[Apolipoprotein E genotype as a risk factor in Japanese patients with early-onset and late-onset Alzheimer's disease].

Recent studies have provided evidence of an association of apolipoprotein E (apoE) epsilon 4 allele and late-onset sporadic Alzheimer's disease (AD). Some studies have shown the possibility that apoE epsilon 4 is a risk factor of developing AD in early-onset type. We have analyzed the apoE gene polymorphism in a sample of 310 Japanese AD subjects and 237 age-matched Japanese controls. We divided the sporadic AD patients into two subgroups of 237 late-onset (> 65 years) and 73 early-onset (< or = 65 years) patients, and into three subgroups according to their apoE genotype, no epsilon 4, one epsilon 4, and two epsilon 4 alleles. Our data confirmed an association between epsilon 4 allele and early-onset AD and late-onset AD. The odds ratios (95% confidence interval) referred to no epsilon 4 allele for AD were 3.4 (1.7-7.0) for one epsilon 4 allele and 20.3 (2.5-166.6) for two epsilon 4 alleles in early-onset type, and 6.7 (3.9-11.3) for one epsilon 4 allele and 19.0 (2.5-145.6) for two epsilon 4 alleles in late-onset type. These ratios were significantly increased in both early-onset AD and late-onset AD. Kaplan-Meier survival analysis, which estimates the age of onset for subjects with no, one, and two epsilon 4 alleles in early-onset and late-onset type, revealed a significant dose effect where each additional epsilon 4 allele made the age of onset earlier (p < 0.0001). The age of onset is 9.7 years earlier for two epsilon 4 bearers and 3.9 years earlier for one epsilon 4 bearers than no epsilon 4 bearers in late-onset AD, 2.9 years earlier for two epsilon 4 bearers and 1.4 years earlier for one epsilon 4 bearers than no epsilon 4 bearers in early-onset AD. Moreover, we studied an association between apoE epsilon 2 allele and early-onset AD and late-onset AD. There was a significantly decreased frequency of apoE epsilon 2 allele in patients with late-onset AD (p = 0.026), although the frequency of apoE epsilon 2 was not changed significantly in early-onset AD (p = 0.360). The odds ratios referred to no epsilon 2 allele for AD were 1.9 (0.6-5.7) for one epsilon 2 allele in early-onset type, and 0.4 (0.2-0.9) for one epsilon 2 allele in late-onset type. Our study suggested the difference in the effect of apoE genotype on developing AD between early-onset and late-onset type in Japanese patients.