Test-retest reliability and validity of body composition methods in adults

Cost-effective and efficient body composition measurement devices that are reliable and valid are necessary for identifying health risk as well as for understanding the effectiveness of lifestyle interventions. The objective of this study was to evaluate the test-retest reliability and validity of three body composition measurement devices. Forty-nine adults (mean age (SD) = 31.5 (10.7) y; BMI = 23.5 (3.0) kg/m² ) completed a reference air displacement plethysmography (ADP) measure, and duplicate measures using skinfold callipers (Lange), ultrasound (BodyMetrix A-mode) and a 3-dimensional photonic scanner (3DPS; Fit3D ProScanner). Skinfold thickness was measured at seven sites using callipers and ultrasound; percent body fat (%BF) was then estimated using population-specific algorithms. The 3DPS was used to measure body circumferences, and then %BF was estimated using its beta-software. While skinfold callipers showed poor absolute reliability (mean differences (Δ) [95% CI] = 0.54% [0.22, 0.87], standard error of measurement (SEM) = 0.63%), ultrasound and the 3DPS showed excellent absolute (Δ = 0.17% [-0.25, 0.58], SEM = 0.78%; and Δ = -0.01% [-0.43, 0.40], SEM = 0.67%, respectively) and relative reliability (ICC_2,1  = 0.988 [0.979, 0.993]; and ICC_2,1  = 0.983 [0.968, 0.991], respectively). Compared to ADP (n = 43), skinfold callipers underestimated %BF (Δ = -4.53 [-7.72, -1.34]; p = 0.003), while ultrasound (Δ = -0.32 [-3.51, 2.87]; p = 0.99) and the 3DPS (Δ = 1.06 [-2.12. 4.26]; p = 0.77) were not significantly different. Bland-Altman plots showed a minimal bias of ultrasound [95% limit of agreement (LOA) = -7.87, 7.23] and the 3DPS [95% LOA = -6.66, 8.79]. In conclusion, estimating %BF from subcutaneous fat measurements using ultrasound and body circumferences using a 3DPS may be reliable and valid methods that require minimal technician expertise.