Background: Recently, GAB2 has been suggested to modify the risk of late-onset Alzheimer's disease (AD) among APOEepsilon4 carriers. However, replication data are inconsistent.
Methods: In a population-based cohort study (n = 5507; age > 55) with 443 incident AD cases, we investigated the association between rs4945261 and AD. Because we used high-density genotyping, we also investigated other polymorphisms within and around GAB2 and performed a meta-analysis with published studies.
Results: We found that rs4945261 was associated with AD among APOEepsilon4 carriers (p = .02) but not among noncarriers (p = .26). Fifteen of the 20 remaining polymorphisms within GAB2 and several polymorphisms in the 250kbp-region surrounding GAB2 were also associated with AD among carriers and only one among noncarriers. For rs2373115, meta-analysis yielded an odds ratio of 1.58 (1.17-2.14) with p = 3.0 * 10(-3) among carriers and 1.09 (.97-1.23) with p = .16 among noncarriers. For rs4945261, the pooled odds ratio was 1.75 (1.21-2.55) with p = 3.0 * 10(-3) among carriers and 1.20 (1.01-1.41) with p = .03 among noncarriers.
Conclusions: We found GAB2 to be associated with AD. Furthermore, the meta-analysis also suggests that GAB2 modifies the risk of AD in APOEepsilon4 carriers.