Premium
P4‐118: Tau phosphorylation pathway genes and cerebrospinal fluid tau levels in Alzheimer's disease
Author(s) -
Bekris Lynn,
Millard Steven,
Lutz Franziska,
Li Ge,
Galasko Douglas,
Farlow Martin,
Quinn Joseph,
Kaye Jeffrey,
Leverenz James,
Tsuang Debby,
Yu Changen,
Peskind Elaine
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.2012.05.1821
Subject(s) - cerebrospinal fluid , fyn , tau protein , single nucleotide polymorphism , phosphatase , phosphorylation , alzheimer's disease , gene , kinase , biology , microbiology and biotechnology , endocrinology , medicine , genetics , disease , neuroscience , genotype , proto oncogene tyrosine protein kinase src
Background: Phenotypes of clinical cognitive decline based on repeated measures provide an alternative to the standard case/control study design, which is confounded by heterogeneity of dementia in cases and subclinical disease in controls. Previous evidence, from candidate SNP and genome wide analyses, suggests that these phenotypes are useful when modeled correctly. We examined multiple techniques for analyzing cognitive change. Methods: We utilized measures of global cognition and episodic memory from two longitudinal cohorts, the Religious Order Study and RushMemory and Aging Project. Three methods for modeling cognitive change were compared: 1) linear mixed effects modeling (LME) with random intercept and slope, 2) change-point modeling (CP), which models different trajectories of decline before and after Alzheimer’s Disease diagnosis and 3) linear spline modeling (LS), which allows multiple change points in the cognitive trajectory. All models controlled for sex, education and age. The fit of the models are compared using AIC. We also compared associations of APOE E4 and CR1 (rs6656401 AA/AG) with cognitive decline. Results: Analyses were limited to 690 ROS and 777 MAP subjects with >2 timepoints of cognitive measures, with mean (SD) age at baseline of 75.8 (7.2) and 81.0 (6.5) respectively. The evaluation of the three models, as applied to the APOE and CR1 loci, show that model fit improves as we go from the LME to the CP and finally the LS model. We see a significant effect of APOE E 4 on episodic memory pre-diagnosis (b1⁄4-0.08, P 1⁄4 1.5x10 -11), but no additional significant effect post-diagnosis (b1⁄4-0.19, P 1⁄4 0.11). Interestingly, the estimate is more than twice as strong post-diagnosis, suggesting the analysis may be underpowered. For global cognition, CR1 appears to exert different effects in the pre-diagnosis (b1⁄4-0.02, P 1⁄4 0.02) and post-diagnosis (b1⁄4-0.06, P 1⁄4 0.40) phases, although again we see lack of power in the post-diagnosis phase. Conclusions: Intermediate phenotypes, such as cognitive change, provide insights into disease progression that we cannot ascertain with simple case/control studies. This can lead to further understanding of the functional consequences of disease-associated variants and allow us to study the aging, non-demented population to explore the pathophysiological processes in AD before and after a terminal, demented state.