wave雅虎社区

613HypertensResVol.
30(2007)No.
7p.
613-619OriginalArticleTheAssociationbetweenAnkle-BrachialIndexandCardiovascularorAll-CauseMortalityinMetabolicSyndromeofElderlyChineseYuanxiXU1),*,JueLI1),*,YingyiLUO1),YongquanWU1),LiqiangZHENG1),JinmingYU1),JunMA1),JianyunGU1),andDayiHU1)Theankle-brachialindex(ABI)isanon-invasive,reliablemeasurementoflower-extremityischemia.
AlowABIisassociatedwithincreasedriskofcoronaryheartdisease,strokeanddeath.
However,therelationshipbetweenABIandall-causemortalityorcardiovasculardisease(CVD)mortalityinpatientswithmetabolicsyndrome(MetS)hasnotbeenwellstudied.
Accordingly,wehereinvestigatedtheassociationbetweenABIandall-causeandCVDmortalityinanelderlyChinesepopulationwithMetS.
Atotalof2,274MetSpatientsdiagnosedunderthecriteriaproposedbytheInternationalDiabetesFederationweredividedintotwogroupsbasedonrepeatedABImeasurementoveraperiodof13.
6months:ABI≤0.
9(n=525)andABI0.
91–1.
4(n=1,749).
Eachofthebaselinecharacteristicsofage,systolicbloodpressure,diabetesmellitusmor-bidity,andsmokinghistoryweresignificantlydifferentbetweenthetwogroups(p102cminmenand>88cminwomen).
MetSwasdiagnosedwhentwoormoreofthefollowingmetabolicabnormalitieswerepresent:SBP≥130mmHgordiastolicbloodpressure(DBP)≥85mmHg,hypertriglyceridemia(serumtriglycerides[TG]≥1.
7mmol/L),lowhigh-densitylipoproteincholesterol([HDL-C]≤0.
9mmol/Linmenand≤1.
1mmol/Linwomen)andhighfastingglucose([FSG]≥5.
6mmol/L).
DiagnosisofessentialhypertensionanddiabetesmellituswerebasedonWHO/ISHguidelines(1999)andADAcriteria(1997),respectively(10).
Table1.
BaselineCharacteristicsoftheStudyPopulationBaselinecharacteristicsABI(n=2,274)pvalue≤0.
90(n=528)0.
91–1.
40(n=1,746)Age(years)71.
4±10.
365.
4±11.
20.
000Men(n(%))50.
854.
80.
107BMI(kg/m2)25.
6±3.
624.
7±3.
60.
626HighBP(n(%))82.
672.
60.
000SBP(mmHg)144±25139±220.
000DBP(mmHg)81.
3±1381.
0±130.
677Dislipidemia(n(%))47.
845.
00.
296TC(mmol/L)4.
65±1.
154.
64±1.
150.
807TG(mmol/L)1.
76±1.
251.
73±1.
090.
650LDL-C(mmol/L)2.
75±0.
912.
79±1.
780.
623HDL-C(mmol/L)1.
15±0.
361.
18±0.
410.
181Serumcreatinine(μmol/L)107.
59±12.
28101.
74±9.
100.
068Diabetes(n(%))42.
231.
70.
000Glu(mmol/L)6.
25±2.
516.
23±2.
640.
859Smokers(n(%))43.
037.
60.
025All-causemortality(%)11.
45.
20.
000CVDmortality(%)5.
72.
00.
000ABI,ankle-brachialindex;BMI,bodymassindex;BP,bloodpressure;SBP,systolicbloodpressure;DBP,diastolicbloodpressure;TC,totalcholesterol;TG,triglycerides;LDL-C,low-densitylipoproteincholesterol;HDL-C,high-densitylipoproteincholesterol;Glu,glu-cose;CVD,cardiovasculardisease.
Xuetal:ABIandCardiovascularorAll-CauseMortality615However,becausetheoriginalcut-offforabdominalobe-sityintheNCEPdefinitionhaspreviouslybeenshowntobeinappropriateforAsianpopulations(11,12),andthenumberofsubjectsinP.
R.
Chinawhometthesecriteriawasextremelylow,thecut-offlimitofwaistcircumferencewasadjustedtothecriteriaof≥94cminmenand≥80cminwomen,whichwerebasedontheriskofobesity-relateddis-ordersinalarge-scalestudyofBeijingcitizens.
ThesecriteriawerealsoconsideredasthestandardofMetSclassificationinChinesepatientsspecifically.
Duringthisexamination,participantswereclassifiedinaccordancewiththepresenceorabsenceofsixpreexistingCVDs:acutemyocardialinfarction(AMI),angina,chronicheartfailure,stroke,transientischemicattack,andintermit-tentclaudication.
ParticipantswithanyofthesesixconditionswereclassifiedashavingprevalentCVD.
ABIwascalculatedastheratiooftheankleSBPtothebra-chialSBPonthehomolateralside.
ABIwasdiagnosedaccordingtothelatestcriteria(USA,2005):1.
00–1.
40inbothlegswasconsiderednormal;0.
91–0.
99wasdefinedasbor-derlinefordevelopmentintoperipheralarterydisease(PAD);restABIbetween0.
41–0.
90inatleastonelegwasdefinedasdecreasedandwasconsideredanindependentpredictorofPADwithonesiteofstenosisinatleastoneleg;andABIof≤0.
40stronglyindicatedmorethanonesiteofstenosisinatleastoneleg(13–15).
StudyDesignandMethodsThisisaretrospectivestudy.
Atthetimeofenrollment(November2004),allpatientspassedastandardizedprotocol,includingahealthquestionnaireoncurrentmedicationuse,pastmedicalhistory,familialvascularhistoryandatheroscle-roticriskfactors.
Height,bodyweight,bodymassindex(BMI),waistcircumferenceandbloodpressureweremea-sured.
Fastingbloodwassampledtodeterminelipidlevels(TG,totalcholesterol[TC],HDL-Candlow-densitylipopro-teincholesterol[LDL-C])andserumglucose.
Thesamemethodswereusedforthelaboratorytestmeasurementsatallinstitutes.
TherestingABIwasmeasuredwiththesubjectinasupinepositionwitha5MHzcontinuouswaveDopplerProbe(CBA1304),andtheinflatablecuffofthesphygmomanome-terwas10cminwidthand40cminlength(Sanofi-AventisCorp.
,Ltd.
,Beijing,P.
R.
China).
ThevalueofthehighestSBPmeasuredattheanklewasdividedbythehighestSBPmeasuredinbotharms.
Theratio(ABI)wascalculatedforbothlegs.
Afterafollow-upperiodof13.
6monthsendinginApril2006,almostallparticipantswereinvestigatedabouttheCVDmortalityandall-causemortalityrespectivelybycon-tactingwiththemselvesortheirrelatives.
IdentificationofDeathsDuetoAll-CauseorCVDMortalityDeathswereidentifiedthroughtherecordsoftheeightpartic-ipatinguniversityhospitalsorbycontactwithparticipantsandtheirfamilies.
Thecauseofdeathwasfurtherinvestigatedusingmedicalrecordsandinformantinterviews.
AllmaterialswerereviewedindependentlybyphysiciansparticipatingintheABIcohortstudytoconfirmthecauseofdeath.
Fig.
1.
PercentagesofCVDmortalityandall-causemortalityinthedifferentABIgroups.
a–c:p0.
05),mainlybecausealltheparticipantsmetthediagnosticcriteriaforMetSandmostofthemwereover-weight.
Similarly,therewerenosignificantdifferencesinfastingbloodglucose(6.
252±2.
508mmol/Land6.
228±2.
642mmol/LfortheabnormalABIgroupandtheborderline/normalABIgroup,respectively)ordyslipidmor-bidity,sincemostofthepatientsweexaminedwereundergo-ingglucoseand/orserumlipiddown-regulationtherapies.
However,themeanage(71.
36±10.
25intheabnormalABIgroupand65.
35±11.
18inthenormalABIgroup)andper-centageofsmokers(43.
0%and37.
6%)weresignificantlydif-ferentbetweenthetwoABIgroups.
Bloodpressure,especiallySBP(143.
65±25.
162mmHgand138.
86±21.
987mmHg),anddiabetesmorbidity(42.
2%,31.
7%)alsoweresignificantlydifferentbetweenthetwogroups(p<0.
01)(Table1).
OutcomesofInterestAsshowninTable1,theratesofbothall-causemortality(dystrophy,tumor,etc.
)andCVDmortality(AMI,stroke,heartfailure,andhemopericardium)weresignificantlydiffer-entbetweenthegroupwithanABI≤0.
90(11.
4%and5.
7%)andthegroupwithanABIof0.
91–1.
40(5.
2%and2.
0%),respectively.
Thenumbersofdeathsduetoall-causemortal-ityinthesetwogroupswere60and91,whilethenumberofCVDmortalitieswere30and35,respectively(AMI,heartfailure,stroke,andhemopericardiumwerethecausesofdeathin12,7,7,and4membersofthegroupwithABI≤0.
90andin10,9,10,and6membersofthegroupwithanABIof0.
91–1.
40).
AsshowninFig.
1,therewasacorrelationbetweentheABIandthepercentagesofbothall-causemortalityandCVDmortality.
Thatis,all-causemortalityandCVDmortalitybothdecreasedgraduallywhenABIchangedfrombelow0.
40toover0.
90(from21.
1to5.
0andfrom10.
5to1.
7,respec-tively).
Astoall-causemortalityandCVDmortality,theresultsadjustedage,gender,historyofhypertension,diabetesandsmokingandfoundalmostthesametrendasinFig.
1(Figs.
2,3).
Themortalityrategraduallydecreasedfromseri-ousPAD(ABI≤0.
40)toarelativelynormalABI(1.
00–1.
40).
Fig.
2.
Relativerisk(RR)ofall-causemortality(95%CI)(adjustedforage,gender,historyofhypertension,diabetesmellitusandsmoking).
65.
6495Q4u12.
32.
65n::2.
382n::1,,~-r71921.
24311.
11.
0140.
5980:::;o.
4o0.
41-0.
900.
91-0.
991.
00-1.
40(n=38)(n=490)(n=305)(n=l441)AB!
GroupXuetal:ABIandCardiovascularorAll-CauseMortality617DiscussionPeripheralarterydiseasecommonlyresultsfromprogressivenarrowingofthearteriesinthelowerextremitiesduetoath-erosclerosis.
ItcanbedeterminedwithhighprecisionusingtheABI,asimple,non-invasivemeasurement(16).
TheresultsoftheNationalHealthandNutritionExaminationSur-vey(NHANES,1999–2000)demonstratedthatthereisahighprevalenceofbothtraditionalandnon-traditionalcardiovas-cularriskfactorsamongpersonswithPAD.
Morethan95%ofindividualswithPADhaveatleastonetraditionalcardio-vascularriskfactor,andthemajoritieshavemultiplefactors.
Thesefindingshighlighttheimportanceofaggressiveprog-nosisandrelatedCVDorall-causemortalityinpersonswithPADorabnormalABIandinthosewithsubnormalornormalABI(17).
Wildetal.
(18)performedasurveyof1,467menandwomenaged35–74,andfoundthat25%ofthestudypop-ulationhadMetS.
Duringthefollow-upperiod,226ofthe1,467participantsdiedfromCVDand462fromnonfatalcar-diovascularevents.
ThisstudyindicatedthatlowABIismoreprevalentamongpeoplewithMetSthanamongthosewithoutit.
TheMetSisamajorpublichealthchallengeworldwide.
Itisnotbenign;itisassociatedwithasubstantiallyelevatedriskoftype2diabetes(5-fold)andofCVD(2-to3-fold),anditsincreasingprevalencecouldpossiblyreversethegainsmadethroughtherecentdeclineinCVDmortality.
OurcurrentstudyclarifiedtherelationshipbetweenlowABIandmortalityinpatientswithMetS.
Wechoseapopula-tionofpatientswithMetSbecausethecomplicationsofMetSvarywidely,andcanincludePAD,particularlyindiabeticpatients.
Across-sectionalstudyperformedbyWaltersetal.
(19)foundan8.
7%prevalenceofPADamongpatientswithtype2diabetesanda23.
5%prevalenceoftype2diabetes.
ItisnotyetcommonpracticetoroutinelyscreenforthediseaseforPADinpatientswithdiabetes,andthehigherprevalenceofinsidioussymptomsexistedinthe"borderline"caseswithanABI0.
80to<0.
90(20).
However,thestandardofABIclassificationwaschangedto0.
91–0.
99intheUSin2005.
Inthepresentstudy,whenMetSpatients,whoseABIwasbelowtheborderlineof0.
91,twomainCVDriskfactors(ageandsmokinghistory),becamestrongcharacteristicsin2groups.
OurresultsfullysupportthefindingsofChoietal.
(21)thatABIwassignificantlyassociatedwiththefeaturesofcardio-vascularriskfactorsandthereforeexacerbatedtheformationofCVDorevendeathrelatedtoit.
Inthepresentstudy,wealsofoundthatinpatientswiththeABI≤0.
90,thepercentageofpatientswithdiabetesandhypertension(highSBP)werehigherthaninthegroupwithanABIof0.
91–1.
40.
ThisresultindicatedthatthecontributionoftheclusteredcomponentsofMetSappearedtobeadditive,withthesubjectshavingmoreCVDriskfactorsshowingasubstantiallylowerABIthanthosewithfewerriskfactorsinbothsexes(22,23).
IndifferentABIgroups,thesametrendappearedwhenABIwasbelow0.
91andanextremelyhighmortalitypercent-ageoccurredwhenABIwasbelow0.
41(groupwithABI≤0.
40).
ThereasonwhythelowerABIwasrelatedtohigherCVDorall-causemortalitymaybethepresenceofatherosclerosisthatledtoarterialstiffening.
Farraretal.
(24)demonstratedinmonkeysthatanatherogenicdietincreasedABIandaorticintimalarea,whileanatherosclerosisregres-siondietdecreasedbothparameters.
Arteriallesionsthencommencedasfattystreaks,progressedtoraisedlesionsandwerecomplicatedbyulceration,calcificationorevenhemor-rhage,whichledtoseverediseasessuchasstroke,PAD,thrombosis,etc.
(25).
Inaddition,theaorticwallundergoesFig.
3.
Relativerisk(RR)ofCVDmortality(95%CI)(adjustedforage,gender,historyofhypertension,diabetesmellitusandsmoking).
1210.
860108u#.
Ln68De:De:4j4.
0923.
645j3.
28022.
3041.
224)298I1.
559o.
7410:,:;o.
4o0.
41-0.
900.
91-0.
991.
00-1.
40(n=38)(n=490)(n=305)(n=1441)AB!
Group618HypertensResVol.
30,No.
7(2007)progressiveaccumulationofcalciumintheelastin-richlayerofthemediaduringaging,especiallyindiabeticpatients,andresultsinmedialarterialcalcification.
ThereforeABI,whichindicatedthedegreeofaorticcalcification,isapredictorofsubsequentcardiovascularmorbidityandmortality(26).
Recently,theFraminghamHeartStudy,aprospectivestudyspanningmorethan20years,reportedthattheseverityofaor-ticcalcificationwascorrelatedwithsubsequentCVDanddeath(27).
Ourstudyalsohadsomepotentiallimitations.
First,wewereunabletodeterminewhethercardiovascularriskfactorswerecasuallyrelatedtotheABIvaluesbecausethisstudywascross-sectional.
Second,wemighthaveunderestimatedtheprevalenceoflowABIowingtoourrelianceonvolun-teers.
Third,ABIisanindirectmakerofincreasedarterialstiffnessordecreasedarterialcomplianceandwecouldnotdeterminetherelativeinfluenceofarterialwallremodelingontherelationshipbetweencardiovascularriskfactorsonarte-rialstiffness.
Fourth,theresponserateinourstudywasrela-tivelylowcomparedwiththatofsimilarsurveys.
Finally,wewerenotabletoprovidetheprevalenceofpatientswithahis-toryofcoronaryarterydiseaseandthereforecouldnotincludethisinformationasaconfoundingfactorintheCoxregressionanalysis(28,29).
Insummary,weresearchedtheassociationbetweenlowABIandtheclusteringofMetScomponentsinapopulationofelderlyChinese.
OurfindingsindicatedthatlowABI(≤0.
90),especiallyABI≤0.
40,maybeausefulmarkerofCVDandpredictorofCVDorall-causemortality.
Thespeci-ficityoflowABItopredictfuturecardiovascularoutcomeishigh,butitssensitivityislow.
ABIshouldbetakenintocon-siderationaspartofthevascularriskassessmentamongselectedindividuals.
Itisasurrogateend-pointinepidemio-logicalstudiesandmayactasatoolforevaluatingCVDriskinclinicalpractice.
BecausePADisanunder-diagnosedandunder-treatedcon-ditioninP.
R.
China,ABImeasurementshouldbearoutinepartoftheclinicalevaluationofhighriskpatients.
Athero-scleroticriskfactorssuchasdiabetes,hypercholesterolemiaandhypertensioncanandshouldbetreatedadequately,andsmokingshouldbestronglydiscouraged.
AcknowledgementsOurgreatestthanksgototheparticipantsinthestudy,andtothedoctors,nursesandadministrativestaffinthehospitalswhoassistedwiththisundertaking.
Wewouldalsoliketoacknowl-edgetheDepartmentofPhylaxiologyofTongjiUniversityandtheSanofi-AventisCorporation,Ltd.
ofP.
R.
Chinafortheirsup-port.
References1.
LijmerJG,HuninkMG,vandenDungenJJ,etal:ROCanalysisofnoninvasivetestsforperipheralarterialdisease.
UltrasoundMedBiol1996;22:391–398.
2.
CurbJD,MasakiK,RodriguezBL,etal:Peripheralarterydiseaseandcardiovascularriskfactorsintheelderly.
TheHonoluluHeartProgram.
ArterioscleroThrombVascBiol1996;16:1495–1500.
3.
StoffersHE,RinkensPE,KesterAD,etal:Theprevalenceofasymptomaticandunrecognizedperipheralarterialocclusivedisease.
IntJEpidimiol1996;25:282–290.
4.
MartynCN,GaleCR,JespersenS,etal:Impairedfetalgrowthandatherosclerosisofcarotidandperipheralarter-ies.
Lancet1998;352:173–178.
5.
TsaiAW,FolsomAR,RosamondWD,etal:Ankle-bra-chialindexand7-yearischemicstrokeincidence:theARICstudy.
Stroke2001;32:1721–1724.
6.
ImanishiR,SetoS,TodaG,etal:Highbrachial-anklepulsewavevelocityisanindependentpredictorofthepresenceofcoronaryarterydiseaseinmen.
HypertensRes2004;27:71–78.
7.
WeatherleyBD,ChamblessLE,HeissG,etal:Thereliabil-ityoftheankle-brachialindexintheatherosclerosisriskincommunities(ARIC)studyandtheNHLBIfamilyheartstudy(FHS).
BMCCardiovascDisord2006;21:7.
8.
KweonSS,ShinMH,ParkKS,etal:Distributionoftheankle-brachialindexandassociatedcardiovascularriskfac-torsinapopulationofmiddle-agedandelderlyKoreans.
JKoreanMedSci2005;20:373–378.
9.
LiS,ChenW,SrinivasanSR,etal:Influenceofmetabolicsyndromeonarterialstiffnessanditsage-relatedchangeinyoungadults:theBogalusaHeartStudy.
Atherosclerosis2005;180:349–354.
10.
ExecutivesummaryofthethirdreportoftheNationalCho-lesterolEducationProgram(NCEP)ExpertPanelonDetec-tion,Evaluation,andTreatmentofHighBloodCholesterolinAdults(AdultTreatmentPanelIII).
JAMA2001;285:2486–2497.
11.
TanCE,MaS,WaiD,etal:CanweapplytheNationalCholesterolEducationProgramAdultTreatmentPaneldefi-nitionofthemetabolicsyndrometoAsiansDiabetesCare2004;27:1182–1186.
12.
JorgensenME,Borch-JohnsenK:Themetabolicsyndrome.
IsoneglobaldefinitionpossibleDiabetMed2004;21:1064–1065.
13.
LengGC,FowkesFG,LeeAJ,etal:Useofanklebrachialpressureindextopredictcardiovasculareventsanddeath:acohortstudy.
BMJ1996;313:1440–1444.
14.
MogtMT,CauleyJA,NewmanAB,etal:Decreasedankle/armbloodpressureindexandmortalityinelderlywomen.
JAMA1993;270:465–469.
15.
SacksD:TheTransatlanticInter-SocietyConsensus(TASC)onthemanagementofperipheralarterialdisease.
JVascIntervRadiol2003;14:S351.
16.
FeigelsonHS,CriquiMH,FronekA,etal:Screeningforperipheralarterialdisease:thesensitivity,specificity,andpredictivevalueofnoninvasivetestsinadefinedpopula-tion.
AmJEpidemiol1994;140:526–534.
17.
SelvinE,ErlingerTP:PrevalenceofandriskfactorsforperipheralarterialdiseaseintheUnitedStates:resultsfromthenationalhealthandnutritionexaminationsurvey,19992000.
Circulation2004;110:738–743.
18.
WildSH,ByrneCD,SmithFB,etal:Lowankle-brachialpressureindexpredictsincreasedriskofcardiovasculardis-Xuetal:ABIandCardiovascularorAll-CauseMortality619easeindependentofthemetabolicsyndromeandconven-tionalcardiovascularriskfactorintheEdinburghArteryStudy.
DiabetesCare2006;29:637–642.
19.
WaltersDP,GatlingW,MulleeMA,etal:Theprevalence,detection,andepidemiologicalcorrelatesofperipheralvas-culardisease:acomparisonofdiabeticandnon-diabeticsubjectsIanEnglishcommunity.
DiabetMed1992;9:710–715.
20.
NewmanAB,SiscovickDS,ManolioTA,etal:Ankle-bra-chialindexasamarkerofatherosclerosisintheCardiovas-cularHealthStudy.
Circulation1993;88:837–845.
21.
ChoiKM,LeeKW,SeoJA,etal:Relationshipbetweenbrachial-anklepulsewavevelocityandcardiovascularriskfactorsofthemetabolicsyndrome.
DiabetesResClinPract2004;66:57–61.
22.
NakanishiN,ShiraishiT,WadaM:Brachial-anklepulsewavevelocityandmetabolicsyndromeinaJapanesepopu-lation:theMinohStudy.
HypertensRes2005;28:125–131.
23.
HasimuB,LiJ,NakayamaT,etal:AnklebrachialindexasamarkerofatherosclerosisinChinesepatientswithhighcardiovascularrisk.
HypertensRes2006;29:23–28.
24.
FarrarDJ,BondMG,RileyWA,etal:Anatomiccorrelatesofaorticpulsewavevelocityandcarotidarteryelasticityduringatherosclerosisprogressionandregressioninmon-keys.
Circulation1991;83:1754–1763.
25.
SolbergLA,StrongJP:Riskfactorsandatheroscleroticlesions:areviewofautopsystudies.
Arteriosclerosis1983;3:187–198.
26.
WittemanJC,KokFJ,vanSaaseJL,etal:Aorticcalcifica-tionasapredictorofcardiovascularmorality.
Lancet1986;2:1120–1122.
27.
WilsonPW,KauppilaLI,O'DonnellCJ,etal:Abdominalaorticcalcificdepositsareanimportantpredictorofvascu-larmorbidityandmortality.
Circulation2001;103:1529–1534.
28.
MeijerWT,HoesAW,RutgersD,etal:Peripheralarterialdiseaseintheelderly:theRotterdamStudy.
ArteriosclerThrombVascBiol1998;18:185–192.
29.
SaijoY,Yoshioka,FukuiT,etal:Metabolicsyndrome,C-reactiveproteinandincreasedarterialstiffnessinJapanesesubjects.
HypertensRes2006;29:589–596.