Objective: To investigate the stability and time course of human CSF amyloid-beta (Abeta) levels over hours.
Methods: Fifteen nondemented participants had CSF sampled hourly for up to 36 hours via indwelling lumbar catheter. CSF Abeta(1-x), Abeta(1-40), and Abeta(1-42) were measured by ELISA in each hourly CSF sample.
Results: Significant variation in Abeta levels of 1.5- to fourfold was detected over 36 hours of serially sampling in individual subjects. Abeta(40), Abeta(42), and Abeta(1-X) are highly correlated over time indicating that similar processes likely regulate the level of these species. On average, the fluctuations of Abeta levels appear to be time of day or activity dependent.
Conclusion: Diagnostic and therapeutic trials that measure Abeta should control for the time of CSF sampling to minimize variability.