Validation of ultrasound techniques applied to body fat measurement
Résumé
AIMS: To evaluate the accuracy of body fat percentage (BF%) estimates from a portable, non-traumatizing ultrasound device with high accuracy and reliability compared to dual-energy X-ray absorptiometry (DEXA), the reference technique. DESIGN: Cross-validation between ultrasound technique (UT), DEXA, air displacement plethysmography (ADP) and bioelectrical impedance (BIA) was developed in the study. SUBJECTS: A total of 89 healthy subjects (41 women, 48 men), aged 48.4 +/- 17.7 (mean +/- SD), with Body mass index (28.5 +/- 7.7 kg/m(2)) and body fat DEXA (29.6 +/- 10.8 kg) participated. METHODS: BF% was measured using an UT associated with anthropometric parameters and simultaneously, with the DEXA reference technique, BIA and ADP. RESULTS: UT estimates of BF% were better correlated with those of DEXA in both males and females (r = 0.98, SEE = 2.0) than with ADP (r = 0.94, SEE = 3.7) or BIA (r = 0.92, SEE = 4.4). The UT in both genders was better (TE = 1.0) than BIA (TE = 2.6) and ADP (TE = 3.0). The 95% limits of agreement were also better for the UT (-2%; 2%) than with BIA (-5.1%; 4.9%) and ADP (-6.3%; 5.3%). CONCLUSIONS: The limits of agreement with BIA and ADP are unacceptably high compared to a DEXA measure criterion. The use of a new portable device based on a UT produced a very accurate BF% estimate in relation to the DEXA reference technique.