Elevated 1-h post-load plasma glucose levels in normal glucose tolerance children with obesity is associated with early carotid atherosclerosis

ContentslistsavailableatScienceDirect

Obesity

Research

&

Clinical

Practice

j o ur na l h o me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / o r c p

Original

Article

Elevated

1-h

post-load

plasma

glucose

levels

in

normal

glucose

tolerance

children

with

obesity

is

associated

with

early

carotid

atherosclerosis

Suna

Kılınc¸

_,

_Tuna

_Demirbas¸

_,

_Enver

_Atay

_,

_Ömer

_Ceran

_,

_Zeynep

_Atay

c_{MedipolUniversityHospital,DepartmentofPediatrics,Turkey}

d_{MedipolUniversityHospital,DepartmentofPediatricEndocrinology,Turkey}

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received8November2019

Receivedinrevisedform3February2020 Accepted5February2020

Keywords: Obesity

1-hpost-loadhyperglycemia Atherosclerosis

Non-alcoholicfattyliverdisease

a

b

s

t

r

a

c

t

Context:Evidencesuggeststhatthe1-hpost-loadplasmaglucose(1-hPG)≥155mg/dLduringanoral glu-cosetolerancetest(OGTT)predictsdevelopmentoftype2diabetes(T2DM)andassociatedcomplications, amongadultswithnormalglucosetolerance(NGT),butrelevantdataonchildrenisscarce.

Objectives:ToinvestigatewhetherNGTchildrenwithobesitywhose1-hPGis≥155mg/dLhavean increasedcarotidintima-mediathickness(IMT)andexhibitnon-alcoholicfattyliverdisease(NAFLD) diagnosedbyultrasonography,ascomparedwithNGTsubjectswith1-hPG<155mg/dLandimpaired glucosetolerance(IGT).

Methods:Cardio-metabolicprofile,OGTT,measurementsofcarotidIMTandliverultrasonographywere analyzedin171non-diabeticchildrenwithobesity.Subjectsweredividedinto3groups:NGTsubjects witha1-hPG<155mg/dL,NGTsubjectswitha1-hPG≥155mg/dL,andIGTsubjects.

Results:AscomparedwithNGTindividualswitha1-hPG<155mg/dL,NGTindividualswitha1-hPG ≥155mg/dLexhibitedhighercarotidIMT(0.75±0.15mmvs.0.68±0.15mm;p<0.05).Nosignificant differenceswereobservedincarotidIMTbetweenIGTandNGTsubjectswitha1-hPG≥155mg/dL (0.75±0.18mmvs0.75±0.15mm;p>0.05).Ofthethreeglycemicparameters,1-hand2-hPG,butnot fastingglucose,weresignificantlycorrelatedwithcarotidIMT.Therewerenosignificantdifferencesfor increasedriskofhavingNAFLDbetweenthethreegroups.

Conclusions:Thesedatasuggestthatavalueof1-hPG≥155mg/dLinchildrenandadolescentswith obesityisasimportantasIGTwithrespecttocardiovascularrisks.

©2020AsiaOceaniaAssociationfortheStudyofObesity.PublishedbyElsevierLtd.Allrights reserved.

Introduction

Theincidenceoftype2diabetesmellitus(T2DM)andrelated dysglycemicconditionsatriskfordiabetes(calledasprediabetes) aregradually increasingworldwide mostly dueto the continu-ouslyincreasingprevalenceofobesity[1].Inadolescence,obesity hasbeenassociatedwithincreasedriskofcoronaryvascular dis-abilityinlaterlife[2].ItisknownthatT2DMisanindependent

∗ Correspondingauthorat:MerkezMahallesi,Dr.SadıkAhmetCaddesi,Bagcılar, Istanbul,Turkey.

E-mailaddresses:sunahancili@hotmail.com(S.Kılınc¸),drtd698@hotmail.com (T.Demirbas¸),atayenver@gmail.com(E.Atay),oceran@medipol.edu.tr(Ö.Ceran), atay.zeynep@yahoo.com(Z.Atay).

riskfactorforcardiovasculardisease(CVD)andnon-alcoholicfatty liverdisease(NAFLD)[3,4].Commoncarotidarteryintima-media thickness(IMT)isawell-establishedmeasureofearly atheroscle-rosis and is broadly used as a surrogate marker for CVD [5]. Severalstudiesreportedthickercarotidwallsindiabeticpatients thaninindividualswithnormalglucosetolerance(NGT)[6].Also, there is a clear association between diabetes and NAFLD [7]. Besidestoallofthese,anincreasingnumberofemergingstudies havedeterminedthatprediabetesstatus,whichincludesimpaired fastingglucose(IFG),impairedglucosetolerance(IGT)and/or gly-catedhemoglobin(HbA1c),werefoundtobeassociatedwithan increasedriskofNAFLDandplayavitalroleinatheroscleroticCVD [8,9].

However,somereportsindicateasignificantriskforCVDand NAFLDeveninsubjectswithNGT[10,11,12].Hanefeldetal. evalu-https://doi.org/10.1016/j.orcp.2020.02.001

atedtheriskfactorsforcarotidIMTinanon-diabeticriskpopulation andfoundthatpostprandialplasmaglucosewasanindependent risk factor for increased carotid IMT [10]. Kimura et al. inves-tigatedwhethersecretionpatternsofglucose andinsulin could influencethehistologicalseverityinNAFLDpatientswithoutprior knownT2DM.Theyfoundthatpostprandialhyperinsulinemia(but notglucoselevels)wasassociatedwithadvancedfibrosis[11].In arecentreport,NGTindividualswith1-hpost-challengeplasma glucose _≥155mg/dL have increasedblood viscosity comparable tothatobservedinsubjectswithIFGand/orIGT.Bloodviscosity hasbeenfoundtobeassociatedwithleftventricularhypertrophy, atherosclerosis,vascular stiffness and NAFLD[12]. Additionally, prospectiveepidemiologicalstudieshavealsodemonstratedthat 30–40%ofindividualswhodevelopT2DMhaveNGTatbaseline, indicatingthatprediabeticconditionsarelimitedintheirability toidentifyindividualsatfutureriskfordevelopingdiabetes[13]. Therefore,forpreventingdiabetesanditsassociatedcomplications itwouldbeimportanttodetecttheindividualswhocarrythe high-estriskatNGT.Recently,acutoffpointof155mg/dL(8.6mmol/L) forthe1-hpost-loadplasmaglucose(PG)duringanoralglucose tolerancetest(OGTT)hasbeensuggestedtobeabletoidentify indi-vidualswithNGTathighriskforfutureT2DM[14,15].Alsoseveral observationalstudies,suchasCATAMERI(Catanzarometabolicrisk factors)study,clearlydemonstratedthatNGTadultswith1-hPG ≥155mg/dLwerepredisposedtoanincreasedriskfordeveloping T2DMovera5yearsperiod[16].InthisCATAMERIstudy,theyalso foundastrongassociationbetweenelevated1-hpost-loadPG lev-elsinadultswithNGTwithcarotidIMTandNAFLD[17,18].Ina mostrecentstudy,Tanakaetal.showedthatPGat60minwas associatedwithincreasedcarotidIMTmorestronglythanPGat 120minby75gOGTTsuggestedthatPGat60mincouldbe consid-eredasapredictorofatherosclerosis[19].Inchildren,ithasbeen demonstratedthat1-hpost-loadPG≥155mg/dLduringanOGTT iscorrelatedwithanunfavorablemetabolicandinflammatory pro-file[20].Accordingtothisstudyacutoffvalueof1-hpost-loadPG ≥155mg/dLcouldbeusedasanadditionalmarkertoidentify chil-drenandadolescentswithobesityatincreasedriskofdeveloping obesitycomplications[20].Because,thereisnodataregardingthe impactofthe1-hpost-loadPGontheriskofatherosclerosisand NAFLDinasymptomaticchildren,wedesignedthisstudytoaddress thequestionifglucosetolerancestatus,andinparticular1-h post-loadPGlevels,mayaffectcarotidIMT andNAFLDinagroupof childrenwithobesity.

Methods

Studydesignandparticipants

Thestudywascarriedoutoveraperiodof12monthsfromJan 2018toJan2019.Thesubjectswereselectedfromthepatients agedbetween10–18yearsthatpresentedtoouroutpatientclinic toassesstheirmetabolicstatusforobesity.Aftera12-hfasting,all subjectsunderwentthoroughaphysicalexaminationanda bio-chemicalevaluation.Patientswithdiabeteswereexcludedfrom thestudy.Additionaly,patientswhohaveachronic gastrointesti-naland/orcardiovasculardisease,anyhematologicdisease,which could alter HbA1c values, a history of drug use, an endocrine pathologyorsuspectedsyndromesassociatedwithobesitywere also excluded. Except those, patients who have a body mass index(BMI)greater than95thpercentileaccordingtothe stan-dardsoftheCentersforDiseaseControlandPrevention(CDC2000) wereconsecutivelyenrolledinthestudy.Thereafter,aftera 12-hfasting1.75g/kg(max75g) OGTTwascarriedout in orderto determine glucose status of 178 eligible participants. Glucose and insulin levels measured 5 time points

(0-,30-,60-,90-,120-min) during a 2-hOGTT. Among 178patients, 7patients were excludedbecausetheyhavediagnosedwithdiabetesaftertheOGTT hadbeenperformed.CarotidIMTwasmeasuredandliver ultra-sonographywereperformedinallpatientsonthesamedaywith OGTT.

Clinicalandbiochemicalmeasurements

Aftera12-hfasting,allindividualsunderwent anthropometri-calevaluationincludingassessmentofBMI,blood pressure(BP) obtainedintheleftarmofthesupineposition,after10minofquiet rest.Abiochemicalevaluationincludingthyroidfunctiontestsand serum cortisolmeasurementsforprobableendocrine pathology wereperformedinallsubjects.HeightwasmeasuredusingaSeca stadiometer witha sensitivityof 0.1cm. Weight wasmeasured usingaSecascalewithasensitivityof0.1kg.Heightandweight wereobtainedwithparticipantsinlightclothesandwithoutshoes. Serum concentrations of fasting plasma glucose (FPG), triglyc-eride, total cholesterol(TC),low-density lipoproteincholesterol (LDL-C),andhigh-densitylipoproteincholesterol(HDL-C) concen-trationsweremeasuredenzymaticallyusingDPModularSystems (RocheDiagnosticCorp.,Indianapolis,IN).Fastinginsulinwas mea-sured using sandwich electrochemiluminescence immunoassay (ECLIA)techniques(ElecsysInsulin,RocheDiagnosticsCorp., Indi-anapolis,IN).Highperformanceliquidchromatography(HPLC)was employedtodeterminethepercentageofHbA1c.

Calculation

BMIwascalculatedbydividingweight(kg)byheightsquared (m2_{).Patientswitha BMIover95thpercentileweredefinedas} obese[21].Glucosetolerancestatus wasevaluatedaccordingto ADAcriteria[22]:individualswerethusclassifiedasNGTwhen FPGwaslessthan100mg/dLand2-hpost-loadPGwaslessthan 140mg/dL,andIGTwhen2-hpost-loadPGwas140−199mg/dL. Then,individualsintheNGTgroupwerefurtherdividedintotwo subgroupsbasedupontheir1-hpost-loadPGconcentration(NGT witha 1-hpost-load PG<155mg/dL, NGTwitha 1-hpost-load PG≥155mg/dL,respectively).

MeasurementofcarotidIMT

HighresolutionB-modeultrasoundimagingwasusedto mea-sureIMTofthecommoncarotidartery(CCA)bya7.5MHzlinear arraytransducer(Aplio500,Toshiba,Japan).IMTwasmeasured fromplaque-free 10mmlinear arterialsegment atthe farwall oftheCCA,10mmproximaltothecarotidbifurcation.Foreach patienttwomeasurementswereperformedbilaterally,thevalues wereaveraged,andpresentedasthemeanofIMTofthecommon carotidartery.Ultrasoundassessmentwasperformedbyan expe-riencedradiologistwhowasunawareofthesubjects’clinicaland laboratoryfindings.

Liverultrasonography

Ultrasonographyoftheliverexecutedbythesameexperienced radiologistwhowasblindtoparticipants’clinicalcharacteristics. Rightintercostal,rightsubcostallongitudinal-transverseand epi-gastriclongitudinal-transversescanningwerecarriedoutusinga 3.5MHzconvextransducer(Aplio500,Toshiba,Japan).The ultra-sonographic criteria used to diagnose fatty liver included liver and kidneyechogenicity discrepancy, thepresence of diffusely increasedhepaticechogenicity,indistinctperiportalechogenicity and/ordisturbedvisibilityofdeeperportionsoftheliver.Since semiquantitativeevaluationofthedegreeoffattyliverby

ultra-soundwasnotavailableparticipantswerecategorizedassubjects withorwithoutfattyliver.

Ethics

Thisstudywasconductedinaccordancewiththedeclarationof HelsinkiandinitiatedaftertheapprovalofthelocalEthics Com-mitteeofMedipolUniversity(approvelnumber:2018-619).Data werecollectedfromtheparticipantsattendingatendocrinology outpatientclinicinUniversityofHealthSciencesIstanbulBagcılar TrainingandResearchHospitalafterthenatureofthestudywas explainedtothemandwritteninformedconsentofthesubjects andparentswasobtainedbeforethestudy.

Statisticalanalysis

Inadditiontodescriptivestatistics(mean,standard-deviation), one-wayanalysisof variance (ANOVA)test wasusedfor group comparisonsofnormallydistributedvariables.Kruskal–Wallistest wasusedforintergroupcomparisonsofnon-normallydistributed variables,Dunn’smultiplecomparisontestwasutilizedinthe com-parisonofsubgroups,andChi-squaretestwasperformedforthe evaluationofqualitativedata.Acorrelationanalysiswasperformed usingSpearman’scorrelationanalysis.Allvariableswereincluded inamultiplelinearregressionanalysistoassesstheindependent determinants.AP-valueof<0.05wasconsideredstatistically sig-nificant.StatisticalanalysiswasperformedusingtheprogramNCSS 2007(NumberCruncherStatisticalSystem,Kaysville,Utah,USA).

Results

Atotalof171non-diabeticsubjectswithobesity(97females, 74males,meanage:13.1_±1.9years)wereincludedinthestudy. Among171patients,69hadimpairedglucosetolerance(IGT).A 1-hpost-loadPGcutoffpointof155mg/dLduringOGTTwasused todividesubjectswithNGTintotwogroups:73subjectswith1-h post-loadPG<155mg/dL,and29subjectswith1-hpost-loadPG ≥155mg/dL.Table1showstheclinicalcharacteristicsand labora-toryfindingsofthethreestudygroups.

Nostatisticallysignificantdifferencewasfoundbetweenthe groupsintermsofage,gender,heightandBMI(p>0.05).Asshown inTable1,IGTsubjectshadaworsemetabolicandcardiovascular riskprofileexhibitingsignificantlyhighertriglycerides,insulin(1-h and2-hpost-challengeinsulin),glucose(fasting,1-hand2-h post-loadPG)andcarotidIMTascomparedwithNGTindividualswith 1-hpost-loadPG<155mg/dL(p<0.05).Bycontrast,nosignificant differenceswereobservedinmetabolicandcardiovascularrisk fac-torsbetweenIGTsubjectsandNGTindividualswith1-hpost-load PG≥155mg/dL(p>0.05),withtheexceptionof2-hpost-challenge glucoselevelswhich wereincreasedintheIGTgroup(p<0.05). Ascomparedwithindividualswith1-hpost-loadPG<155mg/dL, NGTindividualswith1-hpost-loadPG≥155mg/dLexhibited sig-nificantlyhigherFGP,1-hpost-challengeinsulinlevelsandcarotid IMT(p<0.05).Therewerenosignificantdifferencesbetweenthe twogroupsintermsofotherclinicandmetabolicparameters.

Theultrasonographyrevealed NAFLDin a total of 71(41.5%) ofthe subjects;of these,26 (36.6%) had NGT witha 1-h post-loadPG<155mg/dL,11(15.4%)hadNGTwitha1-hpost-loadPG ≥155mg/dL,and34(47.8%)hadIGT.Itisnotablethatindividuals withIGTexhibitanincreasedrisktohaveNAFLD,butincomparison ofthreegroups,nosignificantdifferenceswereobservedinrelation topresenceofNAFLD(Table1).ButpatientswithNAFLDhad signif-icantlyincreasedcarotidIMT(0.76±0.16mmvs.0.69±0.15mm; p=0.003)thanthepatientswithoutNAFLD.

Fig.1.CarotidIMTinthestudypopulationstratifiedaccordingtoglucosestatus.

Fig.2.ProportionofNAFLDdiagnosedbyultrasonographyamongindividuals strat-ifiedaccordingtoglucosestatus.

CarotidIMTinthestudypopulationandproportionofNAFLD amongpatientsstratifiedaccordingtoglucosestatusisshownin Figs.1and2.

In Spearman’s correlation analysis, of the three glycemic parameters,1-hand2-hpost-loadPG,butnotfastingPG,were sig-nificantlypositivelycorrelatedwithcarotidIMT(r=0.215,p=0.005 for 1-h post-load PG; r=0.189,p=0.013 for 2-h post-load PG). Additionally,carotidIMTwascorrelatedwithBMI-SDS(r=0.172, p=0.026), triglycerides (r=0.191, p=0.013) and ALT (r=0.183, p=0.017), while there was a negative correlation with HDL-C (r=−0.190,p=0.013),(Table2).Toestimatetheindependent con-tributionofcarotidIMT,andeliminatetheinfluenceofconfounders, we conducted a multiple linear regression analyses (Table 3). The three variables that remained significantly associated with carotid IMT were BMI-SDS (␤-coefficient=0.155,p=0.040), 1-h post-loadPG(␤-coefficient=0.151,p=0.044)and2-hpost-loadPG (␤-coefficient=0.153,P=0.041).

Table1

Anthropometricandmetaboliccharacteristicsofstudysubjectsstratifiedaccordingtoglucosetolerance. Variables NGTwith1-h PG<155mg/dL NGTwith1-h PG≥155mg/dL IGT P PNGTwith1-h PG<155mg/dL vs.NGTwith1-h PG≥155mg/dL PNGTwith1-h PG<155mg/dL vs.IGT PNGTwith1-h PG≥155mg/dL vs.IGT Gender(n)(male/female) 73(37/36) 29(9/20) 69(28/41) 0.165 0.114 0.298 0.508 Age(years) 13.5±1.8 14.0±2.0 13.3±1.8 0.257 0.448 0.822 0.226 Height-SDS 0.19±1.03 0.08±1.19 0.29±0.97 0.634 0.875 0.833 0.624 BMI(kg/m2) 32.5±3.5 33.1±4.9 32.8±3.3 0.700 0.699 0.853 0.919 BMI-SDS 2.8±0.5 2.9±0.5 2.9±0.4 0.237 0.231 0.555 0.680 SBP(mmHg) 121±12 122±12 125±11 0.131 0.953 0.123 0.416 DBP(mmHg) 75±9 79±8 78±9 0.110 0.145 0.248 0.820 HbA1c 5.7±0.1 5.8±0.1 5.9±0.2 0.311 0.781 0.557 0.318 Totalcholesterol(mg/dL) 181±30 187±23 188±37 0.365 0.698 0.350 0.969 HDL-C(mg/dL) 47±8 46±9 44±7 0.159 0.948 0.147 0.512 LDL-C(mg/dL) 117±25 123±19 121±28 0.512 0.567 0.642 0.940 Triglycerides(mg/dL) 126±36 137±55 150±67 0.044 0.897 0.047 0.058 ALT(UI/L) 27±19 27±18 33±20 0.136 0.999 0.148 0.349 AST(UI/L) 27±15 25±9 30±12 0.211 0.599 0.565 0.196 Fastingglucose(mg/dL) 87±6 90±6 90±8 0.014 0.044 0.016 0.985 1-hglucose(mg/dL) 126±17 165±14 173±27 0.0001 0.0001 0.0001 0.263 2-hglucose(mg/dL) 117±14 122±13 161±16 0.0001 0.276 0.0001 0.0001

Fastinginsulin(␮IU/mL) 20±8 21±9 23±11 0.208 0.858 0.181 0.687

1-hinsulin(␮IU/mL) 112±50 170±80 150±86 0.0001 0.001 0.005 0.453

2-hinsulin(␮IU/mL) 98±58 102±60 139±91 0.003 0.955 0.003 0.068

NAFLD(n) 26(35%) 11(38%) 34(49%) 0.233 0.826 0.340 0.420

CarotidIMT(mm) 0.68±0.15 0.75±0.15 0.75±0.18 0.025 0.043 0.014 0.999 Boldvalues:p<0.05issignificant.

Dataaremeans±SD.NGT,normalglucosetolerance;IGT,impairedglucosetolerance;PG,plasmaglucose;BMI,bodymassindex;SBP,systolicbloodpressure;DBP,diastolic bloodpressure;HDL,high-densitylipoprotein;LDL,low-densitylipoprotein;ALT,Alanineaminotransferase;AST,aspartateaminotransferase;NAFLD,non-alcoholicfatty liverdisease;IMT,intima-mediathickness.

Table2

CorrelationbetweencarotidIMTandcardio-metabolicvariables.

Variables Pearson’scorrelationcoefficient(r) P

Age(years) 0.146 0.057 Height-SDS −0.020 0.799 BMI-SDS 0.172 0.026 SBP(mmHg) −0.050 0.548 DBP(mmHg) 0.088 0.289 HbA1c 0.021 0.787 Totalcholesterol(mg/dL) 0.026 0.734 HDL-C(mg/dL) −0.190 0.013 LDL-C(mg/dL) −0.023 0.769 Triglycerides(mg/dL) 0.191 0.013 ALT(UI/L) 0.183 0.017 AST(UI/L) 0.095 0.217 Fastingglucose(mg/dL) 0.041 0.596 1-hglucose(mg/dL) 0.215 0.005 2-hglucose(mg/dL) 0.189 0.013

Fastinginsulin(␮IU/mL) 0.106 0.167

1-hinsulin(␮IU/mL) 0.136 0.076

2-hinsulin(␮IU/mL) 0.018 0.816

Boldvalues:p<0.05issignificant.

Dataaremeans±SD.BMI,bodymassindex;SBP,systolicbloodpressure;DBP, dias-tolicbloodpressure;HDL,high-densitylipoprotein;LDL,low-densitylipoprotein; ALT,Alanineaminotransferase;AST,aspartateaminotransferase.

Discussion

Inthiscross-sectionalstudy,wefoundthatNGTchildrenwith

obesity,whose1-hpost-loadPGis≥155mg/dL,haveanincreased

carotidIMTascomparedwithNGTsubjectswith1-hpost-loadPG

<155mg/dL,whichisafindingconsistentwiththestudyof

Suc-curroetal.thatwasperformedinadultpatients[17].Inaddition,

itisobservedthatNGTpatientswith1-hpost-loadPG≥155mg/dL have a similar cardiometabolic risk profile,including 1-h post-loadPGlevels,withIGTindividuals,whoareconsideredathigh riskfor both T2DMand CVD.In particular,whenwe compared thethreeglucosehomoeostasisparameters,astrongercorrelation

Table3

MultiplelinearregressionanalyseswithcarotidIMTasthedependentvariable. Independentvariables ␤-coefficient t p

Age(years) 0.129 1.67 0.128 BMI-SDS 0.155 2.07 0.040 SBP(mmHg) −0.101 −1.31 0.180 DBP(mmHg) 0.065 0.81 0.413 Totalcholesterol(mg/dL) 0.136 1.76 0.118 HDL-C(mg/dL) −0.105 −1.34 0.183 LDL-C(mg/dL) −0.142 −1.81 0.071 Triglycerides(mg/dL) 0.067 0.83 0.410 ALT(UI/L) 0.144 1.82 0.070 AST(UI/L) −0.091 −1.21 0.207 Fastingglucose(mg/dL) 0.120 1.53 0.138 1-hglucose(mg/dL) 0.151 2.03 0.044 2-hglucose(mg/dL) 0.153 2.06 0.041

Fastinginsulin(␮IU/mL) −0.118 −1.50 0.140 1-hinsulin(␮IU/mL) −0.032 −0.41 0.521 2-hinsulin(␮IU/mL) −0.040 −0.56 0.507 Boldvalues:p<0.05issignificant.

IMT,intima-mediathickness;BMI,bodymassindex; SBP,systolicblood pres-sure;DBP,diastolicbloodpressure;HDL,high-densitylipoprotein;LDL,low-density lipoprotein;ALT,Alanineaminotransferase;AST,aspartateaminotransferase.

was observed with 1-hpost-load PG and carotid IMT as

com-paredwithFPGand2-hpost-loadPG.Themechanismbywhich

elevated1-hPGisassociatedwithincreasedcarotidIMT,

surro-gate markerfor atherosclerosis, hasnot beenelucidated yet.A

higherdegreeofinsulinresistancecouldbeonepossible

expla-nation as demonstrated by Succurro et al. [17]. In their study

theyfoundthatNGTsubjectswith1-hpost-loadPG≥155mg/dL havelowerinsulinsensitivityascomparedwithNGTindividuals with1-hpost-loadPG<155mg/dL.Inourstudywehaveshown that1-hpost-loadinsulinlevelsaresignificantlyhigher(170±80 ␮IU/mLvs.112±50␮IU/mL)inpatientswithNGTwith≥155mg/dL thanin patientswithNGTwith<155mg/dL. Itwasevenhigher thaninpatientswithIGT.However,insulinlevelswerenot corre-latedwithcarotidIMTinourstudy,thismightsuggeststhatpost

prandialhyperglycemiarather thanhyperinsulinemiaplaysrole inprogressionofatherosclerosis.Thereisawideevidencefrom manystudiesdemonstratingthatsignificantfluctuationsinplasma glucoselevelshavemoredeleteriouseffectoncardiovascular sys-temthancontinuousglucoseelevations.Marfellaetal.reported that maintaining plasma glucose at 15mmol/L for 2h resulted inincreasedmeanheartrate,systolicanddiastolicpressureand plasmacatecholaminelevels[23].Endothelialfunctionmeasured byflow-mediatedendothelium-dependentvasodilationofbrachial artery and myocardial perfusion were also negatively affected bypost-prandialhyperglycemiaintwodifferentstudies[24,25]. Besidestheseclinicalstudies,manyinvitro andanimal studies demonstratedthehazardouseffectofpost-prandialhyperglycemia ontheendothelialcells.Atthecellularlevel,acutehyperglycemia createsoxidativestressthroughoverproductionoffreeradicalsin themitochondriawhichinturnmediatestheapoptoticeffects[26]. Monocyteadhesiontoendothelialcellswhichisanearlyfindingin atherosclerosisincreasesinresponsetoacuteglucoseelevations anditisevenhigherindiabeticratsfedtocreatepostprandial glu-cosespikesthanratsfedatlibitumdespiteahigherHbA1cinthe lattergroup[27].

The pathophysiologyof NAFLD is associated with lipotoxic-ity,inflammatorycytokines,apoptosis,andinsülinresistance[28]. Also,oxidative stressis wellknownasoneof themost impor-tantfactor forinflammation andprogression ofhepatic fibrosis inNAFLDpatients[29].Itissuggestedthathyperinsulinemiaand hyperglycemia,especiallyglycemicvariability,areimportant pre-dictivefactorsinglucoseimpairmentfortheprogressionofhepatic fibrosis in NAFLD [30]. Variability of blood glucose might also inducemonocyteadhesiontoendothelialcells,activate inflamma-torycytokinesandinflammation,andincreaseoxidativestressin theliverofpatientswithNAFLD[30]. Therelationshipbetween NAFLDanddysglycemia includingT2DM,IFGand IGThasbeen demonstratedinbothadultsandchildren.AccordingtoBedogni etal.approximately40%ofthechildrenwithobesityhadNAFLD [31].IGTorT2DMwaspresentin25%ofthechildrenwithNAFLD versus8% ofthose without NAFLD[31]. Earlypost-load hyper-glycemiaat1-hhasalsobeenshowntobeassociatedwithincreased NAFLDriskinadultswithNGT[18].Itwassurprisingthatwecould notdetectanincreasedriskofhavingNAFLDinchildrenwith1-h post-loadPG≥155mg/dL.Theriskwasincreasedbutnot statis-ticallysignificantinourcohort withIGT. Thisfindingmightbe relatedtotherelativelylownumberofcaseswith1-hpost-loadPG ≥155mg/dLinourstudypopulationbesideslimitedsensitivityof USGtodetectearlycasesofNAFLDsinceALTlevelswerepositively correlatedwithcarotidIMT.LiverbiopsyratherthanUSGwouldbe abletodemonstrateasignificantassociationbetweenNAFLDand 1-hpost-loadPG≥155mg/dL.Althoughwecouldnotidentifyan associationbetweenNAFLDand1-hpost-loadPGwecould demon-strateanincreasedcarotidIMTinpatientswithNAFLD.Thiswas expectedbecausepathophysiologicalmechanismsindysglycemia thatleadtocardiovascularandhepaticchangesarebasically simi-lar.

Inconclusion,1-hpost-loadPG≥155mg/dLinchildrenwith obesityisasimportantasIGTwithrespecttocardiovascularrisks. ItmayevencaptureNGTchildrenwithobesitywhoareatriskfor developingcardiovascularproblems.Long-termfollow-upstudies withlargecohortscouldalsodemonstrateitsvalueforpredicting futureriskofmetabolicsyndromeandT2DM.RevisionofADA crite-riaforidentifyingprediabeticconditionsandadding1-hpost-load PG≥155mg/dLtothesecriteriashouldbeconsidered.

Declarationofinterest

None.

Fundingsource

Nofundswereprovidedforthisstudy.

Authorcontributions

S.Kılınc¸andZ.Ataydesignedanddirectedtheproject.S.Kılınc¸ andZ.Ataytooktheleadinpreparingthedraftmanuscriptwith medicalwritingassistance.Allotherauthorscontributedtodata collection,participatedindataanalysisandinterpretation,and crit-icallyreviewedthedraftmanuscript.

Acknowledgement

Theauthorsthankallparticipatinginvestigatorsandsite per-sonnelinvolvedindataacquisition.

AppendixA. Supplementarydata

Supplementarymaterialrelatedtothisarticlecanbefound,in theonline version,at doi:https://doi.org/10.1016/j.orcp.2020.02. 001.

References

[1]InternationalDiabetesFederation,Availablefrom:IDFdiabetesatlas.8thed. Brussels,Belgium:InternationalDiabetesFederation;2017http://www.idf. org/diabetesatlas.

[2]HenrikssonH,HenrikssonP,TyneliusP,EkstedtM,LabayenI,RuizJR,etal. Cardiorespiratoryfitness,muscularstrength,andobesityinadolescenceand laterchronicdisabilityduetocardiovasculardisease:acohortstudyof1 mil-lionmen.EurHeartJ2019,http://dx.doi.org/10.1093/eurheartj/ehz774,pii: ehz774.[Epubaheadofprint].

[3]HaffnerSM,LehtoS,RönnemaaT,PyöräläK,LaaksoM.Mortalityfromcoronary heartdiseaseinsubjectswithtype2diabetesandinnondiabeticsubjectswith andwithoutpriormyocardialinfarction.NEnglJMed1998;339(4):229–34. [4]ShibataM,KiharaY,TaguchiM,TakenoshitaS,IshiyamaN,AkuzawaM,etal.

Nonalcoholicfattyliverdiseaseisariskfactorfortype2diabetesin middle-agedJapanesemen.DiabetesCare2007;30(11):2940–4.

[5]HeissG, Sharett AR,Barnes R,Chambless LE,Szklo M,Alzola C. Carotid atherosclerosis measured by B-mode ultrasoundin populations: associa-tion with cardiovscular risk factors in the ARIC Study. Am J Epidemiol 1991;134(3):250–6.

[6]Temelkova-KurktschievTS,KoehlerC,LeonhardtW,SchaperF, HenkelE, SiegertG,etal.HanIncreasedintimal-medialthicknessinnewlydetectedtype 2diabetes:riskfactors.DiabetesCare1999;22(2):333–8.

[7]Williams KH, Shackel NA, Gorrell MD, McLennan SV, Twigg SM. Dia-betesandnonalcoholocifattyliverdisease:apathogenicduo.EndocrRev 2013;34(1):84–129.

[8]DECODEStudyGroup.Glucosetoleranceandcardiovascularmortality: com-parisonofthefastingandthe2-hourdiagnosticcriteria.ArchInternMed 2001;161(3):397–404.

[9]BaeJC,ChoYK,LeeWY,SeoHI,RheeEJ,ParkSE,etal.Impactofnon-alcoholic fattyliverdiseaseoninsulinresistanceinrelationtoHbA1clevelsin nondia-beticsubjects.AmJGastroenterol2010;105(11):2389–95.

[10]HanefeldM,KöhlerC,SchaperF,FueckerK,HenkelE,Temelkova-Kurktschiev T.Postprandialplasmaglucoseisanindependentriskfactorforincreased carotidintima-mediathicknessinnon-diabeticindividuals.Atherosclerosis 1999;144(1):229–35.

[11]KimuraY,HyogoH,IshitobiT,NabeshimaY,ArihiroK,ChayamaK. Post-prandialinsulinsecretionpatternisassociatedwithhistologicalseverityin non-alcoholicfattyliverdiseasepatientswithoutpriorknowndiabetes melli-tus.JGastroenterolHepatol2011;26(3):517–22.

[12]MariniMA,FiorentinoTV,AndreozziF,ManninoGC,PerticoneM,Sciacqua A,etal.Elevated1-hpost-challengeplasmaglucoselevelsinsubjectswith normalglucosetoleranceorimpairedglucosetoleranceareassociatedwith wholebloodviscosity.ActaDiabetol2017;54(8):775–84.

[13]UnwinN,ShawJ, ZimmetP,Alberti KG.Impairedglucose toleranceand impairedfastingglycemia:thecurrentstatusondefinitionandintervention. DiabetesMed2002;19(9):708–23.

[14]Abdul-GhaniMA,Abdul-GhaniT,AliN,DefronzoRA.One-hourplasmaglucose concentrationandthemetabolicsyndromeidentifysubjectsathighriskfor futuretype2diabetes.DiabetesCare2008;31(8):1650–5.

[15]Abdul-GhaniMA,LyssenkoV,TuomiT,DefronzoRA,GroopL.Fastingversus post-loadplasmaglucoseconcentrationandtheriskforfuturetype2diabetes: resultsfromtheBotniaStudy.DiabetesCare2009;32(2):281–6.

[16]FiorentinoTV,MariniMA,AndreozziF,ArturiF,SuccurroE,PerticoneM,etal. One-hourpostloadhyperglycemiaisastrongerpredictoroftype2diabetes thanimpairedfastingglucose.JClinEndocrinolMetab2015;100(10):3744–51. [17]SuccurroE,MariniMA,ArturiF,GrembialeA,LugaraM,AndreozziF,etal. Elevatedone-hourpost-loadplasmaglucoselevelsidentifiessubjectswith normalglucosetolerancebutearlycarotidatherosclerosis.Atherosclerosis 2009;207(1):245–9.

[18]SestiG,HribalML,FiorentinoTV,SciacquaA,PerticoneF.Elevated1-h post-loadplasmaglucoselevelsidentifyadultswithnormalglucosetolerancebut increasedriskofnon-alcoholicfattyliverdisease.BMJOpenDiabetesResCare 2014;2(1):e000016.

[19]TanakaK,KanazawaI,YamaguchiT,SugimotoT.One-hourpost-load hyper-glycemia by 75g oral glucose tolerance test as a novel risk factor of atherosclerosis.EndocrJ2014;61(4):329–34.

[20]SerbisA,GiaprosV,ChallaA,ChaliasosN,SiomouE.Elevated1-hourpost-load plasmaglucoseidentifiesobeseyouthwithabnormalglucosemetabolismand anunfavorableinflammatoryprofile.ClinEndocrinol(Oxf)2018;89(6):757–64. [21]KuczmarskiRJ,OgdenCL,GuoSS,Grummer-StrawnLM,FlegalKM,MeiZ,etal. 2000CDCGrowthChartsfortheUnitedStates:methodsanddevelopment. VitalHealthStat2002;11(246):1–190.

[22]AmericanDiabetesAssociation.Classificationanddiagnosisofdiabetes: Stan-dardsofMedicalCareinDiabetes.DiabetesCare2018;41(Suppl.1):S13–27. [23]MarfellaR,VerrazzoG,AcamporaR,LaMarcaC,GiuntaR,LucarelliC,etal.

Glutathionereversessystemichemodynamicchangesbyacutehyperglycemia inhealthysubjects.AmJPhysiol1995;268:E1167–73.

[24]KawanoH,MotoyamaT,HirashimaO,HiraiN,MiyaoY,SakamotoT,etal. Hyperglycemia rapidlysuppresses flow-mediatedendothelium-dependent vasodilationofbrachialartery.JAmCollCardiol1999;34(1):146–54. [25]ScognamiglioR,NegutC,DeKreutzenbergSV,TiengoA,AvogaroA.

Postpran-dialmyocardialperfusioninhealthysubjectsandintype2diabeticpatients. Circulation2005;112(2):179–84.

[26]PiconiL,QuagliaroL,AssaloniR,DaRosR,MaierA,ZuodarG,etal. Con-stant and intermittent high glucose enhances endothelialcell apoptosis throughmitochondrialsuperoxideoverproduction.DiabetesMetabResRev 2006;22(3):198–203.

[27]OtsukaA,AzumaK,IesakiT,SatoF,HiroseT,ShimizuT,etal.Temporary hyper-glycaemiaprovokesmonocyteadhesiontoendothelialcellsinratthoracic aorta.Diabetologia2005;48(12):2667–74.

[28]YoonIC,EunJR.Pharmacologictherapyfornonalcoholicsteatohepatitis focus-ingonpathophysiology.YeungnamUnivJMed2019;36(2):67–77.

[29]FujiiH,KawadaN.Inflammationandfibrogenesisinsteatohepatitis.J Gastroen-terol2012;47(3):215–25.

[30]HashibaM,OnoM,HyogoH,IkedaY,MasudaK,YoshiokaR,etal.Glycemic vari-abilityisanindependentpredictivefactorfordevelopmentofhepaticfibrosis innonalcoholicfattyliverdisease.PLoSOne2013;8(11):e76161.

[31]BedogniG,GastaldelliA,MancoM,DeColA,AgostiF,TiribelliC,etal. Relation-shipbetweenfattyliverandglucosemetabolism:across-sectionalstudyin571 obesechildren.NutrMetabCardiovascDis2012;22(2):120–6.