The death toll from being overweight or obese is far less than a controversial government estimate released last year that body fat killed about 400,000 people annually and was poised to outstrip tobacco as the leading preventable cause of death, according to a new study.
The study estimated that obesity killed about 112,000 people, most of whom were extremely obese with body sizes equivalent to a 5-foot-4-inch woman who weighed 204 or more pounds. The effects of milder obesity on death were less severe.
In addition, the scientists found that being underweight killed an estimated 34,000 Americans each year. Risk of death from being underweight was especially high among the elderly.
The scientists reported that people who were merely overweight, as opposed to obese, suffered 86,000 fewer deaths than those whose weight was in the so-called healthy range.
"I think the punch line is obesity is associated with an important number of excess deaths but there has been a little too much hysteria before," said Roland Sturm, a senior economist at Rand Corp., a Santa Monica think tank. "Now we come to some probably more credible, reasonable numbers."
But some scientists said the new study had methodological problems that could underestimate the death toll from fat.
"I think the papers are really naive, deeply flawed and seriously misleading," said Dr. Walter Willett, chairman of the department of nutrition at the Harvard School of Public Health. "This is a huge problem, it's getting worse fast and there's no turnaround in sight."
The paper, published today in the Journal of the American Medical Assn., was written by Katherine Flegal and David Williamson of the Centers for Disease Control and Prevention, and Barry Graubard and Mitchell Gail of the National Cancer Institute.
The team used three national population surveys to calculate the risk of death associated with weight, which was measured using the body mass index, or BMI. The index estimates plumpness based on someone's height and weight. A BMI of 18.5 to 24.9 is considered healthy; 25 to 29.9 overweight; and 30 and above, obese.
The scientists then used a more recent national sample collected between 1999 and 2002 to calculate the total number of people who fell into different BMI categories.
Putting the two together, they could estimate how many people died each year because of their weight.
The scientists reported that the death risk from obesity appeared to be lessening over time.
The trend could be a result of improved lifestyle and treatment of heart disease, the authors said. Heart disease is one of the main causes of death associated with obesity.
Paul F. Campos, a University of Colorado law professor and author who believed obesity has been overstated as a health risk, said the new study called for a reevaluation of the relationship between weight and health.
He pointed to the study's finding that overweight people -- those with a BMI between 25 and 29.9 -- had fewer deaths than people in the healthy weight range.
"I think this is just a bombshell," Campos said. "The real undeniable fact is the entire overweight definition ... is a completely spurious definition."
Stanton Glantz, professor of medicine at UC San Francisco, said the new numbers showed obesity was nowhere near as serious a health concern as tobacco.
Obesity "is presented as a crisis and it's presented as this horrible problem which has exploded onto the scene," Glantz said. "What this paper shows is that it's just not true."
But some scientists said the study was flawed.
One of the main problems, Harvard's Willett said, is that he did not believe it adequately controlled for smokers and the sick. Because those people tend to be thinner -- for unhealthy reasons -- they could make the healthy weight category appear to have a much higher death rate.
In studies that remove smokers and the sick, overweight people are clearly at higher risk for disease and death, Willett said.
The authors said that they did remove smokers and some other high-risk groups from their study, with no significant difference in results.
The authors and other scientists cautioned that the study did not mean that being overweight or obese was harmless.
Carrying extra fat is associated with an increased risk for diabetes, heart disease and certain cancers, which cause complications, medical costs and an eroded quality of life.
"Overweight is associated with other factors than just mortality," Flegal said. "There's quality-of-life issues.... I think we couldn't use our data to really make a statement that overweight is benign."
The study builds on a controversy that has been roiling since an earlier estimate of 400,000 annual obesity deaths was published last March. (That number was reduced to 365,000 in January after the scientists found an error.)
The paper, written by four CDC scientists including the center's head, Dr. Julie L. Gerberding, was heralded with much fanfare. Its publication coincided with the unveiling of a new government anti-obesity campaign.
It was later revealed that some CDC scientists had objected to the paper's methodology before its publication, and the matter eventually led to an internal investigation at the CDC.