Premium
F2‐03‐01: Sleep duration and cognitive function: The Nurses' Health Study
Author(s) -
Devore Elizabeth,
Grodstein Francine,
Schernhammer Eva
Publication year - 2012
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.2012.05.616
Subject(s) - cognition , medicine , cognitive decline , dementia , sleep (system call) , effects of sleep deprivation on cognitive performance , linear regression , repeated measures design , gerontology , demography , audiology , disease , psychiatry , statistics , mathematics , computer science , operating system , sociology
Background: Quantitative endophenotypes for genetic studies provide increased statistical power, may decrease phenotypic heterogeneity and provide more direct biological hypotheses for associated variants. By using this approach it may be possible to identify putative functional SNPs that may explain the GWAS findings reported in case control series and to identify novel genetic factors associated with risk for AD, age at onset or disease progression. Here we demonstrate how CSF biomarkers (Aß42, and, tau phosphorylated at threonine 181 (ptau)) can be useful to characterize the biological effects of known risk variants.Methods: CSFAß42 and ptau levels weremeasured in 1268 samples (371AD cases and 897 controls), fromWUADRC, ADNI and University of Washington. The normalized CSF Aß42a and ptau values were tested for association using an additive model in PLINK. Age, gender, site, and principle components were included as covariates. Results: The test whether we observed an increase in power with a known risk factor we used the same individuals, to test for association between a SNP that tags APOE 4 (rs2075650) and risk for disease (P 1⁄4 1.4e-07), CSF Aß42 (1.2e-29) and tau levels (2.9e-12). Inclusion of APOE genotype, but not APOE 4 dosage, as a covariate completely eliminates this association, indicating that these associations are due to linkage disequilibrium with APOE and that both APOE2 and APOE4 contribute to this association. We next examined whether the novel SNPs from the case control GWAS were also associated with CSF biomarker levels. Several SNPs in PICALM and the MS4A6A gene cluster show a strong association with CSF ptau (P <10-4), suggesting that these genes may influence risk for AD through an tau-dependent mechanism. Conclusions: This work demonstrates that CSF biomarker levels can be a more powerful phenotype than case-control analysis, and that they can help in the interpretation of GWAS findings. Despite the increase in power, a larger CSF series will be necessary to identify novel genes, implicated in AD with genome-wide significant evidence of association.