Testing for Linkage and Association Across the Dihydrolipoyl Dehydrogenase Gene Region with Alzheimer’s Disease in Three Sample Populations
- First Online:
- Cite this article as:
- Brown, A.M., Gordon, D., Lee, H. et al. Neurochem Res (2007) 32: 857. doi:10.1007/s11064-006-9235-3
- 72 Views
Prior case–control studies from our laboratory of a population enriched with individuals of Ashkenazi Jewish descent suggested that association exists between Alzheimer’s disease (AD) and the chromosomal region near the DLD gene, which encodes the mitochondrial dihydrolipoamide dehydrogenase enzyme. In support of this finding, we found that linkage analysis restricted to autopsy-proven patients in the National Institute of Mental Health–National Cell Repository for Alzheimer’s Disease (NIMH–NCRAD) Genetics Initiative pedigree data resulted in point-wise significant evidence for linkage (minimum p-value = 0.024) for a marker position close to the DLD locus. We now report case–control replication studies in two independent Caucasian series from the US and Italy, as well as a linkage analysis from the NIMH–NCRAD Genetics Initiative Database. Pair-wise analysis of the SNPs in the case–control series indicated there was strong linkage disequilibrium across the DLD locus in these populations, as previously reported. These findings suggest that testing for association of complex diseases with DLD locus should have considerable statistical power. Analysis of multi-locus genotypes or haplotypes based upon three SNP loci combined with results from our previous report provided trends toward significant evidence of association of DLD with AD, although neither of the present studies’ association showed significance at the 0.05 level. Combining linkage and association findings for all AD patients (males and females) results in a p-value that is more significant than any of the individual findings’ p-values. Finally, minimum sample size calculations using parameters from the DLD locus suggest that sample sizes on the order of 1,000 total cases and controls are needed to detect association for a wide range of genetic model parameters.