Thursday, September 19, 2013

1 out 2 of us are obese, fat or just breathe oxygen

From a commercial I see regularly on tv:  1 out of 2, hmm doesn't that mean 1/2 of us? Maybe I wasted all of those math classes.

This is taken from the BMI charts

According to my BMI, from a formula  devised in 1832. I am obese. However, so is probably the entire NFL? The formula involves dividing a persons weight by their height, squared. I am 5’7” or 67 inches tall. However in my youth – before I started to lose height due to aging – I was 5’9 or 69 inches tall. My weight is 113.7 Kg or 250.7 pounds. Therefore the according to the formula my BMI is 37.0.

The BMI ranges are:

·               18.5 or less      Underweight
·               18.8  - 24.9      Normal weight
·               25 – 29.9         Overweight
·               30 or greater    Obese

What are the problems with using BMI?
·               Actual measurements of body fat are ignored
·               Bone Structure is ignored
·               Gender is ignored
·               Fat to Muscle ratio is ignored
·               Race is ignored
·               Nutrition is ignored
·               Genetics is ignored
·               Age is ignored

The entire concept of analyzing dead people for their BMI has, to my knowledge, never done.  According to the website The-f-word

[2]The project was intended to describe the standard proportions of the human build.  The equation was largely ignored by the medical community even though insurance companies began using somewhat vague comparisons of height and weight among policyholders beginning in the early twentieth century.

Therefore the most commonly used measurement of human size to weight was instigated as a means of denying life insurance to large men.  I cannot find any reputable study measuring the correlation between height and weight.

What is an appropriate sample size? World population is estimated at seven billion (7,000,000,000) people. If we are looking for a 90% confidence level; assuming that the statistic will be statistically significant[3]; yields a sample size of 15,006.

For an overly obvious example; some assumptions have been made – Geography – do you take the sample in an American community? Do you take the sample in an Asian community? 

Do you see my feelings about BMI? It is impossible to accurately compute. It makes assumptions that are not valid. Most importantly, age, body type and family are ignored.  While BMI is an invalid statistic with no actual statistical validity, it is commonly used by the advertising community.

So, dear doctor, shall we discuss me as an individual and lose the shamanistic bone rattling?

[1] When a real world solution an educated guess can be used as a heuristic approach
[3] Using a p-value of .5

No comments:

Post a Comment