Skinfold reference curves and their use in predicting metabolic syndrome risk in children

www.jped.com.br

ORIGINAL

ARTICLE

Skinfold

reference

curves

and

their

use

in

predicting

metabolic

syndrome

risk

in

children

夽,夽夽

Alynne

C.R.

Andaki

_,

_Teresa

_M.B.

_de

_Quadros

_,

_Alex

_P.

_Gordia

_,

_Jorge

_Mota

_,

Q1

Adelson

L.A.

Tinôco

_,

_Edmar

_L.

_Mendes

a_{UniversidadeFederaldoTriânguloMineiro(UFTM),DepartamentodeCiênciasdoEsporte,Uberaba,MG,Brazil} b_{UniversidadeFederaldoRecôncavodaBahia(UFRB),CentrodeFormac¸ãodeProfessores,CruzdasAlmas,BA,Brazil} c_{UniversidadedoPorto,FaculdadedoDesporto,Porto,Portugal}

d_{UniversidadeFederaldeVic¸osa(UFV),DepartamentodeNutric¸ãoeSaúde,Vic¸osa,MG,Brazil}

Received28July2016;accepted16November2016

KEYWORDS Anthropometry; Metabolicsyndrome X; Cross-sectional studies; Child Abstract

Objectives: Todrawskinfold(SF)referencecurves(subscapular,suprailiac,biceps,triceps)and

Q2

todetermineSFcutoffpointsforpredictingtheriskofmetabolicsyndrome(MetS)inchildren aged6---10yearsold.

Methods: Thiswasacross-sectionalstudywitharandomsampleof1480childrenaged6---10 yearsold,52.2%females,frompublicandprivateschoolslocatedintheurbanandruralareas ofthemunicipalityofUberaba(MG).Anthropometry,bloodpressure,andfastingbloodsamples were taken at school, following specificprotocols. The LMSmethod was used to drawthe referencecurvesandROCcurveanalysistodeterminetheaccuracyandcutoffpointsforthe evaluatedskinfolds.

Results: The four SF evaluated (subscapular, suprailiac, biceps,and triceps)and their sum (4SF)wereaccurateinpredictingMetSforbothgirlsandboys.Additionally,cutoffshave beenproposedandpercentilecurves(p5,p10,p25,p50,p75,p90,andp95)wereoutlinedfor thefourSFand4SF,forbothgenders.

Conclusion: SFmeasurementswereaccurateinpredictingmetabolicsyndromeinchildrenaged 6---10yearsold.Age-andgender-specificsmoothedpercentilescurvesofSFprovideareference forthedetectionofriskforMetSinchildren.

©2017PublishedbyElsevierEditoraLtda.onbehalfofSociedadeBrasileiradePediatria.Thisis anopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/

by-nc-nd/4.0/).

夽 _{Pleasecitethisarticleas:AndakiAC,QuadrosTM,GordiaAP,MotaJ,TinôcoAL,MendesEL.Skinfoldreferencecurvesandtheirusein} predictingmetabolicsyndromeriskinchildren.JPediatr(RioJ).2017.http://dx.doi.org/10.1016/j.jped.2016.11.013

夽夽_{StudycarriedoutatUniversidadeFederaldoTriânguloMineiro,Uberaba,MG,Brazil.} ∗_{Correspondingauthor.}

E-mail:alynneandaki@yahoo.com.br(A.C.Andaki).

http://dx.doi.org/10.1016/j.jped.2016.11.013

0021-7557/©2017PublishedbyElsevierEditoraLtda.onbehalfofSociedadeBrasileiradePediatria.Thisisanopenaccessarticleunder theCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

PALAVRAS-CHAVE Antropometria; SíndromeX; EstudosdeCorte Transversal; Crianc¸as

Curvasdereferênciadedobrascutâneasesuautilizac¸ãonapredic¸ãodoriscode síndromemetabólicaemcrianc¸as

Resumo

Objetivos: Desenharcurvasdereferênciadequatrodobrascutâneas(subescapular,suprailíaca, bíceps,tríceps)edeterminarpontosdecorteparapredizeroriscodeSMemcrianc¸asdeseis a10anosdeidade.

Métodos: Estudoepidemiológicodebasepopulacional,cortetransversal,comamostra prob-abilística,estratificadaporsegmentodeensino,com1.480crianc¸asde6a10anosdeidade, 52,2%dosexofeminino,oriundasdeescolaspúblicaseprivadas,localizadasnazonaurbana eruraldomunicípiodeUberaba(MG).Antropometria(dobrascutâneas),pressãoarterialeas coletasdesangueemjejumforamrealizadasemespac¸oreservadonaescola,seguindo proto-colosespecíficos.OmétodoloLMSfoiutilizadoparadesenharascurvasdereferênciaeanálise decurvaROCparadeterminaraacuráciaepontosdecorteparaasdobrascutâneasavaliadas. Resultados: As quatro DC avaliadas (subescapular, suprailíaca, bíceps e tríceps) e o seu somatório(4DC)foramacuradosnapredic¸ãodaSMparameninasemeninos.Adicionalmente, pontosdecorteforampropostosecurvaspercentílicas(p5,p10,p25,p50,p75,p90ep95)foram delineadasparaasquatroDCeo4DC,paraambosossexos.

Conclusão: Medidasde DC foramacuradas em predizerSM em escolares deseis a10 anos

deidade.AscurvaspercentílicasdeDCdesenhadasporidadeesexofornecemreferênciana detecc¸ãodoriscodeSMemcrianc¸as.

©2017PublicadoporElsevierEditoraLtda.emnomedeSociedadeBrasileiradePediatria.Este ´

eumartigoOpenAccesssobumalicenc¸aCCBY-NC-ND(http://creativecommons.org/licenses/

by-nc-nd/4.0/).

Introduction

Cardiovascular diseases are the main cause of death in Brazil1 _{and worldwide.}2 _{Cardiovascular risk factors, such} astotalandvisceralobesity,dyslipidemias,arterial hyper-tension, hyperglycemia, and hyperinsulinemia have been considereddeterminantsforthedevelopmentof cardiovas-cular diseases. Metabolic syndrome (MetS), characterized by the presence of threeor more of these risk factors,3 hasgainedimportanceduetoitsconsistentassociationwith cardiovascularmorbimortality.4

The presence of cardiovascular risk factors and MetS has been observed in adults5 _{as well as in children and} adolescents.6 _{In recentyears,concerns about the} diagno-sisofand earlyintervention onthesemetabolic disorders hasincreasedduetoevidencethatriskfactorsobservedin childhoodand adolescence tend topersist and worsen in adulthood.7

Inadditiontobloodpressure(BP)andwaist circumfer-encemeasurements,biochemicalanalysesofbloodfractions of HDL-c, triglycerides, and glucose should be performed for thediagnosis ofMetS. Bloodcollectionis an invasive, expensiveanddifficulttoperformtechnique,especiallyin childrenandadolescents.Inturn,theuseofpracticaland low-costmethodsmaybeanimportantalternativeforthe screening of MetS at the population level. In this sense, skinfolds(SFs)wereshowntobeapromisingtoolfor MetS screeninginthepediatricpopulation,duetoitsstrong cor-relationwithsubcutaneousadiposity.8

According to Ali et al.,9 _{the accumulation of} subcuta-neousadiposity is a strong predictor of insulin resistance andhypertriglyceridemiainchildrenand adolescents,and wasastrongerpredictorofcardiometabolicriskfactorsthan

visceral fat. Furthermore, the influence of subcutaneous adiposityonMetSriskispresentinchildrenandadolescents, butnotinadults.9

Percentile curves for SFs have been developed with samples of young North-American,10 _Polish,11 _{and Indian} populations.12 _{It is noteworthy that the World Health} Organization13_{presentedanimportantpublicationin2007,} in which reference curves for anthropometric measure-ments,includingtricepsandsubscapularSFswerecreated asaninternationalreferenceformulti-countrypopulations (Brazil, United States, Ghana,India, Norway, andOman). However, tothe bestof the authors’ knowledge, SF per-centilecurveshavenotbeenproposedforBrazilianchildren, norarethereanySFpercentilecurvesthatcanbeusedto predictMetSinthepediatricpopulation.Referencecurves fromotherpopulationsmaynotbeapplicable toBrazilian childrenduetoethnic,cultural,andsocioeconomic differ-ences.

Therefore,thepresentstudyaimedtodesignreference curvesoffourSFs(subscapular,suprailiac,biceps,and tri-ceps)andtodeterminecutoffstopredicttheriskofMetSin childrenaged6---10years.

Methods

A cross-sectional, population-based epidemiological study wascarriedoutwithaprobabilisticsampleofchildrenaged 6---10 yearsfrompublicandprivateschools locatedin the urbanandruralareasofthecityofUberaba,MG,Brazil.

The sample calculationconsidered the numberof chil-dren enrolled in Elementary School (1st to 9th year of schooling), a prevalence of MetS of 50% (unknown preva-lence in the municipality), tolerable error of 3.5% and

31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116

confidencelevelof95%.Theminimumsamplesizewas768 children; after adding 10% to compensate for losses and refusalsand20%tominimizeconfoundingfactors,the sam-pletotaled1014children.

Forsampleselection,theschoolswerestratified accord-ingtothetypeofschoolasmunicipal,state,orprivate.

TheWorldHealthOrganization14_{recommendsthat10---15} samplecollectionpoints(schools)beusedfor epidemiolog-icalsurveys,andthatthe numberofsubjectsin eachage groupshouldvarybetween25and50foreachsite. There-fore, 15 of the 90 eligible schools of the municipality of Uberaba were randomly selected using the random num-bertable.Fortheadequacyandrepresentativenessofthe local population, the numberof children in each stratum was determined proportionally to the number of enroll-ments,accordingtodataprovidedbytheStateEducation Secretariat.Municipalschoolsaccountedfor43.6%of enroll-ments,41.9%ofstudentswereenrolledfromstateschools and14.5%fromprivateschools.

After approval by the Ethics Committee in Research with Human Beings of Universidade Federal do Triângulo Mineiro (ProtocolCEP/UFTM: 1710), the school principals were contacted to obtain authorization and to schedule thecollections.Studentswhomettheinclusioncriteriaand wereinterestedinparticipatingintheresearchreceivedthe informedconsentformfortheinformationandsignatureof their parents and/or guardians.Anthropometry andblood collectionswere performed at the schoolitself,following specificprotocols.

Thebiceps(BSF),triceps(TSF),subscapular(SSSF),and suprailiac (SISF) SFs were obtained using an adipometer (LangeSkinfoldCaliper,England,UK)exertingconstant pres-sure of 10g/mm2_{. Measurements were performed on the}

right sideof thebody, withthreenon-consecutive repeti-tionsforeachmeasurement.Thefinalmeasureconstituted themeanofthethreevalues.Allmeasurementswere per-formed by twoqualified evaluatorssubmitted toprevious trainingandcalibration.Theresultswereinterpretedalone, aswell asby thesum ofthe fourSFs assessed,and were expressedinmillimeters(mm).

BP was measured using a mercury column sphygmo-manometer (Unitec, São Paulo, Brazil), with cuffs of appropriate sizes for the circumference of the children’s arms, in accordance with the VI Brazilian Guidelines for Hypertension.15_{ThreeBPmeasurementswereperformedat} thefirstvisit, afterwhich thefirstwasdiscardedand the meanofthelasttwowereconsidered.Whenthechildhad analterationinsystolicordiastolicBPabovethe90th per-centile,twofurthermeasurementswereperformedontwo differentdaysfollowingthesameprocedureadoptedonthe firstday.15

Thevolunteerswereinvitedtogotoschool,followinga 12-h fast,oncertaindaysandat certaintimes, accompa-niedbytheirparents.Nursingprofessionalscollectedblood samples(8mL)inBDVacutainerTM_{tubes(Becton,Dickinson}

andCompany,NewJersey,USA).Serumsampleswere ana-lyzedforthe measurementofHDL-c andtriglyceridesand plasmaforglycemia.Thesemi-automatedBio200Fanalyzer (Bioplus,São Paulo, Brazil)wasused.Standardized meth-ods quantitatively determined blood variables, following thestandards andtechnicalspecifications ofthe reagents used.

ThediagnosisofMetS16_{wasdeterminedbythepresence} ofatleastthreeofthefollowingalterations:triglycerides ≥100mg/dL;HDL-c<50mg/dL;glycemia≥110mg/dL;waist circumference ≥75th percentile for age and gender; and alteration in BP (diastolic or systolic) >90th percentile adjustedforage,heightandgender.

Variables were tested for their normality using the Kolmogorov---Smirnov test. Outliers were identified and withdrawn using the interval between quartiles. The Mann---Whitney U-test was used to compare independent groupswithnon-parametricdistribution.

Forcomparisonwithotherstudies,the5th,10th,25th, 50th,75th,90th,and95thpercentileswerechosenas refer-encevalues.ReferencecurveswerecreatedusingtheCole LMSmethod.17 _{TheLMSmethodassumesthat,for} indepen-dent data with positive values, the age-specific Box---Cox transformationcanbeemployedtomakethemhavea nor-maldistribution;the L,M,andS valuesarenatural cubic splines with knot sequence without each age range. The sample,ineachgender,wasseparatedintoagegroupswith 100or moreindividuals,thelowest numberconsidered to beadequatefortheLMSmethod.17

The receiveroperating characteristic (ROC) curve was usedtoevaluatethepredictivecapacityofdiagnostictests. TheareasbelowtheROCcurves(AUC)werecalculatedto assessthediscriminatorypoweroftheSFsintheindication ofmetabolicalterationsconstitutingtheMetS.The sensitiv-ityandspecificityvaluesofthe anthropometricindicators werecalculatedforeachcutoffpresentinthesample.The cutoffthatshowedthebestequilibriumbetween sensitiv-ityandspecificity waschosen tooptimizethe association between these two parameters, showing higher accuracy (lowernumberoffalsenegativeandfalsepositivevalues). Thestatisticalsignificanceofeachanalysiswasverifiedby theAUCandbythelowerlimitofthe95%confidenceinterval >0.5.18

Results

Atotalof1480elementaryschoolstudentsfromtheurban andruralareasofUberaba,MG,Brazil,withameanageof 8.55years(SD=1.53years),ofwhom52.2%werefemales, participatedinthestudy.

Girlshad ahigher prevalence of MetS(12.6% vs.8.5%,

p<0.05) and higher SF values in all assessed anatomical points(p<0.05)whencomparedwithboys(Table1).

Percentilecurves(p5,p10,p25,p50,p75,p90,andp95) werecreatedforthefourassessedSFs,aswellasfortheir sum,forbothgenders(Fig.1).Overall,allSFsshoweda lin-earincreaseaccordingtoageandgender,withhighervalues beingobservedingirls.

The four assessed SFs and their sum were accurate in predictingMetSforgirls(Table2)andboys(Table3).Most ofthesuggested valuesareabovethe75th percentilefor ageandgender.

Discussion

ThepresentstudyshowedpercentileSFcurvesofthe cen-tral and peripheral regions of the body, as well as the sumof the fourSFs topredict the risk of MetSaccording

117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235

Table1 Descriptiveskinfoldcharacteristicsofschoolchildrenaged6---10yearsofage,fromthemunicipalityofUberaba,MG, Brazil,bygender.

Variables Girls(n=773) Boys(n=707) p-Value

Mean(SD) Median(min---max) Mean(SD) Median(min---max)

BSF(mm) 10.14(5.15) 9.00(3.00---29.00) 8.57(5.29) 6.67(3.00---28.00) 0.001 TSF(mm) 14.45(5.66) 13.33(5.00---35.00) 12.71(6.67) 10.67(5.00---35.00) 0.001 SSSF(mm) 11.77(7.63) 9.00(4.00---44.00) 9.85(7.50) 7.00(4.00---44.00) 0.001 SISF(mm) 14.40(10.56) 9.67(3.00---51.00) 11.12(10.23) 6.67(3.00---50.00) 0.001  4SFs(mm) 50.93(28.31) 42.0(10.00---163.00) 42.37(29.79) 31.00(11.00---202.67) 0.001 n,samplenumber;SD,standarddeviation;Min,minimumvalue;Max,maximumvalue;TSF,tricepsskinfold;BSF,bicepsskinfold;SSSF, subscapularskinfold;SISF,suprailiacskinfold;4SFs,sumofthefourskinfoldsevaluated.

Note:Significantdifferencebetweengendersp_≤0.05,Mann---Whitneytest.

Table2 Skinfoldcutoffsforpredictingmetabolicsyndromeingirls,byage,municipalityofUberaba,MG,Brazil.

Q3

Age(years) Predictors AUC p-Value 95%CI Cutoff(mm) S(%) Sp(%)

6 BSF 0.832 0.014 0.709 0.955 11.6 80 79 TSF 0.771 0.044 0.566 0.977 15.3 80 82 SSSF 0.864 0.007 0.760 0.967 9 100 74 SISF 0.805 0.024 0.666 0.943 12.3 80 75  4SFs 0.833 0.013 0.708 0.959 40.3 100 68 7 BSF 0.894 <0.001 0.81 0.977 13.0 88 85 TSF 0.927 <0.001 0.874 0.98 17 100 83 SSSF 0.924 <0.001 0.871 0.978 13.6 100 86 SISF 0.909 <0.001 0.841 0.977 13.6 100 76  4SFs 0.927 <0.001 0.872 0.982 57.3 100 83 8 BSF 0.803 <0.001 0.705 0.902 10 87 65 TSF 0.838 <0.001 0.751 0.924 17.6 73 84 SSSF 0.874 <0.001 0.775 0.974 14.3 80 84 SISF 0.828 <0.001 0.727 0.929 12.3 87 65  4SFs 0.858 <0.001 0.770 0.945 57.6 80 78 9 BSF 0.862 <0.001 0.773 0.951 14.0 86 79 TSF 0.844 <0.001 0.731 0.956 17.6 85 71 SSSF 0.88 <0.001 0.802 0.959 14 86 80 SISF 0.84 <0.001 0.752 0.927 14.3 100 60  4SFs 0.871 <0.001 0.788 0.954 66 86 78 10 BSF 0.764 <0.001 0.627 0.902 14.3 68 80 TSF 0.756 <0.001 0.629 0.884 19 75 81 SSSF 0.823 <0.001 0.701 0.945 15.6 79 78 SISF 0.78 <0.001 0.648 0.911 18.6 80 71  4SFs 0.791 <0.001 0.659 0.923 71.6 80 80

AUC,areaunderthereceiveroperatingcharacteristiccurve;95%CI,95%confidenceinterval;S,sensitivity;Sp,specificity;TSF,triceps skinfold;BSF,bicepsskinfold;SSSF,subscapularskinfold;SISF,suprailiacskinfold;4SFs,sumofthefourassessedskinfolds.

to age and gender in Brazilian children. The curves out-lined here may be a useful strategy to prevent the risk ofMetSinchildhood,withthepossibilityofbeingusedin schools,familyhealthunits,clinics,andhospitals. Accord-ingtothe present findings, both the centraladiposity, as well as the peripheral adiposity SFs and the sum of four SFsshowedcutoffswithhighsensitivityandspecificity val-uestopredict theriskof MetSinchildren. Approximately 75% of the children in the present study with SF values abovethe75thpercentile,regardlessoftheassessedSF,had MetS.

TheprevalenceofMetSinBrazilianchildrenand adoles-cents varied from0% to42.4%, according todata from a systematicreview.19 _{Inthepresentstudy,MetSwas} preva-lent in 12.6% of the girls and 8.5% of the boys, with a significant difference between the genders. The hetero-geneityofdefinitionsandcutoffsfortheMetScomponents mayexplain,atleastpartly,thedifferentprevalencerates reported in the literature.19 _{However, it is true that the} prevalenceofMetShasbeenincreasingamongchildrenand adolescents, witha significantly higher proportionamong thosewhoareobese.20

236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257

Girls Boys Age (years) Age (years) Age (years) Age (years) Age (years) Age (years) Age (years) Age (years) Age (years) Age (years) Biceps skinfold (mm) T riceps skinfold (mm) Subscapular skinfold (mm) Suprailiac skinfold (mm)

Sum of four skinfolds

(mm)

Sum of four skinfolds

(mm) Suprailiac skinfold (mm) Subscapular skinfold (mm) T riceps skinfold (mm) 25 20 15 10 5 0 25 30 40 30 20 10 0 40 30 20 10 0 40 50 30 20 10 150 100 50 0 150 100 50 0 0 40 50 30 20 10 0 40 30 20 10 0 20 15 10 5 0 Biceps skinfold (mm) 25 20 15 10 5 0 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 5 6 7 8 9 10 p5 p5 p5 p5 p5 p5 p5 p5 p5 p5 p50 p50 p95 p95 p95 p95 p95 p95 p50 p50 p50 p50 p50 p50 p50 p50 p95 p95 p95 p95

Figure1 Percentilecurves(p5,p10,p25,p50,p75,p90,andp95)ofthetriceps,biceps,subscapular,andsuprailiacskinfolds,as wellasfortheirsum,ofgirlsandboysaged6---10years,Uberaba,MG,Brazil.

Thepercentiledistributionofthetricepsand subscapu-larSFswereperformedin8568Chineseschoolchildrenaged 7---18yearsofage21_{and32,783North-Americanchildrenand}

adolescents.10 _{The percentile values for TSF and SSSF of} Chineseand American children were lower than the per-centilesofboysandgirlsinthepresentstudy.Thisresultis

258 259 260 261 262 263

Table3 Skinfoldcutoffsforpredictingmetabolicsyndromeingirls,byage,municipalityofUberaba,MG,Brazil.

Age(years) Predictors AUC p-Value 95%CI Cutoff(mm) S(%) Sp(%)

6 BSF 0.95 0.002 0.902 1.00 13.3 100 91.5 TSF 0.95 0.002 0.893 1.00 16.0 100 88.7 SSSF 0.93 0.004 0.854 1.00 8.6 100 82.6 SISF 0.90 0.007 0.807 0.997 11.3 100 81.6  4SFs 0.94 0.003 0.87 1.00 51.3 100 86.1 7 BSF 0.86 0.007 0.726 0.994 8.3 80 72.2 TSF 0.83 0.015 0.681 0.969 11.6 80 72.6 SSSF 0.90 0.003 0.789 1.00 7.6 100 71.6 SISF 0.82 0.016 0.635 1.00 7.0 60 60.7  4SFs 0.86 0.007 0.71 1.00 32.3 100 66.2 8 BSF 0.84 0.044 0.641 1.00 8.3 66.6 63.1 TSF 0.84 0.048 0.626 1.00 11.6 66.6 61.1 SSSF 0.88 0.025 0.756 1.00 9.3 100 73.9 SISF 0.91 0.016 0.821 1.00 13.3 100 82.5  4SFs 0.89 0.024 0.761 1.00 42.6 100 75.5 9 BSF 0.86 0.001 0.767 0.955 10 87.5 68.2 TSF 0.86 0.001 0.769 0.954 14.6 88.8 72.0 SSSF 0.92 0.001 0.868 0.980 12.3 100 83.7 SISF 0.92 0.001 0.852 0.982 14.6 88.8 81.4  4SFs 0.90 0.001 0.832 0.974 55.3 88.8 81.4 10 BSF 0.80 0.001 0.678 0.930 11.3 86.6 70.9 TSF 0.78 0.001 0.645 0.917 17.3 80 79.6 SSSF 0.77 0.002 0.638 0.905 13.3 80 72.3 SISF 0.79 0.001 0.653 0.920 22.0 73.3 86.3  4SFs 0.79 0.001 0.661 0.920 61.6 80.0 80.3

AUC,areaunderthereceiveroperatingcharacteristiccurve;95%CI,95%confidenceinterval;S,sensitivity;Sp,specificity;TSF,triceps skinfold;BSF,bicepsskinfold;SSSF,subscapularskinfold;SISF,suprailiacskinfold;4SFs,sumofthefourassessedskinfolds.

ofconcern,duetotheassociationbetweenSFsandcentral obesity,unfavorablelipidprofile,increasedlevelsofinsulin andBP,andleftventricularmass.22_{Theaccumulationof} sub-cutaneousfat, translatedintohighSFvalues,increasethe chancesofchildrenhavingmetabolicalterations.

Pre-pubertal Mexican children aged 6---10 years had a three-foldhigherchanceofhavingalterationsinMetS com-ponents when the distribution of SSSF was in the fourth quartileofthesample(lowHDL-clevels[OR=3.16;95%CI: 1.41---7.10;p<0.01])andhightriglyceridelevels(OR=3.27; 95%CI:2.02---5.29;p<0.001).23_{However,highvaluesof3SF} (TSF+BSF+SSSF;>90thpercentile)inGermanchildrenaged 3---11yearsincreasedthechanceofhavingthreeormore car-diovascularriskfactorsby1.6-fold(95%CI:1.1---2.2;p<0.05) andthechanceofhavingarterialhypertensionby1.7-fold (95%CI:1.1---2.7;p<0.05)].24

The results ofthe present studydemonstrated thatall isolatedSFmeasurespredict theriskofMetSinboth gen-ders.Inthescientificinvestigationsrelatedtothesubject, studies withTSF andSSSF arepredominant.10,13,21,25 _Many studies have used SF values alone as a predictor of the amount of body fat25---29 _{and isolatedcardiometabolic risk} factors.12,30_{Conversely,studiesinvestigatingthepowerofSF} inpredictingMetSinthepediatricpopulationremainscarce. The use ofthe sumof SFabsolute valuesmaybecome an interesting predictor of MetS, as it minimizes biases in predictive body composition equations, as well as

suggesting values that show equilibrium/disequilibrium of bodyfatdistribution(TSF+BSF+SSSF+SISF).Amongthe dif-ferentanthropometricmeasurestested,4SFwasthemost accuratepredictorofMetSinBraziliangirlsandboys,with AUC=0.908andAUC=0.897,respectively.6

Inthepresentstudy,the4SFshowedanAUCof0.859 for girls and 0.879 for boys, and the suggested cutoffs showedhighsensitivityvalues(>80%).

Tothe bestof theauthors’ knowledge,this is thefirst studywithpercentilecurvesof thecentralandperipheral regionSFs,aswellasthesumoffourSFs,usedtopredict MetSinarepresentativesampleofBrazilianschools. How-ever,thisstudyhadacross-sectionaldesign,whichdidnot allow the determination of whether alterations in subcu-taneousadiposityreflectchangesintheMetScomponents. Thus,longitudinalstudiesarenecessarytodefinethecausal associationbetweenMetSandsubcutaneousadiposity.Itis worthmentioning that,althoughthestudypopulationwas delimitedamongstudentsfromthe1sttothe5thyearof Ele-mentary School,the StateEducation Secretariatprovided updated data onall Elementary School years, which may haveoverestimatedthesamplecalculation.

Referencecurvesfromothercountriescaneither under-estimate or overestimate the disease due to ethnic, socioeconomic,andculturaldifferences.Thedevelopment ofpercentilecurveswithpopulationsofdifferentgeographic areas is a relevant challenge to allow a more accurate

264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317

screening of MetSin young individuals through the evalu-ationofSFs.

Funding

FAPEMIG.

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

References

1.MansurAP,FavaratoD.Mortalityduetocardiovasculardiseases inBraziland inthemetropolitanregionofSaoPaulo:a 2011 update.ArqBrasCardiol.2012;99:755---61.

2.WorldHealthOrganization.Theglobalburdenofdisease:2004 update.Geneva:WHO;2008.

3.Expert panel on detection evaluation, treatment of high bloodcholesterolin adults. Executive summaryof the third reportof thenationalcholesteroleducationprogram(NCEP) expertpanelondetection,evaluation,andtreatmentofhigh bloodcholesterolinadults(AdultTreatmentPanelIII).JAMA. 2001;285:2486---97.

4.IsomaaB, Almgren P, TuomiT, Forsen B, Lahti K, Nissen M, etal.Cardiovascular morbidityandmortalityassociatedwith themetabolicsyndrome.DiabetesCare.2001;24:683---9. 5.SrinivasanSR,MyersL,BerensonGS.Predictabilityofchildhood

adiposityandinsulinfordevelopinginsulinresistancesyndrome (syndromeX)in young adulthood:theBogalusa HeartStudy. Diabetes.2002;51:204---9.

6.AndakiAC,TinocoAL, MendesEL, AndakiJunior R,Hills AP, AmorimPR.Anthropometryandphysicalactivity levelinthe predictionof metabolic syndrome in children. PublicHealth Nutr.2014;17:2287---94.

7.KatzmarzykPT,PerusseL,MalinaRM,BergeronJ,DespresJP, BouchardC.Stabilityofindicatorsofthemetabolicsyndrome fromchildhoodandadolescencetoyoungadulthood:the Que-becFamilyStudy.JClinEpidemiol.2001;54:190---5.

8.MisraA,VikramNK,AryaS,PandeyRM,DhingraV,Chatterjee A,etal.Highprevalenceofinsulinresistanceinpostpubertal AsianIndianchildrenisassociatedwithadversetruncalbodyfat patterning,abdominaladiposityandexcessbodyfat.IntJObes RelatMetabDisord.2004;28:1217---26.

9.AliO, Cerjak D,Kent JW Jr, James R, BlangeroJ, ZhangY. Obesity,centraladiposityand cardiometabolicriskfactors in childrenandadolescents:afamily-basedstudy.PediatrObes. 2014;9:e58---62.

10.AddoOY,HimesJH.Referencecurvesfortricepsand subscapu-larskinfoldthicknessesinUSchildrenandadolescents.AmJ ClinNutr.2010;91:635---42.

11.JaworskiM,Kulaga Z,PludowskiP,Grajda A, GurzkowskaB, NapieralskaE,et al.Population-basedcentile curves for tri-ceps,subscapular,andabdominalskinfoldthicknessesinPolish children and adolescents --- the OLAF study. Eur J Pediatr. 2012;171:1215---21.

12.KhadilkarA, Mandlik R,Chiplonkar S, Khadilkar V,EkboteV, Patwardhan V. Reference centile curves for triceps skinfold thicknessforIndianchildrenaged5 to17yearsandcutoffs forpredicting riskofchildhoodhypertension:amulti-centric study.IndianPediatr.2015;52:675---80.

13.World Health Organization. WHO child growth standards: head circumference-for-age, armcircumference-for-age, tri-cepsskinfold-for-ageandsubscapularskinfold-for-age.Methods anddevelopment.Geneva:WHO;2007.

14.WorldHealthOrganization.Oralhealthsurveys:basicmethods. Geneva:WorldHealthOrganization;1997.

15.Sociedade Brasileira de Cardiologia, Sociedade Brasileira de Hipertensão,SociedadeBrasileiradeNefrologia.VIdiretrizes brasileiras de hipertensão. Arq Bras Cardiol. 2010;95 Suppl. 1:1---51.

16.deFerrantiSD,GauvreauK,LudwigDS,NeufeldEJ,Newburger JW,RifaiN.Prevalence ofthemetabolicsyndromein Ameri-canadolescents:findingsfromtheThirdNationalHealth and NutritionExaminationSurvey.Circulation.2004;110:2494---7. 17.ColeTJ,GreenPJ.Smoothingreferencecentilecurves:theLMS

methodandpenalizedlikelihood.StatMed.1992;11:1305---19. 18.SchistermanEF,FaraggiD,ReiserB,TrevisanM.Statistical

infer-encefortheareaunderthereceiveroperatingcharacteristic curveinthepresenceofrandommeasurementerror.AmJ Epi-demiol.2001;154:174---9.

19.Tavares L, YokooE, Rosa M, FonsecaS. Metabolic syndrome inBrazilianchildrenandadolescents:systematicreview.Cad SaudeColet.2010;18:469---76.

20.FriendA,CraigL,TurnerS.Theprevalenceofmetabolic syn-dromeinchildren:asystematicreviewoftheliterature.Metab SyndrRelatDisord.2013;11:71---80.

21.Ying-XiuZ,Shu-RongW.Distributionofskinfoldthicknessand bloodpressureamong childrenand adolescentsinShandong, China.JTropPediatr.2011;57:258---62.

22.FreedmanDS,SerdulaMK,SrinivasanSR,BerensonGS.Relation ofcircumferencesandskinfoldthicknessestolipidandinsulin concentrationsinchildrenandadolescents:theBogalusaHeart Study.AmJClinNutr.1999;69:308---17.

23.Ramirez-VelezR,Suarez-OrtegonMF,AguilardePlataAC. Asso-ciation between adiposity and cardiovascular risk factors in prepubertalchildren.EndocrinolNutr.2011;58:457---63. 24.HaasGM,LiepoldE,SchwandtP.Predictingcardiovascularrisk

factorsbydifferentbodyfatpatternsin3850Germanchildren: thePEPfamilyheartstudy.IntJPrevMed.2011;2:15---9. 25.KriemlerS, PuderJ,Zahner L,Roth R,Meyer U,Bedogni G.

Estimationofpercentagebodyfatin6-to13-year-oldchildren byskinfoldthickness,bodymassindexandwaistcircumference. BrJNutr.2010;104:1565---72.

26.FreedmanDS,WangJ,OgdenCL,ThorntonJC,MeiZ,Pierson RN,etal.ThepredictionofbodyfatnessbyBMIandskinfold thicknesses among children and adolescents. Ann Hum Biol. 2007;34:183---94.

27.Ayatollahi SM,Mostajabi F.Triceps skinfoldthickness centile chartsinprimaryschoolchildreninShiraz,Iran.ArchIranMed. 2008;11:210---3.

28.BedogniG,IughettiL,FerrariM,MalavoltiM,PoliM,Bernasconi S,etal.Sensitivityandspecificityofbodymassindexand skin-foldthicknessesindetectingexcessadiposityinchildrenaged 8---12years.AnnHumBiol.2003;30:132---9.

29.Pecoraro P, Guida B, Caroli M, Trio R, Falconi C, Princi-patoS, etal. Bodymassindexand skinfoldthickness versus bioimpedanceanalysis: fat masspredictionin children.Acta Diabetol.2003;40:S278---81.

30.Quadros TM,Gordia AP,SilvaRC, SilvaLR. Predictive capac-ityofanthropometricindicatorsfordyslipidemiascreeningin childrenandadolescents.JPediatr(RioJ).2015;91:455---63.

318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432