Aims The aim of the study was to examine the temporal associations between substance use and subclinical psychosis symptoms.
Design Data from a prospective community study sampled within a single cohort over 30 years (1978–2008) were analysed with discrete-time hazard models.
Setting General population-based sample.
Participants At initial sampling in 1978 males (n = 292) were 19 and females (n = 299) were 20 years old.
Measurements Two psychosis syndromes representing ‘schizotypal signs’ and ‘schizophrenia nuclear symptoms’ and various substance use variables including cannabis, alcohol, tobacco and multiple-drug use (i.e. cannabis combined with other drugs).
Findings In bivariate analyses, schizotypal signs were predominantly associated with regular cannabis use in adolescence (OR = 2.29, 95% CI 1.32–3.97). Schizophrenia nuclear symptoms were mainly related to alcohol (OR = 1.84, 95% CI 1.00–3.38) and multiple-drug use (OR = 2.35, 95% CI 1.38–4.02) during adolescence. Multivariate analyses showed that, in particular, regular cannabis use during adolescence was associated with the occurrence of subsequent schizotypal symptoms over a 30-year period (OR = 2.60, 95% CI; 1.59–4.23), whereas multiple-drug use in adolescence was associated predominantly with schizophrenia nuclear symptoms (OR = 1.75, 95% CI 1.01–3.03). Alcohol misuse was only slightly associated with the onset of such symptoms.
Conclusions A significant portion of the occurrence of subclinical psychosis symptoms in adulthood can be attributed to excessive cannabis and multiple-drug use during adolescence. This is in line with the hypothesis that long-term sensitization of dopaminergic brain receptors plays a role in developing psychotic symptoms.