Three methods of body composition assessment were used to estimate percent body fat (%BF) in nonobese (n=77) and obese (n=71) women, 20-72 yrs of age, Skinfolds (SKF), bioelectrical impedance (BIA), and near-infrared interactance (NIR) methods were compared to criterion-derived %BF from hydrostatic weighing (%BFHW). Nonobese subjects had < 30% and obese subjects had >30% . The Jackson, Pollock, and Ward SKF equation and the manufacturer's equations for BIA (Valhalla) and NIR (Futrex-5000) were used. For nonobese women there were no significant differences between mean %BFHW and %BFSKF, %BFB1A, and %BFNIR. The rs and SEEs were 0.65 and 3.4% BF for SKF, 0.61 and 3.6% BF for BIA, and 0.58 and 3.7% BF for NIR for nonobese subjects. For obese women, mean %BPHW was significantly underestimated by the SKF, BIA, and NIR methods. The rs and SEEs for the obese group were 0.59 and 3.4% BF for SKF, 0.56 and 3.5% BF for BIA, and 0.36 and 3.9% BF for NIR. The total errors of the equations ranged from 5.6 to 8.0% BF in the obese group. It is concluded that all three field methods accurately estimate %BF for nonobese women; however, none of the methods is suitable for estimating %BF for obese women.

This study examined whether the predictive accuracy of agespecific bioelectrical impedance (BIA) equations was improved when estimated fat-free body (FFR,,,) was corrected for the influence of FFB size on whole body resistance (WBR) and residual errors of prediction for 152 women, ages 20-72 years. The criterion measure of FFB (FFBHU) was obtained from hydrostatic weighing at residual volume (RV). FFB,, was predicted from age-specific equations. Each subject's FFB,,, was then adjusted for the relationship between FFB and residual scores using the Lohman et al. (1990) Sample reference means were 45.0, 45.3, and 38.8 kg, respectively, for women 20-29 years, 3 0 4 9 years, and 50-70 years of age. The predictive accuracy of the unadjusted (FFBBTA) and adjusted (FFR,,) BIA estimates was analyzed for the total sample and each age group. For the total sample, r', standard error of estimate (SEE), and root mean square error (RMSE) were, respectively, 0.82,2.5 kg, and 2.5 kg for FFB,,.Corresponding values for FFB, , were, respectively, 0.81, 2.5 kg, and 2.6 kg. Across age groups, r2s ranged from 0.66-0.88, and the SEES and RMSEs were between 2.0 kg-2.8 kg. The relationship between FFRII, and residual scores (ry,reJ was significant (P < 0.05) for all age groups. The ry,rea for FFBRTA ranged from 0.37-0.41. For FFBAD,, the ry,y,res was higher (0.62-0.75). Thus, the overall predictive accuracy and systematic prediction error of the agespecific BIA equations were not improved by adjusting BIA estimates for the relationship between FFB size and residual scores. ~CI 1992 Wiley-Liss, Inc.Bioelectrical impedance analysis (BIA) is a rapid, non-invasive, and inexpensive field method for evaluating body composition. The BIA method is based on the relationship between whole body resistance (WBR) and the length (height) and volume (V) of the conductor, and assumes that the conductor (human body) is cylindrical (Lohman, 1989). The volume of total body water, or the fatfree body (FFB), is directly proportional to height squared (HT') and the specific resistivity (p) of the conductor, and inversely proportional to WBR, and is given by the following equation: V = (p)HT2/WBR (Nyboer et al., 1943). Segal et al. (1985) demonstrated that the FFB of men and women could be accurately estimated using the resistance index (HT2/WBR) in combination with body weight. Since 1985, many investigations have been undertaken to validate the BIA method for heterogeneous populations varying in age and levels of adiposity.This research has led t.0 the development of age-, gender-, and fat-specific BIA prediction equations (Boileau et al., 1989;Deurenberg et al., 1989Deurenberg et al., , 1990 Segal et al., 1985Segal et al., , 1988. Typically, hydrodensitometry (HD) has been used to derive criterion measures of FFB to develop and cross-validate these equations. The predictive accuracy of the equations is variable with standard errors of estimate (SEE) ranging from 1.97 kg (Gray

The purpose of this study was to develop a multi-site near-infrared (NIR) model (Model I) and compare its predictive accuracy to single-site models (IIA and IIB). In Model I, the sum of two optical density (OD) measures (Σ2OD), age, body weight, height, and physical activity level were used as potential predictors of body density (D ). In Model IIA, the variables used in the manufacturer's NIR equation (biceps OD and OD , body weight, height, gender, and physical activity level) were the potential predictors. This model was modified by including age as an additional potential predictor in Model IIB. We also examined the test-retest reliability and interrelationships of OD measures taken at 10 anatomical sites, as well as the validity of the manufacturer's NIR equation, for estimating body composition of women. The subjects, 148 women between 20 and 72 years, were hydrostatically weighed to determine criterion D . The Futrex-5000 was used to measure OD and OD at 10 anatomical sites. Only two sites (pectoral OD and biceps OD ) contributed significantly to the variance in D . Thus, the sum of these two ODs (Σ2OD), was used as a potential predictor in the multi-site model. Test-retest reliability was high, with intraclass correlation coefficients ≥0.85 for many of the OD measurements. Intercorrelations of ODs ranged from 0.22 to 0.91. In the multi-site model (I), ΣOD, body weight, age, and height were significant predictors, accounting for 85.7% of the variance in D . The SEE was 0.0076 g/ml or 3.3% BF. In the manufacturer's model (IIA), biceps OD , body weight, and height accounted for 76.3% of the variance in D , and the SEE was 0.0094 g/ml (4.1% BF). When age was included as a predictor (Model IIB), the R increased (86.0%) and the SEE (0.0073 g/ml or 3.1% BF) decreased substantially. Cross-validation of the three equations yielded r s ranging between 0.688 (Model IIA) and 0.748 (Model I) and slightly larger SEEs (0.0094-0.001048 g/ml). There were no significant differences between average criterion D and predicted D for each equation. The manufacturer's equation programmed in the Futrex-5000 yielded a lower r (0.55), higher SEE (5.61% BF), and significantly underestimated criterion % BF by an average of 3% BF. Either the multi-site (model I) or single-site (Model IIB) equations is recommended to estimate body composition of this population. © 1992 Wiley-Liss, Inc.

We examined relationships between skinfold (SKF) and optical density (delta OD) measurements across age and levels of body fatness (%BF) for 151 women, 20 to 72 years. There were significant (p < .05) relationships between delta ODs and SKFs at all sites, except the thigh. The interaction (SKF x Age) was significant (p < .05) for pectoral and biceps delta ODs. Slope comparisons indicated the relationships for younger (29 years) and older (59 years) women differed significantly from zero and each other (p < .05). Analysis of SKF x %BF interactions revealed that relationships between SKFs and delta ODs at the pectoral and biceps sites for leaner (22% BF) women differed significantly from zero (p < .05) and were larger than those for obese (39% BF) women (p < or = .05). Thus, the relationship between SKFs and delta ODs is stronger for younger and leaner women compared to older and fatter women. These findings may reflect differences in fat layering due to age or body fatness and provide insight as to why the manufacturer's near-infrared (NIR) equation significantly underestimates the %BF of obese women.

This study assessed the predictive accuracy of age-and fatnessspecific BIA equations in estimating the fat-free mass (FFM) of a heterogeneous sample (N = 152) of women, 20-72 years (yr), with 11-57% body fat (BF). The criterion method was hydrostatic weighing (HW) at residual volume. The Siri (Siri [196ll Natl. Acad. Sci., pp. 7f3-89) two-component model was used to convert body density into relative body fat (% BF) for calculation of criterion FFWHw Average FFM,, and predicted FFM,, did not differ significantly (P > .05) for the Lohman (Lohman [1981] Hum. and 330% BF) equations. The SEE for these equations ranged from 2.11 to 2.65 kg. All other equations (Lohman 30-49 yr and 5C70 yr; Durenberg (Durenberg et al. [1990] Am. J. Clin. Nutr. 51:3-6) 20-40 yr and 60-83 yr) significantly underestimated ( P < .05) FFM, , by as much as 5 kg, with the SEES ranging from 2.12 to 2.82 kg.. The prediction error of equations developed specifically for young (Lohman, 20-29 yr) and non-obese (Segal, <30% BF) women was less than that for older (Van Loan and Mayclin, 18-64 yr) and obese (Segal, 230% BF; Gray, 19-59% BF) women. In conclusion, Lohman's equation for older (3049 yr) women or Durenberg's equations for younger (20-40 yr) and older (60-83 yr) women are not recommended.