The emotional ageing brain: cognitive mechanisms and neural networks

Abstract

Despite well-documented declines in most cognitive domains, some emotional processes appear to be preserved or even enhanced in late adulthood. A good example of this is the information processing bias older adults show towards positive relative to negative emotional stimuli, often referred to as the age-related positivity effect. The present thesis presents a series of experiments that were designed to better understand the mechanisms that contribute to age-related changes in emotional processing, focusing in particular on the role of cognitive mechanisms and neural networks. In Study 1 aimed to identify the underlying cognitive mechanisms of the positivity effect. The primary focus of this study was to explore the role of distractors during the early attention allocation stage, and to also measure how selective attentional processes during encoding influence later memory outcomes for emotional items. The results showed that consistent with prior literature, a memory positivity effect was found among older relative to younger adults. However, of particular interest was the finding that, participants’ memory for negative targets was not influenced by the presence of positive distractors. This finding suggests that positive distractors did not automatically capture older adults’ attention during encoding for negative items. Importantly, we found that participants’ pupillary responses to negative items mediated the relationship between age and the memory positivity effect, indicating that older adults use their cognitive control resources when encoding negative information, perhaps to down regulate the impact of negative emotions on their memory. Collectively, these two findings provide converging support for the cognitive control account of the positivity effect. Study 2 used a similar paradigm to Study 1 to examine the underlying neural networks involved in processing emotional items during working memory encoding among older and younger adults. Results indicated that a cognitive control network that included fronto-parietal regions, was functionally connected to the left ventrolateral prefrontal cortex during the encoding of negative items among older adults. This network contributed to performance, both accuracy and response times, in older adults’ group. A less distributed network was found for encoding of positive items among older and both items among younger adults. Although older adults recruited a same network that was functionally connected to the amygdala for encoding positive and negative items, younger adults recruited this particular network specifically for encoding negative items. This network facilitated older adults’ higher accuracy and faster response times during retrieval. Taken together, the results from these functional connectivity analyses suggest that there is differential engagement of brain networks connected to these two regions, which are modulated by the emotional valence. While two separate brain networks underlying the encoding of emotionally valence targets are connected to the vlPFC region, one distinct network is functionally connected to the amygdala and subserves the processing of both positive and negative targets. In Study 3 age-related differences in neural substrates involved when processing happy and angry expressions presented with direct versus averted gaze were investigated. This research was motivated by studies that show older adults not only have difficulties processing emotional cues such as facial expression and eye gaze cues, but also have problems integrating these cues. Study 3 provides the first empirical examination of the underlying neural correlates of age-related difficulties in integrating communicative cues. The results showed that for angry facial expressions, younger adults recruited distinct networks while processing direct versus averted eye-gaze cues, however, older adults showed a lack of neural sensitivity to these cues, recruiting a single network for both types of stimuli. In contrast, for happy facial expressions, only older adults showed neural sensitivity to eye gaze cues. Participants’ performance on the scanner task was then correlated with a measure of theory of mind (TOM). Younger (but not older) adults’ performance on a measure of TOM and recognition of angry expressions was differentially correlated with activation in two sets of brain regions as a function of eye gaze. Unlike younger adults, older adults’ performance on TOM was also differentially correlated with the key node of mentalizing brain network during happy expressions as a function of eye gaze. The findings from Study 3 suggest that the age-related difficulties in integrating facial cues could be associated with the recruitment of the mentalizing network when the task imposes high demand on social-cognitive processing. Taken together, the three Studies reported in this thesis provide novel insights into our understanding of age-related differences in the processing of emotionally valenced items, particularly with respect to initial encoding of this information, and how this relates to later memory outcomes. Moreover, for the first time the neural correlates of integrating two important types of facial cue has been identified, and potentially linked to broader social cognitive difficulties. Overall, the findings of this thesis have broad implications for understanding the underlying cognitive mechanisms and neural networks that contribute to age-related differences in the processing of emotional stimuli.