Objectives: We investigated whether apoE genotypes correlate with cognitive functions in clinically healthy persons.
Methods: In 1993 and 1995, we measured information processing speed, delayed free recall and semantic aspects of long-term memory in 227 men and 105 women aged 65 and over, a randomly selected subsample of the prospective Basel Study. Cardiovascular risk factors and education were assessed.
Results: E2 were more prevalent in old-old (>75 years, 23.5% vs. 15%) compared to E4 than in young-old (<75 years, 19.3% vs. 23.5%). Taking into account age and education, subjects with epsilon3/epsilon4 or epsilon4/epsilon4 alleles (E4) performed lowest in all 3 tests compared to those homozygous for epsilon3 (E3) or carriers of one or two epsilon2 alleles (E2) (reaction time P = 0.009, free recall P = 0.05, WAIS-R vocabulary P<0.05). In old-old there was a significant difference between E2 and E4 for reaction time (P = 0.02) and free recall (P<0.02) but not for vocabulary (P = 0.086). In all 3 groups there were no significant changes after 2 years. The subgroup with the genotype epsilon2/epsilon4 performed consistently best in the cognitive tests. Cholesterol was significantly increased in the E4 and E3 group compared to the E2 group.
Conclusion: ApoE genotype correlates with cognitive performance. The increased prevalence of E2 in the old-old and the significantly lower plasma cholesterol levels suggest differential morbidity and mortality as important factors influencing the prevalence of cognitive disorders in late life.