Date of Award


Degree Name

Doctor of Public Health (DrPH)



Second Advisor


Third Advisor



Physical inactivity in adults increases the risk for adverse health outcomes, including non-communicable diseases and premature death. Children who are not sufficiently physically active face increased risk for non-communicable disease risk factors, mental health issues, and poor academic performance. Despite these risks, the majority of people do not achieve sufficient physical activity levels. Social capital is an aspect of the social environment that can shape physical activity behavior throughout different life stages. The three studies of this dissertation sought to assess the association between social capital, from both the cohesion and network perspectives, and physical activity behavior in different life stages. Paper 1 used a representative sample of Texas 2nd grade children to examine the cross-sectional association of neighborhood language-based social cohesion and outdoor play. Using data from the Census, language spoken at home at the school catchment area was determined for each student using Geographic Information Systems (GIS), and multilevel regression was used to assess association of neighborhood language composition and outdoor play with students nested within schools. Interaction models were used to determine if the relation between majority neighborhood language and outdoor play varies by language spoken at home. Paper 2 looked at the interaction between social cohesion and the built environment on adolescent physical activity using a cross-sectional, internet-based study of U.S. adolescents and adults. Built environment features were measured using GIS based variables from GeoFLASHE, and linear regression was used to assess differences in the association of neighborhood social cohesion and out-of-school time MVPA in adolescents by built environment features. Paper 3 assessed network social capital and physical activity trajectories using the longitudinal MoNNET-HA dataset, a cohort of adults age 25 and older living in Montreal. Latent trajectory analysis and multinomial logistic regression was used to assess effect of network social capital on changes in physical activity. Paper 1 found that neighborhood language-based social cohesion was a significant predictor of outdoor play in predominately English speaking neighborhoods. Findings from Paper 2 revealed that social cohesion was significantly associated with physical activity, and this association was stronger for males than females. The association between social cohesion and physical activity differed by built environment features in males, with stronger effects for neighborhoods with unsupportive built environments. Paper 3 found that physical activity trajectories differed by sex and age group, and there were no groups that had increasing levels of physical activity over the five year time period. Network social capital diversity was significantly associated with physical activity in adults, and was also significantly associated with physical activity trajectories. Network social capital reach was significantly associated with physical activity trajectories in adults over 55 years. Overall, this dissertation provides important contributions to the literature base on social capital and physical activity, and these findings provide a foundation for future research to further the field of social capital and physical activity. Interventions and policies that seek to increase physical activity should incorporate both bonding (among like individuals) and bridging (among individuals of different socioeconomic or demographic) groups.