• schizophrenia;
  • adolescent;
  • cohort study;
  • maternal behaviour

Objective:  It has been suggested that influences operating early in life may affect the risk of postpubertal psychosis outcomes. This hypothesis was tested using a broad outcome of psychotic symptoms expressed in adolescence (prevalence: 15.6%).

Method:  Findings are based on a longitudinal, population-based cohort study of 963 adolescents aged 15–20 years and their parents in the area of Munich, Germany. Trained psychologists assessed adolescents with the Munich-Composite International Diagnostic Interview. Independently, direct diagnostic interviews were conducted with the parents.

Results:  A range of medical complications of pregnancy and delivery, including lower birth weight, were not associated with the psychosis outcome. However, a number of maternal health behaviours and experiences did show associations, independent of confounders.

Conclusion:  Not maternally reported medical complications of pregnancy and delivery, but maternal prenatal health behaviours predicted expression of psychosis along a continuum in adolescence. This effect may either be direct or constitute a proxy for later postnatal maternal behaviours associated with psychosis risk in the offspring.