Accuracy of six anthropometric skinfold formulas versus air displacement plethysmography for estimating percent body fat in female adolescents with phenylketonuria.

Background: The reliability of studies investigating biological and therapeutic factors that influence body composition in PKU patients depends on accurate anthropometric measurements.

Objective: To determine the precision of six anthropometric skinfold equations versus air displacement plethysmography (ADP) for predicting body fat (BF) percentage in female adolescents with PKU.

Design: Skinfold and ADP measurements were recorded from a cross section of 59 female patients with PKU, ages 10-19 years. Anthropometric measures were used to calculate percent BF using equations published by Peterson et al., Loftin et al. (TAAG), Slaughter et al., Wilmore and Behnke, Durnin and Womersley, and Jackson et al. Bland-Altman agreement analysis and Lin's concordance correlation coefficient (ρ c) were used to determine the precision of each equation compared with percent BF determined by ADP.

Results: Adolescent females with PKU had a mean BF content of 33% measured by ADP, with an inverse association to birth cohort (r = -0.3, P = 0.016). Based on the Bland-Altman method for evaluating agreement, only Peterson's equation did not differ significantly from ADP percent BF results (P = 0.23). Peterson's skinfold equation yielded percent BF estimates with the smallest mean difference from ADP and the smallest standard deviation (0.76 ± 4.8), whereas Slaughter's equation had the largest (-7.7 ± 7.4). Loftin's TAAG equation had the least mean percent error (2.2%), while Slaughter's equation had the highest (19%). Both TAAG and Peterson's equations had the highest concordance correlation coefficients (ρ c = 0.8, ρ c = 0.8), while Slaughter's equation had the lowest (ρ c = 0.3).

Conclusions: Peterson's equation is a precise surrogate for ADP when estimating percent BF in female adolescents with PKU, though Loftin's TAAG equation is also effective. Observed decreases in adiposity correlating with birth cohort could reflect steady improvements in patient nutrition care.