Governing's analysis of health and census data found a strong correlation between a metro area's overweight/obesity rates and the portion of a workforce walking or biking to work. Read our story on the findings, or view the methodology:.


The Census Bureau’s 2010 American Communities Survey gauges means of transportation to work for various geographical areas. Survey respondents are asked to record the type of transportation used to travel the longest distance during their commute to work. It should be noted that since only a single transportation type for the longest distance can be counted, those walking or biking to rail stations would likely be recorded only as public transportation commuters. Totals for walkers and bikers were added for each metropolitan statistical area (MSA) and divided by the total number of residents age 16 and older who work (not including those working from home) to compute the total percentage of walkers or bike commuters.
The Centers for Disease Control and Prevention tracks a multitude of health risk data for its Behavioral Risk Factor Surveillance System. Using the body mass index (BMI), it groups people into three classifications: obese (30 or greater BMI), overweight (25-29.9 BMI) or neither overweight nor obese (BMI less than 25). For this study, the variable measuring populations considered neither overweight nor obese in 2010 was used.
CDC’s 2010 dataset included estimates for 192 geographic areas. Some of these estimates, such as those for micropolitan statistical areas and metropolitan divisions, were not measured in the census data. A total of 126 metropolitan statistical areas had data from both surveys, and these were used for the analysis.
Governing performed a multiple regression test, with the estimated percentage of an MSA’s population considered neither overweight nor obese used for the dependent variable. The percentage of walk/bike commuters was tested, along with five other independent variables: median household income, percent of population with a bachelor’s degree or higher, population density, unemployment and mean travel time to work. Data for all six variables was obtained from the 2010 American Communities Survey.
The regression test produced the following coefficients table:

Dependent variable = Percentage of residents neither obese nor overweight
Correlations between healthy weight levels and those walking or biking to work, median household income and education were shown to be statistically significant at the .01 level. The education variable had the strongest effect on a population’s weight, followed by biking/walking to work and income.
A scatterplot illustrates the correlation between an MSA's percentages of walkers/bike commuters and those with healthy weights:


The CDC’s Behavioral Risk Factor Surveillance System measures obesity and other health factors. The table below shows 2010 estimates for each geographic region surveyed, most of which are metro areas. Click here for a complete list of areas in included in the 2010 survey, along with specific counties comprising each area.

The following definitions describe the data:

  • Healthy weight: Neither overweight nor obese
  • Overweight: Body mass index of 25-29.9
  • Obese: Body mass index of 30-99.8
  • No physical activity: Respondents reporting doing no physical activity or exercise in past 30 days

Alternative Means of Transportation Map

Governing compiled and analyzed 2010 American Community Survey estimates for means of transportation to work for metropolitan statistical areas. Separate data for more than 400 U.S. cities, towns and other census-designated places is shown on an interactive map. (Click to open map in new window).