Excess weight causes 464,000 deaths per year in US, 4 times more than 2005 CDC estimate

Posted by .(JavaScript must be enabled to view this email address)
Friday, December 15, 2006 12:23 am Email this article
Excess body weight causes an estimated 464,000 excess deaths per year in the U.S. according to a new paper by James Greenberg from Brooklyn College of the City University of New York. This is four times more than the 111,909 estimated last by the CDC (Flegal et al, 2005), which failed to adjust for two statistical biases.

In 2005, the CDC estimate of 111,909 premature deaths per year was far below previous estimates of 280,000 to 580,000.

Bias #1

Bias #1 : Not adjusting for body weight of sick people

One factor is called regression-dilution. This occurs because people who are heavy or thin tend to be heavier or thinner for months or years before and after surgery.

Not adjusting for this causes there to appear to be more deaths among thin people than there really are, and thus, fewer deaths among among overweight people than their really are.

 

Bias #2

Bias #2 : Not adjusting for people who are thin because they have an illness

The CDC paper did not adjust for another bias called reverse causation. This is caused by the fact that very thin people are often thin because of a serious illness, such as cancer, and have a greater risk of death because of this.

However, if this is not adjusted for, and this group of people are included with normal weight people, it makes it appear as if normal weight people are at a great risk of death than they really are, and it also makes it appear as if obese people are at a lower of a risk of dying than they really are.

 

Corrected Estimates

Correction #1 : Deaths attributable to excess weight : 347,000 per year

Correcting for these biases increased the estimated number of deaths 3.1-fold. This means the yearly number of deaths attributable to excess weight—overweight and obesity—would actually be 347,000.

Correction #2 : Deaths attributable to overweight : 464,000 per year

If, rather than using a body mass index (BMI) of 18.5 to 25 as normal weight, they used a BMI of 23 to 25 as normal weight, then the estimated number of deaths attributable to overweight and obesity would be 3.9 times greater than Flegal estimated.

This means the number of deaths attributable to excess weight would actually be 464,000.

 

Corrected Estimates for Overweight

Correction #3 : Deaths attributable to overweight : 211,000 per year

The 2005 CDC paper also estimated that there were 86,094 fewer deaths among people who were overweight compared to if they had been normal weight.

This is nonsense.

Had the paper adjusted for the statistical biases mentioned above, they would have found than there were approximately 211,000 excess deaths among people who were overweight.

To me, this was the obvious problem with the Flegal CDC 2005 study in which she claimed that overweight people were less likely to die than normal weight people, which meant that normal weight people were more likely to die than overweight people, even among those who had never smoked. This just does not make sense.

 

Comment

Comment : What an incredible shame

What an incredible shame that the 2005 CDC study by Flegal did not adjust for these biases. The estimates they gave were bogus.

The problem with this is that people wishing to claim that excess weight is not as much of a health problem as previously thought will point to the Flegal paper for years to come to support their position and say, “It came from the CDC so it must be true.”

Comment : Will the CDC try to justify their mistakes?

I hope this is not the case, but human nature being what it is, I imagine that Katherine Flegal and her coauthors at the CDC will backpeddle and try to justify their paper and their failure to adjust for these biases.

I can only hope that they would simply admit their mistakes, put out a corrected estimates, and have a media campaign to alert the public that excess body weight is a much greater risk than they previous led us to believe, but I wouldn’t bet on it.

REFERENCE

Greenberg J. Correcting biases in estimates of mortality attributable to obesity. Obesity (Silver Spring). 2006 Nov, 14(11):2071-79.

AUTHOR’S CONTACT INFORMATION

James Greenberg
Department of Health and Nutrition Sciences
Brooklyn College of the City University of New York
2900 Bedford Avenue
Brooklyn, NY 11210
.(JavaScript must be enabled to view this email address)

OTHER REFERENCES

Flegal KM, Graubard BI, Williamson DF, Gail MH. Excess deaths associated with underweight, overweight, and obesity. JAMA. 2005, 293(15):1861-67

Articles on the same subject can be found here:

• Deaths from obesity