Background. Clinical studies suggest that psychotic and paranoid states in late life are associated with cognitive dysfunction. However, it is not clear whether this finding would be observed in general population samples of non-demented elderly, particularly after adjustment for potential confounding factors.

Method. A representative sample of non-demented 85-year-olds living in the community or in institutions in Göteborg, Sweden (N=347) was examined using a psychiatric and physical examination (including a medical history), key-informant interview, psychometric testing and review of medical records. Individuals with psychotic symptoms and paranoid ideation were compared with the mentally healthy regarding tests of verbal ability, inductive logical reasoning, spatial ability, perceptual speed, basic arithmetic, primary memory and secondary memory.

Results. Non-demented 85-year-olds with psychotic symptoms or paranoid ideation performed specifically worse on tests measuring verbal ability, logical reasoning and two tests of spatial ability after adjustment for sex, education, hearing impairment, visual deficits, somatic disorders, depression, 3-year-mortality rate and incident dementia.

Conclusions. Psychotic symptoms and paranoid ideation were associated with lower performance on cognitive tests related to verbal ability, logical reasoning and spatial ability in non-demented 85-year-olds after adjustment for potential confounders.