Objective: To test the hypothesis that aspirin and non-steroidal anti-inflammatory drugs (NSAIDs) may prevent dementia or cognitive impairment.
Design: A two-wave longitudinal study over 3.6 years.
Setting: A community survey of elderly persons living in Canberra, Australia.
Participants: There were 1045 elderly persons aged 70 at the start of the study; cognitive assessment was obtained at both waves on 588.
Main outcome measures: Cognitive functioning was measured using the Mini-Mental State Examination, an episodic memory test, a test of mental speed and the National Adult Reading Test. Dementia was assessed using the Canberra Interview for the Elderly.
Results: On cross-sectional data, those who had been taking NSAIDs or aspirin performed no better on the cognitive tests after account had been taken of other confounding variables. There was no interaction with apolipoprotein E genotype. On longitudinal data, no difference was found between NSAID or aspirin users and controls, either in cognitive decline or incidence of dementia.
Conclusions: The results do not support the hypothesis that aspirin or NSAIDs have a protective effect, but it remains possible that various sources of measurement error may have attenuated an effect of clinical significance from either type of drug. Conclusive evidence can be obtained only by a prospective trial.