Date of Award

Fall 12-2019

Degree Type


Degree Name

Doctor of Public Health (DrPH)



Second Advisor


Third Advisor



The overall decline in physical activity levels during adolescence supports the examination of intrapersonal-, interpersonal-, and behavior-level factors that may influence physical activity behavior among this population. The three papers within this dissertation aimed to examine associations of socio-demographics and socio-environmental factors with physical activity among U.S. adolescents. Papers 1 and 3 used data from the Family Life, Activity, Sun, Health, and Eating (FLASHE) Study. Paper 1 examined differences in adolescent physical activity by gender and race/ethnicity, and examined modification by grade and socioeconomic status (SES). Paper 3 examined the association of parent- and adolescent-reported parental social support with adolescent physical activity, and if these relations were modified by parent employment. Paper 2 used data from the National Longitudinal Study of Adolescent to Adult Health (Add Health) to examine the association of adolescent and parent weekly hours worked with adolescent physical activity. The FLASHE study is a cross-sectional study that examined correlates of cancer-related behaviors among U.S. adolescents and their caregivers. Adolescents were eligible if they 1) were 12-17 years and 2) lived in the household for at least 50% of the time. Parents were eligible if they 1) were 18 years or older, 2) lived with at least one child aged 12-17 years for at least 50% of the time, and 3) agreed to be contacted for participation. The Add Health study examined factors that may influence adolescents’ health. Schools served as the primary sampling unit and were eligible if they 1) included an 11th grade and 2) enrolled over 30 students. For this dissertation, only participants with complete data for the primary independent and dependent variables of interest were included. Paper 1 used multiple regression to compare in-school, out-of-school, and weekend MVPA by gender and race/ethnicity, and examine modified by grade level and SES. Paper 2 used multiple regression to examine the relation of adolescent and parent and employment with adolescent physical activity (MET/times/week). Multinomial logistic regression was used to examine the relation of adolescent and parent employment categories with adolescent physical activity tertiles (low, medium, high), using the lowest tertile of physical activity as reference. Paper 3 used multiple regression to examine the relation of parental social support scores with out-of-school and weekend MVPA, and to examine whether these relations were modified by parental employment before and after adjusting for adolescent grade, race/ethnicity, and SES. Paper 1 results suggested that relations of socio-demographics with physical activity are complex and contradict many assumptions made in literature. Paper 2 results show that adolescent physical activity was inversely associated with adolescent hours worked, with a direct association of parent employment with adolescent physical activity. Paper 3 found that both parent- and adolescent-reported parent social support were associated with adolescent physical activity. These results highlight the complexity of adolescent physical activity behavior, and the need to consider multiple levels of influence, potential disparities among socio-demographic subgroups, and differences by periods of physical activity.