I was reading Michael Pollan's article in today's NYT where he made some very provocative claims about the healthcare debate. His main argument is that there is an "elephant in the room" and that most of the blame for the enormous costs of healthcare in the US can be placed on the fact that Americans are far more likely to be obese.
I thought his piece was an interesting "out of the box" look at the issue, and that an interested person should be able to verify his claim by simple searching the web for health statistics. I immediately thought of using wolframalpha.com which is promoted as a "computational knowledge engine" that I had imagined would give a quick, graphic access to widely available statistics. I tried several searches using keywords like, weight, obesity, BMI - all to no avail. So WolframAlpha appears to be useless for comparative health statistics- too bad. Can any major search engine handle the task? A further search using google came up with several not so satisfactory sites. One was a database the WHO Infobase that was difficult and slow. Another was NationMaster site nationmaster.com which gave chart using OECD stats.
This chart showed that US has obesity rates of over 30% with an industrialized country average of 14%. So in essence Americans are twice as likely to be obese, as they spend twice as much on healthcare. Could more (web) data show a stronger correlation? Pollan may be on the right track in his focus on this aspect of the health debate. And here come those corporate interests that he predicted would fight any changes to food industry practices.
Hi Dr. Fred,
ReplyDeleteAnother place to check for this type of data is graphs.gapminder.org , though they're missing obesity data (they have sugar consumption per capita).
I went back to the OECD data used by nationmaster.com here:
http://www.oecd.org/document/16/0,2340,en_2649_34631_2085200_1_1_1_1,00.html
and charted the correlations between obesity vs. spending as a % of GDP and by PPP, and didn't see a correlation.
There is a very clear negative correlation between life expectancy and obesity - to give a rough idea for the nations listed in this document (mostly rich countries), excel showed an R-square of .72 for the linear regression line. For these rich countries, there doesn't appear to be much, if any, correlation between PPP heath spending and life expectancy at birth.
So, for rich countries, my superficial examination of the OECD data indicates little correlation between heath spending and obesity, little correlation between heath spending and life expectancy, and a strong correlation between obesity and life expectancy.