Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta Analysis Consortium.

OBJECTIVE To examine more closely the association between apolipoprotein E (APOE) genotype and Alzheimer disease (AD) by age and sex in populations of various ethnic and racial denominations. DATA SOURCES Forty research teams contributed data on APOE genotype, sex, age at disease onset, and ethnic background for 5930 patients who met criteria for probable or definite AD and 8607 controls without dementia who were recruited from clinical, community, and brain bank sources. MAIN OUTCOME MEASURES Odds ratios (ORs) and 95% confidence intervals (CIs) for AD, adjusted for age and study and stratified by major ethnic group (Caucasian, African American, Hispanic, and Japanese) and source, were computed for APOE genotypes epsilon2/epsilon2, epsilon2/epsilon3, epsilon2/epsilon4, epsilon3/epsilon4, and epsilon4/epsilon4 relative to the epsilon3/epsilon3 group. The influence of age and sex on the OR for each genotype was assessed using logistic regression procedures. RESULTS Among Caucasian subjects from clinic- or autopsy-based studies, the risk of AD was significantly increased for people with genotypes epsilon2/epsilon4 (OR=2.6, 95% CI=1.6-4.0), epsilon3/epsilon4 (OR=3.2, 95% CI=2.8-3.8), and epsilon4/epsilon4 (OR=14.9, 95% CI= 10.8-20.6); whereas, the ORs were decreased for people with genotypes epsilon2/epsilon2 (OR=0.6, 95% CI=0.2-2.0) and epsilon2/epsilon3 (OR=0.6, 95% CI=0.5-0.8). The APOE epsilon4-AD association was weaker among African Americans and Hispanics, but there was significant heterogeneity in ORs among studies of African Americans (P<.03). The APOE epsilon4-AD association in Japanese subjects was stronger than in Caucasian subjects (epsilon3/epsilon4: OR=5.6, 95% CI=3.9-8.0; epsilon4/epsilon4: OR=33.1, 95% CI=13.6-80.5). The epsilon2/epsilon3 genotype appears equally protective across ethnic groups. We also found that among Caucasians, APOE genotype distributions are similar in groups of patients with AD whose diagnoses were determined clinically or by autopsy. In addition, we found that the APOE epsilon4 effect is evident at all ages between 40 and 90 years but diminishes after age 70 years and that the risk of AD associated with a given genotype varies with sex. CONCLUSIONS The APOE epsilon4 allele represents a major risk factor for AD in all ethnic groups studied, across all ages between 40 and 90 years, and in both men and women. The association between APOE epsilon4 and AD in African Americans requires clarification, and the attenuated effect of APOE epsilon4 in Hispanics should be investigated further.