Clinical Medicine & Physiology
Durnin–Womersley Skinfold Equations: A Body Density Example
Explore how the 1974 Durnin–Womersley four-site equations estimate body density, followed by a separate Siri body-fat calculation, using an adult spreadsheet example.
What the original equations estimate
Durnin and Womersley measured four skinfolds and body density by underwater weighing in 209 men aged 17–72 and 272 women aged 16–68. The sample was not random; the authors noted a predominance of moderately sedentary, middle-class participants. Measurements use biceps, triceps, subscapular and suprailiac skinfolds in millimetres.
An equation of the form density = c − m × LOG10(sum of four skinfolds) estimates body density; coefficients differ by the sex and age groups used in the original study. Siri’s separate two-compartment equation estimates body-fat percentage from density. Neither number is a direct measurement of body fat, and results vary with measurement technique and how well the original groups and assumptions fit the individual.
Four illustrative original coefficient pairs
This exercise implements only the original male and female groups aged 20–29 and 30–39 years, with age entered in completed years. For a four-site sum of 45 mm, the unrounded density is used in Siri’s equation. Extra displayed decimal places make the spreadsheet arithmetic reproducible and do not imply individual-level precision.
| Study group | Intercept c; slope m | Estimated density (g/cm³) | Siri estimate (%) |
|---|---|---|---|
| Male, 20–29 | 1.1631; 0.0632 | 1.05862 | 17.59 |
| Male, 30–39 | 1.1422; 0.0544 | 1.05227 | 20.41 |
| Female, 20–29 | 1.1599; 0.0717 | 1.04136 | 25.34 |
| Female, 30–39 | 1.1423; 0.0632 | 1.03782 | 26.96 |
Estimate body density in a spreadsheet
Enter all four positive skinfold thicknesses in B2:B5. Enter 1 for the original study’s male equation or 0 for its female equation in B6, and age in completed years from 20 to 39 in B7. B8 retains the unrounded density so it can be used by the next formula. Values outside these four reference groups are not implemented. Numeric checks do not establish whether measurements are reliable or within a validated clinical range.
=IF(COUNT(B2:B7)<>6,"Enter all numeric inputs",IF(OR(MIN(B2:B5)<=0,AND(B6<>0,B6<>1),B7<>INT(B7),B7<20,B7>39),"Check input values",IF(B6=1,IF(B7<30,1.1631,1.1422),IF(B7<30,1.1599,1.1423))-IF(B6=1,IF(B7<30,0.0632,0.0544),IF(B7<30,0.0717,0.0632))*LOG10(SUM(B2:B5))))Apply the separate Siri conversion
Enter this formula after B8. The result is an approximate two-compartment estimate; the conversion assumes a reference fat-free mass density, and its assumptions may not hold for every individual. Unexpected results require checking measurements and model suitability rather than drawing a health conclusion.
=IF(ISNUMBER(B8),IF(B8>0,ROUND(495/B8-450,2),"Check calculated density"),"Check inputs")Try the example
Use the original values or change them to test your understanding.