The impact of glycemic variability in the progression of renal disease in diabetic patients treated with insulin therapy

www.elsevier.es/rmi

Medicina

e

Investigación

ORIGINAL

ARTICLE

The

impact

of

glycemic

variability

in

the

progression

of

renal

disease

in

diabetic

patients

treated

with

insulin

therapy

S.

Pereira

_,

_J.

_Barros

_,

_A.

_Salgueiro

_,

_N.

_Cardoso

_,

_M.

_Esteves

_,

_J.

_Salomão

_,

D.

Silva

_,

_A.

_Duarte

a_{DepartmentofInternalMedicineI,CentroHospitalardoMedioAve,VilaNovadeFamalicão,Portugal} b_{DepartmentofFamilyMedicine,UnidadeSaúdeFamiliarSantaClara,ViladoConde,Portugal}

Received4October2015;accepted6October2015 Availableonline18March2016

KEYWORDS

DiabetesMellitus; Glycosylated hemoglobin; Chronickidney disease; Coefficientof variability

Abstract

Introduction:Diabeticnephropathyistheleadingcauseofend-stagerenaldiseaseinPortugal. Previousstudiesreportedthatreductionofglycosylatedhemoglobin(HbA1c)inpatientswith DiabetesMellitusisassociatedwithdecreasedmicrovascularcomplications,includingchronic kidneydisease. However,fewstudiesreportedtherelationbetweenHbA1Cfluctuationsand renal disease. This study aims to evaluate the relationship between fluctuations in HbA1c andrenaldiseaseprogressionindiabeticstreatedwithinsulintherapy.

Methods:Thisisaretrospectivecohortstudy.Diabeticpatientstreatedwithinsulintherapy, whowereobservedbetweenJanuaryandDecember2011,wereenrolledfor3yearsfollow-up. Wecalculatedthebaselineandfinalfollow-upestimatedglomerularfiltrationrate(eGFR)using theChronicKidneyDiseaseEpidemiologyCollaborationequationanddefinedpatientswithor withoutrenaldiseaseprogression.Also,weexaminedtherelationshipbetweenfluctuationsin HbA1c(measuredbycoefficientofvariability)andchangesineGFR.

Results:Ofthe538subjectsenrolled,221completedthefollow-upandwereincludedinthe study.24%hadprogressioninrenalstatusand76%maintainedtheeGFR.The coefficientof variability inprogressiongroup versus noprogression group was 0.0973versus 0.0812with p0.019(withstatisticalsignificance).

Abbreviations: CKD,chronickidneydisease; HbA1c,glycosylatedhemoglobin;eGFR,estimatedglomerularfiltrationrate;CKD-EPI, ChronicKidneyDiseaseEpidemiologyCollaborationequation;DM,DiabetesMellitus;NKF,NationalKidneyFoundation;BMI,bodymass index;HTA,arterialhypertension.

∗_{Correspondingauthor.}

E-mailaddress:[email protected](S.Pereira).

http://dx.doi.org/10.1016/j.mei.2016.01.006

Conclusion:Thesedatasuggestthatfluctuationsinglycemiccontrolareassociatedwithkidney diseaseprogressionindiabeticstreatedwithinsulintherapy.

©2016UniversidadAutónomadelEstadodeMéxico.PublishedbyMassonDoymaMéxicoS.A. Thisisanopenaccess articleundertheCCBY-NC-NDlicense(http://creativecommons.org/ licenses/by-nc-nd/4.0/).

PALABRASCLAVE

Diabetesmellitus; Hemoglobina glucosilada;

Lesiónrenalcrónica; Coeficientede variabilidad

Impactodelavariabilidaddelahemoglobinaglucosiladaenlafunciónrenalde enfermosinsulinodependientes

Resumen

Introducción:Lanefropatíadiabéticaeslaprincipalcausadelesiónrenalcrónicaterminalen Portugal.Estudiospreviosreportanquela reduccióndela hemoglobina glucosilada(HbA1c) en diabéticos está asociada a una disminución de las complicaciones microvasculares, incluyendolanefropatía. Apesarde esto,pocos estudiosrelatanla relaciónentrelas fluc-tuacionesdeHbA1cylaenfermedadrenal.Elobjetivodelestudioesevaluarlavariabilidadde laHbA1cendiabéticosinsulinodependienteseinvestigarsuinfluenciaenlafunciónrenal. Métodos: Seprocedióalarealizacióndeunestudioretrospectivo, dediabéticos insulinode-pendientes seguidos en la consultade Diabetología de un centro de Portugal,de eneroa diciembrede2011,durante3a˜nos.Lapoblaciónevaluadafueclasificadaen2grupos:conysin progresióndelalesiónrenal,deacuerdoconlafórmulaChronicKidneyDiseaseEpidemiology Collaboration.Secalculó elcoeficientedevariabilidad delaHbA1c, quesedefine como la razónentreladesviaciónestándardelaHbA1cintrapersonalylamediadelaHbA1c.

Resultados: Seincluyeron538diabéticosinsulinodependientesdeloscuales221completaron elseguimiento.El76%noreflejóprogresióndelalesiónrenalyel24%sílareflejó.Elcoeficiente devariabilidadentreelgrupodeprogresiónyeldenoprogresióndelalesiónrenalfuede0.0973 frentea0.0812,loqueevidenciaresultadosestadísticamentesignificativossiendop=0.019. Conclusión:LavariabilidaddelaHbA1cinfluyesobrelaprogresióndelalesiónrenalen dia-béticosinsulinodependientes.

©2016UniversidadAutónomadelEstadodeMéxico.PublicadoporMassonDoymaMéxicoS.A. EsteesunartículoOpenAccessbajola licenciaCCBY-NC-ND(http://creativecommons.org/ licenses/by-nc-nd/4.0/).

Introduction

Diabeticnephropathyistheleadingcauseofend-stagerenal disease in Portugal and 30% of diabetic patients develop CKD in all stages.1 _{Improvements in glycemic control are}

associatedwithreductionsintheincidenceofmicrovascular complications,includingCKD.2

The diabeticnephropathyisamajorpublichealth pro-blemworldwide andis associated toother cardiovascular diseases,suchasarterialhypertension,dyslipidemia, obe-sity,coronaryheartdiseaseandstroke.3

Theconcernabouttheglycemiccontroltoprevent dia-betic long-term complications was demonstrated in se-veral prospective studies, which used the mean HbA1c to understand who develop renal disease, more often. Recently,therearesomestudieswithestablished relation-shipbetweenfluctuationsinHbA1canditsimpactinrenal function.4_{In fact,HbA1c variabilityisfrequent in}

diabet-icsandit makesdifficult toachieve appropriateglycemic control5_.

Toourknowledge,thisisthefirstPortuguesebased po-pulation study, in which HbA1c variability has been used toexaminediabeticcomplications,suchasnephropathy,in selectedpatientstreatedwithinsulintherapy (minimizing theimpactoforalantidiabeticdrugsinrenalfunction),and withoutarterialhypertension,dyslipidemiaortobaccouse.

Diabetic patients with arterial hypertension were excluded becausethis disease is an important risk factor toproteinuriaandglomerulosclerosis.6,7

Dyslipidemiaisimplicatedinatherosclerosisandin car-diovascularcomplicationsofrenaldisease.8_{Previousstudies}

show that high levels of cholesterol are related to renal disease progression. Thus we excluded patients with dys-lipidemiatoavoidbias.

The tobacco is associated with an increased speed in renal disease progression in diabetic and non-diabetic nephropathybecauseof itsvasoconstrictoreffectandthe increasedriskofthromboembolicevents.9

Thepurposeofthisstudywastoexaminetherelationship betweenfluctuationsinHbA1covertimeandrenaldisease progression in diabetics treated with insulin therapy and withoutotherdiseasesrelatedtonephropathy.

Methods

This retrospective cohort study examined adult patients (>18yearsold)withestablisheddiagnosisoftype1ortype 2DMtreatedwithinsulintherapy,inamedicalcenterinthe NorthofPortugal.ThebaselinetimeperiodwasJanuaryto December2011,witha3yearsfollow-up.

eGFR = 141 X min (Scr/κ,1) X max(Scr/κ,1) X 0.993 X 1.018 [if female] X 1.159

[if black]

α -1.209 Age

Scr is serum creatinine (mg/dL), κ is 0.7 for females and 0.9 for males, α is –0.329for females and –0.411 for males, min indicates the minimum of Scr/κ or 1, and max indicates

the maximum of Scr/κ or 1.

Figure1 ChronicKidneyDiseaseEpidemiologyCollaborationequation.

n=538 diabetics

19% type 1 diabetics

81% type 2 diabetics

n=221 (40.08%) Excluded

Dyslipidemia (8.92%)

HTA (10.78%)

Smokers (22%) HTA and dyslipidemia (39.22%)

Figure2 Patientsenrolledinthecohortstudy.

liquidchromatography)andwe collecteddataofpatient’s medical records such as age, sex, height, weight, serum creatinine levels (determined with Jaffe’s method,

cal-ibrated by isotope dilution mass spectrometry) and

co-morbidities.

The exclusion criteria included patients with arterial hypertension,dyslipidemia andsmokers,tominimize bias inwhatconcernsrenaldiseaseprogression.

TheeGFRatbaselineandatlastfollow-upvisitwere cal-culatedusingtheCKD-EPIequation(Fig.1)andCKDstage wasbasedonNKFcriteria.10

Twogroupsweredefined:onewithrenaldisease progres-sion,whohadworsenedtheirNKFstage;andtheothergroup withoutrenaldiseaseprogression,whomaintainedtheirNKF stageduringthefollow-upperiod.

The glycemic variability is defined by fluctuation of HbA1candthecoefficient of variabilitywascalculatedby theratiobetweenstandarddeviationofintrapersonalHbA1c valuesobtainedandthemeanHbA1cvalue.

ThestatisticalanalysiswasperformedbySPSSStatistics® version 20.0, to compare the influence of HbA1c variabi-lity in progression and no progression groups, using the

Mann---Whitney test, and Chi-square test for categorical

variablescomparison.

Results

From538 patients with DMobserved, 41.07% (n=221) ful-filledthe inclusion criteria(Fig.2). The mean age of the populationwas52±8yearsand58%werewomen(Table1).

BMI>25kg/m2waspresentin67%ofthepopulation.

Table1 Patientdemographiccharacteristics.

Characteristics %

Sex

Male 42.0

Female 58.0

Renalprogression 24.0

Sex

Male 41.5

Female 58.5

TypeofDM

Type1 30.1

Type2 69.9

Typeoftreatment

Insulin 77.4

Insulin+oralantidiabetics 22.6

Norenalprogression 76.0

Sex

Male 42.0

Female 58.0

TypeofDM

Type1 33.3

Type2 66.6

Typeoftreatment

Insulin 80.4

Table2 Homogeneouscharacteristicsofcohortstudy.

Progressionversusnoprogression pvalue

TypeofDM 0.181

Typeoftreatment 0.904

HbA1c

No progression group Progression group

8.536 0.8327 0.0973 0.0812 ∗p<0.05 0.6728 8.103 Mean Standard deviation

Coefficient of variability ∗

∗

∗

Figure3 Mean,standarddeviationandcoefficientof variabi-lityofHbA1cinprogressionandnoprogressiongroup.

ThemeannumberofHbA1cmeasurementswas8.2±1.5 duringthefollow-up.

The progression in renal dysfunction was identified in 24%(58.5%werewomenand30.1%hadtype1DM)and76% maintainedtheeGFR(58%werewomenand33.3%hadDM type1).

InthegroupwithmaintainedeGFR,19.6%weretreated withoral antidiabeticdrugs in additiontoinsulin therapy and, in the progression group, 22.6% were treated with oral antidiabetic drugs; but with the Chi-square test, we obtained p 0.904, which means there are no important differencesbetweenthesetwogroupsrelativelyto antidia-betictherapy(Table2).

ThemeanHbA1cinthegroupwithoutrenaldisease pro-gressionwas8.013,withastandarddeviationof0.7anda coefficientofvariabilityof0.0812.ThemeanHbA1cinthe renaldiseaseprogressiongroupwas8.638,withastandard deviationof0.8327andacoefficientofvariabilityof0.0978.

TheMann---Whitneytestrevelsp0.019comparingthe

coeffi-cientofvariabilityofHbA1candp0.033comparingthemean HbA1cofthetwogroups,presentingstatisticalsignificance (Fig.3).

Discussion

It is important to notice the high prevalence of co-morbidities in diabetic patients enrolled in this study. Morethan 50%of theinitialdiabetic populationhad arte-rialhypertension,dyslipidemiaorwerealsosmokers.This emphasizes that the role of diet, exercise and regular follow-up of these patients are crucial to achieve the glycemiccontrolandtopreventorminimizetheimpactof co-morbidities.11

Importantly,weshowedthatthetwogroupshavesimilar characteristics,namelyintypeofDMandtypeofdrugsused totreatthepatients.

RelativelytoNKF stages,theprogressionwasmore fre-quentfromstage1to2(49.1%),1to3(25.2%)and2to3 (12.5%).

We demonstrated that fluctuations in HbA1c are asso-ciatedwithrenaldiseaseprogressionindiabetic patients, whichisin accordancewithpreviousstudies4,6,12_.Alsothe

mean HbA1c hasan important role in thissubject. These parameters are central in the management of a diabetic patientbecauseoftheirroletominimizetherenaldisease progression.

Thisstudyhasanumberoflimitations,likethenumber ofpatientsincludedandthereduceddurationoffollow-up. Also,we havenotanalyzedthedurationofDM,whichcan influencethetimingofrenaldiseaseworsening.

Conclusion

Renaldiseaseprogressionindiabeticpatientsisassociated with high mean HbA1c and high coefficient of variability of HbA1c,whichmeans thatit isimportanttoreduce the fluctuations ofHbA1c,in ordertominimizethe renal dis-easeprogressionandtopreventthemorbidityrelatedtoDM complications.13_{Thefluctuationsinglycemiccontrolshould}

be evaluated in the context of the patient’s overall risk profile.

Funding

Nofinancialsupportwasprovided.

Conflict

of

interest

Theauthorshavenoconflictsofinteresttodeclare.

References

1.Diabetologia, S.P. Relatório Anual do Observatório Nacional da Diabetes. Portugal. 11/2014; Available from: http://spd.pt/images/od 2014.pdf

2.Cortez-DiasN,etal.Prevalence,managementandcontrolof diabetesmellitusandassociatedriskfactorsinprimaryhealth careinPortugal.RevPortCardiol.2010;29(4):509---37. 3.PowersAC.DiabetesMellitus:Complications.In:Harrison’s

Prin-ciplesofInternalMedicine.19th ed.McGraw Hill Education; 2015.p.2422---30.

4.CummingsDM,etal.Glycemiccontrolpatternsandkidney dis-easeprogression among primary care patientswithdiabetes mellitus.JAmBoardFamMed.2011;24:391---8.

5.Gomes MB, et al. Variabilidade intra-individual do cont-role glicêmico em pacientes com diabetes tipo 1. Arquivos BrasileirosdeEndocrinologia&Metabologia.2003;47:578---83. 6.WadenJ,etal.A1Cvariabilitypredictsincidentcardiovascular

events,microalbuminuria, andovert diabeticnephropathy in patientswithtype1diabetes.Diabetes.2009;58:2649---55. 7.YuanC,etal.Cumulativeeffectsofhypertension,dyslipidemia,

andchronic kidneydiseaseon carotidatherosclerosisin Chi-nesepatients withtype 2diabetesmellitus. JDiabetes Res. 2014:1---7.

8.AlvarengaC.Hipertensãoarterialnadiabetesmellitustipo2 -Evidênciaparaaabordagemterapêutica.RevPortClinGeral. 2005;21:8.

9.EliassonB.Cigarettesmokinganddiabetes.ProgCardiovascDis. 2003;45:405---13.

11.BanachM,etal.Lipid,bloodpressureandkidneyupdate2013. IntUrolNephrol.2014;46:947---61.

12.ChengD,etal.HbA1Cvariabilityandtheriskofrenal status progressionin Diabetes Mellitus:a meta-analysis. PLoSOne. 2014;9(12):1---13.