Federal researchers released a new study on obesity last week which created as much excitement as a Good Humor Man at a Weight Watcher's meeting. The headline grabbing conclusion, as reported by the NY Times - "Some Extra Heft May Be Helpful, New Study Says".
Oh, my. The study confirmed the widely-held belief that the truly obese have high mortality rates. However, the notion that "extra heft may be helpful" followed from the reported result that people classified as "overweight" have a slightly lower mortality rate than the group researchers put in the "normal" weight category.
Pundits immediately leaped to the wrong conclusion. Here is John Tierney of the Times:
After decades of listening to emaciated ascetics lecture us about diet and exercise, it's tempting to return the favor. We could turn into activists ourselves and stand in picket lines outside gyms with signs proclaiming, "StairMaster = Death."
David Brooks, also of the Times, delivered a virtually perfect misreading of the report:
People who work out, eat responsibly and deserve to live are more likely to be culled by the Thin Reaper.
That is simply not what the report said, which led Blogger "Ogged" to chastise Mr. Brooks with energy and commitment.
So, what did the researchers really say, and why the confusion? A few words describing the study will be helpful.
The study looked at health surveys that followed people from the 1970's and forward to observe mortality patterns among different groups. Because of the data collected, researchers were able to control for certain variables, such as age, race, smoking, and alcohol consumption. However, researchers lacked other data such as information on body composition (e.g., percent body fat), or measures of cardiovascular fitness. (These shortcomings have been corrected in surveys from 1999 and forward.)
The researchers divided the survey participants into five groups - thin, normal, overweight, obese, and very obese - based on a simple measure called "Body Mass Index", which is calculated based on a person's height and weight. Tracking mortality for the different groups (and subgroups, based on age, race, smoking, and alcohol use) the researchers reached the headline-grabbing conclusion that the "overweight" group had the lowest mortality rate.
Now, John Tierney and David Brooks are familiar with the use of statistics, so they know a guideline that is critical in understanding this report - Correlation is not Causation. A fun illustration of this truism is the correlation of ice cream sales with shark attacks on swimmers - both are down in December and up in July, but no one has concluded that sharks are hungry for ice cream eaters, or that banning ice cream sales would improve public safety.
Bearing that rule of statistics in mind, what happened in this study?
The first thing to notice is that, although Messrs. Brooks and Tierney concluded that exercise is futile, or even dangerous, the researchers did not control for physical fitness as a separate variable. In fact, fitness buffs may be confounding the data a bit.
There is a bit of a problem with the Body Mass Index, which was used as a proxy for obesity. The people who study these things are satisfied that most people with a high BMI are overweight. Most - but not all. Regular exercisers can add sufficient muscle mass to confound the classifications. For example, First Fitness Fanatic George Bush, with a BMI of roughly 26, would be classified as overweight despite a percentage of body fat that ranks as "lean".
This suggests two problems with the conclusion reached by Brooks and Tierney - first, "fitness" was not a separate control variable, so there is no basis to conclude that fitness is futile. Secondly, there may be some very fit but not-so-fat people slipping into the "overweight" category and boosting the mortality outlook for that group. We should note that "fitness" and "fatness" are independent concepts, although they tend to be correlated (and linked in the imagination of pundits at the Times).
The folks at the Cooper Institute in Dallas have been gathering data on the relationship between weight and cardiovascular fitness for decades. Their conclusion - a person can be overweight, yet still have good cardiovascular fitness. Such a person will have a much lower mortality risk than an equally overweight person with poor cardiovascular fitness. Conversely, a person can be sedentary and slim - such folks have a poorer mortality outlook than the fit fatties.
For the purposes of this study as reported in the popular press, BMI became a proxy for two things - excess fat (rather than muscle), and a sedentary lifestyle - but it is not a perfect proxy for either characteristic. After crunching the numbers, the researchers found that folks classified as "overweight" based on BMI had a better mortality outlook than people in the normal range.
However, all that means is that a group of people which includes the sedentary overweight, the fit fatties, and the muscled up fitness fools experienced lower mortality than a different group with lower BMIs. That is very different from saying, as Messrs. Tierney and Brooks did, that people who are truly overweight and do not exercise have a better health outlook than folks who watch their weight and hit the gym regularly.
So what is the truth? Is exercise a waste of time?
No. The researchers noted that mortality was reduced in part because medical techniques have improved - drugs control high blood pressure and high cholesterol, surgery repairs damaged hearts, and so on.
And the medical consensus is clear, as described by the National Institute of Health - carrying excess pounds is a bad idea, and regular exercise is a good one. Doctors with the President's Council on Physical Fitness also offer a very interesting perspective on the efficacy of dieting - don't worry about the pounds! You may not lose weight, but it's the effort that counts:
Physical activity is clearly viewed as being essential to the prevention of weight gain but fairly ineffective (at least in clinical trials) at promoting weight loss...
On the positive side, recent evidence suggests that physical activity confers health benefits that are largely or entirely independent of changes in body composition. These findings suggest that overweight and obese individuals can obtain the same benefits of physical activity as lean individuals.
...While the public health concerns about the increasing prevalence of obesity are well founded, they may be misplaced. The epidemiological studies reviewed here suggest that the health risks of obesity are largely controlled if a person is physically active and physically fit. The protection appears to come, at least in part, from positive metabolic changes that occur as a result of regular participation in physical activity.
Stop reading Brooks, ignore Tierney, put down that doughnut, and walk to the gym. And walk right past it, if you are not a member. But get moving.