Sunday, November 25, 2007

Congratulations, Austin

We managed to beat out past champ Houston on the Forbes list, coming in at #18; Houston somehow managed to not make the top 20. San Antonio did an impressive turn, ranking #3 in the nation.

Too bad the competition was for most obese high-population city in the US. (Initially, I wrote "most obese large city" but that seemed ambiguous.)

Forbes describes their methodology: "To determine which cities were the most obese, we looked at 2006 data on body mass index, or BMI, collected by the Centers for Disease Control's Behavioral Risk Factor Surveillance System, which conducts phone interviews with residents of metropolitan areas about health issues, including obesity, diabetes and exercise... Though data is collected for roughly 145 metropolitan statistical areas, we looked only at the country's 50 most populated cities and ranked the top 20."

Forbes reports: "Austin wouldn't have entered our list based on 2005 figures, when only 17.2% of its residents were obese. But in 2006, that number shot up to 24.9%, which was enough to surpass neighboring Houston, an erstwhile "fattest" city. The noticeable shift may shock residents who think of their hometown as healthy and active."

Shout out to Okie readers: Oklahoma City makes the list (#8) but Tulsa does not. Since Forbes did not include the list of 50 largest cities that they used to select the top 20, it is unclear whether Tulsa was a contender for the list. This site and this site include Tulsa among the 50 most populous cities, so perhaps Tulsa just didn't manage to make the cut. Better luck next time!

I was pretty surprised to see two California cities on the list, including San Diego. San Diego?! I guess my experience with the place is terribly out of date (15 years), and was limited to the rich Anglo part of town in any case, but I would not have predicted this at all.

One thing that the Forbes article does not discuss is the ethnic composition of the "winning" cities on this list. My very rough-and-dirty calculation comparing percent-Anglo (from US Census data) and percent-obese for the 20 cities on the list shows a correlation of -0.46, which would be considered a moderate correlation between ethnicity and obesity, with a higher percentage of the population non-Anglo (i.e. non-Hispanic white) being associated with higher obesity.

Also, Robert pointed out that the biggest US cities are not represented on the list. Not only Houston (#3 in population), but also NY, LA, Chicago, Phoenix, Philadelphia, Dallas, and San Jose; the only two cities of the top 10 most populated which show up are San Antonio and San Diego. I'm not sure whether this means much of anything, but it is odd. It's hard to credit an explanation like "large cities have better public transportation, which encourages physical activity more than driving does," given the car culture of Houston and Dallas. It may point to something screwy in the disproportionate sampling frame they used for getting representative statewide data when applied to city-level data.

I found the article from this post on Marginal Revolution. Check out comment #6 for some of my thoughts on the methodology and the never-ending "BMI doesn't measure fat" fantasy.

1 comment:

Anonymous said...

Dr. Oz of "You on a Diet" and several other books emphasizes waist measurement as an indicator of obesity/overweight. I believe for women it is anything over 34" and men 40" that is unhealthy.