Comparison of body composition methods in overweight and obese Brazilian women

Abstract

Objective: The purpose of this study was to compare skinfold thickness (SKF) and bioelectrical impedance analysis (BIA) of body composition using three different equations against dual-energy X-ray absorptiometry (DXA) in overweight and obese Brazilian women. Subjects and methods: Thirty-four women (age 43.8 ± 10.9 years; body mass index [BMI] 32.1 ± 4.3 kg/m 2 ) had percentage body fat (BF%), fat mass (FM) and fat-free mass (FFM) estimated by DXA, SKF and BIA (BIA-man: manufacturer's equation; and predictive obesity-specific equations of Segal and of Gray). Regression analysis, Bland-Altman plot analysis and intra-class correlation coefficient (ICC) were used to compare methods. Results: Absolute agreement between DXA and BIA-man was poor for all measures of body composition (BF% -6.8% ± 3.7%, FM -3.1 ± 3.6 kg, FFM 5.7 ± 2.8 kg). BIA-Segal equation showed good absolute agreement with DXA for BF% (1.5% ± 1.5%), FM (1.0 ± 3.2 kg) and FFM (1.5 ± 2.6 kg), albeit the limits of agreement were wide. BIA-Gray equation showed good absolute agreement with DXA for FM (2.3 ± 4.1 kg), and smaller biases for BF% (0.05% ± 4.4%) and FFM (0.2 ± 2.9 kg), although wide limits of agreement. BIA-Gray and DXA showed the highest ICC among the pairs of methods. A good absolute agreement was observed between DXA and SKF for BF% (-2.3% ± 5.8%), FM (0.09 ± 4.7 kg), and FFM (2.4 ± 4.4 kg), although limits of agreement were wider and ICC between DXA and SKF for BF% indicated poor degree of reproducibility. Conclusion: These findings show that both BIA-Segal and BIA-Gray equations are suitable for BF%, FM and FFM estimations in overweight and obese women. Arq Bras Endocrinol Metab. 2010;54(4):398-405 Keywords Absorptiometry, photon; electric impedance; skinfold thickness; body composition; obesity; overweight ResumoObjetivo: Verificar a concordância dos métodos de impedância bioelétrica (BIA) usando três equações diferentes, e medida das pregas cutâneas (PC) com absorciometria de raios-X de dupla energia (DEXA), para análise da composição corporal de mulheres com sobrepeso e obesidade. Sujeitos e métodos: Em 34 mulheres (43,8 ± 10.9 anos; índice de massa corporal [IMC] 32,1 ± 4,3 kg/m 2 ) foram avaliados: percentual de gordura total (%GT), massa gorda (MG) e massa magra (MM) por DEXA, PC, BIA-Fab (equação do fabricante), BIA-Segal e BIA-Gray (equações para obesidade). Foram utilizados: análise de regressão, método de Bland-Altman e coeficiente de correlação intraclasse (CCI). Resultados: A concordância absoluta entre DEXA e BIA-Fab foi fraca para todas as medidas (BF% -6,8% ± 3,7%, FM -3,1 ± 3,6 kg, FFM 5,7 ± 2,8 kg). BIA-Segal apresentou boa concordância absoluta com DEXA para %GT (1,5% ± 1,5%), MG (1,0 ± 3,2 kg) e MM (1,5 ± 2,6 kg), mas com amplos limites de concordância. BIA-Gray teve boa concordância absoluta com DEXA para MG (2,3 ± 4,1 kg) e pequenos vieses para %GT (0,05% ± 4,4%) e MM (0,2 ± 2,9 kg), mas com amplos limites de concordância. BIA-Gray e DEXA tiveram maior CCI entre métodos. Houve boa concordância absoluta entr...

Modified body adiposity index for body fat estimation in severe obesity

“…Bioelectrical impedance analysis (BIA) significantly underestimates BF in morbid obese subjects (7,(10)(11)(12)(13)(14)(15) because the equations developed to calculate BF using BIA resistance are population specific (16) and are considered inappropriate for patients with severe obesity (7,10,13,15,17) . Nevertheless, specific equations for BF estimation in severe obesity were developed and validated by Horie-Waitzberg using BIA data, with a high specificity and sensitivity compared with plethysmography (18) .…”

Obesity in the Natural History of Type 1 Diabetes Mellitus: Causes and Consequences

“…The BIA is an easy, noninvasive, portable, no risk, relatively inexpensive method to measure the percentage of fat and provides results comparable to dual energy X-ray absorptiometry (DXA) (Elberg et al, 2004;Völgyi et al, 2008), that is reliable but expensive, requiring trained operators, individuals exposed to ionizing radiation and is not portable (Thomson et al, 2007). There are data indicating good correlation between BIA and DXA, including Brazilian (Braulio et al, 2010) and T1D subjects (Leiter et al, 1994). Although overestimating the percentage of fat in lean individuals and underestimate it in obese (Sun et al, 2005), proves useful for predicting metabolic risk (including IR) as well as BMI and waist circumference (Lee et al, 2008).…”

Ichthyometry and electrical bioimpedance analysis to estimate the body composition of tambatinga

“…Indirect methods for analyzing the body composition include densitometry, dilution, conductance and bioelectrical impedance, X-ray absorptiometry, computerized tomography, magnetic resonance imaging, and others (Braulio et al 2010;Ellis 2000;Pinto et al 2005). All these techniques are useful for estimating the body composition of animal species (Duncan et al 2007) and enable the study of the growth curve for lean tissue and the yields.…”