All ETDs from UAB

Advisory Committee Chair

Stacey S Cofield

Advisory Committee Members

Suzanne Judd

April P Carson

D Leann Long

James M Shikany

Document Type


Date of Award


Degree Name by School

Doctor of Public Health (DrPH) School of Public Health


Nutrition is a key factor in the development of noncommunicable diseases (NCDs) such as cardiovascular disease, diabetes, obesity, and cancer, which are continuing to increase globally. Research on dietary patterns is crucial for informing what aspects of diet are most critical for preventing and possibly reversing chronic disease. Dietary patterns focus on the types of food that are consumed which reflects the complexity and cultural aspects of diet rather than focusing on a single food or nutrient. Observational cohort studies are able to observe typical dietary intake and thus better reflect the real association between diet and chronic disease. In order to elucidate the association between diet and chronic disease, observational studies can account for a range of factors influencing diet through multivariable modeling. Issues of confounding are one of the primary limitations of observational studies. Even after employing methods such as risk factor adjustment for measured confounding, addressing potential unmeasured confounding remains an issue and can be related to either imperfectly measured variables or unmeasured variables. Given the limitations that have traditionally accompanied sensitivity analysis and other methods for addressing unmeasured confounding, such as lack of comparability, iv interpretability, and potential complexity of assumptions, VanderWeele et al. proposed a new measure, the E-value, for sensitivity analysis that requires minimal assumptions to evaluate the evidence for causality. The outline of this dissertation is as follows. After review of the literature, the first paper undertakes a scoping review to systematically assess how the E-value has been implemented in nutritional epidemiology thus far. The second paper evaluates the utility of the E-value in relation to the association of dietary measures and incident type 2 diabetes by comparing known associations of each potential confounder to an E-value from the model without the confounder. The final paper implements the E-value in relation to the comparison of the association of multiple dietary measures with incident type 2 diabetes in the REasons for Geographic And Racial Differences in Stroke (REGARDS) study. In concluding, public health recommendations and suggestions for future research are provided in relation to the dissertation research.

Included in

Public Health Commons



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.