RESUMO
OBJECTIVE: Since abdominal adiposity has been associated with increased risk for chronic diseases, valid and low cost methods to estimate it are needed for clinical and research purposes. The aim of the current study was to develop and validate, using dual-energy x-ray absorptiometry (DXA) as reference method, a model that estimates abdominal fat mass percentage (AFM%) in white postmenopausal women based on simple and easy-to-apply anthropometric measurements. An additional aim was to validate an abdominal bioelectrical impedance analyzer (ViScan) for estimating waist circumference (WC) and AFM% in this group. METHODS: Ninety-one postmenopausal women (mean age, 61.5 y) with body mass index ranging from 20.9 to 42 kg/m2 were randomly divided into one training set (n = 60) and one testing set (n = 31) to develop and validate a model based on anthropometric measurements estimating abdominal fat mass. Furthermore, in all 91 participants, ViScan estimations of WC and AFM% were validated against tape measurement and DXA results. RESULTS: The model developed was AFM% = 4.496 + (0.318 × WC) + 0.342 × suprailiac skinfold (r = 0.834, P < 0.0001). The model had no significant bias (0.25%) and ± 7.5% limits of agreement. ViScan significantly overestimated WC by 7.04 cm and estimated AFM% with no significant bias (-0.13%) and ± 7.6% limits of agreement. CONCLUSIONS: Both the model and ViScan are equally valid against DXA in estimating AFM%. However, ViScan is not valid in estimating WC in white postmenopausal women.