Date of Award


Document Type

Open Access Dissertation


Epidemiology and Biostatistics

First Advisor

Nancy Fleischer


Introduction: According to the World Health Organization, one-third of the disease morbidity and two-thirds of premature deaths among adults are associated with behaviors that can be traced back to adolescence. These include behaviors resulting in unintentional injury, violent behaviors, alcohol, drug, and tobacco use. The purpose of this research was to explore how differing levels of urbanicity affect youth’s engagement in risk behaviors.

Methods: Analysis was done using a nationally representative sample of 9th-12th graders in the 2003 Youth Risk Behavior Survey (YRBS), the most recent year for which urbanicity is available. The main exposure was urbanicity (classified as urban, suburban, or rural based on location of school the student attended at the time of the survey). Logistic regression was used to measure the main exposure for all risk behaviors. Race/ethnicity, sex, age and geographic region of the country were assessed as potential confounders and/or effect modifiers.

Results: Youth in rural and suburban settings engaged in risk behaviors differently than youth in urban settings. For instance, rural males had twice the odds of urban males for carrying a weapon and suburban males had twice the odds of urban males for not wearing a seatbelt. The association between urbanicity and risk behaviors was often modified by sex and geographic region. Effect modification by sex was important for the least prevalent risk behaviors: weapon carrying and seatbelt use. Males were generally more likely to report both carrying a weapon and not wearing a seatbelt than females were.

Effect modification by geographic region was important for seatbelt use, suicide contemplation and drug use. Where regional interaction was detected, suburban adolescents’ risk behaviors were less impacted by geographic region than rural or urban adolescents. The exception was marijuana use, where suburban youth did experience differences in reporting based on region. Highly prevalent behaviors (like alcohol and tobacco use) were experienced more universally regardless of sex, geography or level of urbanicity.

Conclusions: Given the differences in youth risk behaviors across geography, efforts to reduce risk behaviors may be more effective when tailored to urbanicity, sex and geographic region of the country.