HARVARD GAZETTE ARCHIVES

Staff illustration Georgia Bellas/Harvard News Office

The event, "Weighing the Evidence: A Forum to Examine the Latest News About Overweight, Obesity and Mortality in America," presented a panel of health experts from the Harvard School of Public Health, the American Cancer Society, the Centers for Disease Control and Prevention (CDC), and the University of Texas Southwestern Medical Center.

Michael Thun

The panel examined the methodology of an April 20 study led by Katherine M. Flegal of the CDC's National Center for Health Statistics. Published in the Journal of the American Medical Association, the study showed that being somewhat overweight resulted in a slight reduction in mortality.

After its publication, the study prompted a wave of media headlines questioning the longstanding assertion that gaining weight is bad for your health. Food industry groups also seized on the research, running at least one full-page newspaper advertisement calling the obesity epidemic "hype."

"It is irresponsible for commercial interests to trivialize what has become a major health problem in the United States and globally," said Michael Thun, vice president for epidemiology and surveillance research at the American Cancer Society.

The controversial CDC study comes after years of data showing that the American public is rapidly gaining weight. Just since 1991, the rate of being grossly overweight, or "obese" has increased dramatically. According to CDC data, in 1991 just four states had obesity prevalence rates of 15 to 19 percent and no states had rates of 20 percent or above. By 2003, four states had obesity rates above 25 percent, 31 states had rates of 20 to 24 percent, and 15 states had obesity rates of 15 to 19 percent.

America's weight gain has been linked to a greater danger of developing heart disease, high blood pressure, some cancers, stroke, and diabetes. Earlier studies pegged the number of extra deaths due to obesity as being much higher than that shown by Flegal. As recently as January, the CDC estimated that 365,000 die of obesity-related causes each year, second only to smoking.

The CDC measures America's weight based on a comparison of height to weight called "body mass index," or BMI. People with a body mass index of between 18.5 and 24.9 are considered "normal" weight. Those with a BMI greater than 24.9 are divided into two categories: 25 to 29.9 are considered "overweight" while those with a BMI of 30 or greater are "obese."

For example, someone 5 foot 4 inches tall would be considered "normal" weight at 120 pounds, "overweight" at 150 pounds, and "obese" at 180 pounds. Someone 5 feet 10 inches would be considered "normal" weight at 150 pounds, "overweight" at 180 pounds, and "obese" at 210 pounds.

While BMI is a useful guide, the CDC points out that it is just one of several factors - including level of activity, diet, and family history, among others - that affect one's risk of developing a weight-related disease.

The Flegal study authors said their results may reflect improving health care over time, and particularly a reduction in deaths due to heart disease, which had been a major cause of death due to obesity.

The School of Public Health panel, which was sponsored by the School's Nutrition Department and Communications Office, said the results are related to how the data was collected and analyzed, particularly with respect to smoking and age. While smoking is associated with higher death rates, people who smoke tend to be thinner than nonsmokers. If they were included, panelists said, any data examining mortality and weight would be artificially skewed, giving a higher death rate for thinner people than otherwise would be the case.

Old age is a similar confounding factor, panelists said. The elderly often will lose weight, panelists said, but because of the differential loss of muscle mass and fat and the high prevalence of chronic diseases in the elderly, the weight loss in that group will also skew results by increasing mortality among those with lower body mass indexes.

Other studies that have taken into account those factors, panelists said, show that obesity and mortality are directly related, with the death rate rising steadily with weight gain.

"Overweight and obesity are associated with a substantial burden of disease, both fatal and nonfatal. It is biologically implausible that overweight reduces mortality," said School of Public Health Professor of Epidemiology JoAnn Manson. "Even modest weight loss has been shown to have health benefits."

Associate Professor of Nutrition and Epidemiology Frank Hu said information from the long-running Nurses Health Study shows no evidence of a substantial decline in the influence of obesity on mortality over time, as the Flegal study authors suggested. In fact, Hu said, once corrected to exclude pre-existing illnesses and smoking, there is clear evidence of higher mortality in the overweight group.

"Weight gain during adulthood is directly associated with increased mortality," Hu said.

Scott Grundy, director of the Center for Human Nutrition at the University of Texas Southwestern Medical Center, and Harvard School of Public Health Epidemiology Professor Graham Colditz both voiced concerns that the controversy has created a garbled message for the public on an important issue. With a sole focus on mortality, they said, the clear evidence that being overweight negatively impacts quality of life by increasing the incidence of chronic diseases such as diabetes is being lost.

Donna Stroup, director of the CDC's Coordinating Center for Health Promotion, supported the Flegal study but praised the exchange at the event, saying that obesity-related science is still young and that much work needs to be done.

"I think we're all trying to get to the same side of the issue that obesity is an important public health problem," Stroup said. "Our differences are less important than doing something [to create] interventions that work."