Body composition in adults with cerebral palsy by dual-energy X-ray absorptiometry, bioelectrical impedance analysis, and skinfold anthropometry compared with the 18O isotope-dilution technique.

The aim of this study was to determine in adults with cerebral palsy the accuracy and practicality of standard methods used to estimate body composition. The sample consisted of 20 adults (13 men and 7 women) aged 20-55 y with various degrees of cerebral palsy. Percentage body fat was estimated from skinfold thickness, bioelectrical impedance analysis (BIA), and dual-energy X-ray absorptiometry (DXA), and compared with the reference measure of percentage body fat from total body water by 18O dilution. Values derived from use of BIA and skinfold thickness, estimated by using the Jackson-Pollock equation, were significantly different from those derived with use of 18O dilution (P < 0.001 and P < 0.001, respectively). There was no significant difference between percentage body fat measured with DXA and that measured with 18O. There was favorable agreement between DXA and 18O (mean difference: 0.06 +/- 9.6%), but not between skinfold thickness (mean difference: 6.33 +/- 12.3%) or BIA (mean difference: -6.55 +/- 13.6%) and 18O. Although DXA was the best measure for predicting percentage body fat in the sample, its high cost prohibits its use as a practical method. The best-fitting regression equation specific for this sample by using anthropometric measures to predict percentage body fat was as follows: 8.76 - (7.34 x sex) + (0.32 x weight) + (0.38 x biceps skinfold) (R2 = 0.84, P < 0.001, SEE = 4.85). This equation needs to be cross-validated in an independent sample of adults with cerebral palsy.