Body Composition: SkinFold Measurements
Return to: Tutorial Home Page  ENS 314 Exercise Physiology Lab Web Site
The development of skinfold (anthropometric)
measurements came as the result of investigations for simpler and less
expensive methods of estimating body composition. Body circumferences and/or
skinfold thickness are used in a regression equation, of which there are many
available, for prediction of body composition. Among the most commonly employed
are the generalized equations for use with skinfold measurements developed for
adult males and females by Jackson and Pollock (1978) and Jackson, Pollock, and
Ward (1980), respectively. They are termed generalized because they are most
accurate in predicting body composition of people with average amounts of body
fat. Generalized equations tend to be less accurate with the very lean (e.g.
athletes), obese, young old, or other special populations. Other, more
appropriate, equations have been developed for such populations which increase
their accuracy for prediction of body composition.
The major source of error in anthropometry
lies in the actual skinfold measurement. Making accurate skinfold measurements
is more than simply pinching the skin somewhere around a particular area and
measuring the thickness. There are precise sites on which the measurements are
to be taken. A welltrained technician can obtain results that approach the
precision of underwater weighing. Unfortunately, most people who take skinfold
measurements are not well trained. Obtaining consistently accurate skinfold measurements
requires training and experience.
To take a skinfold measurement, first
determine the correct measurement site. Grab the skin with the thumb and
forefinger about 0.5 inch from the measurement site following the natural fold
of the skin. Lift the skin up from the muscle, apply the calipers and wait for
4 seconds before reading the calipers. Fat is compressible, so reading the
scale before or after the 4sec delay may affect the results.
Triceps  Chest  Abdominal  Suprailiac  Thigh
Along the midline on the back of the triceps of the right arm, determine the midpoint located between the top of the acromial process (top of the shoulder) to the bottom of the olecranon process of the ulna (elbow). Pinch the skin so that the fold is running vertically.
Using a line from the fold of the axillary (armpit) to the nipple, determine the midpoint. Pinch the skin with the fold running in the same direction of the line.
Abdominal Skinfold
Select a site on the right side about 1 inch lateral from and 0.5 inch below the umbilicus (bellybutton). Lift a horizontal fold of skin for the measurement. NOTE: The Jackson and Pollock (1978) and Jackson, Pollock, and Ward (1980) abdominal skinfolds were made using a vertical fold. Consequently, one should stay consistent with how the original equations were validated when using these equations.
Determine the midaxillary line and palpate for the iliac crest (top of the hip bone). Grasp the skin that follows the natural fold which will follow a line of approximately from the suprailiac to the umbilicus (bellybutton), an angle of approximately 30 degrees.
Thigh Skinfold
Determine the midline of the front of the thigh and measure midway between the inguinal crease (the natural crease between the thigh and hip which is at an approximate 45 degrees) and the top of the patella (kneecap). Grasp a vertical fold of skin for the measurement.
Estimation of Body Density
There are two calculation steps for
determining percent fat; determining body density and then using body density
to estimate percent fat. All the skinfold equations estimate body density from
the measurements which is then used to estimate percent fat. Two generalized
equations that have withstood the test of numerous crossvalidation studies are
ones published by Jackson and Pollock (1978) for men and Jackson, Pollock, and
Ward (1980) for women. Each study developed numerous equations based on
different skinfold measurements, although they all were similar in prediction
error. Listed below are a single equation from each study based on the sum of
three skinfold measurements. Standard errors of their formulae ranged from 3.6%
to 3.8%. The potential errors are increased if these equations are used on the
young or old, very lean and muscular or obese. Other, population specific
equations are available and more suitable for these groups.
Females: 
Body Density = 1.0994921  0.0009929*sum + 0.0000023*sum^{2}  0.0001392*age 

(triceps, suprailiac, thigh skinfold measurements) 
Males: 
Body Density = 1.1093800  0.0008267*sum + 0.0000016*sum^{2}  0.0002574*age 

(chest, abdominal, thigh skinfold measurements) 
Estimation of Percent Fat
The two most commonly used equations for
estimating percent fat from body density are the Siri (1961) and Brozek (1963)
formulae. A limitation to these formulae is that they assume the density of
fatfree mass to remain a constant across the population when in fact is
varies. Thus, the actual percent fat tend to be slightly higher than the
measured percent fat in the lean, muscular individual and the opposite effect
in obese individuals.
Siri 
Percent Fat = [(495 / Body Density) 450] * 100 
Brozek 
Percent Fat = [(4.570 / Body Density)  4.142] * 100 