Cardiovascular disease (CVD) affects millions of people and is a leading cause of death worldwide. CVD consists of a broad set of conditions including structural heart disease, coronary artery disease and stroke. Risk for each of these conditions accumulates over long periods of time depending on several risk factors. In order to reduce morbidity and mortality due to CVD, preventative treatments administered prior to first CVD event are critical. According to clinical guidelines, such treatments should be guided by an individual’s total risk within a window of time. A related objective is secondary prevention, or early detection, wherein the aim is to identify and mitigate the impact of a disease that has already taken effect. With the widespread adoption of electronic health records (EHRs), there is tremendous opportunity to build better methods for risk assessment and early detection. However, existing methods which use EHRs are limited in several ways: (1) they do not leverage the full longitudinal history of patients, (2) they use a limited feature set or specific data modalities, and (3) they are rarely validated in broader populations and across different institutions. In this dissertation, I address each of these limitations. In Aim 1, I explore the challenge of handling longitudinal, irregularly sampled clinical data, proposing discriminative and generative approaches to model this data. In Aim 2, I develop a multimodal approach for the early detection of structural heart disease. Finally, in Aim 3, I study how different feature inclusion choices affect the transportability of deep risk assessment models of coronary artery disease across institutions. Collectively, this dissertation contributes important insights towards building better approaches for risk assessment and early detection of CVD using EHR data and systematically assessing their transportability across institutions and populations.

<p>Arterial stiffness is a natural and inevitable process for an ageing artery. In adults and school-aged children, increased stiffness of the central arteries is associated with cardiovascular disease (CVD) and CVD risk factors. Arterial stiffness, and its relationship with adiposity and physical activity (PA), has not been studied in preschool-aged children (3-5 years) despite the high prevalence of obesity and inactivity in this age group. Ninety-eight healthy preschoolers (4.4 ± 0.9 years; 50% boys) participated in this thesis, completing baseline and follow up assessments 12.5 ± 1.1 months apart. Whole-body PWV (carotid to dorsalis pedis; m/s) was used to assess arterial stiffness, body mass index percentile (BMI%ile) was a surrogate measure of adiposity, and PA levels (total, TPA; moderate-to-vigorous, MVPA) were quantified objectively with accelerometers and expressed as a percent of wear time. In our cohort, PWV increased significantly from baseline (4.3 m/s) to follow up (4.8 m/s; p< .001). PWV also tracked fair-to-moderately well (κ=0.25, r=0.37) with no sex differences (χ²=.485, p=.785). Girls had a higher BMI%ile than boys, and the prevalence of overweight/obese preschoolers increased from 18.8% to 21.3% over the 1-year period. Boys were more active than girls and engaged in more MVPA. 75% of preschoolers at baseline, and 70% at follow up, met the current PA guidelines (3-hrs of TPA/day). PWV was not related to BMI%ile or PA at baseline; however, it was weakly related to TPA (r=-0.28, p=.013) and MVPA (r=-0.25, p=.024) at follow up. Furthermore, longitudinal and cross-sectional regression models of sex, age, BMI%ile and TPA or MVPA could not predict PWV. Our findings indicate that adiposity and PA do not influence arterial stiffness in healthy 3 to 5 year old children. Nevertheless, maintaining a healthy body composition and engaging in regular PA has other health benefits and should be encouraged.</p> / Master of Science (MSc)

<p>A comprehensive understanding of the mechanisms regulating vascular structure and function may assist in designing effective strategies to decrease cardiovascular disease risk. The current studies were designed to investigate a) relationships between collagen markers and arterial stiffness and markers of vasoconstriction and inflammation and endothelial function in humans with a wide range of vascular health, including overweight women, elderly healthy men, individuals with coronary artery disease, individuals with spinal cord injury and young healthy men and b) changes in arterial structure and function and circulating serum markers of type I collagen synthesis and degradation, vasoconstriction and inflammation in overweight pre-menopausal women before and after a 16- week diet and exercise intervention. Resting brachial artery flow mediated dilation (FMD), upper limb and/or central pulse wave velocity (PWV_c-r and PWV_c-f) and carotid artery distensibility were assessed at baseline in all groups and, in the overweight population, after the 16-week intervention. Pro-collagen type I C-peptide (PIP), C-telopeptide of type I collagen (CTX), markers of collagen synthesis and degradation respectively, endothelin-1 (ET-1) a vasoconstrictor and interleukin-6 (IL-6) an inflammatory marker were measured. In the spectrum of vascular health, a negative relationship exists between collagen markers and central PWV (CTX–PWV_c-f: r = -0.41, p = 0.001 and PIP – PWV_c-f: r = -0.32, p = 0.01) and a positive relationship between markers and carotid distensibility (CTX: r = 0.59, pc-r increased over time in the overweight population (FMD pre: 4.1 ± 0.5 % vs. post: 6.9 ± 0.7 %, pc-r pre: 8.1 ± 0.3 m/s vs. post: 8.9 ± 0.3 m/s, p</p> / Doctor of Philosophy (PhD)

published_or_final_version / Community Medicine / Master / Master of Public Health

Thesis (Ph. D.)--University of California, San Diego and San Diego State University, 2008. / Title from first page of PDF file (viewed June 12, 2008). Available via ProQuest Digital Dissertations. Vita. Includes bibliographical references (p. 70-81).

No description available.

Thesis (DScMedSc)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: Vascular endothelium forms the first line of defence against harmful stimuli in the circulation. Endothelial dysfunction is a valuable predictor of cardiovascular disease and therapies aimed at improving endothelial function are therefore needed. The anti-dyslipidaemic agents, simvastatin and fenofibrate, are known for their beneficial effects on lipid parameters, however additional pleiotropic effects have been shown for both. These include improved endothelial function due to increased levels of nitric oxide (NO), as well as anti-oxidant and anti-inflammatory actions. NO is produced by the enzyme, nitric oxide synthase (NOS), which exists in the endothelial NOS (eNOS), inducible NOS (iNOS) and neuronal NOS (nNOS) isoforms. Most studies investigating the endothelial effects of simvastatin and fenofibrate are performed on macrovascular-derived endothelial cells, and there is a lack of data on endothelial cells (ECs) from the microcirculation, particularly the cardiac microvessels. This dissertation aimed to investigate and elucidate mechanisms underlying the pleiotropic effects of simvastatin and fenofibrate on ECs and vascular tissue using in vitro, ex vivo and in vivo experimental models. In vitro investigations included flow cytometry-based intracellular measurements of NO, as well as different types of reactive oxygen species (ROS) and cell viability parameters. Signalling pathways involved with these changes were measured by western blot analyses of the expression and phosphorylation of critical proteins involved in vascular function. Results on cardiac microvascular ECs (CMECs) demonstrated that fenofibrate (50 μM) exerted a potent, increasing effect on NO production after short periods (1 and 4 hour treatments), but after 24 hours the effects were less robust. Exhaustive investigations suggested that the NOincreasing effects of fenofibrate in baseline CMECs were NOS-independent, a novel finding as far as we are aware. Fenofibrate’s ability to protect ECs against injury was demonstrated when CMECs incubated with the pro-inflammatory cytokine, TNF-α, were pre-treated with fenofibrate, resulting in increased NO and improved cell viability parameters. Simvastatin (1 μM) increased NO to a lesser extent in baseline CMECs, and resulted in increased apoptosis and necrosis. Following the cell studies, their effects on vascular reactivity was measured by aortic ring isometric tension studies. The effects of acutely administered fenofibrate to pre-contracted aortic rings were investigated, and results showed a modest, but significant NOS-dependent vasodilatory response. Next, an in vivo model of Wistar rats treated with simvastatin (0.5 mg/kg/day) and fenofibrate (100 mg/kg/day) for 6 weeks was established. Data showed that neither drug was able to improve aortic ring contraction and dilation above baseline values. Both drug treatments increased iNOS expression, which is usually associated with harmful actions. However, in our hands, increased iNOS expression was associated with a beneficial anticontractile response in the simvastatin-treated animals. Fenofibrate treatment increased NO bioavailability in the blood of these animals. In conclusion, fenofibrate showed endothelio-protective pleiotropic effects with regards to NO production after short treatment periods in CMECs. These effects were mediated via a NOSindependent mechanism, a novel finding. Fenofibrate pre-treatment was also protective against the harmful effects of TNF-α. Simvastatin did not show pronounced pleiotropic effects in vitro or in vivo on endothelial function. / AFRIKAANSE OPSOMMING: Die vaskulêre endoteellaag is die eerste linie van verdediging teen skadelike stimuli in die bloedsirkulasie. Endoteeldisfunksie is ‘n waardevolle voorspeller van kardiovaskulêre siektes en enige terapeutiese behandeling wat kan bydra tot verbeterde endoteelfunksie is belangrik. Simvastatien en fenofibraat word as anti-dislipidemiese middels voorgeskryf en hoewel hulle primêr gebruik word om cholesterolvlakke te verbeter, toon hulle ook pleiotropiese (cholesterolonafhanklike) eienskappe. Dit sluit in bevordering van endoteelfunksie (via verhoogde stikstofoksied (NO) produksie), asook anti-oksidant en anti-inflammatoriese effekte. NO word vervaardig deur die ensiem, stikstofoksiedsintase (NOS) wat voorkom in drie isovorme: endoteelafgeleide NOS (eNOS), induseerbare NOS (iNOS) en neuronale NOS (nNOS). Die meerderheid studies wat pleiotropiese effekte van simvastatien en fenofibraat ondersoek, gebruik endoteelselle van makrovaskulêre bloedvate, wat beteken daar is ‘n tekort aan data aangaande endoteelselle vanaf mikrovaskulêre vate, veral kardiale mikrovaskulêre vate (CMECs). Hierdie proefskrif het dit ten doel gehad om meganismes betrokke by die pleiotropiese effekte van simvastatien en fenofibraat te ondersoek deur van in vitro, ex vivo en in vivo modelle gebruik te maak. Die in vitro ondersoeke het gefokus op vloeisitometrie-gebaseerde metings van intrasellulêre NO, reaktiewe suurstof-radikale (ROS) en sellewensvatbaarheid. Seintransduksie paaie betrokke by hierdie veranderinge was bepaal deur proteienuitdrukking en -fosforilasie vlakke te meet van belangrike proteïene, met behulp van die Western-blot tegniek. Resultate van die CMEC eksperimente het getoon dat fenofibraat (50 μM) ‘n kragtige en verhogende effek op NO produksie uitgeoefen het na kort behandelingstye (1 en 4 ure), maar na 24 uur was hierdie effek minder uitgesproke. Uitvoerige ondersoeke het getoon dat fenofibraat se basislyn effekte op CMECs deur NOS-onafhanklike meganismes teweeggebring is, en sover ons kennis strek, is dit ‘n nuwe bevinding. Fenofibraat se endoteel-beskermende effekte kon ook aangetoon word deur CMECs vir een uur te behandel voor byvoeging van die pro-inflammatories sitokien, tumor nekrose faktor alpha (TNF-α), wat gelei het tot verhoogde NO vlakke en verbeterde seloorlewing. Simvastatien (1 μM) het tot ‘n mindere mate NO produksie verhoog in CMECs, tesame met pro-apoptotiese en -nekrotiese effekte. Vervolgens was die effekte op vaskulêre reaktiwiteit geëvalueer d.m.v. isometriese spanningsondersoeke. Akute effekte van fenofibraat is gemeet deur byvoeging daarvan tot ‘n vooraf saamgetrekte aorta-ring, wat tot matige, maar beduidende NOS-afhanklike verslapping gelei het. Hierna is ‘n in vivo model opgestel deur Wistar rotte vir ses weke met 0.5 mg/kg/dag simvastatien of 100 mg/kg/dag fenofibraat te behandel. Resultate toon dat geen van die behandelings basislyn kontraksie of verslapping van aorta ringe kon verbeter nie. Beide behandelings het tot verhoogde iNOS uitdrukking gelei, wat gewoonlik met nadelige effekte geassosieer word, maar in ons studies was dit met voordelige, anti-kontraktiele effekte in aortaringe van simvastatien-behandelde rotte geassosieer. Fenofibraat behandeling het die NObiobeskikbaarheid in die rotte se bloed verhoog. Ten slotte, fenofibraat het met endoteel-beskermende, pleiotropiese effekte op endoteelselle gepaard gegaan, veral t.o.v. NO-produksie na akute middeltoediening in die CMECs. Die meganisme was ‘n NOS-onafkanklike proses, wat ‘n nuwe bevinding is. Fenofibraat prebehandeling het teen die skadelike effekte van TNF-α beskerm. Geen uitgesproke pleiotropiese effekte is in vitro of in vivo gevind met simvastatien behandeling nie.

published_or_final_version / Physiology / Doctoral / Doctor of Philosophy

published_or_final_version / Pharmacology / Doctoral / Doctor of Philosophy

published_or_final_version / Physiology / Master / Master of Philosophy