Mortality, dementia, and apolipoprotein E genotype in elderly white women in the United States.

OBJECTIVES To assess the risk of death in relation to apolipoprotein E (APOE) genotype and to evaluate how APOE genotype interacts with dementia and with other major medical conditions to affect survival. DESIGN A 6-year prospective cohort study of dementia, APOE genotype and survival. SETTING Health maintenance organization in southern California. PARTICIPANTS One thousand eight hundred forty-two white women aged 75 and older. MEASUREMENTS Dementia was determined using a multistage assessment procedure, medical record, and death certificate review. RESULTS With women with the APOE 3/3 genotype as the referent, age-adjusted hazard ratios (HRs) for death according to genotype were 1.25 (95% confidence interval (CI)=1.00-1.56) for APOE 2/4, 3/4, or 4/4 and 0.83 (95% CI=0.62-1.13) for APOE 2/3 or 2/2. Survival was associated with APOE genotype (log rank test P=.02). Women with the APOE 2/4, 3/4, or 4/4 genotype died at an earlier age, and those with APOE 2/2 or 2/3 died later than those with the APOE 3/3 genotype. After adjustment for age, education, and hormone use, HRs for death were significantly higher in women with the APOE 2/4, 3/4, or 4/4 genotype who developed dementia (HR=3.74; 95% CI=2.81-4.99) and the APOE 2/3 genotype (HR=3.23; 95%=CI=1.97-5.28) than in women without dementia and the APOE 3/3 genotype. The HRs for death were greater with other medical conditions, but no interaction with any APOE genotype was found. CONCLUSION In this population of elderly women, although having at least one epsilon4 allele increased the chances of an earlier death, having dementia increased the risk of death regardless of APOE genotype.