Is adiponectin a marker of preclinicalatherosclerosis in kidney transplantation?

Cardiovascular disease is one of the leading causesof morbidity and mortality in patients undergoingkidney transplantation (KT) (1).

Abnormalities in glucose homeostasis (AGH)and diabetes mellitus (DM) are recognized asmajor complications following KT. These compli-cations contribute to a greater cardiovascular risk(2) and are associated with a decrease in patientand graft survival (3). Many authors currentlyrecommend performing an oral glucose tolerancetest (OGTT) to determine AGH (including pre-diabetes and DM status), following AmericanDiabetes Association (ADA) criteria, and to

establish the cardiovascular risk profile of thispatient population (4).Adiponectin (ADP) is an adipocytokine secreted

by adipose tissue. In normal conditions, it is foundin the blood at concentrations of 5–30 lg/mL,making up 0.01% of the total circulating plasmaproteins (5). Low adiponectin levels are associatedwith obesity, insulin resistance, and glucose intol-erance, as well as cardiovascular disease in bothtype 1 DM and type 2 DM (6). In previous studies,we demonstrated that low ADP levels prior totransplantation increase insulin resistance andcontribute to the risk of AGH (7). Other studies

Canas L, Bayes B, Granada ML, Ibernon M, Porrini E, Benıtez R,Dıaz JM, Lauzurica R, Moreso F, Torres A, Lampreabe I, Serra A,Romero R. Is adiponectin a marker of preclinical atherosclerosis in kidneytransplantation?.

Abstract: The aim of this study was to analyze the relationship betweenpre-transplant adiponectin (pre-ADP), abnormalities in glucose homeosta-sis (AGH) at three months post-transplantation, and preclinical athero-sclerosis in non-diabetic patients prior to kidney transplantation (KT).Methods: We carried out a multicenter study in 157 non-diabetic KTpatients (66.5% men; age: 50 ± 13 yr). Pre-ADP levels were analyzedusing radioimmunoassay. Carotid ultrasound was performed to determinecarotid intima-media thickness (c-IMT). Oral glucose tolerance test wascarried out to classify patients according ADA criteria.Results: Of the patients, 52.8% had AGH. Median pre-ADP was 19.5 (14–27) lg/mL. An inverse correlation was found between ADP and HOMAindex (r = )0.432; p < 0.001). Median c-IMT was 0.6 (0.48–0.71) mm.Significant inverse correlation existed between ADP and c-IMT on bothsides (p < 0.05). Patients with c-IMT >0.6 mm had more AGH(p = 0.012) and lower ADP levels (p = 0.02). We performed a logisticregression analysis using preclinical atherosclerosis (c-IMT ‡0.6 mm) asdependent variable and sex, age, BMI, ADP, AGH, and HOMA index asindependent variables of altered c-IMT. Age, pre-ADP, and AGH wereindependent risk factors for elevated c-IMT.Conclusions: Patients with AGH have a greater presence of preclinicalatherosclerosis. ADP has an inverse relationship with AGH and is anindependent marker of preclinical atherosclerosis.

Laura Canasa, Beatriz Bayesa,Maria L. Granadab, MeritxellIbernonc, Esteban Porrinid, RosaBenıteze, Juan M. Dıazf, RicardoLauzuricaa, Francesc Moresoc,Armando Torresg, IldefonsoLampreabee, Assumpta Serraa andRamon Romeroa

aNephrology Department, Hospital Germans

Trias i Pujol, UAB, Badalona, Barcelona,bBiochemistry Department, Hospital Germans

Trias i Pujol, Badalona, Barcelona, cNephrology

Department, Hospital de Bellvitge, Barcelona,dResearch Unit, Hospital Universitario de

Canarias, La Laguna, eNephrology Department,

Hospital de Cruces, Bilbao, fNephrology

Department, Fundacio Puigvert, Barcelona andgNephrology Department, Hospital Universitario

de Canarias, University of La Laguna, La Laguna,

Spain

Key words: adiponectin – altered glucose

homeostasis – carotid intima-media thickness –

carotid ultrasound – kidney transplantation

Corresponding author: Laura Canas Sole, Hos-

pital Germans Trias i Pujol, Ctra. De Canyet s/n,

08916 Badalona-Barcelona, Spain.

Tel.: 934978898; fax: 934978852;

e-mail: acanas.germanstrias@gencat.cat

Conflict of interest: There are no conflicts of

interest between the authors of this article.

Accepted for publication 7 April 2011

Clin Transplant 2012: 26: 259–266 DOI: 10.1111/j.1399-0012.2011.01490.xª 2011 John Wiley & Sons A/S.

259

have shown that adiponectin has antiatherogenicand anti-inflammatory properties and could nega-tively regulate the atherogenic process. It is knownthat ADP and TNF-a inhibit each other�s produc-tion in adipose tissue and that ADP can counteractthe proinflammatory effects of TNF-a in vascularcells (5).High-resolution B-mode carotid ultrasound aids

in diagnosing the presence and disease progressionof atherosclerosis (2) by measuring carotid intima-media thickness (c-IMT) and atheromatous plaque(8). The American Heart Association recommendsusing carotid ultrasound to assess cardiovascularrisk (8). c-IMT has also been associated withelevated insulin resistance (9). A recent study atour center demonstrated that kidney transplantpatients with a pathological OGTT have greaterc-IMT than patients with normal glycemia (NG)and, as a result, a higher incidence of preclinicalatherosclerosis (10). The relationship betweenc-IMT and adiponectin has also been studied inthe general population (11), demonstrating that lowlevels of adiponectin are a risk factor for preclinicalatherosclerosis, as measured by c-IMT (12).The aim of this study was to ascertain whether

ADP levels prior to transplantation are a markerfor AGH and preclinical atherosclerosis in non-diabetic patients undergoing KT.

Material and methods

Patients

These analyses were based on a population of 157patients with chronic kidney failure selected fromfive different regions in Spain (66.5% men and33.5% women; mean age 50 ± 13 yr). Prior totransplantation, 15.4% of the patients were smok-ers. The prevalence of hyperlipidemia was 36.6%,hypertension 76%, and left ventricular hyper-trophy 33% (diagnosed using echocardiographycriteria). Before transplantation, 8.8% of thepatients experienced a cardiovascular event.Initial kidney disease was classified in the

patients as follows: 29% chronic glomerulonephri-tis, 11.6% renovascular disease, 9% polycystickidney disease, 10.3% interstitial nephritis, and40% other causes (Alport Syndrome, lupus eryth-ematosus systemic, unknown cause). All patientsreceived renal substitution therapy before trans-plantation: 79.2% with hemodialysis and 20.8%with peritoneal dialysis.Mean body mass index (BMI) three months

after transplantation was 24.87 ± 4.45 kg/m2, andmean waist-to-hip ratio (WHR) at three monthspost-transplantation was 0.93 ± 0.10.

Prior to transplantation, all patients had FPGconcentrations lower than 126 mg/dL on at leasttwo different occasions. None were being treatedwith oral antidiabetic agents or insulin. Patientswith a previous diagnosis of diabetes were excludedfrom the study.

Immunosuppression

Patients received immunosuppression following theclinical practice guidelines of each center; the mostwidely used immunosuppressive regimen was pred-nisone, tacrolimus, and mycophenolate mofetil. Interms of immunosuppressant drugs that favoraltered glucose homeostasis, corticosteroids andtacrolimus were used. All patients were treatedwith corticosteroids (oral prednisone), with a meandose of 10 ± 4 mg/d three months after KT, and60.8% received treatment with tacrolimus. Thetacrolimus dose was adjusted to achieve plasmathrough levels of 8–15 ng/mL. Of the patients,16.1% suffered acute rejection, which was treatedwith a 250–500 mg bolus of intravenous methyl-prednisolone daily, for three d.

This study was approved by the Ethics Com-mittee of Hospital Germans Trias i Pujol, and allpatients gave their written informed consent toparticipate in the study.

Diagnosis of abnormal glucose levels

Diagnosis of abnormalities in glucose prior to KT. Allof the patients included in the study presented withblood glucose levels under 126 mg/dL; however, tocarry out the statistical analysis, patients weresubclassified into two groups: NG (baseline glyce-mia <100 mg/dL) and impaired fasting glucose(IFG) (glycemia ‡100 and <126 mg/dL).

OGTT after transplantation. An OGTT was carriedout in all patients three months after transplanta-tion (75 g glucose). Glucose concentrations weremeasured at baseline (0 min) and at 120 min. Thetest was always carried out in the morning whenpatients had fasted for a minimum of 10 h.

Using the 2006 ADA guidelines (13), patientswere classified as having:

• NG if baseline glycemia was <100 mg/dL andglycemia at 120 min was <140 mg/dL.

• AGH, which included:

1. Prediabetes status:

(i) IFG if baseline glycemia was ‡100 and<126 mg/dL and glycemia at 120 min was<140 mg/dL.

Canas et al.

260

(ii) Impaired glucose tolerance (IGT) if baselineglycemia was <126 mg/dL, and glycemia at120 min was ‡140 and <200 mg/dL.

2. New-onset DM if baseline glycemia was>126 mg/dL and glycemia at 120 min was‡200 mg/dL.

Carotid intima-media thickness

A Duplex–Doppler color carotid ultrasound wasperformed shortly after transplantation (withinthree months) in all patients using an AcusonSequoia 512 with a 7.5–14 MHz high-resolutionlinear transducer for the image in real time and3.5 MHz for the Doppler study. The carotid zonewas screened with the patient in the supine positionwith the neck extended; vessels were examined usinganterior oblique, lateral, posterior oblique, andtransverse views. Four predetermined artery seg-mentswere examined on both sides: internal carotid,external carotid, and carotid bifurcation on bothsides. The results of the c-IMTwere expressed as themean of eight determinations, four on each side.

Junyent et al. (8) described the measurement of c-IMT and determination of the presence of plaque asuseful in assessing preclinical atherosclerosis anddescribed the normal c-IMT values in a cohort ofSpanish subjects. This group described a positivecorrelationbetween age, sex, c-IMT, andpresence ofplaque. In the subjects, ages 46–65 yr, mean c-IMTwas 0.6 mm in men and women. Following thisobservations, in our study, patients were consideredto have preclinical atherosclerosis if c-IMT was>0.6 mm.Furthermore, as shown in the results, thisvalue coincides with the median of our population.

Although all carotid ultrasound studies use thesame general principles to measure c-IMT, themethod shows notable variations in terms ofthe segments examined and the reading procedure.An evaluation of the different centers was made byan external researcher to homogenize the results.

Laboratory measurements

To determine adiponectin levels prior to KT, bloodsamples were obtained from all patients at the lastclinical and analytic control before transplantation(1–3 months pre-transplantation) after at leasteight h of fasting.

Standard blood cell count, plasma glucose, totaland HDL cholesterol, urea, creatinine, and urinarycreatinine were measured with a routine clinicalchemistry laboratory analyser. LDL cholesterolconcentrationswere calculated using theFriedewaldformula. Glycosylated hemoglobin (HbA1c) was

measured in blood samples with EDTA by high-pressure liquid chromatography using a fully auto-mated Menarini ADAMS A1c HA-8160 analyzermanufactured by Arkray (Kyoto, Japan) with aninter-assay coefficient of variation (CV) of 1.9%and1.2% at HbA1c levels of 4.8% and 9%, respectively(reference range: 4.0–5.8%). Serum adiponectinconcentrations were measured by a commercialradioimmunoassay (LincoResearch, Inc., St. Louis,MO,USA). Intra-assayCVwasbelow6.2%and inter-assay CV below 9.2%. Assay sensitivity was 1 lg/L.Serum insulin was measured by an electroche-

miluminescent immunoassay (Modular AnalyticsE-170; Roche Diagnostics, GMBH, Mannheim,Germany). The inter-assay CV was 3.7 and 2.6%at 14.5 and 78.3 lUI/mL, respectively (referencerange: 2.6–24.9 lIU/mL). Insulin resistance wasassessed by the Homeostasis Model Assessment ofInsulin Resistance (HOMA-IR) and calculatedfrom fasting plasma insulin (FI) and FPG concen-trations as follows: HOMA-IR-score = FI (mIU/L) · FPG (mM)/22.5.

Statistical analysis

Results of variables with normal distribution areexpressed as mean ± SD and those of variableswith a non-normal distribution as median andinterquartile range.The chi-square test was used for analysis of the

percentage of patients with AGH and preclinicalatherosclerosis.The non-parametric Mann–Whitney U test was

used to compare adiponectin between patientswith c-IMT <0.6 mm and c-IMT ‡0.6 mm. Whenvariables demonstrated a normal distribution, theparametric Student�s t-test was used. For correla-tion studies, the Spearman correlation coefficientwas used.The Kruskal–Wallis test was applied to compare

differences between groups (ADP quartiles andc-IMT quartiles).For the univariate and multivariate analyses, a

logistic regression analysis was carried out usingc-IMT (‡0.6 mm) as the dependent variable, andsex, age, body mass index, adiponectin, AGH andHOMA index as independent variables.Differences were considered statistically signifi-

cant when p < 0.05.Data were analyzed using the statistics program

SPSS 12.0 (SPSS Inc, Chicago, IL, USA).

Results

The OGTT at three months post-transplantationshowed that 52.8% of patients had AGH (9%

Preclinical atherosclerosis in kidney transplantation

261

IFG, 15.6% IGT, 12.2% IFG + IGT, 16% DM),while 47.2% maintained NG (Fig. 1).Median ADP of the 157 non-diabetic patients

prior to KT was 19.5 (14–27) lg/mL, with higherlevels found in female patients (ADP: 24.41 ±11.22 lg/mL) compared with males (ADP:19.23 ± 9.19 lg/mL; p = 0.005). No significantcorrelation was found between pre-transplantationADP values and age (p = 0.37).With regards to glucose metabolism and the type

of immunosuppressive treatment used, the predni-sone dose used did not allow us to establishdifferences between patients with NG and thosewith abnormalities in glucose. On the other hand,patients with abnormalities in glucose had higherlevels of tacrolimus at three months post-transplantcompared with patients with NG. This differencewas statistically significant (AGH-levels TAC 3 m:9.6 ng/mL; NG-levels TAC 3 m: 9 ng/mL;p = 0.028). There were no differences betweenpatients with NG and those with glucose abnormal-ities in terms of acute rejection episodes (p = NS).

Pre-transplantation abnormalities in glucose,adiponectin and c-IMT

In reference to baseline glycemia in the studypopulation, Table 1 highlights the differencesbetween patients with NG and those with abnor-malities in baseline blood glucose prior to trans-plantation. As shown in the Table, patients withAGH are older, have greater preclinical athero-sclerosis, and have statistically significant lowerlevels of ADP (p < 0.05).

Adiponectin, abnormal glucose homeostasis, andanthropometric parameters at three months aftertransplantation

The relationship between pre-transplantation ADPvalues and variables such as anthropometric

measurements and altered glucose homeostasis-related parameters were analyzed. The analysisshowed an inverse correlation between pre-trans-plant ADP and BMI (r = )0.448; p < 0.001),WHR (r = )0.501; p < 0.001), baseline glucose(r = )0.237; p = 0.002), insulin (r = )0.410;p < 0.001), and HOMA index (r = )0.432;p < 0.001). ADP showed a positive correlationwith HDL cholesterol (r = 0.230; p = 0.007).There was no correlation observed between AGHwith BMI and renal function at three months aftertransplantation.

c-IMT and differences according to age and sex

IMT of the left and right internal carotid arterieswas measured in all patients using carotid ultra-sound, revealing a median c-IMT in our popula-tion of 0.6 mm (0.48–0.71 mm) and a mean c-IMT0.65 ± 0.25 mm. These values coincided withnormal c-IMT values in a Spanish populationreported by Junyent et al. (8), and, consideringthat c-IMT was measured immediately after trans-plantation (<3 months), these results reflect theatherosclerosis that existed before the transplanta-tion. In all cases, the ultrasound study was carriedout three months post-transplantation. c-IMTshowed a positive correlation with age: right-c-IMT–age: r = 0.429; p < 0.001, left c-IMT-age:r = 0.430; p < 0.001. No statistically significantrelationship was observed between c-IMT and sex(right-c-IMT–women: 0.65 ± 0.28 mm, right-c-IMT–men: 0.62 ± 0.22 mm, left c-IMT–women:0.67 ± 0.28 mm, left c-IMT–men: 0.63 ± 0.25 mm).

Abnormal glucose homeostasis three months aftertransplantation and c-IMT

In analyzing the relationship between AGHthree months after transplantation and c-IMT inour population, we observed that patients withc-IMT greater than the median of the population

0,05,0

10,015,020,025,030,035,040,045,050,0

NG IFG IGT IFG+IGT DM

Perc

enta

ge

Fig. 1. Abnormalities in glucose homeostasis at three monthspost-transplant according to OGTT result. OGTT, oral glucosetolerance test; NG, normal glycemia; IFG, impaired fastingglucose; IGT, impaired glucose tolerance; DM, diabetes mellitus.

Table 1. Relationship between pre-TR glycemia, ADP, and c-IMT

Pre-TR glycemia<100 mg/dL(69.6%)

Pre-TR glycemia>100 to <126mg/dL (30.4%) p

Pre-TR ADP(lg/mL)

21.5 ± 10.19 16.3 ± 8.8 0.046

Bilateralc-IMT (mm)

0.62 ± 0.21 0.78 ± 0.28 0.003

Age (yr) 48.14 ± 12.9 56.32 ± 11.7 0.001

Values are expressed as mean ± standard or median deviation.Pre-TR ADP, pre-transplant adiponectin; c-IMT, carotid intima-media thickness;pre-TR glycemia, pre-transplant glycemia.

Canas et al.

262

(c-IMT ‡0.6 mm) had a higher percentage ofabnormal glucose homeostasis (NG 36.3% vs.AHG 63.7%; p = 0.012).

Adiponectin and c-IMT

The correlation between ADP levels before trans-plantation and c-IMT was analyzed, showing aninverse correlation between pre-transplant ADPconcentrations and c-IMT on both sides(r = )0.21; p = 0.01).

Patients with a c-IMT greater than that ofthe mean of the population (c-IMT ‡0.6 mm)had lower pre-transplant ADP concentrations(18.50 ± 8.85 lg/mL) compared with those pa-tients with a lower c-IMT (22.94 ± 10.66 lg/mL)(p = 0.02). After analyzing c-IMT in quartiles, weobserved that patients with a greater c-IMT hadstatistically significant lower levels of pre-trans-plant ADP (Fig. 2).

Characteristics of the study population according toadiponectin quartiles

Characteristics of the study population accordingto ADP quartiles are shown in Table 2. Patientsin Q4 for ADP (the highest ADP levels:36.53 ± 7.57 lg/mL) had a statistically significantlower BMI and higher levels of HDL cholesterolcompared with the remaining patients (p < 0.001and 0.017, respectively).

At three months post-transplantation, patientsin Q4 had lower baseline glycemia (p = 0.025) andlower insulin resistance (lower HOMA index)compared with the rest of the patients (p < 0.001).

In our study, no significant correlation wasfound between pre-transplant ADP and kidneyfunction at three months post-transplantation, asmeasured by creatinine and creatinine clearance.

Logistic regression analysis

Logistic regression analysis, where the dependentvariable was median c-IMT (0.6 mm) and theindependent variables were sex, age, BMI, ADP,AGH, and HOMA index, was used to analyzethe effect of different parameters on preclinicalatherosclerosis. This analysis showed age, ADPpre-transplant and AGH at three months aftertransplantation to be independent markers for anincreased c-IMT, as shown in Table 3. In addition,the analysis showed that for every decreased unit ofpre-transplant ADP, there was a 5% increase in therisk of an elevated c-IMT, regardless of age andAGH.

Discussion

The main findings of this study were the following:first, a high incidence of AGH and DM in kidneytransplant patients, and an inverse relationshipbetween adiponectin and AGH were confirmed;second, patients with AGH had greater preclinicalatherosclerosis, i.e., greater c-IMT; and third, aninverse correlation was found between ADP levelsand c-IMT in the kidney transplant population.Among the classical risk factors described for

cardiovascular disease, obesity is of special interestfor kidney transplant patients. Many studies havereported weight gain in patients following trans-plantation, which is known to play a role inthe development of metabolic disorders such asdyslipidemia, insulin resistance, DM, and themetabolic syndrome.It has recently been found that patients with

obesity present altered secretion of adipocytokines(including adiponectin) by adipocytes, whichfavors insulin resistance, with ADP acting as alink between these two diseases (14). In our study,we demonstrated an inverse correlation betweenobesity-related anthropometric parameters (BMIand WHR) and ADP. ADP can reduce cardiovas-cular disease, by increasing insulin sensitivity andlowering lipid levels in blood (15). Various studieshave shown a close relationship between ADPand glucose metabolism (11, 16). In this study, weconfirm the indirect correlation between ADP andthe HOMA index, in which patients with low ADPlevels have a higher percentage of AGH. In ourstudy, as in that by Tschritter et al. (17), a positivecorrelation was found between plasma ADP levelsand HDL cholesterol.In the general population, the important role of

ADP as a protector against cardiovascular disease(10, 15) and as a predictor of cardiovascular events,

0

5

10

15

20

25

Q1 Q2 Q3 Q4

Quartiles of c-IMT (mm)

Pret

rans

plan

t AD

P (u

g/m

L)

ADP

Fig. 2. Relationship between carotid intima-media thickness(c-IMT) expressed as quartiles and pre-transplant ADP levels.Mann–Whitney U test. p = 0.036. Q1, first quartile(0.4 ± 0.7 mm); Q2, second quartile (0.54 ± 0.35 mm); Q3,third quartile (0.68 ± 0.42 mm); Q4, fourth quartile(0.98 ± 0.24 mm). All were related to c-IMT.

Preclinical atherosclerosis in kidney transplantation

263

owing to its anti-inflammatory and antiatherogenicproperties, has been described (15).The relationship between ADP and cardiovascu-

lar disease has not only been described in kidneytransplant patients, but also in pre-dialysis anddialysis patients. Studies in patients undergoingperitoneal dialysis and pre-dialysis (18) demon-strate a negative correlation between ADP andinflammatory markers (CRP), suggesting that lowplasma ADP levels could be considered a marker ofinflammation in uremic patients. Takemoto et al.(19) showed ADP concentrations to be an indepen-dent risk factor for cardiovascular disease in apopulation of patients undergoing hemodialysis.c-IMT is a measurement of subclinical athero-

sclerosis that has been correlated with the incidenceof coronary heart disease and myocardial infarc-tion in the general population (20) and is consid-ered to be an independent risk factor forcardiovascular mortality in hemodialysis patients(21). In our study, a carotid ultrasound wasperformed shortly after transplantation (threemonths after transplantation), and the differencein time between the analysis of adiponectin and the

carotid ultrasound study was also <3 months. Wethink that the carotid ultrasound three monthsafter transplantation is a reflection of the situationthat exists in the pre-transplantation phase,because it is unlikely that there are major modifi-cations in the vascular system in such a shortperiod of time. This suggests that adiponectin priorto transplantation could be a marker of carotidatherosclerosis in this population.

While discrepancies have been found in thegeneral population with regard to the correlationbetween c-IMT and cardiovascular risk factors suchas diabetes (20), a clear correlation was foundbetween c-IMT and AGH in our population oftransplant patients. In the pre-transplant phase, wecould already observe that patients in our studypopulation with abnormalities in baseline glycemiahad a greater c-IMT compared with patients withNG. This greater preclinical atherosclerosis inpatients with AGH was confirmed after KT, whenwe observed that patients with a c-IMT greater thanthemean of the population (c-IMT ‡0.6 mm) have ahigher percentage of AGH (63.7%) compared withpatientswith a c-IMT less than thepopulationmean.

In a previous study in transplant patients, wedemonstrated the utility of c-IMT as a marker ofpreclinical atherosclerosis and its relationship withnon-classical cardiovascular risk factors such asghrelin (10). Although a study by Nilsson et al.(11) showed an inverse relationship between ADPand c-IMT in a general male population, thepresent study showed that women had higher ADPconcentrations compared with men; however, wedid not observe any statistically significant differ-ences between c-IMT according to sex.

No studies have been published showing the roleof ADP as a marker of atherosclerosis in KT. In

Table 2. Characteristics of kidney transplant patients by quartiles of adiponectin

N

Q1 Q2 Q3 Q4

p-value44 41 40 32

Pre-transplant ADP(lg/mL)

10.29 ± 2.72 17.12 ± 1.79 24.25 ± 2.14 36.53 ± 7.57

Age (yr) 54.16 ± 11.96 49.17 ± 14.05 49.12 ± 14.79 50.87 ± 14.15 NSBMI 3 m (kg/m2) 26.70 ± 3.64 25.47 ± 3.56 24.15 ± 3.51 22.94 ± 3.18 <0.001WHR 3 m (cm) 0.97 ± 0.6 0.94 ± 0.10 0.86 ± 0.12 0.81 ± 0.05 0.002HDL Chol 3 m (mg/dL) 54.05 ± 14.57 60.94 ± 17.61 56.55 ± 16.74 71.00 ± 19.85 0.017Glycemia 3 m (mg/dL) 96.75 ± 13.20 93.59 ± 14.69 94.43 ± 14.90 88.00 ± 11.73 0.025HbA1C 3 m (%) 5.08 ± 0.60 5.05 ± 0.82 5.04 ± 0.70 5.22 ± 0.77 NSHOMA Index 3 m (*) 4.93 (2.33–1.54) 2.74 (1.88–7.89) 2.16 (1.62–2.59) 1.60 (1.06–2.71) <0.001Creatinine 3 m (mg/dL) 1.60 ± 0.49 1.54 ± 0.60 1.45 ± 0.39 1.40 ± 0.39 NSCcr 3 m (mL/min) 59.86 ± 22.70 66.90 ± 32.27 63.25 ± 23.15 59.54 ± 19.42 NS

Values are expressed as mean ± standard or median deviation (P25–P75) (*). The differences between groups were established using the Kruskal–Wallis test.N, total patients; ADP, adiponectin; HDL chol, HDL cholesterol; BMI, body mass index; WHR, waist–hip ratio; HbA1C, glycosylated hemoglobin; Ccr, creatinine clearance.All terms are determined at three months after transplantation (3 m).

Table 3. Risk factors for preclinical atherosclerosis in kidney transplanta-tion

Independentvariable b p-value

Exp(b) CI 95%

Age 0.091 <0.01 1.096 1.050–1.144ADP pre-TR )0.050 0.030 0.951 0.909–0.995AGH-3m )1.214 0.025 0.297 0.103–0.859

Dependent variable: carotid IMT (‡0.6 mm); independent variables: sex, age,body mass index, adiponectin, altered glucose homeostasis, HOMA index.ADP pre-TR, adiponectin pre-transplantation; AGH-3m, abnormalities in glucosehomeostasis at three months post-transplant.

Canas et al.

264

this study, we show that patients in the first quartile(Q1) for ADP (the lowest ADP levels) have agreater c-IMT (0.71 ± 0.25 mm) compared withpatients in the fourth quartile (Q4) for ADP (thehighest ADP levels) (c-IMT 0.58 ± 0.15 mm).Furthermore, c-IMT in the first quartile was greaterand clearly pathological compared with the mean c-IMT in the Catalan population when adjusted forage and sex, as described by Junyent et al. (men0.62 ± 0.12 mm, women 0.59 ± 0.15 mm) (8).

A recent study in a type 1 diabetes populationshowed a clear relationship between ADP andrenal insufficiency progression (22); the concentra-tion of adiponectin is greater when kidney failure ismore severe, because of a decrease in renalclearance. In addition, it has been shown thatduring the first year after transplantation, patientsshow a degree of hyperinsulinemia as a result of anincrease in glomerular filtration rate (23). In ourstudy, we found no relationship between pre-transplant ADP and kidney function at threemonths post-transplantation. We did not havedata for ADP concentrations three months aftertransplantation, and thus cannot demonstrate acorrelation between ADP and renal function.

In summary, our study suggests that ADP is amarker of AGH and demonstrates an inverseassociation between ADP and preclinical athero-sclerosis measured using c-IMT.

Limitations of the study

There are some limitations of the study that shouldbe pointed out.

First of all, an OGTT was not carried out beforetransplantation, and AGH were studied threemonths after transplantation. However, a multi-center study carried out in Catalonia, Spain,showed that AGH found when performing anOGTT in patients on hemodialysis and on atransplant waiting list (31.3%) are very similar toAGH found three months after transplantation(34%), suggesting that alterations found aftertransplantation could reflect the clinical situationprior to transplantation (data pending publication).

In addition, recent published studies demon-strate that the glucose metabolism three monthsafter transplantation reflects the situation thatexists before transplantation (24, 25).

Acknowledgements

This work was supported by grants FIS 04/0988, the RedTematica de Investigacion C 03/03 (Instituto de SaludCarlos III, Spanish Ministry of Health), and the Red deInvestigacion RedInRen.

References

1. Kasiske BL, Guijarro C, Massy ZA, Wiederkehr MR,Ma JZ. Cardiovascular disease after renal transplantation.J Am Soc Nephrol 1996: 7: 158.

2. Alvarez A, Fernandez J, Porrini E et al. Carotid ath-eromatosis in nondiabetic renal transplant recipients: therole of prediabetic glucose homeostasis alterations.Transplantation 2007: 84: 870.

3. Ducloux D, Kazory A, Chalopin JM. Posttransplantdiabetes mellitus and atherosclerotic events in renaltransplant recipients: a prospective study. Transplantation2005: 79: 438.

4. Armstrong KA, Prins JB, Beller EM et al. Should anoral glucose tolerance test be performed routinely in allrenal transplant recipients? Clin J Am Soc Nephrol 2006:1: 100.

5. Palomer X, Perez A, Blanco-Vaca F. Adiponectina: unnuevo nexo entre obesidad, resistencia a la insulina yenfermedad cardiovascular. Med Clin (Barc) 2005: 124:388.

6. Costacou T, Orchard T. Adiponectin: good, bad, or justplain ugly? Kidney Int 2008: 74: 549.

7. Bayes B, Lauzurica R, Granada ML et al. Adiponectinand risk of new-onset diabetes mellitus after kidneytransplantation. Transplantation 2004: 78: 26.

8. Junyent M, Gilabert R, Nunez I et al. Ecografıacarotıdea en la evaluacion de aterosclerosis preclınica.Distribucion de valores del grosor ıntima-media y frecu-encia de placas de ateroma en una cohorte comunitariaespanola. Med Clin (Barc) 2005: 125: 770.

9. Hedblad B, Nilsson P, Janzon L, Berglund G. Rela-tion between insulin resistance and carotid intima-mediathickness and stenosis in non-diabetic subjects. Resultsfrom a cross-sectional study in Malmo, Sweden. DiabetMed 2000: 17: 299.

10. Bayes B, Granada ML, Alonso N et al. Ghrelin, glucosehomeostasis, and carotid intima media thickness in kidneytransplantation. Transplantation 2007: 84: 1248.

11. Nilsson PM, Engstrom G, Hedblad Bo et al. Plasmaadiponectin levels in relation to carotid intima mediathickness and markers of insulin resistance. ArteriosclerThromb Vasc Biol 2006: 26: 2758.

12. Iglseder B, Mackevics V, Stadlmayer A, Tasch G,Ladurner G, Paulweber B. Plasma adiponectin levelsand sonographic phenotypes of subclinical carotid arteryatherosclerosis. Data from SAPHIR study. Stroke 2005:36: 2577.

13. ACE/ADA Task Force on Inpatient Diabetes. Ameri-can College of Endocrinology and American DiabetesAssociation Consensus statement on inpatient diabetesand glycemic control. Diabetes Care 2006: 29: 1955.

14. Fasshauer M, Paschke R. Regulation of adipocytokinesand insulin resistance. Diabetologıa 2003: 46: 1594.

15. Pischon T, Girman CJ, Hotamisligil GS et al. Plasmaadiponectin levels and risk of myocardial infarction inmen. JAMA 2004: 291: 1730.

16. Hojlund K, Frystyk J, Levin K, Flyvbjerg A et al.Reduced plasma adiponectin concentrations may contrib-ute to impaired insulin activation of glycogen synthase inskeletal muscle of patients with type 2 diabetes. Diabet-ologıa 2006: 49: 1283.

17. Tschritter O, Fritsche A, Thamer C, Haap M et al.Plasma adiponectin concentrations predict insulin sensi-tivity of both glucose and lipid metabolism. Diabetes 2003:52: 239.

Preclinical atherosclerosis in kidney transplantation

265

18. Tsao Y, Hsu Y, Chu N et al. Association of plasmaadiponectin and cardiovascular risk profiles in nondiabeticuremic patients on peritoneal dialysis. J Nephrol 2008: 21:744.

19. Takemoto F, Katori H, Sawa N et al. Plasma adipo-nectin: a predictor of coronary heart disease in hemodial-ysis patients – a Japanese prospective eight-year study.Nephron Clin Pract 2008: 111: c12.

20. Preston E, Ellis M, Kulinskaya E et al. Associationbetween carotid artery intima-media thickness and car-diovascular risk factors in CKD. Am J Kidney Dis 2005:46: 856.

21. Nishizawa Y, Shoji T, Maekawa K et al. Intima-mediathickness of carotid artery predicts cardiovascular mortal-ity in hemodialysis patients. Am JKidneyDis 2003: 41: S76.

22. Jorsal A, Tarnow L, Frystyk J et al. Serum adiponectinpredicts all-cause mortality and end stage renal disease inpatients with type 1 and diabetic nephropathy. Kidney Int2008: 74: 649.

23. Porrini E et al. Metabolic syndrome, insulin resistance,and chronic allograft disfunction. Kidney 2010: 78:S42.

24. Sharif A, Ravindran VK, Dunseath G et al. Impendinghyperglycemia in normoglycemic renal transplant recipi-ents—an experimental predictive surrogate. Transplanta-tion 2010: 89: 1341.

25. Iida S, Ishida H, Tokumoto T et al. New-onset diabetesafter transplantation in tacrolimus-treated, living kidneytransplantation: long-term impact and utility of the pre-transplant OGTT. Int Urol Nephrol 2010: 42: 935.

Canas et al.

266