Premium
O1–09–04: Baseline or progression? Biomarkers predictive of memory decline
Author(s) -
Dodge Hiroko,
Zhau Jian,
Harvey Danielle,
Silbert Lisa,
Kaye Jeffrey,
Albin Roger
Publication year - 2013
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.2013.04.101
Subject(s) - cognitive decline , neuropsychology , cognition , biomarker , medicine , dementia , psychology , neuroimaging , alzheimer's disease neuroimaging initiative , oncology , cardiology , disease , neuroscience , biochemistry , chemistry
distribution curves crossed each other. Our clinical cut-point was chosen to differentiate best between subjects with subjective complaints and AD-dementia and was 550 pg/ml.Results:Visual inspection suggested a trimodal distribution (figure). Mixture modeling estimated a bimodal distribution and yielded a cut-point at 639 pg/ml. The best cut-point was lower in subjects younger than 70 years (607 pg/ml) than in subjects older than 70 years (713 pg/ml). Cut-points were comparable in subjects with subjective complaints (649 pg/ml), MCI (664 pg/ml) and dementia (638 pg/ml). Cut-points did not differ with APOE genotype (644 pg/ml in both APOE-e4 carriers and non-carriers). 81% of the subjects with AD had CSF abeta 1-42 concentrations below the cut-point based on clinical diagnosis and 91% below the 639 pg/ml cut-point. Conclusions: Cut-points for abnormal CSF abeta1-42 based on unbiased mixture modeling are higher than those based on clinical diagnosis. This suggests that diagnosis-based cut-points may underestimate the prevalence of abnormal CSF abeta 1-42. Cut-points were dependent on age but independent of degree of cognitive impairment and APOE genotype. Longitudinal studies are needed to determine the outcome of subjects without a clinical diagnosis of AD that had abeta1-42 levels between 550 and 639 pg/ml.