1. ¹Semper AB, Stockholm, Sweden
2. ²Department of Biomedicine and Surgery, University of Linköping, Linköping, Sweden

Correspondence: C Tennefors, Semper AB, Torsgatan 14, SE-105 46 Stockholm, Sweden. E-mail: Catharina.tennefors@arlafoods.com

Guarantor: C Tennefors

Contributors CT was responsible for data collection during a study of energy metabolism and dietary intake in children. CT also analysed the data and prepared the manuscript together with EF.

Received 9 December 2002; Revised 9 May 2003; Accepted 22 May 2003.

Abstract

Objective: To compare body fatness estimated using a skinfold technique and body mass index (BMI) with body fatness estimated using the body water dilution method in healthy Swedish children 9 or 14 months of age.

Methods: Total body fat (TBF) was calculated from total body water, estimated using the doubly labelled water method, and body weight. When expressed in per cent of body weight, these estimates (%TBF-BWD) represented reference values for body fatness. Body fatness was also calculated from skinfold thickness (%TBF-SFT) and as BMI. The children were ranked and grouped into five groups with an increasing level of body fatness using BMI, %TBF-SFT and %TBF-BWD, respectively.

Subjects: A total of 30 infants 9 months of age and 29 children 14 months of age.

Results: On average, the children (n=59) had a BMI=17.51.6 kg/m² and contained 27.83.7 %TBF-SFT and 29.14.4 %TBF-BWD. %TBF-BWD minus %TBF-SFT was=1.354.06%. By measuring %TBF-SFT or BMI, about 35% of the children could be classified in the correct group with respect to body fatness. Serious misclassification (ie two or more groups too high or too low) was, however, more common for %TBF-SFT (29%) than for BMI (17%).

Conclusions: The capacity of BMI to place children in the correct body fatness group was poor although not quite as poor as the corresponding capacity of the skinfold technique. The latter method produced inaccurate and imprecise estimates of body fatness.