Page 1 of 16
European Journal of Applied Sciences – Vol. 11, No. 1
Publication Date: January 25, 2023
DOI:10.14738/aivp.111.13890. Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility
(RR) of Bioelectric Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol -
11(1). 303-318.
Services for Science and Education – United Kingdom
Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass
Cristina Malzoni Ferreira Mangia
Pediatric Critical Care Division, Escola Paulista de Medicina,
Universidade Federal de São Paulo, Brazil.
Alexandre Carneluti
Faculdade de Medicina, FMABC, Brazil.
Benjamin Israel Kopelman
Pediatric Department, Escola Paulista de Medicina,
Universidade Federal de São Paulo, Brazil
Werther Brunow De-Carvalho
Pediatric Critical Care Division, Universidade São Paulo
Maria Cristina Andrade
Pediatric Nephrology Division, Escola Paulista de Medicina,
Universidade Federal de São Paulo, Brazil
ABSTRACT
Background: Bioelectrical analysis measures two bioelectrical vectors: resistance
(R) and reactance (Xc). Resistance is the pure opposition of a biological conductor
to the flow of an alternating current through the intra and extra-cellular ionic
solution and it is inversely related to the dynamics of body fluids and body
composition. Objective: The purpose of this study was to determine the reference
values of the indexes bioelectrical impedance (BI) for children of normal body mass
index in southeastern Brazil of middle-income country. Methods: Two hundred
eighty-one children with normal body mass index were included in the study (135
female and 146 male), aged 4 to 129 months, selected from federal public urban
school in São Paulo, São Paulo, Brazil, where bioelectrical impedance values
resistance (R) and reactance (Xc) values were measured in order to establish
reference values of these parameters. Results: The anthropometric variables, body
mass index, z-scores and bioelectrical impedance parameters were evaluated. For
both genders, the mean and standard deviation of anthropometric variables were
age (months): 73.42 + 34.65; weight (kg): 23.5 + 9.46; height (m): 1.16+0.22; BMI
(kg/m2): 16.65+1,75; Xc (ohms): 63.92+9.6; R (ohms): 749+75.26. For analysis, the
children were stratified into three groups for each gender, being divided by ages: 4
to 23 months; 24 to 71 months and 72 to 129 months. Linear regression analysis
showed R had a significant progressive decrease with age (p=0.0003) while Xc had
a progressive increase (p=0.0065) with age increase. We analyzed by multiple
regression the associations between R and Xc with anthropometric variables by age
group to establish the reference values, confidence intervals and the tolerance
limits for a new individual observation. Test-Retest Repeatability between three
Page 2 of 16
304
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
repeat and consecutive measurements was considered excellent. Intraclass
correlation coefficient and Bland -Altman reproducibility was for reactance 0.755,
Resistance 0.98 and phase angle was 0.93. Conclusion: The BI reference values were
established, in a field where there is a relative lack of publications, and we collected
relevant information about resistance and reactance in a population of middle
income setting that could be used in epidemiologic studies and could be used
reference value in children with altered body composition.
INTRODUCTION
Bioelectrical impedance analysis (BIA) is a fast, and inexpensive method that has been widely
applied to evaluate the body composition for over thirty years [1,2,3,4].
Impedance (Z) is an electrical property of ionic conduction measurable through soft tissue
(except fat and bone) and measures two bioelectrical vectors: resistance (R) and reactance (Xc).
The resistance vector is the opposition of a conductor to the flow of alternating electric current
through intracellular and extracellular ionic solutions and represents the real part of the
impedance. In a biological conductor, current is mainly carried by ions and aqueous solutions.
The reactance vector is the capacitance produced by tissue interfaces and cell membranes and
represents the imaginary part of Z (PICCOLI et al, 1998). It is reciprocal to electrical
capacitance, that is, it is the voltage stored by a biological capacitor for a brief period (LUKASKI
et al, 1996).
When an electric current is applied, when encountering a capacitive element, a phase difference
is created between the current and the applied voltage, determined by the reactive component
of R, which is represented geometrically by the phase angle. (BAUMGARTNER et al, 1988;
PICCOLI et al, 1998). The phase angle (in degrees) is the angle that the impedance vector makes
with its resistive component and is calculated as the arctangent of the ratio of reactance to
resistance.
The use of bioelectrical impedance to estimate body composition variables is based on the
hypothesis that fat-free tissues are good conductors at fixed frequencies and poor conductors
under alternating current, when compared to adipose tissue.
The body bioelectrical impedance technique is useful in the analysis of body composition, as it
allows health professionals to manage and prevent nutritional problems. Additionally, the
growing interest in the study of body composition and its variations as a method of assess in
nutritional status grows over the years as well as recognition of its importance for the
assessment of healthy and sick individuals [7,8].
BI has a hypothetical inverse relationship to the body’s volume and can be used in regression
prediction models to estimate total body water (TBW). It is based on a bi-compartmental
model, which divides the body into lean mass (LM) -high conductivity, a fact that reduces
resistance (R) and fat mass (FM) – low conductivity that increases body resistance (R)
Bioimpedance vectors (resistance and reactance) are rarely referred to in the literature as
vectors that reflect the dynamics of body fluids and the electrical properties of tissues. How
impedance vectors behave in individuals with body composition distortions, such as critically
ill patients, is a challenge that has not yet been fully clarified.
Page 3 of 16
305
Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol - 11(1). 303-318.
URL: http://dx.doi.org/10.14738/aivp.111.13890
However, there has been a growing interest in the use of these body bioimpedance vectors,
reactance (Xc) and resistance (R), as a way to assess hydration (resistance) and cell
permeability disorders (reactance) [5]
The monitoring of primary indices of bioimpedance, resistance (R) and reactance (Xc) has great
applicability in the scenario of critically ill patients, as it can be considered a non-invasive way
of predicting the patient's prognosis, acute malnutrition, assessment of fluid responsiveness,
cumulative fluid balance considering Xc can reflect the from a biophysical point of view changes
in the permeability of the endothelium and cell membrane. [6-10]
There is a relative lack of publications in the field of bioelectrical parameters reference values
on specific population such as low- and middle-income countries. For this reason, few studies
are reported to accurately assess nutritional individual deviations in relation to these
population mean and to analyze the role of bioelectrical parameters on various outcomes in the
clinical setting and epidemiological studies [11,12]
The purpose of this study was to evaluate the repeatability and reproducibility of BIA in
children with normal body mass index in southeastern Brazil.
METHODS
Data were collected in healthy children aged 4months to 129 months at a federal elementary
school in São Paulo city, Brazil. The children belonged to families that have the socio-economic
status of the majority of the Brazilian population, being in the middle-income population of
Brazil. The protocol was approved by the committee of ethics on research and the school’s
authorities.
Study Population
Three hundred, twenty-seven children of both genders were recruited after an interviewing
their parents and obtaining a signed written informed consent. The admission criteria for this
study were: a) z- score between –2 and +2, b) fasting state > 3 hours and c) no vigorous physical
activity in the 24 hours prior to the tests. The exclusion criteria were a) undernutrition [z-score
<-2], b) obesity [z-score > +2], c) acutely ill children, and d) those who were under medications.
Anthropometric Measurement
The anthropometric measurements were obtained by the principal investigator who was
previously trained to perform the measurements. The anthropometric measurement
procedures were undertaken in strict accordance with the methodology described in
previously published papers. [8,9,10]
The body weight was measured to a precision of 0.1 Kg with an electronic beam-balance in
children over 23.9 months of age. In children under 23.9 months, the body weight was
measured to a precision of 0.01 kg using an electronic scale. The body-height was measured by
a stadiometer to a precision of 0.1 cm for all age groups. The children were measured without
shoes and wearing underwear. The age, body weight and height were used to calculate the z- score. We used the relationship weight-for-height (W/H index) for the nutritional assessment
of the children over 23.9 months of age and for children under 23.9 months, the weight-for-age
Page 4 of 16
306
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
(W/A index) and weight-for-height (W/H index). The values obtained were compared to
standard reference values.
We used version 1.02 of the ANTHRO program from the Nutrition Division of the Disease
Control Center (CDC). To calculate the z score, comparisons were made between the z scores
obtained with the curves of the National Center for Health Statistics (NCHS), using cutoff values
to define the nutritional condition ± 2 z scores. We determined the body mass indices (BMI) -
weight (kg) divided by the square of height in meters - for each child, which were also compared
with the NCHS values. Thus, only children with a Z score and BMI within the normal values
established by the NCHS were included in the main study [13]
Bioimpedance Measurements
Whole-body electrical resistance and reactance were measured with a bioimpedance analyser
(Biodynamics model 310; Biodynamics Corporation, Seattle, WA) of alternate current at 800
μA and 50 kHz in tetrapolar arrangement.
Whole-body electrical resistance and reactance were measured with a bioelectrical impedance
analyzer that measure resistance and reactance independently and separately. (Biodynamics
model 310; Biodynamics Corporation, Seattle, WA) of alternate current at 800 μA and 50 kHz
in tetrapolar arrangement. Oil was removed from the skin by cleaning it with alcohol. No direct
contact was made with the child’s skin during measurements, and the children were calm and
relaxed [14,15,16]. For children under 18 months of age (where cooperation was more
difficult), we made a cylindrical non-conducting polyethylene, a type of non-toxic, lightweight,
flexible, and waterproof plastic with the objective of positioning the children correctly, i.e., in
dorsal decubitus with arms and legs separated and in abduction at 30 degrees from the trunk.
That frame was not used with older children, and the supine positioning was maintained. We
positioned the electrodes in pairs on the right side of the body in the following anatomical
positions: 1- Right hand: The current injector electrode was positioned in the middle of the
dorsal surfaces of the hand proximal to the third phalangeal-metacarpal joint. The detector
electrode was placed 4 cm below the wrist (group 1) or medially between the distal bony
prominences of the radius and ulna (group 2 and 3);2- Right foot: the current injector electrode
was positioned in the middle of the dorsal surfaces of the foot to the third metatarsal- phalangeal joint. The detector electrode was placed 4 cm on the ankle (group 1) or medially
between the medial and lateral malleoli at the ankle (group 2 and 3). Before each test, the
master power switch of analyzer was turned off and on. After pressing the on key, the analyzer
performs self-test to check the internal calibration accordance with the recommendation of the
manufacturer.
Repeatability and Reproducibility of Measurements
To preliminarily determine the repeatability and reproducibility of measurements, we
performed 3 consecutive measurements of R and Xc in all children (n = 280). The equipment
was turned off and after pressing the on key waiting for the self-test and reprogramming the
data for a new reading of R and Xc. This procedure was performed sequentially 3 times, and,
during the reprogramming intervals, without the child's movement. In 97 healthy children age
from 72 months to 123 months (n=97) repeatability and reproducibility measurements were
collected on two consecutive days, also.
Page 5 of 16
307
Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol - 11(1). 303-318.
URL: http://dx.doi.org/10.14738/aivp.111.13890
Statistical Analysis
Descriptive analysis was expressed as mean, standard deviation and 95% confidence intervals
(CI). The inferential statistical analyses were performed using GCM and REG procedures of the
statistical software package SAS (Version6.0). Bivariate correlations and stepwise maximum
R2 was performed by multiple linear regression analyses to evaluate the strength and
variability of R and Xc with weight(W), height (H) by age and gender. A p value <0.05 was
considered statistically significant. [18] Multiple regression models and Pearson’s correlation
coefficient were used to assess the strength and relationship between R and Xc with weight
(W), height (H), age and sex. The fitted models were different from each other, according to the
sex and age group. Multiple regression models were then fitted for R and Xc as functions of
weight and height for each sex, considering age groups adapted from the Committee on
Nutrition Advisory to CDC and Waterloo et al [10,17]. Residual analysis was developed to
evaluate the adequacy of the fitted models. The fitted regression models, for each sex and age
group, according to the models:
R= a0 + a1*H + a2*W+ e and Xc = b0 + b1*H + b2*W + e
They were used to predict the average R, average Xc and confidence intervals. The statistical
analysis was accomplished by the SAS system V 6.0(SAS) Institute Inc, 1989. Repeatability and
reproducibility were analyzed using IBM SPSS version 20.0 for Windows (IBM Corp. Armonk,
NY, USA) and EXCEL for Windows version 10.0. Repeatability refers to the variation in repeated
measurements made on the same subject by the same operator under identical conditions
[19,20]. Reproducibility is the additional variability introduced when measurements are made
on same subjects but under different days. In the study was considered two consecutive days.
Repeatability and reproducibility studies can be assessed using reliability (inherent variability
in the true difference between measurements) and /or agreement (the quantified variation
between measurements). The analysis of reliability in this study was determined using
intraclass correlation coefficient (ICC) and agreement using Bland-Altman analysis. ICC values
≤ 0.5, 0.5-0.75, 0.75-0.9 and ≥ 0.9, were indicative of poor, moderate, good, and excellent
reliability. [19,20]
RESULTS
The variables were collected in 327 children during the same period of year. Children excluded
were: six were undernourished, fourteen obese and twenty-six had other exclusion criteria. The
final study population consisted of 281 healthy children with normal body mass index. Children
were previously stratified by three age-group due to low number: a) 4 months to 23 months
(group 1), b)24 months to 71 months (group 2), and c) 72 to129 months (group 3). The subject
characteristics are presented in Table 1. Linear regression analysis was performed to evaluate
if age-group stratification was appropriate to study the variability of the resistance and
reactance in relation to anthropometric variables. Figs. 1 and 2 shows that the stratification
was appropriate. In both graphics there are two inflection points, the first point at 23 months
and the second at 71 months. These two points were interpreted as indicative of the resistance
and reactance variations imposed by growth and development. The straight lines were
significantly different for the resistance (p=0.0003) and reactance (p=0.0065).
Page 6 of 16
308
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
Correlation between Bioelectrical Impedance Components and Anthropometric Variables
Multivariate regression models were used to analyze the correlations between resistance and
reactance with anthropometric variables. The purpose of these models was to establish
confidence intervals for R and Xc for normal children and tolerance intervals for a new
observation. Tables 2 and 3 lists the multivariable regression equations. Due to the small
number of children for each gender in age group 1, one model was adjusted for both genders.
Pearson’s correlation coefficient between anthropometric variables and bioimpedance vector
components are described in Table 4. Weight and height were negatively correlated with
resistance in all age groups. The reactance was positively correlated with weight and height in
females in all age groups. Boys and girls did not differ in age, body weight and body height but
girls had a higher resistance than boys in groups 2 and 3. This difference in body resistance
between boys and girls was not found in the infants (group 1). Reactance increases with age,
having few variations between genders (Table 5). The regression models were used to estimate
Rand Xc mean and 90% to 99% confidence intervals (CI) for age group and gender. In addition,
we used the regression models to estimate the values expected of the impedance vectors and
the tolerance limits 90% to 99% for a new observation.
For analysis of repeatability and reproducibility, 97 children were studied belonging to the
group between 72 to 129 months (group 3). Measurements were taken under ideal conditions
of collaboration according to the technique previously described. The analyzes obtained on
days 1 and 2 were carried out under good conditions of the environment and temperature.
Test-Retest Repeatability between three repeat and consecutive measurements was
considered excellent. For Reactance test-retest was 0,949. For resistance for test re-test was
0,99. For phase angle test-retest repeatability was 0.98.
Intraclass correlation coefficient and Bland -Altman reproducibility between day 1 and Day 2
for reactance was 0.755. Resistance between day 1 and day 2 was 0.98 and phase angle between
day 1 and day 2 was 0.93. The Bland-Altman mean of differences plots are show in figures 4-6.
Simple regression analysis was performed to analyze whether the measurements obtained
agreement and whether there is potential bias in the analysis, that is, whether there is a
tendency, or whether the values tend to be above or below the mean differences. The value of
the mean significance analysis did not show a systematic trend, so there is no proportion bias
Simple regression analysis was performed to analyze whether the measurements obtained are
concordant and whether there is potential bias in the analysis. In other words, if the values tend
to be above or below the mean of differences. The value of the mean significance analysis for
reactance (p =0.785), resistance (p= 0.715) and phase angle (p>0.7) did not show a systematic
trend, so there is no proportion bias
DISCUSSION
Bioelectrical impedance analysis (BIA) is considered a good method for estimating body
composition in the epidemiologic studies and at the bedside. It is safe, non-invasive, reliable,
rapid, inexpensive, portable, and it allows to repeated measures could be taken quickly [5]. We
studied separately R and Xc components grouping by age depending on the sample size and
gender. The three age groups adopted were based on Waterloo et al [10] stratification sampling
criteria, that clustered the children into relatively homogenous subgroups by age.
Page 7 of 16
309
Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol - 11(1). 303-318.
URL: http://dx.doi.org/10.14738/aivp.111.13890
In addition, the skin electrodes were placed on anatomical position and those electrodes had
their patches width reduced in young children because there is a minimal distance required to
avoid interactions between electrodes [16].These criteria adopted by us were similar toother
studies in children where: 1) similar groups of children were considered; 2) skin electrodes
were placed in accordance with the child’s age;3) the children were separated in age–groups;
4) Xc vector component was not neglected; and 5)age-related variability was found in these
studies[1,14,15].The measures demonstrated that resistance measurements were
substantially higher in all age groups than those reported for adults. In healthy American adults,
that means range from432 to 485 ohms for men and 551 to 587 ohms for women and in healthy
Brazilian adults 552 +100 ohms in both genders. Our study demonstrated that resistance values
in young children were higher than older children, and these results are similar to those in the
previous studies [3,5,21]. We observed variability of the resistance and reactance parameters
with growth in our study, reinforcing the importance of the reference values of R and Xc by age
or age-group and gender in healthy populations of children. The variability of parameters might
be reflecting changes during growth as does intra and extra-cellular fluid distribution, cell
growth and changes in body mineral and electrolytic content, therefore, reflecting the
variability of fluids and body composition in children [22]. The study showed that resistance
decreases with age, which might be because the muscular mass of the limbs increases with
growth. These observations reinforce the concept whereby in the infants and toddlers, arms
and legs represent body area with small diameter and length, therefore the resistance is high.
With growth, there is an increase of the diameter and length of the limbs, and R decreases due
to an increase in the cross-sectional area of the extremities. These observations are according
to simple body-composition models where the appendicular skeletal muscles are the primary
electrical conductor [23,24,25,26]. We observed differences in the reactance among the three
study-groups. This might be due to the differences of capacitance properties of the tissue
interfaces and cell membranes. Theoretically, Xc variation among healthy individuals could be
due to differences in the capacitive behavior of the tissues associated with variability of the cell
size, membrane permeability or intracellular composition during growth [25,26].An increase
of interstitial fat (anhydrous, meaning that fat is hydrophobic) during maturation reduces both
the tissue interface permeability and cell membrane interface permeability, producing an
increase in reactance in a critical fixed frequency [27]. The variability of R and Xc might be
explained also by variations that include more and less conductive matter, body temperature,
tissue composition, fluid distribution, ionic concentration, nature of fat, as well anisotropic
effects of muscle fibers. These physiological and structural as well as technical factors affect the
measurement of both bioelectrical impedance vector components, R and Xc [3,4,6]. The
limitation of this study is that the sample cannot be considered representative of all millions of
Brazilian children because there is difference in the nutritional status among specific Brazilian
regions depending on the socioeconomic levels of population in each region of Brazil. In order
to minimize population bias, the epidemiologic procedure performed in this study consisted of
selecting a school with children from families with middle income resources. Our study is the
first and only one study already realized in Brazil to establish bioelectrical impedance vectors
reference values in children for several age groups and gender.
CONCLUSION
In conclusion, we established the normative bivariate 90% to 99% confidence intervals for the
mean impedance indexes by group and gender and the bivariate predictive values 90% to
99%tolerance limits for new individual measurements of the resistance and reactance in
Page 8 of 16
310
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
healthy Brazilian children. Further, changes in resistance and reactance with age are well- established. Our findings add substantial information in a field with relative lack of
publications.
Contributions
This work was carried out in collaboration among all authors. CMFM designed the study,
performed the statistical analysis, wrote the protocol, and wrote the first and final draft of the
manuscript. AC Carneluti A managed the literature searches, performed critical analysis and
edited the final draft of the manuscript. CMFM and MCA wrote the first draft of manuscript.
MCA performed critical analysis of manuscript and reviewed the final draft. BIK and WBC
contributed with ideas, methodology and All authors read and approved the final manuscript
Consent
As per international standard, parental written consent has been collected and preserved by
the author(s).
Ethical
Approval The protocol was approved by the committee of ethics on research of the
Universidade Federal de São Paulo and the school’s authorities.
Acknowledgements
Dr Cristina Mangia MD, PhD would kindly like to thank the Professor João Augusto Mattar MD,
MSc, PhD, FCCM (in memoriam) for his indispensable and valuable teaching on the
bioimpedance in normal subjects and critically ill patients concepts and ever-present support
for my studies on bioelectrical impedance in children. In addition, I would like to thank Gabriela
Stangenhaus, PhD from Statistika Consultoria Inc, Professor Werther Brunow de Carvalho MD,
PhD and Professor Benjamin Israel Kopelman, MD, PhD.].
References
1. Davies PSW. Body composition assessment. Arch Dis Child. 1993; 69:337-338.
2. Hammond J, Rona RJ, Chinn S. Estimation in community surveys of total body fat of children using
bioelectrical impedance or skinfold thickness measurements. Eur J Clin Nutr. 1994;48: 164-171.
3. Lukaski HC, Bolonchuck WW, Hall CB, Siders WA. Validation of tetrapolar bioelectrical impedance method
to assess human body composition. J Appl Physiol. 1986; 60:1327-1332.
4. Lukaski HC. Biological indexes considered in the derivation of the bioelectrical impedance analysis. Am J
Clin Nutr. 1996; 64(Suppl):397-404.
5. Mattar JA and the Brazilian Group for Bioimpedance Study. Application of total body bioimpedance to the
critically ill patients. New Horiz. 1996; 4:493-502.
6. Barbosa-Silva MCG, Barros AJD, Wang J, Heymsfield SB, Pierson JrRN. Bioelectrical impedance analysis:
Population reference values for phase angle by age and sex. Am J Clin Nutr. 2005; 82:49-52.
7. Walker SP, Granthan–McGregor S, Powell C, Fletcher P, Himes JH. Bioelectric impedance, anthropometry
and body composition in stunted and non-stunted children. Eur J Clin Nutr. 1990;44: 763-768.
Page 9 of 16
311
Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol - 11(1). 303-318.
URL: http://dx.doi.org/10.14738/aivp.111.13890
8. Fagundes U, Oliva CAG, Fagundes-Neto U. Evaluation of the nutritional status of Indian children from Alto
Xingu. Brazil. J Pediatr (Rio J). 2002;78(5):383-8.
9. Organización Mundial de la Salud. Medición del Cambio del Estado Nutricional. Directrices para Evaluar el
Efecto Nutricional de Programas de Alimentácion Suplementaria Destinados a Grupos Vulnerables, OMS,
Ginebra. 1983; 1-105.
10. Waterloo JC, Buzina R, Keller W, Lane JM, Nichaman MZ, Tanner JM. The presentation and use of height
and weight data for comparing the nutritional status of group of children under age of 10 years. Bull World
Health Organization. 1977;55: 489-498.
11. De Palo T, Messina G, Edefoni A, Perfumo F, Pisanello L, Peruzzi L, Di Iorio B, Mignozzi M, Vienna A, Conti G,
Penza R, Piccoli A. Normal values of the bioelectrical impedance vector in children and puberty. Nutrition.
2000; 16:417- 424.
12. Piccoli A, Fanos V, Peruzzi L, Schena S et al. Reference values of the bioelectrical impedance vector in
neonates in the first week after birth. Nutrition. 2002; 18:383-387.
13. National Center for Health Statistics, Centers for Disease Control and Prevention. Growth charts. Sept 9,
2010. [Last cited Aug 24,2020]. Available from: https://www.cdc.gov/growthcharts/who_charts.htm.
14. Lukaski HC. Requirements for clinical use of bioelectrical impedance analysis (BIA). Ann New York Acad
Sci. 1999; 873:72-76.
15. Baumgartner RN, Chumlea WMC, Roche AF. Associations between bioelectric impedance and
anthropometric variables. Hum Biol. 1987; 59:235-244.
16. Siervogel RM, Woynarowska B, Chumlea C, Guo S and Roche AF. Bioelectrical Impedance Measures of Body
Composition: their relationship with level of blood pressure in young adults. Human Biol. 1987; 59:245-
255
17. US Food and Nutrition Board, National Academy os Sciences-National Research Council. Committee on
Nutrition Advisory to CDC.Comparision of body weights and lenghts or heights of group of children.
Nutrion Rev 1974;32(9): 284-8.
18. SAS Institute Inc. SAS/STAT User’s Guide. Version 6. 4th ed. Vol 2. Cary, NC: SAS Institute Inc., 1989. pp.846.
19. Bartlett JW and Frost C. Reliability, repeatability and reproducibility: analysis of measurement errors in
continuous variables. Ultrasound Obsttet Gynecol 2008; 31:466-473
20. Bai M, Susic D, O'Sullivan AJ, Henry A. Reproducibility of Bioelectrical Impedance Analysis in Pregnancy
and the Association of Body Composition with the Risk of Gestational Diabetes: A Substudy of MUMS
Cohort. J Obes. 2020; 2020:3128767
21. Nagano M, Suita S, Yamanouchi T. The validity of bioelectrical impedance phase angle for nutritional
assessment in children.J Pediatr Surg 2000;35:1035-39
22. Fuller NJ, Ellia M. Potential use of bioelectrical impedance of the “whole body” and of body segments for
the assessment of body composition: Comparison with densitometry and anthropometry. Eur J Clin Nutr.
1989;43: 779-791.
23. Deuremberg P, Wetstarte JA, Hautvast JGAJ, van der Kooy K. Is the bioelectrical impedance method valid?
Am J Clin Nutr. 1991; 53:179-180.
Page 10 of 16
312
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
24. Organ LW, Bradhan GB, Gore DT, Lozier SL. Segmental bioelectrical impedance analysis: Theory and
application of a new technique. J Appl Physiol. 1994; 77:88-112.
25. Heymsfield SB, Wang ZM, Visser M et al. Techniques used in the measurement of body composition: An
overview with emphasis on bioelectrical impedance analysis. Am J Clin Nutr 1996;64(suppl): 478S-84S.
26. Gregory JW, Greene SA, Scrimgeour CM, Rennie MJ. Body water measurement in growth disorders: A
comparison of bioelectrical impedance and skinfold thickness techniques with isotope dilution. Arch Dis
Child. 1991; 66:220-222.
27. Baumgartener RN, Chumlea WC Roche AF. Bioelectric impedance phase angle and body composition. Am J
Clin Nutr 1988; 48:16-23
Page 11 of 16
313
Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol - 11(1). 303-318.
URL: http://dx.doi.org/10.14738/aivp.111.13890
Table 1. Demographic characteristics of all children
Group Age (months) Height (cm) Weight (Kg) BMI R (Ohm) Xc (Ohm)
Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD
1 Both
gender*
10.3 4.6 71.3 6.3 8.7 1.9 8.71 1.85 801 96 51 8
2 M
(n=39)
51.7 13.6 106.6 10.9 18.4 4.0 16.02 1.35 750 64 63 9
F (n=37) 56 13.1 106.5 9.2 18.4 3.1 16.10 1.17 765 64 65 8
3 M
(n=94)
97 15.8 130.3 9.2 28.8 6.4 16.79 1.86 720 60 67 8
F (n=73) 98.8 15.8 131.5 10.6 29.8 7.8 16.92 2.12 750 75 67 8
Total M
(n=146)
76.48 31.96 1.19 0.20 24.35 8.45 16.67 1.72 729.53 62.98 64.21 9.28
F
(n=135)
70.13 37.18 1.12 0.28 23.95 14.79 16.63 1.80 770.56 81.78 63.61 9.98
Total n=281 73.42 34.65 1.16 0.23 23.5 9.46 16.65 1.75 749 75.26 63.92 9.6
Group 1 = 4 months < age < 23 months; Group 2 = 24 months < age < 71 months; Group 3 = 72 months < age < 123
months. Height, cm; Weight, Kg; R, resistance in ohm (W); Xc, reactance in ohm (W); SD, standard deviation
Fig. 1. Relationship between measured (white) and predicted (black) resistance
values according to age. The regression line predicted for the three age groups
studied were significantly different (p =0.003)
Page 12 of 16
314
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
Fig. 2. Relationship between measured (white) and predicted (black) reactance
values according to age. The regression line predicted for the three age groups
studied were significantly different (p = 0.0065)
Table 2. Prediction of the resistance according to age, body
weight,body height for three study groups by age and genders
Group/sex N a0 a1 a2 r
2 SEE p
G1 Both 38 600.44a 10.86 ns -65.89c 0.41 75.39 0.0001
G 2 Male 39 636.82d 3.22 ns -12.63e 1.14 60.75 0.07
Female 37 608.83d 4.00 ns -14.73ns 0.11 62.71 0.14
G 3 Male 94 467.48d 3.96c -9.14d 0.29 51.37 0.0001
Female 73 268.46f 6.50d -12.54d 0.39 59.43 0.0001
R=a0 +a1*H+a2*W; R = resistance (ohm); H= Height (cm); W = weight (Kg); Group 1 = 4 months < age < 23 months;
Group 2 = 24 months < age < 71 months; Group 3 = 72 months < age < 123 months. ap<0.02; cp<0.002; dp<0.0001;
ep<0.05; fp<0.006. NS = non- significant
Page 13 of 16
315
Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol - 11(1). 303-318.
URL: http://dx.doi.org/10.14738/aivp.111.13890
Table 3. Prediction of reactance according to age, body
weight, body height for three age study-groups and gender
Group Gender N b0 b1 b2 r
2 SEE p
1 Both 38 62.92a -0.24ns 0.60ns 0.008 7.43 0.87
2 Male 39 27.17ns 0.52ns -1.16ns 0.085 8.21 0.21
Female 37 13.15a 0.71 a -1.27ns 0.204 6.87 0.02
3 Male 94 50.73a 0.21ns -0.42ns 0.031 7.91 0.23
Female 73 44.12a 0.33ns -0.73a 0.147 7.16 0.0038
Xc= b0 + b1*H+b2*W; Xc = reactance (ohm); H= Height (cm); W = weight (Kg); Group 1 = 4 months < age < 23
months; Group 2 = 24 months < age < 71 months; Group 3 = 72 months < age < 123 months. ap<0.02; NS = non- significant
Table 4. Correlation of resistance and reactance with
body weight, body height for age study-groups and gender
Group 1 Group 2 Group 3
Both genders Male Female Male Female
Resistance (Ohm)
Height (cm) -0.48a -0.22b -0.07b -0.22b -0.24c
Weight (Kg) -0.60a -0.30e -0.20b -0.44a -0.44a
Reactance (Ohm)
Height (cm) -0.07b 0.08b 0.46d -0.04b -0.04b
Weight (Kg) -0.05b -0.00b 0.32c -0.12b -0.12b
ap<0.001; b p= NS; c p<0.05; dp<0.005; ep<0.06.); Group 1 = 4 months < age < 23 months; Group 2 = 24 months <
age < 71 months; Group 3 = 72 months < age < 123 months
Table 5. Estimates and 95% tolerance intervals for
resistance and reactance for three age-study groups
Group Gender Mean Lower 95% TL Upper 95% TL
Resistance (Ohm)
1 Both 880 707 1053
2 Female 765 744 787
Male 748 728 769
3 Female 749 732 767
Male 721 708 733
Reactance (Ohm)
1 Both 51 48 54
2 Female 65 63 68
Male 63 60 66
3 Female 67 65 69
Male 67 65 68
Group 1 = 4 months < age < 23 months; Group 2 = 24 months < age < 71 months; Group 3 = 72 months < age < 123
months.
TL= tolerance limits; Lower 95% TL= lower limit; Upper 95% TL= TL upper limit. Mean
estimated value was calculated using the regression models presented in Tables 2 and 3.
Tolerance limit for the estimated mean were calculated with the expression:
^
__ 1
0 0 0 ( ) T T
y zs x x X X -
±
Page 14 of 16
316
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
Where x0 = (W, H) in the estimated regression equation and X the model matrix. Z is the
corresponding normal distribution percentile and s the standard error estimate.
Table 6. Descriptive Statistics variables used to Repeatibility and Reprodutibility tests*
Statistic Bias Std. Error 95% Confidence Interval
Lower Upper
AGE N 97 0 0 97 97
Minimum 72
Maximum 129
Mean 96 ,02 1,53 93,25 99,41
Std. Deviation 14,94 -,095 ,825 13,144 16,483
WEIGHT N 97 0 0 97 97
Minimum 18
Maximum 50
Mean 28,94 ,02 ,66 27,65 30,26
Std. Deviation 6,64 -,041 ,515 5,533 7,597
HEIGHT N 97 0 0 97 97
Minimum 112
Maximum 156
Mean 130,47 ,00 1,02 128,50 132,50
Std. Deviation 10,067 -,066 ,615 8,774 11,157
Reactance N 97 0 0 97 97
Minimum 49
Maximum 89
Mean 66 ,008725 ,830427 64,893907 68,065117
Std. Deviation 8,30 -
,0230436
,5734588 7,1403391 9,4329973
Resistance N 97 0 0 97 97
Minimum 559
Maximum 861
Mean 728 -,312203 5,834461 716,942716 739,973619
Std. Deviation 60 -
,6537916
4,1840605 52,0357470 68,6392850
Phase
Angle
N 97 0 0 97 97
Minimum 3,99
Maximum 6,07
Mean 5,211 ,002656 ,058331 5,107796 5,336438
Std. Deviation ,553 -
,0019695
,0362502 ,4855786 ,6255360
Valid N N 97 0 0 97 97
Age group from 72 months to 123 months (n=97)
Page 15 of 16
317
Mangia, C. M. F., Carneluti, A., Kopelman, B. I., De-Carvalho, W. B., & Andrade, M. C. (2023). Repeatibility and Reproducibility (RR) of Bioelectric
Impedance Vectors in Brazilian Children with Normal Body Mass. European Journal of Applied Sciences, Vol - 11(1). 303-318.
URL: http://dx.doi.org/10.14738/aivp.111.13890
Fig3. Bland-Altman plots of differences of Reactance against mean of Reactance
Fig4. Bland-Altman plots of differences of Resistance against mean of Resistance
Page 16 of 16
318
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 1, January-2023
Services for Science and Education – United Kingdom
Fig 5. Bland-Altman plots of differences of Phase Angle against mean of Phase Angle