For years now, the CDC has been putting out figures showing state-by-state changes in the prevalence of obesity over time (e.g., scroll down to the bottom of this link).
It turns out that these state rankings are inaccurate. That's according to some new research just published by researchers in the journal Obesity,
Apparently, the CDC maps were created based on weight information collected from telephone surveys. It has been known for a long time that people tend to under-state their weight when asked on a survey as compared to objective, physical measures taken with a scale.
What appears to have largely gone unnoticed is that there are systematic differences in the extent to which people in different areas of the country under-report. When you look at self-reports from phone surveys, the states in the east south central census region (Kentucky, Tennessee, Alabama, and Mississippi) appear to be the fattest (ranking 1st out of 9 census regions according to the BRFSS survey). BUT, when one looks at actual measured weight, they only rank 5th or 7th depending on which data source you use.
Minnesota and Missouri were only ranked 17th and 19th based on self reports but they rose to numbers 1 and 2 based on actual measurement.
The authors find that the correlation between state-rankings across two data sets (one based on self-reports another based on actual measurements) was only 0.30 and not even significantly different from zero.
All this begs the question: why do some areas of the country under-report their weight more than others? I'm sure there a many potential explanations in addition to the one I hypothesized in the title of this post.