Coronary
calcium
score,
albuminuria
and
inflammatory
markers
in
type
2
diabetic
patients:
Associations
and
prognostic
implications
Akın
Dayan
a,
Burcu
Narin
b,
Murat
Biteker
c,
Sukru
Aksoy
d,
Hakan
Fotbolcu
e,
Dursun
Duman
f,*
aHaydarpas¸aNumuneEducationandResearchHospital,DepartmentofInternalMedicine,Istanbul,Turkey b
Haydarpas¸aNumuneEducationandResearchHospital,DepartmentofRadiology,Istanbul,Turkey
cHaydarpas¸aNumuneEducationandResearchHospital,DepartmentofCardiology,Istanbul,Turkey
dSiyamiErsekThoracicandCardiovascularSurgeryCenter,TrainingandResearchHospital,DepartmentofCardiology,Istanbul,Turkey eCentralHospital,DepartmentofCardiology,Istanbul,Turkey
fMedipolUniversitySchoolofMedicine,DepartmentofCardiology,Istanbul,Turkey
1.
Introduction
Cardiovasculardiseaseisthemajorcauseofdeathin65–70% ofpatientswithdiabetes[1].Diabetesnephropathyoccursin 20–40% ofpeople with diabetesand isone of the leading causesofend-stagerenaldisease[2].Microalbuminuriaand
proteinuria were independently associated with risk of cardiovascular diseaseanddeath [3]. Albuminuriareflects generalizedvasculardamageanditiscloselyassociatedwith inflammationwhichunderlies allstages ofatherosclerotic lesionformation,includingearlyatherogenesis[4].
Itisknownthatdiabeticpatientsarelesssymptomaticin terms of coronary artery disease (CAD), and when CAD
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received16March2012 Receivedinrevisedform 9April2012
Accepted16April2012 Publishedonline15May2012 Keywords:
Diabetesmellitus Albuminuria
Coronarycalciumscore Inflammatorymarker
a
b
s
t
r
a
c
t
Aims:Toinvestigatetherelationshipofcoronaryarterycalcium(CAC)scoreswithcommon carotidarteryintimamediathickness(CCA-IMT),albuminuriaandinflammatoryfactorsin type2diabetes.
Methodsandresults: 128asymptomatictype2diabeticpatients,withatleastone cardiovas-cularriskfactorinadditiontodiabetes,wereincludedinthestudy.CACscores,carotid arteriesplaqueformationandCCA-IMTwereassessed.Thepatientswerefollowedfora meanperiodof36.63.3months.Linearregressionanalysisidentifiedthelogarithmically transformed(Ln)albuminuria(b=0.32,P=0.007),age(b=0.04,P=0.001)andtheuricacid (b=0.13,P=0.04)asindependentdeterminantsoftheCACscore.Duringfollow-upperiod, cardiovasculareventsoccurredin18outof46patientswithCACscore100comparedwith 5outof82patientswithCACscore<100(logrank,P<0.0001).MultivariateCoxproportional hazardsanalysisidentifiedLnCACscore(P<0.0001),LnAlbuminuria(P=0.01)anduricacid (P=0.03)asindependentpredictorsforcardiovascularevents.
Conclusions:There was a significant relationship between CACscore, albuminuria and inflammationinpatientswithtype2diabetes.LnCACscoretogetherwithLnAlbuminuria anduricacidwereidentifiedasindependentpredictorsofcardiovasculareventsinthese patients.
#2012ElsevierIrelandLtd.Allrightsreserved.
*Correspondingauthor.Tel.:+905324234190;fax:+902165452006. E-mailaddress:drduman@excite.com(D.Duman).
ContentsavailableatSciverseScienceDirect
Diabetes
Research
and
Clinical
Practice
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / d i a b r e s
0168-8227/$–seefrontmatter#2012ElsevierIrelandLtd.Allrightsreserved. http://dx.doi.org/10.1016/j.diabres.2012.04.012
appearsclinically,diabetesresultsinaworseprognosis[1].It is important to investigate new risk factors in addition to knownfactorstoaidinearlydiagnosisofpatients,especially asymptomatic ones. Common carotid artery intima media thickness(CCA-IMT)isasurrogatemarkerofatherosclerosis andassociatedwithcardiovascularevents[5].CCAplaqueis easilydetectedandmaybemorecloselyrelatedtosystemic atherosclerosis. There is a relation between albuminuria, CCA-IMT, CCA plaque and atherosclerosis [6]. Recently, coronary arterycalcium (CAC)imaging, asimple,fast, and non-invasivemethodfordetectingsubclinicalcoronaryartery disease,hasbecome apopular method [7]. CAC scorewas demonstratedtobeapredictorofCADinnumerousstudieson generalpopulation[3].Furthermore,therearestudies dem-onstratingthatelevatedlevelsofinflammatorymarkerssuch asIL-6,TNF-aappeartobeassociatedwithagreaternumberof diseasedarteriesand,consequently,theseverityofCAD[8].
In this study, we investigated the relationship of CAC scores with excreted urinary albumin, CCA-IMT, plaque formationandinflammatoryfactors inhighrisktype2DM patientswithnoprevioushistoryofheartdisease.
2.
Methods
Type2diabetesmellituspatients(40–80yearsofage)followed in our hospital between 2008 and 2009, with no previous historyofcardiovasculardiseaseand nocomplaints,butat leastonecardiovascularriskfactoradditiontodiabeteswere includedinthestudy.Patientswerediagnosedwithtype2 diabetesmellitusaccordingtotheAmericanDiabetes Associ-ationcriteriarevisedin2009[9].Exclusioncriteriawereuseof drugs that impair metabolic control (such as cortisol or immunosuppressive treatment), chronic liver disease, any infectious disease, malignancy, hyperthyroidism, hypothy-roidism,familialhypercholesterolemia,diagnosisofdiabetes beforetheageof35years,andpresenceofconnectivetissue diseasesandplannedoractualpregnancy,glomerular filtra-tion rate <30mL/min/1.73m2 according to KDOQI Clinical PracticeGuidelines forChronic KidneyDisease: Evaluation, Classification, and Stratification [10]. Contraindications to electronbeamcomputedtomographyincludedknownallergy toiodinecontrastmedia(supra)ventriculararrhythmiasand renalinsufficiency(serumcreatinine>140mmol/l).
ApprovalwasobtainedfromtheEthicalAdvisory Commit-tee of clinical researches and informed consents were obtainedfromeachpatient.
Medicalhistorywastaken, includingpresenceof cardio-vascular risk factors and duration of diabetes. Physical examinationwasperformed.Height,weight,bodymassindex (BMI)andbloodpressuredatawererecorded.Fastingblood and urine samples were obtained. Glucose, glycosylated hemoglobin(HbA1c),creatinine,fibrinogen,uricacid,blood albumin,urinaryalbuminwereincludedintheanalysis. TNF-a and interleukin-6 levels were measured (Immulite 1000 ImmunoassayAnalyzer,SiemensMedicalSolutions Diagnos-tics,NewJersey,USA).
Multislice computed tomography (MSCT) examinations wereperformedwitha64-sliceToshibamultisliceAquilion64 system (Toshiba Medical Systems, Tokyo, Japan). In all
patients,anoncontrastenhancedscanwasperformedbefore MSCT angiography to assess the total coronary calcium burden. Collimationwas 4mm3.0mm and rotationtime was 500ms. Tube current and voltage were 200mA and 120kV. For the contrast-enhanced scan, collimation was 64mm0.5mm and rotation time was 400 or 450ms, depending on heart rate. Tube current and voltage were 300mA (range 250–400mA) and 120kV (range 100–135kV), respectively.Totalamountofcontrast(Iomeron400,Altana, Konstanz,Germany)was85–105mL,followedbyasalineflush of45mL,bothinjectedat5mL/s.Automateddetectionofpeak enhancement intheaorticrootwasusedtotimethescan. Forty-four 3mm-thick slices of the coronary arterieswere obtainedduringasinglebreathhold(15–20s)synchronizedto 40%oftheR–Rinterval.Inpatientswithaheartrate>65beats/ min, intravenous metoprolol (5–20mg) was administered immediately before MSCT imaging if contraindications for beta-blockade wereabsent. Noadditionalnitroglycerin was given for MSCT imaging. Quantitative CAC score were calculated according tothe methoddescribed by Agatston etal.[11].
BilateralCCA-IMTmeasurementwasperformedusingaB mode ultrasound Toshiba Aplio (Toshiba Medical Systems, Tokyo, Japan) linear array probe (7.5MHz). Subjects were studiedinthesupinepositionwiththeirheadturned458from the side being scanned. Each common carotid artery was evaluated with the subject’s head turned slightly to the contralateralside.Thefielddepth,gain,andnearandfarfield gaincontrolswereoptimizedtoenablevisualizationofthefar wall of the common carotid artery. Carotid arteries were screenedforpresenceofplaquesbilaterallyinthetransverse andlongitudinalplanes,andCCA-IMTwasstudiedintheright andleftcarotidarteries.Carotidarteryplaquewasdefinedasa focal areaofarterialwallthickening >1.5timesthatofthe surroundingarterialwall.CCA-IMTmeasurementsweremade inbothcarotidarteriesatabout3cmafterthebifurcation,at thewidestpointofasegmentdevoidofplaques.CCA-IMTwas defined as the distance between the leading edge of the lumen–intimainterfaceandtheleadingedgeofthe media– adventitia interface of the far wall. In each longitudinal projection,threesites,atthegreatestthicknessandat1cm distalandproximal,wereaveragedandexpressedas mean-CCA-IMT.Therightandleftmeanandmaximumvalueswere determined. All measurementswereperformedbyasingle trainedsonographerwhowasunawareoftheclinicalprofileof thestudysubjects.
Albuminuria was measured as the amount of albumin excretionin24-hurinecollection(urinaryalbuminexcretion, UAE). UAE<30mg/24h was considered as normoalbumi-nuria;30–299mg/24h, microalbuminuriaand 300mg/24h, macroalbuminuria.
Patientswerecategorizedashavingcoronaryheartdisease onthebasisofclinicalandelectrocardiographicevidenceof coronaryarterydiseaseormyocardialinfarction.
Cardiovasculareventsweredefinedasacutemyocardial infarction, hospital admission because of heart failure, intracerebral hemorrhage, occlusionorstenosis ofthe pre-cerebralorcerebralarteries,documentedsignificantcoronary, peripheral artery stenosis(more than 50% inangiography), coronaryarterybypassgraftingorpercutaneoustransluminal
angioplasty/stenting,andothervascularinterventionssuchas percutaneous transluminal angioplasty/stenting or bypass graftingofaortaandperipheralvessels.
3.
Statistical
analysis
SPSS 15.0 for Windows was used in statistical analysis. Categoricalvariableswereexpressedasnumberand percent-age in tables, while descriptive statistics were used for numericalvariables(mean,standarddeviation,median, mini-mum,maximum).Cross-tabulatedstatisticswereprovidedin thecategoricalcomparisonbetweenthegroups,andthelevelof significancewasdeterminedusingtheChi-Squaretest.Mann Whitney and Kruskal Wallis tests were used in numerical comparisonsofindependentgroupsnotfulfillingthecondition ofnormal distribution.In pairwise comparisons,Bonferroni correctionwasappliedbeforeMannWhitneytest.Skeweddata werelogarithmicallytransformed(Ln),andtheLnvalueswere thenusedincorrelationandregressionanalyses.Pearsonrho testwasusedincorrelationanalysis.Multiplelinearregression analysiswasusedforpredictionofLnCACscore.Survivalcurves were estimatedby theKaplan–Meiermethod andevaluated usingaLog-ranktest.Predictivevariablesforoutcomeswere first examined using univariate Cox proportional hazards analysis, and variables with a significant association (P<0.10)wereusedinamultivariateCoxproportionalhazards model.P<0.05wasconsideredsignificant
4.
Results
4.1. Clinicalandlaboratorycharacteristicsofthestudy population
Onehundredandtwentyeightpatients(68.8%female,mean age, 57.58.3 (40–80) years) were included in the study. Mediandurationofdiabeteswas10(1–39)years,andmean BMI was 32.85.4kg/m2. In terms of urinary albumin
excretion, 35 (27.3%) patients were normoalbuminuric, 62 (48.4%) microalbuminuric, and31 (24.2%)were macroalbu-minuric.
When parameters were compared in between normo-, micro-andmacro-albuminuriagroups,statisticallysignificant differences were foundin serum albumin, glucose, HbA1c, HDL,BUN,creatinine,TNF-a,IL-6,CACscoresandCCAplaque formation,aswellasuseofinsulinandoralantidiabeticdrugs (OAD).Theuseofstatins,acetylsalicylicacid,beta-blockers, angiotensinconvertingenzymeinhibitors/angiotensin recep-torblockersandcalciumchannelblockerweresimilarinthree groups. CACscoreandformationofCCAplaquesincreased withincreasinglevelsofalbuminuria.Plaqueformationwas seenin80%ofthepatientsinthemacroalbuminuriagroup, whichisaconsiderablyhighrate.Theseresultsare summa-rizedinTable1.
Total CAC score was statistically significantlyhigher in patientswithplaqueformation(P<0.001).CACscorewas<10
Table1–Patients’sparametersintotalandnormo-,micro-,macroalbuminuriagroups.
Normoalbuminuria (n=35) Microalbuminuria (n=62) Macroalbuminuria (n=31) P Total(n=128) Agea(years) 57.778.85 58.588.67 55.166.72 0.271 57.58.3
Diabetesdurationb(years) 10(1–30) 10(2–33) 14(2–39) 0.170 10(1–39)
BMIb(kg/m2) 33.755.94 32.575.09 32.335.52 0.558 32.85.4
Waistcircumferenceb(cm) 106.5313.51 105.4912.10 105.8411.06 0.827 105.812.1
SystolicBPb(mmHg) 138.423.9 141.421.9 149.925.4 0.164 142.623.5
Albuminuriaa(mg/24h) 19.2(4.9–29.5) 73.9(30.6–294.0) 342.8(317.8–487.6) <0.001 71.2(4.9–487.6)
Creatinineclearanceb(mL/min) 93.0528.59 99.0133.62 84.7735.61 0.543 93.9333.09
HbA1cb(%) 7.51.3 8.71.8 9.11.7 <0.001 8.41.8 HDLcholesterolb(mg/dL) 52.0612.91 51.3713.09 45.1013.76 0.039 50.013.4 LDLcholesterolb(mg/dL) 115.9432.74 108.3634.82 122.5637.12 0.170 113.634.9 Triglyceridea(mg/dL) 152(63–340) 162(50–614) 188(77–602) 0.372 165(50–614) Uricacidb(mg/dL) 4.991.25 4.891.27 5.521.51 0.089 5.01.3 Creatinineb(mg/dL) 0.840.20 0.860.20 1.100.36 0.001 0.910.27 Fibrinogenb(mg/dL) 390.081.4 406.874.0 441.5106.1 0.110 410.686.1 hsCRPa(mg/dL) 0.46(0.5–1.62) 0.42(0.06–1.99) 0.62(0.04–1.87) 0.522 0.44(0.04–1.99) TNF-aa(pg/mL) 10.30(6.69–27.6) 10.55(4.09–31.90) 13.70(7.89–32.90) 0.015 11.00(4.09–32.9) IL-6a(pg/mL) 3.01(2.00–8.07) 3.41(2.00–10.50) 4.41(2.00–8.60) 0.020 3.46(2.00–10.50)
CACscorea(AU) 15.9(0.0–1962.8) 40.6(0.0–2344.0) 160.5(0.0–2916.0) 0.011 46.4(0.0–2916.0)
CCA-IMTbmean(mm) 1.0160.261 0.9900.293 1.0820.249 0.268 1.0740.317
CCA-IMTbmax(mm) 1.0630.278 1.0410.334 1.1550.318 0.307 1.0190.276
CCAplaqueformationc 17(50.0) 32(53.3) 24(80.0) 0.025 73(57)
Smokingc 4(11.4) 11(17.7) 4(12.9) 0.241 19(14.8)
Antidiabeticdrugs
Insulinc 20(57.1) 51(82.3) 29(93.5) 0.001 100(78.1)
OADc 32(91.4) 54(87.1) 18(58.1) 0.001 104(81.3)
Thevalueswereshownasamean(SD),bmedian(interquartilerange),orcn(%).BMI,bodymassindex;BP,bloodpressure;HbA1c,glycosylated
hemoglobin;HDL,high-densitylipoprotein;LDL,low-densitylipoprotein;hsCRP,highsensitivityC-reactiveprotein;TNF-a,tumornecrosis factoralpha;IL-6,interleukin-6;CAC,coronaryarterycalcium;CCA-IMT,commoncarotidarteriesintimamedialthickness;CCA,common carotidarteries;OAD,oralantidiabeticdrugs.
in25.0%,11–100in39.1%,101–400in16.4%,401–1000in10.9% and>1000AUin8.6%.
4.2. Determinantsofcoronarycalcuimscore
LnCACscoreshowedastatisticallypositivecorrelationwith age,LnDurationofdiabetes,systolicbloodpressure, LnAlbu-minuria, BUN, creatinine,uricacid,LnTNF-a, LnIL-6,mean CCA-IMTandanegativecorrelationwithcreatinineclearance, bloodalbumin.TheresultsarepresentedinTable2.
InordertodeterminethefactorsaffectingtheLnCACscore,a modelwascreatedwithage,LnDurationofdiabetes,systolic bloodpressure,LnAlbuminuria,creatinineclearance,uricacid, LnTNF-a,LnIL-6andmeanCCA-IMT,multiplelinearregression analysis was performed. Backward regressionanalysis pre-dicted age, LnAlbuminuria and uric acid as statistically significantfactorsaffectingtheLnCACscore(Table3). 4.3. Fatalandnon-fatalcardiovasculareventsandthe parameters
Themeanfollow-updurationwas36.63.3months.Thesmall numberofdeathsduringthisperioddidnotpermitevaluation ofoverallorcardiovascularmortality.Therefore,weassessed theincidenceofcardiovascularevents.Cardiovascularevents
occurredin23patients(n=18andn=5forpatientswithCAC score100andCACscore<100respectively).The23 cardio-vascular events included ischemic heart disease (n=14; 5 coronaryarteryby-passgraftoperation,3myocardial infarc-tion,4significantcoronarystenosis,2coronaryarterystenting), heartfailure(n=4),peripheralvasculardisease(n=2)stroke (n=2),andsuddendeath(n=1).Toanalyzetherelationship betweenCACscoreandthecardiovascularevents,thepatient populationweredividedintothosewithCACscore100(n=46) andCACscore<100(n=82).PatientswithCACscore100had morecardiovasculareventscomparedtothosewithCACscore <100(39%vs.6%,P<0.0001).Unadjustedcardiovascular event-freesurvivalcurvesforpatientswithCACscore100andCAC score<100arerepresentedinFig.1(P<0.0001byLog-ranktest). InunivariateCoxproportionalhazardsanalysis,LnCACscore, LnAlbuminuria, uric acid and creatinine clearance were significantlyassociated withcardiovascularevents, whereas factorsincludingage,meanCCA,LnTNF-alphaandLnDiabetes durationwerenot(Table4).InamultivariateCoxproportional hazardsmodelusingsignificantvariables(P<0.10)fromthe univariateanalysis,LnCACscore,LnAlbuminuriaanduricacid were identifiedas independent predictors of cardiovascular events(Table4).
5.
Discussion
Type 2 DM is a highly prevalent disease associated with several complications such as retinopathy, neuropathy, nephropathy,cardiovasculardisease,andincreasedmorbidity andmortality[12].Inthisstudy,weinvestigatedthe relation-shipbetweenalbuminuriaand cardiovascularevent predic-tors,CACandinflammatorymarkers,intype2DMpatients.
Table2–CorrelationbetweenLnCACscoresandother parameters.
rho P
Age(years) 0.339 <0.001 LnDurationofdiabetes(years) 0.308 <0.001 Systolicbloodpressure(mmHg) 0.249 0.005 LnAlbuminuria(mg/24h) 0.260 0.003 Creatinineclearance(mL/min) 0.271 0.002 BUN(mg/dL) 0.333 <0.001 Albumin-blood(g/dL) 0.186 0.036 Creatinine(mg/dL) 0.426 <0.001 Uricacid(mg/dL) 0.327 <0.001 LnTNF-a(pg/mL) 0.236 0.012 LnIL-6(pg/mL) 0.236 0.012 CCA-IMTmean(mm) 0.212 0.020 CCA-IMTmax(mm) 0.188 0.039 LnCAC,logarithmicallytransformedcoronaryarterycalcification; BUN,bloodureanitrogen;LnTNF-a,logarithmicallytransformed tumornecrosisfactoralpha;LnIL-6,logarithmicallytransformed interleukin-6;CCA-IMT, commoncarotidarteriesintimamedial thickness.
Table3–SignificantpredictorsofLnCACscoreidentified bymultiplelinearregressionanalysis.
B P 95%confidence interval Constantcoefficient 2.931 <0.001 4.542/ 1.319 Age(years) 0.042 <0.001 0.024/0.061 LnAlbuminuria(mg/24h) 0.323 0.007 0.089/0.558 Uricacid(mg/dL) 0.129 0.036 0.008/0.250 LnTNF-a(pg/mL) 0.896 0.093 0.152/1.944 LnCAC,logarithmicallytransformedcoronaryarterycalcification; LnTNF-a,logarithmicallytransformedtumornecrosisfactoralpha.
Fig.1–Cardiovascularevent-freesurvivalcurvesoftype2 diabeticpatientsduringa44monthperiodaccording coronarycalciumscoreI100(closedsquare)and<100 (opencircle).Therewere46patientsinthecoronary calciumscoreI100groupand82patientsinthecoronary calciumscore<100group.
CAC score has been shown to be proportional with atheroscleroticdiseaseseverityandextent,andpresenceof coronary calcium was shown to be predictive of future coronaryevents[13,14].StudiesinvestigatingCACscoresin variousriskgroupsanddifferentpopulationswerereported and discussed [15–18]. Blaha et al.[19] comparedcoronary heartdiseaseandcardiovasculardiseaseeventrates, multi-variable-adjustedhazardratiosafterstratifyingbyburdenof CAC(scoresof0,1–100,or>100)on444patientsintheMESA JUPITERpopulation.They found74%ofall coronaryevents wereinthe239(25%)ofparticipantswithCACscoresofmore than100.Maliketal.[20]evaluatedCAC,carotidIMTwhich could improve CVD risk stratification over traditional risk factorson6603peoplewithmetabolicsyndromeanddiabetes. CACincreasedtheconcordanceindexes(C-statistic)forevents (P<0.001)overriskfactors.Theyconcludedthatindividuals withMetSordiabeteshavelowrisksforCHDwhenCACor CIMT was not increased and prediction of CHD and CVD eventswasimprovedbyCAC.
Inourstudy,LnCACscore,LnAlbuminuriaanduricacid wereidentifiedasindependent predictorsofcardiovascular events.Urinealbuminexcretionratewasfoundtobe>30mg/ 24hin72.6%of128DMpatientswithnoprevioushistoryof heart disease. There was a positive correlation between urinary albumin levels and CAC score (r=0.260, P=0.003). CACscoreincreasedwithincreasinglevelofalbuminuria,and astatisticallysignificantdifferencewasdeterminedbetween albuminuriagroups(P=0.011).Inpairwisecomparisons,CAC score was significantly higher in macroalbuminuric group compared with normoalbuminuric group (160.5 vs. 15.9, P=0.006).
Elevated serum uric acid levels is a risk factor for cardiovascular and cerebrovascular disease [21–23]. In NHANESIepidemiologicstudiesshowedthatincreasedserum uricacidlevelsareindependentlyandsignificantlyassociated with risk ofcardiovascular mortality [24]. Serum uric acid levelsareanindependentpredictorofdeathinpatientsathigh risk of cardiovascular disease [25]. In present study, we showed asignificant positivecorrelation betweenuricacid and LnCAC score (r=0.36, P<0.001), which has not been reportedbeforeindiabeticpopulation.
InastudyperformedbyFreedmanetal.[6]on588type2DM patients, a strong relationship was demonstrated between albuminuriaandcalcifiedplaquesinthecoronaryandcarotid arteries. Similarly, in our study we determined that CCA plaqueformationincreasedsignificantlywithincreasinglevel
ofalbuminuria.WhilethepresenceofCCAplaquewas50%in thenormoalbuminuricgroup,thisratewasashighas80%in themacroalbuminuricgroup.
The mean CAC scores in patients with CCA plaque formation were also found to be significantlyhigher than thosewithnoplaqueformation(424.18vs.56.08,P<0.001).No significantdifferencewasdeterminedintermsofmeanand maximum CCA-IMT values in the albuminuria groups. However, CAC score had a weak positive correlation with mean and maximum CCA-IMT values. Folsom et al. [13], demonstrated that CAC was better than carotid IMT in predictingcardiovascularevents.Inourstudy,age, albumin-uria and uricacid werefoundtobestatistically significant predictorsofCACscore.
Correlationbetweenmicroalbuminuriaandcardiovascular eventsiswellestablished[26].InastudybyBrunoetal.[27], wheretype2DMpatientswerefollowedoveraperiodof11 years, albumin excretion rate was the main independent predictorofcardiovascularmortality,andthiseffectwasmore pronouncedattheupperlimitsofmicroalbuminuriaandin macroalbuminuria. In agreement these studies, our study show that albuminuria was an independent predictor of cardiovascularevents.Thelevelsofacuteinflammatoryphase reactantsfibrinogen,IL-6andTNF-awerefoundtocorrelate withurinaryalbuminexcretioninDMpatients.Thelevelsof IL-6andTNF-awereshowntobehigherinDMpatientsthanin controls, and theselevels wereshown to increasewithan increaseinexcretionofalbumin[28].Amarkedrelationship was also found between IL-6 and the CAC score [29]. We determined that higher albuminuria levels were correlated with increasingIL-6andTNF-a levels.In pairwise compar-isons,TNF-alevelswereshowntobesignificantlyhigherin themacroalbuminuricgroupthaninthenormoalbuminuric group.(13.70vs.10.30,P=0.004).IL-6levelswerealsofoundto besignificantlyhigherinthemacroalbuminuricgroupthanin the normoalbuminuric group (4.41 vs. 3.01, P=0.005). In present study, we also demonstrated that LnCAC score showed a statistically significant positive correlation with LnTNF-aandLnIL-6.TherelationshipbetweentheCACand cytokine levels in diabetic patients was also assessed by Saremiatal.AmarkedassociationbetweenIL-6andtheCAC scorewasreportedinthisstudy[30].However,theydidnot investigate the interrelationship between urinary albumin excretion, cardiovascularoutcome inadditiontoCACscore andcytokines.Inpresentstudy,weshowedforthefirsttime, the strong relationship between cytokines, CAC score,
Table4–FactorsassociatedwiththecardiovasculareventsinunivariateandmultivariateCoxproportionalhazards regressionanalysis.
UnadjustedHR(95%CI) P-value AdjustedHR(95%CI) P-value
LnCACscore 6.48(2.7–15.5) 0.000 6.82(3.42–13.58) <0.0001 LnAlbuminuria(mg/24h) 2.86(1.29–6.34) 0.01 2.30(1.22–4.35) 0.01 Uricacid 0.62(0.41–1.00) 0.03 0.62(0.42–0.91) 0.02 Creatinineclearance 0.98(0.96–1.00) 0.04 0.98(0.97–1.00) 0.02 Age(years) 1.02(0.95–1.09) 0.64 1.03(0.97–1.09) 0.36 LnDiabetesduration 0.97(0.23–4.16) 0.97 – – MeanCCA(mm) 1.59(0.30–8.44) 0.58 (1.620.31–8.54) 0.57 LnTNF-a 0.54(0.02–13.17) 0.70 (0.550.02–13.18) 0.71 LnCAC,logarithmically transformedcoronary arterycalcification; LnAlbuminuria,logarithmicallytransformed albuminuria; LnDiabetes, logarithmicallytransformeddiabetes;LnTNF-a,logarithmicallytransformedtumornecrosisfactoralpha.
albuminuriaandcardiovasculareventsinthesamediabetic population.
Inconclusion,CACscorewhichreflectssubclinical coro-naryarterydiseaseweresignificantlyassociated with albu-minurialevels,thepresenceofCCAplaqueandinflammatory markersinDMpatients.Furthermore,CACscore,albuminuria and uric acid wereidentified as independent predictors of cardiovascularevents.
Conflict
of
interest
Theauthorsdeclarethattheyhavenoconflictofinterest
r
e
f
e
r
e
n
c
e
s
[1] AnandDV,LimE,LahiriA,BaxJJ.Theroleofnon-invasive imagingintheriskstratificationofasymptomaticdiabetic subjects.EurHeartJ2006;27(8):905–12.
[2] DiagnosisandClassificationofDiabetesMellitusAmerican DiabetesAssociation.Standardsofmedicalcarein diabetes.DiabetesCare2009;32(Suppl.1):13–61.
[3] DiercksGFH,vanBovenAJ,HillegeHL,JanssenWM,Kors TJA,deJongPE,etal.Microalbuminuriaisindependently associatedwithischaemicelectrocardiographic
abnormalitiesinalargenon-diabeticpopulationthe PREVEND(PreventionofREnalandVascularENdstage Disease)study.EurHeartJ2000;21(23):1922–7. [4] FestaA,AgostinoR,HowardG,MykkanenL,TracyRP,
HaffnersM.Inflammationandmicroalbuminuriain nondiabeticandtype2diabeticsubjects:TheInsulin ResistanceAtherosclerosisStudy.KidneyInt 2000;58(4):1703–10.
[5] KwonTG,KimKW,ParkHW,JeongJH,KimKY,BaeJH. Prevalenceandsignificanceofcarotidplaquesinpatients withcoronaryatherosclerosis.KoreanCircJ
2009;39(8):317–21.
[6] FreedmanBI,LangefeldCD,LohmanKK,BowdenDW, CarrJJ,RichSS,etal.Relationshipbetweenalbuminuria andcardiovasculardiseaseintype2diabetes.JAmSoc Nephrol2005;16(7):2156–61.
[7] BonowRO.Clinicalpractice.Shouldcoronarycalcium screeningbeusedincardiovascularpreventionstrategies? NEnglJMed2009;361(10):990–7.
[8] BudoffMJ.Point:diabeticpatientsandcoronarycalcium: riskstratification,compliance,andplaqueprogression. DiabetesCare2003;26(2):541–2.
[9] AmericanDiabetesAssociation.Diagnosisand classificationofdiabetesmellitusAmericanDiabetes Association,definitionanddescriptionofdiabetesmellitus. DiabetesCare2009;32:62–7.
[10] ChenML,HsuCY.ShouldtheK/DOQIdefinitionofchronic kidneydiseasebechanged?AmJKidneyDis
2003;42(October(4)):623–5.
[11] AgatstonAS,JanowitzWR,HildnerFJ,ZusmerNR,Viamonte JrM,DetranoR.Quantificationofcoronaryarterycalcium usingultrafastcomputedtomography.JAmCollCardiol 1990;15(4):827–32.
[12] FoxCS.Cardiovasculardiseaseriskfactors,type2diabetes mellitus,andtheFraminghamHeartStudy.Trends CardiovascMed2010;20(3):90–5.
[13] FolsomAR,KronmalRA,DetranoRC,O’LearyDH,Bild DE,BluemkeDA,etal.Coronaryarterycalcification comparedwithcarotidintima-mediathicknessinthe
predictionofcardiovasculardiseaseincidence:the Multi-EthnicStudyofAtherosclerosis(MESA).ArchInternMed 2008;23:1333–9.
[14] ShawLJ,RaggiP,CallisterTQ,BermanDS.Prognosticvalue ofcoronaryarterycalciumscreeninginasymptomatic smokersandnon-smokers.EurHeartJ2006;27(8):968–75. [15] DiversJ,WagenknechtLE,BowdenDW,CarrJJ,Hightower
RC,XuJ,etal.Ethnicdifferencesintherelationship betweenalbuminuriaandcalcifiedatheroscleroticplaque: theAfricanAmerican-DiabetesHeartStudy.DiabetesCare 2010;33(1):131–8.
[16] DetranoR,GuerciAD,CarrJJ,BildDE,BurkeG,FolsomAR, etal.Coronarycalciumasapredictorofcoronaryeventsin fourracialorethnicgroups.NEnglJMed
2008;358(13):1336–45.
[17] MitchellAR.Coronarycalciumscreening.NEnglJMed 2009;361(25):2490–501.
[18] O’MalleyPG,JonesDL,FeuersteinIM,TaylorAJ.Lackof correlationbetweenpsychologicalfactorsandsubclinical coronaryarterydisease.NEnglJMed2000;343(18): 1298–304.
[19] BlahaMJ,BudoffMJ,DeFilippisAP,BlanksteinR,RiveraJJ, AgatstonA,etal.AssociationsbetweenC-reactiveprotein, coronaryarterycalcium,andcardiovascularevents: implicationsfortheJUPITERpopulationfromMESA,a population-basedcohortstudy.Lancet
2011;20378(9792):684–92.
[20] MalikS,BudoffMJ,KatzR,BlumenthalRS,BertoniAG, NasirK,etal.Impactofsubclinicalatherosclerosison cardiovasculardiseaseeventsinindividualswithmetabolic syndromeanddiabetes:TheMulti-EthnicStudyof Atherosclerosis.DiabetesCare2011;34(10):2285–90. [21] CalbergMJ,KjøllerE.Uricacidasamarkerof
pathophysiologicalmechanismsinpatientswith
cardiovasculardisease.UgeskrLaeger2005;167(40):3771–4. [22] KoenigW,MeisingerC.Uricacid,type2diabetes,and
cardiovasculardiseases:fuelingthecommonsoil hypothesis?ClinChem2008;54(2):231–3.
[23] HaydenMR,TyagiSC.Uricacid:anewlookatanoldrisk markerforcardiovasculardisease,metabolicsyndrome, andtype2diabetesmellitus:theurateredoxshuttle.Nutr Metab2004;191(1):10.
[24] FangJ,AldermanMH.Serumuricacidandcardiovascular mortalitytheNHANESIEpidemiologicFollow-upStudy, 1971–1992.JAMA2000;83(May):18.2004;(Reprinted). [25] IoachimescuAG,BrennanDM,HoarBM,HazenSL,
ByronJHoogwerf.Serumuricacidisanindependent predictorofall-causemortalityinpatientsathighriskof cardiovasculardisease.APreventiveCardiology
InformationSystem(PreCIS)DatabaseCohortStudy. ArthritisRheum2008;58:623–30.
[26] SchiffrinEL,LipmanML,MannJF.Chronickidneydisease: effectsonthecardiovascularsystem.Circulation
2007;116(1):85–97.
[27] BrunoG,MerlettiF,BiggeriA,BargeroG,FerreroS,Pagano G,etal.CasaleMonferratoStudy.FibrinogenandAERare majorindependentpredictorsof11-yearcardiovascular mortalityintype2diabetes:theCasaleMonferratoStudy. Diabetologia2005;48(3):427–4434.
[28] NavarroJF,MoraC,Go´mezM,MurosM,Lo´pez-AguilarC, Garcı´aJ.Influenceofrenalinvolvementonperipheralblood mononuclearcellexpressionbehaviouroftumournecrosis factor-alphaandinterleukin-6intype2diabeticpatients. NephrolDialTransplant2008;23(3):919–26.
[29] SaremiA,AndersonRJ,LuoP,MoritzTE,SchwenkeDC, AllisonM,etal.VADT.AssociationbetweenIL-6and theextentofcoronaryatherosclerosisintheveterans affairsdiabetestrial(VADT).Atherosclerosis
