The apolipoprotein E (APOE) gene is strongly associated to the risk of Alzheimer's disease (AD). More specifically, it seems undisputed that the APOE*4 allele plays an important role in the pathogenesis of AD. However, does this imply that it is ApoE4 which causes the increased susceptibility for AD? Empirical findings in non-Caucasians leave space for at least one alternative hypothesis: not the ApoE4 polymorphism but other defects in, or close to the APOE gene, could be responsible. We discuss this hypothesis, based on population and evolutionary genetic evidence.