METHODS: The STLDOH obtained publicly available height, weight, age, gender and zip code data from the Missouri Department of Motor Vehicles (DMV). Records were cleaned, and addresses were geocoded and body mass index (BMI) computed. BMI values were adjusted for self-reporting bias according to published recommendations in the literature. Data were then analyzed by geography, age, and gender.
RESULTS: A final analytic dataset of 171,894 DMV records from St. Louis City addresses was obtained with ages ranging from 16-90 years. Overweight and obesity varied widely across several factors. Approximately 63% of males and 59% of females were in the overweight or obese categories. Overweight and obesity was correlated with age, with 42% of those under 30 as overweight or obese, while 70% of those over 40 were overweight or obese. Geography was also widely distributed for overweight and obesity. Census block group level distribution of overweight and obesity ranged from 34% to 82%.
CONCLUSIONS: Public health departments are under increasing pressure to do more with less. This approach to obesity surveillance is low cost and applicable to many other cities around the country. DMV records are capable of providing large sample sizes, which assists statistical power. The address level data allows for very granular analyses, revealing patterns previously obscured in larger geographies. Geocoded data can be merged with Census and other data to provide additional information for planning and intervention work.