The question of an emotional memory enhancement in aging, and of a positivity bias in particular, has been the subject of numerous empirical studies in the last decade. However, the roots of such positive preference are not yet well established. Partisans of a motivation-based perspective contend with those arguing that positivity is related to a cognitive or neural degradation. The aim of this study was to introduce some elements concerning positivity effect in aging. We compared immediate (i.e., immediate recall) versus delayed (i.e., delayed recall and recognition) emotional memory performance in 38 young adults, 39 old adults, 37 very old adults, and 41 Alzheimer's disease patients. Moreover, we manipulated the encoding instruction: Either participants received no particular processing instruction, or they had to process the material in a semantic way. The results indicated that the positivity bias is most likely to occur in individuals whose cognitive functions are preserved, after long retention delay, and in experimental conditions that do not constrain encoding. We concluded by highlighting that although these findings seem to be better in line with the motivation, rather than the degradation, perspective, they do not fully support either theory. (PsycINFO Database Record

The self-reference effect (SRE) has been shown to benefit episodic memory in healthy individuals. In healthy aging, its preservation is acknowledged, but in Alzheimer's disease (AD), the jury is still out. Furthermore, there has yet to be a study of the SRE in amnestic mild cognitive impairment (aMCI). As self-reference implies subjective self-representations, and positive information enhance memory performance, we set out to examine the effects of 1) material and 2) identity valence on the SRE across the early stages of AD. Twenty healthy older individuals and 40 patients (20 diagnosed with aMCI and 20 diagnosed with mild AD) performed a memory task. Participants had to judge positive and negative personality trait adjectives with reference to themselves or to another person, or else process these adjectives semantically. We then administered a recognition task. Participants also completed a questionnaire on identity valence. Among healthy older individuals, the SRE benefited episodic memory independently of material and identity valence. By contrast, among aMCI patients, we only observed the SRE when the material was positive. When self-referential material was negative, patients' performance depended on the valence of their self-representations: negative self-representations correlated with poor recognition of negative self-referential adjectives. Finally, performance of patients with mild AD by condition and material valence were too low and inappropriate to be subjected to relevant analyses. The persistence of an SRE for positive adjectives in aMCI suggests the existence of a positivity effect for self-related information, which contributes to wellbeing. The absence of an SRE for negative adjectives, which led aMCI patients to dismiss negative self-related information, could be due to low self-esteem. These results corroborate the mnenic neglect model and point out the importance of the psychoaffective dimension in patients with aMCI, which could constitute a major factor for the preservation of their self-esteem and self-related memory.