CDC 500 Cities Data on PolicyMap


CDC 500 Cities


Centers for Disease Control and Prevention (CDC) 500 Cities

Find on PolicyMap

  • Health
    • Health Status
      • Chronic Conditions
    • Access to Medical Care
      • Preventative Care
    • Risk Factors
      • Overweight & Obesity
      • Physical Activity
      • Alcohol & Tobacco Use
      • Limited Sleep

Recently the CDC Foundation, in collaboration with the Robert Wood Johnson Foundation, began a project that reported 27 health-related measures related to chronic diseases, preventative care and screening, and behavioral risk factors. The census tract-level data, called 500 Cities, is now available on PolicyMap.

The 500 cities included in the project are the top 497 largest American cities and the largest cities in Vermont (Burlington), West Virginia (Charleston), and Wyoming (Cheyenne). This ensured that each state had at least one city represented in the data. The city populations range from 42,400 in Burlington to 8,175,100 in New York City and the combined population across all 500 cities represents approximately 33.4% of the United States population. Because the data is only available for these cities, you need to make sure you’re zoomed in to see it.

Much of the 500 Cities data focuses on chronic health conditions, including information about arthritis, asthma, chronic kidney disease, COPD, coronary heart disease, diabetes, high blood pressure, high cholesterol, hypertension, and strokes.

Preventative care measures in the project cover routine checkups within last year, dental visits, and health screening for cholesterol, colorectal cancer, mammography use, Pap test use, and core preventative services.

Behavioral risk factors are comprised of BMI status, physical inactivity, alcohol consumption, tobacco use, and limited sleep.

This data is somewhat different than most other data on PolicyMap in that the tract-level data is estimated by combining survey data from the larger region with demographic data from the tract (a process known as small area estimation). This means that although the data might provide estimations of the types of health outcomes one might expect from a tract’s population, it shouldn’t be used to look for impacts of a particular local situation, such as an anti-obesity campaign, or a disease hotspot. The estimates are made using the Centers for Disease Control and Prevention (CDC) Behavioral Risk Factor Surveillance System (BRFSS) survey data and Census Bureau population estimates (using data from the ACS and decennial census). The BRFSS survey is reported at larger geographic regions, not census tracts.

The small area estimation techniques employed a multi-level regression and poststratification (MRP) approach that used population demographic and socioeconomic data to estimate local survey responses. The MRP approach could be employed to predict prevalence for each variable at different geographic levels, specifically city and census tract level. Validation of the model included both internal validation using direct BRFSS survey estimates and external validation with direct estimates from 2011 Missouri County-Level Study and American Community Survey data at both the state and county levels. More information about the techniques and validation used can be found here.

If BRFSS sounds familiar, it’s because PolicyMap uses BRFSS data in a similar way to create health indicators, some of which are the same or similar to those in the 500 Cities data. More information about the PolicyMap BRFSS data can be found in this blog post. Although both the CDC BRFSS data and CDC 500 Cities data on PolicyMap use the CDC BRFSS survey as their underlying sources, the different estimation techniques between the two models prohibit comparison across the datasets.

This is the first version of the 500 Cities data, and depending on the feedback it gets and funding available, it could be updated and possibly expanded to more cities. We’ll be keeping a look out to see what happens with this data in the future.