www.jped.com.br
ORIGINAL
ARTICLE
Early
changes
in
adipokines
from
overweight
to
obesity
in
children
and
adolescents
夽
Rafael
Machado
Mantovani
a,
Natália
Pessoa
Rocha
b,
Daniel
Massote
Magalhães
b,
Izabela
Guimarães
Barbosa
b,
Antônio
Lúcio
Teixeira
b,c,
Ana
Cristina
Simões
e
Silva
a,b,∗aUniversidadeFederaldeMinasGerais(UFMG),FaculdadedeMedicina,DepartamentodePediatria,BeloHorizonte,MG,Brazil bUniversidadeFederaldeMinasGerais(UFMG),FaculdadedeMedicina,LaboratórioInterdisciplinardeInvestigac¸ãoMédica,
BeloHorizonte,MG,Brazil
cUniversidadeFederaldeMinasGerais(UFMG),FaculdadedeMedicina,DepartamentodeClínicaMédica,BeloHorizonte,MG,
Brazil
Received11September2015;accepted22February2016 Availableonline7August2016
KEYWORDS
Leptin; Adiponectin; Resistin;
Metabolicmarkers; Solubletumor necrosisfactor receptor
Abstract
Objective: Childhoodobesityhasbeenassociatedwithmetabolicsyndromeand cardiovascu-lar diseases. This study aimedto compare plasma levels oftraditional metabolic markers, adipokinesandsolubletumornecrosisfactorreceptortype1(sTNFR1)inoverweight,obeseand leanchildren.Wealsoassessedtherelationshipsofthesemoleculeswithclassicalmetabolic riskfactors.
Methods: This study included 104 children and adolescents, which were grouped as: lean (n=24),overweight (n=30),andobesesubjects(n=50).Theyweresubjected to anthropo-metrical,clinicalandlaboratorialmeasurements.Allmeasurementswerecomparedbetween groups.Correlationanalyseswerealsoperformedtoevaluatetheassociationbetweenclinical data,traditionalmetabolicmarkers,adipokinesandsTNFR1.
Results: Fastingglucose,insulin,homeostaticmodelassessmentofinsulinresistance (HOMA-IR),LDL-cholesterolandtriglycerideswerecomparableinlean,overweightandobesesubjects. PlasmalevelsofsTNFR1weresimilarinleanandoverweightsubjects,butsignificantlyincreased inobesegroup.Leptin,adiponectinandresistinlevelsdidnotdifferwhenoverweightwere com-paredtoobesesubjects.However,alladipokinesdifferedsignificantlywhenleansubjectswere comparedtooverweightandobeseindividuals.Plasmalevelsofadiponectinwerenegatively correlated with bodymassindex (BMI),whereas leptin,resistin and sTNFR1 concentrations positivelycorrelatedwithBMI.
夽
Pleasecitethisarticleas:MantovaniRM,RochaNP,MagalhãesDM,BarbosaIG,TeixeiraAL,SimõeseSilvaAC.Earlychangesinadipokines fromoverweighttoobesityinchildrenandadolescents.JPediatr(RioJ).2016;92:624---30.
∗Correspondingauthor.
E-mail:[email protected](A.C.SimõeseSilva).
http://dx.doi.org/10.1016/j.jped.2016.02.015
Conclusion: Our resultsshowed significantdifferencesincirculating levelsoftheevaluated markerswhenlean,overweightandobeseindividualswerecompared,suggestingthatthese biomarkersmaychangefromleantooverweightandfromoverweighttoobesity.
©2016SociedadeBrasileiradePediatria.PublishedbyElsevierEditoraLtda.Thisisanopen accessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/
4.0/).
PALAVRAS-CHAVE
Leptina; Adiponectina; Resistina; Marcadores metabólicos; Receptorsolúvel defatordenecrose tumoral
Alterac¸õesprecocesnosníveisdeadipocinasdesobrepesoparaobesidade emcrianc¸aseadolescentes
Resumo
Objetivo: Aobesidadenainfânciatemsidoassociadaàsíndromemetabólicaeadoenc¸as cardio-vasculares.Oobjetivodesteestudofoicompararníveisplasmáticosdemarcadoresmetabólicos tradicionais,adipocinasedoreceptorsolúveldefatordenecrosetumoraltipo1(sTNFR1)em crianc¸ascomsobrepeso,obesasemagras.Tambémavaliamosasrelac¸õesdessasmoléculascom fatoresderiscometabólicoclássicos.
Métodos: Esteestudoincluiu104crianc¸aseadolescentes,agrupadosdaseguinteforma: indiví-duosmagros(n=24),comsobrepeso(n=30)eobesos(n=50).Elesforamsubmetidosamedic¸ões antropométricas,clínicaselaboratoriais.Todasasmedic¸õesforamcomparadasentreosgrupos. Tambémforamrealizadasanálisesdecorrelac¸ãoparaavaliaraassociac¸ãoentredadosclínicos, marcadoresmetabólicostradicionais,adipocinasesTNFR1.
Resultados: Glicemiadejejum,insulina,modelodeavaliac¸ãodahomeostasedaresistênciaà insulina(HOMA-IR),colesterolLDLetriglicerídeosforamcomparáveisemindivíduosmagros, comsobrepesoeobesos.OsníveisplasmáticosdesTNFR1foramsimilaresemindivíduosmagros e com sobrepeso, porém significativamente maiores no grupo obeso. Os níveis de leptina, adiponectinaeresistinanãodiferiramquandoindivíduoscomsobrepesoforamcomparadosaos obesos.Contudo,todasasadipocinasdiferiramsignificativamentequandoindivíduosmagros foramcomparadosaindivíduoscomsobrepesoeobesos.Osníveisplasmáticosdeadiponectina estavam negativamentecorrelacionadosao índice demassacorporal (IMC),aopassoque as concentrac¸õesdeleptina,resistinaesTNFR1estavampositivamentecorrelacionadasaoIMC.
Conclusão: Nossos resultados mostraram diferenc¸as significativas nos níveis circulantesdos marcadoresavaliadosaocompararindivíduosmagros,comsobrepesoeobesos,sugerindoque essesbiomarcadorespoderãomudardeindivíduosmagrosparaindivíduoscomsobrepesoede indivíduoscomsobrepesoparaobesos.
©2016SociedadeBrasileiradePediatria.PublicadoporElsevierEditoraLtda.Este ´eumartigo OpenAccesssobumalicenc¸aCCBY-NC-ND(http://creativecommons.org/licenses/by-nc-nd/4.
0/).
Introduction
Childhoodobesityisincreasinginprevalenceandisstrongly associatedwithobesityinadulthood.1,2Obesechildrenare atriskofmetabolicsyndrome,cardiovasculardiseases,and increasedadultmorbidityandmortality.3,4
Adipokines are hormones secreted by adipose tissue that play a role in metabolic homeostasis.5 Obesity also induces productionof inflammatorycytokinesand infiltra-tion of immune cells into adipose tissue, determining a stateofchroniclow-gradeinflammation.Metabolic inflam-mation has been recognized as a unifying mechanism linking obesity to a broad spectrum of conditions, such asatherosclerosis,type 2diabetes, andsystemicvascular complications.5---7
The currentstudy analyzedplasmalevelsofadipokines (adiponectin,resistin,andleptin),andsolubletumor necro-sisfactor receptortype1(sTNFR1),insulinsensitivity,and lipidmetabolisminoverweightandobesechildrenand ado-lescents in comparisonwith leanindividuals. The authors
alsoassessedtherelationshipsofthesemoleculeswith clas-sical metabolic risk factors. Circulating levels of sTNFR1 are regarded as inflammatory markers, reflecting TNF-␣
activitybetterthanthemeasurementofTNF-␣itself.5---9It
washypothesizedthatadipokines,low-gradeinflammation, insulin sensitivity, and lipid metabolism gradually change fromleantooverweightandfromoverweighttoobesity.
Methods
Subjects
Inclusioncriteria
Thecriterionforobesitywasbodymassindex(BMI)greater thanthe95thpercentilefor ageandsexaccordingtothe National Center for Health Statistics (NCHS) standards.10 OverweightwasdefinedasBMIpercentilesforageandsex of>85thand<95th.Theleangroupincludedsubjectswith BMIpercentiles>10thand<85thforageandsex.
Exclusioncriteria
Subjects with genetic disorders, secondary obesity, and comorbidities such as diabetes, cardiac, pulmonary and neurologicaldiseases, or acute infections were excluded. Subjects receiving medical therapy were also ineligible. Healthstatuswasdeterminedthroughmedicalhistoryand eitherparentalreportorself-report toruleoutchronicor acutediseases.
Studyprotocol
At admission,all subjectsunderwent clinical examination including weight,height, BMI, blood pressure, abdominal circumference(AC),andpubertalstageevaluation.Weight wasobtainedtothenearest0.1kgusingcalibrateddigital scales.Standing height was considered as the average of threemeasures,inthenearestmillimeter,usingafixed sta-diometer.ACwasobtainedwithtapemeasureatjustabove theuppermostlateralborderoftherightileum,attheend ofanormalexpiration,andwasrecordedtothenearest mil-limeter,asrecommendedbytheNCHS.11Pubertalstagewas determinedaccordingtoTannercriteria,12andthepatients wereclassified:(i)pre-pubertal:boysatgenitalstageI,girls at breast stage I; (ii) early-puberty boys at genitalstage II---III,girlsatbreaststageII---III;(iii)late/postpubertal:boys atgenitalstage≥IV,girlsatbreaststage≥IV.Systolicand diastolicbloodpressuresweremeasuredonthreedifferent occasions,intherightarm,aftera10-minrestinthesupine positionusingacalibratedsphygmomanometer,andthetwo valueswereaveraged.Bloodpressuremeasurementswere alsoevaluatedfollowingtherecommendationsoftheFourth TaskForceReport.13
Bloodsampling
After informed consent, all subjects were submitted to blood collection for the measurement of fasting glucose, insulin, total cholesterol, high-density lipoprotein (HDL) cholesterol,low-densitylipoprotein(LDL)cholesterol,very low-density lipoprotein (VLDL) cholesterol, triglycerides, adiponectin, leptin, resistin, and sTNFR1. Blood samp-ling occurred at one occasion after an overnight fast of 12h through peripheral venous puncture. One sample was taken for fasting glucose, insulin, total cholesterol, HDL-cholesterol, LDL-cholesterol, and triglycerides mea-surements, according to standard recommendations. The second sample was for adiponectin, leptin, resistin, and sTNFR1measurements. Forthese measurements, samples wereimmediately immersed in ice, and processed within 30min after collection. Cellswere centrifuged at 700×g
for10minat 4◦C; thensupernatant plasmawascollected andre-spunforanother20minat1300×gtosedimentthe
platelets.Cell-freeplasmawasaliquotedinto0.5mL sam-plesandstoredat−80◦Cuntilmeasurements.
Measurements
Fasting glucose, cholesterol, and triglycerides levels were determined by enzymatic colorimetric assay (Advia Chemistry®; Siemens®; LA, USA), while insulin was
mea-suredbychemiluminescence(referencevalue:<29IU/mL; (Immulite® 2000,LA,USA).Theinsulinresistancewas
esti-mated by the homeostasis model assessment for insulin resistance(HOMA-IR).14Plasmalevelsofadiponectin,leptin, resistin,and sTNFR1weremeasured byspecific ELISAkits (R&DSystems®,Minneapolis,USA),followingthe
manufac-turer’sinstructions.Allsampleswereassayed induplicate during a single assay to avoid inter-assay variation. The intra-assay variation was below 3%. The detection limits were5pg/mLforadiponectin,resistin,andleptin;10pg/mL forsTNFR1.
Statisticalanalysis
Statistical analyses were performed using SPSS software version 19.0 (SPSS Inc, Chicago, USA). Sample size was estimated as at least a total of 96 individuals, consider-ing a power of 0.80(1−ˇ), a differenceof 20% between groups, a significance (˛) level of 0.05, and a standard deviation no greater than 25%. All variables were tested for normality of distribution by the Shapiro---Wilk test. Association between dichotomous variables was assessed by chi-squared or Fisher’s exact test. Differences among groups were comparedusing the Kruskal---Wallis test with Dunn’smultiplecomparisonpost-test.Spearmann’s correla-tionanalyses wereperformedtoexaminetherelationship betweenadipokines,sTNFR1levels,andage,BMI,glucose, insulin,HOMA-IR,HDL-cholesterol,LDL-cholesterol,and tri-glycerides.Thesignificancelevelwassetatp<0.05.
Ethicalissues
TheEthicsCommitteeoftheinstitutionapprovedthestudy accordingtotheprotocolETIC0415/06. Informedconsent wasobtainedfromsubjectsandtheirparents.Theresearch protocol did not interfere withany medical prescriptions and follow-up wasguaranteed evenin casesof refusal to participateinthestudy.
Results
Anthropometrical,clinical,andlaboratorial characteristicsinobese,overweight,andlean groups
Table1 Clinicalandlaboratorycharacteristics(mean±SD)forlean,overweight,andobesesubjects.
Characteristic Lean(n=24)A Overweight(n=30)B Obese(n=50)C p-value
Ageinyears 10.93(±2.57) 12.28(±2.61) 10.65(±2.59) 0.23a
Femalegender(%) 54.2 80.0 44.0 0.007b
Tanner’sstage(%) <0.001b
Pre-pubertal 25.0 0 6.0
Early-puberty 41.7 40.0 72.0
Late/postpubertal 33.3 60.0 22.0
BMI 17.39(±3.65) 23.76(±2.64) 28.60(±7.68) 0.001a
A<B<Cc AC/heightratio 0.46(±0.05) 0.56(±0.03) 0.63(±0.08) <0.001a
A<B<Cc
Bloodpressureclassification(%) 0.003b
Normotensive 100 66.7 48.9
Pre-hypertensive 0 10.0 19.1
Hypertensive 0 6.7 31.9
Glucose(mmol/L) 85.15(±8.74) 82.82(±6.12) 82.08(±7.73) 0.22a Insulin(U/mL) 13.31(±30.78) 11.85(±12.99) 15.77(±27.25) 0.14a
HOMA-IR 3.18(±7.69) 2.49(±2.71) 3.50(±6.45) 0.16a
HDL-cholesterol(mg/mL) 55.15(±12.88) 47.38(±9.53) 45.83(±10.63) 0.03a A=Bc;B=Cc; A>Cc LDL-cholesterol(mg/mL) 87.25(±13.78) 99.32(±23.24) 94.22(±30.41) 0.39a Triglycerides(mg/mL) 60.83(±34.83) 71.92(±31.57) 82.50(±41.37) 0.16a sTNFR1plasmalevels(pg/mL)(±SD) 623.75(±347.14) 624.73(±191.30) 855.14(±269.42) <0.001a
A=Bc;A<Cc; B<Cc
BMI,bodymassindex;HDL,high-densitylipoprotein;HOMA-IR,homeostasismodelassessmentforinsulinresistance;LDL,low-density lipoprotein;SD,standarddeviation;sTNFR,solubletumornecrosisfactor.
a Kruskal---Wallistest. b Pearson’schi-squaredtest. c Dunnpost-test.
subjects (p=0.02). In the univariate analysis, the groups differedaccordingtoTanner stage,BMI,AC/height,blood pressureclassification,andHDLcholesterollevels(p<0.05,
Table 1). The obese group had significantly higher BMI, AC/heightratio,andprevalenceofhypertensionthanlean andoverweightsubjects.HDLcholesterollevelswere signifi-cantlylowerinboththeobeseandoverweightgroupsthanin leanindividuals.However,fastingglucose,insulin,HOMA-IR, LDLcholesterol,andtriglycerideswerecomparableinlean, overweight,andobesesubjects(p>0.05,Table1).
As alsoshowninTable1,plasmalevelsofsTNFR1were similarinleanandoverweightsubjects(p>0.05),butwere significantlyincreasedinobesesubjects(p<0.01overweight vs.obeseandp<0.001leanvs.obese).
Fig.1showsplasmalevelsofadiponectin,resistin,and leptininobese,overweight,andleansubjects.Adipokines didnotdifferwhenoverweightwerecomparedtoobese sub-jects(Fig.1).Adiponectinlevels(pg/mL)weredecreasedin overweightandobesesubjectsincomparisonwithlean sub-jects(5199.18inleansubjects;4537.57 intheoverweight group;3467.57intheobesegroup[p<0.001overweightvs. lean and p<0.001 obese vs. lean, Fig. 1]). Plasma levels of resistin(pg/mL) were increasedin the overweight and obesegroupsincomparisonwithleansubjects(1140.38in theleangroup; 1788.51in theoverweightgroup;2256.39
in the obese group [p<0.001 overweight vs. lean and p<0.001 obese vs. lean, Fig. 1]). The same profile was detectedforleptinlevels(702.54pg/mLintheleangroup; 1933.04pg/mL inthe overweightgroup; 1965.91pg/mL in theobesegroup[p<0.001overweightvs.leanandp<0.001 obesevs.lean,Fig.1]).
In order to evaluate gender influence, this study also comparedplasmalevelsofadipokinesinmalesvs.females for each group (lean, overweight, and obese subjects),
Overweight 6000
5000 4000
3000 pg/mL 2000
1000 0
Adiponectin
∗P<.01 (Kruskal–Walis, Dunn’s post-hoc test). ∗∗P<.001 (Kruskal–Walis, Dunn’s post-hoc test).
∗∗ ∗
∗∗ ∗∗
∗∗ ∗∗
Lean ∗∗P<.001
∗P<.01
Obese
Resistin Leptin
Overweight 6000
5000
4000
3000
2000
1000
0
3000
2000
1000
0
3000
2000
1000
0 Lean
∗∗∗
∗ Male
A
B
C
Female
Adiponectin plasma le
ve
ls
(pg/mL
)
Leptin plasma le
ve
ls
(pg/mL
)
Resistin plasma le
ve
ls
(pg/mL
)
∗P<.005
Obese
Overweight
Lean Obese
Overweight Lean
∗P<.01.
∗∗P<.001 Mann-Whitney test.
Obese
∗∗P<.001
Male Female Male Female
∗∗
Figure 2 Gender influence on plasma levels of adipokines (pg/mL) inlean (n=24),overweight (n=30),and obese sub-jects(n=50).A,adiponectin;B,leptin;C,resistin.*p<0.01. **p<0.001Mann---Whitneytest.
as shown in Fig. 2. In the lean group, males presented higher levels of adiponectin (5642.81 vs. 4983.64pg/mL in females, p=0.005) and lower concentrations of leptin thanfemales(371.79pg/mLin malesvs.1391.26pg/mLin females, p<0.001). No other marker differed when com-paringmalesandfemales.Nodifferencesweredetectedin leptin,resistin,andadiponectinlevelsforthecomparisons betweenmaleandfemalesinoverweightandobesegroups (Fig.2).
ToinvestigateapotentialinterferenceofTanner’sstage, thelevels of adipokines in pre-pubertal vs.early puberty vs. late puberty were compared for each group (lean, overweight, and obese individuals). No differences were detectedinresistinandadiponectinlevelsforthe compar-isons between Tanner’s stages classification in all groups. Leptinlevelsexhibitedamildincreaseaccordingtothe pro-gressionofTanner’sstageonlyintheobesegroup(datanot shown).
Correlationanalysis
Considering all subjects (n=104), plasma levels of adiponectin were negatively correlated with BMI
(p<0.001; =−0.402), insulin (p=0.02; =−0.247), HOMA-IR(p=0.02; =−0.263), andtriglycerides (p=0.02;
=−0.261). Leptin levels were positivelycorrelated with age(p<0.001;=0.425),BMI(p<0.001;=0.577),insulin (p<0.001; =0.500), HOMA-IR (p<0.001; =0.497), tri-glycerides(p=0.003; =0.325),andnegatively correlated with HDL-cholesterol (p=0.03; =−0.240). Resistin con-centrationswerepositivelycorrelatedwithBMI(p<0.001;
=0.469), andnegatively correlatedwithHDL-cholesterol (p=0.02; =−0.247). Plasma levels of sTNFR1 were positively correlated with BMI (p<0.001; =0.338) and negatively correlated with adiponectin levels (p=0.03;
p=−0.212).
Discussion
This study showed that circulating levels of adipokines discriminatedleanfromoverweight/obesesubjects. Over-weight and obese individuals presented similar levels of alltheevaluatedadipokines.Remarkably,adiponectin lev-els were decreased while leptin and resistin levels were increased in overweight and obese groups in comparison withleancontrols. Traditionalmetabolicparameterswere similar in all groups, while sTNFR1 levels were increased in the obesein comparison withthe overweightand lean groups.
There is a paucity of data on the role of adipokines in the transition from leanto overweight and from over-weighttoobesity,especiallyin pediatric patientswithout metabolic syndrome or comorbidities. In comparisonwith the control group, it was found that circulating levelsof adipokines are already changed in overweight subjects, whopresentasimilarprofiletotheobesegroup.Similarly, Ko et al. have shown that leptin levels were associated withincreasedrisk of children beingoverweight.17 Leptin is criticalformetabolic regulation bysignaling nutritional statustothehypothalamus,which,inturn,produces neuro-peptidesandneurotransmitterstomodulatefoodintakeand energyexpenditure.5,23 Leptin levelspositively correlated with insulin, HOMA-IR, and triglycerides; they negatively correlatedwithHDLcholesterol.17Johnsonetal.founda sig-nificantassociationbetweenleptinlevelsandincreasingfat massinchildren.24Twopossibleexplanationsarethat resis-tancetoleptinactionsatthereceptorlevelmightcontribute tooverweightandthathomeostaticalterationsprecedethe occurrenceofobesity.25,26
In the present study, plasma levels of adiponectin negativelycorrelatedwithinsulin,HOMA-IR,and triglyceri-des,whileleptinconcentrationswerepositivelycorrelated with the same metabolic parameters. Adamska and co-workersshowedthatinsulinsensitivitypositivelycorrelated to adiponectin and negatively correlated to sTNFR1 and sTNFR2 levels in obese adults. Soluble TNF receptors and adiponectin also negatively correlated with insulin sensitivity,andthesemarkershavemultipleeffectson glu-cose and lipid metabolism in obesity.9 Plasma levels of adiponectin negatively correlated with insulin, HOMA-IR, andtriglycerides21,27 andpositivelycorrelatedwithinsulin sensitivity.9 The present results also showed a negative correlation between sTNFR1 and adiponectin levels, sup-porting the general idea that increased concentrations of adiponectin may compensate low-grade inflammation. Indeed,adiponectinhasanti-inflammatoryeffectsthat con-tribute to its protective role against metabolic stress in obesity.5,28 In this context, increased concentrations of adiponectin may be related to better metabolic profile despite the greater fat mass. However, the majority of studiesincludedobeseadultsalreadyexhibitingsignificant changesin metabolic parameters,while the present sam-pleonlyhadpediatricindividualswithoutco-morbiditiesor metabolic syndrome. Therefore, reductionof adiponectin levels seems to occur later than changes in leptin con-centrations,butmaycomebeforemetabolicalterationsin pediatricobesity.
Circulating levels of resistin have also been positively associatedwithbodyfatmass.29,30Accordingly,astudywith Mexican children and adolescentsfound negative correla-tionbetweenresistinandHDL-cholesterollevels.18Studies have also associated high resistin levels with increased riskformetabolicandcardiovasculardiseases.5Similarlyto adiponectin,the elevationof resistinlevels probably pre-cedesco-morbiditiesrelatedtoobesityinpediatricpatients. The authors areawareof the limitations of thisstudy. First, the cross sectional design precludes the evaluation of interactions amongvariables. Second, the sample was relatively small, mostly considering overweight and lean subjects. Third, the use of a convenience sample makes homogeneity among the selected groups very difficult to
obtain,sincedifferencesingenderandTanner’sstagemay change adipokine levels. However, no differences were detectedinresistinandadiponectinlevelsforthe compar-isonsbetweenmaleandfemalesandbetweenTanner’sstage intheoverweightandobesegroups.Leptinlevelsonly exhib-itedamild elevationaccordingtoTanner’sstageinobese group.Nevertheless,otheraspectsincreasethestrengthof thefindings,suchasstrictinclusionandexclusioncriteria, well-establishedprotocol for measurements, and a repre-sentativenumberofobeseindividuals.
Inconclusion,thefindingssuggestthatchangesinplasma levelsofadipokines(leptin,resistin,andadiponectin)may come before alterations in traditional metabolic markers of obesity in pediatric patients. Longitudinal studies with largersamplesare necessary toelucidate howadipokines interactwithclinical,metabolic,andinflammatorymarkers inpediatricobesity.
Funding
ThisworkwassupportedbygrantsfromtheResearch Foun-dationoftheStateofMinasGerais(FAPEMIG)andfromthe National Council of Scientific and Technological Develop-ment(CNPq).
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
References
1.EllsLJ,HancockC,CopleyV,MeadE,DinsdaleH,KinraS,etal. PrevalenceofseverechildhoodobesityinEngland:2006---2013. ArchDisChild.2015;100:631---6.
2.WHO.Obesity:preventingandmanagingtheglobalepidemic: reportofaWHOconsultation.WorldHealthOrganTechRepSer. 2000;894:1---253.
3.FreedmanDS,KhanLK,SerdulaMK,DietzWH,SrinivasanSR, BerensonGS.TherelationofchildhoodBMItoadultadiposity: theBogalusaheartstudy.Pediatrics.2005;115:22---7.
4.VanEmmerikNM,RednersCM,vandeVeerM,vanBuurenS,van derBaan-SlootwegOH,Kist-vanHoltheJE,etal.High cardio-vascularriskinseverelyobeseyoungchildrenandadolescents. ArchDisChild.2012;97:818---21.
5.Cao H. Adipocytokines in obesity and metabolic disease. J Endocrinol.2014;220:T47---59.
6.MantovaniRM, RiosDR,Moura LC,Oliveira JM,CarvalhoFF, CunhaSB,etal.Childhoodobesity:evidenceofanassociation betweenplasminogenactivatorinhibitor-1 levelsandvisceral adiposity.JPediatrEndocrinolMetab.2011;24:361---7. 7.NagelG,RappK,WabitschM,Büchele G,KrokeA, ZöllnerI,
et al. Prevalence and cluster ofcardiometabolic biomarkers inoverweightandobeseschoolchildren:resultsfromalarge surveyinsouthwestGermany.ClinChem.2008;54:317---25. 8.AderkaD.Thepotentialbiologicalandclinicalsignificanceof
thesolubletumornecrosisfactorreceptors.CytokineGrowth FactorRev.1996;7:231---40.
10.Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standarddefinitionforchildoverweightandobesityworldwide: internationalsurvey.BrMedJ.2000;320:1240---3.
11.National Health and Nutrition Examination Survey. Anthro-pometryproceduresmanual.Atlanta,GA:CentersforDisease Control and Prevention; 2000. Available at: http://www. cdc.gov/nchs/data/nhanes/bm.pdf[cited16.05.15].
12.TannerJM.Growth and maturationduringadolescence. Nutr Rev.1981;39:43---55.
13.NationalHighBloodPressureEducationProgramWorkingGroup onHighBloodPressureinChildrenandAdolescents2004.The fourth reporton thediagnosis,evaluation, and treatmentof high bloodpressure in children and adolescents. Pediatrics. 2004;114:555---76.
14.MatthewsDR,HoskerJP,RudenskiAS,NaylorBA,TreacherDF, Turner RC. Homeostasismodelassessment: insulinresistance andbeta-cellfunctionfromfastingplasmaglucoseandinsulin concentrationsinman.Diabetologia.1985;28:412---9.
15.SpeiserPW,RudolfMC,AnhaltH,Camacho-HubnerC,Chiarelli F,EliakimA,etal.Childhoodobesity.JClinEndocrinolMetab. 2005;90:1871---87.
16.ReinehrT, Woelfle J,Roth CL.Lack of associationbetween, insulinresistance,cardiovascular riskfactors,and obesity in children:alongitudinalanalysis.Metabolism.2011;60:1349---54. 17.KoBJ,LeeM,ParkHS,HanK,ChoGJ,HwangTG,etal. Ele-vatedvaspinand leptinlevelsare associatedwithobesityin prepubertalKoreanchildren.EndocrJ.2013;60:609---16. 18.HuangF,Del-Río-NavarroBE,Pérez-OntiverosJA,Ruiz-Bedolla
E,Saucedo-RamírezOJ,Villafa˜naS,etal.Effectofsix-month lifestyleinterventiononadiponectin,resistinandsolubletumor necrosis factor-␣ receptors in obese adolescents. Endocr J. 2014;61:921---31.
19.Volberg V, Heggeseth B, Harley K, Ruiz-Bedolla E, Saucedo-Ramírez OJ, Villafa˜na S, et al. Adiponectin and leptin trajectoriesinMexican-Americanchildrenfrombirthto9years ofage.PLOSONE.2013;8:e77964.
20.Klünder-KlünderM,Flores-HuertaS,García-MacedoR, Peralta-RomeroJ,CruzM.Adiponectinineutrophicandobesechildren asabiomarkertopredictmetabolicsyndromeandeachofits components.BMCPublicHealth.2013;13:88.
21.Codo˜ner-FranchP,Tavárez-AlonsoS,Porcar-AlmelaM, Navarro-SoleraM,Arilla-Codo˜nerÁ, Alonso-IglesiasE.Plasmaresistin levelsare associated withhomocysteine, endothelial activa-tion, and nitrosative stress in obese youths. Clin Biochem. 2014;47:44---8.
22.NemetD,WangP,FunahashiT,MatsuzawaY,TanakaS, Engel-manL,etal.Adipocytokines,bodycomposition,andfitnessin children.PediatrRes.2003;53:148---52.
23.BarbosaJA,RodriguesAB,MotaCC,BarbosaMM,SimõeseSilva AC.Cardiovascular dysfunctioninobesityand newdiagnostic imagingtechniques:therole ofnoninvasive image methods. VascHealthRiskManag.2011;7:287---95.
24.Johnson MS, Huang TT, Figueroa-Colon R, Dwyer JH, Goran MI.Influenceofleptin onchangesinbodyfat duringgrowth in African American and white children. Obes Res. 2001;9: 593---8.
25.AkinciG, AkinciB,Coskun S,BayindirP,HekimsoyZ,Ozmen B. Evaluation ofmarkers of inflammation, insulin resistance andendothelialdysfunctioninchildrenatriskforoverweight. Hormones(Athens).2008;7:156---62.
26.Rizzo AC, Goldberg TB, Silva CC, Kurokawa CS, Nunes HR, Corrente JE. Metabolicsyndrome riskfactors in overweight, obese,andextremelyobeseBrazilianadolescents.NutrJ.2013; 12:19.
27.Wardaningsih E, Miida T, Seino U, Fueki Y, Ito M, Nagasaki K,et al.Low adiponectinstate isassociated withmetabolic abnormalities in obese children, particularly depend-ing on apolipoprotein E phenotype. Ann Clin Biochem. 2008;45:496---503.
28.SuSC,PeiD,HsiehCH,HsiaoFC,WuCZ,HungYJ.Circulating pro-inflammatorycytokinesandadiponectininyoungmenwith type2diabetes.ActaDiabetol.2011;48:113---9.
29.OrtegaL,RiestraP,NavarroP,Gavela-PérezT,Soriano-Guillén L,GarcésC.Resistinlevelsarerelatedtofatmass,butnotto bodymassindexinchildren.Peptides.2013;49:49---52. 30.Vozarova de Courten B, Degawa-Yamauchi M, Considine RV,
