Association of plasma amyloid beta with risk of dementia: the prospective Three-City Study.

Abstract

Several lines of evidence indicate that a decrease in the CSF concentration of amyloid beta(42) (Abeta(42)) is a potential biomarker for incident Alzheimer disease. In contrast, studies on plasma Abeta(1-40) and Abeta(1-42) peptide levels have yielded contradictory results. Here, we explored the links between incident dementia and plasma Abeta(1-40) and Abeta(1-42) peptide concentrations in the prospective, population-based Three-City (3C) Study. We also assessed the association between plasma concentrations of truncated Abeta (Abeta(n-40) and Abeta(n-42)) and the risk of dementia. During a subsequent 4-year follow-up period, 257 individuals presented incident dementia from 8,414 participants, and a subcohort of 1,185 individuals without dementia was drawn as a control cohort. Plasma levels of Abeta(1-40), Abeta(1-42), Abeta(n-40), and Abeta(n-42) were measured using an xMAP-based assay technology. The association between plasma Abeta peptide levels and the risk of dementia was assessed using Cox proportional hazard models. Of the various Abeta variables analyzed, the Abeta(1-42)/Abeta(1-40) and Abeta(n-42)/Abeta(n-40) ratios presented the strongest association with the risk of dementia: people with a high Abeta(1-42)/Abeta(1-40) or Abeta(n-42)/Abeta(n-40) ratio had a lower risk of developing dementia. These associations were restricted to individuals diagnosed at 2 years of follow-up and the Abeta(n-42)/Abeta(n-40) ratio was mainly associated with the risk of mixed/vascular dementia. Plasma Abeta peptide concentrations and Abeta(1-42)/Abeta(1-40) and Abeta(n-42)/Abeta(n-40) ratios may be useful markers to indicate individuals susceptible to short-term risk of dementia.