Date of Award


Document Type

Open Access Dissertation


Epidemiology and Biostatistics



First Advisor

Swann A Adams


Antibiotics are one of the most beneficial discoveries in medicine and public health. However, the use, overuse, and misuse of these drugs have led to increases in antibiotic resistant bacterial infections (ARI). Furthermore, previous epidemiological studies have linked antibiotic use to breast cancer, but these studies have not included effects on metabolic organs, such as the liver and kidneys. This dissertation investigates the role of antibiotic use in infections and liver and kidney cancers in the state of South Carolina. Using ecological study methods and Poisson regression to determine relative risk ratios, it was found that antibiotic use is a risk factor for the development of ARIs and kidney cancer, but not liver cancer. Census tracts with higher percentages of black populations were found to be more at risk for these outcomes, including liver cancer. Case-control methodology was used to investigate individual risk for liver and kidney cancer outcomes, and demographic and geographic variables were examined as confounders or effect modifiers between these relationships. Using conditional logistic regression to calculate odds ratios (OR), it was determined that antibiotic usage is not a risk factor for liver cancer, with ORs of 1.07 (0.77-1.49) for 5 to 36 total prescriptions, 1.33 (0.72-1.46) for 64 to 204 total prescriptions, and 1.39 (0.98-1.98) for 205 to 4374 total prescriptions per participant. No association was found between liver cancer and days of use of antibiotics, nor was their increased risk by specific antibiotic classes. Despite these findings, antibiotic usage was associated with higher odds of kidney cancer outcomes, with ORs of 1.50 (1.27-1.78) for 18 to 131 total prescriptions and 1.43 (1.20-1.69) for 132 to 12362 total prescriptions per participant. For days of use of antibiotics, ORs were 1.41 (1.20-1.67) for 116 to 950 total days of use, and 1.46 (1.23-1.73) for 951 to 123588 total days of use per participant. Also, increased kidney cancers risks were associated with certain classes of antibiotics for some or all levels of exposure by total prescription number and days of use. Overall, these findings suggest that antibiotics must be used in a more judicious manner in medical settings.


© 2014, Prea Thathiah

Included in

Public Health Commons