Numerous studies have found that psychotic patients are more likely to consume cannabis than the general population but, until recently, many psychiatrists believed that their patients were taking the drug to counteract negative symptoms or medication effects. In 1987, a follow-up of 50,000 conscripts into the Swedish Army revealed that those who admitted at age 18 to having taken cannabis on more than 50 occasions were 6 times more likely to develop schizophrenia in the ensuing 15 years. These findings were largely ignored but in the last 5 years, 6 other studies have confirmed that cannabis consumption acts to increase later risk of schizophrenia. Firstly, a Dutch study of some 4,000 people in the general population showed that those taking large amounts of cannabis at the initial interview were almost 7 times more likely to have psychotic symptoms three years later.

Critics argued that the findings of the Swedish and Dutch studies could have been caused by those individuals who were already odd and destined to develop schizophrenia, being more likely to use cannabis. Other studies have excluded this. For example, an extension of the Swedish Army study demonstrated that the results held even when initial personality, and use of other drugs, were taken into account. Then, a birth cohort study in Dunedin, New Zealand, found that cannabis use at 15 years carried an odds ratio of 4.5 for developing schizophreniform psychosis by age 26 years; when the presence of quasi-psychotic ideas at age 11 years was taken into account, the risk of schizophrenic symptoms at 26 was diminished but remained significant (Arseneault e.a. 2002). Some individuals appear more vulnerable than others; thus, individuals with the val/val genotype of the comt gene (which metabolises dopamine in the frontal cortex) appeared to be more susceptible than those with the met/met genotype (which results in slower metabolism of frontal dopamine) (Caspi e.a. 2005). Then another Dutch study (Henquet e.a. 2005) and a New Zealand Cohort study (Fergusson e.a. 2005) showed that the direction of effect is from cannabis consumption to psychotic symptoms and not the other way round.

A review by Arseneault e.a. (2004) concluded that cannabis use was a modest contributory cause of schizophrenia, and finally a systematic review from an Edinburgh group (Semple e.a. 2005) concluded that 'the available evidence supports the hypothesis that cannabis is an independent risk factor both for psychosis and the development of psychotic symptoms'.

Now that the epidemiological data are well established, we need to: (a) understand the brain mechanisms involved, and (b) develop educational campaigns aimed at lessening heavy consumption of cannabis, especially its more potent varieties.