This study examines the nineteen demographic, socioeconomic, behavioral, and environmental factors influencing adult obesity across Chicago’s 77 community neighborhoods. Using advanced geographic methods—including geo-visualization, Exploratory Spatial Data Analysis (ESDA), Analysis of Variance (ANOVA), and Geographically Weighted Regression (GWR), the study provides a nuanced understanding of how obesity patterns cluster geographically and how multifaceted risk factors vary spatially across the city.
Data & Methods
The demographic data (percentage of White, African American, Hispanic, and other minority groups) from the American Community Survey (ACS). Adult obesity data came from the Chicago Health Atlas. Physical environment determinants included Sidewalk Quality, Land Use Mix Index (LUMI), Number of Parks, Walkability Index (WI), Social Vulnerability Index (SVI), and density of Fitness Centers. Behavioral data, such as adult smoking, psychological distress, neighborhood safety, and loneliness, as well as socioeconomic variables, including the Hardship Index (HI), unemployment rate, Economic Diversity Index (EDI), poverty, food insecurity, and public assistance income, were also obtained from the Chicago Health Atlas. Together, these variables illustrate a complex interplay of contextual factors determining health outcomes. The analytical framework involved spatial statistics to measure clustering and spatial relationships. Local Moran’s I identified obesity "hotspots" and "coldspots," showing statistically significant geographic controls. GWR allowed coefficients to vary spatially, rather than assuming uniform effects across the city, providing a local rather than a global view of the factors driving obesity.
Spatial Distribution of Adult Obesity
The spatial analysis revealed substantial variation at the neighborhood level. Obesity rates ranged from under 10% in parts of the Near South Side to nearly 60% in places such as Archer Heights. Chicago’s average obesity prevalence (32%) is comparable to the Illinois average but lower than the U.S. average, yet it far exceeds the global average. ANOVA confirmed statistically significant differences across Healthy Chicago Equity Zones (HCEZs), with Central, North, and Far North zones showing significantly lower rates, while Southwest and West zones exhibited the highest burdens. Moran’s, I detected significant positive spatial autocorrelation (I = 0.28), confirming obesity clustering rather than random distribution. Southwestern and western neighborhoods formed consistent high-obesity clusters; northern and central zones formed low-obesity clusters.
Fig. 1 Spatial analysis of the pattern of the adult obese population rate among Chicago neighborhoods: a) adult obese percentage, b) Global Spatial autocorrelation for geographic control, and c) cluster analysis
Disparities in Adult Obesity Determinants in space and context
Race and ethnicity emerged as powerful predictors of spatial location, with the White residents having a strong negative association with obesity rates (R² = 0.39), while minority residents showed strong positive associations (R² = 0.39). The strongest associations among minorities occurred in historically Black neighborhoods on the West and South Sides. Roughly 32% of neighborhoods had little to no association with demographic obesity, particularly in more racially mixed northern communities. On the other hand, Environmental determinants such as sidewalk quality, land use mix, park distribution, walkability, and fitness center density. Among environmental variables, walkability showed the strongest relationship (R² = 0.43), followed by sidewalk quality (R² = 0.33) and fitness centers (R² = 0.23). Surprisingly, the number of parks and the land use mix showed weak associations (R² = 0.06–0.08), suggesting that the mere presence of parks does not guarantee usage, accessibility, or quality, especially in neighborhoods that are not safe for walking and other physical activities.
Table 1: Spatial Relationships of Percentages of Adult Obesity and Demographic Composition of Chicago Neighborhood Communities
Similarly, Behavioral factors like Adult smoking and psychological distress were positively associated with obesity, while neighborhood safety was negatively associated. The findings support broader literature linking obesity to chronic stress, trauma, social isolation, and unsafe environments that limit outdoor physical activity. Lastly, the socioeconomic stressors showed some of the strongest and most statistically significant associations with obesity. Hardship Index and unemployment rate each explained 35% of the variance in obesity rates, while food insecurity, poverty, EDI, and public assistance income also showed moderate associations. Over half of the neighborhoods, however, showed no association, indicating that socioeconomic determinants alone cannot explain all variation, which underscores the multifactorial nature of obesity.
Conclusion
Adult obesity in Chicago communities is shaped by a complex interplay of demographic, environmental, behavioral, and socioeconomic factors operating within spatially distinct contexts. Obesity does not occur randomly; it clusters in neighborhoods characterized by racial segregation, economic hardship, lower walkability, limited fitness infrastructure, higher smoking rates, and psychological distress. Therefore, effective interventions must consider spatial heterogeneity and tailor strategies to local conditions. Policies should target structural inequities, enhance physical environments, increase access to health-promoting resources, and address behavioral health needs. Despite limitations including autocorrelation biases and multicollinearity, the study provides a robust empirical foundation for spatially informed public health planning and equity-focused interventions in Chicago.