Study Overstated Obesity Deaths, Its Authors Say
A highly publicized government study that reported obesity was poised to become the leading preventable killer of Americans contained serious errors that inflated the numbers, the study’s authors conceded in a letter published today in a major medical journal.
But critics of the controversial study remained unsatisfied, saying that the study has even deeper methodological flaws that further exaggerate obesity’s death toll.
The correction in the Journal of the American Medical Assn. came 10 months after the original study was published. The study estimated that 400,000 Americans die annually from obesity -- just 35,000 fewer than those who die from the leading preventable killer, tobacco.
Written by scientists at the Centers for Disease Control and Prevention, including the center’s director, Dr. Julie Gerberding, the study was cited prominently in the March 2004 kickoff of a government anti-obesity campaign. But even before it was published, the report was criticized as methodologically shoddy by some scientists at the CDC and other researchers.
In June, Rep. Henry A. Waxman (D-Los Angeles) requested a Government Accountability Office investigation into the matter. The “CDC needs to disclose how a flawed paper was cleared within the agency and what measures are in place to prevent similar problems in the future,” he said Tuesday.
The CDC conducted an internal investigation to discover why the objections of some of its scientists went unheeded.
That report has not been made public.
The correction downgraded the annual obesity death toll from 400,000 to 365,000 a year. The CDC blamed computational errors for the mistake.
Lead author Ali Mokdad and coauthors wrote in the letter that their principal conclusions remained unchanged: Obesity, caused by inactivity and poor diet, is a leading killer -- and the death toll is increasing.
However, critics disagreed, saying that the actual death toll from obesity could be much lower because of methodological errors in calculating the risks of death.
“They’re still stonewalling and still denying the fundamental problem,” said Stanton Glantz, professor of medicine at UC San Francisco.
Critics said the original study, for example, assumed that the death risk associated with obesity is the same for younger adults and for the elderly. Some studies suggest that older people have a lower risk of obesity-related death.
Last year, two papers co-written by CDC scientists argued that such an assumption would lead to inaccurate estimates.
“There’s absolutely no question that it’s overestimated,” said Daniel McGee, professor of statistics at Florida State University in Tallahassee.
“It just goes beyond common sense that so many deaths in the elderly would be due to obesity,” he said.
The result could be an inflation of the death toll by 100,000 or more, some experts said.
Not all researchers agree. Dr. Michael McGinnis, a senior scholar at the Institute of Medicine, which advises the government on medical issues, said that he believed the methods used in the CDC paper were reasonable and that the numbers they arrived at were likely to be in the ballpark.
Gerberding said the CDC is studying ways to improve its risk estimates.