Some dementias, such as Creutzfeldt-Jakob disease, may result from tra
nsmissible agents. To assess the possibility of a blood-borne transmis
sible agent in Alzheimer's disease (AD), we compared informant-reporte
d history of blood transfusions among 326 newly recognized probable AD
cases and 330 control subjects of similar age and sex, randomly selec
ted from the same health maintenance organization population. Since ri
sk of AD is increased in persons with apolipoprotein E epsilon 4 (APOE
-e4) alleles, we examined whether the transfusion-AD relationship vari
es by APOE genotype. Transfusion was more frequent among controls (17.
9%) than cases (12.0%) in the overall group (crude odds ratio = 0.62;
95% confidence interval = 0.40, 0.97). Similar risks were seen for tra
nsfusions occurring less than or equal to or > 10 years before the ref
erence year. Stratified analysis by APOE-e4 genotype showed no effect
modification. Logistic regression adjusting for age, sex, education, a
nd presence of e4 did not change the transfusion-AD association. We co
nclude that a history of blood transfusion is not associated with incr
eased risk of AD, nor does consideration of APOE-e4 status modify that
conclusion.