Lung Cancer Risk in White and Black Americans.docx

Lung Cancer Risk in White and Black Americans.docx

《Lung Cancer Risk in White and Black Americans.docx》由会员分享，可在线阅读，更多相关《Lung Cancer Risk in White and Black Americans.docx（20页珍藏版）》请在冰豆网上搜索。

LungCancerRiskinWhiteandBlackAmericans

LungCancerRiskinWhiteandBlackAmericans

STEVEND.STELLMAN,PHD,MPH,YUCHEN,MPH,JOSHUAE.MUSCAT,MPH,

MIRJANAV.DJORDJEVICPHD,JOHNP.RICHIE,JR.,PHD,PHILIPLAZARUS,PHD,

SETHTHOMPSON,PHD,NASSERALTORKI,MD,MARIANNEBERWICK,PHD,

MARCL.CITRON,MD,SUSANHARLAP,MD,TAJINDERB.KAUR,MD,

ALFREDI.NEUGUT,MD,PHD,SARAOLSON,PHD,JOHNM.TRAVALINE,MD,

PHILIPWITORSCH,MD,ANDZUO-FENGZHANG,MD,PHD

PURPOSE:

Totestwhetherdifferencesinsmoking-relatedlungcancerrisksinblacksandwhitescanex-plainwhylungcancerincidenceisgreaterinblackmalesthaninwhitemalesbutaboutequalinblackandwhitefemales,giventhatagreaterproportionofblacksaresmokers,butsmokefarfewercigarettesperdaythandowhites.

METHODS:

Ahospital-basedcase-controlstudywasconductedbetween1984and1998thatincludedinterviewswith1,710whitemaleand1,321whitefemalecasesofhistologicallyconfirmedlungcancer,254blackmaleand163blackfemalecases,and8,151controls.Relativeriskswereestimatedviaoddsra-tiosusinglogisticregression,adjustedforage,education,andbodymassindex.

RESULTS.Weconfirmedpriorreportsthatsmokingprevalenceishigherbutoveralldosageisloweramongblacks.OverallORsweresimilarforblacksandwhites,exceptamongtheheaviestsmokingmales（21cigarettesperdayor37.5pack–years）,inwhomORsforblackswereconsiderablygreaterthanforwhites.Long-termbenefitsofcessationweresimilarforwhiteandblackex-smokers.Smokersofmentholflavoredcigaretteswereatnogreaterriskforlungcancerthanweresmokersofunflavoredbrands.

CONCLUSIONS.Lungcancerrisksweresimilarforwhitesandblackswithsimilarsmokinghabits,ex-ceptpossiblyforblackswhowereveryheavysmokers;thissub-groupisunusualinthegeneralpopulationofAfricanAmericansmokers.Explanationsofracialdisparitiesinlungcancerriskmayneedtoaccountformodifyingfactorsincludingtypeofcigarette（yield,mentholation）,diet,occupation,andhostfactorssuchasabilitytometabolizemainstreamsmokecarcinogens.

AnnEpidemiol2003;13:

294–302（C）2003ElsevierScienceInc.Allrightsreserved.

KEYWORDS:

LungCancer,Cigarettes,Smoking,Risk,RacialDifferences,Dosage,Menthol.

LungcancerratesintheUSshowsubstantialunexplainedracialvariability.SEERincidencerateshavebeenreportedhigherforblacksthanwhitesineveryyearsince1973,withratedifferentialsbetween34%and67%.（1,2）.Thepreva-lenceofcigarettesmokinghasbeenconsiderablyhigherinblackthaninwhitemalessince1950.Itwasslightlyhigherinblackthaninwhitefemalesfrom1962until1992,afterwhichtherateshavebeennearlyequal（3–6）.Theseeming

FromtheAmericanHealthFoundation,OneDanaRoad,Valhalla,NY

10595（S.D.S.）;DepartmentofEpidemiology,MailmanSchoolofPublic

Health,ColumbiaUniversity,630West168thStreet,PH-18,NewYork,

NY10032（S.D.S.,Y.C.,A.I.N.）;DivisionofEpidemiology,American

HealthFoundation,OneDanaRoad,Valhalla,NY10595（J.E.M.）;To-

baccoControlResearchBranch,DivisionofCancerControlandPopula-

tionSciences,NationalCancerInstitute,6130ExecutiveBoulevard,EPN

4039,Rockville,MD20852（M.V.D.）;DivisionofNutritionalCarcinogen-

esis,AmericanHealthFoundation,OneDanaRoad,Valhalla,NY10595

（J.P.R.）;DivisionsofCancerControlandMol.Oncology,H.LeeMoffitt

CancerCenter,MRC-2E,12902MagnoliaDrive,Tampa,FL33612（P.L.）;

Bristol-MeyersSquibb,5ResearchParkway,Wallingford,CT06492（S.T.）;

DivisionofThoracicSurgery,NewYorkPresbyterianHospital,525East

68thStreet,NewYork,NY10021（N.A.）;DepartmentofEpidemiology&

Biostatistics,MemorialSloan-KetteringCancerCenter,1275YorkAvenue,

NewYork,NY10021（M.B.,S.O.）;ProHealth,Inc.,2800MarcusAvenue,

LakeSuccess,NY11042（M.L.C.）;Dept.ofObstetricsandGynecologyan

KaplanCancerCenter,NYUMedicalCenter—RoomNBV-9E-2,550Firs

Avenue,NewYork,NY10016（S.H.）;Dept.ofObstetricsandGynecology

TempleUniversityHospital,3401N.BroadStreet,Philadelphia,PA1914

（T.B.K.）;PulmonaryDivision,TempleUniversitySchoolofMedicine

3400N.BroadStreet,Philadelphia,PA19140（J.M.T.）;DivisionofClini

calPharmacology,DepartmentsofMedicineandPharmacology,George

townUniversityMedicalCenter,3900ReservoirRoadNW,Washington

DC20007（P.W.）;DepartmentofEpidemiology,UCLASchoolofPubli

Health,71-225CHS,Box951772,10833LeConteAvenue,LosAngeles

CA90095-1772.（Z-F.Z.）.

Addresscorrespondencesto:

StevenD.Stellman,Ph.D.M.P.H.,Dept

ofEpidemiology,MailmanSchoolofPublicHealth,630West168t

Street–PH-18,NewYork,NY10032,USA.Tel.:

212-305-4911.E-mail

sds91@columbia.edu

ReceivedMay9,2001;revisedMay9,2002;acceptedMay22,2002.

consistencybetweensmokingandlungcancerratesis

weakened,however,bysubstantialdifferencesinthenum-

berofcigarettessmokedperday（CPD）.In1991,29.2%of

adultblackscurrentlysmokedcigarettescomparedwith

25.5%ofwhites,butblackssmokedonaverage15.0CPD

whilewhitessmoked21.0CPD.（7）.Comparablediffer-

enceshavebeenreportedinmanyotherstudies（5,8–10）.

Fewanalyticstudieshaveaddressedthisanomaly.Schwartz

andSwanson（11）concludedthatracialdifferencesininci-

dencecouldbe“entirelyexplained”bysmokinghabits,

basedonanepidemiologicalstudyofover5500casesdiag-

nosedinDetroitareahospitalsin1984to1987.Neverthe-

less,theirconclusiondidnotapplytopersonsunder55yearsof

age,andwasbasedonlargenumbersofproxyinterviews,

whichmighthaveaffectedtheprecisionofreportedORs.

Althoughsmokingistheoverwhelmingcauseoflung

cancer,otherhostandenvironmentalfactorsmayalsomod-

ifyrisk.Modifyingfactorsthathavebeenstudiedinclude

diet（12,13）,geneticpolymorphismsinmetabolizinggenes

（14–18）aswellasmoregeneralfamilialfactors（19,20）,

metabolismdifferences（21,22）,occupation（23）,andnon-

biologicalfactorssuchassocialclass（24）andeducation

（25）.Thehypothesisthatthestrongpreferenceformen-

tholflavoredcigarettesamongblacksmokersmayalso

partlyexplainriskdifferenceshasledtoconflictingresults

amonginvestigators,withnoassociationreportedbyour

group（26）,andapositiveassociationreportedformenbut

notwomenbySidneyetal.（27）.

Tobetterdelineatesmoking-relatedrisksforlungcancer

betweenracialgroups,itisimportanttomakedirectassess-

mentsofriskinrelationtosmokinghabitsasanessential

backdropforinterpretingtheimpactofotherriskfactors,

includingthoseobservedinmetabolicandmolecularstud-

ies.Toaddresstheseissuesweexaminedsmokinghabits

andlungcancerriskinblackandwhiteAmericans.

METHODS

Between1984and1998theAmericanHealthFoundation

performedahospital-basedcase-controlstudyinthethree

majorNewYorkCitycancercentersplusotherhospitalsin

NewYork,Philadelphia,WashingtonDC,andotherUS

cities（seeAcknowledgments）.Onlyincidentcaseswerese-

lected,definedaspersonsdiagnosedwithlungcancerfor

thefirsttimeduringthe12monthsprecedinginterview

（mostwithin2months）.Allcaseswereconfirmedbyhisto-

pathology.Adenocarcinomasweremorecommoninwomen

（46%ofcases）thaninmen（37%）,butdifferedlittleby

race.Controlswereselectedfromthedailyadmissionros-

tersandfrequencymatchedtocasesonthebasisofsex,age

（5y）,hospital,andyearofinterview.Eligiblecontroldi-

agnosesexcludedtobacco-relateddiseasessuchascoronary

heartdisease,stroke,peripheralvasculardisease,chronic

obstructivepulmonarydisease,gastriculcer,cirrhosisofthe

liver,andcancersofthemouth,larynx,esophagus,bladder,

kidney,pancreas,orliver（28）.Controlpatientsforstudies

ofothertobacco-relatedcancersbesideslungwerebeingin-

terviewedatthesametimeascases,sothatalargecontrol

poolofpatientswithnon-tobacco-relateddiseaseswas

available.Approximatelyhalfofmalecontrolshadbenign

ormalignantdiseasesincludingbenignprostatichypertro-

phy（9%）,prostatecancer（8%）,colon-rectumcancer

（11%）;boneandjointdiseases（5%）;kidneystones,ne-

phritis,andotherkidneydiseases（8%）;abscesses（2%）;

sprains,strains,andfracturesincludinghip,andawidevari-

etyofothernon-malignantconditionsrequiringhospital-

ization.Fortypercentoffemalecontrolshadcancerswhich

includedbreast（15%）,colon-rectum（7%）,ovary（4%）,

connectivetissue（2%）,andmelanoma（2%）;otherfemale

controlswerehospitalizedforosteoarthritis（5%）,fractures

includinghip（5%）,genitalprolapse（2%）,abscesses（2%）

andawidevarietyofothernon-malignantconditions.Af-

terprovidingwritteninformedconsentusingaformap-

provedbytheInstitutionalReviewBoardofeachhospital,

everysubjectwasinterviewedbyanAHF-trainedinter-

viewerusingastructuredquestionnairethatelicitedinfor-

mationondemographicvariables,smokinghistory,and

otherpossibleriskfactors.Approximately85%ofeligible

patientswhowereapproachedagreedtobeinterviewed.

The15-yearaccrualintervalwaschosenbecauseitin-

cludedlargenumbersofblackandwhitepatientswhowere

recruitedwithauniformprotocolandinterviewedunder

similarcircumstances.Thepresentanalysisoverlapsand

extendsearlierreports（26,29）thatincludedpatientsinter-

viewedbetween1977and1991.Thisisasignificantexten-

sion,sinceitmakesuseofdataobtainedviafacetoface

interviewswith11,599patients（4192interviewedafter

1991）,ofwhom3448werecasesand8151werecontrols.It

includes417blackcases,254ofwhom