[extropy-chat] Overweight people live longer
hal at finney.org
Wed Apr 20 16:42:54 UTC 2005
According to a new study published in the Journal of the American Medical
Association, moderately overweight people actually have lower mortality
rates than people of supposedly ideal weights. The study is available
for free access at http://jama.ama-assn.org/cgi/content/full/293/15/1861
but here is the most relevant table:
Table 2. Relative Risks by Age Group and BMI Level From the Combined
NHANES I, II and III Data Set
Relative Risk by Age Category
BMI Level 25-59 y 60-69 y >= 70 y
<18.5 1.38 2.30 1.69
18.5 to <25 1.00 1.00 1.00
25 to <30 0.83 0.95 0.91
30 to <35 1.20 1.13 1.03
>=35 1.83 1.63 1.17
<18.5 1.25 2.97 1.50
18.5 to <25 1.00 1.00 1.00
25 to <30 0.66 0.81 0.90
30 to <35 0.77 1.21 1.13
>=35 1.25 2.30 1.12
(Note, I just typed this in by hand and left off the 95% confidence
interval range. See table 2 from the link above to see all the details.)
Lower numbers are better! You can see that the overweight range of 25
to 30 BMI has consistently lower mortality than the "ideal" range of
18.5 to 25. The effect is most pronounced among non-smokers 25 to 59
years old, where being overweight reduces mortality rates by 1/3. Even
being obese with a BMI up to 35 reduces mortality by 1/4 in that group.
A 5'10" person could weigh up to 243 pounds; a 5'4" person could weigh
203, and still have BMI less than 35.
This result is causing considerable press commentary (for example
<http://www.nytimes.com/2005/04/20/health/20fat.html>), much of it
focusing on another result of the paper, which re-evaluated the net
deaths due to obesity. A report last year from the CDC had estimated
400,000 deaths per year from obesity (in the U.S.), almost at the level
due to smoking and adding fuel to the government's new anti-obesity
efforts. That report had been widely attacked as over-stated, however.
The new result suggests that obesity is causing 112,000 deaths, but that
being overweight is preventing 86,000, for a net cost of 26,000 deaths.
This compared to 34,000 excess deaths from being underweight. The lesson
appears to be that Americans are not obese enough!
Extropians may not find these results to be too surprising. For years
list member Doug Skrecky posted similar data on a regular basis.
His position was that it was far more important to be fit than thin.
Physical fitness was, according to the data he had found, a much better
prediction of mortality than weight. "Fit but fat" people actually had
the lowest mortality of all. (However it's not all that easy to get fit
when you're fat, and/or to stay fat when you're fit, so fit but fat is
a pretty small minority.)
This data may call into question the benefits of calorie restriction
(CR) for extending lifespan. The Times article says:
"As for whether there is truly a mortality risk in being underweight,
Dr. Mark Mattson, a rail-thin researcher at the National Institute on
Aging who is an expert on caloric restriction as a means of prolonging
life, said it was not clear that eating fewer calories meant weighing so
little, since some people eat very little and never get so thin. In any
event, while caloric restriction may extend life, Dr. Mattson said,
'there's certainly a point where you can overdo it with caloric
restriction, and we don't know what that point is.'"
It seems odd that he would say that maybe CR is OK because it doesn't
necessarily make you thin, while he himself is described as "rail-thin"
and the unstated implication is that he engages in CR himself.
All in all it is an exciting result and it is always good to see the
conventional wisdom being overturned, especially on a matter where public
health officials have been scolding Americans for years. Hardly a day
goes by when we don't hear some new ominous result about the obesity
epidemic. Now, if it turns out that overweight is actually healthy,
people will be forced to re-evaluate these results. Ambiguity leads to
critical thinking, which is always preferable to mindless dogma.
More information about the extropy-chat