RevPaulPediatr.2016;34(3):343---351
REVISTA
PAULISTA
DE
PEDIATRIA
www.rpped.com.br
ORIGINAL
ARTICLE
Prevalence
of
abdominal
obesity
in
adolescents:
association
between
sociodemographic
factors
and
lifestyle
João
Antônio
Chula
Castro,
Heloyse
Elaine
Gimenes
Nunes,
Diego
Augusto
Santos
Silva
∗ProgramadePós-Graduac¸ãoemEducac¸ãoFísica,UniversidadeFederaldeSantaCatarina(UFSC),Florianópolis,SC,Brazil
Received11August2015;accepted17January2016 Availableonline13June2016
KEYWORDS
Waistcircumference; Lifestyle;
Anthropometry; Epidemiology; Adolescenthealth; Publichealth
Abstract
Objective: Toestimatetheprevalenceofabdominal obesityandverifytheassociationwith sociodemographic factors(gender,schoolshift, ethnicity,age,maternaleducation and eco-nomic status) and lifestyle (alcohol consumption, sleep, soft drink consumption, level of physicalactivityandsedentarybehavior)inadolescentsinSouthernBrazil.
Methods: Thiswasacross-sectionalepidemiologicalstudyof930adolescents(490girls)aged 14---19years,livinginthecityofSãoJosé,SC,Brazil.Aself-administeredquestionnairewasused tocollectsociodemographicandlifestyledata.Abdominalobesitywasmeasuredthroughthe waistcircumferenceandanalyzedaccordingtogenderandage.Descriptivestatistics(absolute andrelativefrequency,meanandstandarddeviation)andbinarylogisticregression,expressed asOddsRatios (OR)and95%confidenceinterval(95%CI)were employed,withp<0.05being consideredstatisticallysignificant;theSPSS17.0softwarewasusedforthestatisticalanalyses. Results: The prevalenceofabdominalobesity was10.6% for thetotal sample(10.5% male, 10.8%female).Adolescentsthatwatchedtelevisiondailyfortwoormorehours(OR=2.11,95%CI 1.08---4.13)hadahigherchanceofhavingabdominalobesityandadolescentswhosemothers hadfewerthaneightyearsofschooling(OR=0.56;95%CIfrom0.35to0.91)hadalowerchance ofhavingabdominalobesity.
Conclusions: Approximatelyonein10adolescentshadabdominalobesity;theassociated fac-torswerematernalschooling(≥8years)andtelevisionscreentime(≥2h/day).
©2016SociedadedePediatriadeS˜aoPaulo.PublishedbyElsevierEditoraLtda.Thisisanopen accessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).
∗Correspondingauthor.
E-mail:diegoaugustoss@yahoo.com.br(D.A.Silva).
http://dx.doi.org/10.1016/j.rppede.2016.01.007
PALAVRAS-CHAVE
Circunferênciada cintura;
Estilodevida; Antropometria; Epidemiologia; Saúdedo adolescente; Saúdepública
Prevalênciadeobesidadeabdominalemadolescentes:associac¸ãoentrefatores sociodemográficoseestilodevida
Resumo
Objetivo: Estimaraprevalênciadeobesidadeabdominaleverificaraassociac¸ãocomfatores sociodemográficos(sexo,turnodeestudo,cordapele, idade,escolaridadematernaenível econômico)eoestilo devida(consumode álcool,sono,consumo derefrigerante,nível de atividadefísicaecomportamentosedentário)emadolescentesdoSuldoBrasil.
Métodos: Estudo epidemiológico descritivotransversal,feito com930 adolescentes(490do sexofeminino)de14---19anosdeSãoJosé,SC,Brasil.Usou-sequestionárioautoadministrado paracoletardadossociodemográficosedoestilodevida.Aobesidadeabdominalfoiavaliada peloperímetrodacintura eanalisadadeacordocomsexo eidade.Empregou-seestatística descritiva(frequênciaabsolutaerelativa,médiaedesviopadrão)eregressãologísticabinária, expressaemOddsRatio(OR)eintervalodeconfianc¸ade95%(IC95%),foisignificativop<0,05e usou-seosoftwareSPSS17.0.
Resultados: Aprevalênciadeobesidadeabdominalfoide10,6%paramostratotal(10,5% mas-culino;10,8%feminino).Adolescentesqueassistiamàtelevisãodiariamenteporduasoumais horas(OR=2,11;IC95%1,08---4,13)apresentarammaioreschancesdeobesidadeabdominaleos adolescentescujasmãestinhamescolaridadeinferioraoitoanos(OR=0,56;IC95%0,35---0,91) tiverammenorchancedeobesidadeabdominal.
Conclusões: Aproximadamenteum acada10adolescentesapresentouobesidadeabdominal, osfatoresassociadosforamaescolaridadematerna(≥8anos)eotempodeteladetelevisão (≥2horas/dia).
©2016SociedadedePediatriadeS˜aoPaulo.PublicadoporElsevierEditoraLtda.Este ´eum artigoOpenAccesssobumalicenc¸aCCBY(http://creativecommons.org/licenses/by/4.0/).
Introduction
Abdominalfataccumulation inadolescents isan indepen-dentrisk factorfor chronicdisease,suchashypertension, fattyliver,insulinresistance,andtypeIIdiabetes,1,2aswell
association with metabolic syndrome in adolescence and
adulthood.2Waistcircumference(WC)isonewaytoassess
abdominal obesity (AO), described as an anthropometric
indicatorofeasyapplicabilityandaccuracy.3
TheliteraturereportsdifferentprevalenceofAO,which
shows differencesand/or cultural and social similarities.4
Parketal.1founddifferenceinprevalenceofAOwhen
com-paringadolescentsaged 12---19 yearsin the UnitedStates
andSouth Korea(34.7% and 8.4%, respectively).Schröder
etal.5describedAOprevalenceof11.6%wheninvestigating
Spanishadolescentsaged12---17years.Thesedifferencesin
AOprevalence are alsoseen in Brazil. Astudy conducted
withadolescentsfromMaranhão(Northeastregion)showed
aprevalenceof22.7%.6Silvaetal.7inastudywith1065
ado-lescents(aged14---17years)found2.1%ofAOprevalencein
theSoutheastregion(MinasGerais)and 6.3%inthe South
region(Santa Catarina). Also in the Southregion, studies
performedinCuritiba8andSaudades9foundAOprevalence
of12.2%and13.3%inadolescents,respectively.
Evidenceof AOassociation withsociodemographic
fac-torsandlifestylearestillunclear.Althoughitis seenthat
femaleadolescentshavehigherpercentagesofbodyfat,10
there is a tendency in the literature to describe higher
prevalence of AO in males,5,7,8 but there is no consensus
onthe relationship between AO and sex in adolescents.4
Therearealso discrepancies inthe findings regardingthe
economiclevel,withstudiesshowingahigherprevalenceof
AOincountrieswithhighereconomic levels,4 atthesame
timethatinvestigationsinregionswithlowereconomic
lev-elsalsoshowedahighprevalenceofAO.7,8Researchesthat
found excessiveconsumption of softdrinks inadolescents
foundnoassociationwithAO,11,12evenknowingthat
inade-quatedietsandhighsugarintakeareassociatedwithhigher
prevalenceofAO.
Takingintoaccount thatAO entails risks tothehealth
of adolescents and implications throughout life and that
thepossiblecombinationsofAOwithsociodemographic
fac-tors and lifestyleare notyet clear, the prevalence of AO
in adolescents and possible associated factors should be
investigated. The aim of this study was to estimate AO
prevalence anditsassociationwithsociodemographic
fac-torsandlifestyleamongadolescentsinacityintheSouth
regionofBrazil.
Method
The population of this epidemiological, cross-sectional study was composed of adolescents, aged 14---19 years, enrolled in high school in São José, Santa Catarina, Brazil.ThisstudywasapprovedbytheInstitutionalReview Board of the Federal University of Santa Catarina (CAAE: 33210414.3.0000.0121).
Abdominalobesityinadolescents 345
students in11 eligibleschools and170 classesdistributed
inhigh schoolgrades(74.8% ofthestudents attendedday
classes) was considered. We adopted a confidence level
of 1.96 (95% confidence interval), tolerance error of 5%,
50%prevalence,anddesigneffectof1.5.13 We added20%
to minimize possible losses and refusals and another20%
tocontrolforpossibleconfoundingvariablesinassociation
studies.14 Withtheseparameters,therequiredsamplesize
was751 students. Adolescents of both sexes, aged14---19
years,attendingpublichighschoolsoftheCityofSãoJosé,
SC,Brazil,wereeligibleforthestudy.Pregnantadolescents,
thosewhohadchildreninthepastsixmonths,adolescents
whodidnotgiveinformedconsentsignedbyparents (age
<18 years) or by themselves (age ≥18 years), refused to
participate, and thosewith physical disabilities that
pre-vent the performance of the physical macroproject tests
were not evaluated. After class conglomerate process in
which all students inselected classes were collected and
met the eligibility criteria, the total sample consisted of
1132students.
Anthropometricmeasurementsandunderstandingofthe
questionnaire werepretested, and a pilotstudy was
con-ductedinJuly2014inPauloLopes,SC,Brazil,with84high
schoolstudentswhoagreedtoparticipate.Datacollection
inSanJose,SC,Brazil,wasmadefromAugusttoNovember
2014. To this end, seven post-graduate and four
gradu-atestudentsinphysicaleducationwereselected,threeof
themhadlevel1certificationfromtheInternational
Soci-etyfortheAdvancementofKinanthropometryandmadethe
anthropometricmeasurement.
WCwasmeasuredatthenarrowestportionofthetrunk,
betweenthe lowercostalmargin andtheiliac crest,with
anthropometric measuring tape.3 To classify adolescents
withAO,thecutoffpointsproposedbyTayloretal.3 were
used,whichdefinedasexcessfatvalueswithZscore≥1.
These cutoffs points are near the 85th percentile,
con-sideredcut-off point foroverweight bybodymassindex.3
Cutoffspointshavebeenproposedaccordingtoageandsex
(sensitivityandspecificityvaluesof84%and94%forfemale
and87%and92%formale,respectively)3(Table1).
Aself-administeredquestionnairewasappliedwith
ques-tions related to sociodemographic variables (gender, skin
color,age,maternaleducation,economiclevel,andschool
shift)andlifestyle(physicalactivity,alcoholconsumption,
soft drink consumption, sleep, and sedentary behavior)
(Table1).
Categories of skin color were collected in accordance
withtherecommendationsoftheBrazilianInstituteof
Geog-raphy and Statistics15 and categorized in brown, black,
yelloworredasacategory,andwhiteinanothercategory.15
Maternaleducationwascategorizedtakingintoaccountthe
averageschoolingofBrazilians(7.2years),16classifiedaslow
(<8years)andhigh(≥8years).Economiclevelwasevaluated
byhouseholdpurchasingpower.Forthisstudythecategories
AandBweredefinedashighandtheremainingaslow.17
To evaluatetheconsumption ofalcohol andsoftdrinks
and the level of physical activity (PA), questions of the
YouthRiskBehaviorSurvey(YRBS)wereused,translatedand
validatedfor Brazil.18 PAcategorizationtookintoaccount
theevidenceshowingthat60minofPAfivedaysperweek
is sufficientto maintainhealth in adolescenceand higher
amounts would provide additional benefits.19 For alcohol
consumption,subjectswerecategorizedinto‘‘no’’ (those
who did not consume in the last 30 days five or more
alcoholicdrinksinoneoccasion)and‘‘yes’’(thosewho
con-sumedsuchdosageinoneormoredaysinthelast30days).
Forsoftdrinkconsumption, adolescentswereclassified as
‘‘no’’(thosewhodidnotconsumeanytimeduringtheweek)
and‘‘yes’’(thosewhoconsumedoneormoretimesduring
theweek).
Sleep was evaluated for quality using the
Fantas-tic Lifestyle questionnaire, translated and validated for
Brazil.20 Individuals who said they slept well ‘‘almost
always’’and‘‘fairlyoften’’wereconsideredasgood
qual-itysleep(yes),whilethosewhoanswered‘‘almostnever’’,
‘‘rarely’’,and‘‘sometimes’’wereconsideredaspoor
qual-itysleep(no).
Toassesssedentarybehavior(SB)onweekdaysand
week-ends,the2-hcutoffpointwasusedforeachbehavior,due
toevidenceofharmtothehealthofyoungpeoplewhohave
SBabovethatvalue.19Whenbehaviorsareadded,the
mul-tiplicationofthe2-hcutoffpointby2isused(totaling4h)
fortotaldisplaytime.21,22
Traveltoschoolwasclassified asactivefor those with
energyexpenditureinthistravelingorpassiveforthosewho
traveledbycarwithoutenergyexpenditure.22
Descriptivestatisticswasusedwithanalysisofabsolute
andrelativefrequency,mean andstandard deviation. For
comparisonofmedians,theMannWhitneyUtestwasusedto
comparefrequencies,thechi-squaretestofheterogeneity.
Toidentifyfactorsassociatedwiththedependentvariable,
binarylogisticregressionwasusedtoestimateOddsRatio
and95% confidenceintervals. Forallstatistical tests,
sig-nificancelevelwassetatp<0.05.Inregressionanalysiswe
usedthe hierarchical model of determination,
hypotheti-callytemporal,fromdistaltoproximaldeterminants,23 in
which the demographic variables (gender, skin color, and
age)wereincludedinthedistalblock,socioeconomic
varia-bles(economiclevel,maternaleducation,andschoolshift)
intheintermediateblock,andlifestylevariables(PA,
alco-holconsumption,softdrinkconsumption,sleep,andSB)in
thedistalblock.Intheregressionmodel,allcovariateswere
treateddichotomously,asdescribedinTable1.The
selec-tionfor input variables in the adjusted model was made
using the backward method. All variables were included
in the adjusted analysis, regardless of the crude analysis
p-value.Adjustmentsweremadeforvariablesatthesame
levelandabovelevelsthatrepresentedp≤0.20intheWald
test(adjusted analysis)andremainedinthemodel.24 The
SPSS17.0softwarewasusedforallanalyzes.
Results
There was a sample loss of 17.8% of adolescents, which includedthosewhoansweredthequestionnaire,butdidnot participateinthemacroprojectanthropometricassessment. Thus,thepresent study sampleincluded 930 adolescents, mean age of 16.1±1.1 years, with female prevalence (n=490) (Table 2). Most adolescents had white skin color
(62.7%), 14---16 years of age (58%), low maternal
educa-tion(56.4%),higheconomiclevel(68.2%),andattendedday
school(69.8%)(Table 2).Nine outof 10adolescentswere
Castro
JA
et
al.
Type Variables Instrument/Question Categorization
Dependent Waistcircumference Anthropometrictape Cutoffpoints
Female: Male:
14years≥77 14years≥79
15years≥78.3 15years≥81.1
16years≥79.1 16years≥83.1
17years≥79.8 17years≥84.9
18years≥80.1 18years≥86.7
19years≥80.1 19years≥88.4
<Cutoffpoint=Normal
Cutoffpoints=Abdominalobesity3
Independents Sex Whatisyoursex? Female/Male
Skincolor TheBraziliancensususesthewordswhite,brown,black,yellowand indigenoustoclassifythecolororraceofpeople.Ifyouhadtoanswerthis question,howwouldyourateyourselfaboutyourcolororethnicity?
White
Black/Brown/Yellow/Indigenous15
Age Howoldareyou? 14---16years
17---19years[30] Maternaleducation Whatisthelevelofeducationcompletedbyyourmother? <8years
≥8years16 Economiclevel Whatisthemonthlyincomeof
yourfamily(thecurrentminimum wageisR$724)?17
A/B=High
C/D/E=Low9
Schoolshift Whatisyourschoolshift? Morning/Afternoon/Fullday=Day
Evening=Evening Physicalactivity Duringatypicalweek,onhowmanydaysyoupracticemoderatetovigorous
physicalactivity(physicalactivityduringleisuretime,atwork,andgoingto school)?18
<5days=Insufficientlyactive
≥5days=Active19 Alcoholconsumption Duringthelast30days,onhowmanydaysdidyoudrinkfiveormore
alcoholicbeveragesinasingleoccasion?18
Noday=No ≥1day=Yes29 Softdrinkconsumption Duringthelast7days,howmanytimesdidyoudrinkabottle,canorglass
ofsodasuchasCoca-Cola,Fanta,Sprite,PepsiandPureza?(Donotconsider dietandlightsoftdrinks)18
Ididnotdrinksoftdrinksinthelast7days=No
Otheroptions=Yes12
Sleep Doyousleepwellandfeelrested?20 Almostnever/Rarely/Sometimes=No
Withrelativefrequency/Almostalways=Yes20 TVTime HowmanyhoursperdaydoyouwatchTVonschooldays(MondaytoFriday)? <2h/≥2h19
HowmanyhoursperdaydoyouwatchTVonweekends(Saturdayand Sunday)?
Computertime Howmanyhoursperdaydoyouusethecomputeronschooldays(Mondayto Friday)?
<2h/≥2h19
Howmanyhoursperdaydoyouusecomputeronweekends(Saturdayand Sunday)?
Videogametime Howmanyhoursperdayyouplayvideogamesonschooldays(Mondayto Friday)?
<2h/≥2h19
Howmanyhoursperdayyouplayvideogamesonweekends(Saturdayand Sunday)?
Totalscreentime Total <4hours/≥4hours21,22
Traveltoschool Howdoyouusuallytraveltoschool? Walking/biking=Active
Abdominalobesityinadolescents 347
Table2 Sampledistribution,waistcircumference(meanandstandarddeviation)andabdominalobesityprevalenceaccording tosociodemographicvariables.
Variables Sample Waistcircumference p-value Abdominalobesity p-value
n(%) M±SD n(%)
Total 930 71.5±8.0 99(10.6)
Sex
Male 440(47.3) 73.8±7.7 46(10.5)
Female 490(52.7) 69.4±7.6 <0.01a 53(10.8) 0.85
Skincolor
White 574(62.7) 71.0±7.7 56(9.8)
Black/brown/yellow/red 342(37.3) 72.0±8.3 0.09 41(12.0) 0.28
Age(years)
14---16 539(58.0) 70.6±7.7 54(10.0)
17---19 391(42.0) 72.7±8.2 <0.01a 45(11.5) 0.46 Maternaleducation
<8years 518(56.4) 71.3±8.0 49(9.5)
≥8years 400(43.6) 71.7±8.0 0.38 50(12.5) 0.14
Economiclevel
High 537(68.2) 71.5±7.7 51(9.5)
Low 250(31.8) 71.2±8.6 0.32 31(12.4) 0.21
Schoolshift
Day 644(69.8) 70.8±7.9 66(10.2)
Evening 278(30.2) 72.9±8.0 <0.01a 33(11.9) 0.46
M,mean;SD,standarddeviation.
a p≤0.05;Mann---WhitneyUtestforindependentsamples.
excessive alcohol (33.9%), eight out of 10 consumed soft drinks (84.1%), and about two thirds did not sleep well (61.4%).RegardingSB,sevenoutof10adolescentswatched twoor morehours oftelevision (78.6%),about two-thirds ofthestudents spenttwoor morehoursonthecomputer (68.7%),andthreeoutof 10playedvideogamesfor more than twohours (28%). Approximately nine out of 10 ado-lescents had screen time above 4h per day (87.2%) and half of thestudents usedpassive transportation toget to school(50.1%)(Table3).Adolescentswhoweremale,older,
attendedeveningclasses, drank alcohol and playedvideo
gamesdailyfor2hormorehadhigherWC(p<0.05).There
wasnodifferenceinAOprevalence betweencategoriesof
theindependentvariables(Tables2and3).
Adolescentswhosemothershadloweducationwereless
likely to have AO in the crude analysis (OR=0.59, 95%CI
0.35---0.98). In theadjusted analysis,the association with
maternaleducationremained(OR=0.56,95%CI0.35---0.91),
andstudentswhowatchedtelevisionfortwohoursormore
were more likelyto have AO (OR=2.11; 95%CI 1.08---4.13)
(Table4).
Discussion
The present studymain findingswere:(i)about onein 10 adolescentshadAO(10.6%);(ii)adolescentswhowatched televisionfor twoor morehoursweremorelikelytohave AO(OR=2.11,95%CI1.08---4.13),and(iii)adolescentswhose mothers had less thaneight yearsof education wereless likely to have AO (OR=0.56; 95%CI 0.35---0.91). Previous studiesinthesameareadescribedAOprevalenceof6.3%,7
13.3%,9and12.8%.8Thisstudybringsasadvancethe
infor-mation that sedentary behavior was the main lifestyle
factor associated with AO. Such behavior is modifiable
throughhealth education initiatives, which may result in
lowerprevalence of AO. In addition tothis variable, the
onlysociodemographicfactorassociatedwithAOwashigh
maternaleducation,whichshowstheneedforspecific
inves-tigationsforthisindicator.
The10.6%prevalenceofAOfoundinthisstudyisbelow
the values found for adolescents in Spain (11.6%)5 and
USA(34.7%),1andsimilartothosefoundfor SouthKorean
adolescents(8.4%).1When comparedtostudiesperformed
in Brazil, the found prevalence of AO was higher than
the values reported for Minas Gerais (Januária) (2.1%)7
andlowerthan thevaluesreportedfor Maranhão(22.7%)6
and Curitiba (12.2%).8 These differences in OA
preva-lencemaydemonstrateculturalandsocialdifferencesand
similarities,4 if we take into account that São José and
Curitibahavethehumandevelopmentindexrankedasvery
high (0.809 and 0.823)16 and Januária ranked as medium
(0.658).16
Anotherfactorthatmayexplainthediscrepancybetween
theprevalencefoundistheuseofdifferentcutoffpointsby
differentstudies.Similartothepresentinvestigation,three
studies5---7usedthecutoffpointsproposedbyTayloretal.3,
which definedAO values with Z score ≥1. The validation
ofWCmeasurementwasperformedusingdualenergyX-ray
absorptiometry(DXA)andshowedhighcorrelationforboth
genders (r=0.92), with 84% sensitivity and 94% specificity
forgirlsand 87%sensitivityand92%fspecificityfor boys.3
Table3 Sampledistribution,waistcircumference(meanandstandarddeviation)andabdominalobesityprevalenceaccording tobehavioralvariables.
Variables Sample Waistcircumference p-value Abdominalobesity p-value
n(%) M±SD n(%)
Total 930 71.5±8.0 99(10.6)
Physicalactivity
Insufficientlyactive 832(92.1) 71.5±8.1 94(11.3)
Active 71(7.9) 72.0±6.7 0.30 5(7.0) 0.27
Alcoholconsumption
Não 610(66.1) 70.9±7.9 61(10.0)
Sim 313(33.9) 72.4±8.1 0.01a 38(12.1) 0.32
Softdrinkconsumption
Não 143(15.9) 72.2±8.7 21(14.7)
Sim 754(84.1) 71.3±7.9 0.45 78(10.3) 0.12
Sleep(sleepwell)
Não 559(61.4) 71.3±8.0 58(10.4)
Sim 351(38.6) 71.7±7.9 0.62 39(11.1) 0.72
TVtime
<2h 192(21.4) 71.6±7.3 15(7.8)
≥2h 705(78.6) 71.4±8.3 0.23 84(11.9) 0.08
Computertime
<2h 281(31.3) 71.6±8.7 36(12.8)
≥2h 616(68.7) 71.4±7.8 0.83 63(10.2) 0.29
Videogametime
<2h 646(72.0) 71±7.9 72(11.1)
≥2h 251(28.0) 72.6±8.3 0.01a 27(10.8) 0.85 Totalscreentime
<4h 115(12.8) 70.7±7.9 13(10.3)
≥4h 782(87.2) 71.6±8.1 0.41 86(11.0) 0.86
Traveltoschool
Active 448(49.9) 71.2±7.8 43(9.6)
Passive 449(50.1) 71.7±8.3 0.58 56(12.5) 0.17
M,mean;SD,standarddeviation.
ap≤0.05;Mann---WhitneyUtestforindependentsamples.
WCpercentiles.AOdiagnosisdependsonthecutoffpoints usedwithhighsensitivityandspecificityvaluesforthe popu-lationstudied.Theuseofreferencesthatdonotadoptthese criteriamayleadtomisclassificationandunderestimateor overestimatethevalues.4
Adolescentswhosemothershadlessthaneightyearsof
educationwere lesslikelytohave AO.Parental education
determinesthechildren’s chanceof educationand family
culturalsphere.25 Duetothepositiverelationshipwiththe
familyincome,thehighertheeducationallevel,thehigher
the householdincome.25 Thus, this finding is in line with
studiesreportingassociationbetweenAOandhigheconomic
level, which is related to obesogenic environments.4,6,7,11
In our study, when the economic level of adolescents
was directly investigated, there was no association with
the prevalence of AO. This fact shows theimportance of
assessing factors associated with AO in different ways to
determinethepresenceobesogenicenvironments.
Televisiontimeequaltoorgreaterthan2hwasassociated
withAO.Theresultofthisstudyissimilartothatfoundby
Byunetal.26whoreportedtheassociationofAOandSB.This
association may beexplainedby thelowerenergy
expen-diturethroughoutthedayinadolescentswhohavegreater
involvementinSB.27Moreover,theliteraturehasshownthat
high-calorie food consumption occurs concomitantly with
theactofwatchingtelevision.27
WCwashigherinmalecomparedtofemaleadolescents,
butthesedifferenceswerenotsustainedwhenassessingthe
prevalenceofAO.Moraesetal.4foundthat,althoughthere
is atendencytohigherWCvaluesinmen,theassociation
betweenAOandsexisnotyetclearinadolescents.Higher
valuesofWCinmaleadolescentsoccurasaresultofsexual
dimorphisminfatdistribution.10 Femaleadolescents,even
withahigherpercentageofbodyfatduetohormonal
dif-ferencesbetweenthesexes, haveincreasedaccumulation
ofadiposetissueinthehipanddecreasedinthewaistarea
comparedwiththeirmalepeers.10
Adolescents who played video games for 2h or more
hadhigher WC,which canbeexplainedby thehigherWC
in maleadolescent,astheyhave video gamescreen time
higherthanfemale.26,28Inourstudy,boyshadhigher
Abdominalobesityinadolescents 349
Table4 OddsRatioand95%confidenceinterval,incrudeandadjustedbinarylogisticregressionanalysis,betweenabdominal obesityandindependentvariables.
Crudeanalysis Adjustedanalysisa
OR(CI) p-value OR(CI) p-value
Sexb 0.70 0.71
Male 1 1
Female 0.90(0.53---1.53) 1.08(0.70---1.65)
Skincolorb 0.10 0.28
White 1 1
Black/brown/yellow/red 1.50(0.92---2.45) 1.26(0.82---1.93)
Age(years)b 0.44 0.49
14---16 0.82(0.50---1.35) 0.86(0.56---1.31)
17---19 1 1
Maternaleducationc 0.04d 0.02d
<8years 0.59(0.35---0.98) 0.56(0.35---0.91)
≥8years 1 1
Economiclevelc 0.06 0.07
High 1 1
Low 1.65(0.97---2.78) 0.64(0.39---1.06)
Shiftc 0.45 0.49
Day 1 1
Evening 1.22(0.72---2.07) 1.18(0.72---1.93)
Physicalactivitye 0.66 0.71
Insufficientlyactive 1.24(0.46---3.35) 1.19(0.45---3.15)
Active 1 1
Alcoholconsumptione 0.56 0.37
No 1 1
Yes 1.16(0.69---1.93) 1.24(0.76---2.03)
Softdrinkconsumptione 0.08 0.11
No 1 1
Yes 0.57(0.30---1.07) 0.62(0.34---1.12)
Sleep(sleepwell)e 0.34 0.46
No 0.78(0.47---1.29) 0.83(0.51---1.35)
Yes 1 1
TVtimee 0.10 0.02d
<2h 1 1
≥2h 1.82(0.88---3.73) 2.11(1.08---4.13)
Computertimee 0.09 0.09
<2h 1 1
≥2h 0.61(0.35---1.08) 0.66(0.40---1.07)
Videogametimee 0.57 0.78
<2h 1 1
≥2h 0.84(0.46---1.52) 0.92(0.53---1.60)
Totalscreentimee 0.44 0.54
<4h 1 1
≥4h 1.40(0.59---3.36) 1.29(0.55---3.01)
Traveltoschoole 0.06 0.09
Active 1 1
Passive 1.60(0.97---2.65) 1.51(0.93---2.44)
OR,oddsratio;CI,confidenceinterval.
a Adjustedanalysisforallindependentvariables.
b Distalleveltooutcomevariables.
c Intermediatelevelvariables.
d p≤0.05.
(53.9±157.7min).Moreover,thepercentageofadolescents who played video games for 2h or more was higher in males (45.5%) than females (15.5%) (Data not shown in table/figures).
OlderadolescentshadhigherWCvalues.Theliterature states that, due to the morphological and physiological development,theWCincreaseswithageandstageof sex-ualmaturation,regardlessofthepresenceofAO.3,10Another
findingofourstudywasthatadolescentswithhigheralcohol
consumptionalso hadhigherWC. Epidemiologicalanalysis
indicatedthatalcoholconsumptionincreaseswithage.29As
olderstudentshadhigherWCvalues,agecanbea
confound-ingvariableintheassociationbetweenalcoholconsumption
andWC.
Regardingschool shift,higherWCwasfoundin
adoles-centsattending eveningclasses. It is speculated thatthis
findingisrelatedtothefactthattheseadolescentsareolder,
haveloweconomiclevelsand,therefore,theywork.Thus,
theyhavelifestylehabitsmoresimilartoadultsandgreater
engagementinSB.22
Thestudywithadolescentsattendingpublichighschool
inSãoJoséwasalimitation,asitimplies thattheresults
maynot be extrapolated toprivate school students who,
inBrazil,havedifferentsocioeconomiccharacteristicsfrom
thoseobserved in young people from publicschools. The
strengthsofthisstudyweretheschool-basedsampleandthe
useofcutoffpointsvalidatedthroughamethodwithstrong
correlationwithbenchmarksinabdominalfatevaluation.3
Additionally,severalBrazilianstudiesusedthesamecutoff
points,6,7whichfacilitatescomparisonbetweenstudies.
Itcanbeconcludedthatahighprevalenceofabdominal
obesitywasfound in approximatelyoneout of10
adoles-cents.Maternaleducation(≥8 years)andthetimesitting
infrontoftelevision(≥2h)wereassociatedwithabdominal
obesity.
Funding
Thisstudydidnotreceivefunding.
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
References
1.Park J,HilmersDC,MendozaJA,StuffJE, Liu Y,NicklasTA. Prevalenceofmetabolicsyndromeandobesityinadolescents aged12to19years:comparisonbetweentheUnitedStatesand Korea.JKorMedSci.2010;25:75---82.
2.HeF,Rodriguez-ColonS,Fernandez-MendozaJ,VgontzasAN, BixlerEO,BergA,etal.Abdominalobesityandmetabolic syn-drome burden in adolescents --- Penn State Children Cohort study.JClinDensitom.2015;18:30---6.
3.Taylor RW, JonesIE, Williams SM,Goulding A. Evaluationof waistcircumference,waist-to-hipratio,andtheconicityindex asscreeningtoolsforhightrunkfatmass,asmeasuredby dual-energyX-rayabsorptiometry,inchildrenaged3---19y.AmJClin Nutr.2000;72:490---5.
4.DeMoraesAC,FadoniRP,RicardiLM,SouzaTC,Rosaneli CF, NakashimaAT,etal.Prevalenceofabdominalobesityin adoles-cents:asystematicreview.ObesRev.2011;12:69---77.
5.SchroderH,RibasL,KoebnickC,FuntikovaA,GomezSF,FitoM, etal.PrevalenceofabdominalobesityinSpanishchildrenand adolescents.Doweneedwaistcircumferencemeasurementsin pediatricpractice?PLoSOne.2014;9:e87549.
6.Nascimento-FerreiraMV,DeMoraesAC,CarvalhoHB,Moreno LA,GomesCarneiroAL,DosReisVM,etal.Prevalenceof car-diovascular risk factors,the association with socioeconomic variablesinadolescentsfrom low-income region.NutrHosp. 2014;31:217---24.
7.Silva DA, Pelegrini A, Silva AF, Grigollo LR, Petroski EL. Abdominalobesityandassociatedfactorsamongadolescents: comparisonoftwoeconomicallydifferentBrazilianregions.Arq BrasEndocrinolMetabol.2012;56:291---9.
8.BozzaR,deCamposW,BacilED,BarbosaFilhoVC,HardtJM, da Silva PM. Sociodemographic and behavioral factors asso-ciatedwithbody adiposity in adolescents. RevPaul Pediatr. 2014;32:241---6.
9.GlanerMF,PelegriniA,CordobaCO,PozzobonME.Associac¸ão entre insatisfac¸ão com a imagem corporal e indicadores antropométricosemadolescentes.RevBrasEducFisEsporte. 2013;27:129---36.
10.TaylorRW,GrantAM,WilliamsSM,GouldingA.Sexdifferencesin regionalbodyfatdistributionfrompre-topostpuberty.Obesity (SilverSpring).2010;18:1410---6.
11.Moraes AC, Falcao MC. Lifestyle factors and socioeconomic variablesassociatedwithabdominalobesity inBrazilian ado-lescents.AnnHumBiol.2013;40:1---8.
12.TheAVENAstudyGómez-MartínezS,Martínez-GómezD,Perez deHeredia F,RomeoJ, Cuenca-GarciaM, Martin-MatillasM, et al. Eating habits and totaland abdominalfat in Spanish adolescents:influenceofphysicalactivity.JAadolescHealth. 2012;50:403---9.
13.LuizRR,MagnaniniMM.Alógicadadeterminac¸ãodotamanho daamostraeminvestigac¸õesepidemiológicas.CadSaudeColet. 2000;8:9---28.
14.Kuhnen M, Boing AF, Oliveira MCd, Longo GZ, Njaine K. Tabagismo e fatores associados em adultos: um estudo de base populacional. Rev Bras Epidemiol. 2009;12: 615---26.
15.World Health Organization [homepage on the Internet].
Handbook on health inequality monitoring with a special
focus on low-and middle-income countries. Available from:
http://www.who.int/gho/healthequity/handbook/en/ [cited 28.07.15].
16.United Nations [homepage on the Internet]. Human devel-opment report 2014. Sustaining human progress: reducing vulnerabilities and building resilience 2014. Available from: http://www.uy.undp.org/content/dam/undp/library/ corporate/HDR/2014HDR/HDR-2014-English.pdf [cited 28.07.15].
17.Associac¸ão das Empresas de Pesquisa [homepage on the Internet]. Critério de Classificac¸ão Econômica Brasil 2014. Availablefrom:http://www.abep.org/codigos-e-guias-da-abep [cited28.07.15].
18.GuedesDP,LopesCC.Validac¸ãodaversãobrasileiradoyouth riskbehaviorsurvey.RevSaudePublica.2010;44:840---50. 19.TremblayMS,LeblancAG,JanssenI,KhoME,HicksA,Murumets
K,etal.Canadiansedentarybehaviourguidelinesforchildren andyouth.ApplPhysiolNutrMetab.2011;36:5---71.
20.RodriguezAnezCR, ReisRS, PetroskiEL.Brazilianversion of alifestylequestionnaire:translationandvalidationforyoung adults.ArqBrasCardiol.2008;91:92---8.
21.De Moraes AC, Fernandes CAM, Elias RGM, Makashima ATA, Reichert FF, Falcão MC. Prevalência de inatividade física e fatores associados em adolescentes. Rev Assoc Med Bras. 2009;55:523---8.
Abdominalobesityinadolescents 351
physicaleducationamonghighschoolstudentsinSantaCatarina State,Brazil.CadSaudePublica.2009;25:2187---200.
23.VictoraCG,Huttly SR,FuchsSC, OlintoMT.Theroleof con-ceptualframeworksinepidemiologicalanalysis:ahierarchical approach.IntJEpidemiol.1997;26:224---7.
24.Maldonado G, Greenland S. Simulationstudy of confounder-selectionstrategies.AmJEpidemiol.1993;138:923---36. 25.KappelDB.ÍndicededesenvolvimentoinfantilnoBrasil:uma
análiseregional.RevBrasEduc.2007;12:232---40.
26.ByunW,DowdaM,PateRR.Associationsbetweenscreen-based sedentarybehavior andcardiovascular diseaseriskfactors in Koreanyouth.JKorMedSci.2012;27:388---94.
27.Chaput JP, Klingenberg L, Astrup A, Sjodin AM. Modern sedentary activities promote overconsumption of food in our current obesogenic environment. Obes Rev. 2011;12: e12---20.
28.Vasconcellos MB,Anjos LA, VasconcellosMT. Nutritional sta-tusandscreentimeamongpublicschoolstudentsinNiteroi, Rio de Janeiro State, Brazil. Cad Saude Publica. 2013;29: 713---22.
