Social Ecological Model (SEM) and Childhood Obesity: What Do the Data Suggest?

Abstract

The SEM is often used to describe the complex etiology of childhood obesity. Variables included in proximal layers of the SEM (age, gender, parent characteristics) influence weight outcomes in youth. Roles of distal factors such as food and physical activity (PA) environments are less clear. This study examines relationships of different layers of the SEM on child weight outcomes. Data on demographics and parental perceptions and children's use of PA and food outlets were collected from parents of 1400 children. Location of child's home and neighborhood food and PA outlets were geocoded. Multivariate analyses tested independent and joint significance of variables in different layers of the SEM. In fully adjusted models with all layers of SEM, as a group, variables in proximal layers were significantly associated with child's weight. Collectively, variables representing the built environment, those representing perception of built environments, and those assessing use of these environments were not significant. Individually, several measures of the built environment such as presence of a convenience store within a quarter mile radius of home (OR=2.16, CI: 1.12–4.16), presence of a park within a half mile radius of home (OR=0.41, CI: 0.18 – 0.94), and use of the main store to purchase fruits and vegetables (OR= 0.48, CI: 0.27–0.85) were significantly associated with child's weight outcomes.