A

iwa

tMoc

e©

K

1

cltcm

0d

Atherosclerosis 196 (2008) 262–269

Low plasma adiponectin exacerbates the risk of prematurecoronary artery disease in familial hypercholesterolemia

Tarek Bouhali a, Diane Brisson a, Julie St-Pierre a,c, Gerald Tremblay a, Patrice Perron a,Catherine Laprise a,b, Marie-Claude Vohl d, Maud N. Vissers e, Barbara A. Hutten f,

Jean-Pierre Despres c, John J.P. Kastelein e, Daniel Gaudet a,∗a Lipid Research Group and University of Montreal Community Genomic Medicine Center, Universite de Montreal,

Chicoutimi Hospital, 305 St. Vallier, Chicoutimi, Quebec G7H 5H6, Canadab Universite du Quebec a Chicoutimi, Canada

c Quebec Heart Institute, Laval Hospital, Sainte-Foy, Canadad Department of Food Science and Nutrition, Laval University, Quebec, Canadae Department of Vascular Medicine, University of Amsterdam, The Netherlands

f Department of Clinical Epidemiology and Biostatistics, University of Amsterdam, The Netherlands

Received 29 March 2006; received in revised form 29 September 2006; accepted 25 October 2006Available online 22 November 2006

bstract

Familial hypercholesterolemia (FH) is characterized by increased risk for premature coronary artery disease (CAD). This risk is exacerbatedn the presence of abdominal obesity and insulin resistance. Low adiponectin is part of the clustering of metabolic abnormalities associatedith abdominal obesity and insulin resistance. The present study, therefore, aims to examine the relationship between plasma adiponectin and

ge at CAD diagnosis in FH patients.Plasma adiponectin was measured by ELISA in 568 non-diabetic FH individuals of French-Canadian origin. CAD was defined according

o strict clinical criteria. Prior to analyses, patients were grouped according to age and gender-specific tertiles of plasma adiponectin levels.ultivariate Cox proportional hazards regression was used to estimate the association between plasma adiponectin levels and age at diagnosis

f CAD. Overall, FH patients in the lowest tertile of plasma adiponectin exhibited CAD at a significantly younger age (hazard ratio = 1.73,

onfidence interval 95%: [1.19–2.53]; p = 0.004).

These results suggest that low plasma adiponectin is associated with an increased risk of premature CAD over and above the alreadyxaggerated risk seen in FH patients.

2006 Elsevier Ireland Ltd. All rights reserved.

olester

silca

eywords: Adiponectin; Adipocytokines; Coronary artery disease; Hyperch

. Introduction

Familial hypercholesterolemia (FH) is a clinical syndromeharacterized by high plasma concentrations of low-densityipoprotein (LDL)-cholesterol and apolipoprotein (apo) B,

endinous xanthomas and an increased risk of prematureoronary artery disease (CAD). Most often, FH is caused byutations in the LDL receptor (LDLR) gene, although this

∗ Corresponding author. Tel.: +1 418 541 1077; fax: +1 418 541 1116.E-mail address: daniel.gaudet@umontreal.ca (D. Gaudet).

gt(ssT

021-9150/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.atherosclerosis.2006.10.035

olemia; Risk factors

yndrome can also be the consequence of genetic variationsn the ARH and NARC-1 genes [1–3]. FH is the most preva-ent monogenic disorder worldwide and an important geneticause of myocardial infarction [3]. In absence of treatment,lmost 50% of men and more than 20% of women, heterozy-ous for LDLR gene mutations, will develop CAD beforehe age of fifty [3,4]. FH affects approximately 1 per 350

0.3%) individuals among Caucasians, but the prevalence isignificantly higher in some population clusters worldwide,uch as in Lebanon, North Africa, South Africa, Finland,he Netherlands and among French-Canadians [5,6]. In the

scleros

S1

saemnC

lwAtTa3laaiprocc

bn

2

2

C

coffjLp(anaiia1iataagtC11atTueut5

TB

WSCHLTA

B

A

M

T. Bouhali et al. / Athero

aguenay region (Quebec, Canada), the prevalence reachesper 80 (1.1%) [7].Hyperapobetalipoproteinemia (hyperapoB) is not exclu-

ively observed in FH, but it is also associated with obesitynd insulin resistance-related phenotypes [8]. Whereas sev-ral complex metabolic pathways are involved in apoBetabolism and modulate CAD risk, numerous genetic and

on-genetic factors are responsible for the heterogeneity ofAD risk in individuals with hyperapoB.

Non-FH subjects with abdominal obesity and hyperinsu-inemia exhibit a CAD risk equivalent to that of FH subjectsith neither abdominal obesity nor hyperinsulinemia [9,10].diponectin has been studied in relation with these fac-

ors and other features of the metabolic syndrome [11].his molecule is an adipocyte-derived plasma protein, highlybundant in plasma, with concentrations ranging from 5 to0 �g/ml (0.01% of total plasma) [12,13]. Its levels corre-ate positively with HDL-cholesterol, independently fromll known correlates of HDL-cholesterol [14], but it is neg-tively associated with the metabolic syndrome [15]. Fornstance, although it is well known that excess visceral adi-osity is associated with reduced HDL-cholesterol levels, aecent study indicated that for any given level of visceralbesity, patients with low plasma adiponectin levels wereharacterized by even more markedly reduced plasma HDL-holesterol [16].

The aim of the present study was to assess the associationetween plasma adiponectin and age at first CAD event inon-diabetic FH patients.

. Methods

.1. Subjects and clinical data

The Saguenay-Lac-Saint-Jean (SLSJ) region (Quebec,anada) has a population of 270,000 inhabitants, which is

wALo

able 1aseline characteristics of the studied FH sample

Without CAD (N = 391)

aist circumference (cm) 84.2 ± 12.3moking habit (%)a 25.3holesterol (mmol/l) 9.06 ± 1.56DL-cholesterol (mmol/l) 1.17 ± 0.33DL-cholesterol (mmol/l) 7.06 ± 1.53riglycerides (mmol/l)b 1.60 ± 1.00polipoprotein B (g/l)c 1.50 ± 0.36

lood pressure (mmHg)Systolic 122.6 ± 17.8Diastolic 78.0 ± 11.5

diponectin (�g/ml)b 7.40 ± 4.11

ean ± S.D. Lipid and adiponectin levels correspond to values without lipid-lowera Proportion of current smoker subjects.b Geometric mean and p-value after log10-transformation were used for triglyceric N = 452.

is 196 (2008) 262–269 263

omparable to the population of Iceland [17]. The prevalencef FH in the SLSJ population is 1.1% and has been estimatedrom three complementary sources: [1] the Hardy Weinbergormula (based on the repertory of known homozygous sub-ects in this region) [7,2] the proportion of individuals withDL-C concentrations above the 99th percentile in the sameopulation as assessed by the Quebec Heart Health Surveya randomly selected public health survey conducted in 1991s part of a pan-Canadian health study) and finally [3] theumber of FH individuals and families having been evalu-ted as part of a health prevention and promotion programn the mid-1990s [18]. All three approaches were concordantn estimating the total FH population in the SLSJ region atpproximately 3000, of which 80% are adults aged between8 and 65 years. Following a systematic public health screen-ng protocol implemented in 1993 [19], more than 60% of FHdults in this age group have since been met and evaluated athe Chicoutimi Hospital Lipid Clinic. Due to founder effect,lmost 70% of all known FH cases in the SLSJ populationre the consequence of two specific variations in the LDLRene: a point (W66G) missense mutation (also known ashe French-Canadian-2 mutation), which originates from theharlevoix region and explains 50% of cases, and a non-sense5 kb deletion in promoter and exon 1 (the French-Canadian-mutation), originating from the Bas-Saint-Laurent region

nd introduced in the SLSJ population in the early 20th cen-ury [20]. This last mutation accounts for 20% of FH cases.hree other mutations explain an additional 15% of molec-lar diagnosed FH in SLSJ and are related to a founderffect as well [7]. After the exclusion of diabetic individ-als (approximately 7% of the adult FH population) andhose with renal failure, this study comprised a sample of68 French-Canadian adult FH subjects (283 men and 286

omen) from the SLSJ region in Quebec, Canada (Table 1).ll these patients were evaluated at the Chicoutimi Hospitalipid Clinic. FH diagnosis was based on: (1) the presencef a mutation in the LDLR gene; or (2) the presence of a

With CAD (N = 177) p-Value

89.8 ± 12.5 <0.00131.6 NS

9.12 ± 1.85 NS1.05 ± 0.30 <0.0017.13 ± 1.80 NS1.81 ± 1.02 0.0071.49 ± 0.42 NS

127.4 ± 19.8 0.00478.9 ± 10.9 NS

6.47 ± 4.62 0.004

ing medication. NS: p > 0.1. Data were adjusted for age and gender.

de and adiponectin levels.

2 scleros

moLacfiMLidCdfioabftAaoctCtiptatcdaatedBpttp1tawwtatbtFti

E(p[siaiC

2

aartecpaPpafHtyygeseweGtaiataw

3

Owwt

64 T. Bouhali et al. / Athero

utation in the LDLR gene in a first degree relative and onef the following: typical tendinous xanthomata or plasmaDL-cholesterol over the 95th percentile in the absence ofsecondary cause of hypercholesterolemia; or (3) LDL-

holesterol over the 95th percentile and the following: aamilial history of raised plasma LDL-cholesterol transmittedn an autosomal dominant pattern and tendinous xanthomata.

olecular screening of FH included the detection of nineDLR gene mutations accounting for the majority of cases

n the province of Quebec. Two deletions (5 and >15 kb) wereetected by Southern blotting [20], and the W66G, E207K,646Y, C152W, R329X, C347R and Y468X mutations wereetected by dot-blot hybridization of genomic DNA ampli-ed by polymerase chain reaction (PCR) with allele specificligonucleotide probes or by PCR-based restriction fragmentnalysis [21–23]. The presence of LDLR gene mutations haseen confirmed in 81.5% of patients. Subjects with renalailure or diagnosed type 2 diabetes, defined according tohe World Health Organization criteria [24], were excluded.mong the 568 FH patients having met the criteria described

bove and who agreed to participate in this study, the presencef CAD was ascertained in 177 individuals on the basis of: (a)linical and ECG criteria of myocardial infarction accordingo the consensus document of the joint European Society ofardiology and the American College of Cardiology commit-

ee [25] or, (b) evidence of coronary stenosis of at least 50%n >1 main coronary arteries following coronary angiogra-hy for the investigation of ischemic heart disease [9]. Morehan half of MI patients and all non-MI (typical angina or

positive exercise tolerance test or both) FH patients ofhis study having been classified as CAD had angiographiconfirmation of their status. Smoking habits were classifiedichotomously as “non-smoker” versus “smoker”. When mett the Chicoutimi lipid clinics for the first time, all FH patientsre evaluated by a multidisiplinary team and asked to stopheir lipid lowering medication for 2–4 weeks in order tostimate the effect and compliance to the prescribed AHAiet, to obtain a complete baseline lipid profile including apoand to establish a familial metabolic profile of risk. This is

erformed under close medical and nutritional monitoring. Inhis study, 35% of CAD patients had never been evaluated athe lipid clinic before their coronary event. Thus, blood sam-les were obtained without lipid-lowering medication after a2 h-overnight fast from the antecubital vein into vacutainerubes containing EDTA. The HDL subfraction was obtainedfter precipitation of LDL (d > 1.006 g/ml) in the infranatantith heparin and MnCl2 [26]. An oral glucose tolerance testas performed following a 12-h fast [27]. Glucose, choles-

erol and triglyceride levels were enzymatically measured onMultiparity Analyzer CX7 (Beckman Instruments, Fuller-

on, CA, USA) and plasma apoB concentration was assessedy nephelometry [28]. Serum creatinine was measured using

he kinetic Jaffe reaction method (Beckman Instruments,ullerton, CA, USA). The Beckman BUN Analyser was used

o determine the urea concentration in whole blood. Fast-ng plasma adiponectin concentrations were measured with

etph

is 196 (2008) 262–269

LISA, a commercial enzyme-linked immunosorbent assayB-bridge International, Inc. San Jose, CA, USA). Anthro-ometric measures were performed as previously described29]. Subjects gave informed consent to participate in thistudy and were assigned a code that systematically denom-nalizes all clinical data and allows to built family trees andssess familial relations without having access to nominalnformation [30]. This project received the approval of thehicoutimi Hospital Ethics Committee.

.2. Statistical analysis

Due to their skewed distribution, plasma triglyceride anddiponectin values were log10-transformed before analyses,nd geometric means are presented. Conditional logisticegression models were constructed in order to investigatehe independent relationship between coronary artery dis-ase, considered as the dependent variable, and adiponectinoncentration, the presence of FH and the effect of covariatesotentially affecting adiponectin concentrations, particularlyge, gender, HDL-cholesterol and waist circumference. Theearson product-moment correlation coefficient was com-uted to quantify the association between plasma adiponectinnd HDL-cholesterol concentrations as well as waist circum-erence. Plasma adiponectin concentrations were adjusted forDL-cholesterol and waist circumference and then ranked in

ertiles according to age groups (under 40 years old, 40–60ears old, and over 60) and by gender for the survival anal-sis (Cox proportional hazards regression) since age andender are known to interfere with plasma adiponectin lev-ls [14]. Subjects in the second tertile (which included theample median and mean adiponectin values) were consid-red as the reference group to which an odds ratio of 1.0as set for comparison purposes to illustrate the modulator

ffect of adiponectinemia on age of the first coronary events.ender, smoking habit, LDL-cholesterol concentrations and

ime interval between CAD events and blood sampling fordiponectin analyses (retrospective events) were includedn multivariate models. CAD events that occurred after thediponectin measure (prospective events) were ranged at aime interval of 0. Effects of discrete variables were evalu-ted dichotomously. All statistical analyses were performedith the SPSS package (release 11.0, SPSS, Chicago III).

. Results

Features of the study population are presented in Table 1.verall, the mean plasma adiponectin concentration in FHas 8.07 ± 4.30 �g/ml. However, the group of FH patientsith CAD had significantly lower adiponectin concentra-

ions than unaffected FH individuals. Both groups presented

xtremely high but comparable plasma apo B, total choles-erol and LDL-cholesterol concentrations. In addition, FHatients with CAD tended to have lower HDL-cholesterol andigher anthropometrics values. As shown in Fig. 1, plasma

T. Bouhali et al. / Atherosclerosis 196 (2008) 262–269 265

F ist circur patients

ap(tlaalwlcFaFo

rarta(mwsjH

TRo

V

F9p

A9p

H9p

W9p

ScW

ig. 1. Association between plasma adiponectin, HDL-cholesterol, and waepresent patients without CAD (N = 391), whereas filled squares represent

diponectin concentrations were significantly correlated withlasma HDL-cholesterol levels and waist circumferencep < 0.001). Thus, all multivariate analyses took into accounthe effect of these covariates. Table 2 presents multivariateogistic analyses performed to determine the contribution ofdiponectin and FH to coronary artery disease before andfter adjustment for the waist girth and HDL-cholesterolevels. In the first model, in which the contribution of FHas examined before the inclusion of adiponectin and corre-

ates of abdominal adiposity, the relative odds of expressingoronary artery disease was 1.7-fold higher than in the non-

H group (p < 0.001). In all models, the contribution ofdiponectin to CAD risk was independant of that of FH.urther adjustment for the potentially confounding effectsf abdominal fat deposition did not substantially weaken the

LvFt

able 2elative risk (odds ratio) of coronary artery disease (<60 years) associated with famf abdominal fat deposition

ariable Odds ratio

Model 1 Mod

H 1.71 1.615% CI [1.40–1.99] [1.30-Value <0.001 <0.00

diponectina (�g/ml) 0.435% CI [0.29-Value <0.00

DL-C (mmol/l)5% CI-Value

aist girth (cm)5% CI-Value

ample size: N = 2938, including the 568 FH individuals of the present study and 2ovariates: age, gender and smoking habits. CI: confidence interval. FH: presenaist-TG: waist circonference >90 cm in men or 85 cm in women in presence of hya Plasma adiponectin values were log10-transformed.

mference in FH, adjusted for the effect of age and gender. Empty squareswith CAD (N = 177).

elation of FH or adiponectin per se to the risk of coronaryrtery disease (models 3 and 4 in Table 2). Multivariate Coxegression revealed that low plasma adiponectin concentra-ions (values in the lowest tertile as defined in Table 3) weressociated with shorter survival without CAD (hazard ratioHR) = 1.73, CI 95%: [1.19–2.53]; p = 0.004) as compared toiddle tertile (Fig. 2). This tendency was also evident whenomen and men were analysed separately, but no longer

ignificant in the female cohort due the low number of sub-ects (data not shown). This association was independent ofDL-cholesterol and waist circumference, gender, smoking,

DL-cholesterol concentration, creatinine, and time inter-al between CAD and plasma adiponectin measure (Fig. 2).ig. 3 compares plasma adiponectin concentrations between

he non-diabetic FH patients of the present study and a sam-

ilial hypercholesterolemia, plasma adiponectin concentration and correlates

el 2 Model 3 Model 4

1.61 1.48–1.99] [1.30–1.99] [1.20–1.84]1 <0.001 <0.001

0.45 0.63–0.62] [0.31–0.67] [0.42–0.94]1 <0.001 0.023

0.30 0.30[0.22–0.39] [0.22–0.39]<0.001 <0.001

1.09[0.60–1.00]NS

370 non-FH and non-diabetics subjects. All models included the followingce of familial hypercholesterolemia due to LDL-receptor gene mutation.pertriglyceridemia (fasting triglycerides >2.0 mmol/l). NS = p > 0.05.

266 T. Bouhali et al. / Atherosclerosis 196 (2008) 262–269

Table 3Tertiles of plasma adiponectin concentrations, according to gender and age groups in FH

Age (years) Men Women

Tertile 1 Tertile 2 Tertile 3 Tertile 1 Tertile 2 Tertile 3

<40N 30 30 30 35 36 35Mean 3.83 6.36 10.8 4.51 7.39 12.18Minimum 1.00 4.96 7.42 2.47 5.99 8.82Maximum 4.79 7.38 20.50 5.79 8.76 22.86

40–60N 58 58 58 47 47 47Mean 4.52 6.63 11.01 4.72 8.36 14.04Minimum 2.11 5.51 8.07 1.31 6.75 10.37Maximum 5.50 8.06 23.71 6.67 10.31 28.20

>60N 6 7 6 12 13 13Mean 5.68 9.12 13.11 5.69 8.87 14.25Minimum 3.00 7.62 10.56 2.83 7.40 11.11

20.33

P rol conc

ppleth

Ffiacadiwawpr

4

Maximum 7.02 10.49

lasma adiponectin concentrations were adjusted for plasma HDL-choleste

le of non-FH diabetic patients originating from the sameopulation. Overall, patients with diabetes had significantlyower adiponectin concentrations than FH patients. How-

ver, 40% of FH patients having adiponectin levels belowhe mean value observed in the CAD-free diabetic groupad CAD.

ig. 2. Analysis of cumulative survival regression (Cox model) of the age atrst CAD symptoms among FH patients according to tertiles of plasmadiponectin concentrations. Adiponectin values were adjusted for HDL-holesterol concentrations and waist circumference, and ranked in tertilesccording to age group (under 40, 40–60 and over 60 years old) and gen-er. Gender, smoking, LDL-cholesterol concentrations, creatinine and timenterval between adiponectinemia measure and retrospective CAD eventsere included in model as covariates. Subjects in the second tertile of

diponectin, according to Table 3, were considered as the reference group tohich an hazard ratio of one was set for comparison, whereas low and highlasma adiponectin levels are used to designate subjects in tertiles 1 and 3,espectively.

sseTatiOiattwifmiwtaitfatnptohvs

7.23 11.02 20.49

entrations and waist circumference.

. Discussion

In the present study, we observed a significant relation-hip between low plasma adiponectin and shorter CAD-freeurvival in a sample of well-characterized patients havingxtremely high plasma LDL-cholesterol and apo B levels.hese results suggest that low plasma adiponectin is associ-ted with an increased risk of premature CAD over and abovehe already exaggerated risk seen in FH patients. Adiponectins an adipocytokine displaying pleiotropic signal effects [11].ver the last few years, adiponectin has generated substantial

nterest due to its potential implication as a modulator of CADnd type 2 diabetes risk. On the one hand, it has been reportedhat plasma adiponectin levels are lower in CAD patients, andhat such low plasma adiponectin levels in diabetic subjectsere a predictor of macroangiopathy [31]. On the other hand,

t has been suggested that high levels of adiponectin are inact cardioprotective [32]. As previously proposed, severalechanisms could explain the protective role of adiponectin

n atherogenesis. Adiponectin accumulates in the vascularall when the endothelial barrier is injured and suppresses

he macrophage-to-foam cell transformation [33]. Kubota etl. also observed that intimal thickness was two-fold greatern adiponectin −/− mice than in the wild-type animals whenhe femoral artery was injured, suggesting a protective roleor adiponectin in neointimal formation [34]. Low plasmadiponectin has also been reported to be associated withhe sequential expression of metabolic syndrome compo-ents and to predict the presence of an atherogenic lipidrofile [11]. Published results have attributed the protec-ive effects of hyperadiponectinemia to its possible action

n plasma HDL-cholesterol [16,35]. Raised HDL-cholesterolas been previously associated with prolonged CAD-free sur-ival in the general population as well as in FH [36]. FHtatus per se is associated with an increased risk of premature

T. Bouhali et al. / Atherosclerosis 196 (2008) 262–269 267

Fig. 3. Plasma adiponectin concentrations in non-diabetic FH patients vs. non-FH patients with Type 2 diabetes. Overall, non-diabetic FH patients tended tohave higher adiponectin values than patients with type 2 diabetes. However, 40% of FH patients of this study who had plasma adiponectin concentration belowthe mean value observed among CAD-free diabetic patients had CAD. This suggests that low plasma adiponectin may exacerbate premature CAD over anda sma LD(

CiosrhaehotcscatednatAewc

p

iyswtestlnmtTaIwcowdcsa

bove the already exaggerated risk seen in FH, a genetic model of raised pla±S.E.M.) are shown.

AD [37,38] but several risk factors – age, smoking, fam-ly history of premature CAD, abdominal obesity, diabetesr insulin resistance, low HDL-cholesterol, high blood pres-ure and lipoprotein (a), among others – contribute to CADisk modulation in FH [38]. To our knowledge, no studiesave specifically assessed the role of adiponectin (alone orssociated with other emerging CAD risk markers) in FHxpression. Although premature CAD risk is significantlyigher among FH subjects, substantial variability has beenbserved in the age at onset of atherosclerotic manifesta-ions, even in patients sharing the same genetic defect andomparable plasmatic LDL-cholesterol levels [3,38]. Thisuggests that a single assessment of raised plasma LDL-holesterol, at a single, specific moment and context, cannotlone explain the prematurity of the CAD manifestations. Inhis regard, we have previously reported that FH subjectsxhibiting hyperinsulinemia and abdominal obesity are atramatically higher risk of CAD than those with neither FHor abdominal obesity [9]. Adiponectin, a correlate of obesitynd insulin resistance, is likely to become a relevant media-or of the relationship between abdominal obesity and CAD.lthough plasma adiponectin concentrations might be influ-

nced by renal function [37], the association of adiponectin

ith premature CAD expression in FH was independent of

reatinine and blood urea nitrogen in this cohort.Some aspects of our study require further discussion. Tem-

orality of association was an important issue to consider

wfmv

L-cholesterol and premature CAD. Geometric means of adiponectin values

n this study. It is documented that over a period of 10ears, adiponectinemia does not fluctuate much in healthyubjects below 70 years of age [39]. In the present study,e have evaluated mean adiponectinemia and its distribu-

ion in FH subjects of various age groups; age structureffect was included in our analysis models. In addition,ince plasma adiponectin levels were assessed relative tohe CAD event, regression models have included the timeapse so as to adjust results for this difference. Most coro-ary events that occurred retrospectively to the adiponectineasure did so within a 5-year interval, and it is unlikely

hat adiponectinemia fluctuated much within that period.o ensure that results suffered no bias due to the relatedpproach, different survival analysis models were elaborated.n fact, the association reported herein remained the samehether the retrospective or prospective CAD events were

onsidered separately or whether only the events that hadccurred less than 3 years before the adiponectin measuresere taken into account. As shown in Table 3, for both gen-ers, there was a significant trend for plasma adiponectinoncentration to increase with age in FH. The relation-hip between adiponectin and age remained significant afterdjustment for plasma HDL-cholesterol concentrations and

aist circumference (p < 0.001). This observation can result

rom a survival effect, FH patients with higher values beingore likely to survive longer. In addition to a potential sur-

ival effect, it is noteworthy that an increase in adiponectin

2 scleros

cSlsepptitwirata[mifia

oeec

D

oaAirrf

A

BftGDonafcILC

(t#

R

[

[

[

[

[

[

[

[

68 T. Bouhali et al. / Athero

oncentration with aging has previously been reported [40].everal factors may contribute to an elevation of adiponectin

evels with aging, including decreased concentration of sexteroid hormones and lower inhibition of testosterone andstrogen on adiponectin production [41,42]. However, theresent study was not designed to answer this issue and arospective study would be more suitable to clarify the rela-ionship between age and plasma adiponectin concentrationn FH. Finally, although a large number of studies have inves-igated the implication of adiponectin in vascular remodeling,e still do not have enough evidence to consider this promis-

ng protein as clinically useful in the assessment of CADisk. Furthermore, methods to quantify this adipocytokinere not yet standardized. The native protein circulates inwo main forms: a lower molecular weight (LMW) hexamernd multimeric structure of high molecular weight (HMW)43]. Further studies are needed to determine the best clinicalarker between HMW and LMW. Taking these limitations

n consideration, our results nevertheless provide, for therst time, evidence that adiponectin concentration might becovariate of premature CAD risk expression in FH.

In summary, our findings are consistent with clinicalbservations suggesting that low plasma adiponectin mayxacerbate premature CAD over and above the alreadyxaggerated risk seen in FH, a model of raised plasma LDL-holesterol and premature CAD.

isclosure

Dr. Despres is a consultant for Sanofi Aventis. The Chairn Cardiometabolic Risk at Universite Laval is supported byn unrestricted grant awarded to Universite Laval by Sanofiventis. The Chair is an academic organisation completely

ndependent from Sanofi Aventis and the authors did noteceive any contribution from Sanofi Aventis to support thisesearch. This project was supported by an unrestricted grantrom AstraZeneca Canada Inc. received by D. Gaudet.

cknowledgments

J. St-Pierre is the recipient of the “Walter & Jessieoyd & Charles Scriver” M.D./Ph.D. Studentship Award

rom the Canadian Institutes of Health Research (CIHR),he Canadian Genetic Diseases Network, the Canadianene Cure Foundation, and Theratechnologies (A. Jeanegrandpre Scholarship Award). C. Laprise is the chairholderf the Canada Research Chair for genetic determi-ants in asthma (http://www.chairs.gc.ca). MC. Vohl isresearch scholar from the FRSQ. J.P Despres is pro-

essor of human nutrition, lipidology and prevention of

ardiovascular disease and the Scientific Director of thenternational Chair on Cardiometabolic Risk at Universiteaval. D. Gaudet is the chairholder of the Canada Researchhair in preventive genetics and community genomics

[

[

is 196 (2008) 262–269

http://www.chairs.gc.ca). This project was supported byhe ECOGENE-21 project from the CAHR program (grantCAR43283) and AstraZeneca Canada Inc.

eferences

[1] Abifadel M, Varret M, Rabes JP, et al. Mutations in PCSK9 cause auto-somal dominant hypercholesterolemia. Nat Genet 2003;34(2):154–6.

[2] Garcia CK, Wilund K, Arca M, et al. Autosomal recessive hyperc-holesterolemia caused by mutations in a putative LDL receptor adaptorprotein. Science 2001;292(5520):1394–8.

[3] Goldstein J. Familial hypercholesterolemia. In: Scriver CB, BeaudetAL, Sly WS, Valle D, editors. The metabolic basis of inherited dis-ease. 8th ed. New York: McGraw-Hill Book Co.; 2000. p. 2863–913.

[4] Moorjani S, Roy M, Torres A, et al. Mutations of low-density-lipoprotein-receptor gene, variation in plasma cholesterol, andexpression of coronary heart disease in homozygous familial hyper-cholesterolaemia. Lancet 1993;341(8856):1303–6.

[5] Lansberg PJ, Tuzgol S, van de Ree MA, Defesche JC, Kastelein JJ.Higher prevalence of familial hypercholesterolemia than expected inadult patients of four family practices in Netherlands. Ned TijdschrGeneeskd 2000;144(30):1437–40.

[6] Moorjani S, Roy M, Gagne C, et al. Homozygous familial hyper-cholesterolemia among French Canadians in Quebec Province.Arteriosclerosis 1989;9(2):211–6.

[7] Vohl MC, Moorjani S, Roy M, et al. Geographic distribution ofFrench-Canadian low-density lipoprotein receptor gene mutations inthe Province of Quebec. Clin Genet 1997;52(1):1–6.

[8] Reaven G. Metabolic syndrome: pathophysiology and impli-cations for management of cardiovascular disease. Circulation2002;106(3):286–8.

[9] Gaudet D, Vohl MC, Perron P, et al. Relationships of abdominal obe-sity and hyperinsulinemia to angiographically assessed coronary arterydisease in men with known mutations in the LDL receptor gene. Cir-culation 1998;97(9):871–7.

10] Jansen AC, van Aalst-Cohen ES, Tanck MW, et al. The contribution ofclassical risk factors to cardiovascular disease in familial hypercholes-terolaemia: data in 2400 patients. J Intern Med 2004;256(6):482–90.

11] Trujillo ME, Scherer PE. Adiponectin—journey from an adipocytesecretory protein to biomarker of the metabolic syndrome. J InternMed 2005;257(2):167–75.

12] Arita Y, Kihara S, Ouchi N, et al. Paradoxical decrease of anadipose-specific protein, adiponectin, in obesity. Biochem Biophys ResCommun 1999;257(1):79–83.

13] Nakano Y, Tobe T, Choi-Miura NH, Mazda T, Tomita M. Isolationand characterization of GBP28, a novel gelatin-binding protein purifiedfrom human plasma. J Biochem (Tokyo) 1996;120(4):803–12.

14] Cnop M, Havel PJ, Utzschneider KM, et al. Relationship of adiponectinto body fat distribution, insulin sensitivity and plasma lipopro-teins: evidence for independent roles of age and sex. Diabetologia2003;46(4):459–69.

15] Kondo H, Shimomura I, Matsukawa Y, et al. Association of adiponectinmutation with type 2 diabetes: a candidate gene for the insulin resistancesyndrome. Diabetes 2002;51(7):2325–8.

16] Cote M, Mauriege P, Bergeron J, et al. Adiponectinemia in visceralobesity: impact on glucose tolerance and plasma lipoprotein and lipidlevels in men. J Clin Endocrinol Metab 2005;90(3):1434–9.

17] Scriver CR. Human genetics: lessons from Quebec populations. AnnuRev Genomics Hum Genet 2001;2:69–101.

18] Gaudet D, Tremblay G, Perron P, et al. Familial hypercholesterolemia ineastern Quebec: a public health problem? The experience of the hyper-lipidemia clinic of Chicoutimi. Union Med Can 1995;124(2):54–60.

19] Gaudet D, Gagne C, Perron P, Couture P, Tonstad S. Ethical issues inmolecular screening for heterozygous familial hypercholesterolemia:

scleros

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

T. Bouhali et al. / Athero

the complexity of dealing with genetic susceptibility to coronary arterydisease. Community Genet 1999;2(1):2–8.

20] Ma YH, Betard C, Roy M, Davignon J, Kessling AM. Identifica-tion of a second “French Canadian” LDL receptor gene deletion anddevelopment of a rapid method to detect both deletions. Clin Genet1989;36(4):219–28.

21] Couture P, Vohl MC, Gagne C, et al. Identification of three mutationsin the low-density lipoprotein receptor gene causing familial hyper-cholesterolemia among French Canadians. Hum Mutat 1998;(Suppl1):S226–31.

22] Leitersdorf E, Tobin EJ, Davignon J, Hobbs HH. Common low-densitylipoprotein receptor mutations in the French Canadian population. JClin Invest 1990;85(4):1014–23.

23] Vohl MC, Couture P, Moorjani S, et al. Rapid restriction fragment anal-ysis for screening four point mutations of the low-density lipoproteinreceptor gene in French Canadians. Hum Mutat 1995;6(3):243–6.

24] W.H.O. Definition, diagnosis and classification of diabetes mellitus andits complications: Report of a WHO Consultation. Part 1. Diagnosis andclassification of Diabetes mellitus. Geneva: World Health Organization;1999.

25] Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarctionredefined—a consensus document of The Joint European Society ofCardiology/American College of Cardiology Committee for the redefi-nition of myocardial infarction. J Am Coll Cardiol 2000;36(3):959–69.

26] Havel RJ, Eder HA, Bragdon JH. The distribution and chemical com-position of ultracentrifugally separated lipoproteins in human serum. JClin Invest 1955;34(9):1345–53.

27] Report of the Expert Committee on the Diagnosis and Classification ofDiabetes Mellitus. Diabetes Care, 1997;20(7):1183–97.

28] Laurell CB. Quantitative estimation of proteins by electrophoresis inagarose gel containing antibodies. Anal Biochem 1966;15(1):45–52.

29] Standardization of anthropometric measurements.Lohman TRA, Mar-torel R, editors. The Airlie (VA) consensus conference. 1988. p. 39–80.

30] Gaudet D, Arsenault S, Belanger C, et al. Procedure to protect confi-

dentiality of familial data in community genetics and genomic research.Clin Genet 1999;55(4):259–64.

31] Hotta K, Funahashi T, Arita Y, et al. Plasma concentrations of a novel,adipose-specific protein, adiponectin, in type 2 diabetic patients. Arte-rioscler Thromb Vasc Biol 2000;20(6):1595–9.

[

is 196 (2008) 262–269 269

32] Pischon T, Girman CJ, Hotamisligil GS, et al. Plasma adiponectin levelsand risk of myocardial infarction in men. JAMA 2004;291(14):1730–7.

33] Ouchi N, Kihara S, Arita Y, et al. Adipocyte-derived plasma pro-tein, adiponectin, suppresses lipid accumulation and class A scavengerreceptor expression in human monocyte-derived macrophages. Circu-lation 2001;103(8):1057–63.

34] Kubota N, Terauchi Y, Yamauchi T, et al. Disruption of adiponectincauses insulin resistance and neointimal formation. J Biol Chem2002;277(29):25863–6.

35] Rothenbacher D, Brenner H, Marz W, Koenig W. Adiponectin, riskof coronary heart disease and correlations with cardiovascular riskmarkers. Eur Heart J 2005;26(16):1640–6.

36] Gaudet D, Bouhali T, Couture P, et al. HDL-cholesterol and othercovariables associated with CAD-free survival among familial hyper-cholesterolemic heterozygotes or homozygotes over 70 years of age.Atherosclerosis Suppl 2003;4(2):333–4 (Abstract).

37] Isobe T, Saitoh S, Takagi S, et al. Influence of gender, age and renalfunction on plasma adiponectin level: the Tanno and Sobetsu study. EurJ Endocrinol 2005;153(1):91–8.

38] Jansen AC, van Wissen S, Defesche JC, Kastelein JJ. Phenotypic vari-ability in familial hypercholesterolaemia: an update. Curr Opin Lipidol2002;13(2):165–71.

39] Adamczak M, Rzepka E, Chudek J, Wiecek A. Ageing and plasmaadiponectin concentration in apparently healthy males and females.Clin Endocrinol (Oxford) 2005;62(1):114–8.

40] CnopF M.H.P., Utzschneider KM, Carr DB, et al. Relationship ofadiponectin to body fat distribution, insulin sensitivity and plasmalipoproteins: evidence for independent roles of age and sex. Diabetolo-gia 2003;46(4):459–69.

41] Nishizawa H, Shimomura I, Kishida K, et al. Androgens decreaseplasma adiponectin, an insulin-sensitizing adipocyte-derived protein.Diabetes 2002;51(9):2734–41.

42] Combs TP, Berg AH, Rajala MW, et al. Sexual differentiation, preg-nancy, calorie restriction, and aging affect the adipocyte-specific

secretory protein adiponectin. Diabetes 2003;52(2):268–76.

43] Pajvani UB, Du X, Combs TP, et al. Structure-function studiesof the adipocyte-secreted hormone Acrp30/adiponectin. Implica-tions for metabolic regulation and bioactivity. J Biol Chem2003;278(11):9073–85.