The February 2004 issue contained a good review of evidence linking cannabis use to risk for developing schizophrenia (Reference Arseneault, Cannon and WittonArseneault et al, 2004). Three plausible causal explanations for this association are given. First, that cannabis and/or related drug use is a causal factor for schizophrenia. Second, that the altered mental state induced by cannabis may be mistaken for schizophrenia. Third, that cannabis use may be increased in individuals with the premorbid features of schizophrenia. Arseneault et al believe that the evidence favours the first alternative, and we agree. However, we call attention to a fourth possibility. Consider two propositions: (a) features of schizophrenia such as negative symptoms and cognitive impairments precede the onset of psychosis and are considered early morbid rather than premorbid; and (b) schizophrenia is associated with high rates of substance misuse. The cause of substance misuse in schizophrenia is not known. We suggest a fourth hypothesis to explain the cannabis/schizophrenia association. Substance misuse may be a morbid manifestation of some forms of schizophrenia. Vulnerability to substance use may be considered similar to vulnerability to psychosis.

The data review by Arseneault et al suggests that the cannabis/schizophrenia association is not based on shared genetic vulnerability. This is of interest to us in that a rodent model of schizophrenia has been developed by one of us (J.I.K.) based on the application of repeated stresses to pregnant rats during the rat equivalent to the second trimester of human pregnancy. The offspring of the stressed dams, once achieving adulthood, manifest the following schizophrenia-like behaviours: diminished cognitive ability on a hippocampal-dependent memory task; impaired gating of event-related potentials and sensory information; augmented behavioural responses to psychostimulants; social apathy and incompetence (Reference Koenig, Elmer and BradyKoenig et al, 2001; further details available from the authors on request).

In addition, adult rats exposed to stressful gestation consume alcohol in excess compared with control animals. We therefore raise the possibility that aspects of the non-genetic environment may contribute simultaneously to increased risk for cannabis use and increased risk for schizophrenia diathesis.

Declaration of interest

A research contract from Novartis Pharma, AG supported development of the rat model.