Familial risks for Alzheimer disease from a population-based series.

Kaplan-Meier risks estimates are calculated and compared for two consecutive series (N1 = 840, N2 = 819) of first-degree relatives of Alzheimer disease (AD) patients diagnosed as either "probable" or "autopsy-confirmed" AD. The consistency of results increases confidence in estimates and suggests consistent case ascertainment over 8 years. Lifetime risk estimates to age 88 for the combined sample (23.4% +/- 3.0%) do not approach the 50% risk compatible with an autosomal dominant model of transmission. These results support our previous finding and suggest that an autosomal dominant gene(s) is not responsible for all cases of AD.