Use of anthropometric variables to predict relative body fat determined by a four-compartment body composition model

Abstract

Objective: To generate equations for the prediction of percent body fat (% BF) via a four-compartment criterion body composition model from anthropometric variables and age. Design: Multiple regression analyses were used to predict % BF from the best-weighted combinations of independent variables. Subjects: In all 79 healthy males ( X X7s.d.: 35.0712.2 y; 84.24712.53 kg; 179.876.8 cm) aged 19-59 y were recruited from advertisements placed in a university newsletter and on community centres' noticeboards. Interventions: The following measurements were conducted: % BF using a four-compartment (water, bone mineral mass, fat and residual) model and a restricted anthropometric profile (nine skinfolds, five girths and two bone breadths). Results: Stepwise multiple regression selected six (subscapular, biceps, abdominal, thigh, calf and mid-axilla) of the nine skinfold measurements to predict % BF and using the sum of these six produced a quadratic equation with a standard error of estimate (SEE) and R 2 of 2.5% BF and 0.89, respectively. The inclusion of age as a predictor further improved the equationHowever, the best equation used only the sum of three skinfold thicknesses (mid-axilla, calf and thigh) and age but also included waist girth and biepicondylar femur breadth as predictors (% BF ¼ À0.00258 Â ( P 3SF) 2 þ 0.558 Â P 3SF þ 0.118 Â age þ 0.282 Â waist girth -2.100 Â femur breadth -2.34; SEE ¼ 1.8% BF, R 2 ¼ 0.94). Analyses of two age groups, o30 and Z30 y, demonstrated that for the same % BF, the former exhibited a higher sum of skinfold thicknesses. Conclusions: Equations were generated for the prediction of % BF via the four-compartment criterion body composition model from anthropometric variables and age. Agewise differences for the sum of skinfold thicknesses may be related to an increase in internal fat for the older subjects. Sponsorship: Australian Research Council (small grants scheme).

Prediction of DXA-determined whole body fat from skinfolds: importance of including skinfolds from the thigh and calf in young, healthy men and women

“…Table 2. These results concur with data from previous studies which have shown strong associations between lower limb skinfolds or adipose tissue thicknesses and measurements of percentage body fat in healthy young adults (Eston et al, 1994(Eston et al, ,1995Stewart & Hannan, 2000;van der Ploeg et al, 2003) and healthy children and youth (Slaughter et al, 1988;Eston & Powell, 2003). The potential importance of measures of lower limb skinfolds as predictors of total subcutaneous adipose tissue mass was also identified in the studies of Martin et al (1985) and Clarys et al (1987), albeit the conclusions were drawn from data derived from 25 middleto old-age cadavers.…”

“…of 2.4% body fat were obtained for the relationship between the 4C model and DXA %fat in our study, although a dependent t-test revealed a significant difference of 1.4% between the two methods. These results are impressive, but not as good as those of van der Ploeg et al (2003), who reported r and SEE values of 0.952 and 1.6%, respectively. Although there are differences in the samples used in these two studies-the mean age and age range of the women's group was considerably higher, and the mean BMI was almost 3 units lower in the former study-the difference in the precision accuracy of the four-compartment model of assessing body composition in the current study should be viewed with some caution.…”

“…This has been observed using a variety of criterion methods, for example, hydrodensitometry (in adults- Jackson & Pollock, 1978;Jackson et al, 1980;Eston et al, 1995;in children-Slaughter et al, 1988;Eston & Powell, 2003); cadaver dissection (Martin et al, 1985;Clarys et al, 1987), ultrasound (Eston et al, 1994); dual energy X-ray absorptiometry (Stewart & Hannan, 2000) and a four compartment model of body composition (van der Ploeg et al, 2003). The correlation between subcutaneous abdominal adipose tissue volume and thigh fat volume (assessed by magnetic resonance imaging) is also reported to be greater than the corresponding correlation for the sum of the biceps, triceps, subscapular and iliac crest (Eliakim et al, 1997).…”

“…It allows for a rapid, noninvasive estimation of body fat with minimal radiation exposure (van der Ploeg et al, 2003) and has the advantage of a three-compartment model of body composition that quantifies fat, soft lean tissue and bone mineral. DXA has been validated against various multicompartment models in young (Prior et al, 1997;Clasey et al, 1999), old (Clasey et al, 1999) and a wide age range of healthy sedentary individuals (Gallagher et al, 2000).…”

“…DXA has been validated against various multicompartment models in young (Prior et al, 1997;Clasey et al, 1999), old (Clasey et al, 1999) and a wide age range of healthy sedentary individuals (Gallagher et al, 2000). Nevertheless, the validity of the method has remained subject to question, particularly with regard to concerns over tissue thickness and hydration levels (Laskey et al, 1992;Jebb et al, 1995;Pietrobelli et al, 1996Pietrobelli et al, , 1998Wang et al, 1998;van der Ploeg et al, 2003), which will vary between individuals and groups of subjects.…”