Validity of the estimated body fat percentage by bioimpedance and skinfolds in middle-aged and elderly women

Objectives: Verify the validity of electrical bioimpedance (BIA) and the skinfold method in estimating the percentage of body fat (% BF) in middle-aged and elderly women, using dual emission X-ray absorptiometry (DXA) as a reference method. Material and methods: The sample consisted of 106 volunteers (middle age, n = 58 [51.3 ± 4.9 years] and elderly, n = 48 [67.2 ± 5.8 years]). The volunteers were submitted to anthropometric assessments and the % BF was subsequently calculated using the Jackson and Pollock 7 skinfold protocol (7SF). The % BF was also measured using BIA and DXA. Pearson’s correlation, Bland and Altman method, effect size and repeated-measure t-test were used to test the hypotheses. Results: Although there is a moderate positive relationship between the methods evaluated for middle-aged women (DXA vs 7SF, r = 0.67; DXA vs BIA, r = 0.62) and moderate to strong for the elderly (DXA vs 7SF, r = 0.57; DXA vs BIA, r = 0.75) (p ≤ 0.05 in all analyzes), the agreement between the methods was weak (limits of agreement > ± 3.5%). Additionally, when assessing the % BF mean using each method within the groups, it was observed that there was no significant difference between the estimates only between DXA and Pollock 7SF in the elderly group. Conclusion: Therefore, both the 7SF method and the BIA are not valid for estimating body fat in the sample evaluated. It is necessary to develop equations for specific skinfolds to estimate the % BF of middle-aged and elderly women and review the equations used by the BIA.