Contributing cause hypothesis

The first cohort study to find that cannabis use may predict subsequent development of psychosis was conducted by Andreasson and colleagues and was based on a sample of 45, 570 male Swedish conscripts (1987). They were assessed at age 18 for substance use and psychiatric diagnosis and were then followed up during a 15-year period for inpatient admission for schizophrenia. Results showed that those individuals who had used cannabis presented a risk for schizophrenia which was 2.4 times higher than non-users and that number raised to a risk which 6.0 times greater if the individuals had used the drug more than 50 times.

Though these results were robust, many critics pointed out the methodological flaws in the study and the question still remained whether there was actually a causal link between cannabis and psychosis. Some of the main criticisms included:

»      The fact that this study used self-administered questionnaires to assess the individuals' cannabis consumption at the time they were drafted and didn't document whether this consumption persisted over the following 15 years

»      That the individuals' diagnoses of schizophrenia were established by simply looking the medical records and could have been confused with cannabis-induced psychoses

»      Other substance use was not taken into account and thus some results could have been attributed to consumption of other drugs and/or alcohol

»      Personality traits which may be known to predispose individuals both to develop psychosis and to use cannabis or other substances were not measured and thus it was difficult for this study to establish whether consumption of the cannabis had triggered the psychosis or whether the drug use was actually a cause of the disease.

As a result of these criticisms scientists got to work on developing more studies that would address these flaws. Through the growing numbers of these studies though, more and more support seemed to be provided for the hypothesis that cannabis may play a key role in the development of chronic psychosis.

For example, In 2002 Zammit and his colleagues reanalyzed, expanded and updated the data from Andreasson's longitudinal study and concluded not only that cannabis use was associated with an increased risk of developing schizophrenia but also that this risk was dose-dependent, meaning that the higher the amounts used, the higher the risk would be for developing the disease. They also noted that this link could be explained by the use of other drugs or by personality traits

Another longitudinal study, also in 2002, was conducted by Van Os et al., which conducted a baseline evaluation of 4104 individuals ages 18 to 64 and followed them for 3 years. The study found that cannabis use increased the risk of psychosis whether they had established psychotic vulnerability or not, though the risk was particularly high for people with an already established psychotic vulnerability. They also found a dose dependent relationship with these results and reported that the association between cannabis use and psychosis was independent of the use of other drugs both at the time of the baseline study and during the follow up period.

Many opponents of this position argue that if cannabis use really does increase the risk of psychosis then shouldn't there be a corresponding number of diagnoses of psychosis in places where cannabis use is particularly high for example?  So, for example if cannabis use is high in Canada then the logic would follow that the number of diagnoses of psychosis should also be relatively high. This simply isn't the case as the percentage of the population who are diagnosed with psychosis is relatively consistent in a Western- approximately 3 people out of 100 or about 3 % while the percentage of people who use cannabis tends to vary depending on the location.  Degenhardt et al (2003) noted for example that although there has been a significant increase in terms of cannabis use in Australia in the last few decades combined with a decline in the initial age of first use, the number of cases of schizophrenia remains relatively consistent. They did note that cannabis can cause psychosis in those individuals who are already vulnerable to this disease (for example, they have a history of mental illness in their family) and can worsen the trajectory of the illness in individuals who are already diagnosed with psychosis.

Along this same vein there are many studies that suggest that using cannabis may increase the number of psychotic symptoms or exacerbate them without necessarily increasing the number of diagnoses of psychosis itself. Numerous studies have reported that cannabis use is a significant risk factor for relapse of psychosis (Hunt, Bergen & Bashir, 2002; Liszen et al., 2003;Lewis, Tarrier & Drake, 2005;  Mamali et al., 2006; Wade et al., 2006 etc.) as well as also being linked to poor treatment compliance, more inpatient stays and poor prognosis in general.

More specifically, in a prospective New Zealand study, authors Arsenault et al (2002) found that individuals who had used cannabis at least three times between the ages of 15 and 18 had an increased risk of developing schizophreniform symptoms by the age of 26 and that earlier use produced a higher risk than later use. However, the authors did not observe an increased incidence of diagnoses of psychosis in the 759 individuals who were included in the study.

Fergusson et al (2003) also noted that higher rates of psychotic symptoms were present in individuals who had previously been regular cannabis users though not to the point of increase in the frequency of diagnoses of psychosis. They followed up on these findings in 2005 to show that direction of causality is likely to be the path from using cannabis to an increase in psychotic symptoms.

Next >