Friday, January 4, 2008

Skeptical remarks on the definition of overweight

Currently, a body mass index of more than 25 is considered overweight, but this should be checked against the newest mortality data from the Centers for Disease Control and Prevention.

In compliance with my good blogging intentions for 2008, I forget all my earlier posts about body weight, reset all my opinions to zero and have a new, fresh and unbiased look at the data. Not just some data, but data from the most recent "big" publication based on more than 2.3 million deaths, by Flegal and co-workers in JAMA 2007;298:2028.

There has been some debate in the skeptic community as to whether we always should take the side of mainstream consensus, about the difference between skepticism and crankery, and about the analysis of rhetoric tactics as a diagnostic tool. My point of view is that facts is what counts, so let's have a look at them.

The body mass indices in this study have been calculated from measured height and weight values which makes them more reliable than the self-declared values used in other studies.

graph

We see that obesity (BMI 30+) is strongly linked to excess deaths from coronary heart disease and from other cardiovascular diseases and is substantially linked to some, but not to all cancers. All other weight classes are not linked to mortality from these diseases. In particular, overweight (BMI 25-29) is not significantly different from normal weight.

graph

Looking at non-heart and non-cancer deaths, we see excess deaths from diabetes and kidney disease linked to overweight and obesity and a reduction of excess deaths from chronic respiratory disease, from injury and from other causes.

To sum up the excess deaths linked to overweight, we see a decrease of 25'000 deaths even when considering the data in favour of the mainstream definition: a maximum of +25'000 from diabetes related diseases, a minimum of -20'000 from chronic respiratory disease, a minimum of -10'000 from injury, and a minimum of -20'000 from other causes. In other words, a body mass index of 25 to 30 is linked to a lower mortality, mainly from causes other than heart disease and cancer. When those "big killers" are considered, overweight shows no link.

May smoking have distorted the data, causing more excess deaths in the normal weight category? Excluding smokers from the analysis did not significantly change the results, according to the authors. Moreover, it never could have explained links of such an extent. And without a pro mainstream bias, the decrease in mortality would even be as many as 95'000 excess deaths.

History of consensus

The old definition of overweight has been lowered from BMI 30+ to 25+, based on mortality data, a couple of years ago. This adjustment shows that the scientific community is ready to take new findings into account. Thus, with updated mortality data, a backward shift should also be possible.

graph

When the data of different study periods are compared, we see a decreasing link of extreme body fat (obesity) to excess deaths from 1975 (NHANES I) to 2000 (NHANES III). An explanation may be difficult, but facts are facts and should be considered.

A skeptic view

As far as I know, there is no broad scientific debate about the definition of overweight taking place in the public health community. On the contrary, we see public campaigns aimed at keeping people below BMI 25. Why? I see two main reasons.

The first one is mass inertia caused by the sheer number of experts dealing with body weight today.

The second one is primary prevention, the idea that "overweight" is the first stage of obesity (BMI 30+) which has been linked to an "unhealthy", sedentary lifestyle and various diseases, see the discussion here. I strongly agree with the idea that prevention is better than cure, and that a lifestyle that favours diseases and obesity (not necessarily diseases caused by obesity) should be changed to the better. But, in my opinion, the restrictive definition of "overweight" and the general weight obsession in modern western societies may be counterproductive. Physical activity, regardless of weight, may be a much better concept. The worst thing in extreme obesity, in my view, is being a handicap to physical activity (see also my post about obesity and fitness in the elderly). Instead, for many people, physical activity is only a means to the real end, weight loss.

In conclusion, I see an intriguing discrepancy between the most recent mortality data and the current definition of overweight, and I see no signs that this may change in the near future. I am not a scientist but only a skeptic watchdog, and all I can do is to bark, hoping to find more fellow skeptics who are ready to bark with me. There is a border between sound public health and unsound healthism, weight obsession and food nazism, a border that must be recognized and not be crossed.

No comments: