Thoughts on Obesity (1): The First Step In Fighting Obesity Is…
…Actually measuring it accurately. This goes for research as well as in clinical practice.
This is a huge pet peeve of mine. Possibly because I was trained in a much more rigorous science before I started on my way in the medical profession (third year medical student). Also possibly because I’m kind of a healthy living nut. And equally as possibly because I’ve been ‘overweight’ my entire adult life despite wearing pants with a 32″ waist. Hard to say really.
The point is that obesity is a serious problem. Not just when it comes to major causes of mortality like heart disease, stroke, and even cancer and dementia, but also major morbidities–ones that often further predispose you to the above-mentioned mortality risks–like arthritis, spinal dysfunction, diabetes, peripheral vascular disease, and recurrent infection. There’s also little question that excess weight substantially reduces quality of life if nothing more than for the simple reason that it’s harder to move.
There is no doubt that fact, and the escalating proportion of the population that fits this definition is an even greater concern. But we actually don’t know and can’t know how serious a problem it is as long as we continue to use the BMI (actually an approximation thereof) as our method of categorization.
The Problem with BMI
BMI is a population-level measurement and has been proven time and time again to be of little to no validity when applied to the individual. Granted, if someone has a BMI of 17.5, they probably are substantially malnourished, but it still shouldn’t be the criterion we use to differentiate anorexia from bulimia, for instance. And if someone has a BMI of 40, chances are they are indeed carrying dangerous levels of body fat and have metabolic syndrome, if not overt diabetes and atherosclerosis. But in between the BMI is nothing if not lacking in precision.
The formula for the BMI is quite simple and is nothing more than our weight divided by our height squared. Yielding units of kg/m^2. And here we have the first problem. Any measurement system we use should be based on actual physical characteristics. The only physical characteristic that would naturally be expressed in terms of m^2 is our surface area. This would actually be a substantially more pertinent (although still limited) calculation. Body Surface Area is a pretty useful quality and finds application in everything from bioenergetics to exercise physiology to pharmacology. A measurement that used BSA and weight would give us an idea of body density, which is the basis for the more accurate body composition tests we have such as immersion. But since density at a given body fat percentage varies based on height (i.e. a 5′ tall person with 7% body fat will have a different density than a 6′ tall person with 7% body fat), a Bodymass Density Index (which Height/BSA would be), still wouldn’t be the most useful thing in the world.
The next problem with the body mass index is that it really doesn’t look at what kind of mass you’re carrying.
It’s a population level statistic and is more of an epiphenomenon of the fact that people are fatter today than they used to be than anything else. In other words, the only reason the BMI tends to work is because in society at large heavier people tend to be fatter. It’s an incidental finding. Amongst my gym buddies, Ken, a 230lb behemoth, sits at 6.5% body fat (based on statistically valid caliper testing). At 5′11″, he’s ‘obese’ according to BMI, yet just about anyone reading this should be envious of his low body fat content. At 5′11″ and 195, I’m only ‘overweight’, but I would bet any amount of money that I’m the one with a higher body fat percentage. And, even though my love handles have been my constant companion since I was about 14, I carry less body fat than most people who weigh 10lbs less than I do or more.
On the other end of the scale, researchers have had to coin the term ‘normal weight obesity‘ to identify those individuals who are not caught by the BMI screen and yet carry a substantial amount of body fat. Using a cutoff of 20% bodyfat for men and 30% for women, Mayo researchers found that over half of the sample of people with BMIs from 18.5-24.9 qualified as ‘normal weight’ obese. Metabolically, they appeared for all intents and purposes similar to those with a BMI over 30. The obesity problem thus isn’t restricted to those with a higher BMI, not by a longshot.
From a research standpoint, we can pretty much invalidate most of the ‘obesity’ research that has occurred thus far, at least as it relates to health risks and etiology. I haven’t been able to find a study looking at ‘normal body fat obesity’, but given that at 5′11″ a person can only weight 178 while still being considered ‘normal’ (which really isn’t very big), I suspect that a substantial proportion of people considered ‘overweight’ and even ‘obese’ by BMI are actually well within healthy limits for body fat. One bit of (slight) corroboration that I’m aware of is a study I found some years back indicating that increases in BMI among adolescent males were linked to decreasedbody fat and increased physical activity.
Current epidemiological and population-level research in obesity is thus lumping together disparate body types and compositions into its various categories. A full half of the ‘normal weight’ group carries an unhealthy amount of body fat, while a good-sized portion of the ‘overweight’ group is actually quite healthy. The major effect is to understate the health risks associated with unhealthy levels of body fat. Other potentially important factors with regard to chronic and major illness, morbidity, and mortality could easily be missed as well, such as whether llghter individuals with high body fat are at greater risk for health problems than heavier individuals with otherwise similar amounts of body fat (perhaps accounting for the occasionally mentioned ‘protective effects’ of body fat), or whether heavier individuals with low body fat are at increased or decreased risk for chronic medical conditions compared to lighter individuals of comparable body composition.
And, again (the third time I’m mentioning it I think), as BMI doesn’t hold for individuals anyway, we find ourselves hampered when it comes to translation of research into practice. It’s literally comparing apples to oranges. There is no way to put the information BMI-based research gives us into practice when it comes to screening and treatment.
The Answer?
Measure body fat. Recent attempts to popularize waist size instead of weight are a step in the right direction. But waist size is still an imperfect proxy. While unlike the BMI it is instantly translatable from research to practice, it still overlooks other possibly important body variables such as total lean body mass.
Well, immersion (hydrostatic) body composition tests are quite obviously impractical. The equipment, facilities, cost, staff training and expertise, and patient discomfort/inconvenience are prohibitive. But the 7-site skinfold method is accurate within 2% or so, training is easy, it’s fast, and it’s cheap. It might add a minute or two of time to patient/subject processing, but the dividends it pays will be tenfold.
Conclusion
Moving to skinfold tests to measure bodyfat in research and clinical practice is the wisest step we can take in understanding, preventing, and combatting the effects of obesity. It is the simplest solution to improved understanding of health and disease as well as more effective targetting of at-risk patients.
One benefit of body fat percentage testing which I’ll only touch on briefly is that it also provides us the ability to measure lean muscle mass. Muscle is an oft overlooked entity when it comes to medicine at large. We look at our patients’ bones, joints, visceral organs, endocrine systems, and brains, but nary a glance do we lend to one of the most amazing and adaptive tissues in the human body. Which is concerning, given that it makes up roughly half of our total mass.
Lean muscle mass is a pretty important variable unfortunately ignored for the most part in research and in practice. Exercise physiologists have known for decades the benefits of lean muscle mass and medicine has just started to take notice of its role in areas from prevention and alleviation of arthritis, to decreased spinal, neck, and back problems, to improved daily functioning in the elderly, to diabetes resistance, and lean body mass may even have a protective effect with regard to dementia–at the very least providing an improvement in symptoms and slow in the rate of cognitive decline.
Whatever the case, in order to properly understand, prevent, and treat obesity and obesity-related illness, we must be able to accurately assess its extent both within the population and in our individual patients. The body is too complex to be defined by a single number, be it BMI or simply weight. And in order to best understand the human body in sickness and in health we need to move away from such simplistic measurements.
Sun Tzu advises us that ‘If you know the enemy and know yourself, you need not fear the result of a hundred battles.” High BMI is not the enemy. And we are not our weight. The enemy is body fat, and body composition is ourselves.
Target weight or unhealthy BMI, either concept is wrong. Rather we should seek a target body and mind, and to such an end we should endeavor to use the most enlightening tools at our disposal, especially when the effort is so small and the reward so large.