When the source of false memory helps to differentiate mild cognitive impairment (MCI) from Alzheimer’s Disease (AD)

juil. 6, 2017 15:00
Potsdam, Germany

Background. Episodic memory impairment is at the core of amnestic Mild Cognitive Impairment (aMCI) and Alzheimer’s Disease (AD). Notably, they both commit more memory errors than control participants. Nonetheless, these two groups of patients exhibit the same kinds of memory errors in recognition. The present study aims at exploring whether the source of the memory errors might help to distinguish aMCI from AD.

Method. Three groups of participants, 25 healthy elderly adults (HE), 20 aMCI patients and 9 AD patients, selectively learn 15 words (underlined) always presented with a distractor (not underlined). Few minutes after, they complete a yes/no recognition task in which the targets are mixed with distractors. Half of the distractors comes from the learning phase (old condition) and the other half are new words (new condition). Old and new distractors could belong to the same category (close condition: kettle-coffe maker), or to a related category (kettle-herbal tea), or to a unrelated category (kettle-spinning top) as the target.

Results. As expected, HE show the best memory performance followed by the aMCI patients and finally the AD patients. The same pattern of results is observed for the memory errors in the new condition. However, aMCI and AD patients exhibit similar performance in the close condition. Reversely, the biggest gap in performance between these groups is reported in the old close distractor condition, whereas similar performance is shown in the other conditions.

Discussion. The results suggest that both quantity and quality of memories should be taken into account in the evaluation of memory performance. AD patients present a more qualitatively impaired memory representation than aMCI patients (Vallet, Rouleau et al., 2016). This study highlights the usefulness of a selective learning procedure to distinguish between aMCI and AD as proposed in the binding memory test (Buschke, 2014).