Obesity: How big a problem is it?

Evan David Rosen, M.D., Ph.D. Assistant Professor of Medicine, Harvard Medical School

As any casual reader of the popular press knows, we're in the middle of an obesity "epidemic" that will increase mortality rates and drive up health care costs for years to come. The severity of this problem has been emphasized by several high profile publications that estimate approximately 400,000 Americans will die each year as a result of being too fat, a number that rivals those associated with cigarette smoking. Obesity exerts these negative effects by causing type 2 diabetes, cardiovascular disease, and certain cancers.

Two new studies by a group from the Centers for Disease Control (CDC), however, cast doubt on some of these assertions by claiming that far fewer obese people are dying than previously thought. In the new estimates, obesity only accounts for approximately 112,000 excess deaths per year, still serious stuff but not even close to past calculations. Furthermore, these papers show that the excess deaths are only seen when body mass index (BMI) gets over 30. The group with a BMI of 25-29.9, which is considered overweight, actually had fewer deaths than folks with a BMI 18.5-24.9, which had been considered ideal. This last finding led to newspaper headlines actually proclaiming the health benefits of being overweight. This was predictably followed by an outcry from folks in the "fat is bad for you" camp, and by acclamations of "I told you so" from people who feel that the stigmatization of overweight has gone too far. This latter group has notably included lobbying groups associated with the junk food industry, who immediately took out ads in newspapers across America trumpeting the new studies. Those who aren't wed to any particular opinion, but just want to know the facts, are mostly just confused.

Several factors are likely to account for the discrepancy between past and current estimates. For one, the new papers used more recent data for evaluation than some of the older papers. The newer data suggest that obesity causes death less often than in the past because we've gotten better at treating the proximal causes of disease, such as high cholesterol and hypertension. Smoking rates have also gone down for both lean and obese people. This is reflected in the significant downward trend in deaths from cardiovascular disease noted in recent years, and increases in overall life expectancy.

The new studies also used different statistical methods to determine the risks of dying from weight gain. I am far from an expert in statistics, but I can say that I've spoken to people on both sides of the debate that are quite certain that their methodology is superior to the other. While that's not particularly satisfying, it is worth pointing out (as the authors of the new study do) that when measuring small effects, even minor changes in the numbers can have big effects. Obesity increases the risk of death for any individual person only very slightly. The reason that the number of excess deaths has been calculated to be so high is because so many people are obese. If you change the risk of dying from obesity from, say 2% to 3%, a relatively small difference, you increase the number of deaths by a lot. As a concrete example, the authors of the new study claim 95% statistical certainty that the number of excess deaths due to obesity is between 54,000 to 170,000, a huge range. If their raw estimates were even a little bit off, the number could potentially be much higher, and might overlap older estimates.

A few critical points need to be made that might otherwise get lost in the statistical arguments. First, the new studies looked at mortality only. But the troubles caused by being overweight are not captured solely by death rates. Obesity causes type 2 diabetes, osteoarthritis, gall bladder disease, and is strongly associated with depression and other psychological ailments. The human and financial costs of these conditions need to be accounted for as well. Second, the reduction in deaths because of improved medical care (e.g. better lipid lowering drugs) belies the fact that more of such care is required for overweight and obese people, which also adds to medical costs. Third, soon-to-be-published data suggests that when obese people get even fatter, their risk of dying goes up further. This means that it is critical to get your weight down as much as possible, even if you remain in the obese range after weight loss. Finally, even the lowest reasonable estimate of excess deaths due to obesity is unacceptable—54,000 deaths is the equivalent of eighteen September 11th attacks every year in human costs!

There will continue to be quibbling about the numbers, but we can't let that distract us from grasping the importance of this problem, and thus from continuing to seek solutions.

Reference:
1. Flegal KM, Graubard BI, Williamson DF, Gail MH. Excess deaths associated with underweight, overweight, and obesity. Journal of the American Medical Association. 2005 Apr 20; 293(15):1861-7.
2. Edward W. Gregg, Yiling J. Cheng, Betsy L. Cadwell, Giuseppina Imperatore, Desmond E. Williams, Katherine M. Flegal, K. M. Venkat Narayan, and David F. Williamson. Secular Trends in Cardiovascular Disease Risk Factors According to Body Mass Index in US Adults. Journal of the American Medical Association. 2005;293:1868-1874.

Viewpoint is an editorial column that expresses the opinion of the specific Medical Director, who is solely responsible for its content. Viewpoint does not represent the views or opinions of Veritas Medicine and does not reflect the opinions of other physicians and researchers.

If you have feedback about this article, Click Here