The amnestic prodrome of Alzheimer's disease

The launch of two new drugs (Aricept® and Exelon®) for the treatment of Alzheimer’s disease heralds the end of an era of therapeutic nihilism in the dementias and emphasizes the need for accurate early diagnosis. It is generally accepted that if disease-modifying treatments of this type are to be effective then it is imperative that we have more sensitive and discriminating methods of establishing a diagnosis before the onset of global cognitive impairment. Numerous papers have been published which address the neuropsychological or imaging findings in early Alzheimer’s disease, but the paper by Foxet al. (1998) is an important step in the quest to determine the earliest cognitive changes in Alzheimer’s disease. In addition to this work its relevance to the question of early diagnosis has important theoretical implications for understanding the pathological changes which occur in Alzheimer’s disease. In order to understand the relevance of Fox et al.’s findings it is necessary to cover some recent history. The concept of dementia, and hence Alzheimer’s disease—the commonest cause of dementia—has undergone revolutionary changes over the past two decades. Sadly, medical students are still taught old definitions of dementia along the lines of ‘progressive global cognitive decline in clear consciousness’, whereas it is increasingly clear that each neurodegenerative disorder has a distinct cognitive signature, at least in the early stages of the disease, which mirrors the distribution of the neuropathological changes. In the case of Alzheimer’s disease, the earliest neuropathological changes (neurofibrillary tangles and amyloid plaques) are typically seen in the medial temporal lobe including the hippocampal complex (Braak and Braak, 1991), although it will be argued below that this view is probably an oversimplification. To reflect this evolving concept, more modern definitions of dementia, such as the widely used NINCDS–ADRDA criteria require ‘deterioration in two or more areas of cognition, including memory, sufficient to interfere with work or social function’. This is obviously an improvement, but still requires the presence of substantial cognitive impairment which arguably delays diagnosis beyond the stage at which disease modification is likely to have a substantial impact. If the pathological changes of Alzheimer’s disease actually begin in a consistent and circumscribed brain area, then it is reasonable to predict that the earliest changes may occur in a single aspect of cognition. A central focus of neuropsychological research in Alzheimer’s disease is, therefore, aimed at addressing the question: ‘what goes first?’ The paper by Fox et al. adds further weight to the argument that the first deficit, at