“Obesity is catching up to tobacco as the leading cause of death in America,” proclaimed Centers for Disease Control and Prevention chief Julie Gerberding this week. “Americans need to understand that overweight and obesity are literally killing us,” added Health and Human Services Secretary Tommy Thompson.
While it’s not disputed that severe obesity (search) may shorten life, the real killer in this case seems to be the CDC’s statistical malpractice.
The excuse for the desperate health warning is a study in the March 10 issue of the Journal of the American Medical Association in which the CDC claims that poor diet and physical inactivity caused 400,000 deaths in 2000. That estimate supposedly represents a 33 percent increase from the 1990 estimate and approaches the 435,000 deaths in 2000 supposedly attributable to smoking.
Now it’s been said that there are two types of statistics ― the kind you look up and the kind you make up. CDC’s body counts are definitely the latter.
The CDC produced its estimates with a statistical ruse called “attributable risk” ― the fearmongers’ method of choice for alarming the public with large body counts. Attributable risk (search) could be the poster child for the saying, “garbage in, garbage out.”
Without getting lost in the depths of statistical formulas, the key components of attributable risk calculations are statistical correlations between potential causes and effects, like overweight/obesity and premature death. But just because overweight/obesity and premature death might have been statistically correlated in some studies doesn’t mean that overweight/obesity has been proven to cause premature death.
In the few studies that have reported correlations between overweight/obesity and premature death, the vast majority of the correlations are small, not statistically significant (that is, they may be due to chance) and, in short, are unreliable. Reported correlations between overweight/obesity with premature death don’t start to inspire even minimal confidence until the obesity in question is extreme ― cases where you only need common sense, not statistical hocus-pocus.
Recklessly plugging unreliable statistical correlations into the attributable risk formula to produce sensational body counts can only be described as junk science.
But you don’t need to take my word about the folly of the CDC’s methodology.
As the New England Journal of Medicine editorialized in 1998, “Although some claim that every year 300,000 deaths in the United States are caused by obesity, that figure is by no means well established. Not only is it derived from weak or incomplete data, but it is also called into question by the methodologic difficulties of determining which of many factors contribute to premature death.”
“Calculations of attributable risk are fraught with problems … [and can produce] a nonsensical result,” noted the Journal.
And if all this is too technical, just ask yourself this question: Is it really plausible that the death rate from overweight and obesity has increased by 33 percent in the last 10 years?
Let’s not forget that despite all the hyperventilating about our health, the CDC reported last month (with much less fanfare) that U.S. life expectancy (search) ― the most objective measure of public health ― reached an all-time high of 77.4 years in 2002, up from about 75.2 in 1990.
So what gives? Why does the CDC insist on nagging us about our waistlines? Two reasons come to mind.
First, the previously mentioned New England Journal of Medicine editorial characterized the obesity obsession as an example of “a tendency to medicalize behavior we do not approve of” ― that is, politically incorrect activities like over-eating, not exercising, smoking, drinking, and gun ownership.
Next, the public health establishment (search) is simply running out of things to do. Preventing and controlling the spread of infectious disease, the traditional and primary mission of public health professionals, has largely been achieved. The relatively small number of infectious disease deaths that still occur annually, excluding AIDS-related deaths, decreased by 25 percent from 1990 to 2000, according to the CDC.
In former President Dwight D. Eisenhower’s famous 1961 speech warning us of a looming military-industrial complex (search), he also said, “The prospect of domination of the nation's scholars by Federal employment, project allocations, and the power of money is ever present and is gravely to be regarded. Yet, in holding scientific research and discovery in respect, as we should, we must also be alert to the equal and opposite danger that public policy could itself become the captive of a scientific-technological elite.”
Were Ike witnessing the health nannies’ apparent desire to control our behavior and bigger budgets, he might warn us of the looming “public health-industrial complex.”
Steven Milloy is the publisher of JunkScience.com, an adjunct scholar at the Cato Institute and the author of Junk Science Judo: Self-Defense Against Health Scares and Scams (Cato Institute, 2001).