Copyright 2001-2002 John P. Hussman, Ph.D.
Direct links to this web page (http://www.hussman.org/fitness/caliper.htm) are encouraged. All rights to text and script are reserved and actively enforced.
A pop-up window should appear below. This page works only with Java-enabled browsers.
You really don't want to know what I think is "fun". Just for fun, I statistically reduced the standard Durnin & Womersly data into the best functional forms I could estimate. I got a 99% fit for both the men's and women's data, so my formulas will be pretty accurate within a fraction of 1%.
For readings from the suprailiac only, the best functional form turns out to be quadratic. Using M as your suprailiac reading in millimeters, my estimated formulas (extremely accurate for readings below 35 mm) are:
Women: Fat % = 1.223 M - 0.0134 Msquared + 0.124 Age + 6.07
Men: Fat % = 1.378 M - 0.0174 Msquared + 0.213 Age - 5.84
Above 35 mm, calculate the result for 35 mm and then add 0.25% for every 1mm above 35.
If you want to measure from more places, you can use the following calculations. They're not programmed in at the moment.
Typical sites are the back of the upper arm (over the tricep, taking the fold vertically as if you were pinching the tricep), the front of the upper arm (over the bicep, vertically ), the subscapular (the back slightly below the shoulder blade, taking the fold at a 45 degree angle, along the line from the opposite shoulder to the nearest elbow), and the suprailiac (the side of the waist, just above the point of the hipbone) In order to do the sites other than the suprailiac, you have to have someone else taking the measurements. You then take the sum of the measurements at all four locations, There are entire tables for the multiple measurements, which have hundreds of entries.
ln( ) refers to the natural logarithm, which is one of the functions on any good calculator. "Sum" is the sum of the 4 readings: tricep, bicep, back and waist (which will range anywhere from 18 to 200 mm). Here is the simplest functional form that gives accurate results.
Women: Fat % = 11.91 x ln(Sum) + 0.0442 x ln(Sum) x Age - 23.54
Men: Fat % = 10.32 x ln(Sum) + 0.0657 x ln(Sum) x Age - 27.03
I hate logarithms too. Look, I just run the numbers.