A study using national health survey data has found varying associations between body mass index and mortality, depending on the cause.
Using data on cause-specific relative risks of mortality from the National Health and Nutrition Examination Survey (NHANES) from 1971–2002, Katherine M. Flegal, Ph.D., and colleagues looked at the association between body mass index (BMI) and excess deaths associated with three different BMI categories: underweight (BMI less than 18.5), overweight (BMI of 25 to less than 30), and obesity (BMI of 30 and over). Deaths were divided into three major categories: cardiovascular disease (CVD), cancer, and all other causes (noncancer, non-CVD causes). The normal-weight category was used as the reference group.
The underweight category was associated with significantly increased mortality from noncancer and non-CVD causes, but was not associated with increased cancer or CVD mortality.
The overweight category, however, was associated with significantly decreased mortality from noncancer and non-CVD causes. This category was not associated with cancer or CVD mortality, but was associated with significant increased mortality from diabetes and kidney disease. The net result was that the overweight category was associated with significantly decreased all-cause mortality overall, the authors reported.
The obese category was associated with significantly increased mortality from CVD, some cancers, and diabetes and kidney disease. There was no significant association between obesity and cancer mortality overall, or with noncancer, non- CVD mortality. But it was associated with increased mortality from obesity-related cancers such as colon, breast, esophageal, uterine, ovarian, kidney, and pancreatic cancer (JAMA 2007;298:2028–37).
These data “indicate that the association of BMI with mortality varies considerably by cause of death,” the authors concluded. These results also help clarify their findings in an earlier study, which found “excess overall mortality associated with underweight and obesity but not with overweight.”
In an interview, Dr. Flegal, senior research scientist at the National Center for Health Statistics, Hyattsville, Md., and lead author of the study, said the study's results were similar to those in other studies and are not intended for clinical use. Instead, they are intended to make estimates of the contribution of obesity and overweight to excess deaths.
The current study is an extension of a study, published in 2005, which determined that, based on national survey data from 2000, all-cause mortality was significantly increased in the underweight and obese categories and significantly decreased in the overweight category when compared with normal-weight categories.
Dr. Flegal said although some media coverage of that study suggested the findings were unusual, the association of overweight with mortality that is similar to or lower than that for normal weight, as well as the idea that being overweight may offer some survival benefits, have been found in other studies.
She emphasized that the relationship between BMI and mortality is complex. The study is not “the arbiter of whether it's OK to be overweight or not.”
Asked to comment on the study's findings, Dr. Jeffrey I. Mechanick said its implications could “easily and dangerously” be distorted and should not be interpreted to mean that the results support allowing oneself to remain overweight or that dieting to achieve a “normal” BMI may not be medically indicated. Overweight people are at risk for diseases associated with a higher morbidity and mortality rate, including diabetes, obesity, and metabolic syndrome, said Dr. Mechanick, director of metabolic support and clinical professor of medicine at Mount Sinai School of Medicine, New York.