ABSTRACT
Introduction. There is a debate about the ability of patients with Alzheimer’s disease to build an up-to-date representation of their memory function, which has been termed mnemonic anosognosia. This form of anosognosia is typified by accurate online evaluations of performance, but dysfunctional or outmoded representations of function more generally. Method. We tested whether people with Alzheimer’s disease could adapt or change their representations of memory performance across three different six-week memory training programs using global judgements of learning. Results. We showed that whereas online assessments of performance were accurate, patients continued to make inaccurate overestimations of their memory performance. This was despite the fact that the magnitude of predictions shifted according to the memory training. That is, on some level patients showed an ability to change and retain a representation of performance over time, but it was a dysfunctional one. For the first time in the literature we were able to use an analysis using correlations to support this claim, based on a large heterogeneous sample of 51 patients with Alzheimer’s disease. Conclusion. The results point not to a failure to retain online metamemory information, but rather that this information is never used or incorporated into longer term representations, supporting but refining the mnemonic anosognosia hypothesis.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 These predictions were also analyzed but are not presented here. We concluded that participants did not understand the prediction of future recognition performance, and in fact predicted recall (see Moulin, Citation2002, for analysis of the inefficacy of predictions of recognition in AD patients as a method to understand their metacognitive abilities).