Indoor air quality and atopic sensitization in primary schools: A follow-up study

(1)

Porto

Biomedical

Journal

h tt p://w w w . p o r t o b i o m e d i c a l j o u r n a l . c o m /

Original

article

Indoor

air

quality

and

atopic

sensitization

in

primary

schools:

A

follow-up

study

João

Cavaleiro

Rufo

a,b,c,d,∗

_,

_Joana

_Madureira

c

_,

_Inês

_Paciência

a,b,c,d

_,

_Lívia

_Aguiar

d,e,f

_,

João

Paulo

Teixeira

d,e,f

_,

_André

_Moreira

a,b,d

_,

_Eduardo

_de

_Oliveira

_Fernandes

c

a_Faculty_of_Medicine_of_the_University_of_Porto,_Porto,_Portugal b_S._João_Hospital_Centre,_Porto,_Portugal

c_Institute_of_Science_and_Innovation_in_Mechanical_Engineering_and_Industrial_Engineering,_Porto,_Portugal d_EPIUnit,_Institute_of_Public_Health_of_the_University_of_Porto,_Porto,_Portugal

e_National_Institute_of_Health,_Porto,_Portugal f_Institute_of_Public_Health,_Porto,_Portugal

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received2May2016 Accepted18July2016 Availableonline10August2016 Keywords:

Schools Indoorairquality Children Follow-upstudy Particulatematter Atopy

a

b

s

t

r

a

c

t

Background:Theonsetandexacerbationofallergicdiseasesandasthmahavebeenassociatedwithpoor indoorairquality(IAQ)insideclassrooms.

Objective:TheaimwastoinvestigatehowIAQchangedinprimaryschoolsafterapplyingindoorairquality recommendations,andtoexplorehowthesechangesinﬂuencedallergicsensitizationonchildren. Methods: Totalvolatileorganiccompounds,PM2.5,PM10,CO2,CO,temperatureandrelativehumidity intheindoorandoutdoorairof20primaryschoolsweremeasuredin2010–2012.Theschoolstaff receivedinstructionsonhowtoimproveIAQinaccordancewiththededicatedguidelines.Atopystatus wasassessedinchildrenattendingtheparticipatingclassroomsbyskinpricktestsandexhalednitric oxide.Afollow-upsamplingcampaignwasperformedin2014–2015inthesameschools.

Results: IndoorPM2.5andPM10concentrationswereapproximately40%lowerinthefollow-up mea-surements (p<0.05). There were no signiﬁcantdifferences regarding outdoor PM concentrations. Nevertheless,PMlevelsfromthefollow-upcampaignstillexceededthereferencevalueestablishedby Portugueselegislation.Moreover,therewerenosigniﬁcantdifferencesinatopicprevalenceandFENO valuesbetweenthecampaigns.

Conclusion:ThesefindingssuggestthatadoptionoftherecommendationsbasedontheSINPHONIE guide-lineswasparticularlysuccessfulinreducingPM2.5andPM10inprimaryschoolsofPorto.Nevertheless, theschoolsfailedtoreducethelevelsofotherIAQpollutants,aswellastheprevalenceofatopicdisease. ©2016PBJ-AssociaçãoPortoBiomedical/PortoBiomedicalSociety.PublishedbyElsevierEspa ña, S.L.U.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/ licenses/by-nc-nd/4.0/).

Introduction

Schoolclassroomsareoneoftheindoorenvironmentswhere children spend most of theirtime.1 _Indoor _air _quality _(IAQ) _in

schoolshasthusbeenrecognizedasoneofthekeyinﬂuenceson children’shealth, given the highpotentialfor a largeexposure to indoor air contaminants in that environment. The develop-mentandexacerbationofallergicsensitization,asthma,rhinitisor

Abbreviations:eNO,exhalednitricoxide;IAQ,indoorairquality;PM,particulate matter;SPT,skinpricktest;TVOC,totalvolatileorganiccompounds;WHO,World HealthOrganization.

∗ Correspondingauthor.

E-mailaddress:[email protected](J.CavaleiroRufo).

rhinoconjunctivitishaveallbeenassociatedwithpoorIAQinside classrooms2,3 _and, _according _to_the _World _Health _Organization

(WHO),36%ofrespiratorydiseasesand22%ofchronicdiseasesare associatedwithapoorindoorenvironment.4

In 2010, to further investigate IAQ in schools, the SIN-PHONIEprojectwasconducted.Itcomprised36environmentand health institutionsfrom 25countries, and assessed severalIAQ parametersin morethan100Europeanschools.These parame-tersincludedtotal volatileorganiccompounds(TVOC),ﬁneand coarseparticulatematter(PM2.5andPM10,respectively),carbon

monoxide(CO),carbondioxide(CO2), temperature,and relative

humidity.5 _The_results_of_the_SINPHONIE_project_led_to

produc-tionofasetofGuidelinesforhealthyenvironmentswithinEuropean schools6 _that, _as _its_name _suggests, _aimed_at _improving_IAQ _in

schools.

http://dx.doi.org/10.1016/j.pbj.2016.07.003

2444-8664/©2016PBJ-AssociaçãoPortoBiomedical/PortoBiomedicalSociety.PublishedbyElsevierEspa ña,S.L.U.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

DuringtheSINPHONIEproject,atotalof20schoolswereaudited inPorto,Portugal,andeachschoolreceivedareportwiththeaudit’s results,along withspeciﬁcrecommendationsbasedonthe SIN-PHONIEguidelines onhowtoimprove IAQ.For instance,ifthe classroomspresentedCO2levelsabove1250ppmduetoair

reno-vationdeﬁciency,thereportrecommendedthatschoolstaffshould openwindowsandpromoteventilationduringrecessperiods.As anotherexample,ifPMconcentrationsinaclassroomwerevery highandtheblackboardwasidentiﬁedasamainemissionsource, therecommendationswouldsuggestamorefrequentblackboard cleansingwithawetcloth,topreventparticlesuspension. How-ever,therewasnosupervisiontoverifyiftherecommendations werecorrectlyadopted.

Theaimof thisstudy wastoinvestigate(a)ifand howIAQ improvedintheauditedschoolsinPorto,afterthe implementa-tionofrecommendations;and(b)tocomparetheatopicstatusof thechildrenattendingtheschoolsbetweenbothcampaigns.

Materials/methods

Studydesign

AftergettingtheconsentfromPortoCityHallandschools’ prin-cipals,19 ofthe20 schoolsthat participated intheSINPHONIE projectinPortowereincludedinthisfollow-upstudy.Themissing schoolhassincebeenclosedduetoextremelyhighconcentrations ofindoorradonandthuswasnotconsideredforthefollow-up.For conveniencereasons,theSINPHONIEandthefollow-upcampaign shallbeabbreviatedinfuturereferencesasT1andT2,respectively. Ineachcampaign,childrenattendingthesampledclassrooms wereinvitedtoparticipateinaclinicalassessmentandaninformed consentrequestwassenttotheirlegalguardians inaccordance withtheHelsinkideclaration.

Samplingcampaign

T1campaignoccurredfrom2011to2012,whileT2occurred from2014to2015.Inordertoreplicatetheauditingconditions, the T2 campaign was similarly conducted during the heating season,i.e.duringwinter.Ineachschool,2to4classroomsand1 outdoorlocationwereselectedforIAQassessment,resultingina totalof67classroomsand19outdoorspaces.Preferencewasgiven toclassrooms withhighdensities of occupation (occupant/m2₎

andfullyoccupiedthroughouttheweek.IdenticallytoT1,TVOC, PM2.5, PM10, CO, CO2, temperatureand relative humiditywere

sampled.Safeand childproof samplinglocationswereselected, complyingwithISO16000-1.7_Indoor_samples_were_collected_near

children’sbreathingzone(approximately0.7–1.5maboveground) andnocloserthan1mtoawall,window,doororactiveheating system. The instruments were placed as far away as possible fromtheblackboard,whenapplicable.Thesamplingprocesswas supervisedbya researcher,avoidingdisturbancesofthenormal classroomactivities.

Volatileorganiccompoundswere collectedby passive diffu-sion,during5daysperclassroom/outdoor(fromMondaytoFriday), usingstainless-steelsorptiontubescontainingasingle-bedTenax® TA(60/80)matrix.Aﬁeldblankwasemployedineveryschoolto controlcontaminationduringtransportandsampling.Allsamples weretakeninduplicatetoverifythereproducibilityof measure-ments.

PortableTSIDustTrakDRXphotometers(model8533,TSIInc.) wereusedfortheassessmentofPM2.5andPM10concentrations.

Thisequipmentmeasuresparticleswithalaserphotometerbased onlightscatteringprinciple.Themeasuringrangeoftheequipment is1–150×103_␮g/m3_with_accuracy_of_±0.1%_for₁_␮g/m3_,

operat-ingwithaﬂowrateof3.0l/minusingabuilt-indiaphragmpump

poweredbyaninternalbattery.Instrumentswereinstalledinside eachclassroomandweresettocontinuouslymeasurePMduringat leastoneschoolday(7h,avoidingMondaysandFridays).Logging intervalsweresetto1min.OutdoorPM2.5andPM10weresampled

byasimilarinstrument.Therespectiveinstrumentswereinstalled intheschoolplaygrounds,ataheightof1–1.2mandprotectedfrom rain.Theinstrumentsetupwasthesameastheoneusedindoors. Thephotometerswerecalibratedexternally,onceperyear,atthe manufacturer.

CO, CO2, temperature and relative humidity levels were

recordedcontinuouslyfor24hperdayduring5days perweek (fromMondaytoFriday)ineachsamplinglocation.These param-eterswereallmeasuredusingtheIAQ-CALCmonitor(model7545, TSI,Inc.).Theequipmentcombinedanelectrochemicalsensorfor CO,aninfrarednon-dispersivesensorforCO2,athermistorfor

mea-suringtemperatureinarangefrom0to60◦Cwithanaccuracy of±0.6◦_C,_and_a_thin-ﬁlm_capacitive_sensor_for_relative_humidity

(rangeof5–95%;accuracy±3.0%).Loggingintervalsweresetto 5min.

Clinicalassessment

Participatingchildrenperformedskinpricktests(SPT)andhad theirexhaledlevelofnitricoxide(FENO)measuredbyatrained professional.

ThevolunteersweresubmittedtoSPTontheirforearmusinga QuickTestTM_applicator_containing_house_dust_mite,_mix_of_weeds,

mixofgrasses,catdander,dogdander,Alternariaalternata,negative control,andapositivecontrolconsistingofhistamineat10mg/ml, allbelongingtothesamebatch(HallAllergy,Netherlands).Results wereread15minafterwardsandatopywasdeﬁnedbyapositive SPTtoatleastoneoftheallergens.8 _If_patients_were_on_tricyclic

antidepressantsorantihistaminesoriftheyhadappliedany prod-uctcontainingcorticosteroidsontheskinwithintheprevious7 days,skinpricktestswouldbepostponed.

TheNObreath(BedfontScientificLtd.,UK)wasusedtoperform theFENOassessmentandtheresultswereexpressedaspartsper billion(ppb).TheFENOvalueswerestratifiedaccordingtothe offi-cialATSguidelinesforchildren.9

Laboratoryanalysis

Thestainless-steelTenaxtubeswerethermallydesorbed(Dani STD33.50)andquantiﬁedusinganon-polarcolumnbygas chro-matography (Agilent Technologies 6890N) coupled to a mass spectrometrydetector(AgilentTechnologies5973),accordingto ISO16000-6.10_TVOC_{concentrations}_were_quantiﬁed_using_toluene

responsefactor, andwerecalculatedasthesumofVOCeluting betweenhexaneandhexadecane(included),expressedastoluene. Statisticalanalysis

StatisticalanalysiswasperformedusingIBM®SPSS®version23. TheKolmogorov-Smirnovtestwasusedtocheckifdatawas nor-mallydistributed.Sincenon-Gaussiandistributionswereobserved forallcontinuousvariablesotherthanage,non-parametrictests were selectedfor inferential statistics. Age betweenboth cam-paignswascomparedusingthet-studenttest,whiletheWilcoxon signed ranks test was used to compare non-parametric data betweencampaigns.

Results

WhencomparingoutdoorairparameterconcentrationsinT1 andT2,theresultsshowedthatcertainmeasuredparameterswere signiﬁcantlydifferentbetweenthetwocampaigns(Table1).For

(3)

Table1

OutdoorconcentrationsofairparametersmeasuredinT1andT2.“p”valuesinboldcorrespondtostatisticallysigniﬁcantdifferencesbetweenthetwocampaigns(Wilcoxon signedrankstest).

Parameters T1 T2 p 25% Median 75% 25% Median 75% TVOC(mg/m3₎ _0.04 _0.05 _0.06 _0.06 _0.08 _0.14 _0.01 PM2.5(mg/m3) 0.04 0.07 0.11 0.03 0.06 0.09 0.64 PM10(mg/m3) 0.05 0.08 0.11 0.03 0.07 0.11 0.59 CO(ppm) 0.06 0.12 0.53 0.13 0.16 0.54 0.87 CO2(ppm) 365 412 520 300 316 400 <0.01 Temperature(◦C) 11.3 12.3 14.0 12.8 14.9 16.8 0.01 Rel.humidity(%) 60.0 65.2 70.7 54.0 66.1 70.1 0.68 Table2

IndoorconcentrationsofairparametersmeasuredinT1andT2.“p”valuesinboldcorrespondtostatisticallysigniﬁcantdifferencesbetweenthetwocampaigns(Wilcoxon signedrankstest).

Parameters T1 T2 p 25% Median 75% 25% Median 75% TVOC(mg/m3₎ _0.11 _0.14 _0.17 _0.15 _0.20 _0.26 _0.13 PM2.5(mg/m3) 0.07 0.09 0.12 0.04 0.06 0.08 0.01 PM10(mg/m3) 0.12 0.13 0.16 0.07 0.09 0.13 <0.01 CO(ppm) 0.20 0.38 0.47 0.39 0.66 0.78 0.01 CO2(ppm) 1175 1666 1941 839 1475 1891 0.18 Temperature(◦C) 19.5 20.7 21.9 18.8 19.5 21.4 0.44 Rel.humidity(%) 48.6 54.2 60.4 52.2 60.0 64.2 0.04

instance,outdoorCO2concentrationsweresigniﬁcantlyhigherin

T1whencomparedtoT2(p<0.01).Conversely,temperatureand TVOCincreasedsigniﬁcantlyinT2(p=0.01,forbothcases).No sig-niﬁcantdifferenceswerefoundbetweenT1andT2inregardsto outdoorCO,relativehumidity,PM2.5andPM10.

Regardingthe indoor air, there were nosigniﬁcant changes intemperaturebetweenthetwocampaigns.Moreover,although therewasaCO2 decreaseandaTVOCincrease duringT2,these

changeswerealsonon-signiﬁcant.However,signiﬁcantdifferences wereobservedforotherindoorairparameters(Table2).

IndoorconcentrationsofCOandrelativehumidityincreased sig-niﬁcantlyinT2(p=0.01andp=0.04,respectively),whilePM2.5and

PM10concentrationsdecreasedsigniﬁcantly(p=0.01andp<0.01,

respectively), being approximately 40% lower when compared toT1. Nevertheless,themedian PM2.5 andPM10 concentrations

obtainedinT2(0.06and0.09mg/m3_,_{respectively)}_still_exceeded

the safety limit established by the Portuguese legislation of 0.025mg/m3_for_PM

2.5and0.05mg/m3forPM10.11Fig.1illustrates

theseresults.

Atotalof1580and1523childrenwereinvitedtoparticipatein T1andT2,respectively.FENOdatawasobtainedfor351children inT1and802inT2.Atotalof330childrenperformedSPTinT1 while786childrenperformedSPT T2.Therewerenosignificant differencesingenderbetweencampaigns(p=0.838)buttherewere inage(p<0.01),althoughthemeanswerequitesimilar(8.6±0.8in T1vs.8.7±0.8inT2).NosignificantdifferenceswerefoundinSPT positivitybetweencampaigns(p>0.05),althoughtheprevalenceof allergicsensitizationwashigherinT1whencomparedtoT2(35.3 and31.7%,respectively).Therewerealsonosignificantdifferences inFENOvaluesbetweenthecampaigns(p>0.05).However,similar toSPTpositivity,childreninT1presentedhigherlevelsofFENO whencomparedtochildreninT2(18.5and16.9ppb,respectively). TheseresultsarepresentedinTable3.

Discussion

Thisstudyconstitutestheﬁrsttimethattheimpactof adopt-ing the recommendations based on the guidelines for healthy

0.150 ∗ ∗∗ 0.125 0.100 0.075 Mean concentr ation (mg/m3) 0.050 0.025 0.000 PM_2.5 PM₁₀ Campaign T1 T2 PM₁₀ limit (Portuguese legislation) PM_2.5 limit (Portuguese legislation)

Fig.1. IndoormeanconcentrationsofPM2.5andPM10inT1andT2.Thedashed

linescorrespondtothesafetylimitsrecommendedbythePortugueselegislation (11).*p=0.01;**p<0.01(Wilcoxonsignedrankstest).

environmentswithinEuropeanschoolswasinvestigated.Theresults showhowIAQchangedinschoolsafterprovidingthe aforemen-tionedrecommendations,whilenotdisregardingthedifferences intheoutdoorenvironment.

The most accentuated changes between campaigns were observedinindoorPMconcentrations.Outdoorairisoneofthe mostpredominantsourcesofindoorPM.2,12_However,_there_were

nosigniﬁcantdifferencesinoutdoorPM2.5 andPM10

concentra-tionsbetweenT1andT2,stronglysuggestingthatthedecreasein thefollow-upmaybeassociatedwithachangeinindoorsources oroccupants’activities.Evenconsideringpossibleconfounding fac-tors,suchasthehigherrelativehumidityinT2,thedecreaseinPM concentrationswashighly signiﬁcant.Aspreviously mentioned, reduction ofindoorsources isthemostprobablecausefor this resultand,althoughtheimplementationoftherecommendations

(4)

Table3

SummarizedresultsoftheclinicaltestsinT1andT2.Thepresented“p”values representthestatisticalsigniﬁcanceofthecomparisonsbetweencampaigns.

Clinicaltest T1 T2 p

N(males)a ₃₅₁₍₁₇₆₎ ₈₀₂₍₄₃₁₎ _0.838b

Age(years) 8.6(±0.8) 8.7(±0.8) <0.01c

FENO(ppb) 18.5(±17.4) 16.9(±20.0) 0.147d

SPTpositivity(%) 35.3 31.7 0.893*

Resultsarepresentedinmean(±SD)exceptifstatedotherwise.

a_Number_of_cases. b_Chi-squared_test. c _t-student_test.

d _Wilcoxon_signed_ranks_test.

basedontheSINPHONIEguidelineswasnotsupervised,itis possi-bletoassumethatcertainactionstoreduceindoorPM2.5andPM10,

suchascleaningthechalkboardsmorefrequentlywithawetcloth toavoidre-suspensionofparticles,weretakenintoconsideration. Unfortunately,thenatureofthepresentstudydoesnotallowthe identiﬁcationofindoorPMsources,thereforeitisnotpossibleto ascertainwhichactionswereresponsibleforthedecreaseofindoor PM2.5andPM10concentrationsintheschools.Despitethe

signiﬁ-cantreductions,medianPMconcentrationsinT2stillexceededthe Portugueselegislationby80to140%,suggestingthatfurther rec-ommendationsandactionsshouldbetakenintoconsiderationin ordertoenhancetheimprovementofIAQ.Moreover,itisimportant topointthatoutdoorPMconcentrationsalsoexceededtheWHO recommendedlimitinbothT1andT2,13_thus_further_efforts_should

bemadetoreduceoutdooremissionsinordertoprevent penetra-tionoflargeconcentrationsofPMintotheindoorenvironments.

AlthoughnosignificantchangesinTVOCconcentrationswere observed betweenT1 and T2, it is possible to notice a certain tendencyforhigherconcentrationsinthefollow-up.Thereason behind this slight increase may be associated withthe signifi-cantlyhigher outdoor concentrationsof TVOCin T2. Moreover, although indoor concentrations were considerably lower than therecommendedvalue establishbythePortuguese legislation (600␮g/m3₎11_,_certain_specific_volatile_organic_compounds

recog-nizedbytheWHOtoberesponsibleforadversehealtheffectsin highconcentrations,suchasbenzeneortoluene,maybeincreased inschools.13_Special_attention_should_be_given_to_these_compounds

infuturestudiesinschools.

DespiteoutdoorCO2 concentrationsbeingsigniﬁcantlylower

inthesecondsamplingcampaign,indoorconcentrations,aproxy fortheventilationrate,showednosigniﬁcantchangesandwere stillabovethePortugueselegislatedlimitof1250ppm.This sug-geststhatventilation intheclassroomsmaystillbeinadequate andCO2 levelsarestillaprobleminschools.Furthermore,even

thoughthe recommendations underlinedtheimportance of air renovation,itseemsthattheschoolstaff’seffortswerenot suf-ﬁcient tosigniﬁcantlyreduce indoorCO2 concentrations. These

resultsareinaccordancewithpreviousstudiesthatidentiﬁed sim-ilarproblemsconcerningIAQinschools.14–17_{Interestingly,}_a_study

byGeelenetal.,2008,whereseveraldifferentapproachestoreduce CO2 inschools (suchasspeciﬁc classroom-advice,CO2 warning

devicesand teachingpackagesbasedonventilation recommen-dations)wereinvestigated,showedthattheinterventionswere unsuccessfulinreducingindoorCO2concentrationstolevelsbelow

1000ppm.Nevertheless,theauthorsshowedthat,tosigniﬁcantly improveairrenovationinschools,CO2warningdeviceand

teach-ingpackagecombinedwithaclass-speciﬁcventilationadviceare effectivetools,whilegivingtheventilationadviceisnoteffective onitsown.18_The_present_study_supports_these_results,_suggesting

thattheschoolstafftendtodisregardtherecommendationsthat aimtopromoteproperventilationinclassrooms.Moreover, inad-equateventilationisariskfactorforincreasedconcentrationsof

indoorairpollutants,suchasPMorTVOC.SinceindoorCO2

con-centrationswerenon-significantly differentbetweenT1andT2, anyalterationintheconcentrationsofindoorairparametersin thefollow-upweremostprobablynotsignificantlyassociatedwith changesinairrenovation.However,thiscannotbecompletely ver-ifiedsinceventilationrateswerenotmeasuredinT1and,therefore, CO2concentrationsarethemostapproximateindicator.

TheindoormedianconcentrationsofCOmeasuredinboth cam-paigns (0.38and0.66in T1andT2,respectively)werefarfrom the protectionlimit of 9.00ppm establishedby thePortuguese legislation,11_and_were,_thus,_not_problematic._Despite_this,_a

sig-niﬁcant increase was observed in T2. Carbon monoxide is one ofthemostcharacteristic trafﬁcpollutants,usually observedin urbanareas.15 _However, _in _the_present _study,_the_median

out-doorconcentrationsofCOwerelowerthanthoseindoorsinboth campaigns, suggesting theexistence of indoor sources suchas combustionactivity,naturalplantdecayorsimplytheoccupants’ exhaledbreathwhichmayhavebeenfurtheraggravatedbylow ventilationratesintheclassroom.19,20

Themeasuredmedianoutdoortemperaturevaluessuggestthat T1 had a colder heating seasonthan T2, although the relative humiditywasnot significantlydifferent betweenthetwo cam-paigns. Regarding the indoor air, however, while temperature registerednosignificantchanges,relative humiditywas signifi-cantlyhigherinT2.Thischangeinindoorrelativehumiditymay notonlyinfluencetheconcentrationofindoorairpollutants(i.e. numberofparticlesinsuspension),butmayalsoaffectthe occu-pants’perceptionofpoorIAQ,21,22_as_suggested_by_Feng_et_al.₍₁₉₉₈₎

whoshowedthattheperceptionofairqualityisnegatively asso-ciatedwithtemperatureandrelative humidity.Therefore,since relativehumiditywashigherinT2,theschoolstaffmayhavehad alowerperceptionofIAQwhencomparedtoT1,whichcouldhave contributedinaminorextenttosomedisregardingofthe recom-mendationsbytheschooloccupants.

Inregardstotheclinicaltrials,theparticipationrateswere con-siderablyhigherinT2whencomparedtoT1(228–238%higher).The reasonbehindthisincreaseinparticipationratesisprobably associ-atedwithanenhancedawarenessoftheparentstotheproblemsof poorIAQinschoolsinT2.Moreover,thefactthatoneoftheschools sampledinT1wasclosedduetoIAQissuesmayhaveincreasedthe parents’concerntothisissue,thusresultinginahighernumberof legalguardiansconsentingtheirchildrentoparticipateinthe sec-ondcampaign.WhiletherewerenosigniﬁcantdifferencesinFENO valuesandSPTpositivitybetweenthecampaigns,bothparameters werelowerinT1.Thisisaninterestingresultsincethistendency foralowerprevalenceofatopicsensitizationoverlappedwithan improvementoftheIAQ.However,itisnotpossibletoestablisha causalrelation.

Thepresentstudyhassomelimitationsmainlyinvolvingbias producedbythepresenceofsomeconfoundingfactors.Themost relevant confounding factor is the ventilation rate, which was notsupervisedineithercampaign,thusCO2 levelswereusedas

themostappropriateindicator.Additionally,theobserved differ-encesinoutdoorairbetweenbothcampaigns mayhavehadan impactontheIAQ.Finally,theimpactofaspecific recommenda-tiononIAQcouldnotbeverifiedsincenosourceapportionment analysiswasconductedinthepresentstudy,thereforeit isnot possibletoidentifywhichrecommendationsor guidelineswere truly important in the improvement of IAQ in the schools. A moredetailedanalysis,withamoreextensivenumberof build-ing/classroomcharacteristics,couldalsohelpsuggestingspecific measuresofprevention.However,thehighnumberofschools re-auditedinthefollow-up(95%),therigorousmimicryofthebaseline auditmethodology,as wellasthelargeamountof participants anddatacollected,allcountasstrengthsthatsupportthestudy results.

(5)

Conclusions

ThepresentworkinvestigatedhowSINPHONIE-based recom-mendationscontributedtoimprovingIAQinschoolsinafollow-up study.Moreover,occupant’ssensitizationstatuswasalsoassessed. Themostsigniﬁcantresultspointedoutthatprimaryschoolswere abletoreduceindoorPM2.5 and PM10 concentrations. This

sig-niﬁcantdecrease in PMwas most likely caused by changes in occupational activities presumably associated with the recom-mendationstransmittedtotheschoolstaff.Nevertheless,median indoorPM2.5and PM10still exceededthePortugueselegislation

by at least 80%, and the levels of other IAQ parameters, such as of CO2, remained high, indicating that adequate ventilation

wasnotguaranteed.Thissuggeststhattheuseofguidelinesfor healthyenvironmentswithinEuropeanschoolscontributedtothe improvementofIAQinschools,butfurtherworkisrequiredinorder tominimizetheadversehealtheffectsonchildren,teachersand schoolstaffassociatedwithpoorIAQ.

Conﬂictsofinterest

Theauthorsdeclarenoconﬂictsofinterest.

Funding

This work was ﬁnanced by the HEBE project (NORTE-01-0145-FEDER-000010), and by Fundac¸ão para a Ciência e TecnologiathroughtheARIAproject(PTDC/DTP-SAP/1522/2012, FCOMP-01-0124-FEDER-028709) and through the scholar-ships SFRH/BD/108605/2015, SFRH/BD/112269/2015 and SFRH/BPD/105100/2014.

AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,atdoi:10.1016/j.pbj.2016.07.003.

References

1.RaysoniAU,StockTH,SarnatJA,MontoyaSosaT,EbeltSarnatS,HolguinF,etal. Characterizationoftrafﬁc-relatedairpollutantmetricsatfourschoolsinElPaso,

Texas,USA:implicationsforexposureassessmentandsitingschoolsinurban areas.AtmosEnviron.2013;80:140–51.

2.MadureiraJ,PaciênciaI,RufoJ,RamosE,BarrosH,TeixeiraJP,etal.Indoorair qualityinschoolsanditsrelationshipwithchildren’srespiratorysymptoms. AtmosEnviron.2015;118:145–56.

3.MohaiP,KweonB-S,LeeS,ArdK.Airpollutionaroundschoolsislinkedtopoorer studenthealthandacademicperformance.HealthAff.2011;30:852–62.

4.WorldHealthOrganization,Copenhagen,DenmarkHealthandenvironmentin Europe:progressassessment;2010.

5.CsobodE,Annesi-MaesanoI,CarrerP,KephalopoulosS,MadureiraJ,RudnaiP, etal.SINPHONIE(SchoolsIndoorPollutionandHealthObservatoryNetworkin Europe):executivesummaryoftheﬁnalreport;2014.

6.KephalopoulosS,CsobodE,BruinendeBruinY,OliveiraFernandesE.Guidelines forhealthyenvironmentswithinEuropeanschools;2014.

7.Indoorairepart1:generalaspectsofsamplingstrategy.2004.

8.GINA.Globalstrategyforasthmamanagementandprevention2015.Available from:http://www.ginasthma.org.

9.DweikRA,BoggsPB,ErzurumSC,IrvinCG,LeighMW,LundbergJO,etal.An ofﬁ-cialATSclinicalpracticeguideline:interpretationofexhalednitricoxidelevels (FENO)forclinicalapplications.AmJRespirCritCareMed.2011;184:602–15.

10.Determinationofvolatileorganiccompoundsinindoorandtestchamberairby activesamplingonTenaxTASorbent,thermaldesorptionandgas chromatog-raphyusingMSorMS-FID;2011.

11.Regulamentodedesempenhoenergéticodosedifíciosdecomércioeservic¸os (recs)requisitosdeventilac¸ãoequalidadedoarinterior;2013.

12.AlmeidaSM,CanhaN,SilvaA,FreitasMdC,PegasP,AlvesC,etal.Children exposuretoatmosphericparticlesinindoorofLisbonprimaryschools.Atmos Environ.2011;45:7594–9.

13.WorldHealthOrganization,Copenhagen,DenmarkWHOguidelinesforindoor airquality:selectedpollutants;2010.

14.Bakó-BiróZ,Clements-CroomeDJ,KochharN,AwbiHB,WilliamsMJ.Ventilation ratesinschoolsandpupils’performance.BuildEnviron.2012;48:215–23.

15.PegasPN,AlvesCA,EvtyuginaMG,NunesT,CerqueiraM,FranchiM,etal.Indoor airqualityinelementaryschoolsofLisboninspring.EnvironGeochemHealth. 2010;33:455–68.

16.MadureiraJ,Alvim-FerrazMCM,RodriguesS,Gonc¸alvesC,AzevedoMC,Pinto E,etal.Indoorairqualityinschoolsandhealthsymptomsamongportuguese teachers.HumEcolRiskAssess:IntJ.2009;15:159–69.

17.MumovicD,PalmerJ,DaviesM,OrmeM,RidleyI,OreszczynT,etal.Winter indoorairquality,thermalcomfortandacousticperformanceofnewlybuilt secondaryschoolsinEngland.BuildEnviron.2009;44:1466–77.

18.GeelenLM,HuijbregtsMA,RagasAM,BretveldRW,JansHW,vanDoornWJ,etal. Comparingtheeffectivenessofinterventionstoimproveventilationbehaviorin primaryschools.IndoorAir.2008;18:416–24.Epub2008/08/05.

19.ZayasuK,SekizawaK,OkinagaS,YamayaM,OhruiT,SasakiH.Increased car-bonmonoxideinexhaledairofasthmaticpatients.AmJRespirCritCareMed. 1997;156:1140–3.Epub1997/11/14.

20.MadureiraJ,PaciênciaI,PereiraC,TeixeiraJP,FernandesEdO.Indoorairquality inPortugueseschools:levelsandsourcesofpollutants.IndoorAir.2015.

21.FangL,ClausenG,FangerPO.Impactoftemperatureandhumidityonthe per-ceptionofindoorairquality.IndoorAir.1998;8:80–90.

22.ChengY,HeK-b,DuZ-y,ZhengM,DuanF-k,MaY-l.Humidityplaysanimportant roleinthePM2.5pollutioninBeijing.EnvironPollut.2015;197:68–75.