Cognitive aging has profound effects on the hippocampus, a brain region that is critical for episodic memory formation and spatial navigation. Accurate episodic memory formation requires pattern separation, a neurocomputational process that orthogonalizes similar stimulus input into distinct neural representations and can be examined using behavioral mnemonic discrimination paradigms in humans. Age-related impairment in hippocampally-dependent cognition emphasizes the importance of identifying modulators of hippocampal plasticity. Critically, studies in older adults have demonstrated that greater cardiorespiratory fitness (CRF) is associated with greater volume of the hippocampus, mitigated age-related decline in spatial mnemonic discrimination, and better visuospatial memory. Nonetheless, how CRF modulates the underlying structural and functional neural correlates of mnemonic discrimination and spatial navigation in cognitive aging remains unknown. Therefore, the overall objective of this dissertation was to examine CRF as a modulator of hippocampal memory system structure and function, specifically regarding the hippocampally-dependent processes of mnemonic discrimination and spatial navigation, in cognitively healthy older adults. In a series of three experiments, we tested the central hypothesis that CRF enhances hippocampal plasticity and modulates the underlying neural correlates of mnemonic discrimination and spatial navigation in older adults. Data for these three experiments came from two studies. In the first, young adults (ages 18-35 years) and older adults (ages 55-85 years) underwent high-resolution fMRI to examine hippocampal subfield blood-oxygenation level-dependent (BOLD) signal during mnemonic discrimination. In the second, older adults (ages 60-80 years) underwent whole-brain, high-resolution fMRI to examine whole-brain BOLD signal during spatial navigation. In both studies, participants performed a submaximal treadmill test to estimate CRF and underwent high-resolution structural MRI to measure hippocampal subfield volumes as markers of neuroplasticity in the hippocampus. The primary goal of Experiment 1 was to examine the prediction that CRF is positively associated with mnemonic discrimination task performance and dentate gyrus (DG)/CA3 volume in older adults, given that these hippocampal subfields are thought to support pattern separation. Contrary to our initial prediction, we did not observe a relationship between CRF and mnemonic discrimination task performance or CRF and DG/CA3 volume. Instead, we observed a significant positive relationship between CRF and the volume of another hippocampal subfield, the bilateral subiculum, in older adult women, but not men. The primary goal of Experiment 2 was to examine the prediction that mnemonic discrimination-related BOLD signal in the hippocampal subfields is modulated by aging and CRF in young and older adults. In line with our initial prediction, there was a significant difference between young and older adults in right DG/CA2-3 BOLD signal during mnemonic discrimination task performance. Most importantly, CRF significantly modulated bilateral subiculum BOLD signal in an opposing fashion in young and older adults. The primary goal of Experiment 3 was to extend the current study beyond the function of the hippocampus in isolation and to examine whole-brain activation in association with an ecologically valid task that engages a large network of brain regions including but not limited to the hippocampus. We predicted that CRF modulates BOLD signal activation patterns driven by the employed spatial navigation task, specifically in brain regions in the frontal, parietal, and temporal cortices, and the cerebellum, given that the previously published literature in older adults has suggested that CRF enhances structural integrity in these regions in addition to the hippocampus. Our results demonstrated that CRF is significantly positively associated with BOLD signal in the right cerebellum lobule VIIa Crus I and Crus II, a region that has been implicated in sequence-based navigation. And, consistent with our results from Experiment 2, this relationship was observed in older adult women. Importantly, the findings of these experiments highlight novel targets of fitness-related neuroplasticity in the older adult brain, including the subiculum subfield of the hippocampus and the cerebellum lobule VIIa Crus I and Crus II. Furthermore, these findings underscore the importance of examining sex as a modulating factor of fitness-related neuroplasticity.

Long-term declarative memory depends on pattern separation, which reduces the degree of overlap between similar representations, to maintain memory specificity, and on pattern completion, which occurs when a degraded cue is used to retrieve a previously stored memory. Previous studies aimed at evaluating the underlying neuronal substrates of these computational processes have used a mnemonic discrimination paradigm and fMRI to focus on the hippocampus, to the exclusion of cortical processing. We aim to investigate the influences extra-hippocampal processes have on pattern separation in the following two studies. Study 1. Computational models of pattern completion suggest it occurs cortically and results in generalized memories whereas pattern separation occurs in the hippocampus and results in memory specificity. It is unknown how the incongruity of these two neuronal processes is resolved. Many studies evaluating the neuronal correlates of pattern separation have used fMRI to evaluate activity in the hippocampus. The sluggish time resolution of fMRI and the restricted spatial focus leave room for considerable differences between pattern completion and pattern separation to go undetected. Here, we use encephalography (EEG) and an event-related potential (ERP) analysis to examine neuronal activity during pattern separation and pattern completion to investigate whether or not cortical processing is employed to resolve the discrepancy between these two neuronal processes. We largely did not observe differences between the ERPs associated with pattern separation and pattern completion. Failure to identify neuronal differences could result from the bulk of neuronal processing differentiating between the two processes occurring deeper in the brain than can be measured by ERPs. Study 2. Extrinsic rewards contingent on memory performance can boost memory and learning. However, the effects of extrinsic rewards on memory specificity, particularly in regards to the process of pattern separation, are not well understood. In this behavioral study, we evaluate how extrinsic rewards affect behavioral performance in a task that taxes pattern separation. Our data show that rewards given for participation at the time of encoding boost mnemonic discrimination between target-lure pairs while rewards given for memory performance at the time of retrieval do not. We hypothesize this is because pattern separation is an encoding dependent process. This boost in discriminability is only seen when the rewarded stimuli are blocked together in separate blocks from the non-rewarded stimuli. When the rewarded and non-rewarded stimuli are interspersed within blocks, discriminability does not significantly differ between the rewarded and non-rewarded trials. Overall, performance was better when rewards were contingent on performance than when rewards independent of performance, although this difference is eliminated when attention during encoding is controlled.

Two experiments are presented in this dissertation to investigate the effect a schema may have on mnemonic discrimination. We developed stimuli composed of a foreground item on a background that was either schematically congruent or incongruent. For the encoding phase of both experiments, these stimuli were presented to 98 participants, who were tasked with determining the congruency of each foreground-background pair. Next, the two experiments diverged for the retrieval phase, where participants were presented with either the same object as before (Target) or one that was similar (Lure). Forty-six participants in Experiment 1 saw stimuli with the same background as initially presented during the retrieval phase. For Experiment 2, fifty-two participants saw the foreground item presented only on a white background. Behavioral, eye tracking, and whole-brain, high-resolution fMRI data were acquired for both experiments and both phases of the task. We found memory discriminability (d-prime) scores were larger for incongruent stimuli when target-lure pairs were less similar and only when the background was present during retrieval. Critically, we found evidence of recognition in the hippocampal subregions as opposed to lure detection. These findings support the notion of a congruency benefit due to the "generate-and-recognize" model and an incongruency benefit due to increased initial attention.

Two projects are presented in this dissertation. First, we investigated the impact of false memories on the original trace and whether recovery of the original is possible. Second, we tested whether mnemonic discrimination for temporal duration is possible. Both projects incorporated fMRI techniques in order to implicate any potential neural correlates of these memory behaviors. Project One. To elicit false memories and study a potential recovery therefrom, thirty-five healthy, young adults performed multiple recognition memory tests where they were induced to make errors in a first test and then participants were tested again in a surprise, second test. These two tests allowed us to determine which memory error would be corrected, if any. Further, fMRI signal associated with the encoding and retrieval processes during the experiment allowed us to implicate the regions associated with false memory correction. We found that false memories do not overwrite the original trace in all instances, as recovery of the original information was possible. Critically, we determined that recovery of the original information was dependent on activity in regions associated with retrieval, saliency attention, and bottom-up attention during the formation of the false memory, and not on processing at the time of encoding or the second test episode. Project Two. We developed a novel paradigm to test episodic memory for temporal duration. Thirty-five healthy, young adults completed a temporal discrimination task that consisted of a continuous-recognition paradigm in which visual objects were presented one at a time for either 1 or 1.5 seconds. Certain items repeated (Targets and Lures) where Targets were presented for the same duration while the duration of Lures was altered by ±0.5 seconds. Participants were asked to identify whether the stimulus duration changed. Whole-brain high-resolution fMRI data were acquired. Behavioral results indicate that participants were sensitive to both increases and decreases in duration. Further, fMRI analyses revealed that the left entorhinal and perirhinal cortices were differentially involved in encoding and retrieval, respectively, of correct duration representations. These findings support the notion of the entorhinal cortex supporting temporal representations in memory as well as the perirhinal cortex representing the conjunction of item and context.