Cannabis,stimulants and psychosis. Commentary on Gururajan et al. (2012): Drugs of abuse and increased risk of psychosis development

Does illicit drug use cause psychosis? In this issue, Gururajan and colleagues (2012) review recent evidence relevant to this question. Their review is timely given current drug policy debates.

There is a strong association between illicit drug use and psychosis. Possible explanations for this association include causation (drugs cause psychosis), self-medication (psychosis leads to drug use) and confounding (shared vulnerabilities underlie both drug use and psychosis). There is increasing evidence that illicit drugs play a causal role in at least some individuals with psychosis, as follows.

Drugs can produce psychotic symptoms in experimental studies

Administration of cannabis (D’Souza et al., 2004) and stimulants (Angrist et al., 1974) to healthy volunteers can produce transient psychotic symptoms. Gururajan and colleagues (2012) summarise animal studies demonstrating the effects of cannabis or stimulant administration on memory, response to novel stimuli, sensory ‘gating’, behavioural activity and hedonic response. While animal models for psychosis have significant limitations, it is interesting that these effects are greatest when exposure to drugs occurs during critical developmental stages.

There is a dose effect

A greater quantity of drug use is associated with more psychotic symptoms in experimental studies and in clinical studies with recreational drug users (McKetin et al., 2006), people with first-episode psychoses (Wade et al., 2007), or more chronic psychotic disorders (Foti et al., 2010).

Illicit drug use usually precedes psychosis

Gururajan summarises several cohort studies which have found that cannabis use acts as an independent risk factor and usually precedes the development of psychosis. These results have been replicated in further well-designed cohort studies.

Drug use is associated with earlier onset of psychosis

A recent meta-analysis summarises evidence that use of cannabis and other drugs is associated with a 2–3-year reduction in the average age of onset of psychotic disorders (Large et al., 2011).

The pattern of drug use affects course and outcome in people with psychosis

Ongoing drug use appears to be a major risk factor for the development of more chronic illness. Following a first psychotic episode in young people, cessation of drug use may be associated with better outcomes, increased remission and reduced relapse (Wade et al., 2007).

There is a plausible theoretical mechanism

Abnormalities of dopamine pathways and the essential role of dopamine blockade to treatment response are some of the most consistent findings in our understanding of psychosis. Gururajan summarises a wide range of evidence linking cannabis and stimulants to dopamine pathways. This includes evidence from clinical and animal studies suggesting that genetic variations in monoamine metabolic pathways such as catechol-O-methyl transferase (COMT) may mediate vulnerability to psychosis on exposure to cannabis or amphetamines.

Taken together these findings suggest that cannabis and other drugs can be a causal factor in psychosis, probably by interacting with genetic, neuro-developmental, personal and/or social vulnerabilities in the individual. Cannabis use may double the risk of development of schizophrenia in vulnerable individuals (Hall and Degenhardt, 2011). It has been estimated that from 7% to 25% of incident cases of schizophrenia may be attributable to cannabis (Hickman et al., 2007).

This brief summary of course oversimplifies. Gururajan’s review also highlights the complex and often conflicting nature of the evidence. Shared vulnerabilities to drug use and the development of psychosis may also explain many of the findings summarized above. The main argument that confounding may be more important than causation is the apparent lack of a substantial increase in the incidence or prevalence of psychosis over the last two decades, despite significant increases in population use of cannabis and stimulants. On the other hand, the main argument against confounding as a sufficient model is that well-designed cohort studies have found that drug use remains a risk factor for psychosis even after controlling for many potential confounders. It is likely that the link between drugs and psychosis involves both direct causative effects and shared vulnerabilities.

While some individuals use substances to manage the distress associated with psychotic disorders, self-medication cannot explain the range of findings above. Illicit drugs have a dose-related effect on the presence of psychotic symptoms, but the reverse is not true and the type or nature of psychotic symptoms or distress is a poor predictor of drug use in people with psychosis.

Gururajan supplements other recent reviews (Curran et al., 2004; Hermens et al., 2009) of the relationship between stimulants and psychosis. This relationship is likely to be at least as strong as that for cannabis. Amphetamines are structural analogues and powerful agonists of dopamine. Unlike many other drugs, their effects sensitize rather than habituate with repeated use. However, because of their less frequent use and almost universal overlap with cannabis and other substances, they have been less systematically studied. The suggestive but conflicting findings presented by Gururajan underline the particular need for further research on stimulants.

Cannabis, stimulants and other drugs do appear to play a causal role in psychosis. However, this does not by itself argue either for or against changes to current legal approaches to drug use. Many legal substances cause harm. Proponents of drug law liberalization argue that, amongst other effects, decriminalization may reduce the high cost of illicit drugs and thereby remove a driver for drug-related criminal activity. However, in the absence of other strategies, reduction in cost may lead to increased drug use, with a consequent increase in drug-related psychosis. This may be a small effect at a population level, but not for the young people affected and their families. On the other hand, proponents of prohibition should be challenged by recent modelling suggesting that more than 1300 men aged 20–24 years would need to be prevented from becoming heavy cannabis users in order to prevent a single case of psychosis (Hickman et al., 2009). Evidence about the impact of drug use on psychosis is relevant to current debates on drug policy, and may help to inform some of the difficult trade-offs inherent in any position on this complex problem.

See Review by Gururajan et al., 2012, 46(12): 1120–1135