Premium
P1‐062: Strong association of a transthyretin snp with late‐life cognitive impairment and Alzheimer's brain lesions at autopsy: The honolulu asia aging study (HAAS)
Author(s) -
White Lon R.,
Vostrov Alexander,
Zarow Chris,
Sonnen Joshua,
Uyehara-Lock Jane,
Launer Lenore J.,
Goldgaber Dmitry
Publication year - 2015
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2015.06.259
Subject(s) - apolipoprotein e , cerebral amyloid angiopathy , senile plaques , neurofibrillary tangle , atrophy , clinical dementia rating , alzheimer's disease , neuropathology , pathology , autopsy , hippocampal formation , cognitive decline , medicine , psychology , dementia , disease
Background:An Alzheimer’s disease (AD) genome wide association study (GWAS) in African-Americans showed that only four of the established AD risk loci reached significant association. Importantly, the specific variants that associated with risk were different from those that achieved significance in GWAS of subjects of European ancestry; underscoring the importance of expanding genetic association studies of AD in African-Americans. Methods: We have tested the association of variants on the NeuroX chip (an Illumina exome array with custom content for neurological diseases) with the risk of AD in 76 African American (AA) patients with clinical diagnosis of AD fromMayo Clinic Florida in Jacksonville. Illumina’s Genome Studio software was used for genotype calling and initial quality control (QC). Samples with a call rate<99% were excluded, and variants with a GenTrain score<0.7 (number between 0 and 1 indicating how well the samples clustered) or in Hardy-Weinberg disequilibrium (p<0.0001) were also excluded. Using a Fisher’s exact test, the allele frequencies in AAAD cases were compared to allele frequencies from AA population controls in the exome variant server (EVS, n1⁄42,203). Results: After QC, 193,310 variants and 75 AA AD samples remained for analysis. False discovery rates (FDRs) were calculated according to Benjamini and Hochberg procedure to correct for multiple tests. After correction, 203 SNPs significantly associated with risk of AD (q-value<0.05). None of those variants have been previously associated with risk of AD. Nonetheless, several variants in genes implicated in AD had nominally significant association with risk of AD in this study. Although the APOE e4tagging SNP (rs429358) did not pass QC on the NeuroX, based on APOE e4 Taqman genotypes from these 75 ADs and the APOE e4 frequency in AA controls from EVS, we are able to detect significant APOE e4 association with AD risk (OR1⁄42.87, p-value1⁄44.05x10), consistent with previous reports. Conclusions: These preliminary results suggest that some genetic risk factors for AD in this population may overlap with those previously reported for AD in Caucasians, while others have not been previously reported as risk factors for AD. Validation of these results in a larger AA cohort is currently underway.