Research ArticleValidity of Four Commercial Bioelectrical Impedance Scales inMeasuring Body Fat among Chinese Children and Adolescents

Lin Wang1 and Stanley Sai-chuen Hui2

1Center for Sports Medicine and Rehabilitation Shanghai University of Sport Shanghai 200438 China2Department of Sports Science and Physical Education The Chinese University of Hong Kong Hong Kong

Correspondence should be addressed to Stanley Sai-chuen Hui hui2162cuhkeduhk

Received 26 January 2015 Accepted 26 May 2015

Academic Editor Abel Romero-Corral

Copyright copy 2015 L Wang and S S-c Hui This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

The aim of the study is to examine the validity in predicting body fat percentage (BF) of different bioelectrical impedance (BIA)devices among Chinese children and adolescents A total of 255 Chinese children and adolescents aged 9ndash19 years old participatedin the study BF was assessed by BIA scales namely Biodynamics-310 (Model A) Tanita TBF-543 (Model B) Tanita BC-545(Model C) and InBody 520 (Model D) Dual-energy X-ray absorptiometry (DXA) was used as the criterion measurement Linrsquosconcordance correlation coefficients of estimated BF between Model A Model B Model C and DXA showed poor agreementsfor both genders Moderate agreements for BF were found between DXA and Model D measurements In boys differences inBF were found between DXA and Model B and Model C No significant BF differences were found between Model A ModelD and DXA However the two BIA analyzers showed a significant positive correlation between the bias and average BF betweenBIA and DXA In girls differences in BF were observed between Model B Model C Model D and DXA Model A and DXAshowed no significant differences of BF however the bias and the average BF between the BIA and DXA had a significantpositive correlation Using embedded equations in BIA devices should be validated in assessing the BF of Chinese children andadolescents

1 Introduction

The World Health Organization has acknowledged thatchildhood obesity is one of the most serious public healthchallenges of the 21st century [1] The International ObesityTask Force (IOTF) reported that 1 in 10 children aged 5ndash17 years is overweight and 30ndash45 million children aged 5ndash17 years (accounting for 2-3 of the global population ofchildren) are obese [2] The prevalence of overweight andobesity has continued to increase at alarming rates in Chinaparticularly among children and adolescents [3] From 1982to 2000 the prevalence of overweight and obesity amongchildren and adolescents increased from 14 to 59 for boysand from 14 to 45 for girls [4]

Obesity is generally defined as abnormal or excessive fataccumulation that presents health risks [1] Childhood obe-sity has significant adverse effects on health including car-diovascular disease type 2 diabetes hypertension stroke andcertain types of cancer [5] In addition to physiologic risks

psychological morbidity is likely to be the most widespreadhealth impact during childhood [6] The measurement ofbody fat is central to many aspects of childhood obesityresearch [7 8] It is important to have valid and reliable toolsto assess body fat Devices that accurately assess body fatpercentage (BF) can be used to determine an individualrsquosbody composition or weight loss over a certain period

Many laboratory methods have been developed to mea-sure body composition Such techniques include underwa-ter weighing air-displacement plethysmography deuteriumdilution dual-energy X-ray absorptiometry (DXA) comput-erized tomography (CT) and magnetic resonance imaging(MRI) [9] These methods have provided researchers withopportunities tomeasure human body fat precisely [10] DXAcan provide acceptable accuracy and reliability in measuringbody composition among children [11]This has been justifiedby its successful validation against the multicomponentmodel for children [12] therefore DXA is popularly usedas a method of measuring childrenrsquos body composition [13]

Hindawi Publishing CorporationBioMed Research InternationalVolume 2015 Article ID 614858 8 pageshttpdxdoiorg1011552015614858

2 BioMed Research International

However because of its inaccessibility the high cost ofequipment and exposure to low-dose radiation the use ofDXA is limited in the field setting Thus simple methodssuch as bioelectrical impedance analysis (BIA) and skinfoldthickness (SKF) measurement are used in field studies [9]

Compared with methods of measuring body compo-sition BIA is relatively simple quick and noninvasive Itgives reliable measurements of body composition with smallintra- and interobserver variability [14] BIA does not requireexposure to radioactivity or submersion in water which iswhy it is regarded as a practicalmeasure of body compositionespecially for field studies involving children Because of itsadvantages many BIA consumer devices are popular withhealth professionals and the general public for the assessmentand monitoring of body fat These devices are characterizedby differences in electrodes (number type and placement)electric current frequency and body position atmeasurement[15] Single-frequency BIA devices with four electrodes suchas a traditional hand-to-foot BIA machine using adheringelectrodes and a foot-to-foot BIA machine with four metalelectrodes have been used widely to assess body composition[15] Multifrequency hand-to-foot BIA machines with eightelectrodes have been developed for assessing segmental andtotal body compositions [16] Multifrequency BIA was sug-gested to be a more accurate method for estimating the bodycomposition of adults [17] Segmental multiple-frequencyBIA devices employ a current pathway that bypasses thearms legs and trunk using current different frequencies[18] The most appropriate combination of frequencies andmultivariate methods of using multi-frequency impedancevalues in estimating body composition are just now beingexplored therefore bioelectrical impedance spectroscopy(BIS) devices are designed to scan a wide range of frequen-cies this approach provides estimates of extracellular waterintracellular water and total body water [19]

Foot-to-foot BIA consumer devices are widely used toassess body composition among children and adolescentsSome studies have investigated the accuracy of single-frequency foot-to-foot BIA devices in assessing body fatContradictory results were found between foot-to-foot BIAand the reference methods in assessing body compositionamong children and adolescents with some studies showingan acceptable level of accuracy [20ndash23] and others reporting apoor level of accuracy [24ndash29] Althoughmultifrequency BIAwas suggested to be a more accurate method for estimatingadultsrsquo body composition [17] its accuracy in evaluating thebody composition of children and adolescents is doubtfulSome studies considered multifrequency BIA as an accurateway to measure body fat in children and adolescents [30 31]whereas others found poor accuracy betweenmultifrequencyBIA and criterion methods [10 32] These discrepanciesmay be associated with differences among BIA consumerdevices and ethnic differences in body composition Previousstudies showed that East Asian and South Asian childrenand adolescents tend to have more body fat than Caucasiansat an equivalent BMI [33ndash35] To date only two studieshave been conducted to investigate the accuracy of foot-to-foot BIA consumer devices for the measurement of bodyfat percentage in Chinese children and adolescents against

the DXA measurement [21 23] A good correlation of bodyfat percentage and acceptable agreement between BIA mea-surement and DXA measurement were identified Howeverthese studies were limited by a small sample size in a wideage range Only single-frequency foot-to-foot BIA deviceswere tested in these studiesThe accuracy of different portableBIA consumer devices in assessing body fat among Chinesechildren and adolescents remains questionable

To keep track of childhood obesity levels effectively it isnecessary to evaluate the obesity level of children in Chinausing objective and practical measurement methods BIAconsumer devices provide a cheap and easy way to measurethe body composition of Chinese children and adolescentsHowever no study has assessed the validity of differentportable BIA devices in relation to a preferred criterionmethod To achieve this the accuracy of BIA consumerdevices in predicting body fat among Chinese children andadolescents should be determined Therefore the aim of thepresent study is to examine the validity in measuring bodyfat

2 Material and Methods

21 Participants A total of 255 healthy Chinese children andadolescents aged 9ndash19 years were recruited from schools inShanghai China Of the total 127 were boys and 128 weregirls A stratified sampling method was used to recruit aheterogeneous sample covering a wide range of ages andbody compositions according to age and gender specificBMI distributions among Chinese children The stratifiedsampling method was used in our previous study Writteninformed consent was obtained from each participant and hisor her parents and the studywas approved by the Local EthicsCommittee

22 Anthropometrics Measurement Body weight and heightwere measured with minimal clothing and bare feet Bodyheight was measured to the nearest 05 cm using a fixedstadiometer (Holtain Crymych Dyfed United Kingdom)Body weight was measured to the nearest 01 kg using astandard scale (Tanita TBF-543 Tanita Tokyo Japan) Atrained investigator performed all the measurements

23 Measurement of Body Composition

231 DXA Measurement Dual-energy X-ray absorptiom-etry (GE Lunar Prodigy software version 1051006 GEHealthcare Madison USA) was used as the criterion mea-surement for body composition in terms of fat mass (FM)lean body mass (LTM) and bone mineral contents (BMC)Each participant was scanned in the supine position using X-ray at two energy sources (40 keV and 70 keV) in fastmode Aseries of transverse scans were done from head to toe at one-centimeter intervals The time spent was about five minutesdepending on the participantrsquos height A trained investigatorcompleted all the measurements of the DXA scan

232 BIA Measurements BIA measurements were under-taken at least two hours after breakfast and with an empty

BioMed Research International 3

bladder Girls having their menses avoided the tests Inthe current study body composition was assessed by fourBIA scales namely Biodynamics-310 (Model A BiodynamicsCorp Seattle USA) Tanita TBF-543 (Model B Tanita CorpTokyo Japan) Tanita BC-545 (Model C Tanita Corp TokyoJapan) and InBody 520 (Model D Biospace Co Ltd SeoulKorea) Manufacturersrsquo equations were used to predict BF(using all scales) and fat-freemass (usingModel A andModelD) Model A was a single-frequency traditional hand-to-footBIA machine using adhering electrodes The measurementwas carried out with the participants lying supine on a couchThe armswere separated from the trunk by about 30∘ the legswere abducted and separated by about 45∘ The skin of theright hand and foot was cleaned with an alcohol pad beforethe electrodes were placed Body composition was measuredat 50 kHz with a tetrapolar arrangement of standard elec-trodes (Red Dot 2330 3M Healthcare Saint Paul USA) Twoelectrodes were placed on the right ankle with one (the sourceelectrode) just proximal to the third metatarsophalangealjoint and the other (the sensing electrode) on the anteriorankle between the medial and the lateral malleolus Twoelectrodes were placed on the right wrist with one (the sourceelectrode) just proximal to the third metacarpophalangealjoint and the other (the sensing electrode) on the posteriorwrist between the styloid processes of the radius and theulna Model B was a bipolar single-frequency (50 kHz) foot-to-foot instrument Participants were instructed to standbarefoot with heel and forefoot placed on the four metalelectrodes Model C was a dual-frequency (50 kHz and625 kHz) andModel Dwas amultifrequency (5 kHz 50 kHzand 500 kHz) BIA device with eight electrodes in a tetrapolararrangementThemeasurements required the participants tostand barefoot on metal electrodes while grasping a pair ofelectrodes fixed on a handle with arms extended in frontof the chest All BIA measurements were completed by atrained investigator according to the device manufacturersrsquoinstructions

24 Statistical Analysis Descriptive statistics were reportedas means plusmn standard deviations (SD) An independent t-test was employed to determine the differences in the par-ticipantsrsquo physical characteristics DXA measurements andBIA measurements between boys and girls The 5 levelwas chosen for statistical significance Linrsquos concordancecorrelation coefficient (rc) was used to measure the bivariaterelationship of BF and FFM obtained from DXA withthose obtained from BIA measurements McBride suggeststhe following descriptive scale for values of the concordancecorrelation coefficient (for continuous variables) lt090 poor090ndash095 moderate 095ndash099 substantial and gt099 almostperfect [36] Differences in body composition between thoseassessed by BIA methods and DXAmeasures were examinedusing a paired t-test Bland-Altman analysis was applied todetermine the agreement between BIA and DXA measure-ments [37] The agreement between methods is representedby the bias estimated by the mean difference and the SD ofthe differences Therefore the 95 limits of agreement werecalculated as the mean difference plusmn196 SD of the differencesbetween methods The pure error in BF between each BIA

device with DXA measurement was assessed using the totalerror (TE) as follows

TE = radicsum(119884 minus 119884

1015840

)

2

119873

(1)

where 119884 = observed values 1198841015840 = predicted values and 119873 =the number of participants in the sample These statisticalanalyses were performed separately on boys and girls

3 Results

31 Characteristics of Participants A total of 255 (128 girlsand 127 boys) Chinese children and adolescents participatedin the study Table 1 presents their physical characteristics andbody compartments (FM LTM FFM BMC and BF) Nosignificant differences in mean age were observed betweengenders The mean body weight height and BMI of boyswere higher than those of girls The mean of the individualbody compartments was significantly different between boysand girls (Table 1) 803 of boys and 828 of girls hadentered puberty (pubertal stage ge2 for pubic hair or breasts(girls)genitalia (boys))

Table 2 shows Linrsquos concordance correlation coefficientsbetween parameters of body composition assessed by dif-ferent BIA devices and DXA methods In boys the FFMdetermined by all BIA devices showed substantial agreementswith that determined by DXA (rc values within 0951ndash0979)In girls the FFM determined by InBody 520 showed substan-tial agreement with that determined by DXA (rc = 0957)moderate agreements for FFM were found on Biodynamics-310 (rc = 0909) and Tanita BC-545 (rc = 0910) and pooragreement for FFMwas found onTanita TBF 543 (rc= 0897)Linrsquos concordance correlation coefficients of estimated BFbetween Biodynamics-310 Tanita TBF-543 Tanita BC-545and DXA showed poor agreements (rc values within 0747ndash0863) for both genders Moderate agreements for BF werefound between DXA and InBody 520 measurements (rc =0926 in boys and 0912 in girls)

In boys significant differences in BF were foundbetween DXA and Model B and Model C (paired t-test 119901 lt005) The mean bias of BF for Model B was 115 (SD =595) with 95 limits of agreement minus1050ndash1281 ModelC underestimates BF the mean bias was minus167 (SD =48) with 95 limits of agreement minus1108ndash774 Howeverno significant BF differences were found betweenModel AModel D and DXA Model A and Model D BIA analyzersshowed a significant positive correlation between the bias andaverage BF between BIA and DXA indicating that BFwas underestimated at low values and overestimated at highvalues in the present sample (Table 3)

In girls significant differences in BF were observedbetween Model B Model C Model D and DXA BFwas underestimated by the three BIA devices The meanbiases were minus263 for Model B (SD = 323 95 limitsof agreement minus896ndash370) minus147 for Model D (SD =299 95 limits of agreementminus732ndash440) andminus205 forModel C (SD= 321 95 limits of agreementminus834ndash424)

4 BioMed Research International

Table 1 Physical characteristics and body compositions of the study population

Boys (119899 = 127) Girls (119899 = 128)

Mean (SD) Range Mean (SD) RangeLower Upper Lower Upper

Age (years) 138 (29) 90 195 137 (28) 90 194Height (cm) 1618 (140) 1305 1860 1543 (98) 1266 1720Weight (kg) 559 (170) 275 1081 474 (113) 235 971BMI (kg2m) 210 (42) 136 371 197 (32) 133 348LTM-DXA (kg) 410 (112) 210 653 310 (55) 187 543BMC-DXA (kg) 20 (07) 09 38 17 (05) 08 27FM-DXA (kg) 127 (95) 33 531 143 (68) 27 406FFM-DXA (kg) 430 (118) 221 690 328 (59) 195 568FFM-Model A (kg) 438 (127) 218 774 339 (65) 186 612FFM-Model B (kg) 432 (117) 208 732 347 (55) 206 517FFM-Model C (kg) 445 (120) 226 732 345 (55) 206 516FFM-Model D (kg) 443 (122) 228 720 340 (59) 200 553BF-DXA () 208 (99) 61 491 279 (77) 118 465BF-Model A () 214 (70) 49 396 274 (60) 136 421BF-Model B () 219 (69) 101 453 253 (78) 93 494BF-Model C () 190 (90) 50 547 258 (77) 103 476BF-Model D () 202 (85) 62 473 263 (81) 64 448Note data are presented as mean (SD) BMI body mass index LTM-DXA lean tissue mass measured by dual-energy X-ray absorptiometry BMC-DXA bonemineral content measured by dual-energy X-ray absorptiometry FM-DXA fat mass measured by dual-energy X-ray absorptiometry BF-DXA body fatpercentage measured by dual-energy X-ray absorptiometry FFM-DXA fat-free mass calculated as the sum of LTM and BMC Model A Biodynamics-310Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Table 2 Linrsquos concordance correlation coefficients between param-eters assessed by different BIA and DXA measurement

BIA devicesModel A Model B Model C Model D

Boys(119899 = 127)

BF-DXA 0765 0750 0855 0926FFM-DXA 0953 0953 0951 0979

Girls(119899 = 128)

BF-DXA 0747 0863 0881 0912FFM-DXA 0909 0897 0910 0957

Note BMI body mass index BF-DXA body fat percentage measured bydual-energy X-ray absorptiometry FFM-DXA fat-free mass measured bydual-energy X-ray absorptiometryModel A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Model A and DXA showed no significant differences of BFhowever the bias and the averageBF betweenBIA andDXAhad a significant positive correlation indicating that the BFwas underestimated at low values and overestimated at highvalues in the present sample (Table 3) Model A Model CandModel D overestimated the FFM in boys All BIA devicesoverestimated the FFM in girls (Table 3)

4 Discussion

Chinese children and adolescents are at increased risk forobesity and associated health risks Determining weightstatus throughmeasurement of body fat percentage is impor-tant in diagnosing childhood obesity Several methods such

as DXA Bod Pod and underwater weighing can provideaccurate body fat measurement However these measure-ments are often inaccessible to the larger scale populationsurvey The search for an accurate diagnostic instrumentfor measuring body fat that is practical and inexpensiveis a challenge for researchers In this sense commercialBIA has become a popular method due to its ease of useand low cost [38] Numerous different BIA commercialdevices can measure body composition but their suitabilityfor Chinese children and adolescents is still unclear Theaccuracy of BIA measurement may be influenced by severalfactors such as body shape hydration status and ethnicity[9] Previous studies have suggested that body compositionfrom commercial BIA devices is derived frommanufacturersrsquoequations which are often not validated for specific ethnicgroups [38ndash40] These discrepancies may be associated withdifferences among the devices and ethnic differences on bodycomposition [10 26] We sought investigating the validityand diagnostics accuracy of several different BIA devicesin assessing excessive body fat in Chinese children andadolescentsThe findings of the current study may determineuseful BIA devices for local health practitioners in assessingobese children

Previous validation studies in children and adolescentshave found that different BIA (foot-foot hand-foot andmultiple frequencies) devices usingmanufacturersrsquo equationstend to overestimate or underestimate BF in relation tocriteria measurement [24ndash28 32] Single-frequency foot-foot or hand-foot BIA devices generally tend to underes-timate or overestimate BF in children and adolescents

BioMed Research International 5

Table 3 Cross-validation of four BIA devices (embedded equations) for the perdition of BFFFM in Chinese children and adolescents

Gender BIAdevices

BF-DXA

BF-BIA

FFM-DXA

FFM-BIA 119901 119905-test Bias (SD) TE 119903

119910minus1199101015840119910

95limits ofagreement

Boys(119899 = 127)

Model A 2077 2134 mdash mdash 0268 058 (586) 586 0513lowastlowast minus1096 to1212

Model B 2077 2192 mdash mdash 0031 115 (595) 604 0528lowastlowast minus1051 to1281

Model C 2077 1904 mdash mdash lt0001 minus167(480) 506 0162 minus1108 to

774

Model D 2077 2022 mdash mdash 0246 minus037(351) 351 0365lowastlowast minus725 to

651

Model A mdash mdash 4302 kg 4380 kg 0019 078(370) kg 377 kg 0261lowastlowast minus647 kg to

803 kg

Model B mdash mdash 4302 kg 4318 kg 0608 016(360) kg 360 kg minus0027 minus690 kg to

722 kg

Model C mdash mdash 4302 kg 4450 kg lt0001 154(335) kg 360 kg 0024 minus503 kg to

811 kg

Model D mdash mdash 4302 kg 4431 kg lt0001 102(217) kg 239 kg 0197lowast minus323 kg to

527 kg

Girls(119899 = 128)

Model A 2785 2737 mdash mdash 0266 minus048(488) 488 0359lowastlowast minus1004 to

908

Model B 2785 2522 mdash mdash lt0001 minus263(323) 415 minus0049 minus896 to

370

Model C 2785 2581 mdash mdash lt0001 minus205(321) 379 0012 minus834 to

424

Model D 2785 2634 mdash mdash lt0001 minus147(299) 331 minus0147 minus733 to

439

Model A mdash mdash 3277 kg 3389 kg lt0001 111 (242) kg 266 kg 0245lowastlowast minus363 kg to585 kg

Model B mdash mdash 3277 kg 3474 kg lt0001 197(177) kg 264 kg minus0255lowastlowast minus150 kg to

544 kg

Model C mdash mdash 3277 kg 3449 kg lt0001 172(177) kg 247 kg minus0219lowastlowast minus175 kg to

519 kg

Model D mdash mdash 3277 kg 3403 kg lt0001 125(122) kg 168 kg minus0022 minus114 kg to

364 kgNote BF body fat percentage FFM fat-free mass DXA dual-energy X-ray absorptiometry BIA bioelectrical impedance analysisBias value obtained from BIA measurement minus DXA measurement SD standard deviation119903119910minus1199101015840119910

Pearsonrsquos correlation coefficient for the relationship between averaged BFFFM [(value of DXA + value of BIA)2] and the biaslowast

119901 lt 005 lowastlowast119901 lt 001Model A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D Inbody 520

In the current study although high correlations in bodycomposition measurements between BIA devices and DXAwere found foot-foot (TBF-543) BIA device overestimatedBF in boys and underestimated BF in girls in comparisonwith DXA measurement Most of previous studies foundthat Tanita BIA devices underestimated BF in both genders[24 25 27 37] with which the findings of the current studywas consistent for girls However in current boys TanitaTBF-543 overestimated BF (bias = 115 plusmn 559) withnegative correlation between bias and averaged BF betweenBIA and DXA measurements The discrepancies may beexplained by different testing population ethnicity and use ofTanita BIA devices with different models The current studyuses a heterogeneous and relatively larger sample coveringa wide range of ages and body compositions according toage and gender specific BMI distributions among Chinese

children To date two studies were conducted to investigatethe accuracy of foot-foot BIA consumer devices inmeasuringbody fat percentage in Chinese children and adolescentscompared with DXA measurement [21 23] however thesestudies had several limitations Lu et al measured BF of64 Chinese obese children aged 10ndash17 years using Tanita-401and no significant differences were found in BF betweenfoot-foot BIA and DXA [21] Moreover BIA overestimatedthe BF in seriously obese children Sung et al foundthat BIA (Tanita TBF 401) slightly overestimated BF in7ndash18-year-old Chinese children (119899 = 49) however theBF bias did not reach statistical significance [23] The twostudies used smaller samples with a wider age range toinvestigate the validity of foot-foot measured BF whichmay significantly explain the different findings between thecurrent and previous studies

6 BioMed Research International

In addition despite no mean bias was found for theBF between the hand-foot BIA device (Biodynamics-310)and DXA measurement in both genders the Bland-Altmananalysis showed that the two BIA devices overestimatedmeasurements at a lowBF and underestimated them at highBF Previous studies investigated the validity of hand-footBIA devices in assessing BF in children [22 27 28 41]Several studies had the same results with that of the presentstudy [22 41]

The eight-tactile electrodes with multiple frequenciesBIA device could estimate body composition in adults moreaccurately than a single-frequency BIA device [17 42] How-ever contradictory findings were revealed on the validityof multiple frequencies BIA devices in measuring bodycomposition in children and adolescents Two studies didnot show statistical significance in BF between multiplefrequencies BIA (InBody) and DXA measurements [3031] However Jensky-Squires et al found that InBody 320overestimated BF in female adolescents from underwaterweighing [10] In two other studies Prins et al found thatTanita BC-418 overestimated BF determined by deuteriumdilution in African children [43] Sluyter et al found thatthe same BIA device underestimated BF in a multiethnicgroup of adolescents Moreover the BIA device tended tounderestimate measurements in lower BF and overestimatethem in higher BF individuals [32] In the current studytwomultiple frequencies BIA devices (InBody 520 and TanitaBC-545) were also used to evaluate body composition TanitaBC-545 underestimated BF in both genders and InBody520 underestimated BF in girls In boys no bias was foundin BF between InBody 520 and DXA measurement but asignificant positive correlation was found between the biasand averaged BF between BIA and DXA InBody 520 thusoverestimated BF at high values and overestimated it at lowvalues in boys For Tanita multiple frequencies BIA devicesthe results of the current and previous studies are consistentThedevices underestimatedBF in children and adolescentsFor the InBody multiple BIA devices the varying results mayhave been caused by different population and ethnicity

Moreover note that although themean biasmay be smallof higher relevance are the 95 limits of agreement whichseem to more accurately reflect the validity of measurementin an individual Numerous studies found that the BIA hasquite wide limits of agreement with criterionmeasurement ofBF [24 26 27 41 44] therefore BIA is not interchangeablewith criterion measurement for body composition As suchis true even for Chinese children and adolescents as in thecurrent study the 95 limits of agreement of all BIA devicesshowed wide ranges in agreement with previous studiesSung et al found that the 95 limits of agreement in BFwere narrow (from minus392 to 061) between foot-foot BIAand DXA in 17 obese and 32 nonobese Chinese childrenaged 7ndash18 years [23] The current study had larger 95 limitsof agreement in the same ethnicity and similar age rangeof population The discrepancy may be partly explained bydifferent samples size sampling method and DXA instru-ments used (Hologic QDR-4500 and GE Lunar Prodigy)Previous studies suggested that if researchers intend tomeasure body composition in large-scale population surveys

then BIAmeasurementmay provide an efficient alternative tolaboratory measurement of body composition when accurateinstruments are unavailable [45] BIA measurement shouldbe used with caution in evaluating or monitoring changes inindividual body composition in healthy and clinical setting[38]

5 Conclusion

Using embedded equations in BIA devices should be vali-dated in assessing the body composition of Chinese childrenand adolescents aged 9ndash19 years old All BIA devices are notdirectly interchanged with DXA measurement in assessingindividual BF and assessing changes of body compositionowing to the wide limits of agreement

Disclaimer

The authors alone are responsible for the content and writingof the paper

Conflict of Interests

The authors report no conflict of interests

Acknowledgments

This study was supported by the Hong Kong Association forthe Study of Obesity (2011) and Natural Science Foundationof Shanghai (15ZR1439300) The funders had no role in studydesign collection analysis and interpretation of data writingof paper or decision to submit for paper

References

[1] World Health Organization ldquoObesity and overweightrdquo FactSheet No 311 2011 httpwwwwhointmediacentrefact-sheetsfs311enindexhtml

[2] T Lobstein L Baur and R Uauy ldquoObesity in children andyoung people a crisis in public healthrdquo Obesity ReviewsSupplement vol 5 no 1 pp 4ndash104 2004

[3] Y Wu ldquoOverweight and obesity in Chinardquo British MedicalJournal vol 333 no 7564 pp 362ndash363 2006

[4] Y P Li E G Schouten X Q Hu Z H Cui D C Luan and GS Ma ldquoObesity prevalence and time trend among youngstersin China 1982ndash2002rdquo Asia Pacific Journal of Clinical Nutritionvol 17 no 1 pp 131ndash137 2008

[5] J J Reilly E Methven Z C McDowell et al ldquoHealth conse-quences of obesityrdquo Archives of Disease in Childhood vol 88no 9 pp 748ndash752 2003

[6] W H Dietz ldquoHealth consequences of obesity in youth child-hood predictors of adult diseaserdquo Pediatrics vol 101 no 3 pp518ndash525 1998

[7] J C KWells andM S Fewtrell ldquoMeasuring body compositionrdquoArchives of Disease in Childhood vol 91 no 7 pp 612ndash617 2006

[8] J C K Wells and M S Fewtrell ldquoIs body compositionimportant for paediatriciansrdquoArchives of Disease in Childhoodvol 93 no 2 pp 168ndash172 2008

BioMed Research International 7

[9] V H Heyward and D R Wagner Applied Body CompositionAssessment Human Kinetics Champaign Ill USA 2004

[10] N E Jensky-Squires CMDieli-Conwright A Rossuello DNErceg SMcCauley and T E Schroeder ldquoValidity and reliabilityof body composition analysers in children and adultsrdquo BritishJournal of Nutrition vol 100 no 4 pp 859ndash865 2008

[11] M Helba and L A Binkovitz ldquoPediatric body compositionanalysis with dual-energy X-ray absorptiometryrdquo PediatricRadiology vol 39 no 7 pp 647ndash656 2009

[12] A B Sopher J C Thornton J Wang R N Pierson Jr S BHeymsfield and M Horlick ldquoMeasurement of percentage ofbody fat in 411 children and adolescents a comparison of dual-energy X-ray absorptiometry with a four-compartment modelrdquoPediatrics vol 113 no 5 pp 1285ndash1290 2004

[13] T G Lohman and C Zhao ldquoDual-energy X-ray adsorptiom-etryrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 63ndash77 HumanKinetics Champaign Ill USA 2nd edition 2005

[14] E O Diaz J Villar M Immink and T Gonzales ldquoBioimpe-dance or anthropometryrdquo European Journal of Clinical Nutri-tion vol 43 no 2 pp 129ndash137 1989

[15] S Demura S Sato and T Kitabayashi ldquoPercentage of total bodyfat as estimated by three automatic bioelectrical impedanceanalyzersrdquo Journal of Physiological Anthropology and AppliedHuman Science vol 23 no 3 pp 93ndash99 2004

[16] A Pietrobelli F Rubiano M-P St-Onge and S B HeymsfieldldquoNew bioimpedance analysis system improved phenotypingwith whole-body analysisrdquo European Journal of Clinical Nutri-tion vol 58 no 11 pp 1479ndash1484 2004

[17] R Martinoli E I Mohamed C Maiolo et al ldquoTotal body waterestimation using bioelectrical impedance a meta-analysis ofthe data available in the literaturerdquo Acta Diabetologica vol 40supplement 1 pp S203ndashS206 2003

[18] K R Foster andH C Lukaski ldquoWhole-body impedancemdashwhatdoes it measurerdquo The American Journal of Clinical Nutritionvol 64 no 3 pp 388Sndash396S 1996

[19] W C Chumlea and S S Sun ldquoBioelectrical impedance anal-ysisrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 79ndash88 HumanKinetics Champaign Ill USA 2nd edition 2005

[20] L Iacopino A Andreoli I Innocente et al ldquoUse of foot-to-foot bioelectrical impedance analysis in childrenrdquo ActaDiabetologica vol 40 supplement 1 pp S210ndashS211 2003

[21] K Lu B Quach T K Tong and PW C Lau ldquoValidation of leg-to-leg bio-impedance analysis for assessing body compositionin obese Chinese childrenrdquo Journal of Exercise Scienceamp Fitnessvol 1 no 2 pp 97ndash103 2003

[22] A Mooney L Kelsey G W Fellingham et al ldquoAssessing bodycomposition of children and adolescents using dual-energyX-ray absorptiometry skinfolds and electrical impedancerdquoMeasurement in Physical Education and Exercise Science vol 15no 1 pp 2ndash17 2011

[23] R Y T Sung P Lau CW Yu P KW Lam and E A S NelsonldquoMeasurement of body fat using leg to leg bioimpedancerdquoArchives ofDisease inChildhood vol 85 no 3 pp 263ndash267 2001

[24] C Azcona N Koek and G Fruhbeck ldquoFat mass by air-displacement plethysmography and impedance in obesenon-obese children and adolescentsrdquo International Journal of Pedi-atric Obesity vol 1 no 3 pp 176ndash182 2006

[25] G S Goldfield P Cloutier R Mallory D Prudrsquohomme TParker and E Doucet ldquoValidity of foot-to-foot bioelectrical

impedance analysis in overweight and obese children andparentsrdquoThe Journal of SportsMedicine and Physical Fitness vol46 no 3 pp 447ndash453 2006

[26] J Hosking B S Metcalf A N Jeffery L D Voss and TJ Wilkin ldquoValidation of foot-to-foot bioelectrical impedanceanalysis with dual-energy X-ray absorptiometry in the assess-ment of body composition in young children the EarlyBirdcohortrdquo British Journal of Nutrition vol 96 no 6 pp 1163ndash11682006

[27] S Lazzer Y Boirie M Meyer and M Vermorel ldquoEvaluationof two foot-to-foot bioelectrical impedance analysers to assessbody composition in overweight and obese adolescentsrdquo BritishJournal of Nutrition vol 90 no 5 pp 987ndash992 2003

[28] L Parker J J Reilly C Slater J C K Wells and Y PitsiladisldquoValidity of six field and laboratory methods for measurementof body composition in boysrdquoObesity Research vol 11 no 7 pp852ndash858 2003

[29] J J Reilly K Gerasimidis N Paparacleous et al ldquoValidationof dual-energy x-ray absorptiometry and foot-foot impedanceagainst deuterium dilution measures of fatness in childrenrdquoInternational Journal of Pediatric Obesity vol 5 no 1 pp 111ndash115 2010

[30] J S Lim J S Hwang J A Lee et al ldquoCross-calibration ofmulti-frequency bioelectrical impedance analysis with eight-point tactile electrodes and dual-energy X-ray absorptiometryfor assessment of body composition in healthy children aged6-18 yearsrdquo Pediatrics International vol 51 no 2 pp 263ndash2682009

[31] O K Yu Y K Rhee T S Park and Y S Cha ldquoComparisons ofobesity assessments in over-weight elementary students usinganthropometry BIA CT and DEXArdquo Nutrition Research andPractice vol 4 no 2 pp 128ndash135 2010

[32] J D Sluyter D Schaaf R K R Scragg and L D PlankldquoPrediction of fatness by standing 8-electrode bioimpedance amultiethnic adolescent populationrdquo Obesity vol 18 no 1 pp183ndash189 2010

[33] K P Navder Q He X Zhang et al ldquoRelationship between bodymass index and adiposity in prepubertal children ethnic andgeographic comparisons between New York City and Jinan City(China)rdquo Journal of Applied Physiology vol 107 no 2 pp 488ndash493 2009

[34] N J Shaw N J Crabtree M S Kibirige and J N FordhamldquoEthnic and gender differences in body fat in British schoolchil-dren as measured by DXArdquo Archives of Disease in Childhoodvol 92 no 10 pp 872ndash875 2007

[35] M A Stone L Williams S Chatterjee M J Davies and KKhunti ldquoEthnic differences in body composition in adoles-centsrdquo Primary Care Diabetes vol 2 no 1 pp 55ndash57 2008

[36] G B McBride ldquoA proposal for strength-of-agreement criteriafor Linrsquos concordance correlation coefficientrdquo NIWA ClientReport HAM2005-062 2005

[37] J M Bland and D G Altman ldquoStatistical methods for assessingagreement between two methods of clinical measurementrdquoTheLancet vol 1 no 8476 pp 307ndash310 1986

[38] M Dehghan and A T Merchant ldquoIs bioelectrical impedanceaccurate for use in large epidemiological studiesrdquo NutritionJournal vol 7 no 1 article 26 2008

[39] B SMcClanahanM B Stockton J Q Lanctot et al ldquoMeasure-ment of body composition in 810-year-old African-Americangirls a comparison of dual-energy X-ray absorptiometry andfoot-to-foot bioimpedance methodsrdquo International Journal ofPediatric Obesity vol 4 no 4 pp 389ndash396 2009

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

2 BioMed Research International

However because of its inaccessibility the high cost ofequipment and exposure to low-dose radiation the use ofDXA is limited in the field setting Thus simple methodssuch as bioelectrical impedance analysis (BIA) and skinfoldthickness (SKF) measurement are used in field studies [9]

Compared with methods of measuring body compo-sition BIA is relatively simple quick and noninvasive Itgives reliable measurements of body composition with smallintra- and interobserver variability [14] BIA does not requireexposure to radioactivity or submersion in water which iswhy it is regarded as a practicalmeasure of body compositionespecially for field studies involving children Because of itsadvantages many BIA consumer devices are popular withhealth professionals and the general public for the assessmentand monitoring of body fat These devices are characterizedby differences in electrodes (number type and placement)electric current frequency and body position atmeasurement[15] Single-frequency BIA devices with four electrodes suchas a traditional hand-to-foot BIA machine using adheringelectrodes and a foot-to-foot BIA machine with four metalelectrodes have been used widely to assess body composition[15] Multifrequency hand-to-foot BIA machines with eightelectrodes have been developed for assessing segmental andtotal body compositions [16] Multifrequency BIA was sug-gested to be a more accurate method for estimating the bodycomposition of adults [17] Segmental multiple-frequencyBIA devices employ a current pathway that bypasses thearms legs and trunk using current different frequencies[18] The most appropriate combination of frequencies andmultivariate methods of using multi-frequency impedancevalues in estimating body composition are just now beingexplored therefore bioelectrical impedance spectroscopy(BIS) devices are designed to scan a wide range of frequen-cies this approach provides estimates of extracellular waterintracellular water and total body water [19]

Foot-to-foot BIA consumer devices are widely used toassess body composition among children and adolescentsSome studies have investigated the accuracy of single-frequency foot-to-foot BIA devices in assessing body fatContradictory results were found between foot-to-foot BIAand the reference methods in assessing body compositionamong children and adolescents with some studies showingan acceptable level of accuracy [20ndash23] and others reporting apoor level of accuracy [24ndash29] Althoughmultifrequency BIAwas suggested to be a more accurate method for estimatingadultsrsquo body composition [17] its accuracy in evaluating thebody composition of children and adolescents is doubtfulSome studies considered multifrequency BIA as an accurateway to measure body fat in children and adolescents [30 31]whereas others found poor accuracy betweenmultifrequencyBIA and criterion methods [10 32] These discrepanciesmay be associated with differences among BIA consumerdevices and ethnic differences in body composition Previousstudies showed that East Asian and South Asian childrenand adolescents tend to have more body fat than Caucasiansat an equivalent BMI [33ndash35] To date only two studieshave been conducted to investigate the accuracy of foot-to-foot BIA consumer devices for the measurement of bodyfat percentage in Chinese children and adolescents against

the DXA measurement [21 23] A good correlation of bodyfat percentage and acceptable agreement between BIA mea-surement and DXA measurement were identified Howeverthese studies were limited by a small sample size in a wideage range Only single-frequency foot-to-foot BIA deviceswere tested in these studiesThe accuracy of different portableBIA consumer devices in assessing body fat among Chinesechildren and adolescents remains questionable

To keep track of childhood obesity levels effectively it isnecessary to evaluate the obesity level of children in Chinausing objective and practical measurement methods BIAconsumer devices provide a cheap and easy way to measurethe body composition of Chinese children and adolescentsHowever no study has assessed the validity of differentportable BIA devices in relation to a preferred criterionmethod To achieve this the accuracy of BIA consumerdevices in predicting body fat among Chinese children andadolescents should be determined Therefore the aim of thepresent study is to examine the validity in measuring bodyfat

2 Material and Methods

21 Participants A total of 255 healthy Chinese children andadolescents aged 9ndash19 years were recruited from schools inShanghai China Of the total 127 were boys and 128 weregirls A stratified sampling method was used to recruit aheterogeneous sample covering a wide range of ages andbody compositions according to age and gender specificBMI distributions among Chinese children The stratifiedsampling method was used in our previous study Writteninformed consent was obtained from each participant and hisor her parents and the studywas approved by the Local EthicsCommittee

22 Anthropometrics Measurement Body weight and heightwere measured with minimal clothing and bare feet Bodyheight was measured to the nearest 05 cm using a fixedstadiometer (Holtain Crymych Dyfed United Kingdom)Body weight was measured to the nearest 01 kg using astandard scale (Tanita TBF-543 Tanita Tokyo Japan) Atrained investigator performed all the measurements

23 Measurement of Body Composition

231 DXA Measurement Dual-energy X-ray absorptiom-etry (GE Lunar Prodigy software version 1051006 GEHealthcare Madison USA) was used as the criterion mea-surement for body composition in terms of fat mass (FM)lean body mass (LTM) and bone mineral contents (BMC)Each participant was scanned in the supine position using X-ray at two energy sources (40 keV and 70 keV) in fastmode Aseries of transverse scans were done from head to toe at one-centimeter intervals The time spent was about five minutesdepending on the participantrsquos height A trained investigatorcompleted all the measurements of the DXA scan

232 BIA Measurements BIA measurements were under-taken at least two hours after breakfast and with an empty

BioMed Research International 3

bladder Girls having their menses avoided the tests Inthe current study body composition was assessed by fourBIA scales namely Biodynamics-310 (Model A BiodynamicsCorp Seattle USA) Tanita TBF-543 (Model B Tanita CorpTokyo Japan) Tanita BC-545 (Model C Tanita Corp TokyoJapan) and InBody 520 (Model D Biospace Co Ltd SeoulKorea) Manufacturersrsquo equations were used to predict BF(using all scales) and fat-freemass (usingModel A andModelD) Model A was a single-frequency traditional hand-to-footBIA machine using adhering electrodes The measurementwas carried out with the participants lying supine on a couchThe armswere separated from the trunk by about 30∘ the legswere abducted and separated by about 45∘ The skin of theright hand and foot was cleaned with an alcohol pad beforethe electrodes were placed Body composition was measuredat 50 kHz with a tetrapolar arrangement of standard elec-trodes (Red Dot 2330 3M Healthcare Saint Paul USA) Twoelectrodes were placed on the right ankle with one (the sourceelectrode) just proximal to the third metatarsophalangealjoint and the other (the sensing electrode) on the anteriorankle between the medial and the lateral malleolus Twoelectrodes were placed on the right wrist with one (the sourceelectrode) just proximal to the third metacarpophalangealjoint and the other (the sensing electrode) on the posteriorwrist between the styloid processes of the radius and theulna Model B was a bipolar single-frequency (50 kHz) foot-to-foot instrument Participants were instructed to standbarefoot with heel and forefoot placed on the four metalelectrodes Model C was a dual-frequency (50 kHz and625 kHz) andModel Dwas amultifrequency (5 kHz 50 kHzand 500 kHz) BIA device with eight electrodes in a tetrapolararrangementThemeasurements required the participants tostand barefoot on metal electrodes while grasping a pair ofelectrodes fixed on a handle with arms extended in frontof the chest All BIA measurements were completed by atrained investigator according to the device manufacturersrsquoinstructions

24 Statistical Analysis Descriptive statistics were reportedas means plusmn standard deviations (SD) An independent t-test was employed to determine the differences in the par-ticipantsrsquo physical characteristics DXA measurements andBIA measurements between boys and girls The 5 levelwas chosen for statistical significance Linrsquos concordancecorrelation coefficient (rc) was used to measure the bivariaterelationship of BF and FFM obtained from DXA withthose obtained from BIA measurements McBride suggeststhe following descriptive scale for values of the concordancecorrelation coefficient (for continuous variables) lt090 poor090ndash095 moderate 095ndash099 substantial and gt099 almostperfect [36] Differences in body composition between thoseassessed by BIA methods and DXAmeasures were examinedusing a paired t-test Bland-Altman analysis was applied todetermine the agreement between BIA and DXA measure-ments [37] The agreement between methods is representedby the bias estimated by the mean difference and the SD ofthe differences Therefore the 95 limits of agreement werecalculated as the mean difference plusmn196 SD of the differencesbetween methods The pure error in BF between each BIA

device with DXA measurement was assessed using the totalerror (TE) as follows

TE = radicsum(119884 minus 119884

1015840

)

2

119873

(1)

where 119884 = observed values 1198841015840 = predicted values and 119873 =the number of participants in the sample These statisticalanalyses were performed separately on boys and girls

3 Results

31 Characteristics of Participants A total of 255 (128 girlsand 127 boys) Chinese children and adolescents participatedin the study Table 1 presents their physical characteristics andbody compartments (FM LTM FFM BMC and BF) Nosignificant differences in mean age were observed betweengenders The mean body weight height and BMI of boyswere higher than those of girls The mean of the individualbody compartments was significantly different between boysand girls (Table 1) 803 of boys and 828 of girls hadentered puberty (pubertal stage ge2 for pubic hair or breasts(girls)genitalia (boys))

Table 2 shows Linrsquos concordance correlation coefficientsbetween parameters of body composition assessed by dif-ferent BIA devices and DXA methods In boys the FFMdetermined by all BIA devices showed substantial agreementswith that determined by DXA (rc values within 0951ndash0979)In girls the FFM determined by InBody 520 showed substan-tial agreement with that determined by DXA (rc = 0957)moderate agreements for FFM were found on Biodynamics-310 (rc = 0909) and Tanita BC-545 (rc = 0910) and pooragreement for FFMwas found onTanita TBF 543 (rc= 0897)Linrsquos concordance correlation coefficients of estimated BFbetween Biodynamics-310 Tanita TBF-543 Tanita BC-545and DXA showed poor agreements (rc values within 0747ndash0863) for both genders Moderate agreements for BF werefound between DXA and InBody 520 measurements (rc =0926 in boys and 0912 in girls)

In boys significant differences in BF were foundbetween DXA and Model B and Model C (paired t-test 119901 lt005) The mean bias of BF for Model B was 115 (SD =595) with 95 limits of agreement minus1050ndash1281 ModelC underestimates BF the mean bias was minus167 (SD =48) with 95 limits of agreement minus1108ndash774 Howeverno significant BF differences were found betweenModel AModel D and DXA Model A and Model D BIA analyzersshowed a significant positive correlation between the bias andaverage BF between BIA and DXA indicating that BFwas underestimated at low values and overestimated at highvalues in the present sample (Table 3)

In girls significant differences in BF were observedbetween Model B Model C Model D and DXA BFwas underestimated by the three BIA devices The meanbiases were minus263 for Model B (SD = 323 95 limitsof agreement minus896ndash370) minus147 for Model D (SD =299 95 limits of agreementminus732ndash440) andminus205 forModel C (SD= 321 95 limits of agreementminus834ndash424)

4 BioMed Research International

Table 1 Physical characteristics and body compositions of the study population

Boys (119899 = 127) Girls (119899 = 128)

Mean (SD) Range Mean (SD) RangeLower Upper Lower Upper

Age (years) 138 (29) 90 195 137 (28) 90 194Height (cm) 1618 (140) 1305 1860 1543 (98) 1266 1720Weight (kg) 559 (170) 275 1081 474 (113) 235 971BMI (kg2m) 210 (42) 136 371 197 (32) 133 348LTM-DXA (kg) 410 (112) 210 653 310 (55) 187 543BMC-DXA (kg) 20 (07) 09 38 17 (05) 08 27FM-DXA (kg) 127 (95) 33 531 143 (68) 27 406FFM-DXA (kg) 430 (118) 221 690 328 (59) 195 568FFM-Model A (kg) 438 (127) 218 774 339 (65) 186 612FFM-Model B (kg) 432 (117) 208 732 347 (55) 206 517FFM-Model C (kg) 445 (120) 226 732 345 (55) 206 516FFM-Model D (kg) 443 (122) 228 720 340 (59) 200 553BF-DXA () 208 (99) 61 491 279 (77) 118 465BF-Model A () 214 (70) 49 396 274 (60) 136 421BF-Model B () 219 (69) 101 453 253 (78) 93 494BF-Model C () 190 (90) 50 547 258 (77) 103 476BF-Model D () 202 (85) 62 473 263 (81) 64 448Note data are presented as mean (SD) BMI body mass index LTM-DXA lean tissue mass measured by dual-energy X-ray absorptiometry BMC-DXA bonemineral content measured by dual-energy X-ray absorptiometry FM-DXA fat mass measured by dual-energy X-ray absorptiometry BF-DXA body fatpercentage measured by dual-energy X-ray absorptiometry FFM-DXA fat-free mass calculated as the sum of LTM and BMC Model A Biodynamics-310Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Table 2 Linrsquos concordance correlation coefficients between param-eters assessed by different BIA and DXA measurement

BIA devicesModel A Model B Model C Model D

Boys(119899 = 127)

BF-DXA 0765 0750 0855 0926FFM-DXA 0953 0953 0951 0979

Girls(119899 = 128)

BF-DXA 0747 0863 0881 0912FFM-DXA 0909 0897 0910 0957

Note BMI body mass index BF-DXA body fat percentage measured bydual-energy X-ray absorptiometry FFM-DXA fat-free mass measured bydual-energy X-ray absorptiometryModel A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Model A and DXA showed no significant differences of BFhowever the bias and the averageBF betweenBIA andDXAhad a significant positive correlation indicating that the BFwas underestimated at low values and overestimated at highvalues in the present sample (Table 3) Model A Model CandModel D overestimated the FFM in boys All BIA devicesoverestimated the FFM in girls (Table 3)

4 Discussion

Chinese children and adolescents are at increased risk forobesity and associated health risks Determining weightstatus throughmeasurement of body fat percentage is impor-tant in diagnosing childhood obesity Several methods such

as DXA Bod Pod and underwater weighing can provideaccurate body fat measurement However these measure-ments are often inaccessible to the larger scale populationsurvey The search for an accurate diagnostic instrumentfor measuring body fat that is practical and inexpensiveis a challenge for researchers In this sense commercialBIA has become a popular method due to its ease of useand low cost [38] Numerous different BIA commercialdevices can measure body composition but their suitabilityfor Chinese children and adolescents is still unclear Theaccuracy of BIA measurement may be influenced by severalfactors such as body shape hydration status and ethnicity[9] Previous studies have suggested that body compositionfrom commercial BIA devices is derived frommanufacturersrsquoequations which are often not validated for specific ethnicgroups [38ndash40] These discrepancies may be associated withdifferences among the devices and ethnic differences on bodycomposition [10 26] We sought investigating the validityand diagnostics accuracy of several different BIA devicesin assessing excessive body fat in Chinese children andadolescentsThe findings of the current study may determineuseful BIA devices for local health practitioners in assessingobese children

Previous validation studies in children and adolescentshave found that different BIA (foot-foot hand-foot andmultiple frequencies) devices usingmanufacturersrsquo equationstend to overestimate or underestimate BF in relation tocriteria measurement [24ndash28 32] Single-frequency foot-foot or hand-foot BIA devices generally tend to underes-timate or overestimate BF in children and adolescents

BioMed Research International 5

Table 3 Cross-validation of four BIA devices (embedded equations) for the perdition of BFFFM in Chinese children and adolescents

Gender BIAdevices

BF-DXA

BF-BIA

FFM-DXA

FFM-BIA 119901 119905-test Bias (SD) TE 119903

119910minus1199101015840119910

95limits ofagreement

Boys(119899 = 127)

Model A 2077 2134 mdash mdash 0268 058 (586) 586 0513lowastlowast minus1096 to1212

Model B 2077 2192 mdash mdash 0031 115 (595) 604 0528lowastlowast minus1051 to1281

Model C 2077 1904 mdash mdash lt0001 minus167(480) 506 0162 minus1108 to

774

Model D 2077 2022 mdash mdash 0246 minus037(351) 351 0365lowastlowast minus725 to

651

Model A mdash mdash 4302 kg 4380 kg 0019 078(370) kg 377 kg 0261lowastlowast minus647 kg to

803 kg

Model B mdash mdash 4302 kg 4318 kg 0608 016(360) kg 360 kg minus0027 minus690 kg to

722 kg

Model C mdash mdash 4302 kg 4450 kg lt0001 154(335) kg 360 kg 0024 minus503 kg to

811 kg

Model D mdash mdash 4302 kg 4431 kg lt0001 102(217) kg 239 kg 0197lowast minus323 kg to

527 kg

Girls(119899 = 128)

Model A 2785 2737 mdash mdash 0266 minus048(488) 488 0359lowastlowast minus1004 to

908

Model B 2785 2522 mdash mdash lt0001 minus263(323) 415 minus0049 minus896 to

370

Model C 2785 2581 mdash mdash lt0001 minus205(321) 379 0012 minus834 to

424

Model D 2785 2634 mdash mdash lt0001 minus147(299) 331 minus0147 minus733 to

439

Model A mdash mdash 3277 kg 3389 kg lt0001 111 (242) kg 266 kg 0245lowastlowast minus363 kg to585 kg

Model B mdash mdash 3277 kg 3474 kg lt0001 197(177) kg 264 kg minus0255lowastlowast minus150 kg to

544 kg

Model C mdash mdash 3277 kg 3449 kg lt0001 172(177) kg 247 kg minus0219lowastlowast minus175 kg to

519 kg

Model D mdash mdash 3277 kg 3403 kg lt0001 125(122) kg 168 kg minus0022 minus114 kg to

364 kgNote BF body fat percentage FFM fat-free mass DXA dual-energy X-ray absorptiometry BIA bioelectrical impedance analysisBias value obtained from BIA measurement minus DXA measurement SD standard deviation119903119910minus1199101015840119910

Pearsonrsquos correlation coefficient for the relationship between averaged BFFFM [(value of DXA + value of BIA)2] and the biaslowast

119901 lt 005 lowastlowast119901 lt 001Model A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D Inbody 520

In the current study although high correlations in bodycomposition measurements between BIA devices and DXAwere found foot-foot (TBF-543) BIA device overestimatedBF in boys and underestimated BF in girls in comparisonwith DXA measurement Most of previous studies foundthat Tanita BIA devices underestimated BF in both genders[24 25 27 37] with which the findings of the current studywas consistent for girls However in current boys TanitaTBF-543 overestimated BF (bias = 115 plusmn 559) withnegative correlation between bias and averaged BF betweenBIA and DXA measurements The discrepancies may beexplained by different testing population ethnicity and use ofTanita BIA devices with different models The current studyuses a heterogeneous and relatively larger sample coveringa wide range of ages and body compositions according toage and gender specific BMI distributions among Chinese

children To date two studies were conducted to investigatethe accuracy of foot-foot BIA consumer devices inmeasuringbody fat percentage in Chinese children and adolescentscompared with DXA measurement [21 23] however thesestudies had several limitations Lu et al measured BF of64 Chinese obese children aged 10ndash17 years using Tanita-401and no significant differences were found in BF betweenfoot-foot BIA and DXA [21] Moreover BIA overestimatedthe BF in seriously obese children Sung et al foundthat BIA (Tanita TBF 401) slightly overestimated BF in7ndash18-year-old Chinese children (119899 = 49) however theBF bias did not reach statistical significance [23] The twostudies used smaller samples with a wider age range toinvestigate the validity of foot-foot measured BF whichmay significantly explain the different findings between thecurrent and previous studies

6 BioMed Research International

In addition despite no mean bias was found for theBF between the hand-foot BIA device (Biodynamics-310)and DXA measurement in both genders the Bland-Altmananalysis showed that the two BIA devices overestimatedmeasurements at a lowBF and underestimated them at highBF Previous studies investigated the validity of hand-footBIA devices in assessing BF in children [22 27 28 41]Several studies had the same results with that of the presentstudy [22 41]

The eight-tactile electrodes with multiple frequenciesBIA device could estimate body composition in adults moreaccurately than a single-frequency BIA device [17 42] How-ever contradictory findings were revealed on the validityof multiple frequencies BIA devices in measuring bodycomposition in children and adolescents Two studies didnot show statistical significance in BF between multiplefrequencies BIA (InBody) and DXA measurements [3031] However Jensky-Squires et al found that InBody 320overestimated BF in female adolescents from underwaterweighing [10] In two other studies Prins et al found thatTanita BC-418 overestimated BF determined by deuteriumdilution in African children [43] Sluyter et al found thatthe same BIA device underestimated BF in a multiethnicgroup of adolescents Moreover the BIA device tended tounderestimate measurements in lower BF and overestimatethem in higher BF individuals [32] In the current studytwomultiple frequencies BIA devices (InBody 520 and TanitaBC-545) were also used to evaluate body composition TanitaBC-545 underestimated BF in both genders and InBody520 underestimated BF in girls In boys no bias was foundin BF between InBody 520 and DXA measurement but asignificant positive correlation was found between the biasand averaged BF between BIA and DXA InBody 520 thusoverestimated BF at high values and overestimated it at lowvalues in boys For Tanita multiple frequencies BIA devicesthe results of the current and previous studies are consistentThedevices underestimatedBF in children and adolescentsFor the InBody multiple BIA devices the varying results mayhave been caused by different population and ethnicity

Moreover note that although themean biasmay be smallof higher relevance are the 95 limits of agreement whichseem to more accurately reflect the validity of measurementin an individual Numerous studies found that the BIA hasquite wide limits of agreement with criterionmeasurement ofBF [24 26 27 41 44] therefore BIA is not interchangeablewith criterion measurement for body composition As suchis true even for Chinese children and adolescents as in thecurrent study the 95 limits of agreement of all BIA devicesshowed wide ranges in agreement with previous studiesSung et al found that the 95 limits of agreement in BFwere narrow (from minus392 to 061) between foot-foot BIAand DXA in 17 obese and 32 nonobese Chinese childrenaged 7ndash18 years [23] The current study had larger 95 limitsof agreement in the same ethnicity and similar age rangeof population The discrepancy may be partly explained bydifferent samples size sampling method and DXA instru-ments used (Hologic QDR-4500 and GE Lunar Prodigy)Previous studies suggested that if researchers intend tomeasure body composition in large-scale population surveys

then BIAmeasurementmay provide an efficient alternative tolaboratory measurement of body composition when accurateinstruments are unavailable [45] BIA measurement shouldbe used with caution in evaluating or monitoring changes inindividual body composition in healthy and clinical setting[38]

5 Conclusion

Using embedded equations in BIA devices should be vali-dated in assessing the body composition of Chinese childrenand adolescents aged 9ndash19 years old All BIA devices are notdirectly interchanged with DXA measurement in assessingindividual BF and assessing changes of body compositionowing to the wide limits of agreement

Disclaimer

The authors alone are responsible for the content and writingof the paper

Conflict of Interests

The authors report no conflict of interests

Acknowledgments

This study was supported by the Hong Kong Association forthe Study of Obesity (2011) and Natural Science Foundationof Shanghai (15ZR1439300) The funders had no role in studydesign collection analysis and interpretation of data writingof paper or decision to submit for paper

References

[1] World Health Organization ldquoObesity and overweightrdquo FactSheet No 311 2011 httpwwwwhointmediacentrefact-sheetsfs311enindexhtml

[2] T Lobstein L Baur and R Uauy ldquoObesity in children andyoung people a crisis in public healthrdquo Obesity ReviewsSupplement vol 5 no 1 pp 4ndash104 2004

[3] Y Wu ldquoOverweight and obesity in Chinardquo British MedicalJournal vol 333 no 7564 pp 362ndash363 2006

[4] Y P Li E G Schouten X Q Hu Z H Cui D C Luan and GS Ma ldquoObesity prevalence and time trend among youngstersin China 1982ndash2002rdquo Asia Pacific Journal of Clinical Nutritionvol 17 no 1 pp 131ndash137 2008

[5] J J Reilly E Methven Z C McDowell et al ldquoHealth conse-quences of obesityrdquo Archives of Disease in Childhood vol 88no 9 pp 748ndash752 2003

[6] W H Dietz ldquoHealth consequences of obesity in youth child-hood predictors of adult diseaserdquo Pediatrics vol 101 no 3 pp518ndash525 1998

[7] J C KWells andM S Fewtrell ldquoMeasuring body compositionrdquoArchives of Disease in Childhood vol 91 no 7 pp 612ndash617 2006

[8] J C K Wells and M S Fewtrell ldquoIs body compositionimportant for paediatriciansrdquoArchives of Disease in Childhoodvol 93 no 2 pp 168ndash172 2008

BioMed Research International 7

[9] V H Heyward and D R Wagner Applied Body CompositionAssessment Human Kinetics Champaign Ill USA 2004

[10] N E Jensky-Squires CMDieli-Conwright A Rossuello DNErceg SMcCauley and T E Schroeder ldquoValidity and reliabilityof body composition analysers in children and adultsrdquo BritishJournal of Nutrition vol 100 no 4 pp 859ndash865 2008

[11] M Helba and L A Binkovitz ldquoPediatric body compositionanalysis with dual-energy X-ray absorptiometryrdquo PediatricRadiology vol 39 no 7 pp 647ndash656 2009

[12] A B Sopher J C Thornton J Wang R N Pierson Jr S BHeymsfield and M Horlick ldquoMeasurement of percentage ofbody fat in 411 children and adolescents a comparison of dual-energy X-ray absorptiometry with a four-compartment modelrdquoPediatrics vol 113 no 5 pp 1285ndash1290 2004

[13] T G Lohman and C Zhao ldquoDual-energy X-ray adsorptiom-etryrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 63ndash77 HumanKinetics Champaign Ill USA 2nd edition 2005

[14] E O Diaz J Villar M Immink and T Gonzales ldquoBioimpe-dance or anthropometryrdquo European Journal of Clinical Nutri-tion vol 43 no 2 pp 129ndash137 1989

[15] S Demura S Sato and T Kitabayashi ldquoPercentage of total bodyfat as estimated by three automatic bioelectrical impedanceanalyzersrdquo Journal of Physiological Anthropology and AppliedHuman Science vol 23 no 3 pp 93ndash99 2004

[16] A Pietrobelli F Rubiano M-P St-Onge and S B HeymsfieldldquoNew bioimpedance analysis system improved phenotypingwith whole-body analysisrdquo European Journal of Clinical Nutri-tion vol 58 no 11 pp 1479ndash1484 2004

[17] R Martinoli E I Mohamed C Maiolo et al ldquoTotal body waterestimation using bioelectrical impedance a meta-analysis ofthe data available in the literaturerdquo Acta Diabetologica vol 40supplement 1 pp S203ndashS206 2003

[18] K R Foster andH C Lukaski ldquoWhole-body impedancemdashwhatdoes it measurerdquo The American Journal of Clinical Nutritionvol 64 no 3 pp 388Sndash396S 1996

[19] W C Chumlea and S S Sun ldquoBioelectrical impedance anal-ysisrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 79ndash88 HumanKinetics Champaign Ill USA 2nd edition 2005

[20] L Iacopino A Andreoli I Innocente et al ldquoUse of foot-to-foot bioelectrical impedance analysis in childrenrdquo ActaDiabetologica vol 40 supplement 1 pp S210ndashS211 2003

[21] K Lu B Quach T K Tong and PW C Lau ldquoValidation of leg-to-leg bio-impedance analysis for assessing body compositionin obese Chinese childrenrdquo Journal of Exercise Scienceamp Fitnessvol 1 no 2 pp 97ndash103 2003

[22] A Mooney L Kelsey G W Fellingham et al ldquoAssessing bodycomposition of children and adolescents using dual-energyX-ray absorptiometry skinfolds and electrical impedancerdquoMeasurement in Physical Education and Exercise Science vol 15no 1 pp 2ndash17 2011

[23] R Y T Sung P Lau CW Yu P KW Lam and E A S NelsonldquoMeasurement of body fat using leg to leg bioimpedancerdquoArchives ofDisease inChildhood vol 85 no 3 pp 263ndash267 2001

[24] C Azcona N Koek and G Fruhbeck ldquoFat mass by air-displacement plethysmography and impedance in obesenon-obese children and adolescentsrdquo International Journal of Pedi-atric Obesity vol 1 no 3 pp 176ndash182 2006

[25] G S Goldfield P Cloutier R Mallory D Prudrsquohomme TParker and E Doucet ldquoValidity of foot-to-foot bioelectrical

impedance analysis in overweight and obese children andparentsrdquoThe Journal of SportsMedicine and Physical Fitness vol46 no 3 pp 447ndash453 2006

[26] J Hosking B S Metcalf A N Jeffery L D Voss and TJ Wilkin ldquoValidation of foot-to-foot bioelectrical impedanceanalysis with dual-energy X-ray absorptiometry in the assess-ment of body composition in young children the EarlyBirdcohortrdquo British Journal of Nutrition vol 96 no 6 pp 1163ndash11682006

[27] S Lazzer Y Boirie M Meyer and M Vermorel ldquoEvaluationof two foot-to-foot bioelectrical impedance analysers to assessbody composition in overweight and obese adolescentsrdquo BritishJournal of Nutrition vol 90 no 5 pp 987ndash992 2003

[28] L Parker J J Reilly C Slater J C K Wells and Y PitsiladisldquoValidity of six field and laboratory methods for measurementof body composition in boysrdquoObesity Research vol 11 no 7 pp852ndash858 2003

[29] J J Reilly K Gerasimidis N Paparacleous et al ldquoValidationof dual-energy x-ray absorptiometry and foot-foot impedanceagainst deuterium dilution measures of fatness in childrenrdquoInternational Journal of Pediatric Obesity vol 5 no 1 pp 111ndash115 2010

[30] J S Lim J S Hwang J A Lee et al ldquoCross-calibration ofmulti-frequency bioelectrical impedance analysis with eight-point tactile electrodes and dual-energy X-ray absorptiometryfor assessment of body composition in healthy children aged6-18 yearsrdquo Pediatrics International vol 51 no 2 pp 263ndash2682009

[31] O K Yu Y K Rhee T S Park and Y S Cha ldquoComparisons ofobesity assessments in over-weight elementary students usinganthropometry BIA CT and DEXArdquo Nutrition Research andPractice vol 4 no 2 pp 128ndash135 2010

[32] J D Sluyter D Schaaf R K R Scragg and L D PlankldquoPrediction of fatness by standing 8-electrode bioimpedance amultiethnic adolescent populationrdquo Obesity vol 18 no 1 pp183ndash189 2010

[33] K P Navder Q He X Zhang et al ldquoRelationship between bodymass index and adiposity in prepubertal children ethnic andgeographic comparisons between New York City and Jinan City(China)rdquo Journal of Applied Physiology vol 107 no 2 pp 488ndash493 2009

[34] N J Shaw N J Crabtree M S Kibirige and J N FordhamldquoEthnic and gender differences in body fat in British schoolchil-dren as measured by DXArdquo Archives of Disease in Childhoodvol 92 no 10 pp 872ndash875 2007

[35] M A Stone L Williams S Chatterjee M J Davies and KKhunti ldquoEthnic differences in body composition in adoles-centsrdquo Primary Care Diabetes vol 2 no 1 pp 55ndash57 2008

[36] G B McBride ldquoA proposal for strength-of-agreement criteriafor Linrsquos concordance correlation coefficientrdquo NIWA ClientReport HAM2005-062 2005

[37] J M Bland and D G Altman ldquoStatistical methods for assessingagreement between two methods of clinical measurementrdquoTheLancet vol 1 no 8476 pp 307ndash310 1986

[38] M Dehghan and A T Merchant ldquoIs bioelectrical impedanceaccurate for use in large epidemiological studiesrdquo NutritionJournal vol 7 no 1 article 26 2008

[39] B SMcClanahanM B Stockton J Q Lanctot et al ldquoMeasure-ment of body composition in 810-year-old African-Americangirls a comparison of dual-energy X-ray absorptiometry andfoot-to-foot bioimpedance methodsrdquo International Journal ofPediatric Obesity vol 4 no 4 pp 389ndash396 2009

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

BioMed Research International 3

bladder Girls having their menses avoided the tests Inthe current study body composition was assessed by fourBIA scales namely Biodynamics-310 (Model A BiodynamicsCorp Seattle USA) Tanita TBF-543 (Model B Tanita CorpTokyo Japan) Tanita BC-545 (Model C Tanita Corp TokyoJapan) and InBody 520 (Model D Biospace Co Ltd SeoulKorea) Manufacturersrsquo equations were used to predict BF(using all scales) and fat-freemass (usingModel A andModelD) Model A was a single-frequency traditional hand-to-footBIA machine using adhering electrodes The measurementwas carried out with the participants lying supine on a couchThe armswere separated from the trunk by about 30∘ the legswere abducted and separated by about 45∘ The skin of theright hand and foot was cleaned with an alcohol pad beforethe electrodes were placed Body composition was measuredat 50 kHz with a tetrapolar arrangement of standard elec-trodes (Red Dot 2330 3M Healthcare Saint Paul USA) Twoelectrodes were placed on the right ankle with one (the sourceelectrode) just proximal to the third metatarsophalangealjoint and the other (the sensing electrode) on the anteriorankle between the medial and the lateral malleolus Twoelectrodes were placed on the right wrist with one (the sourceelectrode) just proximal to the third metacarpophalangealjoint and the other (the sensing electrode) on the posteriorwrist between the styloid processes of the radius and theulna Model B was a bipolar single-frequency (50 kHz) foot-to-foot instrument Participants were instructed to standbarefoot with heel and forefoot placed on the four metalelectrodes Model C was a dual-frequency (50 kHz and625 kHz) andModel Dwas amultifrequency (5 kHz 50 kHzand 500 kHz) BIA device with eight electrodes in a tetrapolararrangementThemeasurements required the participants tostand barefoot on metal electrodes while grasping a pair ofelectrodes fixed on a handle with arms extended in frontof the chest All BIA measurements were completed by atrained investigator according to the device manufacturersrsquoinstructions

24 Statistical Analysis Descriptive statistics were reportedas means plusmn standard deviations (SD) An independent t-test was employed to determine the differences in the par-ticipantsrsquo physical characteristics DXA measurements andBIA measurements between boys and girls The 5 levelwas chosen for statistical significance Linrsquos concordancecorrelation coefficient (rc) was used to measure the bivariaterelationship of BF and FFM obtained from DXA withthose obtained from BIA measurements McBride suggeststhe following descriptive scale for values of the concordancecorrelation coefficient (for continuous variables) lt090 poor090ndash095 moderate 095ndash099 substantial and gt099 almostperfect [36] Differences in body composition between thoseassessed by BIA methods and DXAmeasures were examinedusing a paired t-test Bland-Altman analysis was applied todetermine the agreement between BIA and DXA measure-ments [37] The agreement between methods is representedby the bias estimated by the mean difference and the SD ofthe differences Therefore the 95 limits of agreement werecalculated as the mean difference plusmn196 SD of the differencesbetween methods The pure error in BF between each BIA

device with DXA measurement was assessed using the totalerror (TE) as follows

TE = radicsum(119884 minus 119884

1015840

)

2

119873

(1)

where 119884 = observed values 1198841015840 = predicted values and 119873 =the number of participants in the sample These statisticalanalyses were performed separately on boys and girls

3 Results

31 Characteristics of Participants A total of 255 (128 girlsand 127 boys) Chinese children and adolescents participatedin the study Table 1 presents their physical characteristics andbody compartments (FM LTM FFM BMC and BF) Nosignificant differences in mean age were observed betweengenders The mean body weight height and BMI of boyswere higher than those of girls The mean of the individualbody compartments was significantly different between boysand girls (Table 1) 803 of boys and 828 of girls hadentered puberty (pubertal stage ge2 for pubic hair or breasts(girls)genitalia (boys))

Table 2 shows Linrsquos concordance correlation coefficientsbetween parameters of body composition assessed by dif-ferent BIA devices and DXA methods In boys the FFMdetermined by all BIA devices showed substantial agreementswith that determined by DXA (rc values within 0951ndash0979)In girls the FFM determined by InBody 520 showed substan-tial agreement with that determined by DXA (rc = 0957)moderate agreements for FFM were found on Biodynamics-310 (rc = 0909) and Tanita BC-545 (rc = 0910) and pooragreement for FFMwas found onTanita TBF 543 (rc= 0897)Linrsquos concordance correlation coefficients of estimated BFbetween Biodynamics-310 Tanita TBF-543 Tanita BC-545and DXA showed poor agreements (rc values within 0747ndash0863) for both genders Moderate agreements for BF werefound between DXA and InBody 520 measurements (rc =0926 in boys and 0912 in girls)

In boys significant differences in BF were foundbetween DXA and Model B and Model C (paired t-test 119901 lt005) The mean bias of BF for Model B was 115 (SD =595) with 95 limits of agreement minus1050ndash1281 ModelC underestimates BF the mean bias was minus167 (SD =48) with 95 limits of agreement minus1108ndash774 Howeverno significant BF differences were found betweenModel AModel D and DXA Model A and Model D BIA analyzersshowed a significant positive correlation between the bias andaverage BF between BIA and DXA indicating that BFwas underestimated at low values and overestimated at highvalues in the present sample (Table 3)

In girls significant differences in BF were observedbetween Model B Model C Model D and DXA BFwas underestimated by the three BIA devices The meanbiases were minus263 for Model B (SD = 323 95 limitsof agreement minus896ndash370) minus147 for Model D (SD =299 95 limits of agreementminus732ndash440) andminus205 forModel C (SD= 321 95 limits of agreementminus834ndash424)

4 BioMed Research International

Table 1 Physical characteristics and body compositions of the study population

Boys (119899 = 127) Girls (119899 = 128)

Mean (SD) Range Mean (SD) RangeLower Upper Lower Upper

Age (years) 138 (29) 90 195 137 (28) 90 194Height (cm) 1618 (140) 1305 1860 1543 (98) 1266 1720Weight (kg) 559 (170) 275 1081 474 (113) 235 971BMI (kg2m) 210 (42) 136 371 197 (32) 133 348LTM-DXA (kg) 410 (112) 210 653 310 (55) 187 543BMC-DXA (kg) 20 (07) 09 38 17 (05) 08 27FM-DXA (kg) 127 (95) 33 531 143 (68) 27 406FFM-DXA (kg) 430 (118) 221 690 328 (59) 195 568FFM-Model A (kg) 438 (127) 218 774 339 (65) 186 612FFM-Model B (kg) 432 (117) 208 732 347 (55) 206 517FFM-Model C (kg) 445 (120) 226 732 345 (55) 206 516FFM-Model D (kg) 443 (122) 228 720 340 (59) 200 553BF-DXA () 208 (99) 61 491 279 (77) 118 465BF-Model A () 214 (70) 49 396 274 (60) 136 421BF-Model B () 219 (69) 101 453 253 (78) 93 494BF-Model C () 190 (90) 50 547 258 (77) 103 476BF-Model D () 202 (85) 62 473 263 (81) 64 448Note data are presented as mean (SD) BMI body mass index LTM-DXA lean tissue mass measured by dual-energy X-ray absorptiometry BMC-DXA bonemineral content measured by dual-energy X-ray absorptiometry FM-DXA fat mass measured by dual-energy X-ray absorptiometry BF-DXA body fatpercentage measured by dual-energy X-ray absorptiometry FFM-DXA fat-free mass calculated as the sum of LTM and BMC Model A Biodynamics-310Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Table 2 Linrsquos concordance correlation coefficients between param-eters assessed by different BIA and DXA measurement

BIA devicesModel A Model B Model C Model D

Boys(119899 = 127)

BF-DXA 0765 0750 0855 0926FFM-DXA 0953 0953 0951 0979

Girls(119899 = 128)

BF-DXA 0747 0863 0881 0912FFM-DXA 0909 0897 0910 0957

Note BMI body mass index BF-DXA body fat percentage measured bydual-energy X-ray absorptiometry FFM-DXA fat-free mass measured bydual-energy X-ray absorptiometryModel A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Model A and DXA showed no significant differences of BFhowever the bias and the averageBF betweenBIA andDXAhad a significant positive correlation indicating that the BFwas underestimated at low values and overestimated at highvalues in the present sample (Table 3) Model A Model CandModel D overestimated the FFM in boys All BIA devicesoverestimated the FFM in girls (Table 3)

4 Discussion

Chinese children and adolescents are at increased risk forobesity and associated health risks Determining weightstatus throughmeasurement of body fat percentage is impor-tant in diagnosing childhood obesity Several methods such

as DXA Bod Pod and underwater weighing can provideaccurate body fat measurement However these measure-ments are often inaccessible to the larger scale populationsurvey The search for an accurate diagnostic instrumentfor measuring body fat that is practical and inexpensiveis a challenge for researchers In this sense commercialBIA has become a popular method due to its ease of useand low cost [38] Numerous different BIA commercialdevices can measure body composition but their suitabilityfor Chinese children and adolescents is still unclear Theaccuracy of BIA measurement may be influenced by severalfactors such as body shape hydration status and ethnicity[9] Previous studies have suggested that body compositionfrom commercial BIA devices is derived frommanufacturersrsquoequations which are often not validated for specific ethnicgroups [38ndash40] These discrepancies may be associated withdifferences among the devices and ethnic differences on bodycomposition [10 26] We sought investigating the validityand diagnostics accuracy of several different BIA devicesin assessing excessive body fat in Chinese children andadolescentsThe findings of the current study may determineuseful BIA devices for local health practitioners in assessingobese children

Previous validation studies in children and adolescentshave found that different BIA (foot-foot hand-foot andmultiple frequencies) devices usingmanufacturersrsquo equationstend to overestimate or underestimate BF in relation tocriteria measurement [24ndash28 32] Single-frequency foot-foot or hand-foot BIA devices generally tend to underes-timate or overestimate BF in children and adolescents

BioMed Research International 5

Table 3 Cross-validation of four BIA devices (embedded equations) for the perdition of BFFFM in Chinese children and adolescents

Gender BIAdevices

BF-DXA

BF-BIA

FFM-DXA

FFM-BIA 119901 119905-test Bias (SD) TE 119903

119910minus1199101015840119910

95limits ofagreement

Boys(119899 = 127)

Model A 2077 2134 mdash mdash 0268 058 (586) 586 0513lowastlowast minus1096 to1212

Model B 2077 2192 mdash mdash 0031 115 (595) 604 0528lowastlowast minus1051 to1281

Model C 2077 1904 mdash mdash lt0001 minus167(480) 506 0162 minus1108 to

774

Model D 2077 2022 mdash mdash 0246 minus037(351) 351 0365lowastlowast minus725 to

651

Model A mdash mdash 4302 kg 4380 kg 0019 078(370) kg 377 kg 0261lowastlowast minus647 kg to

803 kg

Model B mdash mdash 4302 kg 4318 kg 0608 016(360) kg 360 kg minus0027 minus690 kg to

722 kg

Model C mdash mdash 4302 kg 4450 kg lt0001 154(335) kg 360 kg 0024 minus503 kg to

811 kg

Model D mdash mdash 4302 kg 4431 kg lt0001 102(217) kg 239 kg 0197lowast minus323 kg to

527 kg

Girls(119899 = 128)

Model A 2785 2737 mdash mdash 0266 minus048(488) 488 0359lowastlowast minus1004 to

908

Model B 2785 2522 mdash mdash lt0001 minus263(323) 415 minus0049 minus896 to

370

Model C 2785 2581 mdash mdash lt0001 minus205(321) 379 0012 minus834 to

424

Model D 2785 2634 mdash mdash lt0001 minus147(299) 331 minus0147 minus733 to

439

Model A mdash mdash 3277 kg 3389 kg lt0001 111 (242) kg 266 kg 0245lowastlowast minus363 kg to585 kg

Model B mdash mdash 3277 kg 3474 kg lt0001 197(177) kg 264 kg minus0255lowastlowast minus150 kg to

544 kg

Model C mdash mdash 3277 kg 3449 kg lt0001 172(177) kg 247 kg minus0219lowastlowast minus175 kg to

519 kg

Model D mdash mdash 3277 kg 3403 kg lt0001 125(122) kg 168 kg minus0022 minus114 kg to

364 kgNote BF body fat percentage FFM fat-free mass DXA dual-energy X-ray absorptiometry BIA bioelectrical impedance analysisBias value obtained from BIA measurement minus DXA measurement SD standard deviation119903119910minus1199101015840119910

Pearsonrsquos correlation coefficient for the relationship between averaged BFFFM [(value of DXA + value of BIA)2] and the biaslowast

119901 lt 005 lowastlowast119901 lt 001Model A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D Inbody 520

In the current study although high correlations in bodycomposition measurements between BIA devices and DXAwere found foot-foot (TBF-543) BIA device overestimatedBF in boys and underestimated BF in girls in comparisonwith DXA measurement Most of previous studies foundthat Tanita BIA devices underestimated BF in both genders[24 25 27 37] with which the findings of the current studywas consistent for girls However in current boys TanitaTBF-543 overestimated BF (bias = 115 plusmn 559) withnegative correlation between bias and averaged BF betweenBIA and DXA measurements The discrepancies may beexplained by different testing population ethnicity and use ofTanita BIA devices with different models The current studyuses a heterogeneous and relatively larger sample coveringa wide range of ages and body compositions according toage and gender specific BMI distributions among Chinese

children To date two studies were conducted to investigatethe accuracy of foot-foot BIA consumer devices inmeasuringbody fat percentage in Chinese children and adolescentscompared with DXA measurement [21 23] however thesestudies had several limitations Lu et al measured BF of64 Chinese obese children aged 10ndash17 years using Tanita-401and no significant differences were found in BF betweenfoot-foot BIA and DXA [21] Moreover BIA overestimatedthe BF in seriously obese children Sung et al foundthat BIA (Tanita TBF 401) slightly overestimated BF in7ndash18-year-old Chinese children (119899 = 49) however theBF bias did not reach statistical significance [23] The twostudies used smaller samples with a wider age range toinvestigate the validity of foot-foot measured BF whichmay significantly explain the different findings between thecurrent and previous studies

6 BioMed Research International

In addition despite no mean bias was found for theBF between the hand-foot BIA device (Biodynamics-310)and DXA measurement in both genders the Bland-Altmananalysis showed that the two BIA devices overestimatedmeasurements at a lowBF and underestimated them at highBF Previous studies investigated the validity of hand-footBIA devices in assessing BF in children [22 27 28 41]Several studies had the same results with that of the presentstudy [22 41]

The eight-tactile electrodes with multiple frequenciesBIA device could estimate body composition in adults moreaccurately than a single-frequency BIA device [17 42] How-ever contradictory findings were revealed on the validityof multiple frequencies BIA devices in measuring bodycomposition in children and adolescents Two studies didnot show statistical significance in BF between multiplefrequencies BIA (InBody) and DXA measurements [3031] However Jensky-Squires et al found that InBody 320overestimated BF in female adolescents from underwaterweighing [10] In two other studies Prins et al found thatTanita BC-418 overestimated BF determined by deuteriumdilution in African children [43] Sluyter et al found thatthe same BIA device underestimated BF in a multiethnicgroup of adolescents Moreover the BIA device tended tounderestimate measurements in lower BF and overestimatethem in higher BF individuals [32] In the current studytwomultiple frequencies BIA devices (InBody 520 and TanitaBC-545) were also used to evaluate body composition TanitaBC-545 underestimated BF in both genders and InBody520 underestimated BF in girls In boys no bias was foundin BF between InBody 520 and DXA measurement but asignificant positive correlation was found between the biasand averaged BF between BIA and DXA InBody 520 thusoverestimated BF at high values and overestimated it at lowvalues in boys For Tanita multiple frequencies BIA devicesthe results of the current and previous studies are consistentThedevices underestimatedBF in children and adolescentsFor the InBody multiple BIA devices the varying results mayhave been caused by different population and ethnicity

Moreover note that although themean biasmay be smallof higher relevance are the 95 limits of agreement whichseem to more accurately reflect the validity of measurementin an individual Numerous studies found that the BIA hasquite wide limits of agreement with criterionmeasurement ofBF [24 26 27 41 44] therefore BIA is not interchangeablewith criterion measurement for body composition As suchis true even for Chinese children and adolescents as in thecurrent study the 95 limits of agreement of all BIA devicesshowed wide ranges in agreement with previous studiesSung et al found that the 95 limits of agreement in BFwere narrow (from minus392 to 061) between foot-foot BIAand DXA in 17 obese and 32 nonobese Chinese childrenaged 7ndash18 years [23] The current study had larger 95 limitsof agreement in the same ethnicity and similar age rangeof population The discrepancy may be partly explained bydifferent samples size sampling method and DXA instru-ments used (Hologic QDR-4500 and GE Lunar Prodigy)Previous studies suggested that if researchers intend tomeasure body composition in large-scale population surveys

then BIAmeasurementmay provide an efficient alternative tolaboratory measurement of body composition when accurateinstruments are unavailable [45] BIA measurement shouldbe used with caution in evaluating or monitoring changes inindividual body composition in healthy and clinical setting[38]

5 Conclusion

Using embedded equations in BIA devices should be vali-dated in assessing the body composition of Chinese childrenand adolescents aged 9ndash19 years old All BIA devices are notdirectly interchanged with DXA measurement in assessingindividual BF and assessing changes of body compositionowing to the wide limits of agreement

Disclaimer

The authors alone are responsible for the content and writingof the paper

Conflict of Interests

The authors report no conflict of interests

Acknowledgments

This study was supported by the Hong Kong Association forthe Study of Obesity (2011) and Natural Science Foundationof Shanghai (15ZR1439300) The funders had no role in studydesign collection analysis and interpretation of data writingof paper or decision to submit for paper

References

[1] World Health Organization ldquoObesity and overweightrdquo FactSheet No 311 2011 httpwwwwhointmediacentrefact-sheetsfs311enindexhtml

[2] T Lobstein L Baur and R Uauy ldquoObesity in children andyoung people a crisis in public healthrdquo Obesity ReviewsSupplement vol 5 no 1 pp 4ndash104 2004

[3] Y Wu ldquoOverweight and obesity in Chinardquo British MedicalJournal vol 333 no 7564 pp 362ndash363 2006

[4] Y P Li E G Schouten X Q Hu Z H Cui D C Luan and GS Ma ldquoObesity prevalence and time trend among youngstersin China 1982ndash2002rdquo Asia Pacific Journal of Clinical Nutritionvol 17 no 1 pp 131ndash137 2008

[5] J J Reilly E Methven Z C McDowell et al ldquoHealth conse-quences of obesityrdquo Archives of Disease in Childhood vol 88no 9 pp 748ndash752 2003

[6] W H Dietz ldquoHealth consequences of obesity in youth child-hood predictors of adult diseaserdquo Pediatrics vol 101 no 3 pp518ndash525 1998

[7] J C KWells andM S Fewtrell ldquoMeasuring body compositionrdquoArchives of Disease in Childhood vol 91 no 7 pp 612ndash617 2006

[8] J C K Wells and M S Fewtrell ldquoIs body compositionimportant for paediatriciansrdquoArchives of Disease in Childhoodvol 93 no 2 pp 168ndash172 2008

BioMed Research International 7

[9] V H Heyward and D R Wagner Applied Body CompositionAssessment Human Kinetics Champaign Ill USA 2004

[10] N E Jensky-Squires CMDieli-Conwright A Rossuello DNErceg SMcCauley and T E Schroeder ldquoValidity and reliabilityof body composition analysers in children and adultsrdquo BritishJournal of Nutrition vol 100 no 4 pp 859ndash865 2008

[11] M Helba and L A Binkovitz ldquoPediatric body compositionanalysis with dual-energy X-ray absorptiometryrdquo PediatricRadiology vol 39 no 7 pp 647ndash656 2009

[12] A B Sopher J C Thornton J Wang R N Pierson Jr S BHeymsfield and M Horlick ldquoMeasurement of percentage ofbody fat in 411 children and adolescents a comparison of dual-energy X-ray absorptiometry with a four-compartment modelrdquoPediatrics vol 113 no 5 pp 1285ndash1290 2004

[13] T G Lohman and C Zhao ldquoDual-energy X-ray adsorptiom-etryrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 63ndash77 HumanKinetics Champaign Ill USA 2nd edition 2005

[14] E O Diaz J Villar M Immink and T Gonzales ldquoBioimpe-dance or anthropometryrdquo European Journal of Clinical Nutri-tion vol 43 no 2 pp 129ndash137 1989

[15] S Demura S Sato and T Kitabayashi ldquoPercentage of total bodyfat as estimated by three automatic bioelectrical impedanceanalyzersrdquo Journal of Physiological Anthropology and AppliedHuman Science vol 23 no 3 pp 93ndash99 2004

[16] A Pietrobelli F Rubiano M-P St-Onge and S B HeymsfieldldquoNew bioimpedance analysis system improved phenotypingwith whole-body analysisrdquo European Journal of Clinical Nutri-tion vol 58 no 11 pp 1479ndash1484 2004

[17] R Martinoli E I Mohamed C Maiolo et al ldquoTotal body waterestimation using bioelectrical impedance a meta-analysis ofthe data available in the literaturerdquo Acta Diabetologica vol 40supplement 1 pp S203ndashS206 2003

[18] K R Foster andH C Lukaski ldquoWhole-body impedancemdashwhatdoes it measurerdquo The American Journal of Clinical Nutritionvol 64 no 3 pp 388Sndash396S 1996

[19] W C Chumlea and S S Sun ldquoBioelectrical impedance anal-ysisrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 79ndash88 HumanKinetics Champaign Ill USA 2nd edition 2005

[20] L Iacopino A Andreoli I Innocente et al ldquoUse of foot-to-foot bioelectrical impedance analysis in childrenrdquo ActaDiabetologica vol 40 supplement 1 pp S210ndashS211 2003

[21] K Lu B Quach T K Tong and PW C Lau ldquoValidation of leg-to-leg bio-impedance analysis for assessing body compositionin obese Chinese childrenrdquo Journal of Exercise Scienceamp Fitnessvol 1 no 2 pp 97ndash103 2003

[22] A Mooney L Kelsey G W Fellingham et al ldquoAssessing bodycomposition of children and adolescents using dual-energyX-ray absorptiometry skinfolds and electrical impedancerdquoMeasurement in Physical Education and Exercise Science vol 15no 1 pp 2ndash17 2011

[23] R Y T Sung P Lau CW Yu P KW Lam and E A S NelsonldquoMeasurement of body fat using leg to leg bioimpedancerdquoArchives ofDisease inChildhood vol 85 no 3 pp 263ndash267 2001

[24] C Azcona N Koek and G Fruhbeck ldquoFat mass by air-displacement plethysmography and impedance in obesenon-obese children and adolescentsrdquo International Journal of Pedi-atric Obesity vol 1 no 3 pp 176ndash182 2006

[25] G S Goldfield P Cloutier R Mallory D Prudrsquohomme TParker and E Doucet ldquoValidity of foot-to-foot bioelectrical

impedance analysis in overweight and obese children andparentsrdquoThe Journal of SportsMedicine and Physical Fitness vol46 no 3 pp 447ndash453 2006

[26] J Hosking B S Metcalf A N Jeffery L D Voss and TJ Wilkin ldquoValidation of foot-to-foot bioelectrical impedanceanalysis with dual-energy X-ray absorptiometry in the assess-ment of body composition in young children the EarlyBirdcohortrdquo British Journal of Nutrition vol 96 no 6 pp 1163ndash11682006

[27] S Lazzer Y Boirie M Meyer and M Vermorel ldquoEvaluationof two foot-to-foot bioelectrical impedance analysers to assessbody composition in overweight and obese adolescentsrdquo BritishJournal of Nutrition vol 90 no 5 pp 987ndash992 2003

[28] L Parker J J Reilly C Slater J C K Wells and Y PitsiladisldquoValidity of six field and laboratory methods for measurementof body composition in boysrdquoObesity Research vol 11 no 7 pp852ndash858 2003

[29] J J Reilly K Gerasimidis N Paparacleous et al ldquoValidationof dual-energy x-ray absorptiometry and foot-foot impedanceagainst deuterium dilution measures of fatness in childrenrdquoInternational Journal of Pediatric Obesity vol 5 no 1 pp 111ndash115 2010

[30] J S Lim J S Hwang J A Lee et al ldquoCross-calibration ofmulti-frequency bioelectrical impedance analysis with eight-point tactile electrodes and dual-energy X-ray absorptiometryfor assessment of body composition in healthy children aged6-18 yearsrdquo Pediatrics International vol 51 no 2 pp 263ndash2682009

[31] O K Yu Y K Rhee T S Park and Y S Cha ldquoComparisons ofobesity assessments in over-weight elementary students usinganthropometry BIA CT and DEXArdquo Nutrition Research andPractice vol 4 no 2 pp 128ndash135 2010

[32] J D Sluyter D Schaaf R K R Scragg and L D PlankldquoPrediction of fatness by standing 8-electrode bioimpedance amultiethnic adolescent populationrdquo Obesity vol 18 no 1 pp183ndash189 2010

[33] K P Navder Q He X Zhang et al ldquoRelationship between bodymass index and adiposity in prepubertal children ethnic andgeographic comparisons between New York City and Jinan City(China)rdquo Journal of Applied Physiology vol 107 no 2 pp 488ndash493 2009

[34] N J Shaw N J Crabtree M S Kibirige and J N FordhamldquoEthnic and gender differences in body fat in British schoolchil-dren as measured by DXArdquo Archives of Disease in Childhoodvol 92 no 10 pp 872ndash875 2007

[35] M A Stone L Williams S Chatterjee M J Davies and KKhunti ldquoEthnic differences in body composition in adoles-centsrdquo Primary Care Diabetes vol 2 no 1 pp 55ndash57 2008

[36] G B McBride ldquoA proposal for strength-of-agreement criteriafor Linrsquos concordance correlation coefficientrdquo NIWA ClientReport HAM2005-062 2005

[37] J M Bland and D G Altman ldquoStatistical methods for assessingagreement between two methods of clinical measurementrdquoTheLancet vol 1 no 8476 pp 307ndash310 1986

[38] M Dehghan and A T Merchant ldquoIs bioelectrical impedanceaccurate for use in large epidemiological studiesrdquo NutritionJournal vol 7 no 1 article 26 2008

[39] B SMcClanahanM B Stockton J Q Lanctot et al ldquoMeasure-ment of body composition in 810-year-old African-Americangirls a comparison of dual-energy X-ray absorptiometry andfoot-to-foot bioimpedance methodsrdquo International Journal ofPediatric Obesity vol 4 no 4 pp 389ndash396 2009

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

4 BioMed Research International

Table 1 Physical characteristics and body compositions of the study population

Boys (119899 = 127) Girls (119899 = 128)

Mean (SD) Range Mean (SD) RangeLower Upper Lower Upper

Age (years) 138 (29) 90 195 137 (28) 90 194Height (cm) 1618 (140) 1305 1860 1543 (98) 1266 1720Weight (kg) 559 (170) 275 1081 474 (113) 235 971BMI (kg2m) 210 (42) 136 371 197 (32) 133 348LTM-DXA (kg) 410 (112) 210 653 310 (55) 187 543BMC-DXA (kg) 20 (07) 09 38 17 (05) 08 27FM-DXA (kg) 127 (95) 33 531 143 (68) 27 406FFM-DXA (kg) 430 (118) 221 690 328 (59) 195 568FFM-Model A (kg) 438 (127) 218 774 339 (65) 186 612FFM-Model B (kg) 432 (117) 208 732 347 (55) 206 517FFM-Model C (kg) 445 (120) 226 732 345 (55) 206 516FFM-Model D (kg) 443 (122) 228 720 340 (59) 200 553BF-DXA () 208 (99) 61 491 279 (77) 118 465BF-Model A () 214 (70) 49 396 274 (60) 136 421BF-Model B () 219 (69) 101 453 253 (78) 93 494BF-Model C () 190 (90) 50 547 258 (77) 103 476BF-Model D () 202 (85) 62 473 263 (81) 64 448Note data are presented as mean (SD) BMI body mass index LTM-DXA lean tissue mass measured by dual-energy X-ray absorptiometry BMC-DXA bonemineral content measured by dual-energy X-ray absorptiometry FM-DXA fat mass measured by dual-energy X-ray absorptiometry BF-DXA body fatpercentage measured by dual-energy X-ray absorptiometry FFM-DXA fat-free mass calculated as the sum of LTM and BMC Model A Biodynamics-310Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Table 2 Linrsquos concordance correlation coefficients between param-eters assessed by different BIA and DXA measurement

BIA devicesModel A Model B Model C Model D

Boys(119899 = 127)

BF-DXA 0765 0750 0855 0926FFM-DXA 0953 0953 0951 0979

Girls(119899 = 128)

BF-DXA 0747 0863 0881 0912FFM-DXA 0909 0897 0910 0957

Note BMI body mass index BF-DXA body fat percentage measured bydual-energy X-ray absorptiometry FFM-DXA fat-free mass measured bydual-energy X-ray absorptiometryModel A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D InBody 520

Model A and DXA showed no significant differences of BFhowever the bias and the averageBF betweenBIA andDXAhad a significant positive correlation indicating that the BFwas underestimated at low values and overestimated at highvalues in the present sample (Table 3) Model A Model CandModel D overestimated the FFM in boys All BIA devicesoverestimated the FFM in girls (Table 3)

4 Discussion

Chinese children and adolescents are at increased risk forobesity and associated health risks Determining weightstatus throughmeasurement of body fat percentage is impor-tant in diagnosing childhood obesity Several methods such

as DXA Bod Pod and underwater weighing can provideaccurate body fat measurement However these measure-ments are often inaccessible to the larger scale populationsurvey The search for an accurate diagnostic instrumentfor measuring body fat that is practical and inexpensiveis a challenge for researchers In this sense commercialBIA has become a popular method due to its ease of useand low cost [38] Numerous different BIA commercialdevices can measure body composition but their suitabilityfor Chinese children and adolescents is still unclear Theaccuracy of BIA measurement may be influenced by severalfactors such as body shape hydration status and ethnicity[9] Previous studies have suggested that body compositionfrom commercial BIA devices is derived frommanufacturersrsquoequations which are often not validated for specific ethnicgroups [38ndash40] These discrepancies may be associated withdifferences among the devices and ethnic differences on bodycomposition [10 26] We sought investigating the validityand diagnostics accuracy of several different BIA devicesin assessing excessive body fat in Chinese children andadolescentsThe findings of the current study may determineuseful BIA devices for local health practitioners in assessingobese children

Previous validation studies in children and adolescentshave found that different BIA (foot-foot hand-foot andmultiple frequencies) devices usingmanufacturersrsquo equationstend to overestimate or underestimate BF in relation tocriteria measurement [24ndash28 32] Single-frequency foot-foot or hand-foot BIA devices generally tend to underes-timate or overestimate BF in children and adolescents

BioMed Research International 5

Table 3 Cross-validation of four BIA devices (embedded equations) for the perdition of BFFFM in Chinese children and adolescents

Gender BIAdevices

BF-DXA

BF-BIA

FFM-DXA

FFM-BIA 119901 119905-test Bias (SD) TE 119903

119910minus1199101015840119910

95limits ofagreement

Boys(119899 = 127)

Model A 2077 2134 mdash mdash 0268 058 (586) 586 0513lowastlowast minus1096 to1212

Model B 2077 2192 mdash mdash 0031 115 (595) 604 0528lowastlowast minus1051 to1281

Model C 2077 1904 mdash mdash lt0001 minus167(480) 506 0162 minus1108 to

774

Model D 2077 2022 mdash mdash 0246 minus037(351) 351 0365lowastlowast minus725 to

651

Model A mdash mdash 4302 kg 4380 kg 0019 078(370) kg 377 kg 0261lowastlowast minus647 kg to

803 kg

Model B mdash mdash 4302 kg 4318 kg 0608 016(360) kg 360 kg minus0027 minus690 kg to

722 kg

Model C mdash mdash 4302 kg 4450 kg lt0001 154(335) kg 360 kg 0024 minus503 kg to

811 kg

Model D mdash mdash 4302 kg 4431 kg lt0001 102(217) kg 239 kg 0197lowast minus323 kg to

527 kg

Girls(119899 = 128)

Model A 2785 2737 mdash mdash 0266 minus048(488) 488 0359lowastlowast minus1004 to

908

Model B 2785 2522 mdash mdash lt0001 minus263(323) 415 minus0049 minus896 to

370

Model C 2785 2581 mdash mdash lt0001 minus205(321) 379 0012 minus834 to

424

Model D 2785 2634 mdash mdash lt0001 minus147(299) 331 minus0147 minus733 to

439

Model A mdash mdash 3277 kg 3389 kg lt0001 111 (242) kg 266 kg 0245lowastlowast minus363 kg to585 kg

Model B mdash mdash 3277 kg 3474 kg lt0001 197(177) kg 264 kg minus0255lowastlowast minus150 kg to

544 kg

Model C mdash mdash 3277 kg 3449 kg lt0001 172(177) kg 247 kg minus0219lowastlowast minus175 kg to

519 kg

Model D mdash mdash 3277 kg 3403 kg lt0001 125(122) kg 168 kg minus0022 minus114 kg to

364 kgNote BF body fat percentage FFM fat-free mass DXA dual-energy X-ray absorptiometry BIA bioelectrical impedance analysisBias value obtained from BIA measurement minus DXA measurement SD standard deviation119903119910minus1199101015840119910

Pearsonrsquos correlation coefficient for the relationship between averaged BFFFM [(value of DXA + value of BIA)2] and the biaslowast

119901 lt 005 lowastlowast119901 lt 001Model A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D Inbody 520

In the current study although high correlations in bodycomposition measurements between BIA devices and DXAwere found foot-foot (TBF-543) BIA device overestimatedBF in boys and underestimated BF in girls in comparisonwith DXA measurement Most of previous studies foundthat Tanita BIA devices underestimated BF in both genders[24 25 27 37] with which the findings of the current studywas consistent for girls However in current boys TanitaTBF-543 overestimated BF (bias = 115 plusmn 559) withnegative correlation between bias and averaged BF betweenBIA and DXA measurements The discrepancies may beexplained by different testing population ethnicity and use ofTanita BIA devices with different models The current studyuses a heterogeneous and relatively larger sample coveringa wide range of ages and body compositions according toage and gender specific BMI distributions among Chinese

children To date two studies were conducted to investigatethe accuracy of foot-foot BIA consumer devices inmeasuringbody fat percentage in Chinese children and adolescentscompared with DXA measurement [21 23] however thesestudies had several limitations Lu et al measured BF of64 Chinese obese children aged 10ndash17 years using Tanita-401and no significant differences were found in BF betweenfoot-foot BIA and DXA [21] Moreover BIA overestimatedthe BF in seriously obese children Sung et al foundthat BIA (Tanita TBF 401) slightly overestimated BF in7ndash18-year-old Chinese children (119899 = 49) however theBF bias did not reach statistical significance [23] The twostudies used smaller samples with a wider age range toinvestigate the validity of foot-foot measured BF whichmay significantly explain the different findings between thecurrent and previous studies

6 BioMed Research International

In addition despite no mean bias was found for theBF between the hand-foot BIA device (Biodynamics-310)and DXA measurement in both genders the Bland-Altmananalysis showed that the two BIA devices overestimatedmeasurements at a lowBF and underestimated them at highBF Previous studies investigated the validity of hand-footBIA devices in assessing BF in children [22 27 28 41]Several studies had the same results with that of the presentstudy [22 41]

The eight-tactile electrodes with multiple frequenciesBIA device could estimate body composition in adults moreaccurately than a single-frequency BIA device [17 42] How-ever contradictory findings were revealed on the validityof multiple frequencies BIA devices in measuring bodycomposition in children and adolescents Two studies didnot show statistical significance in BF between multiplefrequencies BIA (InBody) and DXA measurements [3031] However Jensky-Squires et al found that InBody 320overestimated BF in female adolescents from underwaterweighing [10] In two other studies Prins et al found thatTanita BC-418 overestimated BF determined by deuteriumdilution in African children [43] Sluyter et al found thatthe same BIA device underestimated BF in a multiethnicgroup of adolescents Moreover the BIA device tended tounderestimate measurements in lower BF and overestimatethem in higher BF individuals [32] In the current studytwomultiple frequencies BIA devices (InBody 520 and TanitaBC-545) were also used to evaluate body composition TanitaBC-545 underestimated BF in both genders and InBody520 underestimated BF in girls In boys no bias was foundin BF between InBody 520 and DXA measurement but asignificant positive correlation was found between the biasand averaged BF between BIA and DXA InBody 520 thusoverestimated BF at high values and overestimated it at lowvalues in boys For Tanita multiple frequencies BIA devicesthe results of the current and previous studies are consistentThedevices underestimatedBF in children and adolescentsFor the InBody multiple BIA devices the varying results mayhave been caused by different population and ethnicity

Moreover note that although themean biasmay be smallof higher relevance are the 95 limits of agreement whichseem to more accurately reflect the validity of measurementin an individual Numerous studies found that the BIA hasquite wide limits of agreement with criterionmeasurement ofBF [24 26 27 41 44] therefore BIA is not interchangeablewith criterion measurement for body composition As suchis true even for Chinese children and adolescents as in thecurrent study the 95 limits of agreement of all BIA devicesshowed wide ranges in agreement with previous studiesSung et al found that the 95 limits of agreement in BFwere narrow (from minus392 to 061) between foot-foot BIAand DXA in 17 obese and 32 nonobese Chinese childrenaged 7ndash18 years [23] The current study had larger 95 limitsof agreement in the same ethnicity and similar age rangeof population The discrepancy may be partly explained bydifferent samples size sampling method and DXA instru-ments used (Hologic QDR-4500 and GE Lunar Prodigy)Previous studies suggested that if researchers intend tomeasure body composition in large-scale population surveys

then BIAmeasurementmay provide an efficient alternative tolaboratory measurement of body composition when accurateinstruments are unavailable [45] BIA measurement shouldbe used with caution in evaluating or monitoring changes inindividual body composition in healthy and clinical setting[38]

5 Conclusion

Using embedded equations in BIA devices should be vali-dated in assessing the body composition of Chinese childrenand adolescents aged 9ndash19 years old All BIA devices are notdirectly interchanged with DXA measurement in assessingindividual BF and assessing changes of body compositionowing to the wide limits of agreement

Disclaimer

The authors alone are responsible for the content and writingof the paper

Conflict of Interests

The authors report no conflict of interests

Acknowledgments

This study was supported by the Hong Kong Association forthe Study of Obesity (2011) and Natural Science Foundationof Shanghai (15ZR1439300) The funders had no role in studydesign collection analysis and interpretation of data writingof paper or decision to submit for paper

References

[1] World Health Organization ldquoObesity and overweightrdquo FactSheet No 311 2011 httpwwwwhointmediacentrefact-sheetsfs311enindexhtml

[2] T Lobstein L Baur and R Uauy ldquoObesity in children andyoung people a crisis in public healthrdquo Obesity ReviewsSupplement vol 5 no 1 pp 4ndash104 2004

[3] Y Wu ldquoOverweight and obesity in Chinardquo British MedicalJournal vol 333 no 7564 pp 362ndash363 2006

[4] Y P Li E G Schouten X Q Hu Z H Cui D C Luan and GS Ma ldquoObesity prevalence and time trend among youngstersin China 1982ndash2002rdquo Asia Pacific Journal of Clinical Nutritionvol 17 no 1 pp 131ndash137 2008

[5] J J Reilly E Methven Z C McDowell et al ldquoHealth conse-quences of obesityrdquo Archives of Disease in Childhood vol 88no 9 pp 748ndash752 2003

[6] W H Dietz ldquoHealth consequences of obesity in youth child-hood predictors of adult diseaserdquo Pediatrics vol 101 no 3 pp518ndash525 1998

[7] J C KWells andM S Fewtrell ldquoMeasuring body compositionrdquoArchives of Disease in Childhood vol 91 no 7 pp 612ndash617 2006

[8] J C K Wells and M S Fewtrell ldquoIs body compositionimportant for paediatriciansrdquoArchives of Disease in Childhoodvol 93 no 2 pp 168ndash172 2008

BioMed Research International 7

[9] V H Heyward and D R Wagner Applied Body CompositionAssessment Human Kinetics Champaign Ill USA 2004

[10] N E Jensky-Squires CMDieli-Conwright A Rossuello DNErceg SMcCauley and T E Schroeder ldquoValidity and reliabilityof body composition analysers in children and adultsrdquo BritishJournal of Nutrition vol 100 no 4 pp 859ndash865 2008

[11] M Helba and L A Binkovitz ldquoPediatric body compositionanalysis with dual-energy X-ray absorptiometryrdquo PediatricRadiology vol 39 no 7 pp 647ndash656 2009

[12] A B Sopher J C Thornton J Wang R N Pierson Jr S BHeymsfield and M Horlick ldquoMeasurement of percentage ofbody fat in 411 children and adolescents a comparison of dual-energy X-ray absorptiometry with a four-compartment modelrdquoPediatrics vol 113 no 5 pp 1285ndash1290 2004

[13] T G Lohman and C Zhao ldquoDual-energy X-ray adsorptiom-etryrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 63ndash77 HumanKinetics Champaign Ill USA 2nd edition 2005

[14] E O Diaz J Villar M Immink and T Gonzales ldquoBioimpe-dance or anthropometryrdquo European Journal of Clinical Nutri-tion vol 43 no 2 pp 129ndash137 1989

[15] S Demura S Sato and T Kitabayashi ldquoPercentage of total bodyfat as estimated by three automatic bioelectrical impedanceanalyzersrdquo Journal of Physiological Anthropology and AppliedHuman Science vol 23 no 3 pp 93ndash99 2004

[16] A Pietrobelli F Rubiano M-P St-Onge and S B HeymsfieldldquoNew bioimpedance analysis system improved phenotypingwith whole-body analysisrdquo European Journal of Clinical Nutri-tion vol 58 no 11 pp 1479ndash1484 2004

[17] R Martinoli E I Mohamed C Maiolo et al ldquoTotal body waterestimation using bioelectrical impedance a meta-analysis ofthe data available in the literaturerdquo Acta Diabetologica vol 40supplement 1 pp S203ndashS206 2003

[18] K R Foster andH C Lukaski ldquoWhole-body impedancemdashwhatdoes it measurerdquo The American Journal of Clinical Nutritionvol 64 no 3 pp 388Sndash396S 1996

[19] W C Chumlea and S S Sun ldquoBioelectrical impedance anal-ysisrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 79ndash88 HumanKinetics Champaign Ill USA 2nd edition 2005

[20] L Iacopino A Andreoli I Innocente et al ldquoUse of foot-to-foot bioelectrical impedance analysis in childrenrdquo ActaDiabetologica vol 40 supplement 1 pp S210ndashS211 2003

[21] K Lu B Quach T K Tong and PW C Lau ldquoValidation of leg-to-leg bio-impedance analysis for assessing body compositionin obese Chinese childrenrdquo Journal of Exercise Scienceamp Fitnessvol 1 no 2 pp 97ndash103 2003

[22] A Mooney L Kelsey G W Fellingham et al ldquoAssessing bodycomposition of children and adolescents using dual-energyX-ray absorptiometry skinfolds and electrical impedancerdquoMeasurement in Physical Education and Exercise Science vol 15no 1 pp 2ndash17 2011

[23] R Y T Sung P Lau CW Yu P KW Lam and E A S NelsonldquoMeasurement of body fat using leg to leg bioimpedancerdquoArchives ofDisease inChildhood vol 85 no 3 pp 263ndash267 2001

[24] C Azcona N Koek and G Fruhbeck ldquoFat mass by air-displacement plethysmography and impedance in obesenon-obese children and adolescentsrdquo International Journal of Pedi-atric Obesity vol 1 no 3 pp 176ndash182 2006

[25] G S Goldfield P Cloutier R Mallory D Prudrsquohomme TParker and E Doucet ldquoValidity of foot-to-foot bioelectrical

impedance analysis in overweight and obese children andparentsrdquoThe Journal of SportsMedicine and Physical Fitness vol46 no 3 pp 447ndash453 2006

[26] J Hosking B S Metcalf A N Jeffery L D Voss and TJ Wilkin ldquoValidation of foot-to-foot bioelectrical impedanceanalysis with dual-energy X-ray absorptiometry in the assess-ment of body composition in young children the EarlyBirdcohortrdquo British Journal of Nutrition vol 96 no 6 pp 1163ndash11682006

[27] S Lazzer Y Boirie M Meyer and M Vermorel ldquoEvaluationof two foot-to-foot bioelectrical impedance analysers to assessbody composition in overweight and obese adolescentsrdquo BritishJournal of Nutrition vol 90 no 5 pp 987ndash992 2003

[28] L Parker J J Reilly C Slater J C K Wells and Y PitsiladisldquoValidity of six field and laboratory methods for measurementof body composition in boysrdquoObesity Research vol 11 no 7 pp852ndash858 2003

[29] J J Reilly K Gerasimidis N Paparacleous et al ldquoValidationof dual-energy x-ray absorptiometry and foot-foot impedanceagainst deuterium dilution measures of fatness in childrenrdquoInternational Journal of Pediatric Obesity vol 5 no 1 pp 111ndash115 2010

[30] J S Lim J S Hwang J A Lee et al ldquoCross-calibration ofmulti-frequency bioelectrical impedance analysis with eight-point tactile electrodes and dual-energy X-ray absorptiometryfor assessment of body composition in healthy children aged6-18 yearsrdquo Pediatrics International vol 51 no 2 pp 263ndash2682009

[31] O K Yu Y K Rhee T S Park and Y S Cha ldquoComparisons ofobesity assessments in over-weight elementary students usinganthropometry BIA CT and DEXArdquo Nutrition Research andPractice vol 4 no 2 pp 128ndash135 2010

[32] J D Sluyter D Schaaf R K R Scragg and L D PlankldquoPrediction of fatness by standing 8-electrode bioimpedance amultiethnic adolescent populationrdquo Obesity vol 18 no 1 pp183ndash189 2010

[33] K P Navder Q He X Zhang et al ldquoRelationship between bodymass index and adiposity in prepubertal children ethnic andgeographic comparisons between New York City and Jinan City(China)rdquo Journal of Applied Physiology vol 107 no 2 pp 488ndash493 2009

[34] N J Shaw N J Crabtree M S Kibirige and J N FordhamldquoEthnic and gender differences in body fat in British schoolchil-dren as measured by DXArdquo Archives of Disease in Childhoodvol 92 no 10 pp 872ndash875 2007

[35] M A Stone L Williams S Chatterjee M J Davies and KKhunti ldquoEthnic differences in body composition in adoles-centsrdquo Primary Care Diabetes vol 2 no 1 pp 55ndash57 2008

[36] G B McBride ldquoA proposal for strength-of-agreement criteriafor Linrsquos concordance correlation coefficientrdquo NIWA ClientReport HAM2005-062 2005

[37] J M Bland and D G Altman ldquoStatistical methods for assessingagreement between two methods of clinical measurementrdquoTheLancet vol 1 no 8476 pp 307ndash310 1986

[38] M Dehghan and A T Merchant ldquoIs bioelectrical impedanceaccurate for use in large epidemiological studiesrdquo NutritionJournal vol 7 no 1 article 26 2008

[39] B SMcClanahanM B Stockton J Q Lanctot et al ldquoMeasure-ment of body composition in 810-year-old African-Americangirls a comparison of dual-energy X-ray absorptiometry andfoot-to-foot bioimpedance methodsrdquo International Journal ofPediatric Obesity vol 4 no 4 pp 389ndash396 2009

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

BioMed Research International 5

Table 3 Cross-validation of four BIA devices (embedded equations) for the perdition of BFFFM in Chinese children and adolescents

Gender BIAdevices

BF-DXA

BF-BIA

FFM-DXA

FFM-BIA 119901 119905-test Bias (SD) TE 119903

119910minus1199101015840119910

95limits ofagreement

Boys(119899 = 127)

Model A 2077 2134 mdash mdash 0268 058 (586) 586 0513lowastlowast minus1096 to1212

Model B 2077 2192 mdash mdash 0031 115 (595) 604 0528lowastlowast minus1051 to1281

Model C 2077 1904 mdash mdash lt0001 minus167(480) 506 0162 minus1108 to

774

Model D 2077 2022 mdash mdash 0246 minus037(351) 351 0365lowastlowast minus725 to

651

Model A mdash mdash 4302 kg 4380 kg 0019 078(370) kg 377 kg 0261lowastlowast minus647 kg to

803 kg

Model B mdash mdash 4302 kg 4318 kg 0608 016(360) kg 360 kg minus0027 minus690 kg to

722 kg

Model C mdash mdash 4302 kg 4450 kg lt0001 154(335) kg 360 kg 0024 minus503 kg to

811 kg

Model D mdash mdash 4302 kg 4431 kg lt0001 102(217) kg 239 kg 0197lowast minus323 kg to

527 kg

Girls(119899 = 128)

Model A 2785 2737 mdash mdash 0266 minus048(488) 488 0359lowastlowast minus1004 to

908

Model B 2785 2522 mdash mdash lt0001 minus263(323) 415 minus0049 minus896 to

370

Model C 2785 2581 mdash mdash lt0001 minus205(321) 379 0012 minus834 to

424

Model D 2785 2634 mdash mdash lt0001 minus147(299) 331 minus0147 minus733 to

439

Model A mdash mdash 3277 kg 3389 kg lt0001 111 (242) kg 266 kg 0245lowastlowast minus363 kg to585 kg

Model B mdash mdash 3277 kg 3474 kg lt0001 197(177) kg 264 kg minus0255lowastlowast minus150 kg to

544 kg

Model C mdash mdash 3277 kg 3449 kg lt0001 172(177) kg 247 kg minus0219lowastlowast minus175 kg to

519 kg

Model D mdash mdash 3277 kg 3403 kg lt0001 125(122) kg 168 kg minus0022 minus114 kg to

364 kgNote BF body fat percentage FFM fat-free mass DXA dual-energy X-ray absorptiometry BIA bioelectrical impedance analysisBias value obtained from BIA measurement minus DXA measurement SD standard deviation119903119910minus1199101015840119910

Pearsonrsquos correlation coefficient for the relationship between averaged BFFFM [(value of DXA + value of BIA)2] and the biaslowast

119901 lt 005 lowastlowast119901 lt 001Model A Biodynamics-310 Model B Tanita TBF 543 Model C Tanita BC-545 Model D Inbody 520

In the current study although high correlations in bodycomposition measurements between BIA devices and DXAwere found foot-foot (TBF-543) BIA device overestimatedBF in boys and underestimated BF in girls in comparisonwith DXA measurement Most of previous studies foundthat Tanita BIA devices underestimated BF in both genders[24 25 27 37] with which the findings of the current studywas consistent for girls However in current boys TanitaTBF-543 overestimated BF (bias = 115 plusmn 559) withnegative correlation between bias and averaged BF betweenBIA and DXA measurements The discrepancies may beexplained by different testing population ethnicity and use ofTanita BIA devices with different models The current studyuses a heterogeneous and relatively larger sample coveringa wide range of ages and body compositions according toage and gender specific BMI distributions among Chinese

children To date two studies were conducted to investigatethe accuracy of foot-foot BIA consumer devices inmeasuringbody fat percentage in Chinese children and adolescentscompared with DXA measurement [21 23] however thesestudies had several limitations Lu et al measured BF of64 Chinese obese children aged 10ndash17 years using Tanita-401and no significant differences were found in BF betweenfoot-foot BIA and DXA [21] Moreover BIA overestimatedthe BF in seriously obese children Sung et al foundthat BIA (Tanita TBF 401) slightly overestimated BF in7ndash18-year-old Chinese children (119899 = 49) however theBF bias did not reach statistical significance [23] The twostudies used smaller samples with a wider age range toinvestigate the validity of foot-foot measured BF whichmay significantly explain the different findings between thecurrent and previous studies

6 BioMed Research International

In addition despite no mean bias was found for theBF between the hand-foot BIA device (Biodynamics-310)and DXA measurement in both genders the Bland-Altmananalysis showed that the two BIA devices overestimatedmeasurements at a lowBF and underestimated them at highBF Previous studies investigated the validity of hand-footBIA devices in assessing BF in children [22 27 28 41]Several studies had the same results with that of the presentstudy [22 41]

The eight-tactile electrodes with multiple frequenciesBIA device could estimate body composition in adults moreaccurately than a single-frequency BIA device [17 42] How-ever contradictory findings were revealed on the validityof multiple frequencies BIA devices in measuring bodycomposition in children and adolescents Two studies didnot show statistical significance in BF between multiplefrequencies BIA (InBody) and DXA measurements [3031] However Jensky-Squires et al found that InBody 320overestimated BF in female adolescents from underwaterweighing [10] In two other studies Prins et al found thatTanita BC-418 overestimated BF determined by deuteriumdilution in African children [43] Sluyter et al found thatthe same BIA device underestimated BF in a multiethnicgroup of adolescents Moreover the BIA device tended tounderestimate measurements in lower BF and overestimatethem in higher BF individuals [32] In the current studytwomultiple frequencies BIA devices (InBody 520 and TanitaBC-545) were also used to evaluate body composition TanitaBC-545 underestimated BF in both genders and InBody520 underestimated BF in girls In boys no bias was foundin BF between InBody 520 and DXA measurement but asignificant positive correlation was found between the biasand averaged BF between BIA and DXA InBody 520 thusoverestimated BF at high values and overestimated it at lowvalues in boys For Tanita multiple frequencies BIA devicesthe results of the current and previous studies are consistentThedevices underestimatedBF in children and adolescentsFor the InBody multiple BIA devices the varying results mayhave been caused by different population and ethnicity

Moreover note that although themean biasmay be smallof higher relevance are the 95 limits of agreement whichseem to more accurately reflect the validity of measurementin an individual Numerous studies found that the BIA hasquite wide limits of agreement with criterionmeasurement ofBF [24 26 27 41 44] therefore BIA is not interchangeablewith criterion measurement for body composition As suchis true even for Chinese children and adolescents as in thecurrent study the 95 limits of agreement of all BIA devicesshowed wide ranges in agreement with previous studiesSung et al found that the 95 limits of agreement in BFwere narrow (from minus392 to 061) between foot-foot BIAand DXA in 17 obese and 32 nonobese Chinese childrenaged 7ndash18 years [23] The current study had larger 95 limitsof agreement in the same ethnicity and similar age rangeof population The discrepancy may be partly explained bydifferent samples size sampling method and DXA instru-ments used (Hologic QDR-4500 and GE Lunar Prodigy)Previous studies suggested that if researchers intend tomeasure body composition in large-scale population surveys

then BIAmeasurementmay provide an efficient alternative tolaboratory measurement of body composition when accurateinstruments are unavailable [45] BIA measurement shouldbe used with caution in evaluating or monitoring changes inindividual body composition in healthy and clinical setting[38]

5 Conclusion

Using embedded equations in BIA devices should be vali-dated in assessing the body composition of Chinese childrenand adolescents aged 9ndash19 years old All BIA devices are notdirectly interchanged with DXA measurement in assessingindividual BF and assessing changes of body compositionowing to the wide limits of agreement

Disclaimer

The authors alone are responsible for the content and writingof the paper

Conflict of Interests

The authors report no conflict of interests

Acknowledgments

This study was supported by the Hong Kong Association forthe Study of Obesity (2011) and Natural Science Foundationof Shanghai (15ZR1439300) The funders had no role in studydesign collection analysis and interpretation of data writingof paper or decision to submit for paper

References

[1] World Health Organization ldquoObesity and overweightrdquo FactSheet No 311 2011 httpwwwwhointmediacentrefact-sheetsfs311enindexhtml

[2] T Lobstein L Baur and R Uauy ldquoObesity in children andyoung people a crisis in public healthrdquo Obesity ReviewsSupplement vol 5 no 1 pp 4ndash104 2004

[3] Y Wu ldquoOverweight and obesity in Chinardquo British MedicalJournal vol 333 no 7564 pp 362ndash363 2006

[4] Y P Li E G Schouten X Q Hu Z H Cui D C Luan and GS Ma ldquoObesity prevalence and time trend among youngstersin China 1982ndash2002rdquo Asia Pacific Journal of Clinical Nutritionvol 17 no 1 pp 131ndash137 2008

[5] J J Reilly E Methven Z C McDowell et al ldquoHealth conse-quences of obesityrdquo Archives of Disease in Childhood vol 88no 9 pp 748ndash752 2003

[6] W H Dietz ldquoHealth consequences of obesity in youth child-hood predictors of adult diseaserdquo Pediatrics vol 101 no 3 pp518ndash525 1998

[7] J C KWells andM S Fewtrell ldquoMeasuring body compositionrdquoArchives of Disease in Childhood vol 91 no 7 pp 612ndash617 2006

[8] J C K Wells and M S Fewtrell ldquoIs body compositionimportant for paediatriciansrdquoArchives of Disease in Childhoodvol 93 no 2 pp 168ndash172 2008

BioMed Research International 7

[9] V H Heyward and D R Wagner Applied Body CompositionAssessment Human Kinetics Champaign Ill USA 2004

[10] N E Jensky-Squires CMDieli-Conwright A Rossuello DNErceg SMcCauley and T E Schroeder ldquoValidity and reliabilityof body composition analysers in children and adultsrdquo BritishJournal of Nutrition vol 100 no 4 pp 859ndash865 2008

[11] M Helba and L A Binkovitz ldquoPediatric body compositionanalysis with dual-energy X-ray absorptiometryrdquo PediatricRadiology vol 39 no 7 pp 647ndash656 2009

[12] A B Sopher J C Thornton J Wang R N Pierson Jr S BHeymsfield and M Horlick ldquoMeasurement of percentage ofbody fat in 411 children and adolescents a comparison of dual-energy X-ray absorptiometry with a four-compartment modelrdquoPediatrics vol 113 no 5 pp 1285ndash1290 2004

[13] T G Lohman and C Zhao ldquoDual-energy X-ray adsorptiom-etryrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 63ndash77 HumanKinetics Champaign Ill USA 2nd edition 2005

[14] E O Diaz J Villar M Immink and T Gonzales ldquoBioimpe-dance or anthropometryrdquo European Journal of Clinical Nutri-tion vol 43 no 2 pp 129ndash137 1989

[15] S Demura S Sato and T Kitabayashi ldquoPercentage of total bodyfat as estimated by three automatic bioelectrical impedanceanalyzersrdquo Journal of Physiological Anthropology and AppliedHuman Science vol 23 no 3 pp 93ndash99 2004

[16] A Pietrobelli F Rubiano M-P St-Onge and S B HeymsfieldldquoNew bioimpedance analysis system improved phenotypingwith whole-body analysisrdquo European Journal of Clinical Nutri-tion vol 58 no 11 pp 1479ndash1484 2004

[17] R Martinoli E I Mohamed C Maiolo et al ldquoTotal body waterestimation using bioelectrical impedance a meta-analysis ofthe data available in the literaturerdquo Acta Diabetologica vol 40supplement 1 pp S203ndashS206 2003

[18] K R Foster andH C Lukaski ldquoWhole-body impedancemdashwhatdoes it measurerdquo The American Journal of Clinical Nutritionvol 64 no 3 pp 388Sndash396S 1996

[19] W C Chumlea and S S Sun ldquoBioelectrical impedance anal-ysisrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 79ndash88 HumanKinetics Champaign Ill USA 2nd edition 2005

[20] L Iacopino A Andreoli I Innocente et al ldquoUse of foot-to-foot bioelectrical impedance analysis in childrenrdquo ActaDiabetologica vol 40 supplement 1 pp S210ndashS211 2003

[21] K Lu B Quach T K Tong and PW C Lau ldquoValidation of leg-to-leg bio-impedance analysis for assessing body compositionin obese Chinese childrenrdquo Journal of Exercise Scienceamp Fitnessvol 1 no 2 pp 97ndash103 2003

[22] A Mooney L Kelsey G W Fellingham et al ldquoAssessing bodycomposition of children and adolescents using dual-energyX-ray absorptiometry skinfolds and electrical impedancerdquoMeasurement in Physical Education and Exercise Science vol 15no 1 pp 2ndash17 2011

[23] R Y T Sung P Lau CW Yu P KW Lam and E A S NelsonldquoMeasurement of body fat using leg to leg bioimpedancerdquoArchives ofDisease inChildhood vol 85 no 3 pp 263ndash267 2001

[24] C Azcona N Koek and G Fruhbeck ldquoFat mass by air-displacement plethysmography and impedance in obesenon-obese children and adolescentsrdquo International Journal of Pedi-atric Obesity vol 1 no 3 pp 176ndash182 2006

[25] G S Goldfield P Cloutier R Mallory D Prudrsquohomme TParker and E Doucet ldquoValidity of foot-to-foot bioelectrical

impedance analysis in overweight and obese children andparentsrdquoThe Journal of SportsMedicine and Physical Fitness vol46 no 3 pp 447ndash453 2006

[26] J Hosking B S Metcalf A N Jeffery L D Voss and TJ Wilkin ldquoValidation of foot-to-foot bioelectrical impedanceanalysis with dual-energy X-ray absorptiometry in the assess-ment of body composition in young children the EarlyBirdcohortrdquo British Journal of Nutrition vol 96 no 6 pp 1163ndash11682006

[27] S Lazzer Y Boirie M Meyer and M Vermorel ldquoEvaluationof two foot-to-foot bioelectrical impedance analysers to assessbody composition in overweight and obese adolescentsrdquo BritishJournal of Nutrition vol 90 no 5 pp 987ndash992 2003

[28] L Parker J J Reilly C Slater J C K Wells and Y PitsiladisldquoValidity of six field and laboratory methods for measurementof body composition in boysrdquoObesity Research vol 11 no 7 pp852ndash858 2003

[29] J J Reilly K Gerasimidis N Paparacleous et al ldquoValidationof dual-energy x-ray absorptiometry and foot-foot impedanceagainst deuterium dilution measures of fatness in childrenrdquoInternational Journal of Pediatric Obesity vol 5 no 1 pp 111ndash115 2010

[30] J S Lim J S Hwang J A Lee et al ldquoCross-calibration ofmulti-frequency bioelectrical impedance analysis with eight-point tactile electrodes and dual-energy X-ray absorptiometryfor assessment of body composition in healthy children aged6-18 yearsrdquo Pediatrics International vol 51 no 2 pp 263ndash2682009

[31] O K Yu Y K Rhee T S Park and Y S Cha ldquoComparisons ofobesity assessments in over-weight elementary students usinganthropometry BIA CT and DEXArdquo Nutrition Research andPractice vol 4 no 2 pp 128ndash135 2010

[32] J D Sluyter D Schaaf R K R Scragg and L D PlankldquoPrediction of fatness by standing 8-electrode bioimpedance amultiethnic adolescent populationrdquo Obesity vol 18 no 1 pp183ndash189 2010

[33] K P Navder Q He X Zhang et al ldquoRelationship between bodymass index and adiposity in prepubertal children ethnic andgeographic comparisons between New York City and Jinan City(China)rdquo Journal of Applied Physiology vol 107 no 2 pp 488ndash493 2009

[34] N J Shaw N J Crabtree M S Kibirige and J N FordhamldquoEthnic and gender differences in body fat in British schoolchil-dren as measured by DXArdquo Archives of Disease in Childhoodvol 92 no 10 pp 872ndash875 2007

[35] M A Stone L Williams S Chatterjee M J Davies and KKhunti ldquoEthnic differences in body composition in adoles-centsrdquo Primary Care Diabetes vol 2 no 1 pp 55ndash57 2008

[36] G B McBride ldquoA proposal for strength-of-agreement criteriafor Linrsquos concordance correlation coefficientrdquo NIWA ClientReport HAM2005-062 2005

[37] J M Bland and D G Altman ldquoStatistical methods for assessingagreement between two methods of clinical measurementrdquoTheLancet vol 1 no 8476 pp 307ndash310 1986

[38] M Dehghan and A T Merchant ldquoIs bioelectrical impedanceaccurate for use in large epidemiological studiesrdquo NutritionJournal vol 7 no 1 article 26 2008

[39] B SMcClanahanM B Stockton J Q Lanctot et al ldquoMeasure-ment of body composition in 810-year-old African-Americangirls a comparison of dual-energy X-ray absorptiometry andfoot-to-foot bioimpedance methodsrdquo International Journal ofPediatric Obesity vol 4 no 4 pp 389ndash396 2009

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

6 BioMed Research International

In addition despite no mean bias was found for theBF between the hand-foot BIA device (Biodynamics-310)and DXA measurement in both genders the Bland-Altmananalysis showed that the two BIA devices overestimatedmeasurements at a lowBF and underestimated them at highBF Previous studies investigated the validity of hand-footBIA devices in assessing BF in children [22 27 28 41]Several studies had the same results with that of the presentstudy [22 41]

The eight-tactile electrodes with multiple frequenciesBIA device could estimate body composition in adults moreaccurately than a single-frequency BIA device [17 42] How-ever contradictory findings were revealed on the validityof multiple frequencies BIA devices in measuring bodycomposition in children and adolescents Two studies didnot show statistical significance in BF between multiplefrequencies BIA (InBody) and DXA measurements [3031] However Jensky-Squires et al found that InBody 320overestimated BF in female adolescents from underwaterweighing [10] In two other studies Prins et al found thatTanita BC-418 overestimated BF determined by deuteriumdilution in African children [43] Sluyter et al found thatthe same BIA device underestimated BF in a multiethnicgroup of adolescents Moreover the BIA device tended tounderestimate measurements in lower BF and overestimatethem in higher BF individuals [32] In the current studytwomultiple frequencies BIA devices (InBody 520 and TanitaBC-545) were also used to evaluate body composition TanitaBC-545 underestimated BF in both genders and InBody520 underestimated BF in girls In boys no bias was foundin BF between InBody 520 and DXA measurement but asignificant positive correlation was found between the biasand averaged BF between BIA and DXA InBody 520 thusoverestimated BF at high values and overestimated it at lowvalues in boys For Tanita multiple frequencies BIA devicesthe results of the current and previous studies are consistentThedevices underestimatedBF in children and adolescentsFor the InBody multiple BIA devices the varying results mayhave been caused by different population and ethnicity

Moreover note that although themean biasmay be smallof higher relevance are the 95 limits of agreement whichseem to more accurately reflect the validity of measurementin an individual Numerous studies found that the BIA hasquite wide limits of agreement with criterionmeasurement ofBF [24 26 27 41 44] therefore BIA is not interchangeablewith criterion measurement for body composition As suchis true even for Chinese children and adolescents as in thecurrent study the 95 limits of agreement of all BIA devicesshowed wide ranges in agreement with previous studiesSung et al found that the 95 limits of agreement in BFwere narrow (from minus392 to 061) between foot-foot BIAand DXA in 17 obese and 32 nonobese Chinese childrenaged 7ndash18 years [23] The current study had larger 95 limitsof agreement in the same ethnicity and similar age rangeof population The discrepancy may be partly explained bydifferent samples size sampling method and DXA instru-ments used (Hologic QDR-4500 and GE Lunar Prodigy)Previous studies suggested that if researchers intend tomeasure body composition in large-scale population surveys

then BIAmeasurementmay provide an efficient alternative tolaboratory measurement of body composition when accurateinstruments are unavailable [45] BIA measurement shouldbe used with caution in evaluating or monitoring changes inindividual body composition in healthy and clinical setting[38]

5 Conclusion

Using embedded equations in BIA devices should be vali-dated in assessing the body composition of Chinese childrenand adolescents aged 9ndash19 years old All BIA devices are notdirectly interchanged with DXA measurement in assessingindividual BF and assessing changes of body compositionowing to the wide limits of agreement

Disclaimer

The authors alone are responsible for the content and writingof the paper

Conflict of Interests

The authors report no conflict of interests

Acknowledgments

This study was supported by the Hong Kong Association forthe Study of Obesity (2011) and Natural Science Foundationof Shanghai (15ZR1439300) The funders had no role in studydesign collection analysis and interpretation of data writingof paper or decision to submit for paper

References

[1] World Health Organization ldquoObesity and overweightrdquo FactSheet No 311 2011 httpwwwwhointmediacentrefact-sheetsfs311enindexhtml

[2] T Lobstein L Baur and R Uauy ldquoObesity in children andyoung people a crisis in public healthrdquo Obesity ReviewsSupplement vol 5 no 1 pp 4ndash104 2004

[3] Y Wu ldquoOverweight and obesity in Chinardquo British MedicalJournal vol 333 no 7564 pp 362ndash363 2006

[4] Y P Li E G Schouten X Q Hu Z H Cui D C Luan and GS Ma ldquoObesity prevalence and time trend among youngstersin China 1982ndash2002rdquo Asia Pacific Journal of Clinical Nutritionvol 17 no 1 pp 131ndash137 2008

[5] J J Reilly E Methven Z C McDowell et al ldquoHealth conse-quences of obesityrdquo Archives of Disease in Childhood vol 88no 9 pp 748ndash752 2003

[6] W H Dietz ldquoHealth consequences of obesity in youth child-hood predictors of adult diseaserdquo Pediatrics vol 101 no 3 pp518ndash525 1998

[7] J C KWells andM S Fewtrell ldquoMeasuring body compositionrdquoArchives of Disease in Childhood vol 91 no 7 pp 612ndash617 2006

[8] J C K Wells and M S Fewtrell ldquoIs body compositionimportant for paediatriciansrdquoArchives of Disease in Childhoodvol 93 no 2 pp 168ndash172 2008

BioMed Research International 7

[9] V H Heyward and D R Wagner Applied Body CompositionAssessment Human Kinetics Champaign Ill USA 2004

[10] N E Jensky-Squires CMDieli-Conwright A Rossuello DNErceg SMcCauley and T E Schroeder ldquoValidity and reliabilityof body composition analysers in children and adultsrdquo BritishJournal of Nutrition vol 100 no 4 pp 859ndash865 2008

[11] M Helba and L A Binkovitz ldquoPediatric body compositionanalysis with dual-energy X-ray absorptiometryrdquo PediatricRadiology vol 39 no 7 pp 647ndash656 2009

[12] A B Sopher J C Thornton J Wang R N Pierson Jr S BHeymsfield and M Horlick ldquoMeasurement of percentage ofbody fat in 411 children and adolescents a comparison of dual-energy X-ray absorptiometry with a four-compartment modelrdquoPediatrics vol 113 no 5 pp 1285ndash1290 2004

[13] T G Lohman and C Zhao ldquoDual-energy X-ray adsorptiom-etryrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 63ndash77 HumanKinetics Champaign Ill USA 2nd edition 2005

[14] E O Diaz J Villar M Immink and T Gonzales ldquoBioimpe-dance or anthropometryrdquo European Journal of Clinical Nutri-tion vol 43 no 2 pp 129ndash137 1989

[15] S Demura S Sato and T Kitabayashi ldquoPercentage of total bodyfat as estimated by three automatic bioelectrical impedanceanalyzersrdquo Journal of Physiological Anthropology and AppliedHuman Science vol 23 no 3 pp 93ndash99 2004

[16] A Pietrobelli F Rubiano M-P St-Onge and S B HeymsfieldldquoNew bioimpedance analysis system improved phenotypingwith whole-body analysisrdquo European Journal of Clinical Nutri-tion vol 58 no 11 pp 1479ndash1484 2004

[17] R Martinoli E I Mohamed C Maiolo et al ldquoTotal body waterestimation using bioelectrical impedance a meta-analysis ofthe data available in the literaturerdquo Acta Diabetologica vol 40supplement 1 pp S203ndashS206 2003

[18] K R Foster andH C Lukaski ldquoWhole-body impedancemdashwhatdoes it measurerdquo The American Journal of Clinical Nutritionvol 64 no 3 pp 388Sndash396S 1996

[19] W C Chumlea and S S Sun ldquoBioelectrical impedance anal-ysisrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 79ndash88 HumanKinetics Champaign Ill USA 2nd edition 2005

[20] L Iacopino A Andreoli I Innocente et al ldquoUse of foot-to-foot bioelectrical impedance analysis in childrenrdquo ActaDiabetologica vol 40 supplement 1 pp S210ndashS211 2003

[21] K Lu B Quach T K Tong and PW C Lau ldquoValidation of leg-to-leg bio-impedance analysis for assessing body compositionin obese Chinese childrenrdquo Journal of Exercise Scienceamp Fitnessvol 1 no 2 pp 97ndash103 2003

[22] A Mooney L Kelsey G W Fellingham et al ldquoAssessing bodycomposition of children and adolescents using dual-energyX-ray absorptiometry skinfolds and electrical impedancerdquoMeasurement in Physical Education and Exercise Science vol 15no 1 pp 2ndash17 2011

[23] R Y T Sung P Lau CW Yu P KW Lam and E A S NelsonldquoMeasurement of body fat using leg to leg bioimpedancerdquoArchives ofDisease inChildhood vol 85 no 3 pp 263ndash267 2001

[24] C Azcona N Koek and G Fruhbeck ldquoFat mass by air-displacement plethysmography and impedance in obesenon-obese children and adolescentsrdquo International Journal of Pedi-atric Obesity vol 1 no 3 pp 176ndash182 2006

[25] G S Goldfield P Cloutier R Mallory D Prudrsquohomme TParker and E Doucet ldquoValidity of foot-to-foot bioelectrical

impedance analysis in overweight and obese children andparentsrdquoThe Journal of SportsMedicine and Physical Fitness vol46 no 3 pp 447ndash453 2006

[26] J Hosking B S Metcalf A N Jeffery L D Voss and TJ Wilkin ldquoValidation of foot-to-foot bioelectrical impedanceanalysis with dual-energy X-ray absorptiometry in the assess-ment of body composition in young children the EarlyBirdcohortrdquo British Journal of Nutrition vol 96 no 6 pp 1163ndash11682006

[27] S Lazzer Y Boirie M Meyer and M Vermorel ldquoEvaluationof two foot-to-foot bioelectrical impedance analysers to assessbody composition in overweight and obese adolescentsrdquo BritishJournal of Nutrition vol 90 no 5 pp 987ndash992 2003

[28] L Parker J J Reilly C Slater J C K Wells and Y PitsiladisldquoValidity of six field and laboratory methods for measurementof body composition in boysrdquoObesity Research vol 11 no 7 pp852ndash858 2003

[29] J J Reilly K Gerasimidis N Paparacleous et al ldquoValidationof dual-energy x-ray absorptiometry and foot-foot impedanceagainst deuterium dilution measures of fatness in childrenrdquoInternational Journal of Pediatric Obesity vol 5 no 1 pp 111ndash115 2010

[30] J S Lim J S Hwang J A Lee et al ldquoCross-calibration ofmulti-frequency bioelectrical impedance analysis with eight-point tactile electrodes and dual-energy X-ray absorptiometryfor assessment of body composition in healthy children aged6-18 yearsrdquo Pediatrics International vol 51 no 2 pp 263ndash2682009

[31] O K Yu Y K Rhee T S Park and Y S Cha ldquoComparisons ofobesity assessments in over-weight elementary students usinganthropometry BIA CT and DEXArdquo Nutrition Research andPractice vol 4 no 2 pp 128ndash135 2010

[32] J D Sluyter D Schaaf R K R Scragg and L D PlankldquoPrediction of fatness by standing 8-electrode bioimpedance amultiethnic adolescent populationrdquo Obesity vol 18 no 1 pp183ndash189 2010

[33] K P Navder Q He X Zhang et al ldquoRelationship between bodymass index and adiposity in prepubertal children ethnic andgeographic comparisons between New York City and Jinan City(China)rdquo Journal of Applied Physiology vol 107 no 2 pp 488ndash493 2009

[34] N J Shaw N J Crabtree M S Kibirige and J N FordhamldquoEthnic and gender differences in body fat in British schoolchil-dren as measured by DXArdquo Archives of Disease in Childhoodvol 92 no 10 pp 872ndash875 2007

[35] M A Stone L Williams S Chatterjee M J Davies and KKhunti ldquoEthnic differences in body composition in adoles-centsrdquo Primary Care Diabetes vol 2 no 1 pp 55ndash57 2008

[36] G B McBride ldquoA proposal for strength-of-agreement criteriafor Linrsquos concordance correlation coefficientrdquo NIWA ClientReport HAM2005-062 2005

[37] J M Bland and D G Altman ldquoStatistical methods for assessingagreement between two methods of clinical measurementrdquoTheLancet vol 1 no 8476 pp 307ndash310 1986

[38] M Dehghan and A T Merchant ldquoIs bioelectrical impedanceaccurate for use in large epidemiological studiesrdquo NutritionJournal vol 7 no 1 article 26 2008

[39] B SMcClanahanM B Stockton J Q Lanctot et al ldquoMeasure-ment of body composition in 810-year-old African-Americangirls a comparison of dual-energy X-ray absorptiometry andfoot-to-foot bioimpedance methodsrdquo International Journal ofPediatric Obesity vol 4 no 4 pp 389ndash396 2009

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

BioMed Research International 7

[9] V H Heyward and D R Wagner Applied Body CompositionAssessment Human Kinetics Champaign Ill USA 2004

[10] N E Jensky-Squires CMDieli-Conwright A Rossuello DNErceg SMcCauley and T E Schroeder ldquoValidity and reliabilityof body composition analysers in children and adultsrdquo BritishJournal of Nutrition vol 100 no 4 pp 859ndash865 2008

[11] M Helba and L A Binkovitz ldquoPediatric body compositionanalysis with dual-energy X-ray absorptiometryrdquo PediatricRadiology vol 39 no 7 pp 647ndash656 2009

[12] A B Sopher J C Thornton J Wang R N Pierson Jr S BHeymsfield and M Horlick ldquoMeasurement of percentage ofbody fat in 411 children and adolescents a comparison of dual-energy X-ray absorptiometry with a four-compartment modelrdquoPediatrics vol 113 no 5 pp 1285ndash1290 2004

[13] T G Lohman and C Zhao ldquoDual-energy X-ray adsorptiom-etryrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 63ndash77 HumanKinetics Champaign Ill USA 2nd edition 2005

[14] E O Diaz J Villar M Immink and T Gonzales ldquoBioimpe-dance or anthropometryrdquo European Journal of Clinical Nutri-tion vol 43 no 2 pp 129ndash137 1989

[15] S Demura S Sato and T Kitabayashi ldquoPercentage of total bodyfat as estimated by three automatic bioelectrical impedanceanalyzersrdquo Journal of Physiological Anthropology and AppliedHuman Science vol 23 no 3 pp 93ndash99 2004

[16] A Pietrobelli F Rubiano M-P St-Onge and S B HeymsfieldldquoNew bioimpedance analysis system improved phenotypingwith whole-body analysisrdquo European Journal of Clinical Nutri-tion vol 58 no 11 pp 1479ndash1484 2004

[17] R Martinoli E I Mohamed C Maiolo et al ldquoTotal body waterestimation using bioelectrical impedance a meta-analysis ofthe data available in the literaturerdquo Acta Diabetologica vol 40supplement 1 pp S203ndashS206 2003

[18] K R Foster andH C Lukaski ldquoWhole-body impedancemdashwhatdoes it measurerdquo The American Journal of Clinical Nutritionvol 64 no 3 pp 388Sndash396S 1996

[19] W C Chumlea and S S Sun ldquoBioelectrical impedance anal-ysisrdquo in Human Body Composition S B Heymsfield T GLohman Z Wang and S B Going Eds pp 79ndash88 HumanKinetics Champaign Ill USA 2nd edition 2005

[20] L Iacopino A Andreoli I Innocente et al ldquoUse of foot-to-foot bioelectrical impedance analysis in childrenrdquo ActaDiabetologica vol 40 supplement 1 pp S210ndashS211 2003

[21] K Lu B Quach T K Tong and PW C Lau ldquoValidation of leg-to-leg bio-impedance analysis for assessing body compositionin obese Chinese childrenrdquo Journal of Exercise Scienceamp Fitnessvol 1 no 2 pp 97ndash103 2003

[22] A Mooney L Kelsey G W Fellingham et al ldquoAssessing bodycomposition of children and adolescents using dual-energyX-ray absorptiometry skinfolds and electrical impedancerdquoMeasurement in Physical Education and Exercise Science vol 15no 1 pp 2ndash17 2011

[23] R Y T Sung P Lau CW Yu P KW Lam and E A S NelsonldquoMeasurement of body fat using leg to leg bioimpedancerdquoArchives ofDisease inChildhood vol 85 no 3 pp 263ndash267 2001

[24] C Azcona N Koek and G Fruhbeck ldquoFat mass by air-displacement plethysmography and impedance in obesenon-obese children and adolescentsrdquo International Journal of Pedi-atric Obesity vol 1 no 3 pp 176ndash182 2006

[25] G S Goldfield P Cloutier R Mallory D Prudrsquohomme TParker and E Doucet ldquoValidity of foot-to-foot bioelectrical

impedance analysis in overweight and obese children andparentsrdquoThe Journal of SportsMedicine and Physical Fitness vol46 no 3 pp 447ndash453 2006

[26] J Hosking B S Metcalf A N Jeffery L D Voss and TJ Wilkin ldquoValidation of foot-to-foot bioelectrical impedanceanalysis with dual-energy X-ray absorptiometry in the assess-ment of body composition in young children the EarlyBirdcohortrdquo British Journal of Nutrition vol 96 no 6 pp 1163ndash11682006

[27] S Lazzer Y Boirie M Meyer and M Vermorel ldquoEvaluationof two foot-to-foot bioelectrical impedance analysers to assessbody composition in overweight and obese adolescentsrdquo BritishJournal of Nutrition vol 90 no 5 pp 987ndash992 2003

[28] L Parker J J Reilly C Slater J C K Wells and Y PitsiladisldquoValidity of six field and laboratory methods for measurementof body composition in boysrdquoObesity Research vol 11 no 7 pp852ndash858 2003

[29] J J Reilly K Gerasimidis N Paparacleous et al ldquoValidationof dual-energy x-ray absorptiometry and foot-foot impedanceagainst deuterium dilution measures of fatness in childrenrdquoInternational Journal of Pediatric Obesity vol 5 no 1 pp 111ndash115 2010

[30] J S Lim J S Hwang J A Lee et al ldquoCross-calibration ofmulti-frequency bioelectrical impedance analysis with eight-point tactile electrodes and dual-energy X-ray absorptiometryfor assessment of body composition in healthy children aged6-18 yearsrdquo Pediatrics International vol 51 no 2 pp 263ndash2682009

[31] O K Yu Y K Rhee T S Park and Y S Cha ldquoComparisons ofobesity assessments in over-weight elementary students usinganthropometry BIA CT and DEXArdquo Nutrition Research andPractice vol 4 no 2 pp 128ndash135 2010

[32] J D Sluyter D Schaaf R K R Scragg and L D PlankldquoPrediction of fatness by standing 8-electrode bioimpedance amultiethnic adolescent populationrdquo Obesity vol 18 no 1 pp183ndash189 2010

[33] K P Navder Q He X Zhang et al ldquoRelationship between bodymass index and adiposity in prepubertal children ethnic andgeographic comparisons between New York City and Jinan City(China)rdquo Journal of Applied Physiology vol 107 no 2 pp 488ndash493 2009

[34] N J Shaw N J Crabtree M S Kibirige and J N FordhamldquoEthnic and gender differences in body fat in British schoolchil-dren as measured by DXArdquo Archives of Disease in Childhoodvol 92 no 10 pp 872ndash875 2007

[35] M A Stone L Williams S Chatterjee M J Davies and KKhunti ldquoEthnic differences in body composition in adoles-centsrdquo Primary Care Diabetes vol 2 no 1 pp 55ndash57 2008

[36] G B McBride ldquoA proposal for strength-of-agreement criteriafor Linrsquos concordance correlation coefficientrdquo NIWA ClientReport HAM2005-062 2005

[37] J M Bland and D G Altman ldquoStatistical methods for assessingagreement between two methods of clinical measurementrdquoTheLancet vol 1 no 8476 pp 307ndash310 1986

[38] M Dehghan and A T Merchant ldquoIs bioelectrical impedanceaccurate for use in large epidemiological studiesrdquo NutritionJournal vol 7 no 1 article 26 2008

[39] B SMcClanahanM B Stockton J Q Lanctot et al ldquoMeasure-ment of body composition in 810-year-old African-Americangirls a comparison of dual-energy X-ray absorptiometry andfoot-to-foot bioimpedance methodsrdquo International Journal ofPediatric Obesity vol 4 no 4 pp 389ndash396 2009

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

8 BioMed Research International

[40] V J Tyrrell G Richards P Hofman G F Gillies E Robinsonand W S Cutfield ldquoFoot-to-foot bioelectrical impedance anal-ysis a valuable tool for the measurement of body compositionin childrenrdquo International Journal of Obesity vol 25 no 2 pp273ndash278 2001

[41] H Fors L Gelander R Bjarnason K Albertsson-Wiklandand I Bosaeus ldquoBody composition as assessed by bioelectricalimpedance spectroscopy and dual-energy X-ray absorptiom-etry in a healthy paediatric populationrdquo Acta PaediatricaInternational Journal of Paediatrics vol 91 no 7 pp 755ndash7602002

[42] M Dittmar ldquoComparison of bipolar and tetrapolar impedancetechniques for assessing fat massrdquo American Journal of HumanBiology vol 16 no 5 pp 593ndash597 2004

[43] M Prins S Hawkesworth AWright et al ldquoUse of bioelectricalimpedance analysis to assess body composition in rural Gam-bian childrenrdquo European Journal of Clinical Nutrition vol 62no 9 pp 1065ndash1074 2008

[44] D Radley C B CookeN J Fuller et al ldquoValidity of foot-to-footbio-electrical impedance analysis body composition estimatesin overweight and obese childrenrdquo International Journal of BodyComposition Research vol 7 no 1 pp 15ndash20 2009

[45] S H Wong S S Hui and S H Wong ldquoValidity of bioelectricalimpedance measurement in predicting fat-free mass of Chinesechildren and adolescentsrdquoMedical Science Monitor vol 20 pp2298ndash2310 2014

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Molecular Biology International

GenomicsInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioinformaticsAdvances in

Marine BiologyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Signal TransductionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

BioMed Research International

Evolutionary BiologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Genetics Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Enzyme Research

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

Microbiology