metabolic syndrome features in android obese subjects

(1)

SHORT REPORT

Weight loss improves serum mediators and

metabolic syndrome features in android obese subjects

Ching-Ya Chou

^a^,^b

, Hui-Fen Lang

^a^,^b

, Wanye Huey-Herng Sheu

^b^,∗

, Jin-Yuarn Lin

^a^,∗∗

aDepartmentofFoodScienceandBiotechnology,NationalChungHsingUniversity,250Kuo-KuangRoad, Taichung40227,Taiwan,ROC

bTaichungVeteransGeneralHospital,No.160Sec.3,Chung-KangnRoad,Taichung40705,Taiwan,ROC

Received20April2011 ;receivedinrevisedform14September2011;accepted3October2011

KEYWORDS Androidobesity;

Interleukin(IL)-6;

IL-10;

Metabolicsyndrome;

Tumornecrosisfactor (TNF)-␣

Summary Serumanti-/pro-inﬂammatorymoleculessuchasadiponectin,IL-6,IL- 10,andTNF-␣,andmetabolicsyndrome(MetSyn)featuresin15androidobese(6 MetSynand9non-MetSyn)subjectswereassessedduringan8-weekweightcontrol program. The resultsshowedthat thebody massindex,weight,leanbodymass, triglyceride,totalcholesterol/highdensitylipoproteincholesterolratio,andTNF-␣

inMetSynsubjectsweresigniﬁcantly(P<0.05)improved.Thisstudysuggeststhat weightreductionin androidobesesubjectsmaybebeneﬁcialin reducingcardio- vascular diseases via improving serumIL-6and TNF-␣ levels,aswell asMet Syn features.

Obesity causes physiological changes in the body, accompanied with Met Syn features, although obesity is caused by multiple factors. Obese subjects deposit excessive adipose cells in the body. Unfortunately, these adipose cells may further secret pro-inﬂammatory cytokines, such as

∗Correspondingauthor.Tel.:+886423592525;

fax:+886423595046.

∗∗ Correspondingauthor.Tel.:+886422851857;

fax:+886422851857.

E-mailaddresses:[email protected] (W.H.-H.Sheu),[email protected](J.-Y.Lin).

tumornecrosisfactor(TNF)-␣andinterleukin(IL)- 6[1].Secretedpro-inflammatorycytokinesfurther induce a macrophage accumulation, resulting in the over-production of TNF-␣, IL-6 and mono- cyte chemo-attractant protein-1 (MCP-1) by the recruited macrophages, causing chronic diseases due to inflammation damage [2]. In addition, obesesubjectsaregenerallyaccompaniedwithglu- coseintolerance,suggestingthatpro-inflammatory cytokinesmaycauseinsulinresistanceandcardio- vasculardiseases(CVD).

Adiposefat invisceral obesityis astronger risk factor for CVD than subcutaneous fat [3]. It has

doi:10.1016/j.orcp.2011.10.003

(2)

been found that visceral obese subjects, regard- lessofanormalwaistcircumference(WC),havea higher carotid artery far-wall intima-media thick- ness compared with those with increased WC, but less visceral fat, suggesting that a direct estimation for visceral fat may be necessary in assessing atherosclerotic burdenin menwithtype 2 diabetes [4]. In contrast to pro-inflammatory cytokines,adiposetissuesalsosecretaproteinhor- mone,adiponectin,thatplaysananti-inflammatory roleviainhibitingpro-inflammatorycytokinesTNF-

␣ and IL-6, but increasing an anti-inﬂammatory cytokineIL-10productionsbymacrophagesinvitro [5].Adiponectinexclusivelysecretedfromadipose tissue into the bloodstream modulates metabolic processes including glucose regulation and fatty acidcatabolism[6].Adiponectinisnegativelyasso- ciated with Met Syn in middle aged and elderly ChineseindependentlyofBMI,physicalactivityand life habits[7]. Undoubtedly, hypoadiponectinemia is a risk factor for Met Syn [8], and is associated with indices of subclinical atherosclerosis, suchasintimamediathicknessandarterialcompli- anceinobesepatients[9].Therefore,apernicious balance between pro-inﬂammatory cytokines and adiponectin secreted by adipose tissues in obese subjectsdeterioratemetabolicsyndromefeatures.

Android obesity plays an important role in Met Syn. Abdominal fat seems to release some activating factors that modulate metabolism in vivo. Android obesesubjectshave lower serum adiponectin and IL-10 levels compared to normal subjects[10].MetSynsubjectsseemtohavelower plasmaIL-10[11],buthigherIL-6levels[12].How- ever,the realrole ofIL-6in MetSynhasnotbeen clariﬁed, yet. This study tried to investigate the relationship amongadiponectin,IL-10,TNF-␣,IL-6 andMetSyn.

This study analyzed the effects of an 8-week weight control program on serum inflammatory markersandtheriskfactorsforMetSyninandroid obese subjects.Based on the WHOcriteria (2004) for androidobesity in the Asia-Pacificregion[13], android obese subjects were recruited for the program at Taichung Veterans General Hospital, Taiwan, ROC. Allapplicable institutional and gov- ernmental regulations concerning the ethical use of human volunteers were approved by the Insti- tutional Review Board (IRB) of Taichung Veterans General Hospital at the 61st committee meeting of the IRB. Fifteen obese subjects including four males and eleven females were qualified for this study. Anthropometric evaluation was measured using traditionalmethod.Biochemicalmarkersfor MetSynfromparticipantsubjectsweredetermined using a colorimetric method. Met Syn or non-Met

Syn subjects were primarily identiﬁed with the updated National Cholesterol Education Program AdultTreatment Panel III (NCEP ATP III,2001) cri- teria[14],and aslight adoptionbyDepartmentof Health,Taiwan,ROC.DeﬁnitionofMetSynrequires the presence of at least three of the following:

WC≥90cmformen,or≥80cmforwomen;triglycerides (TC) ≥150mg/dl; high density lipoprotein (HDL) cholesterol <40mg/dl for men or<50mg/dl forwomen;bloodpressure≥130/85mmHg;fasting glucose≥100mg/dl [15].Tocompare intervention effects of the 8-week weight control program on MetSynandnon-MetSynsubgroups,datacollected were also subdivided into twosubgroups (Met Syn versusnon-MetSyn)forcomparison.

The intervention of the 8-week weight control programwhichcombineddietaryguidanceandaer- obicexercisetrainingwasimplemented.Adietician inthehospitalcorrectedthe3-daydietrecordand taught the participant subjectsto select low-fat, low-sugar,low-salt,high-fiber,andlow-caloriediet asusualfoods[16].Thesubjectswerealsoencour- agedtotakeregularexercisetoincreasetheweight loss. Changes in weight loss, WC, body fat, blood pressure,fastingplasmaglucoselevel,and fasting serumlipidprofiles,inflammatorycytokinesinclud- ing IL-6 and TNF-␣, as well as anti-inflammatory cytokines including adiponectin and IL-10, were calculated as the difference between the partici- pantsubjects’baseline(week0)andthelastweek of attendance (week 8). Plasma fasting glucose levels andserumlipidsincludingTG, totalcholesterol(TC),andHDLwereprocessedbytheCentral Laboratory Services at Taichung Veterans General Hospital using enzymatic methods. The body fat was measured using a machine with tetra-polar segmentalbioelectrical impedanceanalysis (SBIA) (Zeus 9.9 Jawon medical Co. Ltd., Seoul, Korea).

Baselinedatawerecollectedwithin1weekpriorto theprogram.Post-programsampleswereobtained within 1 week after the program ﬁnished [15].

Serum adiponectin, TNF-␣, IL-6, and IL-10 levels weredeterminedusingsandwichELISAkits,respectively.The adiponectinconcentrationwasassayed usinganELISAkit(QuantikineHumanadiponectin, R&DSystems,Inc.,Minneapolis,MN,USA).TheTNF-

␣,IL-6,andIL-10concentrationswereassayedusing high-sensitivity ELISA kits (Quantikine HS Human TNF-␣, IL-6, IL-10, R&D Systems, Inc., Minneapo- lis, MN,USA).Thesensitivityofadiponectin assay was about <15.6pg/ml. The sensitivity of TNF-

␣, IL-6,and IL-10 assays was about <0.039pg/ml, 0.039pg/ml,and0.500pg/ml,respectively.

Valueswereexpressedasmeans±SD.Datawere ﬁrst analyzed using the Kolmogorov—Smirnov test torecognizeanormaldistribution,followedbythe

(3)

Table1 Anthropometricandbiochemicalcharacteristicsofallobesesubjectsaswellastheirgroupingincluding metabolicsyndrome(MetSyn)andnon-MetSynobesesubjects.

Variable Allsubjects

(n=15)

MetSyngroup (n=6)

Non-MetSyn group(n=9)

P-Value(MetSyn versusnon-Met Synsubjects)

Age(years) 41±11 44±10 39±11 0.346

Gender(M/F) 4/11 2/4 2/7 0.645

Waistcircumference(cm) 101±7.5 99±7.5 103±7.3 0.154

Weight(kg) 79±13 78±16 80±11 0.480

Body-massindex(kg/m²) 30.3±3.5 29.3±2.8 30.9±3.9 0.596

Bodyfat(%) 34.1±4.2 33.1±3.1 34.7±4.8 0.346

Subcutaneousfat(kg) 22.4±4.0 21.3±3.2 23.1±4.5 0.556

Visceralfat(kg) 4.38±1.03 4.15±0.88 4.53±1.14 0.515

Leanmass(kg) 47.6±8.9 47.8±11.8 47.4±7.2 0.768

Fatmass(kg) 26.8±5.0 25.5±4.0 27.6±5.7 0.479

Waist/Hipratio(WHR) 0.9±0.04 0.91±0.05 0.90±0.03 0.906

SystolicBP(mmHg) 126±13 133±14 121±10 0.099

DiastolicBP(mmHg) 79±12 82.2±12.9 76.9±11.5 0.515

Fastingglucose(mg/dl) 97.3±27.2 111±47 87.9±6.1 0.238

HDLcholesterol(mg/dl) 47.1±9.9 43±6 49.8±11.3 0.214

Triglycerides(mg/dl) 162±89 224±87 120±67 0.013^*

Adiponectin(␮g/ml) 1.26±0.52 1.32±0.5 1.23±0.56 0.814

IL-10(pg/ml) 1.79±1.66^a ND 1.79±1.66 0.000^*

TNF-␣(pg/ml) 1.27±0.59 1.48±0.56 1.13±0.59 0.346

IL-6(pg/ml) 2.76±1.63 1.73±0.90 3.33±1.75 0.053

Valuesaremeans±SD.Themeasurementswere,respectively,madeattheﬁrstdaybeforetheweightcontrolprogram.Statistical signiﬁcancewascalculatedbyMann—WhitneyUtest.ND,notdetectable.

aTheIL-10valueswereobtainedfrom5subjectsinthenon-MetSyngroup.TheminimumdetectabledoseofhumanIL-10ELISA kitusedinthisstudyistypicallylessthan0.5pg/ml.

*P<0.05.

Pearsoncorrelation for a two-tailed test between two variables. Changes in all subjects between the beginning and post-treatment were analyzed usingpairedStudent’st-test.Intra-changesinMet Synor non-Met Synsubgroups betweenthe beginning and post-treatment were analyzed using the WilcoxonSigned-Ranktest.Inter-changesbetween Met Syn and non-Met Syn subgroups through the programwereanalyzed usingthe Mann-WhitneyU test.Differencesbetweengroupswereconsidered statistically signiﬁcant if P<0.05. Statistical tests wereperformedusingSPSSversion11.5(SPSS,Inc., Chicago,IL,USA).

Results

To characterize participant subjects of the program,anthropometricandbiochemicalcharacteris- ticsofallobesesubjects,aswellastheirsubgroups including MetSyn and non-Met Syn subjectswere surveyed before the weight control program was administered (Table 1). There were no signiﬁ- cantdifferencesinanthropometricandbiochemical

characteristics,exceptTGandIL-10,betweenMet Synandnon-MetSynsubgroups.However,MetSyn- positivesubjectshadsigniﬁcantlyhigherTGlevels, but slightlylowerIL-6 levels in serathan those in non-Met Synsubjects. Particularly,the IL-10 level inMetSynsubjectswastoolowtobedetectable.

To determine the correlation among serum pro-inflammatory,anti-inflammatorymediatorsand Met Syn features in subjects, the association amongserumpro-inflammatory, anti-inflammatory mediators and body compositions of all subjects beforethe weight controlprogramwere analyzed (Table 2). Serum adiponectin concentrations in all subjects showed statistically (P<0.05) nega- tivecorrelationswithBMI,bodyweight,leanbody mass,subcutaneousfat,visceralfat,bodyfat,and WC.Asignificantlynegativecorrelationalsoexisted betweenserumTNF-␣andHDLconcentration,but astatisticallypositivecorrelationexistedbetween serumTNF-␣ and blood pressure(BP). SerumIL-6 concentrationexhibitedstatistically(P<0.05)posi- tivecorrelationswiththesubcutaneousfat,visceral fat,bodyfatandWC,suggestingthatexcesslevels ofIL-6 mayresult fromadipose tissuesin android obesesubjects. However,there wasno significant

(4)

Table2 Correlationsamongindividualinﬂammatorymediatorsinseraandbodycompositionsofallsubjectsbefore theweightcontrolprogram.

R Adiponectin(n=15) IL-10^a(n=5) TNF-␣(n=15) IL-6^b(n=14)

BMI(kg/m²) −0.693^* −0.412 0.341 0.491

Bodyweight(kg) −0.672^** 0.092 0.463 0.499

Leanmass(kg) −0.583^* 0.700 0.501 0.304

Subcutaneousfat(kg) −0.547^* −0.652 0.192 0.606^*

Visceralfat(kg) −0.526^* −0.584 0.205 0.658^*

Bodyfat(%) −0.546^* −0.641 0.196 0.620^*

WHR −0.459 0.369 0.359 0.449

Waistcircumference(cm) −0.514^* −0.423 0.326 0.637^*

TG(mg/dl) 0.038 0.372 0.210 −0.329

HDLcholesterol(mg/dl) 0.416 0.105 −0.669^** −0.028

SystolicBP(mmHg) 0.071 0.819 0.519^* 0.232

Fastingglucose(mg/dl) −0.200 −0.166 0.453 −0.137

Lean/fatmass −0.115 0.832 0.300 −0.153

Cholesterol(mg/dl) 0.205 −0.472 −0.096 0.293

TC/HDL −0.183 −0.316 0.388 −0.108

Adiponectin(␮g/ml) 1 −0.664 −0.171 −0.354

IL-10(pg/ml) −0.182 1 0.273 0.443

TNF-␣(pg/ml) −0.246 0.468 1 −0.111

IL-6(pg/ml) −0.362 0.458 −0.111 1

AlldataareanalyzedbyPearsoncorrelation.BMI,bodymassindex;WHR,waist/hipratio.

aTheIL-10valueswereobtainedfrom5subjectsinthenon-MetSyngroup.

bTheIL-6valueswereobtainedfrom8subjectsinthenon-MetSyngroup.

* MeanssigniﬁcantatthelevelofP<0.05.

**MeansigniﬁcantatthelevelofP<0.01.

associationamongMetSynfeaturesandserumIL-10 levels.

The Met Syn features of all subjects, including WC, diastolic BP, and TC were significantly improved(P<0.05)throughthe8-weekweightcontrol program (Table 3). However, serum anti- or pro-inflammatory mediators, such as adiponectin, IL-10, IL-6and TNF-␣,didnot significantlychange (P>0.05),although anti-inflammatoryadiponectin and IL-10 slightly increased, pro-inflammatory IL- 6 and TNF-␣ slightly decreased after the 8-week weight control program (Table 3). In non-Met Syn subjects, the weight, BMI, lean body mass (Table 4), WC and TC (Table 5) were also significantly improved. The improvement in MetSyn features,suchasWCandTClevels innon-MetSyn subjectswasbetterthanthoseinMetSynsubjects (Table5).Interestingly,serumTNF-␣levelsinnon- MetSynsubjectssignificantlyincreasedthroughthe 8-week weight control program, while IL-10 also slightlyincreased(Table5).

Discussion

We hypothesized that Met Syn responders them- selves may hinder the remission of Met Syn.

In general, Met Syn may result in an increased risk for developing type 2 diabetes, cardiovascu- lar diseases, psychiatric comorbidity, stress, and a reduced quality of life [17]. The 8-week program indeed achieved some outcomes to both Met Syn and non-Met Syn subjects, although the weight control program was still a short- term program. Unfortunately, serum anti- or pro-inflammatory mediators, such asadiponectin, IL-10, IL-6 and TNF-␣, did not markedly change, although anti-inflammatory adiponectin and IL- 10 slightly increased and pro-inflammatory IL-6 and TNF-␣ slightly decreased after the 8-week weightcontrolprogram(Table3).We suggestthat the weight loss intervention for remission of Met Syn and inflammatory mediators in obese subjects shouldbe extendedfor atleast 15weeks to avoid significant decrease in lean mass [15], but to increase anti-inflammatoryadiponectin and IL- 10and decrease pro-inflammatoryIL-6 and TNF-␣

levels. Nicklas et al. (2005) reported that inflammatory mediators in obese subjects may have a significantdecreasebyweightlossthroughbehav- ior modification for 12 months [18]. In addition, a significant weight loss via weight loss intervention is required, too. Previous studies performed in patients after gastric by-pass show a signifi- cantdecreaseinalltheseparametersaftermassive

(5)

Table3 Effectsofthe8-weekweightcontrolprogramonthebodycompositions,MetSynfeaturesandinﬂamma- torymediatorsofobesesubjects.

Variable (n=15)

Before After Change P

Weight(kg) 78.8±12.5 76.6±12.5 −2.19±1.65^** <0.001

BMI(kg/m²) 30.3±3.5 29.4±3.4 −0.82±0.60^** <0.001

Fatmass(kg) 26.8±5.0 25.7±5.2 −1.02±1.49^* 0.019 Bodyfat(%) 34.1±4.2 33.7±4.2 −0.43±1.27 0.206 WHR 0.90±0.04 0.89±0.05 −0.01±0.02 0.084

Subcutaneousfat(kg) 22.4±4.0 21.6±4.2 −0.82±1.11^* 0.013

Visceralfat(kg) 4.38±1.03 4.18±1.09 −0.20±0.40 0.062

Leanmass(kg) 47.6±8.9 46.5±8.5 −1.07±0.84^** 0.001

Lean/fatmass 1.81±0.37 1.85±0.39 +0.04±0.11 0.207

Waistcircumference(cm) 101±7.5 97.3±7.4 −4.1±2.5^** 0.001

SystolicBP(mmHg) 126±13 122±16.8 −3.07±14.59 0.429

DiastolicBP(mmHg) 79±12 74.2±11.2 −4.73±7.65^* 0.031

Fastingglucose(mg/dl) 97.3±27.2 94.1±13.8 −3.2±6.6 0.108

Cholesterol(mg/dl) 204±38 189±30 −17.3±22.2^* 0.002

HDLcholesterol(mg/dl) 47.1±9.9 45.1±7.9 −1.93±4.01 0.830

Triglycerides(mg/dl) 162±89 156±78 −6.27±36 0.512

TC/HDL 4.49±1.18 4.35±1.15 −0.14±0.38 0.165

Adiponectin(␮g/ml)(n=15) 1.26±0.52 1.37±0.71 +0.11±0.44 0.350

IL-10(pg/ml)(n=5) 1.79±1.66 1.89±1.21 +0.08±2.79 0.276

IL-6(pg/ml)(n=14) 2.76±1.63 2.65±1.7 −0.10±1.39 0.950

TNF-␣(pg/ml)(n=15) 1.27±0.59 1.16±0.64 −0.11±0.67 0.540

Valuesaremeans±SD.StatisticalsigniﬁcancewasanalyzedbypairedStudent’st-test.Themeasurementswere,respectively, madeattheﬁrstdaybeforeandaftertheweightcontrolprogram.Minussign(−)representsadecrease;plussign(+)represents anincrease.

*P<0.05.

**P<0.01.

weightloss,includingresistin,adiponectin,ghrelin, leptin, and pro-inflammatory cytokines [19]. The participantsubjectsthroughtheprogramonlylost about2kg(Table3),resultinginnosignificantdif- ferencesininflammatorycytokines.

It has been found that hypoadiponectinemia is ariskfactorforMetSyn[8],and maybe anindex forsubclinicalatherosclerosis,suchasintimamedia thicknessandarterialcomplianceinobesepatients [9]. In this study, we found that the adiponectin concentrations in all participant subjects demon- stratedstatistically(P<0.05)negativecorrelations with BMI, body weight, lean body mass, subcutaneousfat,visceralfat,bodyfat,andWC(Table2), suggesting that mild or severe obesity also may cause hypoadiponectinemia. Unfortunately, the short-term weight control program could not sig- niﬁcantlychangeserumadiponectinlevelsinobese subjects(Table5).

It is reported that obesity strongly associated with increased circulating TNF-␣ levels, while weightlosslowerssystemiclevels[20].Bodyweight reductioninobesesubjects(BMI=39.3±2.2kg/m²) which resulted in improved insulin sensitivity was

also associated with a decrease in TNF-␣ mRNA expressioninfattissue[21].TNF-␣isalsoexpressed atahigher level in themuscletissue frominsulin resistant and diabetic subjects [22]. Our results alsoshowedthatTNF-␣levels in MetSynsubjects weresignificantly(P<0.05)decreasedthroughthe weight loss program (Table 5). The results are in accordance with previousstudies [20—22], suggesting that weight loss in Met Syn patients may effectivelydecrease systemiclow-level inflammation [20]. Particularly, we first found that serum TNF-␣ levels in non-Met Syn subjects markedly (P<0.05)increasedafterthe8-weekweightcontrol program(Table5). Changesin serumTNF-␣ levels between Met Syn patients and non-Met Syn subjects were quite different. We hypothesized that aerobicexercise innon-Met Synsubjectsmaypro- ducepro-inflammatorycytokinessuchasTNF-␣or IL-6athigherlevelsthanthoseinnon-MetSynsub- jects, although the values may be considered as minor increases. However, the immunoregulatory mechanism on circulating TNF-␣ levels in normal ornon-MetSynsubjectsduringweightlossthrough exerciseremainstobe furtherclarified. Allserum

(6)

Table4 Effectsofthe8-weekweightcontrolprogramonbodycompositionsinMetSynandnon-MetSynsubgroups.

Variable MetSyngroup(n=6)

Weight(kg) 77.7±16.1 75.7±16.8 −1.98±1.88^* 0.049

BMI(kg/m²) 29.3±2.8 28.6±3.3 −0.75±0.69^* 0.045

Fatmass(kg) 25.5±4.02 24.4±4.9 −1.07±1.68 0.180 Bodyfat(%) 33.1±3.1 32.5±3.2 −0.62±1.26 0.283 WHR 0.91±0.05 0.90±0.07 −0.02±0.02 0.076

Subcutaneousfat(kg) 21.3±3.2 20.5±3.8 −0.87±1.28 0.158

Visceralfat(kg) 4.15±0.88 3.95±1.14 −0.2±0.4 0.275

Leanmass(kg) 47.8±11.8 46.9±11.5 −0.85±0.42^** 0.004

Lean/fatmass 1.87±0.28 1.93±0.27 +0.06±0.12 0.344

Variable Non-MetSyngroup(n=9)

Weight(kg) 79.5±10.5 77.2±9.8 −2.32±1.58^** 0.002

BMI(kg/m²) 30.9±3.9 30.0±3.6 −0.87±0.56^** 0.002

Fatmass(kg) 27.6±5.7 26.6±5.5 −0.99±1.47 0.078 Bodyfat(%) 34.7±4.8 34.4±4.8 −0.31±1.34 0.502 WHR 0.90±0.03 0.89±0.03 −0.01±0.02 0.274

Subcutaneousfat(kg) 23.1±4.5 22.3±4.4 −0.78±1.07 0.059

Visceralfat(kg) 4.53±1.14 4.33±1.10 −0.21±0.41 0.167

Leanmass(kg) 47.4±7.2 46.2±6.6 −1.21±1.03^** 0.008

Lean/fatmass 1.77±0.44 1.80±0.46 +0.03±0.11 0.440

Valuesaremeans±SD.StatisticalsigniﬁcancewasassayedbyWilcoxonsigned-ranktest.Themeasurementswere,respectively, madeattheﬁrstdaybeforeandaftertheweightcontrolprogram.Minussign(−)representsadecrease;plussign(+)represents anincrease.

* P<0.05.

**P<0.01.

cytokine except adiponectin levels are extremely lowinnormalbutnotsepticsituations.Therefore, adiponectin plays an anti-inflammatory role via inhibiting pro-inflammatory cytokines TNF-␣ and IL-6,but increasingananti-inflammatorycytokine IL-10productions.Inourpreviousstudy,asubgroup analysis showed that obese subjects whose origi- nal BMI was <30kg/m² had significantly increased serumadiponectinlevelsduringweightlossprocess [23]. We assume that adiponectin, but not IL-10, playsamainroleinmetabolicsyndrome.

There are somelimitations to this study. First, numbersofparticipantsubjectswerelimited.Sec- ond, the 8-week weight loss program is still a short-termfor signiﬁcantweight reduction.Third, many variables, such as waist, adiponectin, and HDL-cholesterol, areheavilygender-dependent.If possible, theyshouldbeanalyzed gender-speciﬁc.

However, this study population is too small to allow this. There is a big difference by gender in the number of patients recruited in this study. In our previous study, we discovered that the change patterns ofmostselectedbiomarkers, includingadiponectinandTNF-␣,in eitherwomen ormenobesesubjectswerequitesimilarduringthe

weight-lossprocess[23].Therefore,we pooledall datafromwomenandmensubjectstominimizethe possibledeviationduetothenumberlimitofatotal of4obesemenin this study. Fourth, thesubjects might not strictly followthe dietary and exercise program. Therefore,themagnitudeofweightloss through the program on serum markers has been discussedinanotherarticle[23].Largerandlonger- term studies should be conducted in the future.

Nevertheless,we achievedsomeimportantresults in the present study. Undoubtedly, obesity is the most important risk for the development of diabetes and predisposes individuals to hypertension and dyslipidaemia [24]. This study exhibited that obese subjects through a short-term weight control program may have a great beneﬁt to reduce CVDviaimprovingserumIL-6andTNF-␣levels,as wellasMetSynfeatures.Inaddition,thisstudyhas shownprofoundsigniﬁcanceofmetabolicsyndrome in Asian people. However, it should be paid care- ful attention to racialdifferences when we make an interpretation of the data provided. In general, visceral obesity and/or insulin resistance is generally upstream of metabolic syndrome. How- ever,theimpairmentofinsulinsecretionoftenhas

(7)

Table5 Effectsof the8-week weightcontrolprogramonMetSynfeaturesandinﬂammatorymediatorsinMet Synandnon-MetSynsubgroups.

Variable MetSyngroup(n=6)

Waistcircumference(cm) 98.7±7.5 96.1±8.7 −2.58±1.36^** 0.006

SystolicBP(mmHg) 133±14 121±22 −11.5±12.8 0.080

DiastolicBP(mmHg) 82±13 76.8±14.4 −5.33±7.09 0.125

Fastingglucose(mg/dl) 111±47 103±43 −8.17±5.46 0.139

Cholesterol(mg/dl) 214±53 197±37 −17.3±22.2 0.125

HDLcholesterol(mg/dl) 43.0±5.9 43.3±4.6 +0.33±4.59 0.866

TG(mg/dl) 224±87 188±101 −35.5±24.9^* 0.018

TC/HDL 4.99±1.15 4.58±0.98 −0.41±0.35^* 0.046

Adiponectin(␮g/ml) 1.32±0.49 1.71±0.73 +0.38±0.42 0.075

TNF-␣(pg/ml) 1.48±0.56 0.71±0.4 −0.76±0.31^* 0.028

IL-10^a(pg/ml) ND ND 0 1.000

IL-6^b(pg/ml) 1.73±0.9 2.00±0.83 +0.27±1.18 0.463

Variable Non-MetSyngroup(n=9)

Waistcircumference(cm) 103±7.32 98.2±6.75 −5.11±2.64^** <0.001

SystolicBP(mmHg) 121±10 124±14 +2.56±13.4 0.583

DiastolicBP(mmHg) 76.9±11.5 72.6±8.9 −4.34±8.4 0.160

Fastingglucose(mg/dl) 87.9±6.1 88±6.32 +0.11±5.11 0.950

Cholesterol(mg/dl) 197±25 185±26 −12.7±7.3^** 0.001

HDLcholesterol(mg/dl) 49.8±11.3 46.3±9.7 −3.44±2.92^** 0.008

TG(mg/dl) 120±67 134±55 +13.2±28.5 0.202

TC/HDL 4.17±1.13 4.20±1.28 +0.03±0.30 0.859

Adiponectin(␮g/ml) 1.23±0.56 1.15±0.65 −0.07±0.368 0.859

TNF-␣(pg/ml) 1.13±0.59 1.46±0.64 +0.33±0.41^* 0.049

IL-10^a(pg/ml) 1.79±1.66 1.89±1.21 +0.08±2.79 0.276

IL-6^b(pg/ml) 3.33±1.75 3.17±1.94 −0.16±1.62 0.515

Valuesaremeans±SD.StatisticalsigniﬁcancewasassayedbyWilcoxonsigned-ranktest.Themeasurementsweremadeatthe ﬁrstdaybeforeandaftertheweightcontrolprogram.Minussign(−)representsadecrease;plussign(+)representsanincrease.

ND,notdetectable.

aTheIL-10valueswereobtainedfrom5subjectsinthenon-MetSyngroup.

bTheIL-6valueswereobtainedfrom8subjectsinthenon-MetSyngroup.

*P<0.05.

**P<0.01.

a strong effect on the pathogenesis of metabolic syndrome in Asian people. Because insulin secre- tionandresistancewerenotincludedinmetabolic syndromefeatures[15],wedidnotmeasureserum insulin levels in the present study. Metabolic syndrome usually accompanies insulin resistance.

However, metabolic syndrome subjects may have thelowerinsulinsecretionratherthantheseverer insulin resistance.Insulinsecretionlevels in blood and insulin resistance such as homeostatic model assessment(HOMA)forinsulinresistance(IR)should be determinedin the futuretofurtherclarifythe relationship between insulin secretion and resis- tanceinmetabolicsyndromesubjects.Inaddition, moredatafromdifferentraceorregionsshouldbe accumulated to further unravel racialdifferences ofmetabolicsyndrome.

Overall, the weight, BMI, and WC in subjects withandroidobesitywereindeed improvedasthe weightcontrolprogramfinished.Theimprovement in net changes in TG, TC/HDL, and TNF-␣ in Met Synsubjectwassignificant.Thisstudysuggestthat weightlosscontrolmaybebeneficialtoreduceCVD viaimprovingserumIL-6andTNF-␣levels,aswell as MetSyn features. The knowledge about serum IL-6 and TNF-␣ levels in android obese subjects improvedbyshort-termweightlossisstillnew;the resultsconcerningMetSynimprovedbyweightloss areaconfirmationofpreviousreports.

Conﬂicts of interest statement

Theauthorsdeclarenoconﬂictofinterest.

(8)

Acknowledgements

ThisstudywassupportedbyresearchgrantsNSC95- 2313-B-005-059 fromthe National ScienceCouncil and IRBTCVGHNo:C05082/561fromtheTaichung Veterans General Hospital, Taichung, Taiwan, ROC.

References

[1] Tataranni PA, Ortega E. A burning question: does an adipokine-inducedactivationoftheimmunesystemmedi- atetheeffectofovernutritionontype2diabetes?Diabetes 2005;54:917—27.

[2]FantuzziG.Adiposetissue,adipokines,andinﬂammation.

JAllergyClinImmunol2005;115:911—9.

[3]Despres JP, Lemieux I.Abdominal obesity and metabolic syndrome.Nature2006;444:881—7.

[4]Kim SK, Park SW,Kim SH, ChaBS, Cho YW. Visceral fat amountisassociatedwithcarotidatherosclerosisevenin type2diabeticmenwithanormalwaistcircumference.Int JObes2009;33:131—5.

[5] Kumada M, KiharaS, Ouchi N, KobayashiH, Okamoto Y, OhashiK, etal. Adiponectinspeciﬁcally increasedtissue inhibitor of metalloproteinase-1 through interleukin- 10 expression in human macrophages. Circulation 2004;109:2046—9.

[6]DiezJJ,IglesiasP.Theroleofthenoveladiponectin-derived hormoneadiponectin inhumandisease.EurJEndocrinol 2003;148:293—300.

[7]WangJ,LiH,FrancoOH,YuZ,LiuY,LinX.Adiponectinand metabolic syndromein middle-agedand elderlyChinese.

Obesity2008;16:172—8.

[8]RenaldiO,PramonoB,SinoritaH,PurnomoLB,AsdieRH, AsdieAH.Hypoadiponectinemia:ariskformetabolicsyn- drome.ActaMedIndones2009;41:20—4.

[9]Shargorodsky M,BoazM, GoldbergY, MatasZ,GavishD, FuxA,etal.Adiponectinandvascularpropertiesinobese patients:isitanovelbiomarkerofearlyatherosclerosis?

IntJObes2009;33:553—8.

[10]Manigrasso MR,Ferroni P,Santilli F, Taraborelli T, Guag- nanoMT,MichettiN,etal.Associationbetweencirculating adiponectin and interleukin-10 levels in android obesity: effects of weight loss. J Clin Endocrinol Metab 2005;90:5876—9.

[11]EspositoK, PontilloA, GiuglianoF,Giugliano G,Marfella R, Nicoletti G, et al. Association of low interleukin-10

levelswiththemetabolicsyndromeinobesewomen.JClin EndocrinolMetab2003;88:1055—8.

[12]ChoiKM,RyuOH,LeeKW,KimHY,SeoJA,KimSG,etal.

Serumadiponectin,interleukin-10levelsandinﬂammatory markersinthemetabolicsyndrome.DiabetesResClinPract 2007;75:235—40.

[13]WHOExpertPanel.Appropriatebody-massindexforAsian populationsanditsimplicationsforpolicyandintervention strategies.Lancet2004;363:157—63.

[14] Expert Panelon Detection, Evaluation and Treatmentof HighBloodCholesterolinAdults.Executivesummaryofthe thirdreportoftheNationalCholesterolEducationProgram (NCEP)ExpertPanelonDetection,Evaluation,and Treat- mentofHighBloodCholesterolinAdults(AdultTreatment PanelIII).JAMA2001;285:2486—97.

[15]Lundgren JD, MalcolmB, Binks M, O’NeilPM.Remission of metabolic syndrome following a 15-week low-calorie lifestyle change program for weight loss. Int J Obes 2009;33:144—50.

[16]Culling KS, Neil HAW, Gilbert M, Frayn KN. Effects of short-termlow-andhigh-carbohydratedietsonpostpran- dialmetabolisminnon-diabeticanddiabeticsubjects.Nutr MetabCardiovascDis2009;19:345—51.

[17]TsaiAG,WaddenTA,SarwerDB,BerkowitzRI,WombleLG, HessonLA,et al.Metabolicsyndromeandhealth-related qualityoflifeinobeseindividualsseekingweightreduction.

Obesity2008;16:59—63.

[18] Nicklas BJ, You T, Pahor M. Behavioral treatments for chronic systemicinﬂammation:effects ofdietary weight lossandexercisetraining.CMAJ2005;172:1199—209.

[19]Vendrell J, Broch M, Vilarrasa N, Molina A, Gómez JM, GutiérrezC,etal. Resistin,adiponectin, ghrelin,leptin, and proinﬂammatory cytokines: relationships in obesity.

ObesRes2004;12:962—71.

[20]BruunsgaardH.Physicalactivityandmodulationofsystemic low-levelinﬂammation.JLeukocBiol2005;78:819—35.

[21]HotamisligilGS,ArnerP,CaroJF,AtkinsonRL,Spiegelman BM.Increasedadiposetissueexpressionoftumornecrosis factor-alphainhumanobesityandinsulinresistance.JClin Invest1995;95:2409—15.

[22]Saghizadeh M, Ong JM,Garvey WT, Henry RR, Kern PA.

The expression ofTNF␣ byhuman muscle.J ClinInvest 1996;97:1111—6.

[23]LangHF,ChouCY,SheuWHH,LinJY.Weightlossincreased serumadiponectinbutdecreasedlipidlevelsinobesesub- jectswhosebodymassindexwaslowerthan30kg/m².Nutr Res2011;31:378—86.

[24]NiswenderK.Diabetesandobesity:therapeutictargeting and risk reduction—–a complex interplay. Diabetes Obes Metab2010;12:267—87.

Availableonlineatwww.sciencedirect.com