitive behaviour therapy that concentrated on Cannabis use, found no beneficial effect of either intervention on either psychotic symptoms or amount of Cannabis use.
Interest in the relationship between Cannabis use and psychosis has increased dramatically in recent years, in part because of concerns related to the growing availability of Cannabis and potential risks to health and human functioning. There now exists a plethora of scientific articles addressing this issue, but few provide a clear verdict about the causal nature of the Cannabis-psychosis association. The Professors’ greatest concern is not that someone might be advised to stop using Cannabis. They are concerned that a misunderstanding of the relation between Cannabis use and psychotic behaviour leads to an oversimplification of the complex developmental nature of substance use and mental disorders. Furthermore, they proposed that future studies that limit their data collection to focus exclusively on the Cannabis-psychosis association only will do little to enhance understanding of the complexity of this comorbidity. Research studies of this ilk will therefore be of little value in the quest to better understand psychosis and how and why it occurs.
“I have to make sure I don’t engage in conversations with
people who don’t abide by the rules of evidence”, Carl Hart
Adapted from Correspondence in the Lancet, Psychiatry, May 2016, and Cannabis and Psychosis: a Critical Overview of the Relationship