Cardiovascular risk profile of high school students: A cross-sectional study

(1)

Revista

Portuguesa

de

Cardiologia

Portuguese

Journal

of

Cardiology

www.revportcardiol.org

ORIGINAL

ARTICLE

Cardiovascular

risk

proﬁle

of

high

school

students:

A

cross-sectional

study

Teresa

Rocha

a,∗

_,

_Evangelista

_Rocha

a

_,

_Ana

_Catarina

_Alves

b

_,

Ana

Margarida

Medeiros

b

_,

_Vânia

_Francisco

b

_,

_Sónia

_Silva

b

_,

Isabel

Mendes

Gaspar

c

_,

_Quitéria

_Rato

d

_,

_Mafalda

_Bourbon

b

a_Unidade_de_{Epidemiologia,}_Instituto_de_Medicina_Preventiva,_Faculdade_de_Medicina_de_Lisboa,_Lisboa,_Portugal

b_Grupo_de_Investigac_¸ão_{Cardiovascular,}_UID,_Departamento_de_Promoc_¸ão_da_Saúde_e_Prevenc_¸ão_de_Doenc_¸as_Não_{Transmissíveis,} InstitutoNacionaldeSaúdeDr.RicardoJorge,Lisboa,Portugal

c_Servic_¸o_de_Genética_Médica,_Hospital_de_Santa_Maria,_Centro_Hospitalar_de_Lisboa_Ocidental,_EPE,_Lisboa,_Portugal d_Servic_¸o_de_Cardiologia,_Hospital_S._Bernardo,_Centro_Hospitalar_de_Setúbal,_EPE,_Setúbal,_Portugal

Received8December2013;accepted2January2014 Availableonline12September2014

KEYWORDS Physicalactivity; Smoking; Obesity; Cardiovascular; Risk; Adolescents Abstract

Introduction:Diseaseprevention shouldbegininchildhoodandlifestyles areimportantrisk determinantsofcardiovasculardisease.Awarenessandmonitoringofriskisessentialin pre-ventivestrategies.

Aim:Tocharacterize cardiovascular riskand therelationshipsbetween certain variables in adolescents.

Methods:Inacross-sectionalstudy,854adolescentschoolchildrenweresurveyed,meanage 16.3±0.9years.Datacollectionincludedquestionnaires,physicalexamination,chartsfor 10-yearrelativeriskofmortality,andbiochemicalassays.In thestatisticalanalysiscontinuous variableswerestudiedbytheStudent’sttestandcategoricalvariablesbythechi-squaretest andFisher’sexacttest,andeachriskfactor wasenteredasadependentvariableinlogistic regressionanalysis.

Results:Physicalactivitywasinsufﬁcientin81%ofstudents.Thedailyconsumptionofsoup, saladorvegetables,andfruitwas,respectively,37%,39%and21%.Aminority(6%)took≤3 and77%took≥5mealsaday.Theprevalenceofeachriskfactorwasasfollows:overweight 16%;smoking13%;hypertension11%;impairedglucose metabolism9%;hypertriglyceridemia 9%;andhypercholesterolemia5%.Out-of-schoolphysicalactivity,hypertensionandoverweight weremoreprevalentinmales(p<0.001).Femaleshadhigherlevelsofcholesterol(p<0.005) andtriglycerides(p<0.001).Aquarteroftheadolescentshadarelativeriskscorefor10-year cardiovascularmortalityof≥2.Overweightshowedapositiveassociationwithbloodpressure, changesinglucosemetabolismandtriglycerides,andanegativeassociationwithnumberof dailymeals.

∗_{Corresponding}_author.

E-mailaddresses:teresa.rocha1778@gmail.com,teresa.rocha@episystem.eu(T.Rocha).

http://dx.doi.org/10.1016/j.repc.2014.01.024

(2)

Conclusions: Theresultsdemonstratetheneedforactioninprovidingandencouraginghealthy choicesforadolescents,withanemphasisonbehavioralandlifestylechangesaimedat indi-viduals,familiesandcommunities.

Perﬁlderiscocardiovasculardeadolescentesescolarizados:umestudotransversal Resumo

Introdução: Aprevençãodevecomeçarnainfânciaeoestilodevidaéumdeterminante impor-tantedadoençacardiovascular.Oconhecimentoemonitorizaçãodoriscocardiovascularsão essenciaisnumaestratégiapreventiva.

Objetivo: Caracterizaroriscocardiovasculareasrelac¸õesentrealgumasvariáveisem adoles-centesescolarizados.

Métodos: Foramobservados854estudantes,16,3±0,9anosdeidade,numestudotransversal. Acolheitadedadosincluiuquestionários,examefísico,tabeladoriscorelativodamortalidade cardiovascularadezanos,doseamentosbioquímicos.Aanáliseestatísticaincidesobrevariáveis contínuasecategoriaisefatoresderiscocomovariáveisdependentes(regressãologística).

Resultados: Atividadefísicainsuficiente81%.Consumodiáriodesopa,salada/legumesefrutas caracterizou37,39e21%dosalunos.Umaminoria(6%)faziadiariamente≤3refeiçõese77% ≥5refeições.Prevalênciadosfatoresderisco:excessodepeso16%;tabagismo13%;hipertensão arterial 11%;anomalias do metabolismo daglicose 9%; hipertrigliceridemia 9%; hipercoles-terolemia5%.Atividadefísicaextracurricular,hipertensãoeexcessodepesomaisprevalentes nosexomasculino(p<0,001),enquantonasadolescentesforamahipercolesterolemia(p<0,005) ehipertrigliceridemia(p<0,001).Umquartodosadolescentescomriscorelativodemortalidade cardiovascularadezanos≥2.Excessodepesorevelourelaçãopositivacompressãoarterial, alteraçõesmetabolismodaglucoseetriglicerídeosenegativacomnúmeroderefeiçõesdiárias.

Conclusões: Os resultados evidenciam a necessidade de ac¸ões de fomento e incentivo de escolhassaudáveispelosadolescentes,comênfasenasmudanc¸asdecomportamentoaonível dosindivíduos,famíliasecomunidades.

Introduction

Despitetheadvancesofthelast50years,strategiesfor car-diovascularpreventionhave nothadanequivalentimpact to the progress of knowledge. Although the majority of deathsfromcardiovasculardisease(CVD)arepreventableor treatable,CVDremainstheleadingcauseofdeathinmost developedcountries.

The development of atherosclerotic disease is a con-tinuum in which the vessel passes from a healthy state, through an intermediate stage that is inﬂuenced by tra-ditional, genetic and environmental risk factors, until it reachestheclinicalstage.Inrecentdecadesevidencehas accumulated on cardiovascular risk throughout life: risk exposurebeginswithinﬂuencesduringpregnancyand con-tinuesintochildhood,adolescenceandadulthood.1_Children

and young people whohave a high value of a risk factor willkeep itthroughout life, but achievinglowerlevelsof thisfactor when young willhave a greaterimpactonthe disease than if it is detected and treated in adulthood.2

Nevertheless, strategies for CVD prevention are designed mainly for patients and high-risk individuals, particularly

middle-aged and elderly adults. Furthermore, the most commonapproachhasbeenbasedontheindividualandon medication. Attempts toraise awareness and topromote changes in lifestyles have had littlesuccess, even in sec-ondary prevention,3_but _prevention_in _young_people_could

have many beneﬁts. Its guiding principles are those of a community-basedapproachtoprevention,basedonhealth educationandindividualattitudesthatarealsodependent ontheenvironment.4_Indeed,_deviant_behaviors_result_from

the dynamics between biological and social factors that favororlimitthediversityofindividualcharacteristics.At thesametime,theWorldHealthOrganization(WHO) recog-nizesthatthemostprominentnon-communicablediseases (NCDs), namely cardiovascular disease, cancers, chronic respiratorydiseasesanddiabetes,arelinkedtotobaccouse, alcoholabuse,anunhealthydiet,andphysicalinactivity.5_As

acorollary,theWHO,workinginpartnershipwithother orga-nizationswithdecision-makingpowersinpublichealth,has developedstrategicactionplansforthepreventionand con-trolofNCDs.6_Obesity_also_merits_special_attention_because

itresultsfromsomeofthesamefactorsandisimplicated in other NCDs. This does not mean that all factors are

(3)

associated with each disease to the same extent, but they are considered together in action plans and educa-tion strategies toemphasize their common causes andto highlightpotentialsynergiesinprevention.

To facilitate comparative studies the WHO and the US Centersfor DiseaseControlandPrevention havedrawnup 10model questionnaires,the GlobalSchool-basedStudent Health Survey.7 _Nevertheless _the _{single-factor} _approach

is still predominant and estimates of total cardiovascu-lar risk in adolescents are limited. The present study,

Coração Jovem (‘‘Young Heart’’), developed by the Por-tuguese National Institute of Health (INSA), is guided by the same principles as the National Network of Health-PromotingSchools(REPS),whichissupportedbytheCouncil of Europe and the European Commission via the Schools for Health in Europe network. The purposeof REPS is to establishmodelschoolstodemonstratetheimpactofhealth promotion,throughhealtheducationwiththeinvolvement oftheeducationcommunity.TheCoraçãoJovemstudy cov-ers threeareas: health behaviors (diet, physical activity, and smoking), seven cardiovascular risk factors, and the relationshipbetweenthesecharacteristics.Italsohasa ped-agogicalcomponent,itsresultsbeingpresentedanddebated ineachschoolwiththeinvolvementofthestudentsinthese actions. Although its purpose was basically diagnostic, it alsoaimedtomotivateadolescentschoolchildrentochange theirbehaviorswithregardtohealth.Specifically,themain aimsweretocharacterizephysicalactivity,dietand smok-ing;todefinecardiovascularriskprofiles;andtoassessthe relationshipbetweenvariablesandtheirimplicationsin car-diovascularpreventionandoverallhealth.

Methods

This cross-sectional study was approved by the National CommissionforDataProtectionandtheInstitutionalEthics Committee.Theminimumnumberofparticipantswas cal-culated (n=777) onthe basis of 95% conﬁdence intervals, populationsize,8_estimated_prevalence_of_overweight_(15%)

anda2.5%standarderror.Schoolswereselectedfromthe list of private andpublic secondary schools in the Lisbon regioninordertoachievea balancebetweenparticipants byschool type.Allstudents meetingtheinclusioncriteria (15---18 years old) were invited to participate. The mea-suringinstruments consistedof structuredself-completion questionnaires (on physical activity, eating and smoking habits for students; on premature cardiovascular events forparents),anOmronsphygmomanometer(bloodpressure measured twice),andscalesfor weight(inkg) andheight (incm).Biochemicalassaysonfastingcapillaryblood sam-ples measured total cholesterol, triglycerides (Accutrend GCT) and glucose (Accu-Chek Aviva). The chart for rela-tive riskof 10-year cardiovascularmortality(SCORE)9 _was

used.Theﬁeldworkwasprecededbyvalidationofthe bio-chemicalequipment(CobasIntegraanalyzer)andapre-test performed on 30 volunteers. The ﬁeldwork took place in threestages:(1)presentation andplanningof theproject in schools; (2) visits to obtain the list of participants and their informed consent and application of the ques-tionnaires to parents; and (3) physical examinations and biochemicalassays in fastingstate andapplication of the

questionnaire to students. Data collection took place in 2009.Biologicalvariableswereclassiﬁedaccordingto inter-nationalrecommendations.9---11 _Tests_of_{independence,}_the

chi-squaretest, Fisher’sexact test and logisticregression analysis were used to analyze the relationships between thevariables.Avalueofp<0.05wasconsideredsigniﬁcant. SPSS®_version_15.0_was_used.

Results

Studypopulation

Of the total of 2549 students aged between 15 and 18yearsenrolledintheeightselectedschoolsinfourLisbon municipalities,854participated(54%femaleand46%male), corresponding to 34% adherence. Participation in private schools (n=3) was higher than in public schools (51% vs. 26%).Meanagewas16.3±0.9years,similarinboys(16.1± 0.9years)andgirls(16.3±0.8years).Thestudypopulation wascomposedof209adolescentsaged15years,291aged 16years,281aged17yearsand73aged18years,witha pre-dominanceofstudentsfromtheprivatesector(n=492,58%).

Qualitycontrol

Comparingthe measurementsof cholesterol,triglycerides and glucose in capillary blood and plasma, the differ-ence between mean cholesterol was minimal (−0.3± 11.9mg/dl),notsigniﬁcant,butwasgreaterfortriglycerides (35.9±34.69)andglucose(12.6±10.1)(p<0.001).

Lifestyles

Dataonhealthbehaviors(lifestyles)areshowninTable1.

Physicalactivity.Whenaskedhowmuchtimetheyspent inextracurricularphysical activitieswithan intensitythat leftthemoutofbreathorsweating,29%responded‘‘never’’ and53%responded between30min andlessthan4hours. Only 19% did this type of exercise for four hours a week or more, of whom morewere at private schools (22% vs. 17%,p=0.002)andmorewereboys(30%vs.10%,p<0.001). The majority (54%) had less than one hour per week of out-of-schoolphysicalactivityornone,especiallystudents aged18.

Sedentarytime.Regardingthenumberofhourswatching television,55%spentmorethanonehour,and8%spentthree hoursor moreperday. There werenostatistically signiﬁ-cantdifferencesaccordingtoschooltype(publicorprivate), genderorage.Everyday,61%spentoveronehoursitting, usingcomputersand/orgamesconsoles,and14%spentover threehours.The differencebetweenthe proportionusing suchdevices for over one hourdaily in private (67%) and publicschools(58%)wasstatisticallysigniﬁcant (p=0.007), aswasthe differencebetween boys(69%) andgirls (55%) (p<0.001).Thepercentageofschoolchildrenwith≥2hours screentime(TVand/orcomputer/gamesconsole)dailywas almost80%,themajorityattendingprivateschools(83%vs. 76%,p=0.03).

Diet.Soup,saladorvegetables,andfruitwereconsumed daily,respectively,by63%,61%and76%ofstudentsandless

(4)

Table1 Teenagers’lifestylesbygenderandschooltype(privateorpublic).

Lifestyles Total Gender Schooltype

Female Male Private Public

n % n % n % n % n %

Out-of-schoolphysicalactivity(weekly)

Never 226 28.6 164 38.1 62 17.2 66 21.6 160 33.0

30minto<4hours 415 52.5 224 52.1 191 52.9 173 56.5 242 49.9

≥4hours 150 19.0 42 9.8 108 29.9 67 21.9 83 17.1

Total 791 100 430 100 361 100 306 100 485 100

pa _--- _NS _NS

Screentime(daily)

<2hours 164 21.3 96 23.2 68 19.2 51 17.2 113 23.9 ≥2hours 605 78.7 318 76.8 287 80.8 245 82.8 360 76.1 Total 769 100 414 100 355 100 296 100 473 100

pa _--- _NS _0.030

Numberofmealsperday

≤3 48 6.2 25 5.9 23 6.6 20 6.7 28 5.9 4 131 16.9 64 15.1 67 19.1 55 18.5 76 15.9 ≥5 596 76.9 336 79.1 260 74.3 222 74.7 374 78.2 Total 775 100 425 100 350 100 297 100 478 100 pa _--- _NS _NS Smoking Non-smokers 681 86.5 365 85.1 316 88.3 271 89.4 410 84.7 Formersmokers 2 0.3 2 0.5 0 0.0 0 0.0 2 0.4 Occasionalsmokers 44 5.6 27 6.3 17 4.7 17 5.6 27 5.6 Dailysmokers 60 7.6 35 8.2 25 7.0 15 5.0 45 9.3 Total 787 100 429 100 358 100 303 38.5 484 61.5 pa _NS _NS

a_Fisher’s_exact_test.

thanonceaweekby14%,13%and6%.Differencesingender, ageandtypeofschoolwerenotsignificant. Regardingthe numberof dailymeals, alarge majority (77%)had fiveor moremealsadaybut6%hadthreeorfewer,while17%had four.Thedifferencereachedsignificanceonlyforage,with atendencytodecreasewithage(p<0.01).

Smoking. Most (87%) did not smoke. Of the 13% who smoked,5% didsooccasionallyand 8%everyday, withno differences according to type of school or gender. Mean dailyconsumptionwas5±6cigarettes.Theyreported hav-ingstartedsmokingat14±1.6yearsandthemodalclasswas 18years. Therewasa signiﬁcant increasein consumption withageinbothsexes(p<0.01).

Riskfactors

The maindataoncardiovascularrisk factorsareshownin Tables1and2andFigure1.

Overweight/obesity.Mostrespondents (81%)were nor-mal weight; only 3% were low weight, while 11% were overweight and 5% were obese, giving a ﬁgure for over-weight/obesityof 16%in theseadolescents,witha higher prevalenceinboysthaningirls(21%vs.12%,p=0.001). Dif-ferencesbyschooltypeandagewerenotsigniﬁcantbutthe highestprevalenceofoverweightwasin15-year-olds.

Hypertension.Bloodpressurewashigh-normalin28%and withcriteriaforhypertensionin11%.Systolicbloodpressure wasmoredeterminantthandiastolicbloodpressureinthe classiﬁcationofhypertension(46%vs.30%).Theprevalence ofhypertensioninboyswashigherthaningirls(15%vs.7%, p<0.001).

Cholesterol.Totalcholesterolwasnormalinmost adoles-cents(73%)but5%hadvalues≥200mg/dl.Theprevalence of hypercholesterolemia in girls was higher than in boys (7%vs.3%,p<0.005).

Triglycerides. The prevalence of hypertriglyceridemia (≥150 mg/dl)ingirlswashigherthaninboys(12%vs.5%, p<0.001).

Impairedglucosemetabolism.AccordingtotheAmerican Diabetes Associationclassiﬁcation9%hadimpairedfasting glucose (IFG) and 0.5% were diabetic. The prevalence of IFGanddiabeteswashigherinboys(11%)thaningirls(7%) (p=0.06).

Family history of premature CVD. Approximately 6% of the students’ parents reported a history of premature cardiovascular events, more of them mothers (aged 46± 4.7 years) than fathers (aged 48±5.7 years). This family historywasmorefrequentwitholderstudents(p<0.05).

Figure 1 shows the prevalence of individual risk factors in both sexes. The leading modiﬁable cardio-vascular risk factors in boys were overweight/obesity,

(5)

Table2 Biologicalcharacteristicsandfamilyhistoryofcardiovasculardiseasebygenderandschooltype(privateorpublic). Biologicalcharacteristics

andfamilyhistoryof cardiovasculardisease

Total Gender Schooltype

Female Male Private Public

n % n % n % n % n %

Bodymassindex

Underweight 23 2.7 11 2.4 12 3.1 6 1.7 17 3.5 Normalweight 690 81.2 393 85.4 297 76.2 299 83.3 391 79.6 Overweight 95 11.2 46 10.0 49 12.6 41 11.4 54 11.0 Obese 42 4.9 10 2.2 32 8.2 13 3.6 29 5.9 Total 850 100 460 100 396 100 359 42.2 491 57.8 pa _--- _0.001 _NS Bloodpressure Normal 521 61.1 356 77.6 165 41.9 229 63.4 292 59.3 High-normal 242 28.4 71 15.5 171 43.4 98 27.1 144 29.3 Grade1HT 59 6.9 24 5.2 35 8.9 24 6.6 35 7.1 Grade2HT 31 3.6 8 1.7 23 5.8 10 2.8 21 4.3 Total 853 100 459 100 394 100 361 42.3 492 57.7 pa _--- _<0.001 _NS Cholesterol Normal 607 72.6 296 65.2 311 81.4 252 71.2 355 73.7 Borderline 187 22.4 126 27.8 61 16.0 84 23.7 103 21.4 HC 42 5.0 32 7.0 10 2.6 18 5.1 24 5.0 Total 836 100 454 100 382 100 354 42.3 482 57.7 pa _--- _0.004 _NS Triglycerides Normal 721 91.3 376 87.9 345 95.3 303 91.8 418 90.9 HTG 69 8.7 52 12.1 17 4.7 27 8.2 42 9.1 Total 790 100 428 100 362 100 330 41.8 460 58.2 pa _--- _<0.001 _NS Glucometabolicstatus Normal 720 90.8 399 92.6 321 88.7 303 91.5 417 90.3 IFG 69 8.7 29 6.7 40 11.0 26 7.9 43 9.3 Diabetes 4 0.5 3 0.7 1 0.3 2 0.6 2 0.4 Total 793 100 431 100 362 100 331 41.7 462 58.2 pa _--- _NS _NS

Familyhistoryofprematurecardiovasculardisease

No 616 94.5 329 94.0 287 95.0 277 95.8 339 93.4

Yes 36 5.5 21 6.0 15 5.0 12 4.2 24 6.6

Total 652 100 350 100 302 100 289 44.3 363 55.7

pa _--- _NS _NS

HC:hypercholesterolemia;HT:hypertension;HTG:hypertriglyceridemia;IFG:impairedfastingglucose.

a _Fisher’s_exact_test.

hypertensionandsmoking,whileingirlstheyweresmoking, overweight/obesity and hypertriglyceridemia. Moreover, approximatelyhalf (49%) alreadyhadat least one modiﬁ-ableriskfactor(35%hadone,11%hadtwoand3%had≥3 riskfactors).

Relationshipsbetweenvariables

The relationships between the variables studied in the 15---18age-groupobtainedbymultivariatelogisticregression analysis are shown in Figure 2. Overweight/obesity was

associatedwithnumberof meals(≤3mealsdaily), abnor-malbloodpressurelevels(grade2hypertension),impaired glucose metabolism and hypertriglyceridemia. The varia-bles associated withhypertension were male gender and overweight/obesity.Impairedglucosemetabolismwas asso-ciated with hypertension, especially grade 2, and with overweight/obesity.Variablesrelatedtoelevated triglyce-rides were female gender and overweight/obesity, while thoseassociatedwithhighcholesterolwerefemalegender andimpairedglucosemetabolism.Smokingincreasedwith age,especiallyatage17.

(6)

Prevalence of risk factors (%)

0 5 10

Male Female

Hyper cholesterolemia

Family history of premature cardiovascular disease

Hypertriglyerceridemia

Impaired glucose metabolism

Hypertension

Smoking

Overweight/obesity

Total

15 20 25

Figure1 Prevalenceofcardiovascularriskfactorsinadolescents(15---18yearsofage)bygender.

Ten-yearrelativeriskforcardiovascularmortality Accordingtothetableofrelativerisk(RR)of cardiovascu-larmortalitybasedonbloodpressure,totalcholesteroland smoking, notdiscriminatedby gender,27% ofrespondents hadRR≥2.Ofthese,3%hadRR≥3.Estimated cardiovascu-larriskinboyswassigniﬁcantlyhigherthaningirls(38%vs. 17%,p<0.001)andoverallincreasedfrom21%atage15to 43%atage18(p<0.005)(Table3).

Discussion

Participantswereincludedfrombothpublicandprivate sec-ondary schools in orderto test the hypothesis that there were differences in students attending these two school types. The sample was balanced for school type, gender and age. The measuring instruments and laboratory tests were subjected to quality control to assess the reliabil-ity of theresults. Operational deﬁnitionsof the variables

Table3 Ten-yearrelativeriskforcardiovascularmortalityinadolescents(15---18yearsofage).

Relativerisk Total Gender Schooltype Age(years)

Female Male Private Public 15 16 17 18

1 n 608 374 234 276 332 158 208 201 41 % 73.3 83.1 61.7 78.4 69.6 79.4 73.8 72.8 56.9 2 n 194 71 123 65 129 36 64 67 27 % 2.4 15.8 32.5 18.5 27.0 18.1 22.7 24.3 37.5 3 n 24 3 21 10 14 5 10 6 3 % 2.9 0.7 5.5 2.8 2.9 2.5 3.5 2.2 4.2 4 n 3 2 1 1 2 0 0 2 1 % 0.4 0.4 0.3 0.3 0.4 0.0 0.0 0.7 1.4 p --- <0.001a _0.005a _0.003b

TestsconductedcomparingRR=1toRR≥2.

a_Fisher’s_exact_test. b _Pearson’s_chi-square_test.

(7)

Overweight/ obese Smoking High BP HTG HC Female gender IFG/diabetes Female gender Overweight/obesity Overweight/obesity high-normal BP/ pre-hypertension Grade 2 hypertension Male gender Overweight/obesity Borderline cholesterol Age: 17 years Age: 18 years OR=3.8 OR=1.7 p<0.001 p=0.042 p<0.001 p=0.023 p=0.004 p=0.011 p=0.008 p=0.012 p=0.043 p=0.001 p=0.015 p=0.010 p=0.018 p=0.039 p=0.005 p<0.001 p=0.018 p<0.001 p=0.028 95% CI=[1.90; 7.43] 95% CI=[1.02; 2.84] 95% CI=[1.25; 3.11] 95% CI=[1.24; 5.21] 95% CI=[1.40; 9.14] 95% CI=[1.19; 4.06] 95% CI=[1.02; 3.75] 95% CI=[1.43; 4.04] 95% CI=[1.14; 3.51] 95% CI=[1.19; 3.61] 95% CI=[1.55; 5.51] 95% CI=[1.15; 6.32] OR=2.9 OR=2.9 OR=2.4 OR=2.0 OR=2.1 OR=3.6 OR=2.2 OR=4.4 OR=2.9 OR=2.1 OR=2.0 OR=2.5 OR=3.6 OR=2.2 OR=2.0 Number of daily meals ≤3

Number of daily meals =4 High-normal BP/ pre-hypertension Grade 1 hypertension Grade 2 hypertension IFG/ diabetes Hypertriglyceridemia Impaired glucose metabolism 95% CI=[1.14; 3.74] 95% CI=[1.03; 3.33] 95% CI=[1.55; 11.95] 95% CI=[1.88; 6.79] 95% CI=[1.14; 4.11] 95% CI=[1.77; 10.80] 95% CI=[1.12; 7.51] OR=1.8 OR=4.3

Figure 2 Characteristics associated with cardiovascular risk factors (logistic regression). BP: blood pressure;CI: conﬁdence interval;HC:hypercholesterolemia;HTG:hypertriglyceridemia;IFG:impairedfastingglucose;OR:oddsratio.

and cut-offs werethose deﬁnedby medical societies and internationalorganizations.9---11_The_fact_that_all_the_schools

invitedagreedtoparticipateisanindicationthatthey rec-ognizedtheimportanceoftheresearch,goingsofarasto makeitpartoftheprojectselementofthecurriculum,even thoughonlyoneschoolwasamemberofREPSandsupported thestudyundertheaegisofthePromotionandEducationfor HealthprojectandoftheSchoolHealthAgenda.8

Theassessmentofphysicalactivityincludedthe180min perweekofphysicaleducationincludedinthecurriculum. On this basis, 19% of the participants met the WHO rec-ommendation that children and young people aged 5---17 shouldaccumulateatleast60mindailyofmoderateto vig-orous intensityphysicalactivity,andthatamountsgreater thanthatprovideadditionalhealthbeneﬁts.12_This_ﬁgure_is

slightlyhigherthanthe15%ofthe2006HealthBehaviorin School-agedChildren(HBSC)study13_and_conﬁrms_that_males

takemorephysicalactivitythanfemales,14_a_tendency_that

can be explained by different habits, tastes and goals of leisureactivities.However,severalfactorsinﬂuencelevels ofphysicalactivity:self-esteem,opportunities,motivation, childhoodandadolescencehabits,localresourcesandcosts oftransportandequipment,andothers.Someofthesemay underliethehigherlevelsofphysicalactivityinstudentswho

attendprivateschools.Nevertheless,asshowninother stud-ies,levelsofphysicalactivity andsports tendtodecrease withage, particularlyin females, andso schools must be abletomotivateyoungpeopletoremainactive.

With regard to screen time, the time spent at com-puters and games consoles reﬂects children’s interest in technologicalinnovationsandcontent. Genderdifferences suggest thatsuch activitiesare morepopular amongboys and imply that, in this case, accessibility is not a reason for thedifferences between thosewho attendpublic and privateschools.Theseresultsaresimilartothoseofother studies, in which young people spend an average of 3---5 hoursperdayusingthesemedia.15

The role ofdiet in obesity andatherosclerosis preven-tion is well deﬁned in reports and guidelines.9,16,17 _One

of the recommendations is that the ﬁrst balanced meal shouldbetakensoonafterwakingup,but11%ofthe stu-dentsreportednottakingbreakfast.Soup,acharacteristic componentof theMediterranean dietthatis knowntobe protectiveagainstsomechronicdiseases,includingcoronary heartdiseaseandcancer,18_was_reported_as_never_consumed

ornomorethanonceaweekby37%oftheparticipants.Diet shouldbe variedand dividedbetween at least ﬁvemeals a day,17 _but _almost _a _quarter _of _the _participants _did _not

(8)

followthisguidance.The negativerelation between over-weightandnumberofmealscorroboratestheevidenceand reinforcestherecommendation.

In this population, 13% smoked, although 5% only did sooccasionally. Theseﬁguresarehigherthanthoseofthe HBSC study, both in 2002 (8.5%) and in 2006 (5%),13 _but

thosestudentswerealsoyoungerthanthoseinthepresent study.However,theyarelowerthaninthe15---24age-group (24%)inthe4thPortugueseNationalHealthSurvey(INS),19

althoughinlinewiththeupwardtrendintheproportionof femalesmokers.Thehigherprevalenceofsmokingin pub-lic than inprivate schools is probablynot bychance, but the effectof differences in health educationand promo-tion.Determining themost common age tostart smoking (15years)identiﬁesariskgroupthatshouldbeapriorityin smokingprevention,sinceitis easiernottostartsmoking thantoquit.

The prevalence of body mass index (BMI) above the 85thpercentilewashigh (16%) andthatof obesity (>95th percentile, 5%) was higher than that reported in the 4th INS in the15---24 age-group (<2%). However,these ﬁgures are similar to those obtained in 2002 in Portugal.20 _The

percentage of overweight or obesity was higher in boys thanin girls,asalsofound in population studies21 _as_well

as the 4th INS, up to age 25, although the opposite was found over age 55.19 _In _the _four _years _between _ages ₁₅

and 18 there were no differences in overweight/obesity, since this problem starts earlier, before age 15, as can be seen in the COSI study.22 _In _Portugal, _obesity _has

increased tenfold relative to the 1970s.23 _The _obstacles

toreversing this trendareenormous, and nocountry has succeeded in overcoming them.6 _However, _unless _there

are changes, more health problems (psychological, car-diovascular, osteoarticular and others) can be expected. Urbanizationandtechnologicaldevelopmenthave brought great beneﬁts but are contributing to worsening dietary habits, increasing physical inactivity andstress in various forms,fromchildhoodonwards.Whatisneededisan envi-ronmentthat promoteshealth andreducesobesity,which requiresconcertedpoliciesandcollaboration,both nation-allyandinternationally.

The prevalence of hypertension (11%) lies within the rangeofother studies(7.5%and16.5%),11,24 _including_one

onuniversity students (10%).25 _It _was_higher _in _males, _as

inanotherstudy ofyoungadults.26 _This _may_be_explained

by hormonal changes after puberty, although the effect of overweight/obesity, more prevalent in males and pos-itively associated with blood pressure,27 _should _not _be

ignored.

According to the criteria of the American Academy of Pediatrics11 _the _prevalence _of _elevated _cholesterol

(>200 mg/dl) was 5%, lower than in US studies (13%) but closetovaluesinthePortuguesepopulationobservedina systematicreview thatincluded childrenandadolescents, and that found that hypercholesterolemia wasalso more commoninfemales.28_As_plasma_cholesterol_is_the_most

sta-blelipidvariableandsinceitshowedthe bestagreement according to quality control of the biochemical studies, this proportion in particular should be close to the true ﬁgure in this age-group. The prevalence of hypertriglyc-eridemia(7%),higheringirlsthaninboys,isslightlylower thaninastudyofchildrenagedbetween5and14(11%).28

Comparisonsarenoteasysincetriglycerides,althoughnot presentingsigniﬁcantcircadianorseasonalvariations,have far greater intra-individualvariability thancholesterol,so results on triglycerides should be regarded with caution. Moreover, the pre-test reliability of the method applied was not optimal, so this value may differ from the true prevalence.

Bloodglucoselevelscorrespondingtoimpairedglucose metabolism (9%)weremorefrequentthan expected,with 0.5% being classiﬁed as diabetic. Overall, there was a high prevalence of all glucose metabolism abnormalities, although astudyin2006of4370Americanteenagersaged between12and19,withafastingperiodofatleast8hours, obtained similarvalues(0.5%diabetic andabout11% with IFG).29 _Most _studies _in _this _area _are _in _the _adult

popula-tion and epidemiological information on type 1 diabetes is limited. The incidence calculated for Portugal in joint international studieswasapproximately8per10000 pop-ulation,lowerthaninothercountries.However,according totheDiabetesAtlasoftheInternationalDiabetes Federa-tionin2010,itwas13.2per10000populationinthoseaged under15years.Evenso,theprevalenceoftype1diabetes inourpopulationwashigher.Theproportionofyoung peo-plewithimpairedglucosemetabolismmaybeoverestimated becauseofthemethod’sunreliability(qualitycontrol agree-mentwasnotideal)and/orfailuretoobservethe12-hour fastingperiod.

Certainrelatedcharacteristicsarefrequentlyassociated (obesity, inﬂammatory factors, hereditary factors, hyper-tension, dyslipidemia, impaired glucose metabolism, and lifestyles).9 _Such _{relationships} _are_more _thoroughly

docu-mentedinepidemiologicalandclinicalstudiesinadultsthan inchildrenandadolescents,butthepathophysiology under-lyingtheseassociationsremainsthesame.Althoughfurther longitudinalstudiesfrominfancytoadulthoodarerequired toclarifyhowthesecharacteristicschangeovertime,ithas been conﬁrmedthatthereisadegree oftrackinginsome cases,forexampleinweightandbloodpressure,themore sowhenhighbloodpressurelevelsinchildhoodare associ-atedwithobesityorafamilyhistoryofhypertension.30 _The

relationship between overweight/obesity and blood pres-sure levelswasconfirmed inourstudybutthe association withimpairedglucosemetabolismwasasurprise,although itwouldnothavebeenifthestudyhadbeenperformedin the adult population.The association ofimpaired glucose metabolism, obesity and documentedhypertension,three componentsofthemetabolicsyndrome,suggeststhatthis syndromeisalreadyprevalentattheseages.Thehigher lev-elsoflipids,triglyceridesandtotalcholesterolobservedin girls isundoubtedlyrelatedtohormonalinfluences (estro-gen)duringadolescence,butthenegativeinfluenceofless physicalactivitycannotbeexcluded.However,asboyshad higherBMIandmorephysicalactivity,itappearsthatdiet hasmoreinfluenceonweightandBMIthanphysical activ-ity.Nevertheless,atatimewhenalltheseteenagershadat least180minofphysicalactivityperweek,thedifference betweenthisdurationanddailyexercisemaynotbe suffi-cienttoseparatethegroupsintermsofinfluenceonweight evolution. Itshouldbenotedthatoverweight/obesity was associated withthreerisk factors(hypertension, impaired glucose metabolism, andincreased triglycerides) and was dependentondietaryhabits(<4mealsaday).

(9)

Conclusions

For adolescents at school aged 15---18 years, the data obtainedshowthat:

(a) Mostwerephysicallyinactive(81%hadinsufﬁcient phys-icalactivityand79%spent ≥2hoursperday watching TVoratacomputer/gamesconsole);

(b) Manydidnotmeetdietaryrecommendationsconcerning choicesoffoodandnumberofmealsperday;

(c) Approximately half (49%) already had at least one modiﬁable risk factor (in descending order: over-weight/obesity 16%, smoking daily/occasionally 13%, hypertension 11%, diabetes/IFG 9.5%, hypertriglyc-eridemia9%,familyhistoryofprematurecardiovascular events6%,andhypercholesterolemia5%);

(d) ≥2riskfactorscoexistedin14%oftheparticipants; (e) Approximatelyaquarter(27%)hada≥210-yearrelative

riskoffatalcardiovascularevents;

(f) Overweight/obesityshowedapositiveassociationwith blood pressure,impaired glucosemetabolism and ele-vated triglycerides, and a negative association with numberofmealsperday.

Theresults demonstrate theneed for actionin provid-ingandencouraginghealthychoices, withanemphasison behavioralandlifestylechanges,particularlydiet,physical activity, and smoking. Preventive actions should be inte-gratedandinterventionsshouldbeaimedatindividualsand at familiesandcommunities.Regular monitoring oflevels and patterns of risk factors is required for planning and evaluating preventive interventions. Prevention based on lifelonghealthbehaviorscanbeeffectiveandshouldbeseen asavitalinvestmentinhealth.

Ethical

disclosures

Protection of human and animal subjects.The authors declarethatnoexperimentswereperformedonhumansor animalsforthisstudy.

Conﬁdentialityofdata.Theauthorsdeclarethattheyhave followedtheprotocolsoftheirworkcenteronthe publica-tionofpatientdata.

Righttoprivacyandinformedconsent.Theauthorshave obtained the written informedconsent of thepatients or subjectsmentionedinthearticle.Thecorrespondingauthor isinpossessionofthisdocument.

Funding

The Corac¸ão Jovemstudywasfundedby aresearchgrant fromtheFoundationAstraZeneca,obtainedfollowinga com-petitivecall.

Conﬂicts

of

interest

Theauthorshavenoconﬂictsofinteresttodeclare.

Acknowledgements

Theauthorsaregratefultoallwhomadethisstudypossible.

References

1.ForsenT,ErikssonJG,TuomilehtoJ,etal.Growthinuteroand during childhoodamong womenwho developcoronary heart disease:longitudinalstudy.BMJ.1999;319:1403---7.

2.BerensonGS,SrinivasanSR,BaoW,etal.Associationbetween multiple cardiovascular risk factors and the early develop-mentofatherosclerosis BogalusaHeartStudy. NEnglJMed. 1998;338:1650---6.

3.KotsevaK,WoodD,DeBackerG,etal.,onbehalfofEUROASPIRE study group. Cardiovascular prevention guidelines in daily practice:acomparisonofEUROASPIREIIIandIIIsurveysineight Europeancountries.Lancet.2009;372:929---40.

4.RoseG.Strategyofprevention:lessonsfromcardiovascular dis-ease.BMJ.1981;282:1847---51.

5.Fernstrom MH, Reed KA, Rahavi EB, et al. Communication strategiestohelpreducetheprevalenceofnon-communicable diseases:proceedingsfromtheinauguralIFICFoundationGlobal Dietand PhysicalActivityCommunicationsSummit.NutrRev. 2012;70:301---10.

6.WHORegionalCommitteeforEuropeEUR/RC61/12.ActionPlan forimplementationoftheEuropeanStrategyforthePrevention and Control of NoncommunicableDiseases2012---2116.Baku: WHORegionalOfﬁceforEurope;2011.

7.WHO/CDC.GlobalSchool-basedStudentHealthSurvey(GSHS). Available at: http://www.cdc.gov/gshs/questionnaire/index. htm

8.MS-DGS. Programa Nacional de Saúde Escolar. Despacho n◦12.045/2006(2asérie).D.R.n◦110.2006Jun7.

9.GrahamI,AtarD,Borch-JohnsenK,etal.Europeanguidelines oncardiovasculardiseasepreventioninclinicalpractice: exec-utivesummaryFourthJointTaskForceoftheEuropeanSociety ofCardiologyandothersocietiesoncardiovasculardisease pre-ventioninclinicalpractice(constitutedbyrepresentativesof ninesocietiesand byinvitedexperts).EurJCardiovascPrev Rehabil.2007;14Suppl.2:E1---40.

10.AAP. National Cholesterol Education Program: report of the expert panel on blood cholesterol levels in children and adolescents.Pediatrics.1992;89:525---84.

11.ADA.Diagnosisandclassiﬁcationofdiabetesmellitus.Diabetes Care.2005;28:S37---42.

12.WHO.GlobalRecommendationsonPhysicalActivityforHealth: 5---17 years old; 2011. Available at: http://www.who.int/ dietphysicalactivity/pa/en/index.html

13.MatosMG,SimõesC,ToméG,etal.,EquipadoProjetoAventura Social.Aventura Social & Saúde--- ASaúde dosadolescentes portugueses---Hojeeem8anos.RelatórioPreliminardoEstudo HBSC.FMH/UTL;2006.

14.HallalPC,BoAndersenL,BullFC,etal.,fortheLancet Phys-ical Activity Series Working Group. Global physical activity levels: surveillanceprogress,pitfalls,and prospects.Lancet. 2012;380:247---57.

15.GouveiaC,SilvaL,VirellaD,etal.Actividadefísicae seden-tarismoemadolescentesescolarizadosdoconcelhodeLisboa. ActaPediatrPort.2007;38:7---12.

16.WorldHealthAssembly55.23.Diet,physicalactivityandhealth. Geneva:WorldHealthOrganization;2002.

17.Conselho Nacional de Alimentação e Nutrição, Comissão de EducaçãoAlimentar.Recomendaçõesparaaeducação alimen-tardapopulaçãoportuguesa.Availableat:http://www.dgs.pt/ wwwbase/raiz/mlkImprimir.aspx?condigoms=.2007

(10)

18.Sánchez-MorenoC,CanoMP,AncosB,etal.Intakeof Mediter-raneanvegetablesouptreatedbypulsedelectricﬁeldsaffects plasmavitamin Cand antioxidantbiomarkers inhumans. Int JFoodSciNutr.2005;56:115---24.

19.INSA-INE.4◦ InquéritoNacionaldeSaúde---2005---2006;2007. Availableat:http://www.ine.pt;www.insa.pt

20.MatosMG,CarvalhosaSF,SimõesC,etal.,EquipadoAventura Social.AventuraSocial&Saúde---Asaúdedosadolescentes por-tugueses---EstudoNacionaldaRedeEuropeiaHBSC/OMS.FMH, ProgramadeEducac¸ãoparaTodos-Saúde;2005.

21.Carmo I, Santos O, Camolas J, et al. Overweight and obe-sityinPortugal:nationalprevalencein2003---2005.ObesRev. 2008;9:11---9.

22.RitoA,WijnhovenTMA,RutterH,etal.Prevalenceofobesity amongPortuguesechildren(6---8yearsold)usingthreedeﬁnition criteria:COSIPortugal2008.PediatrObes.2012;7:413---22. 23.SociedadePortuguesaparaoEstudodaObesidade.Prevalência

eMonitorizac¸ãodaObesidadeemPortugal;2008.Availableat:

http://www.speo-obesidade.pt/CDA/DOCSpublicos.aspx 24.Macedo MAE [PhD dissertation] Estudo epidemiológico da

pressãoarterialemcrianc¸asportuguesas.Faculdadede Medi-cinadoPorto;1988.

25.RochaE,FreitasMG,GomesP,etal.Níveisdepressão arte-rial dos estudantes da Universidade de Lisboa. Clín Saúde. 2003;I:45---52.

26.BurtVL,WheltonP,RoccellaEJ,etal.Prevalenceof hyperten-sionintheUSadultpopulation:resultsfromtheThirdNational HealthandNutritionExaminationSurvey,1988---91. Hyperten-sion.1995;25:305---13.

27.RahmouniK,CorreiaMLG,HaynesWG,etal.Obesity-associated hypertension. New insights into mechanisms. Hypertension. 2005;45:9---14.

28.CostaJ,BorgesM,OliveiraE,etal.Incidênciaeprevalênciada hipercolesterolémia em Portugal: Uma revisão sistemática daLiteratura.ParteII.RevPortCardiol.2003;22:683---702. 29.DuncanGE.Prevalence ofdiabetesand impairedfasting

glu-coselevelsamongUSadolescents.NationalHealthandNutrition Examination Survey, 1999---2002. Arch Pediatr Adolesc Med. 2006;160:523---8.

30.MacedoME,PizarroM,LimaMJ,etal.Índicesdetrackingda pressãoarterialnumestudolongitudinaldecrianc¸ase adoles-centes.RevPortCardiol.1998;17:243---9.