Working memory in Alzheimer's disease and frontotemporal dementia.

Abstract

1. Background Working memory deficits are a recognised feature of Alzheimer’s disease (AD) and can be dissociated from episodic memory impairment, with distinct patterns of breakdown in individuals [1]. Distinctions in clinical profile may have neurobiological relevance. We have previously shown [2] that patients presenting with a dense yet relatively circumscribed classical amnesia (“amnestic-AD”, accounting for 10% in a series of 523 patients) were older and more likely to carry apolipoprotein 4 alleles than those presenting with a constellation of working memory, language and perceptuospatial deficits (labelled “typical- AD” on the basis that they accounted for 61% of cases). Such findings highlight the importance of careful characterisation of individual patients in experimental research. A goal of the present study was to understand the working memory deficit in AD whilst recognising these important distinctions. Existing research focuses on difficulties on dual-task paradigms, encouraging the notion of central executive dysfunction, and promoting a relationship between working memory deficits and frontal lobe pathology in AD [3]. Failures on standard tests of attention and executive function reinforce this interpretation. However, characteristic neuroimaging changes in early onset AD are in posterior hemispheres rather than frontal lobes. Moreover, ‘frontal’ behavioural characteristics are absent. Characteristic qualities of persistence, motivation, and concern for accuracy contrast markedly with the economy of effort and lack of engagement demonstrated by patients with frontotemporal dementia (FTD), the prototypical neurodegenerative disease of the frontal lobes. The presence of qualitative differences in cognitive profile [4] suggests that there are distinct contributions to performance failure. We took the novel approach of exploring the frontal dysexecutive contribution to working memory by comparing test performance in patients with AD and FTD. If the basis of the impairment is primarily ‘frontal dysexecutive’, then one would expect similar profiles in the two groups. However given the ‘posterior’ abnormalities shown on AD patients’ imaging and absence of ‘frontal’ behavioural signs, we predicted that there ought to be qualitative differences on tests of working memory, attention, and executive function.