Many health conditions (such as diabetes and heart disease) disproportionately affect individuals in lower socioeconomic groups. These conditions lead to higher morbidity and premature mortality, making the issue of health disparities an important one to address. Childhood overweight has recently come to the attention of the public health field as one condition needing further understanding. While some of the risk factors and pathways for childhood overweight are known, such as decreased physical activity and higher caloric diet, other contributing factors are still under exploration. One of the most promising areas of inquiry within social epidemiology is contextual level analysis of health outcomes—that is, how do neighborhoods and communities, in both their physical and social structures, influence health statuses and outcomes for individuals. This dissertation examines the role of aggregate-level socioeconomic status and social capital for neighborhoods in Alameda County upon childhood overweight. In order to estimate separate effects of individual/household and neighborhood socioeconomic status and social cohesion/capital, we tested logistic regression models and a random effect logistic regression models.

Our findings did not support clustering of childhood overweight by ZIP code; however, we did find a strong relationship between household level poverty and childhood overweight. While there was no effect modification of social capital on ZIP code level socioeconomic status, it was an effect modifier upon household income levels.

This study contributed to the literature by providing evidence that household level poverty remains more predictive of childhood overweight than neighborhood level socioeconomic status. It is nonetheless important to consider how schools, communities, and neighborhoods can support the healthy development of children, including creating access to nutritious, whole foods, opportunities for recreation and exercise, and regular primary care.