Lipid accumulation product index in HIV-infected patients: a marker of cardiovascular risk

(1)

w w w . e l s e v i e r . c o m / l o c a t e / b j i d

The

Brazilian

Journal

of

INFECTIOUS

DISEASES

Original

article

Lipid

accumulation

product

index

in

HIV-infected

patients:

a

marker

of

cardiovascular

risk

Milena

Maria

Moreira

Guimarães,

Dirceu

Bartolomeu

Greco,

Allyson

Nogueira

Moreira,

Nathalia

Sernizon

Guimarães

∗

,

Cláudia

Maria

Vilas

Freire,

Bruna

Guimarães

Rohlfs,

Lucas

José

de

Campos

Machado

UniversidadeFederaldeMinasGerais,BeloHorizonte,MG,Brazil

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received10January2018 Accepted26March2018 Availableonline22April2018

Keywords:

HIV

Hyperlipidemia Atherosclerosis

Lipidaccumulationproductindex Cardiovascularrisk

Insulinresistance

a

b

s

t

r

a

c

t

Thelipidaccumulationproduct(LAP)indexisanemergingcardiovascularriskmarker.We aimedtoassesstheaccuracyofthisindexasamarkerofcardiovascularriskinHIV-infected patients.Across-sectionalstudyof133HIV-infectedpatientsonantiretroviraldrugsand20 non-infectedcontrolswasconductedattheoutpatientclinicofareferralcenterofinfectious andparasitic diseases.EvaluationsincludedLAPindex,homeostasismodelassessment (HOMA)index,anthropometricmeasurements,bloodpressure,glucosetolerancetest,and cholesterolandtriglyceridelevels.Bodymassindex(BMI)wassimilarinbothgroups; how-ever,waistcircumferencewasgreaterintheHIV-infectedpatients.Triglyceridelevelswere significantlyhigher(p<0.001)andHDLcholesterollevelswerelowerinHIV-infectedpatients (p<0.001).Plasmaglucose(p=0.01)andinsulin(p=0.005)levelstwohoursafteraglucose load,HOMA-IRindex(p<0.001)andLAPindex(p<0.001)werehigherintheHIV-infected patients. Apositiveandsignificant correlationwasfoundbetweenHOMA-IRindexand LAP(r=0.615;p<0.01),BMI(r=0.334;p<0.01)andwaistcircumference(r=0.452;p<0.01)in theHIV-infectedpatients.InmaleHIV-infectedpatientsandcontrols,ROCcurveanalyses revealedthatthebestcut-offvalueofLAPtodefinethepresenceofinsulinresistancewas 64.8(sensitivity86%,specificity77%andareaunderthecurve0.824).Theseresultsconfirm thatinsulinresistanceismorecommoninHIV-patientsonantiretroviraldrugsthanin HIV-negativecontrols.ApositiveandsignificantcorrelationwasfoundbetweentheLAPindex andtheHOMAindex,withLAP≥64.8constitutinganadditionalriskfactorforcardiovascular diseaseinmaleHIVpatients.

∗ _{Corresponding}_author.

E-mailaddress:nasernizon@hotmail.com(N.S.Guimarães).

https://doi.org/10.1016/j.bjid.2018.03.006

(2)

Introduction

The introduction of highly-active antiretroviral therapy (HAART)hassigniﬁcantlyimprovedtheclinicaloutcome of individuals withthe humanimmunodeﬁciency virus (HIV), resultinginincreasedsurvivalrates1–3_;_however,_the_success

ofantiretroviraltherapyistemperedbylong-termsideeffects that include dyslipidemia, insulin resistance, overt type 2

diabetesmellitus,and changes infatdistribution(peripheral lipoatrophyandvisceraladiposity).4–6

Thepatternofthesemetabolicabnormalitiesinpatients receiving antiretroviral therapy resembles that of the metabolic syndrome, which is known to increase the risk ofcardiovasculardisease.However,whetherand howsoon theseantiretroviraltherapy-inducedabnormalitiesmayresult inaclinicallydetectableincreasedriskofcardiovascular dis-easeremainsunknown,asdoestheimpactoftheunderlying HIV infection per se. Indeed,while results from long-term observationalstudiesontheriskofcardiovasculardiseasein HIV-infectedpatientsremainunavailableintheliterature, cal-culatingthepredictedriskofcardiovasculardiseasemayprove usefulfortheclinicalmanagementofthesepatients.

Insulinresistanceisanindependentriskfactorfor cardio-vasculardiseaseandtheearlyrecognitionofinsulinresistance inthesepatientsisimportantforthepreventionof cardiovas-cularinvolvement.

Euglycemic-hyperinsulinemic clamping is currently the goldstandardformeasuringinsulinresistance.However,itis unsuitableforclinicalpractice,sinceitiscomplex,expensive andunfeasibleforlargepopulations.7 _Taking_into

consider-ation the technicallimitations ofclampingand alternative methods (those that rely on the measurement of insulin itself)for identifyinginsulin resistance, someinvestigators havespeculatedthatinsulin resistanceand,therefore, car-diovascular risk could be determined based on variables associatedwiththeeffectsofinsulininstead ofmeasuring insulindirectly.8,9

The lipid accumulation product (LAP) index, which is basedonacombinationofwaistcircumferenceandfasting triglyceride levels, could be useful inthis situation. LAP is determinedbythefollowingequationforwomen: (circumfer-enceofwaist[cm]–58)×(triglycerides[mmol/L]);formen: [waist (cm) – 65] multiplied by triglyceride level (mmol/L). Resultsareexpressedincmmmol/L.10_Kahn10_was_the_ﬁrst_to

describeLAP,reportingintheNationalHealthandNutrition ExaminationSurvey(NHANESIII)thatLAPwasabetter indi-catorofcardiovascularriskinadultsthanbodymassindex (BMI).

Therefore,theaimofthisstudywastodeterminethe accu-racyoftheLAP indexasamarkerofcardiovascularriskin HIV-infectedpatientsandnon-infectedcontrols.

Patients

and

methods

Designandpatients

Thiscross-sectionalstudywasconductedattheoutpatient clinic of a referral and training Center in Infectious and

ParasiticDiseasesinBrazil.Thesamplewasselectedfor con-venience.Samplesizecalculationofthecasegroupassumed: (a)numberofpatientsinfectedbyHIVin2013,agegroup19–40 years,BeloHorizonte(n=157)11_;_(b)_prevalence_of

lipodystro-phy(84%)12_;_(c)_variation_of_5%;_(d)_conﬁdence_level_equal_to

95%.Theminimumsamplesizewas90individuals.

Thecasegroupcomprised133HIV-infectedpatientsaged between18and55years,whohadbeenreceivingantiretroviral (ARV)drugsforaminimumofthreemonthspriorto admis-sion,regardlessofthetimeofinfectiondiagnosis.Thecontrol groupconsistedof20HIV-negativecontrols.

Exclusion criteriaconsisted ofmetabolic disorders such ashyperlipidemia,diabetesmellitusandlipodystrophyprior to the diagnosis ofHIV-infection; use ofglucocorticoids or anyothersteroids,growthhormones,beta-blockers,thiazides or any drugs associated withmetabolic abnormalities and body fat redistribution. Other criteria included any rele-vant clinical event atthe time ofenrollment tothe study; refusaltoparticipate;pregnancyorbreastfeeding;andalcohol abuse.

The study was approved by the COEP and conducted according to the norms of the code of ethics for human research,NationalHealthCouncil,Resolutionn◦466/2012and allparticipantsgavetheirwritteninformedconsent.

Studyprotocol

Anthropometricdataweremeasuredaccordingtothe proce-dures standardizedbytheWorld Health Organization.13 _To

measureheight,thestadiometeroftheanthropometricscale ofthebrand“Filizzolla®”wasused.Bodyweightwas deter-minedonananthropometricscaleoftheTanitabrandwith acapacityof150kg.Waistcircumferencewasdeﬁnedasthe smallestmeasurementatthemidpointbetweenthe lateral iliaccrestandthelowestrib.14 _BMI_(weight_in_kg_divided_by

heightinmeterssquared)wascalculated.

Theskinfoldsweremeasuredwiththeaidofa plicome-ter(Lange®caliper,SantaCruz,CA,USA)withanaccuracyof 0.1mm.Bicipital(DCB)andtricepsskinfoldsweremeasured intheanteriorandposteriorregion.Thesubscapularskinfold (DCSE)wasmeasuredobliquelyinrelationtothelongitudinal axis,followingtheorientationofthecostalarcs,beinglocated twocentimetersbelowtheangle.Thesupra-iliaccutaneous foldwasobtainedobliquelyinrelationtothelongitudinalaxis, inthemidpointbetweenthelastcostalarchandtheiliaccrest, onthemedianaxillaryline.15

Blood pressure was measured in the right arm, at the level of the heart, using a Welch Allyn Tycos® sphygmo-manometer,model705014(NewYork,NY,USA).Valuesabove 140×90mmHgwereconsideredhigh.

Thesubjects were asked about thepractice, type, dura-tion and frequencyofphysicalactivity (FA)being classified asveryactive(FA>5days/weekand≥30min/session),active (FA 3or 4days/weekand≥20min/session),notveryactive (those individualswhoperformphysicalactivity,but insuf-ficienttobeclassifiedasactivebecauseitdoesnotcomply with the recommendations regarding frequency or dura-tion),andsedentary(thosewhodonotperformanyphysical activity).16

(3)

Assays

Blood sampleswere taken after a 12-h overnight fast and two hours after a 75g oral glucose load. Plasma glu-cose, total cholesterol and triglycerides were determined enzymatically (Vitros ChemistryProducts, Johnson & John-sonClinical Diagnostics®, Rochester, USA).HDL-cholesterol was measured using an immuneinhibition assay (Labtest Diagnosis®, LagoaSanta,Brazil),andinsulinwasmeasured by chemiluminescence (Diagnostic Products Corporation®, LosAngeles,CA,USA).LDL-cholesterolwascalculated.LDL wasnotevaluatedinpatientswithtriglyceridelevelsabove 400mg/dl.

Insulinresistancewasestimatedusing theHomeostasis ModelAssessmentofInsulinResistance(HOMA-IR)basedon thefollowingequation:fastinginsulinlevelin␮U/mL multi-pliedbyfastingglucoselevelinmmol/Ldividedby22.5.17_The

cutoffpointtodeﬁneIRwasarbitrarilydeﬁnedasaHOMA index≥2.71.18 _The_LAP_index_was_calculated_using_the

for-mula[waist(cm)–58]multipliedbytriglyceridelevel(mmol/L) forwomenand[waist(cm)–65]multipliedbytriglyceridelevel (mmol/L)formen.10

Statisticalanalysis

Thedatawerestoredusingdouble-keypunchingwithsoftware

Epidata,version3.1.Quantitativevariablesarepresentedas means±standarddeviationsofthemeanormedians, accord-ing totheirdistribution assessedbytheShapiro–Wilk test. Avariablewasconsideredashavinganormaldistributionif

p<0.05andasnon-parametricifp>0.05.Categoricalvariables arepresentedaspercentages.

Thechi-squaretestwas performedtocompare dichoto-mous variables, while differences between the mean (or median) values of continuous variables were assessed using Student’s t-test or the Mann–Whitney test, the lat-ter being used inthe caseof variables with non-Gaussian distribution.

Receiveroperatingcharacteristic(ROC)curveswere gener-atedfortheLAPindex,BMIandwaistcircumference,witha HOMA-IRindexof2.71asamarkerofinsulinresistance. Sen-sitivityandspecificityforLAPwerecalculatedbasedonthe pointofinflectionintheROCcurve.Sensitivityandspecificity forBMI,waistcircumferenceandHOMAindexwerecalculated usingvalidatedcutoffs(25forBMIand94forwaist circum-ference).Sincetherearegenderdifferencesinthevalidated cut-offlevelsforwaistcircumference, theROCcurveswere generatedonlyforHIV-infectedmen.

Thenon-parametricSpearmanrankcorrelationcoefficient was used to assess agreement between LAP and the con-tinuous independent variables. A two-tailed p-value <0.05 wasconsideredstatisticallysignificant.Theclassification sug-gestedbyLandisandKoch19_was_used_as_follows:_poor-to-fair

agreement(kappa<0.40),moderateagreement(kappaof0.41 to0.60), substantial agreement(kappaof0.61 to0.80), and excellentagreement(kappaof0.81to1.0).Atwo-tailedp-value

<0.05wasconsideredstatisticallysigniﬁcant.

AllstatisticalanalyseswereperformedusingSPSS,version 18.0forWindows(Chicago,IL,USA).

Results

The study included 133 HIV-infected patients and 20 non-infected individuals. There were no statistically signiﬁcant differencesbetweentheHIV-infectedpatientsandthecontrol group in relationto their family history ofdiabetes melli-tus,stroke,coronaryarterydisease,dyslipidemiaorarterial hypertension.Table1summarizestheclinicalandmetabolic proﬁlesofthetwogroups.

BMIwassimilarinbothgroups;however,theHIV-infected patientshadagreaterwaistcircumference.Bicepsandtriceps skinfoldthicknesswassignificantlythinnerandscapularskin foldthicknesswasgreaterintheHIV-infectedpatients com-pared tothe controls. Triglyceride levels were significantly higher and HDL cholesterol levels were lower in the HIV-infectedgroup.Plasmaglucoseandinsulinlevelstwohours afteraglucoseload,HOMA-IRindex,andLAPindexwere sig-nificantlyhigherintheHIV-infectedpatients.

Apositiveandsigniﬁcantcorrelationwasfoundbetween theHOMAindexandLAP,BMIandwaistcircumferencein HIV-infectedpatients(Table2).

InmaleHIV-infectedpatients, analysisoftheROCcurve revealedthatthebestcutoffvalueforLAPtodefinethe pres-enceofinsulinresistancewas64.8(sensitivity86%,specificity 77%andareaunderthecurve[AUC]0.824).ROCcurveswere alsogeneratedforBMIandwaistcircumference(WC)using thecutoffsof25forBMI(sensitivity57%,specificity80%and AUC0.688)and94cmforWC(sensitivity50%,specificity87% andAUC0.717)(Figs.1–3).

Discussion

Inthepresentstudy,theLAPindexwasinvestigatedforthe ﬁrsttimeinHIV-infectedpatients.LAPindexeswerefoundto behigherinHIV-infectedpatientscomparedtocontrols, sug-gesting increasedinsulinresistanceandcardiovascularrisk inHIV-infectedpatientsonantiretroviraltherapy.Overrecent decades,it has becomeclearthatHIV-infectedpatients on antiretroviralsdevelopmetabolicdisturbancesthatresemble thosefoundwiththemetabolicsyndrome4–6_and,_moreover,

severalcohortstudieshavereportedahigherriskof myocar-dialinfarctioninHIV-infectedpatients.20,21

Identifyinginsulinresistanceinclinicalpracticeremains a challenge. The gold standard,euglycemic hyperinsuline-micclampingisclearlyinadequate inviewofitshigh cost andcomplexity.Alternativemethodsthatusefastinginsulin levelscanleadtoamisdiagnosisofinsulinresistance,since thereisnowell-establishedreferencerangefornormalinsulin levels.22_For_this_reason,_indexes_that_correlate_well_with_the

clampanddependonfastinginsulin,suchastheHOMA-IR index,aredifﬁculttoimplementasaclinicaltest.

TheLAPindexisbasedonacombinationofWC measure-ment and fasting triglyceride levels.10 _It_was _developed _to

reﬂectthe combinedanatomicaland physiologicalchanges associated with lipid over-accumulation in adults. In that study,KahncomparedtheLAPindextoBMIintermsofthe abilitytoidentifycardiovascularriskinadults,usingdatafrom theNHANESIII.ComparedtoBMI,correlationsbetweenthe

(4)

Table1–ClinicalandmetabolicfeaturesofHIV-infectedpatientsonhighly-activeantiretroviraltherapy(HAART)and non-infectedcontrols.

Variables HIV/HAARTpatients(n=133) Controls(n=20) p-Value

Age(years) 39.90(±8.81) 33.71(±8.30) 0.004 (t-test)

Sex

Male 68.4% 35.0% 0.004

Female 31.6% 65.0% (Chi-squaretest)

Schooling

<8years 47.4% 10.0% 0.002

≥8years 52.6% 90.0% (Chi-squaretest)

Maritalstatus:Single

Yes 63.2% 45% 0.121

No 36.8% 55% (Chi-squaretest)

Currentsmoker

Yes 31.6% 5.0% 0.014

No 68.4% 95.0% (Chi-squaretest)

Highbloodpressure

Yes 8.3% 5.0% 0.612 No 91.7% 95.0% (Chi-squaretest) Sedentarylifestyle Yes 79.7% 45.0% 0.001 No 20.3% 55.0% (Chi-squaretest) Weight(kg) 65.5(±12.0) 68.7(±11.6) 0.323 (t-test) BMI(kg/m2₎ _23.8_(±3.8) _24.2_(±3.2) _0.666 _(t-test)

Waistcircumference(cm) 85.5(±9.4) 80.5(±9.9) 0.03 (t-test)

Bicepsskinfoldthickness(mm) 6.5(±3.9) 9.1(±3.9) 0.007 (t-test)

Tricepsskinfoldthickness(mm) 8.9(±5.2) 16.4(±6.5) <0.001 (t-test)

Supra-iliacskinfoldthickness(mm) 12.1(±6.5) 13.8(±5.5) 0.29 (t-test)

Scapularskinfoldthickness(mm) 20.1(±8.9) 16.9(±5.9) 0.05 (t-test)

Systolicbloodpressure(mmHg) 120.0(±15.2) 113.5(±11.9) 0.054 (t-test)

Diastolicbloodpressure(mmHg) 75.9(±8.9) 72.5(±6.2) 0.101 (t-test)

Triglycerides(mg/dl) 150.5(IQR157.0) 80.0(IQR42.0) <0.001(Mann–Whitneytest)

Totalcholesterol(mg/dl) 194.6(±61.1) 181.1(±29.2) 0.332 (t-test)

LDL-cholesterol(mg/dl) 117.1(±39.9) 101.5(±26.0) 0.095 (t-test)

HDL-cholesterol(mg/dl) 35.8(±12.5) 61.3(±14.2) <0.001 (t-test)

Fastingglucose(mg/dl) 89.1(±19.4) 86.5(±8.7) 0.56 (t-test)

Plasmaglucoselevels2hafteraglucoseload(mg/dl) 119.7(±54.0) 87.7(±28.8) 0.01 (t-test)

Fastinginsulin(U/mL) 8.86(±8.9) 4.7(±4.45) 0.05 (t-test)

Plasmainsulinlevels2hafteraglucoseload(U/mL) 38.65(IQR56.10) 21.55(IQR32.80) 0.005(Mann–Whitneytest)

HOMA-IRindex 1.384(IQR1.260) 0.765(IQR1.000) 0.001(Mann–Whitneytest)

LAPindex 45.95(IQR64.44) 19.96(IQR14.53) <0.001(Mann–Whitneytest)

Boldvalueindicatesap-valuelessthan0.05.

Table2–Spearman’srankcorrelationcoefﬁcientbetweenHOMAindexandlipidaccumulationproduct(LAP),bodymass index,andwaistcircumference.

Independentvariables Spearman’srankcorrelationcoefﬁcient(r) p-Value

Lipidaccumulationproduct 0.615 <0.01

Bodymassindex 0.334 <0.01

Waistcircumference 0.452 <0.01

LAPindexandlipidriskvariableswerestronger,suggesting thatLAPmightbeabetterpredictoroftheincidenceof car-diovasculardisease.

Previousstudies haveevaluatedthe LAPindexin differ-ent insulin-resistant populations such as individuals with

type 2 diabetes mellitus23 _and _polycystic _ovary _syndrome

(PCOS)24,25_;_however,_this_is_the_ﬁrst_time_that_the_LAP_index

hasbeenevaluatedinHIVpatients.Resultsfromthepresent studyshowastrongcorrelationbetweentheLAPindexand

the HOMA-IRindexinHIV-infectedpatient.Inouropinion, screeningthispopulationthatispronetodevelopinginsulin resistance-related comorbidities, including cardiovascular diseases,priortotheonsetofsymptomsmayproveuseful.

ItisinterestingtonotethatalthoughBMIwassimilarin thetwogroups,WCwasgreaterintheHIV-infectedpatients. Inaddition,totalcholesterolandLDLcholesterollevelswere similarinthetwogroups;however,triglyceridelevelswere sig-niﬁcantlyhigherintheHIV-infectedpatients.Previousstudies

(5)

Fig.1–ROCcurveforbodymassindexwithHOMAindex of2.71asamarkerofinsulinresistance.

Fig.2–ROCcurveforwaistcircumferencewithHOMA

indexof2.71asamarkerofinsulinresistance.

Fig.3–ROCcurveforlipidaccumulationproduct(LAP)with HOMAindexof2.71asamarkerofinsulinresistance.

haveevaluatedincreasedWCandhightriglyceridelevelsin differentpopulationsasasurrogatemarkerofcardiovascular risk,withresultsshowingthatthesetwovariablesweremore sensitivethan themetabolicsyndromeitself asmarkersof greatercardiovascularrisk.26,27

AccordingtotheROCcurve,withaLAPindex≥64.8, sensi-tivity(86%)andspeciﬁcity(77%)wereadequatefordetecting astateofinsulinresistanceandperformedbetterthanBMIat acutoffpointof25andWCatacutoffpointof94cm.

Somelimitationsofthisstudymustbementioned, princi-pallywithrespecttothesmallsamplesizeofcontrolgroup; thus,thefindingofhigherLAPlevelsinHIV-infectedpatients comparedtocontrolshastobeinterpretedwithcaution. How-ever,this findingisinlinewiththe resultsofother studies thatshowedanincreasedmetabolicriskinthispopulation. Another limitation ofthe present study isthat euglycemic hyperinsulinemic clamping was not performed because of thetechnicaldifficultiesandhighcost.Nonetheless,thereis astrongcorrelationbetweentheHOMAIRindex,usedasa referencestandard,andeuglycemichyperinsulinemicclamp results.28

Insummary,theLAPindexmaybeconsideredareliable markerofinsulinresistanceandariskfactor to cardiovas-culardiseaseinmaleHIV-infectedpatientsonantiretroviral therapy.Earlyrecognitionofpatientswhoareproneto devel-opinginsulinresistanceandrelatedmetabolicdisturbances will allow therapeuticinterventions to be implemented to reducefuturecardiovascularrisk.

Conﬂicts

of

interest

Theauthorsdeclarenoconﬂictsofinterest.

r

e

f

e

r

e

n

c

e

s

1.HoggRS,O’ShaughnessyMV,GataricN,etal.Declinein deathsfromAIDSduetonewantiretrovirals.Lancet. 1997;349:1294.

2.PalellaFJJr,DelaneyKM,MoormanAC,etal.Declining morbidityandmortalityamongpatientswithadvanced humanimmunodeﬁciencyvirusinfection.HIVOutpatient StudyInvestigators.NEnglJMed.1998;338:853–60.

3.MocroftA,VellaS,BenﬁeldTL,etal.Changingpatternsof mortalityacrossEuropeinpatientsinfectedwithHIV-1. EuroSIDAStudyGroup.Lancet.1998;352:1725–30.

4.CarrA,CooperDA.Adverseeffectsofantiretroviraltherapy. Lancet.2000;356:1423–30.

5.GuimarãesMM,deOliveiraARJr,PenidoMG,etal. Ultrasonographicmeasurementofintra-abdominalfat thicknessinHIV-infectedpatientstreatedornotwith antiretroviraldrugsanditscorrelationtolipidandglycemic proﬁles.AnnNutrMetab.2007;51:35–41.

6.GuimarãesMM,GrecoDB,FigueiredoSM,etal.

High-sensitivityC-reactiveproteinlevelsinHIV-infected patientstreatedornotwithantiretroviraldrugsandtheir correlationwithfactorsrelatedtocardiovascularriskandHIV infection.Atherosclerosis.2008;201:434–9.

7.DeFronzoRA,TobinJD,AndresR.Glucoseclamptechnique:a methodforquantifyinginsulinsecretionandresistance.AmJ Physiol.1979;237:E214–23.

(6)

8. WiltgenD,BenedettoIG,MastellaLS,etal.Lipidaccumulation productindex:areliablemarkerofcardiovascularriskin polycysticovarysyndrome.HumReprod.2009;24:1726–31.

9. WehrE,GruberHJ,GiulianiA,etal.Thelipidaccumulation productisassociatedwithimpairedglucosetolerancein PCOSwomen.JClinEndocrinolMetab.2011;96:E986–90.

10.KahnHS.Thelipidaccumulationproductperformsbetter thanthebodymassindexforrecognizingcardiovascularrisk: apopulation-basedcomparison.BMCCardivascDisord. 2005;5:26.

11.PrefeituradeBeloHorizonte.SecretariaMunicipaldeSaúde, GerênciadeEpidemiologiaeInformac¸ão.AIDS–2000–2013. Publicadoem25/10/2013.

12.MinistériodaSaúde,2015.DST:AIDS,HepatitesVirais. Lipodistroﬁa.Brasília,2015.Availablefrom:

http://www.aids.gov.br/pcdt/10.

13.WorldHealthOrganization.Physicalstatus:theuseand interpretationofanthropometry.ReportofaWHOExpert Committee.Geneva:WHO;1995.

14.LeiteCC,MatsudaD,WajchenbergBL,CerriGG,HalpernA. Correlac¸ãodamedidadeespessuraintra-abdominalmedida pelaultra-sonograﬁacomosfatoresderiscocardiovascular. ArqBrasEndocrinolMetab.2000;44:49–56.

15.DurninJV,WomersleyJ.Bodyfatassessedfromtotalbody densityanditsestimationfromskinfoldthickness: measurementson481menandwomenagedfrom16–72 years.BrJNutr.1974;32:77–97.

16.BaumanA,AinsworthBE,SallisJF,etal.Thedescriptive epidemiologyofsitting.A20-countrycomparisonusingthe InternationalPhysicalActivityQuestionnaire(IPAQ).AmJ PrevMed.2011;41:228–35.

17.MatthewsDR,HoskerJP,RudenskiAS,NaylorBA,TreacherDF, TurnerRC.Homeostasismodelassessment:insulinresistance and␤-cellfunctionfromfastingplasmaglucoseandinsulin concentrationsinman.Diabetologia.1985;28:412–9.

18.GelonezeB,RepettoEM,GelonezeSR,TambasciaMA, ErmeticeMN.Thethresholdvalueforinsulinresistance (HOMA-IR)inaadmixturepopulationIRintheBrazilian MetabolicSyndromeStudy.DiabetesResClinPract. 2006;72:219–20.

19.LandisJR,KockCG.Themeasurementofobserveragreement forcategoricaldata.Biometrics.1977;33:159–74.

20.WormSW,SabinC,WebweR,etal.Riskofmyocardial infarctioninpatientswithHIVinfectionexposedtospeciﬁc individualantiretroviraldrugsfromthe3majordrugclasses: thedatacollectiononadverseeventsofanti-HIVdrugs (D:A:D)study.JInfectDis.2010;201:318–30.

21.BrouwerES,NapravnikS,EronJJJr,etal.Effectsof

combinationantiretroviraltherapiesontheriskofmyocardial infarctionamongHIVpatients.Epidemiology.2014;

25:406–17.

22.ChevenneD,TrivinF,PorquetD.Insulinassaysandreference values.DiabetesMetab.1999;25:459–76.

23.BozorgmaneshM,HadaeghF,AziziF.Diabetesprediction, lipidaccumulationproduct,andadipositymeasures;6-year follow-up;Tehranlipidandglucosestudy.LipidsHealthDis. 2010;9:45.

24.WehrE,GruberHJ,GiulianiA,etal.Thelipidaccumulation productisassociatedwithImpairedGlucoseTolerancein PCOSWomen.JClinEndocrinolMetab.2011;96:986–90.

25.WiltgenD,BenedettoIG,MastellaLS,SpritzerPM.Lipid accumulationproductindex:areliablemarkerof cardiovascularriskinpolycysticovarysyndrome.Hum Reprod.2009;24:1726–31.

26.LemiexI,PascotA,CouillardC,etal.Hypertricyridemicwaist: amarkeroftheatherogenicmetabolictriad

(hyperinsulinemia;hyperapolipoproteinB;small,denseLDL) inmen?Circulation.2000;102:179–84.

27.TankóBL,BaggerYZ,QinG,AlexandersenP,LarsenPJ, ChristiansenC.Enlargedwaistcombinedwithelevated triglyceridesisastrongpredictorofaccelerated atherogenesisandrelatedcardiovascularmortalityin Postmenopausalwomen.Circulation.2005;111: 1883–90.

28.BonoraE,TargherG,AlberichM,etal.Homeostasismodel assessmentcloselymirrorstheglucoseclamptechniquein theassessmentofinsulinsensitivity:studiesinsubjectswith variousdegreesofglucosetoleranceandinsulinsensitivity. DiabetesCare.2000;23:57–63.