A webcast is available at http://www.hsph.harvard.edu/weighing_the_evidence/. The event was moderated by Meir Stampfer, chair of the Department of Epidemiology, and Walter Willett, chair of the Department of Nutrition, and was sponsored by the Department of Nutrition and the HSPH Office of Communications.

Confusion and Counting

The CDC team, led by statistician Katherine Flegal, published the paper in the April 20, 2005 issue of the Journal of the American Medical Association (JAMA). In addition to the purported health benefit of carrying around some extra pounds, the study also cut the estimated number of deaths in the U.S. related to being overweight or obese by more than half, conflicting with other CDC statistics. Earlier, a separate CDC team indicated that about 365,000 Americans die each year due to problems linked to extra weight. The Flegal study asserted the number was closer to 112,000, and the number of excess deaths could be reduced further if one considered the number of lives extended by extra weight. Flegal and her co-authors were unable to attend the HSPH symposium. Instead, Donna Stroup, Director of CDC's Coordinating Center for Health Promotion, represented the CDC at the School's event.

CDC Director Julie Gerberding later apologized for any confusion created by the emerging science on obesity at a press conference in Atlanta a week after the HSPH symposium, stressing that extra weight harms health, even if the exact relationship between weight and premature death remains uncertain. In answering a question about the Flegal study and the relationship between being overweight and poor health outcomes, Gerberding said that, "there are some statistical aspects of the way the study was designed and the data sources used in that, that the author herself would not claim that overweight as protective of ill health. So we want to be very clear-obesity and overweight-and I'm sorry to say overweight because I know a lot of people were hoping that CDC was going to come out and say it was okay to be overweight, but we're not saying that. It is not okay to be overweight." A transcript is available at http://www.cdc.gov/od/oc/media/.

A Difference in Method

The results of the Flegal study "caused a lot of sensation in the press [and] caused a lot of confusion on the part of individuals and health care providers, so we thought it was really important to go into this topic in depth and really look at the science underlying the relationship between body weight and health-mortality being part of that picture, but certainly not being the only part of that picture," said Willett at the HSPH symposium.

Two HSPH panelists, JoAnn Manson and Frank Hu, suggested that the conclusions reached by the Flegal study were the logical outcomes of flawed methodology.

The Flegal study based its estimates on three National Health and Nutrition Examination Surveys (NHANES). The average periods of follow-up were 19 years for NHANES I, 14 years for NHANES II, and 9 years for NHANES III. The NHANES surveys are nationally representative and include direct measurements of height and weight.

Manson explained that she thought methodologic limitations, particularly lack of control for pre-existing diseases that lead to weight loss, are the crux of what she perceives as a problem with the recent study and may help explain discrepancy in findings. Manson is a professor of medicine at Harvard Medical School and a professor in the Department of Epidemiology at HSPH.

She and Hu pointed out that the Flegal study did not effectively exclude patients with chronic diseases or adequately control for smoking at baseline at the same time in the analyses. Smokers and those with chronic illness tend to be lean-but live shorter lives than people who have never smoked or who are healthy-artificially inflating the premature death rate for those of normal weight and making the slightly overweight group look good in comparison. Both problems need to be addressed simultaneously.

"It's biologically implausible, at least in my view, that overweight reduces mortality," said Manson, "and, I believe and think many of us agree, that this finding is more likely due to methodologic bias and artifact."

Manson and Hu, who conduct research using data from the Nurses' Health Study, reiterated that data from that study clearly indicates increased mortality in people who are overweight compared to those who are normal weight. Meanwhile, weight gained during adulthood significantly predicted increased mortality, said Hu.

More than Death Rates

CDC representative Stroup spoke of the impact of the weight problem on this country. "Obesity-related diseases are costly and diminish quality of life," she said. The Flegal study adds to the knowledge of the relationship of body weight and mortality, she said, but "there will be further studies that bring us further in our knowledge."

Stroup outlined challenges the CDC sees to obesity prevention, including the need to clarify the scientific basis of weight-related mortality and morbidity. The agency also strives to bridge the gap between clinical care and public health practice related to weight, realign priorities for program delivery, and emphasize the prevention of early risk factors.

Discussant Scott Grundy, director of the Center for Human Nutrition at the University of Texas Southwestern Medical Center, also touched upon the burden of morbidity caused by excess weight. He spoke from Dallas via a live video conference link.

"Just because it may be difficult to define the role of obesity or overweight [in mortality], that doesn't mean to the individual that there won't be a consequence to being overweight," Grundy said.

The percentage of obese adults in the U.S. has more than doubled in nearly 40 years, accompanied by a tripling in prevalence of overweight in children and adolescents. Being overweight or obese has been associated with increased risks for a multitude of health risks, such as diabetes and heart attacks.

Michael Thun, vice president of epidemiology and surveillance research at the American Cancer Society, shared findings from the Cancer Prevention Study II (CPS-II), which involves 1.2 million men and women in 50 states. Those findings suggest an increased relative risk of mortality for breast cancer among obese women and for colon cancer among obese men, and an increased risk of overall mortality among both overweight and obese people.

Graham Colditz, professor of medicine at HMS and of epidemiology at HSPH, spoke of another aspect of obesity-the burden it places on U.S. society in terms of cost. "Direct health costs related to obesity run at more than seven percent of total health care expenditures in the U.S.," he said. "Indirect costs due to lost productivity may well exceed this direct cost."

--PHC