Association of low plasma Aβ42/Aβ40 ratios with increased imminent risk for mild cognitive impairment and Alzheimer disease

Background: To develop preventive therapy for Alzheimer disease ( AD), it is essential to develop AD-related biomarkers that identify at-risk individuals in the same way that cholesterol levels identify persons at risk for heart disease. Objective: To determine whether plasma levels of amyloid beta protein (A beta 40 and A beta 42) are useful for identifying cognitively normal elderly white subjects at increased risk for mild cognitive impairment (MCI) and AD. Design: Using well-established sandwich enzyme-linked immunosorbent assays, plasma A beta 40 and A beta 42 levels were analyzed at baseline in a prospective, elderly white cohort followed up for 2 to 12 (median, 3.7) years to detect incident cases of MCI or AD. Setting: Cognitively normal, community-based white volunteers recruited from primary care settings into the Mayo Rochester Alzheimer Disease Patient Registry. Patients: We followed up 563 cognitively normal white volunteers (median age, 78 years; 62% female) who had at least 1 follow-up visit after measurement of baseline plasma A beta levels. Main Outcome Measures: The primary outcome was time to development of MCI or AD. The secondary outcome was the annualized rate of cognitive change in patients for whom we had 2 Mattis Dementia Rating Scale evaluations 3 to 7 years apart. Results: During follow-up, 53 subjects developed MCI or AD. Subjects with plasma A beta 42/A beta 40 ratios in the lower quartiles showed significantly greater risk of MCI or AD (P = .04, adjusted for age and apolipoprotein E genotype). Comparison of subjects with plasma A beta 42/A beta 40 ratios in the lowest vs the highest quartile gave a relative risk of 3.1 (95% confidence interval, 1.1-8.3). After adjusting for age and apolipoprotein E genotype, regression analysis using annualized changes in the Dementia Rating Scale scores as an outcome variable showed that participants with lower A beta 42/A beta 40 ratios had greater cognitive decline (P = .02). Conclusion: The plasma A beta 42/A beta 40 ratio may be a useful premorbid biomarker for identifying cognitively normal elderly white subjects who are at increased risk for developing MCI or AD.