Health effects of air pollution on length of respiratory cancer survival

MISSING IMAGE

Material Information

Title:
Health effects of air pollution on length of respiratory cancer survival
Physical Description:
Unknown
Creator:
Xu, Xiaohui
Ha, Sandie
Kan, Haidong
Hu, Hui
Curbow, Barbara A.
Lissaker, Claudia TK
Publisher:
Bio-Med Central (BMC Public Health)
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
System ID:
AA00017954:00001

Full Text




















Year




Full Text

PAGE 1

RESEARCHARTICLEOpenAccessHealtheffectsofairpollutiononlengthof respiratorycancersurvivalXiaohuiXu1*,SandieHa1,HaidongKan2,HuiHu1,BarbaraACurbow3andClaudiaTKLissaker1AbstractBackground: Airpollutionhasbeenextensivelyandconsistentlylinkedwithmortality.However,nostudyhas investigatedthehealtheffectsofairpollutiononlengthofsurvivalamongdiagnosedrespiratorycancerpatients. Methods: Inthisstudy,weconductedapopulation-basedstudytoinvestigateifairpollutionexposurehasadverse effectsonsurvivaltimeofrespiratorycancercasesinLosAngeles(LA),CAandHonolulu,HI.WeselectedallWhite respiratorycancerpatientsinthetwostudyareasfromthe1992 – 2008SurveillanceEpidemiologyandEndResults cancerdata.Deathfromrespiratorycancerandlengthofsurvivalwerethemainoutcomes. Results: Kaplan-Meiersurvivalanalysisshowsthatallrespiratorycancercasesexposedtohighairpollutionreferring totheindividualsfromLAhadasignificantlyshortersurvivaltimethanthelowpollutionexposuregroupreferring tothosefromHonoluluwithoutadjustingforothercovariates(p<0.0001).Moreover,theresultsfromtheCox Proportional-Hazardsmodelssuggestthatexposuretoparticleslessthan10micrometersindiameter(PM10)was associatedwithanincreasedriskofcancerdeath(HR=1.48,95%CI:1.44-1.52per10 g/m3increaseinPM10)after adjustingfordemographicfactorsandcancercharacteristics.Similarresultswereobservedforparticleslessthan2.5 micrometersindiameterandozone. Conclusion: Ourstudyindicatesthatairpollutionmayhavedeleteriouseffectsonthelengthofsurvivalamong Whiterespiratorycancerpatients.Thisstudycallsforattentiontopreventiveeffortfromairpollutionforthis susceptiblepopulationinstandardcancerpatientcare.Thefindingsfromthisstudywarrantfurtherinvestigation. Keywords: Airpollution,Cancer,Cancersurvival,Respiratorycancer,LungcancerBackgroundCanceristhesecondleadingcauseofdeathintheUS, accountingforover500,000deathsannually[1].Worldwide,over7.6millionpeoplediedfromcancerin2008 basedontheGLOBOCANestimates[2].Respiratory cancer,especiallylungcancer,isoneofthemostcommonlydiagnosedcancersaswellastheleadingcauseof cancerdeath.IntheUnitedStatesalone,over200,000 newlungcancercaseswerediagnosed,andofwhom nearly161,250diedin2011[3].Reducingthedeathrate fromrespiratorycancerisstillachallengingmissionalthoughaslightprogresshasbeenmadeoverthepast years. Epidemiologicalstudieshaveshownconsistentevidenceofshort-termhealtheffectsofairpollutionon cardiopulmonarymorbidityandmortalityusingtimeseriesanalysesandcase-crossoverdesigns[4-7].Moreover,airpollutionasariskfactorofrespiratorycancer hasbeenreportedinseveralstudies[8-12].However, previousstudiesrevealedthatmajoreffectsofairpollutionhavebeenobservedamongthesusceptiblegroups suchaschildren,elderlyandpersonswithchronicconditionsincludingdiabetesandheartdisease[13,14]. Therefore,forresearchonairpollutionandrespiratory cancer,itmaybealsointerestingtoinvestigatetheeffectsofairpollutiononthelengthofsurvivalamongrespiratorycancerpatientswhoarealreadyseverely stressed. Toourknowledge,nostudieshavebeenconductedto examinethelong-termeffectsofairpollutiononany typeofcancersurvival.Inhalationofpollutedairmay *Correspondence: xhxu@phhp.ufl.edu1DepartmentofEpidemiology,CollegeofPublicHealthandHealth ProfessionsandCollegeofMedicine,UniversityofFlorida,POBox100231, Gainesville,FL32610,USA Fulllistofauthorinformationisavailableattheendofthearticle 2013Xuetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and reproductioninanymedium,providedtheoriginalworkisproperlycited.Xu etal.BMCPublicHealth 2013, 13 :800 http://www.biomedcentral.com/1471-2458/13/800

PAGE 2

havesignificantimpactsonsurvivalamongrespiratory cancercases.Asindustrializationandurbanizationcontinuestoadvance,globalincreaseinindustrialprocesses andenergyconsumptionresultsinburningalarge amountoffossilfuelincludinggasoline,coalandnatural gas.Theseactivitiesarecontinuingtoreleasetoxicpollutantsintotheairwebreathe[15].DespitethesignificantefforttoachievecleanerairfollowingtheCleanAir Actin1970intheUSA[16],airpollutionremainsasignificantpublichealthproblem[17].Accordingtothe StateoftheAir2012reportbytheAmericanLungAssociation,aboutfouroutoftenpeoplearelivingin countiesthatreceiveanFforairqualityintheUnited States[18].Giventhehighprevalenceofbothairpollutionandrespiratorycancer,itisurgentandcriticalto understandifexposuretoairpollutionhasadverseeffectsonrespiratorycancersurvival. Thepurposeofthispopulation-basedstudywastoinvestigateadverseeffectsofairpollutiononrespiratory cancersurvivalusingthe1992 – 2008SurveillanceEpidemiologyandEndResults(SEER)cancerdatainthe U.S.Inaddition,thisstudyfurtherinvestigatedwhether airpollutionhasimpactsonsurvivalofcancerataspecificsiteofrespiratorysystem.MethodsStudylocationsInthisstudy,weselectedHonolulu,HIandLosAngeles (LA),CAastwostudyareasbecausethesetwolocationsprovideddifferentairpollutionexposurelevels. HonoluluisoneofthecleanestareaswhileLAisonthe listofmostpollutedcitiesinU.S.[18].Weanalyzedambientairpollutionlevelsinthesetwostudyareasusing theU.SEnvironmentalProtectionAgencymonitoredair data.Themeansofannualaveragesofconcentrationsof criteriaairpollutantsincludingparticleslessthan10micrometersindiameter(PM10),particleslessthan2.5micrometersindiameter(PM2.5)andOzone(O3)between 1992 – 2008inLA,CAandHonolulu,HIarelistedin Additionalfile1:TableS1.ThetrendsofannualconcentrationsofPM10,PM2.5andO3arepresentedin Additionalfile2:FigureS1.Inthisanalysis,theresults suggestthatLA,CAhadsignificantlyhigherairpollutionlevelsthanHonolulu,HI.RespiratorycancercasesAllrespiratorycancercasesamongWhitesinthetwo selectedstudyareaswereidentifiedfromtheSEERcancerregistrydatafrom1992 – 2008.TheSEERcancer registry,operatedbytheNationalCancerInstitute (NCI),isasystemofcancerregistriesthatincludes20 differentgeographicareasandcovers28%oftheU.S population.Respiratorycancerswereidentifiedbased onprimarysiteusing InternationalClassificationof DiseasesforOncology, ThirdEdition(ICD-O-3).Cancers beingincludedwerenose,nasalcavityandmiddleear (C300-C301,C310-C319);larynx(C320-C329);lungand bronchus(C340-C349);pleura(C384);andtrachea, mediastinumandotherorgans(C339,C381-C383,C388, C390,C398,C399).Inthisstudy,weonlyselectedthe yearsof1992 – 2008duetothefactthatLosAngeles joinedtheSEERregistriesstartingin1992.Furthermore, sincethenumberofBlackandotherminorityracesin Hawaiiarelow,andthegroupAsian/PacificIslanders consistsofmultiplenationalitieswhichmayhavesignificantdifferencesincultureanddietaryhabits,thisstudy onlyfocusesontheWhitepopulation.OutcomeassessmentThecausesofdeathwereobtainedfromSEER,andfurthercategorizedintotwogroups:1)cancer-specific death,and2)deathofothercauses.Overallmortality wasthemainoutcomeofinterest,andwealsoexamined allcompetingrisks.Survivaltime(unit:months)wascalculatedastheintervalfromthetimeofdiagnosistothe timeofdeathortotheendofthestudy.AirpollutionexposureassessmentTwomethodswereusedtoassessairpollutionexposure, 1)ecologicalexposuremeasurement,and2)individual exposuremeasurement.Inthefirstmethod,cancercases inHonoluluCountyweredefinedasalowexposure group;andthosefromLosAngelesCountyweredefined asahighexposuregroupbecausetheanalysisofairpollutionlevelsbetweentwostudyareassuggestedthat temporalvariationofairpollutionwasverysmallin bothstudyareasanddifferencesbetweenthetwostudy areasweresignificant. Inthesecondmethod,individualexposuretoairpollutantsofPM10,PM2.5,andO3wasestimatedbasedon dateofdiagnosis(yearandmonth),survivaltimeor follow-uptime(unit:months)andcountyofresidence. Dailyairpollutionmonitoreddatainthestudyareas during1992 – 2008wereobtainedfromtheU.S.EnvironmentalProtectionAgency ’ sAirQualitySystem.CountylevelmonthlymeansofairpollutantsincludingPM10, PM2.5andO3werecalculatedusingthedatafromall monitorswithinthecounty.Foreachcancercase,individualairpollutionexposureduringsurvivaltimeor follow-uptimewasestimatedbyaveragingmonthly meansofairpollutantsduringtheperiodinthecounty wherethesubjectlived.Bothcontinuousvariablesand categoricalvariablesofairpollutantswereusedinthe followingregressionanalyses.Thecategoricalvariables ofairpollutantswerecategorizedashighandlowlevels usingthemedianconcentrationsofairpollutants.Xu etal.BMCPublicHealth 2013, 13 :800 Page2of9 http://www.biomedcentral.com/1471-2458/13/800

PAGE 3

CovariatesForeachcancercase,SEERcontainsinformationregardingbasicdemographicsincludingage(<54,55 – 69,70 – 84, and85+yearsold),gender(malevs.female),maritalstatus atdiagnosis(unmarriedmarried,separated/divorce, windowed,andunknown)andinformationoncancer characteristicsincludingdateofdiagnosis(1992 – 1996, 1997 – 2001,2002 – 2006,2006+)site(nose,nasalcavity andmiddleear,larynx,lungandbronchus,andothers), andstagesofcancer(Insitu,localized,regional,distant, andunknown),etc.Theseimportantfactorswereincluded asconfoundersinthemodels.Inaddition,informationon deathincludeddateofdeathifapplicable,thecauseof death,andotherinformation.StatisticalanalysesDescriptivestatisticssuchasChi-squaretestandt-test wereappliedforcomparingthedistributionsofcategoricalandcontinuousvariablesamongrespiratorycancer casesintwostudyareas.Kaplain-Meierlifetableanalyseswereconductedtoshowsurvivalcurvesbetween twogroupsandBreslowtestswereperformedtotestthe significanceofthedifferenceofsurvivalbetweentwo groups.Coxproportionalhazardmodelswereusedto assesshowtheinfluenceofairpollutionwererelatedto timetooverallmortality.Wealsoestimatedthemodel forallcompetingrisks.Thecause-specifichazardfunctionisthefundamentalconceptincompetingrisk models,whichisthehazardofdeathfromagivencause inthepresenceofthecompetingevents.Caseswitha specificcauseofdeatharecomparedwithallthosewho died,butspecificcausesofdeatharenotcompared.Specifically,weexamineddeathsduetorespiratorycancer andcausesotherthanrespiratorycancer.Furthermore, weexaminedairpollution ’ seffectsonrespiratorycancer survivalamongcasesonlyfromLosAngeles(over90% oftotalsamplewerefromLA)toavoidpotential confoundingbycity-levelfactors(e.g.,smokingrates,socioeconomicstatus,qualityofmedicalcare),which coulddifferbetweenLAandHonolulu.Sensitivityanalysiswasfurtherconductedoncaseswithlungand bronchuscancersonly.Allstatisticalanalyseswere conductedusingSASversion9.3(Cary,NC).ResultsFrom1992 – 2008,therewere58,586respiratorycancer casesamongWhitesincluding2,393inHonolulu County,HIand56,193inLosAngelesCounty,CA. Table1presentscharacteristicsofallrespiratorycancer casesamongWhitesbypollutionlevel.Overall,themajorityofcancercaseswerelung/bronchuscancercases (morethan90%)amongWhitesinbothareas.Thedistributionsofageatdiagnosis,sex,maritalstatusat diagnosis,andcancerstageatdiagnosisweresignificantlydifferentbetweentwostudyareas. Figure1showstheresultsofKaplan-Meiersurvival analysisofallrespiratorycancercasesbetween Honolulu,HIandLosAngeles,CA.Itsuggeststhatall respiratorycancercaseslivinginanareawithhighair pollutionhadasignificantlyhigheroverallandcancerspecificmortalityratethanthoselivinginanareawith lowairpollution(p<0.0001)withoutadjustingforother covariates,whilemortalityduetocausesotherthanrespiratorycancerwasnotsignificant. Figure2showstheresultsofKaplan-Meiersurvival analysisforallrespiratorycancercasessurvivalbetween highandlowairpollutantexposurelevels.Itsuggests thatcancercasesexposuretohighlevelsofairpollutants includingPM10andPM2.5hadsignificantlylowsurvival ratethanthoseexposuretolowlevels.Inaddition,cases exposuretohighlevelsofO3hadsignificantlyhigher overallandcancer-specificmortalityratethanthosewith lowexposurelevels(p<0.0001). Table2presentstheunadjustedandadjustedhazard ratios(HR)and95%confidenceintervals(CI)fordeath amongWhites.SubjectsinLosAngelesCountyhada statisticallysignificantincreaseof14%inoverallmortalityratecomparedtothosewholivedinHonolulu Countyintheunadjustedmodel(HR:1.14,95%CI: 1.08-1.20).Afteradjustmentforimportantfactorssuch asageatdiagnosis,maritalstatusatdiagnosis,gender, cancerstageatdiagnosis,andprimarysites,theratesof overallmortalityfromLosAngelesCountyremained statisticallysignificantlyhigher(HR:1.07,95%CI:1.02, 1.13).Consistentresultswereobservedforrespiratory cancer-specificdeathafteradjustingforconfounders (HR:1.08,95%CI:1.02-1.14). Table3presentstheunadjustedandadjustedhazardratios(HR)and95%confidenceintervals(CI)ofdeathdue toeachairpollutantsamongWhites.Theresultssuggest thatexposuretoairpollutantsincludingPM10,PM2.5,and O3wassignificantlyassociatedwithbothoverallmortality andcause-specificmortalityforrespiratorycancerpatients. ConsistentresultswereobservedwhenonlycasesinLos Angeleswereincluded.Furthermore,similarresultswere foundinthesensitivityanalysiswhenonlycaseswithlung andbronchuscancerwereincluded(resultsnotshown).DiscussionOurstudysuggeststhatamongd iagnosedrespiratorycancer cases,thoselivinginaheavily-p ollutedareahadsignificantly shortersurvivaltimescomparedtothoselivinginalesspollutedarea.Further,basedontheanalysisofindividual exposureassignment,wealsofoundthatexposuretoair pollutantsincludingPM10,PM2.5andO3hadadverseeffects onthelengthofrespiratorycancersurvival.Toourknowledge,thisstudyisthefirstandlargestpopulation-basedXu etal.BMCPublicHealth 2013, 13 :800 Page3of9 http://www.biomedcentral.com/1471-2458/13/800

PAGE 4

Table1CharacteristicsofrespiratorycancercasesamongWhitefrom1992 – 2008bypollutionregion(n=58,586)CharacteristicsRespiratorycancercases(%)PValue* HonoluluLosAngeles (n=2,393)(n=56,193) Ageatdiagnoses 54257(10.7)5,860(10.4)0.03 55-69857(35.8)19,715(35.1) 70-841,128(47.1)26,127(46.5) 85+151(6.3)4,489(8.0) SexMale1,419(59.3)30,602(54.5)<0.01 Female974(40.7)25,591(45.5) MaritalstatusatdiagnosesUnmarried273(11.4)7,567(13.5)<0.01 Married1,170(48.9)28,622(50.9) Separated/Divorce413(17.3)7,137(12.7) Windowed477(19.9)11,807(21.0) Unknown60(2.5)1,060(1.9) CancerstageInsitu10(0.42)280(0.50)<0.01 Localized534(22.3)10,199(18.2) Regional629(26.3)13,049(23.2) Distant1,057(44.2)26,037(46.3) Unknown163(6.8)6,628(11.8) PrimarysiteNose,nasalcavityandmiddleear29(1.2)756(1.4)0.55 Larynx169(7.1)3,672(6.5) Lungandbronchus2,188(91.4)51,503(91.7) Others7(0.3)262(0.4) Yeardiagnosed1992-1996697(29.1)17,977(32.0)<0.01 1997-2001683(28.5)16,724(29.8) 2002-2006728(30.4)15,490(27.6) 2007-2008285(11.9)6,002(10.7)*P-valuefromChi-squaretest. Figure1 RespiratorycancersurvivalprobabilitybystudyareasamongWhitesduring1992 – 2008(n=58,586). Xu etal.BMCPublicHealth 2013, 13 :800 Page4of9 http://www.biomedcentral.com/1471-2458/13/800

PAGE 5

studytoinvestigatelong-termeffectsofairpollutiononrespiratorycancersurvival.Ourstudyraisedconcernsthat airpollutionexposuremayhaveagreatimpactonthe lengthofsurvivalforrespiratorycancerpatients.Moreimportantly,thisstudyalsobringsattentiontotheneedfor preventiveeffortstoprotectcancerpatientsfromairpollution,whichiscurrentlyoverlookedincancerpatientcare. Theeffectofairpollutiononthelengthofsurvivalof respiratorycancerpatientsisbiologicallyplausible. Long-termexposuretoairpollutionhasbeenextensively linkedwithmortalityandcancer-specificmortality [19-23].Previousstudiesalsosuggeststhatairpollution, especiallyPM2.5andO3,hasbeenassociatedwithearly mortalityinsusceptiblepopulationswithchronicconditionssuchasCOPD,diabetes,heartfailure,ormyocardialinfarction[24,25].Theseresultsareconsistentwith thefindingsfromthisstudy.Thebiologicalmechanisms bywhichairpollutioncanimpactthelengthofsurvival ofrespiratorycancerpatientsisstillunclear.However,it isclearthattherespiratorysystemisanorganthatis mostdirectlyaffectedbyairpollutionwhichcarries manytypesoftoxicchemicalsincludingthosewithcarcinogenicpotential[26].Thesepollutantscouldreach thewallofrespiratorysystemandevenintotheblood andotherorgans,andinducesystematicinflammation [27].Respiratorycancerpatientsareatincreasedriskfor Figure2 RespiratorycancersurvivalprobabilitybyairpollutantsamongWhitesduring1992 – 2008(n=58,586). Xu etal.BMCPublicHealth 2013, 13 :800 Page5of9 http://www.biomedcentral.com/1471-2458/13/800

PAGE 6

impairedrespiratoryfunctionbecauseofcancer.Furthermore,sideeffectssuchasdepressedimmunityand decreasedresistancetoinfectionthataccompaniescancertherapymakethisgroupevenmoresusceptibleto contaminantsintheambientair.Thus,exposuretoair pollutioncouldfurtherreducetheirrespiratoryfunction, causerespiratoryproblemsandothercomplications and/ormakethemvulnerabletootherriskfactors[28]. Consequently,theadverseeffectofairpollutioncould significantlydecreasesurvivaltimeforcancerpatients andcauseanearlierdeath. Inthisstudy,weselectedtwostudyareaswithdocumentedsignificantlydifferentairpollutionlevels.InLos Angeles,CA,theaverageannualPM2.5was18.1 g/m3during1999 – 2008,whichiswellabovetheU.S.EPA ’ s annualPM2.5standardof15 g/m3.Meanwhile,the averageannualPM2.5during1999 – 2008inHonolulu, HIwasonly4.3 g/m3,whichismuchbelowtheannual standard(PleaserefertoAdditionalfile1:TableS1). Overthestudyperiod,airpollutionlevelhasdecreased intheLAbutlittletemporalvariationwasobservedin Honolulu.Overall,thedifferenceofairqualitybetween twoplacesremainedconsistentlysignificantduring 1992 – 2008(PleaserefertoAdditionalfile2:FigureS1). WeselectedLAasahighly-pollutedareaandHonolulu asalow-pollutedareainthisstudyandthisecological exposureassignmentmightreducemisclassificationof airpollutionexposurebecauseofasignificantdifference andlowtemporalvariationofairpollutionlevelsbetweentwoareas,i.e.variationsofairpollutionbetween counties(LAvs.Honolulu)weremuchgreaterthanvariationswithinacounty.Inaddition,cancerpatientswere unlikelytomove,whichwouldalsominimizethepotentialerrorsinexposureassessmentwhenweusedthis ecologicalexposureassignment.However,wecannot ruleoutthepossibilitythattheobserveddifferencesin Table2Adjustedandunadjustedhazardratios(HR)and95%confidenceintervals(CI)forrespiratorycancerdeath from1992 – 2008amongWhites(n=58,586)Characteristics OveralldeathCancer-specificdeath UnadjustedHRAdjustedHRaUnadjustedHRAdjustedHRa(95%CI)(95%CI)(95%CI)(95%CI) AirpollutionlevelbLow1.001.001.001.00 High1.14(1.08 – 1.20)*1.07(1.02 – 1.13)*1.15(1.09 – 1.22)*1.08(1.02 – 1.14)* Ageatdiagnoses 541.001.001.001.00 55 – 691.29(1.25 – 1.34)*1.30(1.25 – 1.35)*1.21(1.16 – 1.25)*1.21(1.16 – 1.26)* 70 – 841.72(1.66 – 1.78)*1.73(1.67 – 1.80)*1.49(1.43 – 1.54)*1.49(1.44 – 1.55)* 85+2.70(2.58 – 2.83)*2.57(2.43 – 2.71)*2.25(2.14 – 2.36)*2.12(2.00 – 2.25)* MaritalstatusUnmarried1.001.001.001.00 Married0.89(0.86 – 0.91)*0.82(0.79 – 0.85)*0.91(0.88 – 0.94)*0.86(0.83 – 0.89)* Separated/Divorce1.01(0.97 – 1.05)0.97(0.93 – 1.01)1.03(0.99 – 1.07)0.99(0.95 – 1.04) Widowed1.26(1.22 – 1.30)*1.02(0.98 – 1.06)1.23(1.19 – 1.28)*1.06(1.01 – 1.10)* SexMale1.001.001.001.00 Female0.97(0.95 – 0.99)*0.82(0.81 – 0.84)*0.99(0.97 – 1.01)0.84(0.82 – 0.86)* CancerstageInsituorLocalized1.001.001.001.00 Regional1.98(1.91 – 2.06)*1.96(1.89 – 2.03)*2.54(2.43 – 2.65)*2.46(2.36 – 2.57)* Distant5.26(5.09 – 5.44)*5.04(4.87 – 5.21)*7.20(6.91 – 7.49)*6.64(6.37 – 6.92)*PrimarysiteLungandbronchus1.001.001.001.00 Nose,nasalcavityandmiddleear0.38(0.34 – 0.42)*0.49(0.44 – 0.55)*0.32(0.29 – 0.36)*0.43(0.38 – 0.49)* Larynx0.32(0.31 – 0.34)*0.54(0.51 – 0.57)*0.23(0.21 – 0.24)*0.41(0.38 – 0.44)* Others0.36(0.30 – 0.42)*0.38(0.30 – 0.48)*0.38(0.31 – 0.46)*0.40(0.31 – 0.52)* Yeardiagnosed2007 – 20081.001.001.001.00 2002 – 20061.04(1.00 – 1.09)1.25(1.20 – 1.32)*1.06(1.01 – 1.11)*1.09(1.04 – 1.15)* 1997 – 20011.04(0.99 – 1.08)1.18(1.12 – 1.24)*1.04(0.99 – 1.09)1.20(1.14 – 1.26)* 1992 – 19961.06(1.02 – 1.11)*1.25(1.20 – 1.32)*1.05(1.00 – 1.10)1.27(1.21 – 1.34)**Statisticallysignificantatalpha=0.05.aAdjustedhazardratiosareadjustedforAgeatDiagnoses,Sex,Maritalstatusatdiagnoses,Cancerstage,Primarysite,andYeardiagnosed.bSubjectsfromLosAngeles,CAweredefinedashighairpollutionexposuregroupwhilethosefromHonolulu,HIweredefinedaslowairpollution exposuregroup.Xu etal.BMCPublicHealth 2013, 13 :800 Page6of9 http://www.biomedcentral.com/1471-2458/13/800

PAGE 7

cancersurvivaltimebetweentwolocationscouldbealso duetootherfactorsthatdifferacrossthetwolocations, suchassmokingrate,socioeconomicstatusandquality ofmedicalcareetc.Furthermore,individualmeasurementsofairpollutionwerealsoestimatedusingthe county-levelairpollutiondatafromtheU.S.EPAair monitoringsystem.Theresultsfromtheanalysesofindividualexposureassignmentalsosuggestthatexposure tohigherairpollutionmayhaveadverseeffectsonthe lengthofsurvivalfromrespiratorycancersinthisstudy. However,sinceHonoluluhasmuchlowerpollution levelsthanLA,mostoftheindividualswithhighpollutionexposurewouldbefromLAandmostwithlowpollutionexposurewillbefromHonolulu.Asaresult,any observedassociationbetweenpollutionandcancersurvivalcouldstillbeattributedtootherfactorsthatdiffer betweenthetwocities.Forexample,solarultraviolet-B (UVB)andvitaminDwerelinkedtocancermortalityin manystudies[29-33].SolarUVBdosesaremuchhigher inHonoluluthaninLA.Theseunselectedfactorsmay explaintheobservedassociations.Tofurthereliminate thispossibility,weperformedasensitiveanalysistoinvestigatetheassociationbetweenairpollutionand cancersurvivalineachcity.AsLAhadover95%of casesandhigherairpollutionlevel,welimitedouranalysisinLA.Theresultsfromthisanalysisremainconsistent.Therefore,theobservedassociationbetweenair pollutionandrespiratorycancersurvivalinthisstudyis unlikelyexplainedbypotentialconfoundingduetocitylevelfactors.Finally,althoughindividualdemographical factorsandstageofcancerwereadjustedforinthis study,otherunselectedfactorssuchasdietary,medical care,andlifestylefactorsofsmokingandalcoholconsumptionmayalsoconfoundtheobservedassociation. Asinformationonthesefactorsisnotavailableinthis study,wecouldnotcontrolfortheirpotential confoundingeffects.However,althoughdemographic distributionsbetweenLAandHonoluluaremuchdifferent,weonlystudiedoneraceofWhiteinthisstudy. Thus,theinfluencesofdifferentcultureanddietary habitsareunlikelytoexplainthesignificantdifferences inlengthofsurvivalfromrespiratorycancer.Moreover, bothareasinthisstudyareurbanareaswiththesame healthcaresystemintheUnitedStates,whichisalso improbabletoaccountforthissignificantdifference. Inaddition,ascancerpatients,currenttobaccouse Table3Hazardratios(HR)ofdeathbyairpollutantsforrespiratorycancercasesamongWhitesfrom1992-2008AirpollutantaOveralldeathCancer-specificdeath UnadjustedHRAdjustedHRbUnadjustedHRAdjustedHRb(95%CI)(95%CI)(95%CI)(95%CI) HonoluluandLosAngeles(n=58,586) PM10Low(<29.6 g/m3)1.001.001.001.00 High( 29.6 g/m3)1.41(1.39 – 1.44)*1.63(1.58 – 1.67)*1.39(1.36 – 1.42)*1.56(1.51 – 1.61)*Continuousvariable(per10 g/m3)1.40(1.37 – 1.42)*1.48(1.44 – 1.52)*1.37(1.35 – 1.40)*1.43(1.39 – 1.46)*PM2.5Low(<17.6 g/m3)1.001.001.001.00 High( 17.6 g/m3)1.08(1.05 – 1.11)*2.04(1.96 – 1.12)*1.07(1.04 – 1.10)*1.97(1.89 – 2.05)*Continuousvariable(per5 g/m3)1.23(1.21 – 1.25)*1.57(1.53 – 1.61)*1.20(1.18 – 1.22)*1.49(1.45 – 1.53)*O3Low(<0.041PPM)1.001.001.001.00 High( 0.041PPM)1.07(1.05 – 1.09)*1.04(1.01 – 1.06)*1.07(1.05 – 1.10)*1.04(1.02 – 1.07)*Continuousvariable(per10PPB)1.08(1.06 – 1.09)*1.04(1.03 – 1.06)*1.09(1.07 – 1.11)*1.06(1.04 – 1.07)*LosAngeles(n=56,193) PM10Low(<29.7 g/m3)1.001.001.001.00 High( 29.7 g/m3)1.46(1.43 – 1.49)*1.77(1.72 – 1.83)*1.44(1.41 – 1.48)*1.70(1.64 – 1.76)*Continuousvariable(per10 g/m3)1.51(1.48 – 1.54)*1.91(1.85 – 1.98)*1.47(1.44 – 1.50)*1.79(1.73 – 1.86)*PM2.5Low(<17.9 g/m3)1.001.001.001.00 High( 17.9 g/m3)1.09(1.06 – 1.12)*2.24(2.15 – 2.33)*1.07(1.04 – 1.10)*2.13(2.04 – 2.22)*Continuousvariable(per5 g/m3)1.42(1.39 – 1.45)*2.51(2.43 – 2.59)*1.35(1.32 – 1.39)*2.28(2.20 – 2.36)*O3Low(<0.041PPM)1.001.001.001.00 High( 0.041PPM)1.15(1.13 – 1.17)*1.12(1.09 – 1.15)*1.15(1.13 – 1.18)*1.12(1.09 – 1.15)*Continuousvariable(per10PPB)1.07(1.05 – 1.09)*1.04(1.02 – 1.06)*1.09(1.07 – 1.11)*1.06(1.04 – 1.08)**Statisticallysignificantatalpha=0.05.aHighandlowlevelswerecategorizedusingthemedianconcentrationofairpollutant.bAdjustedhazardratiosareadjustedforAgeatDiagnoses,Sex,Maritalstatusatdiagnoses,Cancerstage,Primarysite,andYeardiagnosed.Xu etal.BMCPublicHealth 2013, 13 :800 Page7of9 http://www.biomedcentral.com/1471-2458/13/800

PAGE 8

andalcoholconsumptionarenotcommon;thusthe confoundingeffectsofthesefactorsshouldbelimited [34].Further,thepovertyratesamongWhitesinthetwo studyareas,accordingtotheU.Scensusdata,arevery close,i.e.10%inHonolulu,HIand11%inLosAngeles, CA[35].Therefore,ourpreliminarydataarepromising andconvincing,suggestingthatthisfieldisworthyof furtherinvestigation.ConclusionInsummary,thisstudyrevealed thatairpollutionexposure mayhavedeleteriouseffectsonlengthofsurvivalfromrespiratorycancerpatients.With12.5millionpeopleliving withcancerintheU.S.andanannualincidenceandmortalityratesofapproximately5 00and200per100,000,respectively,itisimportanttopayattentiontopotentialeffectsof airpollutiononcancersurviva l.Carefulassessmentofthe potentialdeleteriouseffectsofairpollutionamongthissusceptiblegroupisalsonecessa ryforimprovementofcancer survivalandestablishmentofsoundregulatorypolicyto promotethepublichealthandwel fare.Asseverallimitations existinthisstudy,additionalresearchisclearlywarranted.WhatthispaperaddsAirpollutionisaglobalenvironmentalissueand,toour knowledge,nostudyhasbeenconductedtoinvestigate healtheffectofairpollutiononlengthofsurvivalamong cancerpatients.Inaddition,p rotectionofthissusceptible groupofcancerpatientsfrom environmentalthreatshas largelybeenoverlookedinstandardclinicalcare.Inthis study,weselectedtwostudyarea swithsignificantlydifferent airpollutionlevels:Honolulu,HI(lowlevelsofairpollution) vs.LosAngeles,CA(highlevelsofairpollution),andthen estimatedindividualairpollut ionexposureduringsurvival periodorstudyperiodusingairmonitordataandfurtherinvestigatedifexposurestopar ticularmatterandozonehave significantimpactsonthelengthofcancersurvival.Our studysuggeststhatexposuretohighlevelsofairpollution hadadverseeffectsonlengthofsurvivalcomparedwithlow levelsofairpollutionafteradjustingforimportantconfounderssuchage,gender,race,diagnosisstageofcancer andprimarysites.Thisstudyalsocallsforattentiontoprotectionsfromenvironmentalcontaminantsforthissusceptiblegroup.AdditionalfilesAdditionalfile1:TableS1. MeansofannualaveragesofPM10,PM2.5andO3between1992 – 2008inLosAngeles,CAandHonolulu,HI. Additionalfile2:FigureS1. Annualaverageconcentrationsofair pollutantsbetweenLosAngeles,CAandHonolulu,HIin1992 – 2008. Competinginterests Theauthorsdeclarethattheyhavenocompetinginterests. Authors ’ contributions XXconceivedofthestudy,andparticipatedinitsdesignandcoordination anddraftthemanuscript.SHparticipatedinthedesignofthestudy, performedthestatisticalanalysis,anddraftedthemanuscript.HKwas involvedindesignofthestudy,resultsinterpretation,manuscriptdrafting. HHperformedthestatisticalanalysis,andhelpedrevisethemanuscript.BC helpedinterpretresultsanddraftthemanuscript.CLhelpedanalyzethedata anddraftedthemanuscript.Allauthorsreadandapprovedthefinal manuscript. Acknowledgements TheprojectdescribedwassupportedbyGrantNumberK01ES019177from theNationalInstituteofEnvironmentalHealthSciences.Thecontentissolely theresponsibilityoftheauthorsanddoesnotnecessarilyrepresentthe officialviewsoftheNationalInstituteofEnvironmentalHealthSciencesor theNationalInstitutesofHealth.Ms.SandieHaandMs.ClaudiaLissakerwere alsofundedbyafellowshipfromtheUniversityofFloridaGraduateSchool. Therearenoothersourcesoffunding. Authordetails1DepartmentofEpidemiology,CollegeofPublicHealthandHealth ProfessionsandCollegeofMedicine,UniversityofFlorida,POBox100231, Gainesville,FL32610,USA.2DepartmentofEnvironmentalHealth,Schoolof PublicHealth,FudanUniversity,Shanghai,China.3DepartmentofBehavioral ScienceandCommunityHealth,CollegeofPublicHealthandHealth Professions,UniversityofFlorida,Gainesville,FL32610,USA. Received:4March2013Accepted:20August2013 Published:3September2013 References1.MurphyMLXJ,KochanekKD: Deaths:preliminarydatafor2010. NatlVital StatRep 2012, 60 (4):1 – 51. 2.JemalA,BrayF,CenterMM,FerlayJ,WardE,FormanD: Globalcancer statistics. CACancerJClin 2011, 61 (2):69 – 90. 3.SiegelR,WardE,BrawleyO,JemalA: Cancerstatistics,2011:theimpactof eliminatingsocioeconomicandracialdisparitiesonprematurecancer deaths. CACancerJClin 2011, 61 (4):212 – 236. 4.HuangW,ZhuT,PanX,HuM,LuSE,LinY,WangT,ZhangY,TangX: Air pollutionandautonomicandvasculardysfunctioninpatientswith cardiovasculardisease:interactionsofsystemicinflammation, overweight,andgender. AmJEpidemiol 2012, 176 (2):117 – 126. 5.HuangYF,MaoC: Exposuretoairpollutantsandmyocardialinfarction risk. JAMA 2012, 307 (22):2370.authorreply2370. 6.Raaschou-NielsenO,AndersenZJ,JensenSS,KetzelM,SorensenM,Hansen J,LoftS,TjonnelandA,OvervadK: Trafficairpollutionandmortalityfrom cardiovasculardiseaseandallcauses:aDanishcohortstudy. EnvironHealth 2012, 11: 60. 7.RichDQ,KipenHM,HuangW,WangG,WangY,ZhuP,Ohman-StricklandP, HuM,PhilippC,DiehlSR, etal : Associationbetweenchangesinair pollutionlevelsduringtheBeijingOlympicsandbiomarkersof inflammationandthrombosisinhealthyyoungadults. JAMA 2012, 307 (19):2068 – 2078. 8.DemetriouCA,Raaschou-NielsenO,LoftS,MollerP,VermeulenR,PalliD, Chadeau-HyamM,XunWW,VineisP: Biomarkersofambientairpollution andlungcancer:asystematicreview. OccupEnvironMed 2012, 69 (9):619 – 627. 9.TurnerMC,KrewskiD,PopeCA3rd,ChenY,GapsturSM,ThunMJ: Longtermambientfineparticulatematterairpollutionandlungcancerina largecohortofnever-smokers. AmJRespirCritCareMed 2011, 184 (12):1374 – 1381. 10.VermeulenR,RothmanN,LanQ: Coalcombustionandlungcancerriskin XuanWei:apossibleroleofsilica? LaMedicinadellavoro 2011, 102 (4):362 – 367. 11.Raaschou-NielsenO,AndersenZJ,HvidbergM,JensenSS,KetzelM, SorensenM,HansenJ,LoftS,OvervadK,TjonnelandA: Airpollutionfrom trafficandcancerincidence:aDanishcohortstudy. EnvironHealth 2011, 10: 67.12.BulatP,IvicML,JovanovicMB,PetrovicSD,MiljusD,TodorovicT,MiladinovMikovMM,BogdanovicM: Cancerincidenceinapopulationlivingneara petrochemicalfacilityandoilrefinery. CollAntropol 2011, 35 (2):377 – 383.Xu etal.BMCPublicHealth 2013, 13 :800 Page8of9 http://www.biomedcentral.com/1471-2458/13/800

PAGE 9

13.vanLeeuwenDM,PedersenM,HendriksenPJ,BoorsmaA,vanHerwijnen MH,GottschalkRW,Kirsch-VoldersM,KnudsenLE,SramRJ,BajakE, etal : Genomicanalysissuggestshighersusceptibilityofchildrentoair pollution. Carcinog 2008, 29 (5):977 – 983. 14.ZanobettiA,SchwartzJ: Arediabeticsmoresusceptibletothehealth effectsofairborneparticles? AmJRespirCritCareMed 2001, 164 (5):831 – 833. 15.SchuenemanJJ: Airpollutionproblemsandcontrolprogramsinthe UnitedStates. JAirPollutControlAssoc 1963, 13: 116 – 125. 16.USEnvironmentalProtectionAgency: CleanAirAct:progresscleaningtheair andimprovingpeople ’ shealth. Washington,DC:USEnvironmentalProtection Agency.http://www.epa.gov/air/caa/progress.html.Accessed8/22/2013. Lastupdated8/15/2003. 17.RebeccaB,DanielLC,RobertD,DouglasWD,MarkWF,WilliamL,Jonathan MS,FrankES,MarkU: Healtheffectsofoutdoorairpollution.Part2. Committeeoftheenvironmentalandoccupationalhealthassemblyof theAmericanthoracicsociety. AmJRespirCritCareMed 1996, 153 (2):477 – 498. 18.AmericanLungAssociation: StateoftheAirReport2012. WashingtonDC: AmericanLungAssociation;2012.Availableathttp://www.stateoftheair.org/ 2012/assets/state-of-the-air2012.pdf.Accessed8/22/2013. 19.PuettRC,HartJE,SuhH,MittlemanM,LadenF: Particulatematter exposures,mortality,andcardiovasculardiseaseinthehealth professional ’ sfollow-upstudy. EnvironHealthPerspect 2011, 119 (8):1130 – 1135. 20.JerrettM,BurnettRT,PopeCA3rd,ItoK,ThurstonG,KrewskiD,ShiY,Calle E,ThunM: Long-termozoneexposureandmortality. NEnglJMed 2009, 360 (11):1085 – 1095. 21.PopeCA3rd,BurnettRT,ThunMJ,CalleEE,KrewskiD,ItoK,ThurstonGD: Lungcancer,cardiopulmonarymortality,andlong-termexposuretofine particulateairpollution. JAMA 2002, 287 (9):1132 – 1141. 22.GrantWB: AirpollutioninrelationtoU.S.cancermortalityrates:an ecologicalstudy;likelyroleofcarbonaceousaerosolsandpolycyclic aromatichydrocarbons. AnticancerRes 2009, 29 (9):3537 – 3545. 23.LewtasJ: Airpollutioncombustionemissions:characterizationof causativeagentsandmechanismsassociatedwithcancer,reproductive, andcardiovasculareffects. MutationRes 2007, 636 (1 – 3):95 –133. 24.ZanobettiA,BindMA,SchwartzJ: Particulateairpollutionandsurvivalin aCOPDcohort. EnvironHealth 2008, 7: 48. 25.ZanobettiA,SchwartzJ: Ozoneandsurvivalinfourcohortswith potentiallypredisposingdiseases. AmJRespirCritCareMed 2011, 184 (7):836 – 841. 26.LadenF,NeasLM,DockeryDW,SchwartzJ: Associationoffineparticulate matterfromdifferentsourceswithdailymortalityinsixU.S.cities. EnvironHealthPerspect 2000, 108 (10):941 – 947. 27.RivaDR,MagalhaesCB,LopesAA,LancasT,MauadT,MalmO,ValencaSS, SaldivaPH,FaffeDS,ZinWA: Lowdoseoffineparticulatematter(PM2.5) caninduceacuteoxidativestress,inflammationandpulmonary impairmentinhealthymice. InhalToxicol 2011, 23 (5):257 – 267. 28.TagerIB,BalmesJ,LurmannF,NgoL,AlcornS,KunzliN: Chronicexposure toambientozoneandlungfunctioninyoungadults. Epidemiol 2005, 16 (6):751 – 759. 29.GrantWB,PeirisAN: DifferencesinvitaminDstatusmayaccountfor unexplaineddisparitiesincancersurvivalratesbetweenAfricanand whiteAmericans. Dermato-Endocrinol 2012, 4 (2):85 – 94. 30.PorojnicuAC,RobsahmTE,DahlbackA,BergJP,ChristianiD,BrulandOS, MoanJ: Seasonalandgeographicalvariationsinlungcancerprognosis inNorway.DoesvitaminDfromthesunplayarole? Lungcancer 2007, 55 (3):263 – 270. 31.TretliS,SchwartzGG,TorjesenPA,RobsahmTE: Serumlevelsof25hydroxyvitaminDandsurvivalinNorwegianpatientswithcancerof breast,colon,lung,andlymphoma:apopulation-basedstudy. Cancer CausesControl 2012, 23 (2):363 – 370. 32.ZhouW,SukR,LiuG,ParkS,NeubergDS,WainJC,LynchTJ,GiovannucciE, ChristianiDC: VitaminDisassociatedwithimprovedsurvivalinearlystagenon-smallcelllungcancerpatients. CancerEpidemiolBiomarkers Prev 2005, 14 (10):2303 – 2309. 33.ZhouW,HeistRS,LiuG,AsomaningK,NeubergDS,HollisBW,WainJC, LynchTJ,GiovannucciE,SuL, etal : Circulating25-hydroxyvitaminDlevels predictsurvivalinearly-stagenon-small-celllungcancerpatients. JClinOncol 2007, 25 (5):479 – 485. 34.PintoBM,TrunzoJJ: Healthbehaviorsduringandafteracancer diagnosis. Cancer 2005,104 (11Suppl):2614 – 2623. 35.BureauUSC: Summaryfile4:USCensus2000. USCensusBureau;2000.doi:10.1186/1471-2458-13-800 Citethisarticleas: Xu etal. : Healtheffectsofairpollutiononlengthof respiratorycancersurvival. BMCPublicHealth 2013 13 :800. Submit your next manuscript to BioMed Central and take full advantage of: € Convenient online submission € Thorough peer review € No space constraints or color “gure charges € Immediate publication on acceptance € Inclusion in PubMed, CAS, Scopus and Google Scholar € Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Xu etal.BMCPublicHealth 2013, 13 :800 Page9of9 http://www.biomedcentral.com/1471-2458/13/800