| This dissertation project asked questions about how metabolomics can advance our understanding of human nutrition, diseases and their relations in translational population research. The study participants were a randomly selected population of African Americans from Jackson filed center of the large population-based Atherosclerosis Risk in Communities (ARIC) Study. Several statistical methods, such as general linear regression analysis and different survival analyses, were applied in the project.;This research project evaluated cross-sectional associations between dietary food/alcohol consumption habits and human serum metabolome (Chapter 3--4), as well as longitudinal associations between baseline human serum metabolome and future hypertension and heart failure development (Chapter 5--6). An initial pilot study (Chapter 2) was conducted to evaluate the over-time reliability of the human serum metabolome, and the result of this pilot study served as inclusion criteria to select reliable metabolites, which entered into the analysis of the longitudinal disease studies (Chapter 5--6). Results from these analyses highlight the application of metabolomics in exploring the mechanism of effects from dietary/alcohol consumption (such as oxidative stress and inflammation) on human metabolome, and thus on disease; as well as the role of human metabolome as a mediator in the etiology of disease development. |
