J. Pediatr. (Rio J.) vol.92 número4

www.jped.com.br

ORIGINAL

ARTICLE

Prediction

equations

for

spirometry

in

four-to

six-year-old

children

夽

Danielle

Corrêa

Franc

¸a

_,

_Paulo

_Augusto

_Moreira

_Camargos

_,

Marcus

Herbert

Jones

_,

_Jocimar

_Avelar

_Martins

_,

_Bruna

_da

_Silva

_Pinto

_Pinheiro

_Vieira

_,

Enrico

Antônio

Colosimo

_,

_Karla

_Morganna

_Pereira

_Pinto

_de

_Mendonc

_¸a

_,

Raíssa

de

Oliveira

Borja

_,

_Raquel

_Rodrigues

_Britto

_,

_Verônica

_Franco

_Parreira

a_{RehabilitationSciencesGraduateProgram,UniversidadeFederaldeMinasGerais(UFMG),BeloHorizonte,MG,Brazil} b_{PediatricPulmonologyUnit,HospitalUniversitário,UniversidadeFederaldeMinasGerais(UFMG),BeloHorizonte,MG,Brazil} c_{PediatricRespirologyDivision,HospitalSãoLucas,PontifíciaUniversidadeCatólicadoRioGrandedoSul(PUCRS),PortoAlegre,}

RS,Brazil

d_{HospitalArnaldoGavazza,PonteNova,MG,Brazil}

e_{PhysiotherapyDepartment,UniversidadeFederaldeMinasGerais(UFMG),BeloHorizonte,MG,Brazil} f_{DepartmentofStatistics,UniversidadeFederaldeMinasGerais(UFMG),BeloHorizonte,MG,Brazil} g_{PhysiotherapyDepartment,UniversidadeFederaldoRioGrandedoNorte(UFRN),Natal,RN,Brazil}

Received26March2015;accepted19October2015 Availableonline7May2016

KEYWORDS

Spirometrychildren; Child;

Predictionequations; Qualitycontrol; Preschool; Equipment

Abstract

Objective: Togeneratepredictionequationsforspirometryin4-to6-year-oldchildren. Methods: Forcedvital capacity,forcedexpiratoryvolume in0.5s,forcedexpiratoryvolume inonesecond,peakexpiratoryflow,andforcedexpiratoryflowat25---75%oftheforcedvital capacitywereassessedin195healthychildrenresidinginthetownofSeteLagoas,stateofMinas Gerais,SoutheasternBrazil.Theleastmeansquaresmethodwasusedtoderivetheprediction equations.Thelevelofsignificancewasestablishedasp<0.05.

Results: Overall, 85% of the children succeeded inperforming the spirometric maneuvers. In the prediction equation, height was the single predictor of the spirometric variables as follows: forced vital capacity=exponential [(−2.255)+(0.022×height)], forced expira-tory volume in 0.5s=exponential [(−2.288)+(0.019×height)], forced expiratory volume in onesecond=exponential [(−2.767)+(0.026×height)],peak expiratory flow=exponential

夽

Pleasecitethisarticleas:Franc¸aDC,CamargosPA,JonesMH,MartinsJA,VieiraBP,ColosimoEA,etal.Predictionequationsforspirometry infour-tosix-year-oldchildren.JPediatr(RioJ).2016;92:400---8.

∗_{Correspondingauthor.}

E-mails:veronicaparreira@yahoo.com.br,veronica.parreira@pq.cnpq.br(V.F.Parreira).

http://dx.doi.org/10.1016/j.jped.2015.10.005

0021-7557/©2016SociedadeBrasileiradePediatria.PublishedbyElsevierEditoraLtda.ThisisanopenaccessarticleundertheCCBY-NC-ND

Predictionequationsforspirometricparameters 401

[(−2.908)+(0.019×height)],andforcedexpiratoryflowat25---75%oftheforcedvital capac-ity=exponential[(−1.404)+(0.016×height)].Neitheragenorweightinfluencedtheregression equations.Nosignificantdifferencesinthepredictedvaluesforboysandgirlswereobserved. Conclusion: The predicted values obtained in the present study are comparable to those reportedforpreschoolersfrombothBrazilandothercountries.

©2016SociedadeBrasileiradePediatria.PublishedbyElsevierEditoraLtda.Thisisanopen accessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/ 4.0/).

PALAVRAS-CHAVE

Espirometriainfantil; Crianc¸a;

Equac¸õesde predic¸ão; Controlede qualidade; Pré-escola; Equipamentos

Equac¸õesdepredic¸ãodaespirometriaemcrianc¸asdequatroaseisanos

Resumo

Objetivo: Gerarequac¸õesdepredic¸ãodaespirometriaemcrianc¸asdequatroaseisanos. Métodos: Capacidadevitalforc¸ada,volumeexpiratórioforc¸adoem0,5segundo,volume expi-ratórioforc¸adoem1segundo,picodefluxoexpiratórioefluxoexpiratórioforc¸adocom25-75% dacapacidadevitalforc¸adaforamavaliadosem195crianc¸assaudáveisqueresidemnacidade deSeteLagoas,EstadodeMinasGerais,SudestedoBrasil.Ométododosmínimosquadrados foiusadoparaderivarasequac¸õesdepredic¸ão.Oníveldesignificânciafoiestabelecidocomo p<0,05.

Resultados: Nogeral,85%dascrianc¸asforambem-sucedidasaofazerasmanobras espirométri-cas. Na equac¸ão de predic¸ão, a estatura foi a única variável preditora das variáveis espirométricas,daseguinteforma:capacidadevitalforc¸ada=exponencial[(-2,255)+(0,022 x estatura)], volume expiratório forc¸ado em 0,5 segundo = exponencial [(-2,288) +(0,019 x estatura)], volume expiratório forc¸ado em 1 segundo = exponencial [(-2,767) + (0,026 x estatura)],picodofluxoexpiratório=exponencial[(-2,908)+(0,019xestatura)]efluxo expi-ratório forc¸ado com25-75%da capacidade vital forc¸ada= exponencial [(-1,404)+(0,016 x estatura)].Nemaidadenemopesoinfluenciaramasequac¸õesderegressão.Nãofoiobservada diferenc¸asignificativanosvaloresprevistosemmeninosemeninas.

Conclusão: Osvaloresprevistosobtidosnesteestudosãocomparáveisàqueles relatadosem crianc¸asemidadepré-escolartantodoBrasilquantodeoutrospaíses.

©2016SociedadeBrasileiradePediatria.PublicadoporElsevierEditoraLtda.Este ´eumartigo OpenAccesssobumalicenc¸aCCBY-NC-ND(http://creativecommons.org/licenses/by-nc-nd/4. 0/).

Introduction

Pulmonary functiontestssupplement theclinical histories and physical examinations of individuals with respiratory problems and contribute to the diagnosis, prognosis, and monitoring ofrespiratory diseases,aswell asassessments ofthetherapeuticeffectsofinterventions.1

Spirometryisauniversallyacceptedmethodforassessing pulmonaryfunction,includinglungvolumesandflows,andis themostwidelyusedmethodfordetectingfunctional alter-ations in the pulmonary functions of adults, adolescents, andschoolchildren.1,2_{However,spirometryhasnotyetbeen} widely studied in Latin American or Brazilian preschool populations.1,3

The earliest studies that assessed spirometric maneu-versinpreschoolerswerepublishedapproximately20years ago.4,5_{Currently,itiswellestablishedintheliteraturethat} 75---86% ofpreschoolers areabletoacceptablyand repro-duciblyperformspirometricmaneuvers1,3,6---10 _asevidenced by the availability of prediction equations for preschool-ers from several countries.6,8,9,11---17 _{Additionally, a recent} multicenterstudy16_{proposedpredictionequationsfor} spiro-metricvariablesfrompreschoolersbasedondatacollected from the children from 11 different countries (n=3777),

including Brazil. An update was published in 2012 and included new multi-ethnic equations.18 _{Recently, Burity} et al.19 _{described reference values based on 135} north-easternBrazilianpreschoolersandidentifiedheightandsex aspredictorsof lungfunctionparameters.However,these equationsarenot necessarilyrepresentative ofthe entire populationofBrazilianpreschoolers.

Duetotheimportanceofspirometryinallagegroupsand thescarcityof predictionequations for spirometric varia-blesforpreschoolersinBrazil,thepresent studyaimedto describethepredictionequationsforspirometryin4-to 6-year-oldchildrenresidinginthetownofSeteLagoas,state ofMinasGerais,southeasternBrazil.

Methods

Settingandsampling

DiseaseATS/DLD/78/C-questionnaire(i.e.,totalscoresthat are equal to or lower than 6)6,20_{; body mass index (BMI)} betweenthe3rdand95thpercentilesforage21_;birthweight greaterthan2500gatagestationalageof37---42weeks20_; nofluepisodeswithintheprevioussevendays;nohistoryof maternalsmokingduringpregnancyorexposuretotobacco smokingover the years preceding the spirometric assess-ment;nopasthistoryofthoracicorabdominalsurgery;no thoracicmalformations,geneticsyndromes,metabolic dys-functions,heart diseases,neuromusculardiseases, mental disorders,orcognitivedeficits;andnouseofmedicineson aregularbasis.

The study protocol and informedconsent (providedby theparentsorlegalguardians)wereapprovedbytheEthics inResearchCommitteeoftheUniversidadeFederaldeMinas Gerais(ETICNo.0612.0.203.000-09).

Measurementinstrumentsandprocedures

The ATS-DLD-78-C questionnaire wasused torule outthe presenceof respiratorydiseases and exposureto environ-mentalpollutants.Thisquestionnaireconsistsofquestions about the symptoms, other illnesses and hospitalizations relatedtorespiratorytractdiseases,anditwascompleted bytheparentsorlegalguardians.Theoverallscoresranged from0to22, andscores greater thanor equal to7were indicativeofachronicrespiratorydisease.6_{Supplementary} questions, for example, questions about gestational age, the neonatalperiod and concomitant diseases, were also included.

A handheld Koko spirometer (Koko®_{, PFT type; nSpire}

HealthInc., CO, USA) wasused toassess thespirometric variablesobtainedfromthe forcedexpiratorymaneuvers. This equipment meets all of the requirements of the ATS/ERS1_{anddisplaysreal-timegraphsoftheflow/volume} andvolume/time curves.Following the ATS recommenda-tions,adisposablemouthpieceandantibacterialfilterthat togetherresultin adeadspace oflessthan2mL/kgwere used. The accuracy of the equipment was verified daily

viacalibrationusingathree-litermanualsyringe(COSMED, PulmonaryFunctionEquipment---Italy).Thefollowing spiro-metricvariableswereanalyzed:FVC,FEV0.5,FEV1,FEF25---75,

andPEF.Thedatafortheforcedexpiratorytime(FET)and back-extrapolatedvolume(BEV)were collectedtoensure thequalitiesofthecurves.1

The data from all of the questionnaires were ana-lyzed,andthechildrenwhomettheinclusioncriteriawere invited to participate in the study. Children whose par-ents/guardiansauthorized theirparticipation inthe study andsignedaninformedconsentformwereassessedduring regularschoolactivities.

The physical examinationswereinitiated by measuring theweightusingasinglescale(Filizola---SãoPaulo,Brazil) andtheheightusingaHarpenden®_{stadiometer(Harpenden}

SkinfoldCaliper,MediflexSurgicalProducts,USA) (measure-mentrangefrom60.0cmto210.0cm,wallmounted,witha precisionof0.1cm,withthechildstanding,barefoot, pos-itionedinsuchamannerthatthehead,shoulders,buttocks, andheelslightlytouchedtheverticalsurfaceof the mea-suring device). Subsequently, blood pressure, heart rate, respiratoryrate, oxygensaturation, andlung auscultation

were assessed,and staticand dynamic inspections of the breathingpatternwereperformed toconfirm theabsence ofanyclinicalabnormality.

To perform spirometry, the recommended acceptabil-ityandreproducibilitycriteriawerefollowed.1_Beforethe test, eachchild receivedage-appropriateinstructions and then performed the maneuvers. The measurements were collectedwiththesubjectseatedusingadisposable well-adjusted mouthpiece(to preventleaks), a nose clip, and computer-animation programs that provided instructions andstimulitotheparticipants.2_{Thesubjectswereaskedto} breatheinthetidalvolumeandtheninstructedtofilltheir lungsasmuchaspossibleandblowasstrongly,rapidlyand longaspossible.Theacceptabilitycriteriawereasfollows: the flow-volume andvolume-time curves were inspected, andmaneuversthatwerevisiblyinadequatewereexcluded,

i.e.,whentheflow-volumecurvedidnotexhibitarapidrise tothepeakflowandasmoothdescendinglimbwithno evi-dence ofacough orglottic closure;theBEVwasequalor lowerthan80mLorlowerthan12.5%oftheFVC;andthe FETwasgreaterthan10%ofthePEFrate.The reproducibil-itycriteriawereasfollows:thedifferencebetweentheFVC and FEV1 was lower than 10% or 100mL for at least two

reproduciblemaneuvers.1_{Thetestwaslimitedto20minto} ensureclinicalapplicability,andthesubjectswereallowed oneminuteofrestbetweentwoconsecutivemaneuversthat wereconductedbythesameinvestigator(DCF)in all sub-jects.

Thevolume---timeandflow---volumecurveswerevisually examinedbytwoindependentobserverswhowereblinded to the children’s past and current health status. Specifi-cally,themaximumvaluesoftheFVC,FEV0.5,FEV1,andPEF

wereobtainedusingtheappropriateandreproduciblecurve (themaximumvaluesforeachofthevariablescouldcome fromdifferentcurves).TheFEF25---75valuewasselectedfrom

thecurvethatexhibitedthelargestsumof FVCandFEV1,

andtheFEV0.5valuewasobtainedfromthecurvewiththe

greaterFEV1valueorthegreaterFEV0.5whentheFETwas

≤1s.1

Statisticalanalysis

Accordingtotheageranges(i.e.,48---59,60---71,and72---83 months)andsex(boysandgirls),180healthypreschoolers werearbitrarilypre-definedwithinthetargetagerangeas theminimumthatshouldbeassessedinthepresentstudy. Acluster randomizationwasperformed basedonthe offi-cial list of the existing public and private schools in the cityofSeteLagoas.Elevenschoolswererandomlyselected (sevenpublicandfourprivateinstitutions)while maintain-ingapublic/privateschoolratioof1.75,which wasbased ontheproportionsofregisteredstudents.Then,all4-to 6-year-oldchildrenregisteredineachschoolwereinvitedto participateinthestudy.

Predictionequationsforspirometricparameters 403

theanthropometricdataandvariablesfromthepulmonary functiontestsbetweenthe4-to6-year-oldchildren.

Togeneratethepredictionequationsforthespirometric variables,theleastmeansquaresmethodwasused.First, logarithmictransformations wereappliedtothe spiromet-ricvariableswithBox---Coxanalyses.Linearregressionswere performed betweenthespirometric(dependent) variables and the height, weight, sex, and age (exploratory varia-bles).Then,amultipleregressionanalysiswasperformed. Following the evaluation of the multiple regression anal-ysis, the normality and homoscedasticity of the residuals wereestablishedusingQ---Qplots.Theresidualassumptions weresatisfied.Theresultswereexpressedasz-scores,i.e., multiplesof the standard deviationof the mean,and the graphswiththeregressioncurvesandz-scoreswereplotted tofacilitatetheuseoftheequationsinclinicalpracticeand research.

Alltestswereassessedatanalphaof0.05withno adjust-ments for multiple comparisons. The data were analyzed usingtheRstatisticallanguage(FreeSoftwareFoundation --- Boston,MA,USA)andSPSS(SPSSInc.Released2009.SPSS forWindows,Version15.0,IL,USA).

Results

Fig. 1 depictsthe study flowchart and indicates that 195 children were initiallyassessed andthat 131 (67%) public schooland64(33%)privateschoolstudentswereenrolledin thestudyfromJune2010toAugust2011.

Table 1 presents the coefficients of determination (r2₎

of the regression equations for the spirometric variables withtheindependentvariables ofheight, height+weight, height+age,andheight+weight+age.Ther2_values

exhib-ited little variation (approximately 1%) when height was used as the single predictor of the spirometric variables withheight+weight+age as the independentvariables in theregressionequations.Table1alsopresentsthe regres-sion equations and the lower normal limits (i.e., −1.64

z-scores) for the spirometric variables withheight as the singlepredictor.

The linear regression between the spirometric param-eters (dependent variables) and height, weight, and age (independentvariables)exhibitedastatisticallysignificant association. However, the spirometric variables did not presentastatisticallysignificantassociationwithsex.

The graphs of the spirometric variables according to height in which the z-scores are plotted and shown in Fig. 2. As illustrated, shorter heights were associated withless variabilityin eachofthespirometricparameters studied.

Table 2 presents the proportion of children who suc-cessfully performed spirometry, the anthropometric data, and the spirometric parameters by age of the 195 chil-dren. Overall, 85% of thechildren successfully performed the spirometricmaneuvers,and thesuccess rate progres-sivelyincreasedwithage.Thesamplefromwhichthecurves weregenerated and theequations were derived was pre-dominantlyofmixedethnicity(134outof166,i.e.,80.7%), white(27outof166,i.e.,16.3%),andthoseAsian/African descendants(onlyfiveoutof166,i.e.,3.0%).This informa-tionwasobtainedfromparents’report.Thecomparisonsof

Questionnaires sent to all four- to six-year-old children in seven public

and four private schools. (n=983)

Questionnaires returned (n=704)

Children who met the inclusion criteria

(n=225)

Children whose parents consented to participate

in the study (n=212)

Children who did not meet the inclusion criteria

(n=479)

- 390: incomplete questionnaires (81.4%);

- 67: presence of a cold or allergic rhinitis (14.0%);

- 22: surgical procedure (4.6%).

Children excluded: (n=17)

- 4: changed schools (23.5%);

- 6: missed school on the day of the test (35.3%);

- 7: exhibited flu along with allergic rhinitis (41.2%). Children assessed in the

study (n=195)

Figure1 Studyflowchart.

theanthropometricdataandspirometricvariablesbetween the boys and girls did not reveal significant differences (p>0.05,independentt-tests).Nonetheless,theageranges indicatedthatweight,height,and mostofthe pulmonary functionparametersexhibitedprogressiveandstatistically significantincreaseswithage. RegardingtheFET,a signif-icantincreasewasobservedonlyamongthechildren aged 4---6 years. Only the BEV, FVC, FEV0.5/FVC, and FEV1/FVC amongall of theinvestigatedage rangesfailedtoexhibit statistically significant differences according to Hochberg test.

Discussion

Tothebestoftheauthors’knowledge,thepresentreportis thesecondstudydesignedtodescribepredictionequations for spirometricvariables exclusively for Brazilian 4- to 6-year-oldpreschoolers.

Franc

¸a

DC

et

al.

Table1 Coefficientofdeterminationofthemultipleregressionanalyzesforspirometricvariables,consideringtheexplanatoryvariables:height,height+weight,height+age, andheight+weight+age;aswellasthelowerlimitofnormalandpredictionequationsforspirometricvariablesaccordingtoheight.

Spirometricvariables Coefficientofdetermination(r2_{) Predictionequations Lowerlimitofnormal}

H H+W H+A H+W+A

FVC 0.56a _0.57a _0.56a _0.57a _{Exp[(−2.255)+(0.022×height)] Exp(predictedvalue−0.253)}

FEV0.5 0.51a 0.51a 0.51a 0.52a Exp[(−2.288)+(0.019×height)] Exp(predictedvalue−0.225)

FEV1 0.45a 0.44a 0.44a 0.44a Exp[(−2.767)+(0.026×height)] Exp(predictedvalue−0.294)

PEF 0.35a _0.34a _0.35a _0.35a _{Exp[(−2.908)+(0.019×height)] Exp(predictedvalue−0.220)}

FEF25---75 0.24a 0.24a 0.24a 0.25a Exp[(−1.404)+(0.016×height)] Exp(predictedvalue−0.191)

H,height(cm);W,weight(kg);A,age(months);Exp,exponential;FVC,forcedvitalcapacity(L);FEV0.5,forcedexpiratoryvolumein0.5s(L);FEV1,forcedexpiratoryvolumeinone

P rediction equations for spirometric parameters 405

Table2 Anthropometricdata,successrates,andobservedvaluesofthespirometricvariablesaccordingtoage,amongthe195assessedchildren.

Age(n) Success

rate(%)

Height (cm)

Weight (kg)

FVC(L) FEV0.5(L) FEV0.5/FVC FEV1(L) FEV1/FVC PEF

(L/min)

FEF25---75

(L/s)

FET(s) BEV(%)

Boys

4(30) 76 107±5

(105---110)

19±3 (17---20)

1.15±0.20 (1.06---1.24)

0.86±0.15 (0.80---0.94)

0.76±0.08 (0.72---0.80)

1.04±0.26 (0.92---1.16)

0.90±0.18 (0.81---0.98)

151±29 (139---166)

1.59±0.39 (1.39---1.75)

1.90±0.60 (1.62---2.18)

3.39±1.08 (2.91---3.87)

5(31) 81 115±6

(113---117)

21±4 (19---22)

1.33±0.21 (1.24---1.42)

0.99±0.13 (0.94---1.04)

0.75±0.08 (0.72---0.79)

1.24±0.17 (1.16---1.31)

0.93±0.05 (0.91---0.95)

173±36 (157---188)

1.74±0.32 (1.61---1.88)

1.88±0.60 (1.62---2.13)

3.22±1.29 (2.67---3.76)

6(31) 90 118±5

(116---120)

22±3 (20---23)

1.42±0.23 (1.33---1.51)

1.01±0.16 (0.95---1.08)

0.72±0.06 (0.70---0.74)

1.28±0.20 (1.21---1.36)

0.91±0.04 (0.89---0.92)

172±33 (159---185)

1.69±0.36 (1.55---1.83)

2.32±0.56 (2.10---2.54)

3.24±1.03 (2.84---3.64)

Girls

4(35) 81 109±5

(107---111)

19±4 (18---21)

1.10±0.24 (1.01---1.20)

0.81±0.16 (0.75---0.87)

0.74±0.07 (0.71---0.76)

1.01±0.21 (0.94---1.09)

0.92±0.05 (0.91---0.94)

135±32 (123---147)

1.41±0.34 (1.28---1.54)

2.09±0.66 (1.83---2.33)

3.13±1.05 (2.73---3.54)

5(36) 89 115±5

(113---117)

20±3 (19---21)

1.25±0.18 (1.18---1.31)

0.93±0.13 (0.88---0.98)

0.75±0.07 (0.73---0.78)

1.16±0.15 (1.10---1.21)

0.93±0.05 (0.91---0.95)

166±34 (154---179)

1.67±0.40 (1.53---1.82)

1.89±0.69 (1.64---2.14)

3.60±1.17 (3.17---4.03)

6(32) 100 120±7

(117---122)

24±6 (22---26)

1.48±0.25 (1.39---1.57)

1.08±0.19 (1.01---1.14)

0.73±0.06 (0.71---0.75)

1.35±0.22 (1.27---1.43)

0.91±0.04 (0.90---0.93)

185±36 (172---197)

1.86±0.47 (1.69---2.03)

2.26±0.70 (2.01---2.51)

3.41±6.80 (3.12---3.70)

Total(195) 85 114±7a,b,c

(113---115)

21±4b,c

(20---22)

1.30±0.26

(1.26---1.33)

0.95±0.18a,b,c

(0.93---0.98)

0.74±0.07

(0.73---075)

1.19±0.24a,b,c

(1.15---1.22)

0.92±0.08

(0.91---0.93)

164±30a,c

(159---170)

1.66±0.41a,c

(1.60---1.73)

2.07±0.66b

(1.97---2.17)

3.34±1.07

(3.17---3.50)

FVC,forcedvitalcapacity;FEV0.5,forcedexpiratoryvolumein0.5s;FEV0.5/FVC,ratiobetweenFEV0.5andFVC;FEV1,forcedexpiratoryvolumeinonesecond;FEV1/FVC,ratiobetween

FEV1andFVC;PEF,peakexpiratoryflow;FEF25---75,forcedexpiratoryflowat25---75%oftheFVC;FET,forcedexpiratorytime;BEV,back-extrapolatedvolume.

Thedataarepresentedasthemeans±thestandarddeviationswiththe95%confidenceintervals(inparentheses).

a _{p<0.05betweenthe4-and5-year-oldchildren.} b _{p<0.05betweenthe5-and6-year-olds.}

4.50 2.50 2.00 1.50 1.00 0.50 0.00 450 400 350 300 250 200 150 100 50 0 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 5.00 4.50 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00

90 100 110 120

Height (cm) FVC (L) FEV 1 (L) FEV 0.5 (L) PEF (L/min) FEF 25-75 (L/sec)

130 140 150

+2 +1 0 –1 –2 +2 +1 0 –1 –2 +2 +1 0 –1 –2 +2 +1 0 –1 –2 +2 +1 0 –1 –2

90 100 110 120

Height (cm)

130 140 150 90 100 110 120

Height (cm)

130 140 150

90 100 110 120

Height (cm)

130 140 150

90 100 110 120

Height (cm)

130 140 150

Figure2 Graphs ofthez-scoresfor thespirometricvariablesaccording toheight. FVC,forcedvital capacity;FEV0.5,forced expiratoryvolumein0.5s;FEV1,forcedexpiratoryvolumeinonesecond;PEF,peakexpiratoryflow;FEF25---75,forcedexpiratory flowat25---75%oftheFVC.

maneuvers.In contrast to the present study, the statisti-calanalysesincludedlogarithmicandlinearregressionsand didnotincludetheleastmeansquaresmethod.Thelatter methodallowsfortheconstructionoffunnel-shapedcurves thatbetterreflect thecorrelationsbetween physical lung growthacrosstheanalyzedages(4,5,and6yearsoflife) andfunctionalparameters.

The main results of the present study are the follow-ing:progressiveandstatisticallysignificantincreasesinthe majorityofthespirometricparameterswereobservedwith increasing height, and the resulting prediction equations incorporated height as a single predictor. These results agreewithpreviouslyreportedresults.6,9,15 _Thus,the pre-dictionequationsforspirometricvariablesdescribedinthe presentstudyweregeneratedbyconsideringheightasthe single independent variable. This choice of a single pre-dictorwasbasedonthefollowingfactors:thestatistically non-significant variation in the coefficients of determina-tion withthe addition of the other possible independent variables (i.e.,age, weight, and sex), the ease of use of theequationswithsinglepredictorsinclinicalpractice,and the previous use of height as a single independent vari-ablebyseveral authorstoderivepredictionequations for

spirometricparametersamongpreschoolers.6,8,11,13,15 Addi-tionally,thevaluesforthecoefficientofdetermination(r2₎

thatwerefoundinthepresentstudy(Table1)aresimilarto thosepreviouslydescribed,9,12,15 _{includingthoseofastudy} thatenrollednortheasternBrazilianpreschoolers.19_Overall, ther2_{valuesreportedintheliteratureforFVCrangedfrom}

0.76to0.53;forFEV0.5,from0.73to0.42;for FEV1,from

0.78 to0.55; for PEF,from0.78to0.38; andfor FEF25---75,

from0.37to0.17.9,12,15,19

The graphs of the spirometric variables according to heightinwhichthez-scoreswereplottedreflectthedirect correlations between the spirometric variables and the increaseinheight.Therefore,withaprogressiveincreasein height,thevariablesthatwereassessedinthepresentstudy presentedwidevariability.Thegraphsfrompreviousstudies resembledthoseofthepresentstudyinappearance.16,17

Predictionequationsforspirometricparameters 407

differencesbetweenpreschoolmalesandfemalesandthe smallsamplesize.

Additionally, strong agreements (i.e.,z-scores closeto zero)wereobservedinthecomparisonsofthespirometric valuesgeneratedfromourpredictionequationsandthe val-uesthatwererecentlyreportedbyQuanjerandcoworkers basedon97,759recordsofhealthynonsmokersaged2.5---95 years from around the world.18 _{For example, the mean}

z-scores for the FVC, FEV1, FEV1/FVC ratio, and FEF25---75

obtainedinthepresentstudywere−0.8,−0.00,+0.1,and +0.05,respectively; i.e.,all oftheparametersfell within theacceptedrangeof±0.5z-scores.Inotherwords,these valueswerelowerthanapproximately1---2%ofthepredicted values reported by Quanjer et al.,18 _{which are clinically} irrelevant. However,thepredictionequations proposedin thepresentstudyappeartobemoreapplicableto popula-tions similartothat of thepresent study,simplybecause theseequationsweregeneratedfromsuchapopulation.

ComparedtothereferencevaluesdescribedbyRosenthal etal.4 _{andBurityetal.}19 _{andaccountingfor thechildren} with heights of 110cm, the present predicted values are closetothosethathavebeenpreviouslyreported,i.e.,FVC of 1.20L, 1.10L,and 1.18L,respectively, FEV1 of 1.10L,

1.05L,and1.10L,respectively,andFEV1/FVCof0.90,0.95,

and0.93,respectively.

Oneofthelimitationsofthepresentstudywasthatitdid notinclude3-year-oldchildren,althoughthepreschoolage rangeisconsidered tobeginat thatage.However,access to these children would have been restricted due to the smallnumber of schools in Sete Lagoas, particularly pub-licschools,for childrenyoungerthan4yearsold.Another limitationistherelativelysmallsamplesizeofthisstudy.

In summary,predictionequationsfor spirometric varia-blesweregeneratedfor4-to6-year-oldBrazilianchildren residinginthestateofMinasGerais.Theseresultsmaybe of importance tothe field of pediatric respiratory health becausethesevaluescanbeusedtooptimizethe diagno-sisandclinicalmanagementofchildren withsuspectedor confirmedrespiratorydiseases.Pulmonaryfunctiontestsare importanttoelucidatediagnostic hypothesesandimprove clinical management. Reflecting the importance of their use, these tests are increasingly being used among the preschoolpopulation.Theresultsofthepresentstudymay helptobroadentheuseofspirometryamongpreschoolers, astheseresultscorroboratethefindingsofpreviousstudies thathavereportedhighsuccessratesintheperformanceof spirometricmaneuversby preschoolersandan increasein thesuccessratewithage.3,6,7,9,10,15,22

Given the continental characteristics of Brazil, further studies on this topic across the five macroregions of the countryshouldbeencouraged,andthesestudiesshouldaim to obtain lung function data from a more representative populationofBrazilianchildren.

Funding

Pró-Reitoria de Pesquisa da Universidade Federal de Minas Gerais, Brazil; Brazilian Research Agencies, Con-selhoNacionaldeDesenvolvimentoCientíficoeTecnológico (CNPq,GrantNo.309494/2013-3[toVFP]and 303396/2012-1[toPAMC]),andFundac¸ãodeAmparoàPesquisadoEstado

de Minas Gerais (FAPEMIG, Grant No. PPM-00065-14 [to PAMC]andPPM-00287-15[toVFP]).

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

Acknowledgments

The authors wish to thank Maria Ângela Gonc¸alves de OliveiraRibeiroandElianeViana Mancuzoforreviewing a preliminaryversionofthemanuscriptandfortheir invalu-ableinput.

References

1.BeydonN,DavisSD,LombardiE,AllenJL,AretsHG,AuroraP, etal.AnofficialAmericanThoracicSociety/European Respira-torySocietystatement:pulmonaryfunctiontestinginpreschool children.AmJRespirCritCareMed.2007;175:1304---45.

2.BeydonN.Pulmonaryfunctiontestinginyoungchildren. Paedi-atrRespirRev.2009;10:208---13.

3.VerasTN,PintoLA.Feasibilityofspirometryinpreschool chil-dren.JBrasPneumol.2011;37:69---74.

4.RosenthalM,BainSH,CramerD,HelmsP,DenisonD,BushA, et al. Lung function in white childrenaged 4---19years: I ---spirometry.Thorax.1993;48:794---802.

5.KanengiserS,DozorAJ.Forcedexpiratorymaneuversin chil-drenaged3---5years.PediatrPulmonol.1994;18:144---9.

6.EigenH,BielerH,GrantD,ChristophK,TerrillD,HeilmanDK, etal.Spirometricpulmonaryfunctioninhealthypreschool chil-dren.AmJRespirCritCareMed.2001;163:619---23.

7.MarosticaPJ,WeistAD,EigenH,AngelicchioC,ChristophK, SavageJ,et al.Spirometryin3-to6-year-old childrenwith cysticfibrosis.AmJRespirCritCareMed.2002;166:67---71.

8.Vilozni D, BarakA, Efrati O, Augarten A, Springer C, Yahav Y, et al. The role of computer gamesin measuring spirom-etry in healthy and asthmatic preschool children. Chest. 2005;128:1146---55.

9.NystadW,SamuelsenSO,NafstadP,EdvardsenE,StensrudT, JaakkolaJJ.Feasibilityofmeasuringlungfunctioninpreschool children.Thorax.2002;57:1021---7.

10.AuroraP,StocksJ,OliverC,SaundersC,CastleR,Chaziparasidis G,et al.Qualitycontrolfor spirometryinpreschoolchildren with and without lungdisease. Am J RespirCrit Care Med. 2004;169:1152---9.

11.Pesant C, Santschi M, Praud JP, Geoffroy M, Niyonsenga T, Vlachos-MayerH. Spirometricpulmonary function in3-to 5-year-oldchildren.PediatrPulmonol.2007;42:263---71.

12.PiccioniP,BorraccinoA,FornerisMP,MiglioreE,CarenaC, Big-naminiE,etal.Referencevaluesofforcedexpiratoryvolumes and pulmonary flows in 3---6year children: a cross-sectional study.RespirRes.2007;8:14.

13.Zapletal A, Chalupová J. Forced expiratory parameters in healthypreschoolchildren(3---6yearsofage).PediatrPulmonol. 2003;35:200---7.

14.American Thoracic Society/European Respiratory Society. ATS/ERSStatementonrespiratorymuscletesting.AmJRespir CritCareMed.2002;166:518---624.

15.JengMJ,ChangHL,TsaiMC,TsaoPC,YangCF,LeeYS,etal. Spirometric pulmonary function parameters of healthy Chi-nese children aged 3---6 years in Taiwan. Pediatr Pulmonol. 2009;44:676---82.

theAsthmaUKCollaborativeInitiative.AmJRespirCritCare Med.2009;180:547---52.

17.QuanjerPH,StanojevicS,StocksJ,HallGL,PrasadKV,Cole TJ, et al. Changes in the FEV1/FVC ratio during childhood and adolescence: an intercontinental study. Eur Respir J. 2010;36:1391---9.

18.QuanjerPH,StanojevicS,ColeTJ,BaurX,HallGL,CulverBH, etal.Multi-ethnicreferencevaluesforspirometryforthe 3---95-yragerange:thegloballungfunction2012equations.EurRespir J.2012;40:1324---43.

19.BurityEF,PereiraCA,RizzoJA,BritoMC,SarinhoES.Reference valuesforspirometryinpreschoolchildren.JPediatr(RioJ). 2013;89:374---80.

20.BurityEF, PereiraCA,Rizzo JÂ,SarinhoES,Jones MH.Early termination of exhalation: effect on spirometric parame-tersinhealthypreschool children.JBrasPneumol.2011;37: 464---70.

21.World Health Organization (WHO). Multicentre Growth References Study Group. WHO Child Growth Standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-heightandbodymassindex-for-age:methodsand development.Geneva:WHO;2006.