Background: Female gender and later onset of psychosis are both associated with better outcome. However whether their effects are independent, is not known.
Method: In 379 incident cases of psychoses, from an epidemiologically defined catchment area, admixture analysis was employed to generate age of onset classes. Five year course and outcome measured across clinical and social domains were used as dependent variables in regression analyses, to estimate associations of outcomes with gender, age of onset and gender by age of onset interaction.
Results: Three age of onset classes were identified: early (14-41 years), late (42-64 years) and very late onset psychosis (65-94 years). Overall, women had better outcomes, including milder delusions, fewer negative symptoms, less deterioration from baseline functioning, fewer hospital readmissions and shorter psychotic episodes. Later age of onset was also associated with better outcome, although in the very late onset class the results were mixed. There was a statistically significant gender by age of onset interaction (in the ratio scale) within this sample with men displaying poorer outcome in the early/late onset class, whereas women tended to have a worse outcome in the very late onset class.
Conclusions: The favourable outcome in women becomes reversed in old age, suggesting gender-age-related differences in the distribution of aetiological factors for psychosis.