# So That’s Why I Scare Children

My pal Ken Arneson wrote in to tell me that not only does my math stink (yes, I screwed up a simple math problem in my last post. After getting my Brian Giles body, I’ll next attempt to get a Nate Silver mind for math … or not. The former is much more likely.) but also that my information may be bad.

After reading my last post, Ken says:

"So that’s why I scare children. I’m 5’10.5", 160.

Those BMI "ideal weight" stats you cite turned out to be wrong (as you suspected).  That’s because the CDC had done its BMI calculations wrong, and overcounted the obesity mortality risk by a factor of 14. So what was categorized as "overweight" actually turns out to be optimal.  "Normal" weight people like me are at slightly more risk for mortality than "overweight" people, and being "underweight" is more dangerous than being "obese".  So the preferable weight categories are, in order:  overweight, normal, obese, underweight.

In other words, going after the Brian Giles body is probably better
for you than going after the Ken Arneson body."

Good to know. I feel slightly less obese now that my mind is full(er).

## 1 Comment

Will,
Those BMI stats are not ‘wrong’, they just need to be understood in context. They’re based on a statistical correlation of BMI with body fat % for the entire world population. Body fat % correlates well with increased risk for coronary disease and diabetes. For the most part, the correlation is quite good. But the problem is that there are many racial phenotypes out there that are averaged into this, so a 5’10”, 190 lb American may not be overweight, but a 5’10” 190lb Vietnamese man with a smaller frame might be. If you consider the size of the smaller framed populations in Asia and India, it makes sense that the correlation is skewed somewhat for larger framed people. (I would bet that somewhere out there there is a BMI for Americans, which correlates BMI to body fat for the typical American person.) Also, because this is a statistical correlation over a huge population sample, and most people aren’t active, the usefulness of BMI goes all to **** when you’re an athlete. The goal of BMI was to create a simple, easy to calculate metric for people who are not savvy about their diet or exercise to know whether they might be at risk. As in, ‘oh wow, I never thought of myself as ‘fat’, but my BMI says I’m obese, maybe I should have a chat with my doctor’.

I’m really interested to see how your experiment turns out, as its my belief that normal people can look a lot like athletes without steroids if they exercise like its their job (as athletes do).

_Rich Hanna