Introduction. To trace the prodromal evolution of Alzheimer's disease using a comprehensive computerized cognitive battery and an evaluation of performance of everyday activities.
Materials and methods. A general population cohort of 397 normal elderly persons hypothesized to have an increased risk of developing Alzheimer's disease was examined over a 3-year period. Twenty-two subjects developed AD in the third year and their performance was compared retrospectively with 150 subjects who had remained free of senile dementia and other psychiatric disorders throughout the period of the study.
Results. Adjusting for eects of age, education and incident illness, signiยฎcant dierences were found between the two groups on a wide range of tests 2 years before diagnosis. For half of these tests the decline was largely limited to persons with low levels of education. Education eects closer to the time of diagnosis were limited to a small number of verbal tasks.
Conclusion. Patterns of cognitive impairment may be detected in cases of AD 2 years before formal clinical diagnosis, although high education level prolongs competence, principally on verbal tasks. The results suggest that a number of brief cognitive tests may have face validity and high discriminability independent of education level in the evaluation of cognitive complaints in early stage AD in the general practice setting.