LACK OF ASSOCIATION OF APOLIPOPROTEIN-E ALLELE EPSILON-4 WITH LATE-ONSET ALZHEIMERS-DISEASE AMONG FINNISH CENTENARIANS

No association between Alzheimer's disease (AD) and apolipoprotein E t ype epsilon 4 (ApoE epsilon 4) phenotype was found among centenarians in Finland (N = 179). The data are based on ascertainment of all cente narians in Finland in 1991. All examinations were conducted during 199 1. The diagnoses of dementia and AD were based on clinical grounds, co nforming to DSM-III-R and NINCDS-ADRDA criteria, The percentage of Apo E epsilon 4 alleles among the centenarians was 8.7% (31 of 358 alleles ). This is significantly lower than percentages found in younger Finni sh populations, Thirty (16.8%) of the 179 centenarians were epsilon 4 allele carriers, One hundred fifty-one (84.4%) of the centenarians wer e women. Twenty-eight (18.5%) of the women had at least one epsilon 4 allele, as did two (7.1%) of the men. The prevalence of clinically dia gnosed AD was 26.8%; 44% of the subjects were cognitively normal, 23% had signs of cognitive decline or at most mild dementia (with no diffe rential diagnosis), and 6% had a dementia clinically diagnosed as bein g due to some cause other than AD. For AD cases versus cognitively nor mal subjects, the odds ratio associated with being a carrier of the ep silon 4 allele was 1.34 (p = 0.64; 95% CI = [0.5, 3.3]). Among women, the odds ratio was 0.99 (p = 1.0; 95% CI = [0.4, 2.6]). There were few er, but not significantly so, epsilon 4 carriers among subjects with c ognitive decline or at most mild dementia (12.2%) than there were amon g the cognitively normal subjects (16.5%). Ten (10.4%) of the 96 allel es belonging to AD cases were epsilon 4, and 8.9% (14/158) of the alle les belonging to the cognitively normal subjects were epsilon 4. This difference is highly nonsignificant. There was only one individual who was homozygous for the epsilon 4 allele. She was cognitively normal. Among the 28 men in the study there were only two carriers of the epsi lon 4 allele; consequently, analyses for men have little power. These results clearly show that the epsilon 4 allele does not necessarily le ad to AD even near the (current) upper age-limit of life. When combine d with previous findings, these results suggest that the association o f the ApoE epsilon 4 allele with AD may be age-dependent and that the ApoE epsilon 4 allele might accelerate the AD dementing process rather than be a direct etiologic agent or a predisposing genetic factor.