Background: Fat-free mass (FFM) is lower in obese subjects with Prader-Willi syndrome (PWS) than in obese subjects without PWS. FFM prediction equations developed in non-PWS subjects may, thus, not work in PWS subjects.
Aim: To test whether the estimation of FFM from bioelectrical impedance analysis (BIA) in PWS subjects requires population-specific equations.
Methods: Using dual-energy X-ray absorptiometry, this study measured FFM in 27 PWS and 56 non-PWS obese women and evaluated its association with the impedance index at 50 kHz (ZI50), i.e. the ratio between squared height and whole-body impedance at 50 kHz.
Results: At the same level of ZI50, PWS women had a lower FFM than non-PWS women. However, when PWS-specific equations were used, FFM was accurately estimated at the population level. An equation employing a dummy variable coding for PWS status was able to explain 85% of the variance of FFM with a root mean squared error of 3.3 kg in the pooled sample (n = 83).
Conclusion: Population-specific equations are needed to estimate FFM from BIA in obese PWS women.
Keywords: Bioelectrical impedance analysis; Prader-Willi syndrome; body composition; dual energy X-ray absorptiometry; prediction equations.