Effects of older age on the neural basis of memory generalization - PROJECT SUMMARY Integration is a key memory ability that allows us to generalize across experiences to make inferences and accumulate knowledge. For example, your doctor might tell you to take a new medication with food, and you might later read on the label that the medication can cause an upset stomach. Integrating these memories leads to the inference that taking the medication with food prevents the upset stomach. Failing to integrate may make you less likely to follow the instructions, more likely to experience the negative side effect, and less likely to take a needed medication. Thus, integration failures prevent us from ‘reading between the lines’ of our experiences and can lead to less informed decisions. Memory and aging theories make conflicting predictions about how older age should impact integration abilities, and there is a critical lack of data to resolve these conflicts. To fill this gap, we will conduct three experiments to test the overarching hypothesis that hippocampal integration mechanisms are most negatively impacted by older age, whereas those of the ventromedial prefrontal cortex (VMPFC) are relatively spared. In each of the three experiments, healthy young (aged 18-30) and older adults (aged 65-80) will undergo fMRI in conjunction with variations on an inference task that allows us to probe behavioral and neural signatures of integration at different memory stages (encoding, retrieval, and in between these stages). Our first aim is to identify age differences in the neural systems recruited to support integration during learning. We will focus first on the hippocampus and VMPFC based on their known role in integration in young adults, but we will also conduct whole brain analyses to identify regions that older adults uniquely recruit to support integration. Our second aim is to determine how age differences in the ability to reactivate existing knowledge during new learning impacts the ability to integrate the past and present to make successful inferences. The third aim is to determine the extent to which the hippocampus and VMPFC integrate as a result of making an inferential decision above-and-beyond their role in integration during learning, and to characterize age differences in retrieval-based integration. These experiments will provide novel empirical data to characterize age-related differences in integration, resolving competing predictions from aging theories. The completion of this work will have a positive impact because it has the potential to identify behavioral and neural intervention targets that could enhance the ability to generalize across experiences in older age. Conversely, our current lack of knowledge about generalization mechanisms in older adults risks us developing memory interventions that promote some forms of memory but backfire by hindering integration and generalization. Thus, filling this knowledge gap may lead to more effective interventions that help older adults glean information across experiences to support informed decision-making.
