Brian Chen, PhD, MS, MSAS
Our project explores diabetes in the United States and Illinois through a two-part data analysis. We move from a high-level, data-driven perspective to actionable, local insights. We start with a national view using AI and then transition to a focused geospatial analysis of Illinois.
This dashboard highlights the social and environmental factors most associated with diabetes prevalence across U.S. counties. The Food Environment Index appears as a leading signal, alongside air pollution, traffic volumne, and access to parks.
Our AI model identifies the most important factors influencing diabetes prevalence. The Food Environment Index is the single greatest predictor, with Air Pollution also being highly significant. This shows that environment, especially food access, is a major driver of risk nationwide.
We map the model’s predictions and compare them to observed prevalence. The Prediction Accuracy map spotlights counties where the model is strongest and where unmeasured factors may exist, guiding deeper public‑health research.
An interactive page lets you drill into any county’s raw data and context, providing a deeper look at the information behind our models.
Tip: Use the toolbar in the bottom-right of the embed to download, share, or view PDF.
Our national AI analysis provides a clear mandate: The Food Environment Index is a priority. The question is, how do we use this insight to create a plan for one specific state? This is where we transition to our Illinois dashboard.
The Illinois dashboard takes the national finding and makes it practical for local public health officials. We move from a broad, predictive view to a specific, granular analysis. This view localizes the national signals. Explore county-level diabetes in Illinois with SDOH overlays like food insecurity, unemployment, and physical inactivity to identify priority areas for intervention.
We start with a Diabetes prevalence map and identify the top and bottom counties for diabetes, giving us a clear picture of key areas of concern within the state.
We examine potential drivers using 7 maps showing key Social Determinants of Health (SDOH) factors from the Healthy People 2030 framework, which allows us to investigate factors like unemployment, food insecurity, and physical inactivity.
This is the most crucial page for action. It uses bivariate maps to show the direct correlation between SDOH factors and diabetes rates at the county level. For example, we can see counties that are high in both diabetes prevalence and food insecurity, which are prime targets for intervention.
Use filters to focus on counties, compare SDOH layers, and inspect bivariate patterns.
The National dashboard gives us our strategic priority: The 'Food Environment Index' is the single most important predictor. The Illinois dashboard gives us our tactical roadmap: It uses that national insight to drill down into a local factor, 'Food Insecurity,' and shows us exactly which counties need help.
Together, these dashboards provide a powerful new methodology. They allow us to move from a general understanding of the problem to a very specific, evidence-based strategy for intervention.