Using meta–analysis to explain the diversity of results in genetic studies of late–onset Alzheimer’s disease and to identify high–risk subgroups

Abstract

In late-onset Alzheimer’s disease, there is a puzzling inconsistency between the findings of case-control studies of most proposed risk genes, except apolipoprotein E ϵ4. This inconsistency may stem from the failure to define the genetic and non-genetic interactions that affect the disease association of each particular susceptibility gene. Such interactions will limit the influence of the gene to a ‘relevant subset’ of vulnerable people. The relevant subsets for many risk genes will be narrow, compared to that of apolipoprotein E ϵ4. Studies may therefore miss the association or even suggest that a risk gene is protective. In these circumstances, the precise composition of a cohort is critical and defining the relevant subset is crucial. We illustrate how such definition may be achieved through meta-analysis. We take as an example the butyrylcholinesterase K variant, whose association with Alzheimer’s disease may now be provisionally defined. This analysis leads to the identification of a potentially high-risk group: over 75 year old male carriers of both apolipoprotein E ϵ4 and butyrylcholinesterase K variant.