Abstract
There is increasing interest in finding markers of Alzheimer's disease (AD) that are discriminative even at an early, pre-dementia stage. This interest is driven partly by a desire to improve clinical diagnosis in more mildly affected individuals, and also by the recent paradigm shift in thinking about clinical trials for AD. This shift is a result of concern that the recent failures of high-profile clinical trials conducted in patients with mild to moderate AD may have been because therapy was “too little, too late.” The implication being that if only treatments had been trialled earlier they would have had a greater chance of success. Certainly, lessons from other aspects of medicine have shown that treatments may be most, or in some cases only, effective if given early in disease. If we did have therapies that could slow disease progression at a very early stage that would increase the interest in early markers of disease. Ideally, such therapies would be given when the minimum of functional decline and irreversible neuronal loss had already occurred. From economic and public health standpoints, delaying symptom onset would be very important: a delay of five years has been estimated to reduce projections for prevalence of symptomatic AD by about 50% (Brookmeyer et al., 1998).