Validity and accuracy of body fat prediction equations using anthropometric measurements in children 7 – 10 years old

Validade e acurácia de equações de predição de gordura corporal usando medidas antropométricas em crianças de 7 a 10 anos

abstract

Children with a deficit of growth because of perinatal malnutrition present specificities in the percentage of body fat (%BF) that could not be detected by previous fat mass-based equations. This study developed and validated predictive equations of the %BF derived from anthropometric variables in children aged 7 to 10 living in Northeast Brazil, using dual-energy x-ray absorptiometry (DXA) as a reference. Body composition data from 58 children were utilized. DXA was used as a reference. A stepwise (forward) multiple regression statistical model was used to develop the new equations. The Bland-Altman analysis (CI: 95%), paired Student's t-test, and the intraclass correlation coefficient (ICC) was used to validate and compare the developed equations. Two new equations were developed for either gender: boys: %BF: 13.642 + (1.527*BMI) + (-0.345*Height) + (0.875*Triceps) + (0.290* Waist Circumference) and girls: %BF: -13.445 + (2.061*Tight). The Bland-Altman analysis showed good agreement, with limits ranging from -1.33 to 1.24% for boys and -3.35 to 4.08% for girls. The paired Student’s t-test showed no difference between %BF-DXA and the two new equations (p> 0.05), and the ICC was 0.948 and 0.915, respectively. DXA-based anthropometric equations provide an accurate and noninvasive method to measure changes in the %BF in children.

Key words:
Anthropometry; Body composition; Child

resumo

Crianças com déficit de crescimento por desnutrição perinatal apresentam especificidades na distribuição do percentual de gordura corporal (%GC) que não puderam ser detectadas por equações anteriores baseadas no %GC. Este estudo desenvolveu e validou equações preditivas do %GC derivadas de variáveis ​​antropométricas em crianças de 7 a 10 anos residentes no Nordeste do Brasil, utilizando como referência a absorciometria radiológica de dupla energia (DXA). Foram utilizados dados de composição corporal de 58 crianças. O DXA foi usado como modelo de referência. Um modelo estatístico de regressão múltipla stepwise (forward) foi usado para desenvolver as equações. A análise de Bland-Altman (IC: 95%), teste t de Student pareado e o coeficiente de correlação intraclasse (CCI) foram utilizados para validar e comparar as equações. Duas novas equações foram desenvolvidas para ambos os sexos: meninos: %GC: 13,642 + (1,527*IMC) + (-0,345*Altura) + (0,875*Tríceps) + (0,290* Circunferência da cintura) e meninas: %GC: - 13,445 + (2,061*coxa). A análise de Bland-Altman mostrou boa concordância, com limites variando de -1,33 a 1,24% para meninos e -3,35 a 4,08% para meninas. O teste t de Student pareado não mostrou diferença entre %GC-DXA e as duas novas equações (p>0,05), e o CCI foi de 0,948 e 0,915, respectivamente.

Palavras-chave:
Antropometria; Composição corporal; Criança

INTRODUCTION

In children, accurate measurement of body composition during childhood can be a useful strategy to provide information on growth and nutrition¹. Body mass index (BMI) is currently the most widely used tool to classify childhood overweight/obesity². However, BMI does not distinguish fat and lean tissue, and previous studies have also demonstrated that the relationship between children’s BMI and body fat varies with age and between ethnic groups³. Thus, to overcome these limitations, more accurate methods to estimate body composition need to be used.

Several methods to assess body composition have been developed for children with varying degrees of accuracy⁴. Laboratory methods, such as dual-energy x-ray absorptiometry (DXA), are widely accepted as reference analysis⁵. However, the high cost and transport infeasibility of the DXA equipment is evident limitations in many field settings⁶. A more practical, quick, and inexpensive alternative, although providing indirect evidence, is the use of equations to predict body fatness using easily measurable anthropometric variables⁷.

Many equations based on anthropometric variables have been developed to estimate body fat percentage (%BF) in children^8,9. Slaughter et al.¹⁰ developed body fat prediction equations for Afro- and Euro-American children using anthropometric variables. In this study, the equations are based on skinfold thickness (%BF boys = 0.735 [Tri + Calf] + 1.0 and %BF girls = 0.610 [Tri + Calf] + 5.1). However, children with growth deficits because of perinatal malnutrition or living in a region of nutrition transition present a relatively particular fatness distribution¹¹. Given these limitations, some disagreement is expected among Brazilian children, particularly because the distribution of overweight and obesity is heterogeneous across the country¹².

In the Northeast region of Brazil, the overweight rate of children is one of the highest in the country (28.8%), and obesity increased by 4.5-fold in the last two decades¹³. Vitória de Santo Antão, located in the State of Pernambuco, witnessed a reduction in infant undernutrition rates with a simultaneous increase in the prevalence of children with obesity¹⁴. This shift is driven by a single cause, the nutritional transition. Recently published data from our research group suggests that Vitória-PE is in an advanced phase of nutritional transition, where 43.7% of the calories consumed are coming from ultra-processed foods¹⁵. Thus, the goal of this study is to develop and validate predictive equations of the percentage of body fat derived from anthropometric variables in children aged 7 to 10 years old living in Vitoria de Santo Antão-Brazil.

METHODS

Sample

The study was conducted in Vitória de Santo Antão, a city in the Northeastern region of Brazil belonging to the Pernambuco state. The sample size was calculated using the G-Power software (version 3.1.9.4). An effect size of f² was assumed: 0.15; a significance level of 0.05%, and the power of the adopted test was 90%. For a multiple regression analysis (two tails; the number of independent variables = 4), a minimum sample size of 58 participants (31 boys and 27 girls) was necessary to identify a medium effect size.

Ethical aspects

Ethical approval was obtained from the Research Ethics Committee with Human Subjects of the institution Federal University of Pernambuco (CAAE 91338718.0.0000.5208), and the protocol was written according to the standards established by the Declaration of Helsinki. All subjects provided written informed consent before the study and their rights were protected.

Anthropometry and body composition

Lohman described standardization as used to measure body weight, height and waist, and hip circumferences¹⁶. Bodyweight was collected using a digital scale (Omron HBF-214/HBF214-LA, São Paulo, Brazil). Height was assessed using a portable stadiometer (compact MD - HT / 01). Both measurements were performed with the child barefooted. Circumferences were measured using an anthropometric measuring tape (Cescorf, Porto Alegre, Rio Grande do Sul). Body mass index (BMI) was calculated using the formula: body mass (Kg) / height (m²). Eight skinfolds were collected using a Cescorf Digital caliper (0.1mm, Porto Alegre, Brazil)¹⁶.

Dual-energy X-ray Absorptiometry (DXA)

Total body composition was assessed using the DXA (HOLOGIC QDR WI Bedford, MA, USA software APEX™). The radiation dose received was less than 1.0 mRem, ensuring the assessment was accurate, safe, and valid for the child population¹⁷. The device was previously calibrated using a phantom manufacturer provided. To avoid the presence of systematic bias and to guarantee reliability between assessments, a single scanner was performed per child¹⁷. Children were instructed to wear light clothing without metal adornments.

Reliability

The technical error of measurement (TEM) was performed in 10% of the total sample. A total of 6 children were evaluated by two evaluators on a single collection day. Both evaluators had the measurements compared and expressed in values: absolute (mm) and relative (%) (Supplementary Material Table 1).

Statistical analysis

The normality of the data distribution was verified by the Kolmogorov-Smirnov test and by visual inspection of histograms. All analyses were performed separately for both sexes. Mean and Standard Deviation (SD) were used for descriptive statistics. For the development and validation of the equations, 58 children were stratified for both sexes and randomly divided into two groups: development group (n = 14 boys; n = 12 girls) and validation group (n = 17 boys; n = 15 girls). The anthropometric characteristics of the groups were compared using the student’s t-test. The % BF derived from DXA measurements was used as a criterion for the development of predictive equations.

The stepwise multiple regression analysis (forward) was used to choose the best predictors. BMI, height, tricipital skinfold, waist circumference (boys; model R²: 0.94), and thigh skinfold (girls; model R²: 0.85) were considered independent variables. The variation inflation factor (VIF) ≤ 10 was adopted to justify the absence of multicollinearity among the predictors. The Durbin-Watson statistic was used to verify the presence of dependence between the residuals of the predictor variables.

The developed equations were used to predict the %BF-DXA in the validation group and compared with equations published in the literature. The choice of previously published equations followed these selection criteria: (i) studies with equations that differed from the age range of the current study were excluded; (ii) equations developed for different populations as athletes or groups of specific diseases were excluded. Thus, only studies that were methodologically similar to the present study were selected, as shown in Supplementary Material Table 2.

The Bland-Altman analysis (CI: 95%), paired Student's t-test, and the Intraclass correlation coefficient (CCI) was used in the set of tests to validate the developed equations¹⁸. The magnitude of the CCI was weak: 0.10 ≤ 0.30; moderate: 0.30 ≤ 0.50; strong: 0.50 ≤ 0.70; very strong: 0.70 ≤ 0.90; and perfect: 1.0. All analyses were performed in the Statistical Package for the Social Science (SPSS), version 22, the level of significance established p ≤ 0.005.

RESULTS

Table 1 presents descriptive sample characteristics. There were no differences in the anthropometric characteristics between the two groups for both sexes. Two new predictive equations for %BF were developed for children (both sexes) aged 7-10 years from anthropometric variables inserted in the statistical stepwise multiple regression model (forward).

Table 1
Mean and standard deviation of demographic variables and body composition of children (both sexes) aged 7 - 10 years.

	Development group				Validation group
	Boys (n=14)		Girls (n=12)		Boys (n=17)		Girls (n=15)
Variables	Mean	±SD	Mean	±SD	Mean	±SD	Mean	±SD	P Boys	P Girls
%BF-DXA (%)	30.9	9.5	32.6	7.8	31.5	8.6	31.7	6.6	0.633	0.658
Abdominal (mm)	16.6	4.1	15.4	3.2	19.5	1.3	15.3	2.7	0.443	0.795
Axillary (mm)	16.8	3.8	14.9	3.4	12.7	2.6	11.4	2.7	0.190	0.308
Bicipital (mm)	16.3	3.6	15.4	2.9	9.5	1.7	9.9	1.4	0.240	0.207
Waist.C (cm)	61.9	8.4	62.8	7.9	60.1	5.7	54.7	3.9	0.614	0.115
BMC (g)	991.9	227.2	964.3	135.4	950.1	136.2	939.2	190.1	0.733	0.960
Hip.C (cm)	67.0	8.9	72.8	6.7	68.4	5.2	64.0	3.7	0.227	0.428
Thigh (mm)	22.8	7.1	22.3	3.5	19.7	2.6	20.4	2.0	0.342	0.555
BMD (g/cm2)	788.6	89.6	762.3	71.6	762	59.0	733.4	84.0	0.409	0.329
Height (cm)	127.6	8.4	132.9	7.5	137.7	3.9	136.8	3.7	0.130	0.302
Age (y)	9.4	1.9	9.4	1.9	9.2	1.0	9.0	1.0	0.210	0.732
BMI (kg/m2)	19.7	4.4	17.2	3.3	21.2	4.6	20.2	3.0	0.711	0.185
FM (g)	528.3	211.3	1506.7	257	1819.1	542.9	2035	779	0.211	0.117
LM (g)	2461.9	553.6	2015.7	446.1	2276.1	474.0	1924	338	0.483	0.607
Calf (mm)	21.1	5.0	19.3	3.4	14.7	2.2	16.8	2.6	0.320	0.513
Weight (kg)	37.6	7.5	31.4	8.4	36.8	4.9	31.3	3.9	0.937	0.898
Subscapular (mm)	15.2	4.6	14.7	3.4	11.5	1.6	12.3	2.4	0.184	0.683
Suprailiac (mm)	20.5	3.6	16.1	17.2	11.0	2.7	11.5	2.4	0.442	0.242
Tricipital (mm)	17.1	4.0	3.0	3.2	12.8	1.4	12.8	1.5	0.238	0.472

• Note. %BF-DXA: Percentage of body fat estimated by DXA; Waist.C: Waist circumference (cm); Hip.C: Hip circumference (cm); Subcutaneous skinfolds expressed in millimeters (mm); BMI: body mass index; BMC: Bone mineral content (g); BMD: Bone mineral density (g / cm²); FM: Fat mass (g); LM: Lean mass (g).

Both models included the variables: BMI, height, tricipital skinfold, waist circumference, and thigh skinfold which explained 94.0%, and 85.0% of the percentage of BF-DXA variance, respectively (Table 2). The values of the Durbin-Watson statistic (1.765 – 1.859) indicated the absence of autocorrelation (dependence) between the residues. The VIF values (1.644 - 4.050; 1,000) between the predictor variables demonstrated the lack of multicollinearity. For both equations, the SEE values were less than 3.20% (Table 2).

Table 2
% BF predictive equations for children of both sexes 7 - 10 years old generated in this study.

	%BF predictive equations	R	R2	SEE	DW	VIF
Boys (n = 14)	%BF: 13.642 + (1.527*BMI) + (-0.345*H) + (0.875*Tri) + (0.290* Waist C.)	97.0	94.0	2.818	1.859	1.644 – 4.050
Girls (n = 12)	%BF: -13.445 + (2.061*Thi)	92.2	85.0	3.200	1.765	1.000

• Note. %BF = percentage of body fat; BMI = body mass index; H= Height; Tri = tricipital skinfold (mm); Thi = thigh skinfold (mm); Waist.C = Waist circumference; SEE= Standard error of the estimate; DW = Durbin-Watson statistics; VIF= Variance inflation factor.

Figure 1 shows the analysis agreement Bland-Altman method between %BF-DXA and the %BF-new equations developed for both sexes. The central lines did not show significant deviations from zero for both equations (Boys = bias: 0.05; Girls = bias: -0.4; p ≤0.005). The Bland-Altman analysis showed good agreement between both methods, with limits of agreement ranging from -1.33 to 1.24% for boys and -3.35 to 4.08% for girls.

Figure 1
Limits of agreement between %BF-DXA and two new %BF equations in the validation sample (boys n = 17; girls n = 15) to children of both sexes 7-10 years. Boys (A) and Girls (B). Results obtained by Bland-Altman analysis.

The agreement between %BF-DXA and the eight predictive %BF equations previously published in the literature is shown in Figure 2. Initially, the central lines showed significant deviations from zero for the six equations (Figure 2A to 2F). The biases ranged from -8.71 to - 10.24; p ≤0.005 for boys and from-8.45 to 10.09; p ≤0.005 for girls. Only Cameron equations did not show significant deviations from zero between the central lines (Boys = bias: 0.89; Girls = bias: -1.14; p ≤0.005) (Figure 2G and 2H). Six %BF equations (Figure 2A and 2F) previously developed for both sexes showed discordant limits of agreement compared to the reference method. Such limits ranged from -1.97 to 22.44% for boys and from -6.26 to 23.16% for girls (Figure 2A and 2B); -1.42 to 19.62% for boys and -27.73 to 7.54% for girls (Figure 2C and 2D); -6.81 to 24.23% for boys and -2.34 to 23.88% for girls (Figure 2E and 2F). Only Cameron equations showed good agreement between both methods. The limits of agreement ranged from -1.33 to 1.24% for boys and from -3.35 to 4.08% for girls (Figure 2G and 2H).

Figure 2
Limits of agreement between %BF-DXA reference values and %BF predicted equations previously published in the validation sample (boys n=17; girls n=15). Figure 2A and 2B shows the Slaughter I equations agreement. Figure 2C and 2D shows the Slaughter II equations agreement. Figure 2E and 2F shows the Bray equation's agreement. Figure 2G and 2H shows Cameron's equations agreement. Boys (left) Girls (right). Results were obtained through the Bland-Altman analysis.

Table 3 demonstrates the comparison of %BF-DXA and all predictive equations. There was no difference between %BF-DXA and the two new equations developed for both sexes (p> 0.05). The Intraclass correlation coefficient (ICC) was classified as very strong between DXA, and the two new equations developed for both sexes (ICC = 0.948; 0.915). The ICC was moderate for both Slaughter I equations (ICC = 0.558; 0.455). Both Slaughter II equations, the ICC varied between moderate (ICC = 0.329) and strong (ICC = 0.655), respectively. Bray’s equations, the ICC were classified as moderate (ICC = 0.366) and strong (ICC = 0.679). Among Cameron equations, ICC was classified as very strong (ICC = 0.700; 0.764). %BF was overestimated in the equations of Slaughter I - Boys (10.17; p <0.005) and Slaughter I - Girls (8.42; p <0.005); Slaughter II - Boys (9.10; p <0.005) Slaughter II - Girls (10.06; p <0.005); Bray-Boys (8.72; p <0.005) Bray - Girls (10.74; p> 0.005). Cameron's equations - Boys (-0.65; p <0.005) and Cameron - Girls (-1.13; p> 0.005) showed no differences (Table 3).

Table 3
Comparison of the body fat percentage, measured by DXA and the predictive equations.

	Authors	N	BF (%) Equations	BF (%) DXA	Intraclass correlation coefficient (ICC)	t	p
			Mean (±SD)	Mean (±SD)
1	%BF-New equations (boys)	17	32.29 (9.29)	31.94 (8.72)	0.948	-3.53	0.728
2	%BF-New equations (girls)	15	31.73 (7.95)	32.13 (6.56)	0.915	0.40	0.717
3	%BF-Slaughter I (boys)	17	21.29 (7.38)	31.94 (8.63)	0.558	10.17	0.000
4	%BF-Slaughter I (girls)	15	23.70 (7.93)	32.13 (6.56)	0.455	8.42	0.001
5	%BF-Slaughter II (boys)	17	22.37 (7.66)	31.94 (8.63)	0.655	9.10	0.000
6	%BF-Slaughter II (girls)	15	22.06 (8.90)	32.13 (6.56)	0.329	10.06	0.001
7	%BF-Bray (boys)	17	22.76 (7.44)	31.94 (8.63)	0.679	8.72	0.000
8	%BF-Bray (girls)	15	21.39 (6.77)	32.13 (6.56)	0.366	1.74	0.000
9	%BF-Cameron (boys)	17	31.54 (7.67)	31.94 (8.63)	0.700	-0.65	0.974
10	%BF-Cameron (girls)	15	32.25 (8.89)	32.13 (6.56)	0.764	-1.13	0.951

DISCUSSION

The present study proposed a novel and simple method for estimating %BF in a group of children aged 7-10 years using the DXA as a reference. Our data demonstrated that the %BF calculated by the two new anthropometric equations was highly associated with DXA (97.0 and 92.2), remarked by the high power of explanation (R²: 94.0 and 85.0) and low errors (SEE: 2.818 and 3.200) (Table 2). The descriptive dataset values from BMC, FM, LM, and BMD assessed by DXA demonstrate valuable information for further studies (Table 1). Another aim of this study was to analyze the validity of previously developed equations. Thus, except for the Cameron equations, none of the previously published equations was valid in our validation sample (Figure 2A to 2H; Table 3). This indicates that specific predictive equations are not universal and must be used in the context of the population characteristics.

This study demonstrated that equations developed for boys included BMI, Height, Triceps, and Waist circumference as independent variables. The inclusion of waist circumference is not surprising in our predicted model because the association between waist circumference and %BF is a phenomenon well recognized. Indeed, it was reported a positive correlation between waist circumference and DXA-%BF in Danish children¹⁹. In Brazilian children, it was demonstrated that the waist circumference at the midpoint had the best correlation with %BF in boys and girls aged 6 to 9 years²⁰. Of all independent predictors, only height had a negative regression coefficient (-0.345), which means the tallest children tend to have lower values of %BF while the smaller children tend to do the inverse. A similar finding was found by Ortiz-Hernández et al.²¹, where the equations were developing for Mexican boys containing only height as a negative coefficient. BMI was also included in our regression model. Although BMI does not provide direct information on body composition, as the measured variables are weight and height, the strength of the linear relationship with %BF is well-described²². In this sense, studies developed by Costa-Urrutia et al. and Srdić et al.^23,24 demonstrated a stronger correlation between BMI and %BF and children aged 7. The triceps skinfold also was included in our final model, corroborating with previous studies^12,25. More recently, two prediction equations for healthy children utilizing the triceps skinfold as an independent variable demonstrated a high correlation with %BF-DXA and lower bias²⁶. Thus, the final predictive model included an interesting combination of the circumference and skinfold thickness. Investigations about these interactions have suggested good values for calculating the percentage of body fat²⁷.

This study also demonstrated that the simplest model developed for girls included only one skinfold thickness as an independent variable. In this sense, only thigh skinfold thickness was a better predictor of %BF-DXA than any other anthropometric variable provided in the present study. Other studies have reported a high correlation between the thickness of adipose tissue of the lower limb (e.g., thigh and calf thickness) with %BF in children and adolescents using different criteria methods (e.g., DXA, hydrostatic weighing, and 4C model)¹². Another study investigated the relations between %BF assessed by DXA with upper body and lower limb skinfold in young healthy. As a result, the lower limb skinfold was more strongly related to %BF than upper body skinfolds²⁸. A broad range of studies showed that the inclusion of lower limb sites from the thigh and calf isolated or combined with other sites from the upper body, for example, was significantly related to total body fat²⁹. The British Olympic Association (BOA) recognized and recommended that the anterior thigh skinfold thickness should be included in the Durin and Womersley equation to increase the estimated capacity of total body fat³⁰. Therefore, our equation can provide an advantage in time, accuracy, and clinical practice, as well as facilitate the assessment of % BF in studies with children.

This study suggests that the two newly developed equations, specific for boys and girls, can be appropriate for the assessment of %BF in children living in a region with the nutritional transition. These equations can play a crucial role in monitoring the changes imposed by the environmental impacts of nutritional transition over body composition and may help in predicting illnesses associated with nutrition-related non-communicable diseases.

CONCLUSIONS

The present study provided two new equations developed and validated in children living in a region witnessing a nutritional transition. Also, this study has descriptive body composition data from these children utilizing a reference model (DXA). All six equations except for the Cameron equations used in this study may not be suitable for assessing %BF in children living in Vitória de Santo Antão. Hence, our newly proposed equations provide a better reflection of %BF than any other previously proposed equation and provide a better predictor of excessively high or low levels of percentage body fat. In this context, a further cross-validation study on independent children’s cohorts is essential to assess the predictability of developed equations in different samples.

COMPLIANCE WITH ETHICAL STANDARDS

SUPPLEMENTARY MATERIAL

Supplementary material accompanies this paper.

Supplementary File 1: Free access in https://osf.io/fgcdb/

Supplementary File 2: Free access in https://osf.io/fgcdb/

These materials are available as part of the online article from 10.1590/1980-0037.2022v24e86719

ACKNOWLEDGEMENTS

This study was supported by National Council for Scientific and Technological Development [CNPq, 312079/2018-4], Coordination for the Improvement of Higher Level -or Education- Personnel [CAPES 977/20], and State of Pernambuco Science and Technology Support Foundation [FACEPE, APQ: 0797-4.05/14]. We thank all children and their families for participating in this study.

• How to cite this articleSantos RM, Nobre IG, Santos GCJ, Oliveira TLPSA, Ribeiro IC, Santos MAM, Pirola L, Leandro CG. Validity and accuracy of body fat prediction equations using anthropometric measurements in children 7 – 10 years old. Rev Bras Cineantropom Desempenho Hum 2022, 24:e86719. DOI: http://doi.org/10.1590/1980-0037.2022v24e86719
• Funding This research was funded in part by National Council for Scientific and Technological Development [CNPq, 312079/2018-4], Coordination for the Improvement of Higher Level -or Education- Personnel [CAPES 977/20], and State of Pernambuco Science and Technology Support Foundation [FACEPE, APQ: 0797-4.05/14].

REFERENCES

• ¹
  Sampaio ADS, Epifanio M, Costa CAD, Bosa VL, Benedetti FJ, Sarria EE, et al. Evidence on nutritional assessment techniques and parameters used to determine the nutritional status of children and adolescents: systematic review. Cien Saude Colet. 2018;23(12):4209-19. http://dx.doi.org/10.1590/1413-812320182312.31502016 PMid:30540004.
  » http://dx.doi.org/10.1590/1413-812320182312.31502016
• ²
  Gutin I. In BMI we trust: reframing the body mass index as a measure of health. Soc Theory Health. 2018;16(3):256-71. http://dx.doi.org/10.1057/s41285-017-0055-0 PMid:31007613.
  » http://dx.doi.org/10.1057/s41285-017-0055-0
• ³
  Belarmino G, Horie LM, Sala PC, Torrinhas RS, Heymsfield SB, Waitzberg DL. Body adiposity index performance in estimating body fat in a sample of severely obese Brazilian patients. Nutr J. 2015;14:130. http://dx.doi.org/10.1186/s12937-015-0119-8 PMid:26717977.
  » http://dx.doi.org/10.1186/s12937-015-0119-8
• ⁴
  Kuriyan R. Body composition techniques. Indian J Med Res. 2018;148(5):648-58. http://dx.doi.org/10.4103/ijmr.IJMR_1777_18 PMid:30666990.
  » http://dx.doi.org/10.4103/ijmr.IJMR_1777_18
• ⁵
  Borga M, West J, Bell JD, Harvey NC, Romu T, Heymsfield SB, et al. Advanced body composition assessment: from body mass index to body composition profiling. J Investig Med. 2018;66(5):1-9. http://dx.doi.org/10.1136/jim-2018-000722 PMid:29581385.
  » http://dx.doi.org/10.1136/jim-2018-000722
• ⁶
  Ponti F, Plazzi A, Guglielmi G, Marchesini G, Bazzocchi A. Body composition, dual-energy X-ray absorptiometry and obesity: the paradigm of fat (re)distribution. BJR Case Rep. 2019;5(3):20170078. http://dx.doi.org/10.1259/bjrcr.20170078 PMid:31555464.
  » http://dx.doi.org/10.1259/bjrcr.20170078
• ⁷
  Forte GC, Rodrigues CAS, Mundstock E, Santos TSD, Detoni A Fo, Noal J, et al. Can skinfold thickness equations be substituted for bioimpedance analysis in children? J Pediatr (Rio J). 2021;97(1):75-9. http://dx.doi.org/10.1016/j.jped.2019.12.006 PMid:32084440.
  » http://dx.doi.org/10.1016/j.jped.2019.12.006
• ⁸
  Almeida SM, Furtado JM, Mascarenhas P, Ferraz ME, Silva LR, Ferreira JC, et al. Anthropometric predictors of body fat in a large population of 9-year-old school-aged children. Obes Sci Pract. 2016;2(3):272-81. http://dx.doi.org/10.1002/osp4.51 PMid:27708844.
  » http://dx.doi.org/10.1002/osp4.51
• ⁹
  Stevens J, Ou FS, Cai J, Heymsfield SB, Truesdale KP. Prediction of percent body fat measurements in Americans 8 years and older. Int J Obes. 2016;40(4):587-94. http://dx.doi.org/10.1038/ijo.2015.231 PMid:26538187.
  » http://dx.doi.org/10.1038/ijo.2015.231
• ¹⁰
  Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Stillman RJ, Van Loan MD, et al. Skinfold equations for estimation of body fatness in children and youth. Hum Biol. 1988;60(5):709-23. PMid:3224965.
• ¹¹
  Santos FK, Moura Dos Santos MA, Almeida MB, Nobre IG, Nobre GG, Ferreira e Silva E, et al. Biological and behavioral correlates of body weight status among rural Northeast Brazilian schoolchildren. Am J Hum Biol. 2018;30(3):e23096. http://dx.doi.org/10.1002/ajhb.23096 PMid:29341385.
  » http://dx.doi.org/10.1002/ajhb.23096
• ¹²
  Fradkin C, Valentini NC, Nobre GC, Dos Santos JOL. Obesity and overweight among brazilian early adolescents: variability across region, socioeconomic status, and gender. Front Pediatr. 2018;6(6):81. http://dx.doi.org/10.3389/fped.2018.00081 PMid:29682495.
  » http://dx.doi.org/10.3389/fped.2018.00081
• ¹³
  Niehues JR, Gonzales AI, Lemos RR, Bezerra PP, Haas P. Prevalence of overweight and obesity in children and adolescents from the age range of 2 to 19 years old in Brazil. Int J Pediatr. 2014;2014(7):583207. http://dx.doi.org/10.1155/2014/583207 PMid:24995019.
  » http://dx.doi.org/10.1155/2014/583207
• ¹⁴
  Nobre GG, de Almeida MB, Nobre IG, Dos Santos FK, Brinco RA, Arruda-Lima TR, et al. Twelve weeks of plyometric training improves motor performance of 7- to 9-year-old boys who were overweight/obese: a randomized controlled intervention. J Strength Cond Res. 2017;31(8):2091-9. http://dx.doi.org/10.1519/JSC.0000000000001684 PMid:27787471.
  » http://dx.doi.org/10.1519/JSC.0000000000001684
• ¹⁵
  Oliveira T, Ribeiro I, Jurema-Santos G, Nobre I, Santos R, Rodrigues C, et al. Can the consumption of ultra-processed food be associated with anthropometric indicators of obesity and blood pressure in children 7 to 10 years old? Foods. 2020;9(11):1567. http://dx.doi.org/10.3390/foods9111567 PMid:33126771.
  » http://dx.doi.org/10.3390/foods9111567
• ¹⁶
  Lohman TG. Applicability of body composition techniques and constants for children and youths. Exerc Sport Sci Rev. 1986;14(1):325-57. http://dx.doi.org/10.1249/00003677-198600140-00014 PMid:3525188.
  » http://dx.doi.org/10.1249/00003677-198600140-00014
• ¹⁷
  Shepherd JA, Wang L, Fan B, Gilsanz V, Kalkwarf HJ, Lappe J, et al. Optimal monitoring time interval between DXA measures in children. J Bone Miner Res. 2011;26(11):2745-52. http://dx.doi.org/10.1002/jbmr.473 PMid:21773995.
  » http://dx.doi.org/10.1002/jbmr.473
• ¹⁸
  Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1(8476):307-10. http://dx.doi.org/10.1016/S0140-6736(86)90837-8 PMid:2868172.
  » http://dx.doi.org/10.1016/S0140-6736(86)90837-8
• ¹⁹
  Wohlfahrt-Veje C, Tinggaard J, Winther K, Mouritsen A, Hagen CP, Mieritz MG, et al. Body fat throughout childhood in 2647 healthy Danish children: agreement of BMI, waist circumference, skinfolds with dual X-ray absorptiometry. Eur J Clin Nutr. 2014;68(6):664-70. http://dx.doi.org/10.1038/ejcn.2013.282 PMid:24473457.
  » http://dx.doi.org/10.1038/ejcn.2013.282
• ²⁰
  Sant’Anna MSL, Tinoco ALA, Rosado LEFPL, Sant’Ana LFR, Mello AC, Brito ISS, et al. Body fat assessment by bioelectrical impedance and its correlation with different anatomical sites used in the measurement of waist circumference in children. J Pediatr (Rio J). 2009;85(1):61-6. http://dx.doi.org/10.2223/JPED.1871 PMid:19198741.
  » http://dx.doi.org/10.2223/JPED.1871
• ²¹
  Ortiz-Hernández L, Vega López AV, Ramos-Ibáñez N, Cázares Lara LJ, Medina Gómez RJ, Pérez-Salgado D. Equations based on anthropometry to predict body fat measured by absorptiometry in schoolchildren and adolescents. J Pediatr (Rio J). 2017;93(4):365-73. http://dx.doi.org/10.1016/j.jped.2016.08.008 PMid:28132762.
  » http://dx.doi.org/10.1016/j.jped.2016.08.008
• ²²
  Kehoe SH, Krishnaveni GV, Lubree HG, Wills AK, Guntupalli AM, Veena SR, et al. Prediction of body-fat percentage from skinfold and bio-impedance measurements in Indian school children. Eur J Clin Nutr. 2011;65(12):1263-70. http://dx.doi.org/10.1038/ejcn.2011.119 PMid:21731039.
  » http://dx.doi.org/10.1038/ejcn.2011.119
• ²³
  Costa-Urrutia P, Vizuet-Gámez A, Ramirez-Alcántara M, Guillen-González MÁ, Medina-Contreras O, Valdes-Moreno M, et al. Obesity measured as percent body fat, relationship with body mass index, and percentile curves for Mexican pediatric population. PLoS One. 2019;14(2):e0212792. http://dx.doi.org/10.1371/journal.pone.0212792 PMid:30802270.
  » http://dx.doi.org/10.1371/journal.pone.0212792
• ²⁴
  Srdić B, Obradović B, Dimitrić G, Stokić E, Babović SS. Relationship between body mass index and body fat in children-Age and gender differences. Obes Res Clin Pract. 2012;6(2):e91-174. http://dx.doi.org/10.1016/j.orcp.2011.08.153 PMid:24331257.
  » http://dx.doi.org/10.1016/j.orcp.2011.08.153
• ²⁵
  Goran MI, Driscoll P, Johnson R, Nagy TR, Hunter G. Cross-calibration of body-composition techniques against dual-energy X-ray absorptiometry in young children. Am J Clin Nutr. 1996;63(3):299-305. http://dx.doi.org/10.1093/ajcn/63.3.299 PMid:8602584.
  » http://dx.doi.org/10.1093/ajcn/63.3.299
• ²⁶
  Wendel D, Weber D, Leonard MB, Magge SN, Kelly A, Stallings VA, et al. Body composition estimation using skinfolds in children with and without health conditions affecting growth and body composition. Ann Hum Biol. 2017;44(2):108-20. http://dx.doi.org/10.3109/03014460.2016.1168867 PMid:27121656.
  » http://dx.doi.org/10.3109/03014460.2016.1168867
• ²⁷
  Simões M, Severo M, Oliveira A, Ferreira I, Lopes C. Predictive equations for estimating regional body composition: a validation study using DXA as criterion and associations with cardiometabolic risk factors. Ann Hum Biol. 2016;43(3):219-28. http://dx.doi.org/10.3109/03014460.2015.1054427 PMid:26226974.
  » http://dx.doi.org/10.3109/03014460.2015.1054427
• ²⁸
  Sun SS, Chumlea WC, Heymsfield SB, Lukaski HC, Schoeller D, Friedl K, et al. Development of bioelectrical impedance analysis prediction equations for body composition with the use of a multicomponent model for use in epidemiologic surveys. Am J Clin Nutr. 2003;77(2):331-40. http://dx.doi.org/10.1093/ajcn/77.2.331 PMid:12540391.
  » http://dx.doi.org/10.1093/ajcn/77.2.331
• ²⁹
  Eston RG, Rowlands A, Charlesworth S, Davies A, Hoppitt T. Prediction of DXA-determined whole body fat from skinfolds: Importance of including skinfolds from the thigh and calf in young, healthy men and women. Eur J Clin Nutr. 2005;59(5):695-702. http://dx.doi.org/10.1038/sj.ejcn.1602131 PMid:15798775.
  » http://dx.doi.org/10.1038/sj.ejcn.1602131
• ³⁰
  Eston R. Prediction of body fat from skinfolds: the importance of including sites from the lower limb. J Sports Sci. 2003;21(5):369-70. http://dx.doi.org/10.1080/0264041031000071236 PMid:12800858.
  » http://dx.doi.org/10.1080/0264041031000071236

Publication Dates

• Publication in this collection
  28 Oct 2022
• Date of issue
  2022

History

• Received
  28 Mar 2022
• Accepted
  09 Sept 2022