Obesity is one of the most pressing global population health issues, and importantly one that affects racial/ethnic minorities and those of low socioeconomic status disproportionately. Obesity tracks from childhood into adulthood and is related to serious medical and economic consequences throughout the life course. Childhood obesity is well recognized as a complex and multifaceted problem influenced by broader social, geographic and environmental factors. A social ecological framework that is transdisciplinary is needed to address individual-level influences and choices that lead to energy imbalance, but also to address the complex interactions among home and neighborhood environmental features that provide the context for health-related behavior change. The overall goal of this dissertation is to investigate multilevel factors that could influence outcomes of an intervention program aimed at addressing childhood obesity. Including studies that investigated the use of clear communication strategies to facilitate parents comprehension, home media environment related to change in children weight status and the neighborhood environmental context of families. This dissertation draws upon a social ecological model that acknowledges multiple levels of human interaction with the environment and represents a comprehensive approach to designing, implementing and evaluating interventions that which target multiple influences on health behaviors. The iChoose is a 3-month family-based childhood obesity treatment program developed under a community-based participatory research approach led by the Partnering for Obesity Planning and Sustainability (POPS) Community Advisory Board (CAB). The studies within this dissertation use clear communication evaluation and qualitative feedback from focus groups, intervention outcome data and home environmental surveys, as well as environmental audits of neighborhoods. To investigate multilevel factors could influence intervention outcomes, we used both multiple statistical analytical techniques and a Geographic Information System (GIS) spatial analysis to evaluate the iChoose program. / Ph. D.

This study evaluated the associations of individual and neighborhood level risk factors with physical health, mental health, and stress in a diverse urban county. Relatively little research has attempted to disentangle the interactive individual characteristics and neighborhood conditions underlying health outcomes and disparities. To address this, survey data were collected and analyzed from 1,107 residents living in one of the 114 census tracts in DeKalb County, GA. Using multilevel structural equation modeling techniques, this study found that neighborhood level measures of the social and built environment were not associated with the health outcomes under study after controlling for neighborhood level income and education. Alternatively, individual level perceptions of the social and built environment and measures of access to health care were significantly associated with physical health, mental health, and perceived stress. This study also found that the association between low individual income and poor physical health was more pronounced for participants who lived in low-income neighborhoods than participants who lived in high-income neighborhoods. Additionally, this study found that Black residents reported significantly better mental health compared to White residents when they lived in high-income neighborhoods, and Black participants reported significantly more stress compared to White participants when they lived in low-income neighborhoods. Results of this study further scientific understanding of the role of neighborhood processes in health disparities and potentially help inform the development of programs and policies related to neighborhood conditions and health disparities.