Last week the New York Times and many other reputable news sources reported on a Canadian study that claims people with a normal body mass index (BMI) had a slightly increased risk of dying over a 12-year period than those with a BMI in the overweight range (25-29).
The use of the phrase “overweight was protective” landed this article just a hair’s width from being labeled Summer Tomato’s B.S. of the week on Friday. An observational study cannot determine cause and effect, as implied by the word “protective.” This study does not prove that extra body weight protects against all cause mortality, and saying so is irresponsible.
Studies (and reporting) like this have instigated wide-spread confusion about health and body weight. First people are told they are too heavy and should lose weight for health, then in the same breath they are told a little extra weight might not be so bad.
What is the average person supposed to believe? How should we act?
If you want to understand the facts it important to know exactly what the data does and does not say. Indeed, some studies (including one on Japanese men reported in the same issue of Obesity) have reported lower or equal risk of mortality for people with an overweight BMI compared to people of normal BMI (18.5-24). However, this is not the whole story.
First, the alleged benefit of being overweight has only been found in older individuals and does not apply to healthy, young people. Second, although it appears in some cases that overall mortality may be reduced, disease incidence is notably higher in overweight individuals compared to people of normal weight.
To point, a study in the most recent issue of Obesity (same journal, 2 weeks later) examines the relationship of BMI with many measures of cardiovascular disease in healthy, athletic men. In this study, those on the lowest end of BMI in the normal range (18.5-22.5) had a much lower risk of dying from or developing cardiovascular disease than normal weight men with a slightly higher BMI (22.5-25).
Men with the lowest healthy BMIs also had lower risk of hypertension, lower cholesterol and half the risk of diabetes. While the length of this study was only 7.7 years (compared to the 12 years reported in the Times story), there were more than double the number of participants (28,776 vs. 11,834).
(Why did this story not make the news? My guess is that it makes for a less compelling storyline and people would rather not hear it.)
Mortality is certainly an important measure in any study, but it is arguably not the most relevant endpoint. Disease and excess body weight can severely impact quality of life, particularly for older individuals (as illustrated by another study in the latest issue of Obesity). While I cannot speak for everyone, it seems probable that quality of life is equally if not more important than longevity alone. Thus it is questionable how much stock to put into studies that ignore these other factors.
It is also critical to remember that BMI is a measure that was designed to describe people at a population level, not as individuals. While large cohort studies can tell us useful things about relative risk, they are not directly applicable to individual people.
The inconsistency of the data related to BMI and mortality may in fact be an indication of its inadequacy as a general measurement. Remember that BMI represents a ratio between height and weight, making it possible to compare people of various body sizes. Normalizing for height may, however, be deceptive.
Decades of data on caloric restriction consistently show that smaller body size (irrespective of body fat levels and, possibly, BMI) is associated with longer life and decreased risk of nearly all diseases. This is true in all animals from yeast, to worms and flies, to mice and monkeys. While humans are certainly different from all these model organisms, there is tremendous evolutionary precedent indicating smaller body size as the best for health.
The principle of parsimony tells us the simplest hypothesis–that smaller body size is beneficial–is probably correct. Substantial evidence must be accumulated before this hypothesis can be rejected, and I have yet to see that data.
Furthermore, while the research on the risk of overweight may be slightly ambiguous, the data on obesity is not. It is painfully clear that the dangers of obesity are profound and on par with those of smoking cigarettes. Overweight is a necessary step to becoming obese, and according to the National Population Health Survey nearly a quarter of Canadians who were overweight in 1994/1995 were obese by 2002/2003. Since overweight is still a substantial risk factor for becoming obese, misleading public health messages about the benefits of body fat are especially dangerous.
As a consumer of information, the most important thing you can do is be skeptical of what you read. Just because something is printed in the New York Times does not make it true. In fact, many of our most trusted sources of health information do not base their recommendations on rigorous scientific thinking, which is probably the reason for the health disaster we are currently facing.
Thanks to Jan from Quest for Health for sparking this discussion.
What does your gut tell you about the relationship between health and body fat?