Evaluation of body composition using dual-energy X-ray absorptiometry, skinfold thickness and bioelectrical impedance analysis in Japanese female college students.

We compared three methods for evaluating body composition: dual-energy X-ray absorptiometry (DXA), skinfold thickness (Skinfolds), and bioelectrical impedance analysis (BIA). Subjects were 155 healthy young college-aged Japanese females whose mean+/-SD (range) age, body height, body weight and body mass index (BMI) were 20.1+/-0.3 (19.6-21.1) y, 158.9+/-4.7 (145.4-172.6) cm, 52.0+/-6.8 (39.4-84.6) kg and 20.6+/-2.3 (16.5-32.5), respectively. Their mean skinfold thickness at the triceps and subscapular were 16.9+/-4.7 (8.0-31.0) and 16.0+/-5.7 (7.0-40.0) mm, respectively. Mean body fat mass percentages evaluated by DXA, Skinfolds and BIA were 29.6+/-5.1, 22.8+/- 5.3 and 25.8+/-4.7%, respectively. Body fat mass was 15.4+/-4.4, 12.1+/-4.5 and 13.6+/-4.5 kg, respectively. Simple correlation coefficients between the three methods for body fat mass percentages provided the following coefficients: r=0.741 for DXA vs. Skinfolds, r=0.792 for DXA vs. BIA and r=0.781 for Skinfolds vs. BIA. Simple correlation coefficients for body fat mass were as follows: r=0.898 for DXA vs. Skinfolds, r=0.927 for DXA vs. BIA and r=0.910 for Skinfolds vs. BIA (all p<0.001). There were significant differences in the values among the three methods with the Skinfolds providing the lowest body fat mass and percentage, and DXA the highest (p<0.001). They all appear to be strongly correlated for evaluating body composition: however, different cut-off values for defining obese and lean need to be defined for each method.