OBJECTIVE: To investigate the relationship of percent body fat (%fat), assessed by dual energy-X-ray absorptiometry (DXA) or a four-compartment model, with upper body and lower limb skinfolds. DESIGN: Cross-sectional design involving forward stepwise and hierarchical multiple regression analyses to assess the relationship of %fat with skinfolds and a combination of four commonly used upper body skinfolds (biceps, triceps, subscapular and iliac crest) with the calf and thigh skinfolds. SETTING: University research laboratory. SUBJECTS: In all, 31 females, mean age 20.9 (+/-2.0) y, and 21 males, mean age 22.3 (+/-5.5) y volunteered for this study, which was approved by the Ethics Committee of the School of Sport, Health and Exercise Sciences, University of Wales, Bangor. MEASUREMENTS: %fat from DXA in both groups, and %fat from a four-compartment (water, bone mineral mass, fat and residual) model (%fat4C) in females only. Skinfolds were measured at the abdomen, iliac crest, biceps, triceps, subscapular, calf and thigh. RESULTS: All skinfolds were positively associated with DXA estimates of %fat (P < 0.01). In males and females, the thigh skinfold had the highest correlation with %fat. This was also observed when %fat4C was used as the criterion in females. Stepwise multiple regression analysis using %fatDXA as the criterion selected the thigh (R(2) = 0.82), calf (R(2) change 0.04) and iliac crest (R(2) change = 0.03) for females, and the thigh (R(2) = 0.79), iliac crest (R(2) change = 0.11) and abdomen (R(2) change = 0.03) for males (all P < 0.01). When %fat4C was used as the criterion in the females, only the thigh was selected as a significant predictor (R(2) = 0.76). Independent prediction factors were created from the sum of biceps, triceps, subscapular and iliac crest (sigma4skf) and from the sum of the thigh and calf (sigmathigh + calf). These factors were then entered into a hierarchical multiple linear regression analysis to predict percent fat. Order of entry was varied to allow the assessment of unique variance accounted for by each predictor. The sum of the thigh and calf explained more variance in %fatDXA than that explained by the sigma4skf alone, irrespective of the order of entry in both males and females. This was also observed when %fat4C was used as the criterion in the females. CONCLUSIONS: The results of this study confirm that lower body skinfolds are highly related to percent body fat in fit and healthy young men and women, and uphold current recommendations by the British Olympic Association to include the thigh skinfold with sigma4skf. Conventional use of the sigma4skf to estimate percent body fat is significantly enhanced by the inclusion of the thigh and calf skinfolds, either independently or in combination. In this group of males and females, the sum of the thigh and calf skinfolds accounted for the most variance in percent fat.
OBJECTIVE: To investigate the relationship of percent body fat (%fat), assessed by dual energy-X-ray absorptiometry (DXA) or a four-compartment model, with upper body and lower limb skinfolds. DESIGN: Cross-sectional design involving forward stepwise and hierarchical multiple regression analyses to assess the relationship of %fat with skinfolds and a combination of four commonly used upper body skinfolds (biceps, triceps, subscapular and iliac crest) with the calf and thigh skinfolds. SETTING: University research laboratory. SUBJECTS: In all, 31 females, mean age 20.9 (+/-2.0) y, and 21 males, mean age 22.3 (+/-5.5) y volunteered for this study, which was approved by the Ethics Committee of the School of Sport, Health and Exercise Sciences, University of Wales, Bangor. MEASUREMENTS: %fat from DXA in both groups, and %fat from a four-compartment (water, bone mineral mass, fat and residual) model (%fat4C) in females only. Skinfolds were measured at the abdomen, iliac crest, biceps, triceps, subscapular, calf and thigh. RESULTS: All skinfolds were positively associated with DXA estimates of %fat (P < 0.01). In males and females, the thigh skinfold had the highest correlation with %fat. This was also observed when %fat4C was used as the criterion in females. Stepwise multiple regression analysis using %fatDXA as the criterion selected the thigh (R(2) = 0.82), calf (R(2) change 0.04) and iliac crest (R(2) change = 0.03) for females, and the thigh (R(2) = 0.79), iliac crest (R(2) change = 0.11) and abdomen (R(2) change = 0.03) for males (all P < 0.01). When %fat4C was used as the criterion in the females, only the thigh was selected as a significant predictor (R(2) = 0.76). Independent prediction factors were created from the sum of biceps, triceps, subscapular and iliac crest (sigma4skf) and from the sum of the thigh and calf (sigmathigh + calf). These factors were then entered into a hierarchical multiple linear regression analysis to predict percent fat. Order of entry was varied to allow the assessment of unique variance accounted for by each predictor. The sum of the thigh and calf explained more variance in %fatDXA than that explained by the sigma4skf alone, irrespective of the order of entry in both males and females. This was also observed when %fat4C was used as the criterion in the females. CONCLUSIONS: The results of this study confirm that lower body skinfolds are highly related to percent body fat in fit and healthy young men and women, and uphold current recommendations by the British Olympic Association to include the thigh skinfold with sigma4skf. Conventional use of the sigma4skf to estimate percent body fat is significantly enhanced by the inclusion of the thigh and calf skinfolds, either independently or in combination. In this group of males and females, the sum of the thigh and calf skinfolds accounted for the most variance in percent fat.
Authors: Ursula Barandun; Beat Knechtle; Patrizia Knechtle; Andreas Klipstein; Christoph Alexander Rüst; Thomas Rosemann; Romuald Lepers Journal: Open Access J Sports Med Date: 2012-07-02
Authors: Annika Willems; Thomas A W Paulson; Mhairi Keil; Katherine Brooke-Wavell; Victoria L Goosey-Tolfrey Journal: Front Physiol Date: 2015-11-27 Impact factor: 4.566