Concordance in prediction body fat percentage of Brazilian women in reproductive age between different methods of evaluation of skinfolds thickness

ABSTRACT Objective To assess the utility of bioimpedance (BIA) and skinfolds thickness (SF) in body fat percentage measuring (%BF) compared to the reference method dual-energy x-ray absorptiometry (DXA) in Brazilian reproductive age women, as well as to estimate of inter- and intra-observer precision for SF. Subjects and methods 170 women aged 18-37 years with BMI between 18 and 39.9 kg/m2 were selected for this cross-sectional study. Body density was evaluated through equations proposed by Jackson, Pollock and Ward (1980) (EqJPW) and Petroski (1995) (EqPET), and %BF was estimated by BIA, DXA and Siri's formula (1961). The SF were measured by two separate observers: A and B (to determine inter-observer variability), who measured the folds at three times with 10-minute interval between them (to determine intra-observer variability - we used only observer A). Results The %BF by DXA was higher than those measured by SF and BIA (p<0.01, for all) of 90 volunteers. The Lin coefficient of agreement was considered satisfactory for %BF values obtained by EqJPW and BIA (0.55) and moderate (0.76) for sum of SF (ΣSF) values obtained by EqJPW and EqPET. No agreement was observed for the values obtained by SF (EqJPW and EqPET), BIA and DXA. Analysis of inter- and intra-observer of 59 volunteers showed that different measures of SF thickness met acceptability standards, as well as the % BF. Conclusion BIA and SF measurements may underestimate %BF compared with DXA. In addition, BIA and SF measurements are not interchangeable with DXA. However, our results suggest the equation proposed by Jackson, Pollock and Ward (three skinfolds) compared to BIA are interchangeable to quantify the %BF in Brazilian women in reproductive age. Furthermore, our results show acceptable accuracy for intra- and inter-observer skinfold measurements. Arch Endocrinol Metab. 2020;64(3):257-68