The recent paper contradicting this campaign comes from the scientific branch of CDC, and is available in full text from JAMA. Even so, the authors bury their main finding in the abstract:
"Overweight was not associated with excess mortality (–86 094 deaths; 95% CI, –161 223 to –10 966). "
[The full text of the abstract is reproduced below, to enable the reader to draw the reader's conclusions on distortion].
BrooklynDodger is mystified why an author would write "not associated with excess mortality" when the actual result is a decrease in mortality which is statistically significant. Most of the words are devoted to obese vs. normal, rather than overweight vs. normal.
For a six foot tall person, 135 lbs borders underweight, 185 borders overweight, and 235 borders obese. In this survey, barely 1/3 of the population was of "normal" weight. The benefits of "overweight" compared to "normal" were greater among non-smokers. The study took smoking into account, since lower weight strata are richer in smokers than higher strata. BrooklynDodger would not be surprised if there were something unhealthy in 6 foot persons weighing less than 135 lbs.
BrooklynDodger thinks the op-ed piece below misses the boat. It's not moralism and scolding which has driven the latest campaign. The motivation is more pernicious. It's a campaign to blame victims, to promote removing benefits from victims.
http://jama.ama-assn.org/cgi/content/full/293/15/1861
Excess Deaths Associated With Underweight, Overweight, and Obesity
Katherine M. Flegal, PhD; Barry I. Graubard, PhD; David F. Williamson, PhD; Mitchell H. Gail, MD, PhD
JAMA. 2005;293:1861-1867
Results Relative to the normal weight category (BMI 18.5 to <25),>
http://www.nytimes.com/2005/04/23/opinion/23tierney.html?incamp=article_popular_3
OP-ED COLUMNIST
Fat and HappyBy JOHN TIERNEY
Published: April 23, 2005
1 comment:
Interesting - the authors do appear to brush off the decrease in mortality in the overweight group. Even in the Comments section they do not mention the decrease in mortality. In fact, it appears (and this is only on second glance, so don't quote me) that the conclusions about obesity (>30) in the abstract are due to combining the 30-35 and >35 groups, whereas Figure 2 appears to indicate little excess mortality in 30-35.
It deserves a little more reading than I have time for tonight, but it does appear to be insufficiently discussed.
Post a Comment