Government analysts downgraded the annual death toll from obesity yesterday in a study that is certain to bewilder a public already obsessed with dieting and nutrition.
In fact, they inexplicably found that people who weigh a few pounds more than the ideal are less likely to die than those who weigh a few pounds less.
Taken together, the findings will undoubtedly leave scientists and consumers arguing over obesity's true role in mortality - though no one argues that being overweight is good for you.
The latest report by scientists with the U.S. Centers for Disease Control and Prevention estimates that obesity kills about 112,000 people a year, only a third of the number estimated just four months ago.
But Dr. Kathleen Flegal, who led the study reported in today's Journal of the American Medical Association, said the lower death estimate should not make consumers complacent about their expanding waistlines.
The study looked at death only, she said - not the role that girth plays in triggering illness or lowering a person's quality of life.
"I'm not sure what the public should take from it," said Flegal, who is with the CDC's National Center for Health Statistics. "I don't see it as the final answer to anything."
In a second study, CDC scientists report that people across all weight categories suffer less from high cholesterol and high blood pressure and smoke less than they did 30 and 40 years ago.
Obese people still suffer more from these diseases, but less so than before because of new treatments such as anti-hypertensive medications and cholesterol-lowering drugs.
Though the downgraded death toll could confuse a public already barraged with conflicting evidence about the roles that vitamins, herbs, fats and fiber play in their health, experts pointed out that there's little argument over the notion that obesity triggers diabetes, heart complications and arthritis.
Flegal said she did not deliberately set out to lower the CDC's controversial 2004 estimate that obesity kills 400,000 people annually - a figure the agency lowered to 350,000 after discovering a calculation error.
She said she had already spent years developing a new statistical method that more closely reflects the health of the general population and teases out the competing forces that age, smoking, drinking and obesity can bring to bear on mortality.
When she applied the new methodology, she said, the estimate simply came out lower.
More or less
The study, which mines data from three national health surveys spanning the 1970s through 2000, also found that people who are slightly overweight actually have a lower chance of dying than people who weigh a few pounds less.
The same cannot be said for people who are truly obese; they face a greater risk the more weight they gain.
Health experts agree that American are becoming fatter with each passing year, leading to an epidemic of diabetes, even among children.
Still, yesterday's announcement will hearten critics who have argued that government agencies, which spend millions convincing Americans to eat less and exercise more, have inflated obesity's death toll.
"A lot of the so-called health risks associated with being overweight have been oversold and overstated," said Glenn A. Gaesser, a University of Virginia physiologist and longtime critic of the earlier studies.
He said previous estimates failed to control for fat people who increased their risks by engaging in "yo-yo" dieting, taking diet pills or not exercising.
Those people, Gaesser said, should focus instead on things that are known to contribute to longevity - eating a balanced, healthy diet and getting regular exercise.
Diane Becker, an epidemiologist with the Johns Hopkins School of Medicine, said estimating the number of people killed by obesity is virtually impossible.
That's because scientists don't understand the relative roles played by obesity and other related afflictions - such as heart disease - in a person's death.
But Becker said the government's earlier estimates sound nearer the mark.
"That sounds closer to me, given what I work with every day in my research arena," said Becker, also a professor with the Johns Hopkins Bloomberg School of Public Health.
"People should not be assured by this."
In a separate study, Becker reports this week that being overweight raises the risk of suffering a "cardiovascular event" - such as a heart attack - by a factor of five. Being obese raises the risk 15-fold.
Body mass index
By the government's definition, overweight is defined as having a body mass index of 25 to 29, while obesity starts with a BMI of 30.
Body mass index is a function of height and weight. A person who is 5 feet 10 and weighs 174 pounds, for example, has a BMI of 25 and is simply overweight. If the person gains 25 pounds, his BMI is 30 and he is considered obese.