The validity of the amyloid cascade hypothesis of Alzheimer's disease