PEDIATRICOBESITY
ORIGINALRESEARCH
ORIGINALRESEARCH
doi:10.1111/ijpo.263
Bioelectrical impedance as a measure of change in body composition in young children K. A. Meredith-Jones1, S. M. Williams2 and R. W. Taylor1
1 Department of Medicine, University of Otago, Dunedin, New Zealand; 2Department of Preventive and Social Medicine, University of Otago, Dunedin, New Zealand
Address for correspondence: Dr KA Meredith-Jones, Department of Medicine, University of Otago, PO Box 56, Dunedin 9056, New Zealand. E-mail: [email protected] Received 20 March 2014; revised 22 June 2014; accepted 22 July 2014
Summary Background and objectives: The ability of bioelectrical impedance (BIA) to measure change in body composition in children has rarely been examined. Methods: Body composition was estimated by BIA (Tanita BC-418) and dualenergy x-ray absorptiometry (DXA) in 187 children aged 4–8 years at baseline and at 12 months. Change in body composition was compared between the two methods using mixed models. Results: Estimates of change in fat mass did not differ between BIA and DXA for overweight girls (mean difference between methods, 95% confidence interval: 0.04 kg, −0.19 to 0.28) or boys (0.07 kg, −0.14 to 0.27). BIA was also able to accurately detect change in fat-free mass, with no significant differences between methods (−0.14 kg, −0.10 to 0.38 in girls and −0.07 kg, −0.35 to −0.20 in boys). Change in percentage fat produced similar estimates in both genders (0.18%, −0.82 to 0.46 in girls and 0.38%, −0.37 to 1.13 in boys). BIA/DXA comparisons in normal weight children were also not significantly different, with the exception of percentage fat in girls, where BIA slightly underestimated change compared with DXA (0.7%, 0.02–0.37). Conclusion: BIA performed well as a measure of change in body composition, providing confidence for its use as an outcome measure in children. Keywords: Bioelectrical impedance, body composition, change, dual-energy x-ray absorptiometry. Abbreviations: BIA, bioelectrical impedance; BMI, body mass index; DXA, dual energy x-ray absorptiometry.
Introduction Overweight and obesity remain a significant problem for children worldwide, with more than one in three children in developed countries affected by excess body weight (1). The ongoing epidemic of obesity in children has highlighted the importance of measuring various aspects of body composition, including body fat and lean mass. Although the traditional method of categorizing overweight by body mass index (BMI) has been considered appropriate at a population level (2), BMI has limitations as an index of adiposity for monitoring treatment efficacy. Firstly, BMI correlates with both the fat and lean components of weight, so cannot differentiate between them and hence may give misleading information on body fat content in children (3,4). BMI also conveys negligible information about changes in body composition during growth with increases in BMI levels largely the result of increases in fat-free mass (FFM) rather than body fatness in normally growing children aged 8–18 years (5). Furthermore, in overweight and obese children undergoing weight loss treatment adiposity loss is often masked by growth.
© 2014 World Obesity. Pediatric Obesity 10, 252–259
As the prime medical importance of obesity lies in decreasing excess body fat without sacrificing lean tissue, measuring changes in body composition rather than BMI is required. Although several methods are available for measuring body composition in children, not all are widely accessible, particularly for clinicians working in routine primary care. Dual-energy x-ray absorptiometry (DXA) is precise and accurate and is suitable in children because of the minimal radiation exposure and relative comfort (6). However, its cost and availability limit its use in clinical settings and large-scale studies. In contrast, bioelectrical impedance (BIA) is quick, inexpensive and easy to use, making it suitable in a variety of settings. However, a recent review on the use of BIA to assess body composition in children (7) has shown that most assessments of the validity of BIA have been conducted in normal-weight populations, with few studies in overweight children (7). Furthermore, crosssectional studies have generally indicated that BIA produces significant magnitude bias and wide limits of agreement in comparison with DXA measures of body composition (8,9).
However, because clinicians and researchers may be more interested in tracking change in body composition, this relative inaccuracy in absolute body composition estimates may not be important if the different methods agree in the absolute change in body composition over time. To date there appear to be only two small longitudinal studies evaluating the ability of BIA to correctly measure change in body composition in children. Although Haroun et al. (8) reported small differences between BIA- and DXAmeasured fat mass and FFM (mean bias 0.2 kg), the sample size was only 17, meaning that differences between girls and boys or across BMI groups could not be ascertained. Another small study (n = 86) in normal and overweight children found that BIA significantly overestimated change in percentage body fat by approximately 1.4%, with the measurement error increasing as the change in the percentage fat increased (10). Thus, further examination of whether BIA can adequately measure change in body composition over time is required in larger samples of children from less homogenous groups. The aim of this study was to determine whether estimates of change in body composition (FFM, fat mass and percentage body fat) over 1 year differed in normal weight and overweight children measured using BIA and DXA.
Methods Participants The participants were part of the Motivational Interviewing in Treatment (MInT) randomized controlled trial aimed at improving health outcomes in young children (11). MInT consisted of a screening initiative to identify participants for a 2-year lifestyle intervention. All children aged 4–8 years enrolled at several Dunedin general practices were invited for baseline measurements (n = 1093) and using Centers for Disease Control reference norms (12), only those children identified as overweight or obese (BMI ≥85th percentile, n = 271) were invited to participate in the intervention. All children had estimates of body composition using BIA. For the specific purpose of validating BIA for measuring change in body composition, 40% of participants were randomly selected to undergo a DXA scan. This practice continued throughout the 17 months of recruitment. However, during the last third of the study, we invited every overweight child to undergo a DXA scan to ensure that we had relatively even numbers of normal and overweight participants for comparison. A total of 268 children underwent baseline DXA scans and all 268 children were invited to have a follow-up DXA scan at 12 months. Attempts were made to ensure measures of body composition by BIA were undertaken within 14 days of the DXA measurements, with a mean difference between BIA and DXA measures of 6.6 (5.9) days at baseline and 2.3 (8.4) days at follow-up.
Anthropometry and body composition Height and weight were measured without shoes and in light clothing using a stadiometer (SECA Leicester height measure, Invicta Plastics Ltd, Leicester, UK), and the BIA
| 253
(BC-418, Tanita Corp., Tokyo, Japan), respectively. Body weight was measured to the nearest 0.1 kg and height to the nearest 0.1 cm in duplicate. A third measurement was obtained if the difference between the first two measures was greater than the tolerance limit (0.7 cm for height and 0.5 kg for weight). The BC-418 (Tanita Corp.) was used for BIA measurements. The system consists of four stainless-steel rectangular foot-pad electrodes fastened to a metal platform set on force transducers for weight measurement, and two handgrips with an anterior and posterior electrode, giving eight electrodes overall. Measurements were carried out at 50 kHz with a 0.8-mA sine wave constant current. The child’s age, gender and height were entered, and body type was entered as ‘standard’ for all children. As children were measured wearing minimal clothing (i.e. shorts or underwear, light t-shirt) a correction for clothing weight was not required. Children were then asked to stand barefoot on the metal footplates of the machine while holding the handles for ∼1 min. Prediction equations supplied in the software were used in this study. The prediction equations used in this model are based on bioimpedance, weight, height and age, and were derived from calibration studies against whole-body DXA (13). A paediatric validation of the BC-418 model against DXA and air displacement plethysmography has also been performed (14), and the same BIA system has been used to develop FFM and body fat reference curves in children aged 5–18 years (15). Given the age of participants at baseline, it was not possible to undertake fasting measurements and voiding of the bladder was not always feasible. Children younger than 7 years of age had to have 7 entered as their age as this was the minimum age that could be selected for this model. DXA scans were performed and analysed by one experienced operator using a Lunar Prodigy scanner (GE Medical Systems Lunar, Madison, WI, USA); standard scanning procedures were used (6) and scans were analysed using the Lunar software package version 13.6 with the paediatric option selected. DXA FFM was calculated as the sum of lean body mass and bone mass. Comparisons between DXA and BIA were made for whole-body FFM, fat mass and percentage body fat.
Statistical analysis Bland and Altman (16) plots for FFM, fat mass and percentage fat were used to compare the DXA and BIA data collected at phase I and are provided simply for comparison with existing cross-sectional research. Bland and Altman plots were also provided to illustrate the pattern of the differences, the limits of agreement and magnitude bias. Mean change (95% confidence interval [CI]) in FFM, fat mass and fat percentage over 1 year are provided separately for BIA and DXA and were based on a paired t-test. The main analysis used mixed models with a random effect for participant to compare the differences between DXA and BIA (DXA-BIA) measures of body composition. The model included terms for BMI category, DXA or BIA and an interaction term. Correlations between the baseline values and
© 2014 World Obesity. Pediatric Obesity 10, 252–259
ORIGINALRESEARCH
BIA and change in children's body composition
ORIGINALRESEARCH
254 |
K. A. Meredith-Jones et al.
Table 1 Characteristics of the children in the study at baseline (n = 187)
Age (years) Height (cm) Weight (kg) BMI z-score Height z-score Weight z-score Weight status, n (%) Normal weight Overweight Obese Ethnicity, n (%) NZ European Maori Pacific Other Decline to answer
Girls (n = 106) Mean (SD)
Boys (n = 81) Mean (SD)
6.5 (1.5) 119.5 (11.2) 26.0 (7.7) 0.9 (0.9) 0.3 (0.9) 0.8 (0.9)
6.3 (1.4) 118.6 (10.3) 24.4 (5.3) 0.8 (1.0) 0.4 (1.1) 0.7 (1.1)
50 (47.2) 35 (33.0) 21 (19.8)
42 (51.9) 23 (28.4) 16 (19.8)
84 (79.2) 15 (14.2) 2 (1.9) 5 (4.7) 0
64 (79.0) 11 (13.6) 1 (1.2) 4 (4.9) 1 (1.2)
on average and there were no differences between genders for weight, height, weight status or ethnicity. Cross-sectionally, percentage body fat measured by DXA was highly correlated with estimates of percentage body fat by BIA at baseline (r = 0.916, P < 0.001). Similarly, fatfree and fat mass measured by DXA were highly correlated with BIA measures at baseline (FFM r = 0.956, P < 0.001; fat mass r = 0.974, P < 0.001). However, as Fig. 1 demonstrates, considerable magnitude bias was observed with BIA overestimating FFM and underestimating fat mass at higher values in both girls and boys.
Change in body composition
the change in values over 1 year were examined to provide guidance on the most appropriate way to examine change in body composition. As the number of obese children was relatively small (n = 37), the overweight and obese groups were combined for all analyses. Therefore, from this point in this manuscript, ‘overweight’ refers to both overweight (BMI 85– 8 kg and those who lost 95th centile), which limited our ability to assess changes specifically within this group and the lack of longer term data. Secondly, the results obtained for the comparison between methods may not be applicable to other BIA or DXA devices or software and as the age of participants meant it was not possible to undertake fasting measurements, BIA measures may have been affected. The proprietary prediction equation has not previously been validated in children under the age of 7 years. However, a significant portion of our sample (61%) was less than 7 years and as we observed no evidence of magnitude bias for those under age 7, our results suggest that this equation is valid in this age group. It must also be noted that although no significant mean differences were observed between DXA and BIA in overweight children, when assessing percentage fat change, individual biases were still present as indicated by wide limits of agreement (−4.6 to 5.1%). Thus, use of the technology to make single measurements in individuals is subject to error. However, our study also has several strengths including the longitudinal design and relatively large sample size for repeat measures of DXA. Although mean changes in fat mass, FFM and percentage fat were small, there was a wide range of change in body composition at the individual level across normal weight and overweight children. Treatment programmes to manage excess weight in children promote both dietary modifications and an increase in physical activity. Therefore, techniques that can accurately quantify both fat mass and FFM in paediatric overweight individuals are needed. The use of the BIA to estimate changes in body composition at an individual level during growth may differ from those assessed by DXA. However, as a low cost, easy to use alternative to DXA, BIA may still be used to assess changes in body composition at the group level.
Conflict of Interest Statement No conflict of interest was declared.
Acknowledgements Funding was provided from the Health Research Council of New Zealand and the University of Otago, New Zealand.
Author contributions KAM-J carried out the study, was responsible for data collection, drafted the initial manuscript and approved the final manuscript as submitted. SMW performed the data analyses, interpreted the results and contributed to the writing of the manuscript. RWT is the principal investigator of the study, provided critical input in all phases of the manuscript production, reviewed and revised the manuscript, and approved the final manuscript as submitted.
References 1. Wang Y, Lobstein TIM. Worldwide trends in childhood overweight and obesity. Int J Pediatr Obes 2006; 1: 11–25. 2. Freedman DS, Sherry B. The validity of BMI as an indicator of body fatness and risk among children. Pediatrics 2009; 124(Suppl. 1): S23–S34. 3. Wells JCK. A Hattori chart analysis of body mass index in infants and children. Int J Obes Relat Metab Disord 2000; 24: 325–329. 4. Ogden C, Carroll M, Kit B, Flegal K. Prevalence of obesity and trends in body mass index among US children and adolescents, 1999–2010. JAMA 2012; 307: 483–490. 5. Demerath EW, Schubert CM, Maynard LM, et al. Do changes in body mass index percentile reflect changes in body composition in children? Data from the Fels longitudinal study. Pediatrics 2006; 117: e487–e495. 6. Goulding A, Taylor RW, Jones IE, Lewis-Barned NJ, Williams SM. Body composition of 4- and 5-year-old New Zealand girls: a DXA study of initial adiposity and subsequent 4-year fat change. Int J Obes Relat Metab Disord 2003; 27: 410. 7. Talma H, Chinapaw MJM, Bakker B, HiraSing RA, Terwee CB, Altenburg TM. Bioelectrical impedance analysis to estimate body composition in children and adolescents: a systematic review and evidence appraisal of validity, responsiveness, reliability and measurement error. Obes Rev 2013; 14: 895–905. 8. Haroun D, Croker H, Viner RM, et al. Validation of BIA in obese children and adolescents and re-evaluation in a longitudinal study. Obesity 2009; 17: 2245–2250. 9. Lazzer S, Boirie Y, Meyer M, Vermorel M. Evaluation of two foot-to-foot bioelectrical impedance analysers to assess body composition in overweight and obese adolescents. Br J Nutr 2003; 90: 987–992. 10. Elberg J, McDuffie JR, Sebring NG, et al. Comparison of methods to assess change in children’s body composition. Am J Clin Nutr 2004; 80: 64–69. 11. Taylor R, Brown D, Dawson A, et al. Motivational interviewing for screening and feedback and encouraging lifestyle changes to reduce relative weight in 4–8 year old children: design of the MInT study. BMC Public Health 2010; 10: 271.
| 259
12. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat 11 2002; 246: 1–190. 13. Nunez C, Gallagher D, Visser M, Pi-Sunyer FX, Wang Z, Heymsfield SB. Bioimpedance analysis: evaluation of leg-to-leg system based on pressure contact footpad electrodes. Med Sci Sports Exerc 1997; 29: 524–531. 14. Pietrobelli A, Rubiano F, Wang J, Wang Z, Heymsfield SM. Validation of contact electrode bioimpedance analysis in a pediatric population. Eur Congress Obes 2005; 6: 132. Abstract. 15. McCarthy HD, Samani-Radia D, Jebb SA, Prentice AM. Skeletal muscle mass reference curves for children and adolescents. Pediatr Obes 2014; 9: 249–259. 16. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–310. 17. Cole TJ, Faith M, Pietrobelli A, Heo M. What is the best measure of adiposity change in growing children: BMI, BMI%, BMI z-score or BMI centile? Eur J Clin Nutr 2005; 59: 419–425. 18. Cleary J, Daniells S, Okely AD, Batterham M, Nicholls J. Predictive validity of four bioelectrical impedance equations in determining percent fat mass in overweight and obese children. J Am Diet Assoc 2008; 108: 136–139. 19. Eisenkolbl J, Kartasurya M, Widhalm K. Underestimation of percentage fat mass measured by biolectrical impedance analysis compared to dual energy X-ray absorptiometry method in obese children. Eur J Clin Nutr 2001; 55: 423. 20. Goldfield GS, Cloutier P, Mallory R, et al. Validity of foot-to-foot bioelectrical impedance analysis in overweight and obese children and parents. J Sports Med Phys Fitness 2006; 46: 447–453. 21. Frisard MI, Greenway FL, Delany JP. Comparison of methods to assess body composition changes during a period of weight loss. Obes Res 2005; 13: 845–854. 22. Tan YX, Nunez C, Sun Y, Zhang K, Wang Z, Heymsfield SB. New electrode system for rapid wholebody and segmental bioimpedance assessment. Med Sci Sports Exerc 1997; 29: 1269–1273. 23. Neovius M, Udden J, Hemmingsson E. Assessment of change in body fat percentage with DXA and eightelectrode BIA in centrally obese women. Med Sci Sports Exerc 2007; 39: 2199–2203. 24. Jebb SA, Siervo M, Murgatroyd PR, Evans S, Frühbeck G, Prentice AM. Validity of the leg-to-leg bioimpedance to estimate changes in body fat during weight loss and regain in overweight women: a comparison with multi-compartment models. Int J Obes (Lond) 2007; 31: 756–762. 25. Madlyn IF, Frank LG, James PD. Comparison of methods to assess body composition changes during a period of weight loss. Obes Res 2005; 13: 845–854. 26. Thomson R, Brinkworth GD, Buckley JD, Noakes M, Clifton PM. Good agreement between bioelectrical impedance and dual-energy X-ray absorptiometry for estimating changes in body composition during weight loss in overweight young women. Clin Nutr 2007; 26: 771–777.
© 2014 World Obesity. Pediatric Obesity 10, 252–259
ORIGINALRESEARCH
BIA and change in children's body composition
ORIGINALRESEARCH
ORIGINALRESEARCH
doi:10.1111/ijpo.263
Bioelectrical impedance as a measure of change in body composition in young children K. A. Meredith-Jones1, S. M. Williams2 and R. W. Taylor1
1 Department of Medicine, University of Otago, Dunedin, New Zealand; 2Department of Preventive and Social Medicine, University of Otago, Dunedin, New Zealand
Address for correspondence: Dr KA Meredith-Jones, Department of Medicine, University of Otago, PO Box 56, Dunedin 9056, New Zealand. E-mail: [email protected] Received 20 March 2014; revised 22 June 2014; accepted 22 July 2014
Summary Background and objectives: The ability of bioelectrical impedance (BIA) to measure change in body composition in children has rarely been examined. Methods: Body composition was estimated by BIA (Tanita BC-418) and dualenergy x-ray absorptiometry (DXA) in 187 children aged 4–8 years at baseline and at 12 months. Change in body composition was compared between the two methods using mixed models. Results: Estimates of change in fat mass did not differ between BIA and DXA for overweight girls (mean difference between methods, 95% confidence interval: 0.04 kg, −0.19 to 0.28) or boys (0.07 kg, −0.14 to 0.27). BIA was also able to accurately detect change in fat-free mass, with no significant differences between methods (−0.14 kg, −0.10 to 0.38 in girls and −0.07 kg, −0.35 to −0.20 in boys). Change in percentage fat produced similar estimates in both genders (0.18%, −0.82 to 0.46 in girls and 0.38%, −0.37 to 1.13 in boys). BIA/DXA comparisons in normal weight children were also not significantly different, with the exception of percentage fat in girls, where BIA slightly underestimated change compared with DXA (0.7%, 0.02–0.37). Conclusion: BIA performed well as a measure of change in body composition, providing confidence for its use as an outcome measure in children. Keywords: Bioelectrical impedance, body composition, change, dual-energy x-ray absorptiometry. Abbreviations: BIA, bioelectrical impedance; BMI, body mass index; DXA, dual energy x-ray absorptiometry.
Introduction Overweight and obesity remain a significant problem for children worldwide, with more than one in three children in developed countries affected by excess body weight (1). The ongoing epidemic of obesity in children has highlighted the importance of measuring various aspects of body composition, including body fat and lean mass. Although the traditional method of categorizing overweight by body mass index (BMI) has been considered appropriate at a population level (2), BMI has limitations as an index of adiposity for monitoring treatment efficacy. Firstly, BMI correlates with both the fat and lean components of weight, so cannot differentiate between them and hence may give misleading information on body fat content in children (3,4). BMI also conveys negligible information about changes in body composition during growth with increases in BMI levels largely the result of increases in fat-free mass (FFM) rather than body fatness in normally growing children aged 8–18 years (5). Furthermore, in overweight and obese children undergoing weight loss treatment adiposity loss is often masked by growth.
© 2014 World Obesity. Pediatric Obesity 10, 252–259
As the prime medical importance of obesity lies in decreasing excess body fat without sacrificing lean tissue, measuring changes in body composition rather than BMI is required. Although several methods are available for measuring body composition in children, not all are widely accessible, particularly for clinicians working in routine primary care. Dual-energy x-ray absorptiometry (DXA) is precise and accurate and is suitable in children because of the minimal radiation exposure and relative comfort (6). However, its cost and availability limit its use in clinical settings and large-scale studies. In contrast, bioelectrical impedance (BIA) is quick, inexpensive and easy to use, making it suitable in a variety of settings. However, a recent review on the use of BIA to assess body composition in children (7) has shown that most assessments of the validity of BIA have been conducted in normal-weight populations, with few studies in overweight children (7). Furthermore, crosssectional studies have generally indicated that BIA produces significant magnitude bias and wide limits of agreement in comparison with DXA measures of body composition (8,9).
However, because clinicians and researchers may be more interested in tracking change in body composition, this relative inaccuracy in absolute body composition estimates may not be important if the different methods agree in the absolute change in body composition over time. To date there appear to be only two small longitudinal studies evaluating the ability of BIA to correctly measure change in body composition in children. Although Haroun et al. (8) reported small differences between BIA- and DXAmeasured fat mass and FFM (mean bias 0.2 kg), the sample size was only 17, meaning that differences between girls and boys or across BMI groups could not be ascertained. Another small study (n = 86) in normal and overweight children found that BIA significantly overestimated change in percentage body fat by approximately 1.4%, with the measurement error increasing as the change in the percentage fat increased (10). Thus, further examination of whether BIA can adequately measure change in body composition over time is required in larger samples of children from less homogenous groups. The aim of this study was to determine whether estimates of change in body composition (FFM, fat mass and percentage body fat) over 1 year differed in normal weight and overweight children measured using BIA and DXA.
Methods Participants The participants were part of the Motivational Interviewing in Treatment (MInT) randomized controlled trial aimed at improving health outcomes in young children (11). MInT consisted of a screening initiative to identify participants for a 2-year lifestyle intervention. All children aged 4–8 years enrolled at several Dunedin general practices were invited for baseline measurements (n = 1093) and using Centers for Disease Control reference norms (12), only those children identified as overweight or obese (BMI ≥85th percentile, n = 271) were invited to participate in the intervention. All children had estimates of body composition using BIA. For the specific purpose of validating BIA for measuring change in body composition, 40% of participants were randomly selected to undergo a DXA scan. This practice continued throughout the 17 months of recruitment. However, during the last third of the study, we invited every overweight child to undergo a DXA scan to ensure that we had relatively even numbers of normal and overweight participants for comparison. A total of 268 children underwent baseline DXA scans and all 268 children were invited to have a follow-up DXA scan at 12 months. Attempts were made to ensure measures of body composition by BIA were undertaken within 14 days of the DXA measurements, with a mean difference between BIA and DXA measures of 6.6 (5.9) days at baseline and 2.3 (8.4) days at follow-up.
Anthropometry and body composition Height and weight were measured without shoes and in light clothing using a stadiometer (SECA Leicester height measure, Invicta Plastics Ltd, Leicester, UK), and the BIA
| 253
(BC-418, Tanita Corp., Tokyo, Japan), respectively. Body weight was measured to the nearest 0.1 kg and height to the nearest 0.1 cm in duplicate. A third measurement was obtained if the difference between the first two measures was greater than the tolerance limit (0.7 cm for height and 0.5 kg for weight). The BC-418 (Tanita Corp.) was used for BIA measurements. The system consists of four stainless-steel rectangular foot-pad electrodes fastened to a metal platform set on force transducers for weight measurement, and two handgrips with an anterior and posterior electrode, giving eight electrodes overall. Measurements were carried out at 50 kHz with a 0.8-mA sine wave constant current. The child’s age, gender and height were entered, and body type was entered as ‘standard’ for all children. As children were measured wearing minimal clothing (i.e. shorts or underwear, light t-shirt) a correction for clothing weight was not required. Children were then asked to stand barefoot on the metal footplates of the machine while holding the handles for ∼1 min. Prediction equations supplied in the software were used in this study. The prediction equations used in this model are based on bioimpedance, weight, height and age, and were derived from calibration studies against whole-body DXA (13). A paediatric validation of the BC-418 model against DXA and air displacement plethysmography has also been performed (14), and the same BIA system has been used to develop FFM and body fat reference curves in children aged 5–18 years (15). Given the age of participants at baseline, it was not possible to undertake fasting measurements and voiding of the bladder was not always feasible. Children younger than 7 years of age had to have 7 entered as their age as this was the minimum age that could be selected for this model. DXA scans were performed and analysed by one experienced operator using a Lunar Prodigy scanner (GE Medical Systems Lunar, Madison, WI, USA); standard scanning procedures were used (6) and scans were analysed using the Lunar software package version 13.6 with the paediatric option selected. DXA FFM was calculated as the sum of lean body mass and bone mass. Comparisons between DXA and BIA were made for whole-body FFM, fat mass and percentage body fat.
Statistical analysis Bland and Altman (16) plots for FFM, fat mass and percentage fat were used to compare the DXA and BIA data collected at phase I and are provided simply for comparison with existing cross-sectional research. Bland and Altman plots were also provided to illustrate the pattern of the differences, the limits of agreement and magnitude bias. Mean change (95% confidence interval [CI]) in FFM, fat mass and fat percentage over 1 year are provided separately for BIA and DXA and were based on a paired t-test. The main analysis used mixed models with a random effect for participant to compare the differences between DXA and BIA (DXA-BIA) measures of body composition. The model included terms for BMI category, DXA or BIA and an interaction term. Correlations between the baseline values and
© 2014 World Obesity. Pediatric Obesity 10, 252–259
ORIGINALRESEARCH
BIA and change in children's body composition
ORIGINALRESEARCH
254 |
K. A. Meredith-Jones et al.
Table 1 Characteristics of the children in the study at baseline (n = 187)
Age (years) Height (cm) Weight (kg) BMI z-score Height z-score Weight z-score Weight status, n (%) Normal weight Overweight Obese Ethnicity, n (%) NZ European Maori Pacific Other Decline to answer
Girls (n = 106) Mean (SD)
Boys (n = 81) Mean (SD)
6.5 (1.5) 119.5 (11.2) 26.0 (7.7) 0.9 (0.9) 0.3 (0.9) 0.8 (0.9)
6.3 (1.4) 118.6 (10.3) 24.4 (5.3) 0.8 (1.0) 0.4 (1.1) 0.7 (1.1)
50 (47.2) 35 (33.0) 21 (19.8)
42 (51.9) 23 (28.4) 16 (19.8)
84 (79.2) 15 (14.2) 2 (1.9) 5 (4.7) 0
64 (79.0) 11 (13.6) 1 (1.2) 4 (4.9) 1 (1.2)
on average and there were no differences between genders for weight, height, weight status or ethnicity. Cross-sectionally, percentage body fat measured by DXA was highly correlated with estimates of percentage body fat by BIA at baseline (r = 0.916, P < 0.001). Similarly, fatfree and fat mass measured by DXA were highly correlated with BIA measures at baseline (FFM r = 0.956, P < 0.001; fat mass r = 0.974, P < 0.001). However, as Fig. 1 demonstrates, considerable magnitude bias was observed with BIA overestimating FFM and underestimating fat mass at higher values in both girls and boys.
Change in body composition
the change in values over 1 year were examined to provide guidance on the most appropriate way to examine change in body composition. As the number of obese children was relatively small (n = 37), the overweight and obese groups were combined for all analyses. Therefore, from this point in this manuscript, ‘overweight’ refers to both overweight (BMI 85– 8 kg and those who lost 95th centile), which limited our ability to assess changes specifically within this group and the lack of longer term data. Secondly, the results obtained for the comparison between methods may not be applicable to other BIA or DXA devices or software and as the age of participants meant it was not possible to undertake fasting measurements, BIA measures may have been affected. The proprietary prediction equation has not previously been validated in children under the age of 7 years. However, a significant portion of our sample (61%) was less than 7 years and as we observed no evidence of magnitude bias for those under age 7, our results suggest that this equation is valid in this age group. It must also be noted that although no significant mean differences were observed between DXA and BIA in overweight children, when assessing percentage fat change, individual biases were still present as indicated by wide limits of agreement (−4.6 to 5.1%). Thus, use of the technology to make single measurements in individuals is subject to error. However, our study also has several strengths including the longitudinal design and relatively large sample size for repeat measures of DXA. Although mean changes in fat mass, FFM and percentage fat were small, there was a wide range of change in body composition at the individual level across normal weight and overweight children. Treatment programmes to manage excess weight in children promote both dietary modifications and an increase in physical activity. Therefore, techniques that can accurately quantify both fat mass and FFM in paediatric overweight individuals are needed. The use of the BIA to estimate changes in body composition at an individual level during growth may differ from those assessed by DXA. However, as a low cost, easy to use alternative to DXA, BIA may still be used to assess changes in body composition at the group level.
Conflict of Interest Statement No conflict of interest was declared.
Acknowledgements Funding was provided from the Health Research Council of New Zealand and the University of Otago, New Zealand.
Author contributions KAM-J carried out the study, was responsible for data collection, drafted the initial manuscript and approved the final manuscript as submitted. SMW performed the data analyses, interpreted the results and contributed to the writing of the manuscript. RWT is the principal investigator of the study, provided critical input in all phases of the manuscript production, reviewed and revised the manuscript, and approved the final manuscript as submitted.
References 1. Wang Y, Lobstein TIM. Worldwide trends in childhood overweight and obesity. Int J Pediatr Obes 2006; 1: 11–25. 2. Freedman DS, Sherry B. The validity of BMI as an indicator of body fatness and risk among children. Pediatrics 2009; 124(Suppl. 1): S23–S34. 3. Wells JCK. A Hattori chart analysis of body mass index in infants and children. Int J Obes Relat Metab Disord 2000; 24: 325–329. 4. Ogden C, Carroll M, Kit B, Flegal K. Prevalence of obesity and trends in body mass index among US children and adolescents, 1999–2010. JAMA 2012; 307: 483–490. 5. Demerath EW, Schubert CM, Maynard LM, et al. Do changes in body mass index percentile reflect changes in body composition in children? Data from the Fels longitudinal study. Pediatrics 2006; 117: e487–e495. 6. Goulding A, Taylor RW, Jones IE, Lewis-Barned NJ, Williams SM. Body composition of 4- and 5-year-old New Zealand girls: a DXA study of initial adiposity and subsequent 4-year fat change. Int J Obes Relat Metab Disord 2003; 27: 410. 7. Talma H, Chinapaw MJM, Bakker B, HiraSing RA, Terwee CB, Altenburg TM. Bioelectrical impedance analysis to estimate body composition in children and adolescents: a systematic review and evidence appraisal of validity, responsiveness, reliability and measurement error. Obes Rev 2013; 14: 895–905. 8. Haroun D, Croker H, Viner RM, et al. Validation of BIA in obese children and adolescents and re-evaluation in a longitudinal study. Obesity 2009; 17: 2245–2250. 9. Lazzer S, Boirie Y, Meyer M, Vermorel M. Evaluation of two foot-to-foot bioelectrical impedance analysers to assess body composition in overweight and obese adolescents. Br J Nutr 2003; 90: 987–992. 10. Elberg J, McDuffie JR, Sebring NG, et al. Comparison of methods to assess change in children’s body composition. Am J Clin Nutr 2004; 80: 64–69. 11. Taylor R, Brown D, Dawson A, et al. Motivational interviewing for screening and feedback and encouraging lifestyle changes to reduce relative weight in 4–8 year old children: design of the MInT study. BMC Public Health 2010; 10: 271.
| 259
12. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat 11 2002; 246: 1–190. 13. Nunez C, Gallagher D, Visser M, Pi-Sunyer FX, Wang Z, Heymsfield SB. Bioimpedance analysis: evaluation of leg-to-leg system based on pressure contact footpad electrodes. Med Sci Sports Exerc 1997; 29: 524–531. 14. Pietrobelli A, Rubiano F, Wang J, Wang Z, Heymsfield SM. Validation of contact electrode bioimpedance analysis in a pediatric population. Eur Congress Obes 2005; 6: 132. Abstract. 15. McCarthy HD, Samani-Radia D, Jebb SA, Prentice AM. Skeletal muscle mass reference curves for children and adolescents. Pediatr Obes 2014; 9: 249–259. 16. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–310. 17. Cole TJ, Faith M, Pietrobelli A, Heo M. What is the best measure of adiposity change in growing children: BMI, BMI%, BMI z-score or BMI centile? Eur J Clin Nutr 2005; 59: 419–425. 18. Cleary J, Daniells S, Okely AD, Batterham M, Nicholls J. Predictive validity of four bioelectrical impedance equations in determining percent fat mass in overweight and obese children. J Am Diet Assoc 2008; 108: 136–139. 19. Eisenkolbl J, Kartasurya M, Widhalm K. Underestimation of percentage fat mass measured by biolectrical impedance analysis compared to dual energy X-ray absorptiometry method in obese children. Eur J Clin Nutr 2001; 55: 423. 20. Goldfield GS, Cloutier P, Mallory R, et al. Validity of foot-to-foot bioelectrical impedance analysis in overweight and obese children and parents. J Sports Med Phys Fitness 2006; 46: 447–453. 21. Frisard MI, Greenway FL, Delany JP. Comparison of methods to assess body composition changes during a period of weight loss. Obes Res 2005; 13: 845–854. 22. Tan YX, Nunez C, Sun Y, Zhang K, Wang Z, Heymsfield SB. New electrode system for rapid wholebody and segmental bioimpedance assessment. Med Sci Sports Exerc 1997; 29: 1269–1273. 23. Neovius M, Udden J, Hemmingsson E. Assessment of change in body fat percentage with DXA and eightelectrode BIA in centrally obese women. Med Sci Sports Exerc 2007; 39: 2199–2203. 24. Jebb SA, Siervo M, Murgatroyd PR, Evans S, Frühbeck G, Prentice AM. Validity of the leg-to-leg bioimpedance to estimate changes in body fat during weight loss and regain in overweight women: a comparison with multi-compartment models. Int J Obes (Lond) 2007; 31: 756–762. 25. Madlyn IF, Frank LG, James PD. Comparison of methods to assess body composition changes during a period of weight loss. Obes Res 2005; 13: 845–854. 26. Thomson R, Brinkworth GD, Buckley JD, Noakes M, Clifton PM. Good agreement between bioelectrical impedance and dual-energy X-ray absorptiometry for estimating changes in body composition during weight loss in overweight young women. Clin Nutr 2007; 26: 771–777.
© 2014 World Obesity. Pediatric Obesity 10, 252–259
ORIGINALRESEARCH
BIA and change in children's body composition
