The Influence of Dimensional Affective Representations on Age-Related Positivity Effects in Memory

Previous studies have demonstrated an age-related preference in attention and memory for positive over negative information. Existing theories explaining this pattern have been difficult to unify and cannot account for some inconsistencies in the literature. We propose that the positivity effect emerges from a weakening of categorical affective representations leading to an increased reliance on frontal valence representations when processing affective information. This process may be enhanced by a selective de-differentiation of negative valence, referring to an age-related loss in the distinctiveness of negative representations. We will test this model in a multi-step procedure on a large scale, publicly available, cross-sectional dataset (CamCAN). While undergoing fMRI, participants encoded neutral foreground objects paired with negative, neutral, and positive background images. After a short delay, several incidental memory measures were taken, of which the free recall of background images is of main interest to this study. To assess the general presence of frontal valence representations and their potential de-differentiation, a representational similarity analysis will be implemented, testing whether the orbitofrontal cortex represents normative valence and whether this representation shows age-dependent changes in valence differentiation. We expect that age will be associated with a selective de-differentiation of negative valence. Next, a structural equation modelling framework will be used to examine a moderated mediation model. This model will be used to test whether age moderates the indirect effect of frontal valence representations on the relative memory for positive over negative stimuli. We propose that this indirect effect occurs through representational connectivity with brain regions that represent categorical affective information, and expect that the influence of valence representations on memory is increasingly mediated by connectivity as people get older. Finally, we will investigate whether the strength of frontal valence representations influences recall of emotional items at the trial level, allowing us to examine whether the effect of valence representations on memory is specific to positive or negative stimuli. Further details on the background, methodology, and hypotheses are discussed in the attached pdf document. Together, these analyses test a new mechanistic account of the positivity effect that has the potential to integrate existing theories and provide a framework for future research on age-related changes in affective processing.