Assessing body composition among 3- to 8-year-old children: anthropometry, BIA, and DXA

• Published in: Obesity (Silver Spring, Md.) Vol. 12; no. 10; pp. 1633 - 1640
• Main Authors: Eisenmann, J.C, Heelan, K.A, Welk, G.J
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01-10-2004
• Subjects: Absorptiometry, Photon - methods; Absorptiometry, Photon - standards; Adipose Tissue - anatomy & histology; Adipose Tissue - physiology; adiposity; adiposity rebound; Analysis of Variance; anthropometric measurements; Anthropometry - methods; bioelectrical impedance; Body Composition; body composition assessment; body fat; Body Mass Index; Child; Child, Preschool; children; dual-energy X-ray absorptiometry; Electric Impedance; fatness; Female; Humans; Male; Obesity; overweight; Predictive Value of Tests; Reproducibility of Results; Sensitivity and Specificity; skin folds
• ISSN: 1071-7323; 1930-7381; 1550-8528; 1930-739X
• DOI: 10.1038/oby.2004.203

Objective: To examine the inter‐relationships of body composition variables derived from simple anthropometry [BMI and skinfolds (SFs)], bioelectrical impedance analysis (BIA), and dual energy x‐ray (DXA) in young children.
Research Methods and Procedures: Seventy‐five children (41 girls, 34 boys) 3 to 8 years of age were assessed for body composition by the following methods: BMI, SF thickness, BIA, and DXA. DXA served as the criterion measure. Predicted percentage body fat (%BF), fat‐free mass (FFM; kilograms), and fat mass (FM; kilograms) were derived from SF equations [Slaughter (SL)1 and SL2, Deurenberg (D) and Dezenberg] and BIA. Indices of truncal fatness were also determined from anthropometry.
Results: Repeated measures ANOVA showed significant differences among the methods for %BF, FFM, and FM. All methods, except the D equation (p = 0.08), significantly underestimated measured %BF (p < 0.05). In general, correlations between the BMI and estimated %BF were moderate (r = 0.61 to 0.75). Estimated %BF from the SL2 also showed a high correlation with DXA %BF (r = 0.82). In contrast, estimated %BF derived from SFs showed a low correlation with estimated %BF derived from BIA (r = 0.38); likewise, the correlation between DXA %BF and BIA %BF was low (r = 0.30). Correlations among indicators of truncal fatness ranged from 0.43 to 0.98.
Discussion: The results suggest that BIA has limited utility in estimating body composition, whereas BMI and SFs seem to be more useful in estimating body composition during the adiposity rebound. However, all methods significantly underestimated body fatness as determined by DXA, and, overall, the various methods and prediction equations are not interchangeable.