Abstract
Objective
To determine accuracy of the Tanita SC-240 body composition analyzer to measure pediatric percent body fat (%BF).
Methods
Eighty-nine African American and white 5–18 year-olds participated in this study. %BF was estimated by DXA and by the Tanita SC-240.
Results
Overall %BF was 33.5 ± 10.5% (Tanita SC-240) versus 34.5 ± 8.7% (DXA). There was no significant difference between the two measures (p = 0.52, average error = −1.0%, average absolute error = 3.9%). The Tanita mean %BF estimates significantly differed from the DXA mean %BF in white boys (p = 0.001, Cohen’s d = 0.40) and white girls (p = 0.006, Cohen’s d = 0.48), but differences were of small effect. No differences in %BF estimates were found for African American boys or girls.
Conclusions
In this sample the Tanita SC-240 demonstrated acceptable accuracy for estimating %BF when compared to DXA, supporting its use in field studies.
Keywords: bioelectrical impedance, body fat percentage, dual energy x-ray absorptiometry, children and adolescents, accuracy
Researchers are faced with a dilemma between cost, practicality and accuracy in body composition assessment (1–3). The current standards for measurement are dual-energy x-ray absorptiometry (DXA), magnetic resonance imaging (MRI), or underwater weighing. These methods are accurate but costly, time consuming, and of limited availability for field studies. Bioelectrical impedance analysis (BIA) scales are practical, safe and inexpensive ways to analyze body composition in the field (4). The Tanita SC-240 is a portable professional grade BIA scale that could be practically employed in field situations, but to our knowledge its accuracy has not been independently tested. The purpose of this study was to determine the accuracy of the Tanita SC-240 to measure body fat percentage (%BF) in a biracial sample of boys and girls.
Participants were a sub-sample of 5–18 year-old children involved in a cross-sectional study of factors related to abdominal adiposity. There were 89 participants in the present analysis (mean age = 12.0 ± 3.6 y). Parents or guardians provided signed informed consent, and the children provided written assent. All procedures were approved by the institutional review board.
The participants came to the clinic in the morning following a 10 hour fast. There were no instructions provided to the participants regarding exercise the day before the test. Participants were measured wearing light clothing and barefoot. Height and weight were measured twice to the nearest 0.1 cm and 0.1 kg, respectively, and the average was used in the analysis. Body mass index (BMI) was computed as weight (kg)/height(m)2, and BMI percentiles were calculated (5, 6). Sexual maturation was physician-assessed for participants <9 years and self-assessed for participants ≥9 years using the Tanner criteria (7).
%BF was estimated by DXA using a Hologic QDR 4500A whole-body scanner (Bedford, MA) and by the Tanita SC-240 body composition analyzer (TANITA Corporation, Japan), which is a lightweight portable scale that uses tetra-polar BIA with four electrodes at the feet. The scale has a measurement frequency of 50 kHz, a measurement current of 90 μA, and a measurement range of 150–1200 Ω. It has a maximum capacity of 200 kg and a graduation of 0.1 kg. Participants’ information (i.e., height, age, and sex) was entered, and they were asked to step onto the scale and remain still.
The %BF assessed by the Tanita SC-240 was compared to the DXA %BF using a Bland-Altman plot (8). In addition to the calculation of error (Tanita %BF – DXA %BF) and absolute error (|Tanita %BF – DXA %BF|), paired samples t-tests were used to determine if there were differences between %BF estimations by the Tanita versus DXA for the whole sample and for each sex-by-race group. Cohen’s d effect sizes were calculated to determine the meaningfulness of the differences using guidelines of d = 0.2 (small effect), d = 0.5 (medium effect), and d > 0.8 (large effect) (9). All data analyses were completed on SPSS 19. A p-value <0.05 was considered significant.
The participants’ characteristics, Tanita and DXA %BF assessments, and error calculations by sex and race, are presented in Table 1. Weight was similar between the Tanita and DXA with an average error of 0.2 ± 0.2 kg, where the largest difference was 0.9 kg and only seven participants had a difference >0.5 kg. Based on sexual maturation assessment, 28% were prepubertal (stage 1), 17% were early pubertal (stages 2–3), and 55% were late pubertal or mature (stages 4–5). In the overall sample, there was no significant difference in %BF between the Tanita SC-240 (mean = 33.5 ± 10.5%) versus the DXA (mean = 34.5 ± 8.7%), p = 0.52, d = 0.11. The average error was −1.0%, and the average absolute error was 3.9%. Four participants had errors beyond the 95% limits of agreement (Figure 1). The Tanita %BF estimates were significantly different from the DXA %BF in the white boys (p = 0.001, d = 0.40) and white girls (p = 0.006, d = 0.48). No differences in %BF estimates were found for African American boys and African American girls.
Table 1.
Descriptive characteristics of the sample and comparison of body fat percentages estimated by bioimpedance (Tanita SC-240) versus DXA (Hologic QDR 4500A).
| White Boys (n=15) | White Girls (n=21) | African American Boys (n=20) | African Amerian Girls (n=33) | |||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Mean (s.d) | Min-Max | Mean (s.d.) | Min-Max | Mean (s.d.) | Min-Max | Mean (s.d.) | Min-Max | |
| Age, years | 10.9 (3.7) | 5–17 | 11.8 (3.9) | 5–17 | 11.9 (3.2) | 5–18 | 12.6 (3.6) | 5–18 |
| Height, cm | 148.0 (22.1) | 108.0–180.4 | 151.3 (19.8) | 112.1–178.5 | 151.3 (16.0) | 115.9–173.0 | 154.0 (13.4) | 119.5–179.1 |
| Tanita weight, kg | 61.9 (34.6) | 17.6–119.8 | 58.7 (17.6) | 23.0–95.4 | 62.9 (29.2) | 19.5–109.3 | 70.1 (23.5) | 28.2–124.7 |
| DXA weight, kg | 62.1 (34.6) | 17.7–119.5 | 58.9 (17.6) | 23.0–95.5 | 63.2 (29.2) | 19.6–109.6 | 70.3 (23.6) | 28.3–124.9 |
| BMI, kg/m2 | 25.8 (8.5) | 14.4–38.07 | 26.0 (8.1) | 15.6–40.21 | 25.1 (4.3) | 17.12–33.7 | 28.9 (6.9) | 16.8–45.5 |
| BMI, percentile | 82.7 (25.8) | 18.5–99.6 | 84.2 (17.6) | 45.4–99.6 | 90.8 (7.3) | 75.9–99.5 | 90.8 (17.4) | 21.1–99.9 |
| Tanita, %BF | 28.2* (9.3) | 17.1–50.3 | 34.4* (5.5) | 26.6–47.5 | 28.7 (14.6) | 10.0–67.2 | 38.1 (8.4) | 18.9–53.2 |
| DXA, %BF | 31.8 (8.5) | 17.5–46.2 | 37.0 (5.3) | 26.3–46.3 | 28.7 (10.5) | 13.8–51.4 | 37.6 (7.4) | 22.6–50.9 |
| Error | −3.6 (3.5) | −9.6–4.1 | −2.6 (3.9) | −8.7–6.7 | 0.1 (5.8) | −11.7–15.8 | 0.5 (4.7) | −10.6–8.7 |
| Absolute error | 4.3 (2.6) | 0.2–9.6 | 3.7 (2.7) | 0.3–8.7 | 4.1 (4.0) | 0.3–15.8 | 3.8 (2.7) | 0.6–10.6 |
= significantly lower than DXA %BF.
Note. %BF indicates body fat percentage; BMI indicates body mass index; DXA indicates dual energy x-ray absorptiometry; s.d. indicates standard deviation, Error = Tanita %BF minus DXA %BF; Absolute error = absolute value of Tanita %BF minus DXA %BF.
Figure 1.
Bland-Altman plot for error of bioelectrical impedance analysis (Tanita SC-240) versus DXA (Hologic QDR 4500A) to estimate body fat percentage in white and African American children and adolescents.
Note. WB = white boys, WG = white girls, AAB = African American boys, AAG = African American girls.
Overall the Tanita SC-240 body composition analyzer demonstrated acceptable accuracy for estimating %BF when compared to estimates obtained from DXA in this biracial sample of boys and girls. The overall mean absolute error of 3.9% was close to the expected error for skinfold measurements (±3.3%), which is another method commonly used in field studies (4). This result is in contrast with other studies that found significant difference in %BF estimates between BIA devices and DXA in youth (10–13).
Considering the Bland-Altman plot and the error and absolute error calculations, for African American boys and girls the Tanita SC-240 did not present a specific pattern of error and did not consistently under- or over-estimate %BF compared to DXA. Those results differ from previous comparisons between BIA and DXA in African American girls where significant bias was found (10, 11). In contrast, for white boys and white girls the Tanita SC-240 consistently under-estimated %BF compared to DXA, though the bias was of small effect (d < 0.5) and lower than the BIA bias found by other researchers which ranged from 4.5 to 18.8 %BF (12, 13).
This study was carefully conducted but is not free of limitations. The equations used to estimate %BF are proprietary data. Therefore it is impossible to determine if the difference in measurement error between whites and African Americans is due to limitations with the Tanita SC-240 measurement of impedance or if it was because the same equation was used for both races. Because the errors found in the measurements were of small magnitude (d < 0.5), these limitations may be inconsequential. Because validation is a continuous process, additional studies should be conducted with the Tanita SC-240.
The Tanita SC-240 body composition analyzer demonstrated acceptable accuracy for estimating %BF in this study. There were significant differences between the DXA and the Tanita SC-240 scale for white boys and girls, but those differences were of small effect (d < 0.5). There were no significant differences between %BF measures for African American boys or girls.
Acknowledgments
Source of Support: NIH-NIDDK-1RC1DK086881-01. This project was partially supported by a Nutrition Obesity Research Center (NIH 2P30DK072476) center grant from the National Institutes of Health.
TVB was involved in the study design, data analysis, literature search and generation of figures. AES was involved in the data analysis and literature search. PTK was involved in the study design. All authors were involved in data interpretation, writing the paper, and had final approval of the submitted and published versions.
Footnotes
Conflict of Interest: The authors have declared that no competing interests exist
References
- 1.Hemmingsson E, Udden J, Neovius M. No apparent progress in bioelectrical impedance accuracy: Validation against metabolic risk and DXA. Obesity. 2009;17:183–187. doi: 10.1038/oby.2008.474. [DOI] [PubMed] [Google Scholar]
- 2.Volgyi E, Tylavsky FA, Lyytikainen A, Suominen H, Alen M, Cheng S. Assessing body composition with DXA and bioimpedance: effects of obesity, physical activity, and age. Obesity. 2008;16:700–705. doi: 10.1038/oby.2007.94. [DOI] [PubMed] [Google Scholar]
- 3.Sluyter JD, Schaaf D, Scragg RK, Plank LD. Prediction of fatness by standing 8-electrode bioimpedance: A multiethnic adolescent population. Obesity. 2010;18:183–189. doi: 10.1038/oby.2009.166. [DOI] [PubMed] [Google Scholar]
- 4.Brodie D, Moscrip V, Hutcheon R. Body composition measurement: A review of hydrodensitometry, anthropometry, and impedance methods. Nutrition. 1998;14:296–310. doi: 10.1016/s0899-9007(97)00474-7. [DOI] [PubMed] [Google Scholar]
- 5.Centers for Disease Control and Prevention (CDC) [Accessed January 30, 2012];A SAS program for the CDC growth charts. 2011 http://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm.
- 6.Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC growth charts: United States. Adv Data. 2000;8:1–27. [PubMed] [Google Scholar]
- 7.Tanner JM. Normal growth and techniques of growth assessment. Clinics in endocrinology and metabolism. 1986;15:411–51. doi: 10.1016/s0300-595x(86)80005-6. [DOI] [PubMed] [Google Scholar]
- 8.Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–310. [PubMed] [Google Scholar]
- 9.Cohen J. Statistical power analysis for the behavioral sciences. 2. L. Erlbaum Associates; Hillsdale, N.J: 1988. [Google Scholar]
- 10.Newton RL, Alfonso A, White MA, et al. Percent body fat measured by BIA and DEXA in obese, African-American adolescent girls. Int J Obes. 2005;29:594–602. doi: 10.1038/sj.ijo.0802968. [DOI] [PubMed] [Google Scholar]
- 11.McClanahan BS, Stockton MB, Lanctot JQ, et al. Measurement of body composition in 8–10-year-old African-American girls: A comparison of dual-energy X-ray absorptiometry and foot-to-foot bioimpedance methods. Int J Pediatr Obes. 2009;4:389–396. doi: 10.3109/17477160902763358. [DOI] [PubMed] [Google Scholar]
- 12.Hosking J, Metcalf BS, Jeffery AN, Voss LD, Wilkin TJ. Validation of foot-to-foot bioelectrical impedance analysis with dual-energy X-ray absorptiometry in the assessment of body composition in young children: the EarlyBird cohort. Br J Nutr. 2006;96:1163–1168. doi: 10.1017/bjn20061960. [DOI] [PubMed] [Google Scholar]
- 13.Eisenkolbl J, Kartasurya M, Widhalm K. Underestimation of percentage fat mass measured by bioelectrical impedance analysis compared to dual energy X-ray absorptiometry method in obese children. Eur J Clin Nutr. 2001;55:423–429. doi: 10.1038/sj.ejcn.1601184. [DOI] [PubMed] [Google Scholar]