Cardiovascular disease is the primary cause of death in the United States and presents a significant challenge for researchers and clinicians due to the complexity and non-uniformity of structure-function relationships in the heart. The largest of the four heart chambers, the left ventricle, must pump efficiently and adapt to changes in demand or blood flow, which the normal heart does remarkably well. However, in disease states, the compensatory response of the myocardium and persistent overloading can cause remodeling to become counter-productive.

Two of the major diseases contributing to overall cardiac morbidity and mortality are hypertension and myocardial infarction. While much is understood about the forms of remodeling and the associated dysfunction that occur in these disease states, they are usually measured at the scale of the entire organ (e.g. ejection fraction), or only in a select region of tissue. Many studies have shown regional variations in structural properties in the normal heart, but a comprehensive understanding of regional remodeling and the resultant functional consequences in disease states is lacking. Thus, the primary aims of this dissertation are to:

1. Determine the effect of pressure overload on regional distributions of electrophysiological and calcium handling proteins in the rat left ventricle.

2. Test the hypothesis that regional gradients in structural properties such as myocyte geometry are diminished in pressure overload.

a. Determine whether regional variations in these structural features could be comprehensively and non-invasively quantified using DT-MRI.

3. Compare the measured size of a scar after myocardial infarction using traditional methods to the size of the corresponding region of dysfunction using a novel 3D MRI-based approach.

A better understanding of regional variations in normal and diseased ventricles will be valuable in identifying potential targets for therapeutic intervention to slow, prevent, or reverse adverse remodeling in patients with cardiac diseases such as hypertension or infarction, and provide insight as to the mechanisms by which regional variations contribute to overall pump function.

Alzheimer’s disease is a debilitating, slowly progressive neurodegenerative disorder affecting over 50 million people globally, a figure expected to increase to 152 million by 2050. Not only is there a long prodromal period wherein individuals show early signs and risk factors for the disease, which are often mistaken as effects of normal aging, but insidious pathological processes are thought to begin taking place in the brain decades before individuals show severe cognitive impairments. Suitably, recent research efforts have focused on detecting signs of ad earlier in life to identify the disease preclinically and before irreversible damage occurs. However, knowing what to focus on in this preclinical period is proving to be a considerable challenge. The locus coeruleus (LC) is a small brainstem nucleus critical for healthy brain functioning that has been implicated in the pathogenesis of Alzheimer’s, and recent advances in neuroimaging allow for in-vivo examination. This dissertation aims to characterize the LC in-vivo using magnetic resonance imaging (MRI), neuropsychological testing, and biomarker assessment, with the goal of bettering our understanding of prodromal AD and AD risk. In Chapter 1 I detail our work optimizing a novel diffusion MRI analysis tool to establish a reliable method for reconstructing an LC projection pathway to the transentorhinal cortex, a region heavily affected by AD pathology. In Chapter 2, I explore how equilibrium probability within this pathway relates to cognition in the context of tau and amyloid biomarkers. Finally, in Chapter 3 I use an LC-sensitive anatomical MRI scan to measure relative LC signal, potentially a marker of nuclear integrity, and examine its associations with cognition in the context of tau and amyloid. Importantly, I show that the relationship between these MRI markers and cognition depends on tau and amyloid status, implicating a more complex relationship between macroscopic imaging measures and cognitive outcomes. This work has implications for preclinical AD research as well as clinical neuroimaging research aiming to uncover in-vivo disease markers in general.

Purpose

Methods

Results

Conclusion

Safe and convenient access to healthy foods for all populations is a fundamental transportation and environmental justice concern. Emerging evidence suggests that residents of lower income communities have less access to healthy food choices than those in higher income areas. Most studies to date rely on an as- sumed level of food quality generalized across different types of food outlets (e.g., grocery versus convenience stores) mapped in space. The current study includes a detailed audit of food quality offered in 302 food establishments in four communities in the Atlanta Region and compares proximity to these outlets in differing urban and demographic settings. The analyses focus on a middle and elementary school in each community and compare the spatial relationships between schools and sit-down and fast food restaurants and between grocery and convenience stores. Road network distances from school sites to each food outlet were calculated in a geographic information system. Results suggest that food quality varies across neighborhoods by income, but not by walkability. Results also suggest the potential for food quality to vary differentially with distance from schools in higher versus lower income communities. Walking or biking to get food is difficult in auto-oriented envi- ronments which has important implications on sustainability. Youth, elderly, and other populations which do not drive are more reliant on the food choices offered in their immediate environments, such as in schools or assisted living facilities. Methods employed can be expanded to examine associations between food outlet quality, urban form, travel and activity patterns, dietary behavior, and health outcomes.