An Analysis of Individual Level and Neighborhood Level Factors on Childhood Obesity

Document Type


Publication Date




A Leader, and R McIntire, Jefferson College of Population Health, Thomas Jefferson University, Philadelphia, PA.


In the U.S., childhood obesity has more than doubled among children and quadrupled among adolescents since 1980. Root causes of obesity are complex and evolving, while inconsistent strategic priorities persist. Some advocates favor behavior modification tactics, such as widespread SNAP (Supplemental Nutrition Assistance Program) reform. Others argue that adverse environments inhibit healthy lifestyles. The purpose of this study was to identify individual- and neighborhood-level predictors of childhood obesity to inform more efficient focus areas for intervention. A secondary data analysis of Public Health Management Corporation’s 2012 Household Health Survey and Trade Dimensions: 2013 Grocery Retail Locations database was conducted among a sample of 590 children (per parent proxy) ages 6-17 living in Philadelphia, PA. The selection of covariates for binomial logistic regression began with bivariate analyses (chi-square/Spearman’s R) to evaluate relationships between individual-level (demographics, behaviors, SNAP status) and neighborhood-level (supermarket locations) factors and obesity. ArcGIS mapped supermarket locations and obesity rates across Philadelphia’s 12 Planning Analysis Sections (PAS). The majority of the sample was male (51.5%) and Black (non-Latino) (55.7%), with a mean age of 11. Obese children comprised 21.7% of the sample. A negative association between supermarket density and obesity rates was observed, both visually (GIS mapping) and quantitatively (bivariate analysis). Diet (servings of fruits and vegetables/sugary-sweetened beverage consumption), screen- time (T.V.), physical activity, household SNAP status, parental education level, and total children in the household, were also significant predictors of obesity. Regression analyses showed that children from households that: 1) Are located in a PAS with low supermarket density; 2) Maintain a high school education level; 3) Have >1 child; and 4) Receive SNAP, had significantly greater odds of being obese. Findings demonstrate the need for synergy across the compound impressionistic foundations of a child. Conclusions support the alignment of childhood obesity priorities in order to exact the most impactful outcomes.

Presentation: 28 minutes