Nutrition, Metabolism & Cardiovascular Diseases (2016) 26, 223e231

Available online at www.sciencedirect.com

Nutrition, Metabolism & Cardiovascular Diseases

journal homepage: www.elsevier.com/locate/nmcd

Metabolic syndrome and 10-year cardiovascular disease incidence:

The ATTICA study
C.-M. Kastorini a, D.B. Panagiotakos a,*, E.N. Georgousopoulou a, A. Laskaris a,
N. Skourlis a, A. Zana a, C. Chatzinikolaou a, C. Chrysohoou b, P.E. Puddu c, D. Tousoulis b,
C. Stefanadis b, C. Pitsavos b, the ATTICA Study Group
a

Department of Nutrition and Dietetics, Harokopio University, Athens, Greece

First Cardiology Clinic, School of Medicine, University of Athens, Greece
Laboratory of Biotechnologies Applied to Cardiovascular Medicine, Department of Cardiovascular, Respiratory, Nephrological, Anesthesiological and
Geriatrical Sciences, Sapienza, University of Rome, Italy

b
c

Received 8 August 2015; received in revised form 12 December 2015; accepted 15 December 2015
Available online 23 December 2015

KEYWORDS
Metabolic syndrome;
Cardiovascular
disease;
Incidence;
Inammatory
markers;
Renal markers

Abstract Aims: To evaluate the inuence of metabolic syndrome (MetS) as well as inammatory and renal markers on cardiovascular disease (CVD) incidence.
Methods and results: During 2001e2002, 1514 men and 1528 women (>18y) without any clinical
evidence of CVD or any other chronic disease, at baseline, living in greater Athens area, Greece,
were enrolled. In 2011e2012, the 10-year follow-up was performed in 2583 participants (15% of
the participants were lost to follow-up). Incidence of fatal or non-fatal CVD was dened according to WHO-ICD-10 criteria. MetS was dened using three denitions, provided by the National
Cholesterol Education Program Adult Treatment panel III (revised NCEP ATP III), the International
Diabetes Federation (IDF) or the Harmonized denition. Furthermore, the contributory predictive
role of C-reactive protein (CRP), inteleukin-6, uric acid and estimated glomerular ltration rate in
the aforementioned models was evaluated. History of MetS-NCEP was positively associated with
CVD, adjusting for potential confounding factors (OR:1.83, 95%CI:1.24e2.72). Not statistically signicant associations with CVD incidence were observed when using the IDF or the Harmonized
denition. Additionally, none of the added inammatory and renal function markers mediated
the inuence of MetS on CVD incidence (all ps from Sobel test >0.40). C-statistic values for
the MetS denitions used exceeded 0.789 (CI:0.751e0.827), indicating fair-to-good predictive
probability of the models.
Conclusion: Results of the present work revealed the negative impact of MetS-NCEP, but not of
the other MetS denitions, on CVD incidence, a key-point that may help in better understanding
the role of IDF and Harmonized MetS denitions on CVD.
2015 The Italian Society of Diabetology, the Italian Society for the Study of Atherosclerosis, the
Italian Society of Human Nutrition, and the Department of Clinical Medicine and Surgery, Federico II University. Published by Elsevier B.V. All rights reserved.

Introduction
The metabolic syndrome (MetS) is a complex disorder,
characterized by a cluster of interacting parameters

including insulin resistance, central obesity, dyslipidemia,

endothelial dysfunction, elevated blood pressure levels,
hypercoagulation and chronic stress [1,2]. It is a common
condition, present in approximately 20e30% of the global

* Corresponding author. 46 Paleon Polemiston St. Glyfada, Attica, 166 74, Greece. Tel.: 30 210 9549332; fax: 30 210 9600719.
E-mail address: d.b.panagiotakos@usa.net (D.B. Panagiotakos).
http://dx.doi.org/10.1016/j.numecd.2015.12.010
0939-4753/ 2015 The Italian Society of Diabetology, the Italian Society for the Study of Atherosclerosis, the Italian Society of Human Nutrition, and the Department of
Clinical Medicine and Surgery, Federico II University. Published by Elsevier B.V. All rights reserved.

224

adult population [3]. During the previous years, the scientic evidence has suggested that the MetS is associated
with elevated levels of various inammatory markers [4].
In fact, low grade inammation is considered as one of the
main characteristics of the syndrome, caused by the
interaction of genetic and environmental factors [1]. Taking into consideration the signicant impact of the syndrome on public health, as its presence has been
associated with the development of atherosclerosis [5] and
most importantly with increased risk of cardiovascular
disease (CVD) [6,7], better understanding of its underlying
parameters is of crucial importance. In particular, understanding the impact of oxidative and inammatory
markers in the predictive ability of the cluster of the
known components might enable earlier diagnosis or
better identication of high-risk patients. Furthermore,
hyperuricemia has been associated with increased risk of
coronary heart disease and stroke [8,9]. Equally important,
impaired kidney function, expressed as estimated
glomerular ltration rate (eGFR) < 60 ml/min/1.73 m2 increases the risk of all cause and CVD mortality [10]. At the
same time, subjects with MetS have higher risk of renal
dysfunction, while uric acid levels have been suggested to
exert a causal role for MetS development [11,12].
Thus, the aim of the present work was to evaluate
whether the aforementioned markers inuence the predictive ability of the MetS regarding CVD development. In
more detail, the inuence of inammatory markers, uric
acid and eGFR, on the CVD incidence was examined in the
10 year follow-up of the ATTICA study [13]. The inuence
of MetS denitions, provided by the National Cholesterol
Education Program Adult Treatment panel III (revised
NCEP ATP III) [14], the International Diabetes Federation
(IDF) [15] or the International Diabetes Federation Task
Force on Epidemiology and Prevention, the National Heart,
Lung, and Blood Institute (NHLBI), the American Heart
Association (AHA), the World Heart Federation, the International Atherosclerosis Society and the International Association for the Study of Obesity (Harmonized criteria) [2]
were evaluated. Specically, the research hypothesis was
to reveal which of the inherent, pathophysiological characteristics that differentiate the metabolic syndrome definitions, may better predict CVD incidence; furthermore, to
reveal if cardiometabolic markers, and particularly those
associated with inammatory effects and renal function,
may have an additional explanatory value to the 10-year
CVD risk. Both research hypotheses have been rarely
investigated in the literature.
Methods
Sampling procedure at baseline examination
The ATTICA study was carried out in the greater metropolitan Athens area (including 78% urban and 22% rural
regions) during 2001e2002 [16]. According to the protocol, the sampling procedure anticipated enrolling only one
participant per household; it was random, multistage and
based on the age (5 strata), sex (2 strata) distribution of the

C.-M. Kastorini et al.

Attica region (27 strata, census of 2001). Of the 4056

invited individuals at baseline examination, 3042 agreed
to participate (75% participation rate); 1514 of the participants were men (18e87 years) and 1528 were women
(18e89 years). All participants were interviewed by
trained personnel (cardiologists, general practitioners, dieticians and nurses) who used a standard questionnaire.
Exclusion of CVD at baseline evaluation was performed
through a detailed clinical evaluation by the physicians of
the study, following standard criteria. The examination
was performed in the individuals homes or working place.
Baseline measurements
The baseline evaluation included information about sociodemographic characteristics (age, sex, nancial status and
years of school), personal and family history of hypertension, hypercholesterolemia and diabetes, family history of
CVD, dietary and other lifestyle habits (i.e., smoking status
and physical activity). The evaluation of the nutritional
habits was based on a validated semi-quantitative foodfrequency questionnaire [17], the EPIC-Greek questionnaire that was kindly provided by the Unit of Nutrition of
Athens Medical School. All participants were asked to
report the average intake (per week or day) of several food
items that they consumed (during the last 12 months). The
MedDietScore was also applied (range 0e55) to evaluate
adherence to the Mediterranean diet [18]. Smokers were
dened as those who were smoking at least one cigarette
per day during the past year or had recently stopped
smoking (within the last 12 months); the rest of the participants were dened as non-smokers. For the ascertainment of physical activity status the International Physical
Activity Questionnaire was used (IPAQ, [19,20]), as an
index of weekly energy expenditure using frequency
(times per week), duration (in minutes per time) and intensity of sports or other habits related to physical activity
(in expended calories per time). Participants who did not
report any physical activities were dened as physically
inactive (sedentary lifestyle). Waist circumference was
measured in the middle between the lowest rib and the
iliac crest using an inelastic measuring tape to the nearest
0.5 cm. Hip circumference (to the nearest 0.5 cm) was also
measured. Height was measured, to the nearest 0.5 cm,
without shoes, back square against the wall tape, eyes
looking straight ahead, with a right-angles triangle resting
of the scalp and against the wall. Weight was measured
with a lever balance, to the nearest 100 g, without shoes,
in light undergarments. Body mass index (BMI) was
calculated as weight (in kilograms) divided by standing
height (in meters squared). Obesity was dened as BMI
greater than 29.9 kg/m2. Arterial blood pressure (3 recordings) was measured at the end of the baseline physical
examination in a sitting position after resting for at least
30 min. Participants whose average blood pressure levels
were greater or equal to 140/90 mmHg or were under
antihypertensive medication were classied as having
hypertension. Blood samples were collected from the
antecubital vein between 8 and 10 a.m., in a sitting

Metabolic syndrome and 10-year cardiovascular disease incidence

position after 12 h of fasting and alcohol abstinence. Total

serum cholesterol, HDL-cholesterol, and triglycerides were
measured using chromatographic enzymic method in a
Technicon automatic analyzer RA-1000 (Dade Behring,
Marburg, Germany). Hypercholesterolemia was dened as
total cholesterol levels greater than 200 mg/dl or the use
of lipids lowering agents. Blood glucose levels (mg/dl)
were measured with a Beckman Glucose Analyzer (Beckman Instruments, Fullerton, CA, USA). Diabetes mellitus
(type 2) was dened according to the American Diabetes
Association diagnostic criteria (i.e., blood glucose levels
greater than 125 mg/dl classied participants as having
diabetes). High sensitivity C-reactive protein (CRP) was
assayed by particle-enhanced immunonephelometry (N
Latex, Dade-Behring Marburg GmbH, Marburg, Germany)
with a range from 0.175 to 1100 mg/dl. Furthermore, participants with chronic inammation or CRP levels above
10 mg/L were not included in the analyses. Interleukin-6
(Il-6) was measured with high sensitivity enzyme linked
immunoassay (R & D Systems Europe Ltd, Abingdon, U.K.)
with a range from 0.156 to 10 pg/ml. The intra-assay and
inter-assay coefcient of variation was <5% for CRP and
<10% for interleukin-6. Serum creatinine was measured in
serum, using a colorimetric method (BioAssay Systems,
Hayward, CA). Serum uric acid was measured using an
enzymatic colorimetric test through the uricaseperoxidase method (UA plus, Roche Diagnostics, Manheim, Germany). The measuring range was 0.2e25 mg/dl
and the inter- and intra-assay variability were 0.5% and
1.7% respectively. Renal function was evaluated according
to eGFR, which describes the ow rate of ltered uid
through the kidney. However, as data on eGFR were not
available in this study, the eGFR was calculated using the
Chronic Kidney Disease Epidemiology Collaboration (CKDEPI) Creatinine Equation: GFR Z [141  min(Scr/k, 1)
a  max(Scr/k, 1)-1.209  0.993Age  1.018 (if
female)  1.159 (if black)], where Scr is serum creatinine, k
is 0.7 for females and 0.9 for males, a is 0.329 for females
and 0.411 for males, min indicates the minimum of Scr/
kor 1, and max indicates the maximum of Scr/k or 1 [21].
Further details about the aims and procedures of the
ATTICA epidemiological study may be found elsewhere
[16].
Denition of the metabolic syndrome
Participants were classied as having the MetS or not,
according to the following three denitions provided by:

225

150 mg/dl; HDL cholesterol level <40 mg/dL for

males or <50 mg/dL for females; blood pressure 130/
85 mmHg; fasting blood glucose 100 mg/dL [15].
c) the Harmonized criteria; three or more of the following
metabolic components present: waist circumference
94 cm for males or 80 cm for females (using the IDF
cut points); triglyceride level 150 mg/dl;
HDL cholesterol level <40 mg/dL for males or <50 mg/
dL for females; blood pressure 130/85 mmHg;
fasting blood glucose 100 mg/dL [2].
Participants under drug treatment for elevated triglycerides levels or reduced HDL-cholesterol, or elevated
blood pressure levels or for hyperglycemia were considered as meeting the aforementioned criteria, respectively.
Follow-up examination (2011e2012)
During 2011e12, the ATTICA Studys investigators performed the 10-year follow-up (mean follow-up time
8.41 y) [13]. Of the n Z 3042 initially enrolled participants,
n Z 2583 were found during the follow-up (85% participation rate). Of the individuals that were lost to follow-up
(i.e., n Z 459), n Z 224 were not found because of missing
or wrong addresses and telephone numbers that they have
provided at baseline examination and n Z 235 because
they denied being re-examined. Regarding CVD evaluation
at follow-up clinically accurate data were obtained from
n Z 2020 participants.
In order to participate in the follow-up all participants
were initially appointed through telephone calls. Afterwards, the investigators approached the n Z 2583 participants that were allocated in the follow-up and
performed a detailed evaluation of their medical records.
In particular, information about participants: (a) vital
status (death from any cause or due to CVD), (b) development of coronary heart disease (CHD, i.e., myocardial
infarction, angina pectoris, other identied forms of
ischemia -WHO-ICD coding 410e414.9, 427.2, 427.6-, heart
failure of different types, and chronic arrhythmias -WHOICD coding 400.0e404.9, 427.0e427.5, 427.9-), (c) development of stroke (WHO-ICD coding 430e438), (d) development of hypertension, hypercholesterolemia and
diabetes, as well as management of these conditions, (e)
assessment of body weight and height, and (f) lifestyle
habits, including physical activity and smoking status, as
well as dietary habits, was assessed.
Statistical analysis

a) NCEP ATP III (revised ); three or more of the following

metabolic components present: waist circumference
102 cm for males or 88 cm for females; triglyceride level 150 mg/dl; HDL cholesterol level
<40 mg/dL for males or <50 mg/dL for females;
blood pressure 130/85 mmHg; fasting blood
glucose 100 mg/dL [14].
b) the IDF Epidemiology Task Force group; increased waist
circumference (94 cm for males or 80 cm for females) plus two of the following: triglyceride level

Non-fatal and fatal incidence rates of combined CVD (i.e.,

CHD or stroke) were calculated as the ratio of new cases to
the number of people that participated in the follow-up.
Distribution of the continuous variables were presented as
mean values  standard deviation and categorical variables
are presented as frequencies. The continuous variables were
tested for normality through PeP plots. Associations between categorical variables were tested using the chi-square
test. Comparisons between mean values of normally

p**

34 (1.1%)
4.07  3.09
1.76  0.33
5.50  1.29
76.38  17.30
33.1  5.36
110.03  11.90
151.85  68.84
18.88  6.41
261.35  124.56
36.53  5.40
147.42  14.04
92.27  8.30
159 (5.2%)
2.87  2.72
1.69  0.40
5.32  1.55
74.73  18.09
29.7  3.82
105.46  12.55
112.02  32.86
15.21  3.08
219.39  105.88
36.94  8.05
138.46  17.32
88.97  11.44
414 (13.6%)
2.77  2.72
1.59  0.55
4.85  1.39
77.35  17.04
29.6  4.57
102.69  11.89
99.40  24.68
14.11  4.79
163.62  123.59
43.65  16.00
131.57  18.04
84.13  10.76
2435 (80%)
1.69  2.26
1.41  0.55
4.08  1.29
81.91  16.42
25.5  4.06
86.37  13.75
89.90  20.32
12.74  3.46
100.10  62.92
50.33  14.13
119.99  17.59
77.13  10.86

4/5

5/5

<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001

p-trend*

mU: microUnit, cm: centimeter, DBP: diastolic blood pressure, dL: deciliter, eGFR: estimated Glomerular ltration rate, L: liter, m: meter, min: minute, mg: milligram, mL: milliliter, mm: millimeter, pg:
picogram, SBP: systolic blood pressure, WC: waist circumference. *p-values for the comparisons between participants with 0, 3, 4 or 5 metabolic syndrome components; **p-values for the comparison
between participants with the metabolic syndrome and metabolic syndrome free subjects. z KruskaleWallis test.

During the 10-year follow-up, n Z 317 (15.7%) had a fatal

or non-fatal CVD event; of them, n Z 198 (19.7%) cases
were men and n Z 119 (11.7%) cases were women (p for
gender difference <0.001). In Tables 2e4, results from
multiple logistic regression analyses are presented,
evaluating the likelihood of having a cardiovascular event
according to the MetS presence, based on different denitions. When evaluated with the NCEP denition, history of MetS was signicantly associated with
cardiovascular disease events, even after adjusting for
several potential confounding factors (Model 2). However, when using the IDF or the newest Harmonized

N (%)
C-Reactive Protein (mg/L)z
Interleukin-6 (pg/mL)
Uric acid (mg/dL)
eGFR (mL/min/1.73 m2)
BMI (kg/m2)
WC (cm)
Glucose (mg/dL)
Insulin (mU/mL)z
Triglycerides (mg/dL)z
HDL-cholesterol (mg/dL)
SBP (mmHg)
DBP (mmHg)

Cardiovascular disease incidence

3/5

At baseline, prevalence of the MetS was 20.0% according

to the revised NCEP ATP III denition, 48.9% according to
the IDF denition and 51.0% according to the Harmonized
criteria using the IDF cut points for waist circumference.
Furthermore, levels of inammatory markers and other
parameters, according to the number of MetS components, based on the NCEP denition, are presented in
Table 1.

Metabolic syndrome presence (NCEP-R) (number of MetS components)

Baseline characteristics

MetS free

Results

Table 1 Levels of inammatory indicators and other factors, according to the number of MetS components (based on the NCEP-2005 denition).

distributed variables between those who developed an

event and the rest of the participants were performed using
Students t-test, after controlling for equality of variances
using the Levenes test, while comparisons of continuous
variables that did not follow the normal distribution were
performed using the non-parametric ManneWhitney-UTest. Comparisons between normally distributed continuous variables and groups of people were performed by the
calculation of ANOVA test, after controlling for equality of
variances, whereas comparisons of continuous variables
that did not follow the normal distribution were performed
using the non-parametric KruskaleWallis test. Estimations
of the relative odds of having a cardiovascular event according to history of MetS and other covariates were performed through multiple logistic regression analysis;
results are presented as odds ratios and the corresponding
95% condence intervals (CI). The renal and inammatory
markers were included in each model separately in order to
avoid collinearity. HosmereLemeshow statistic was calculated to evaluate models goodness-of-t. The estimating
accuracy of the models in correctly classifying the participants as those who developed a CVD event and those who
did not was measured using the C-statistic [22]. Sobels test
was applied to test the signicance of the mediation effects
of various biological and clinical markers on the tested hypotheses [23]. All reported p-values are based on two-sided
tests and compared to a signicance level of 5%. SPSS
version 21 (Statistical Package for Social Sciences, SPSS Inc,
Chicago, IL, U.S.A.) software was used for all the statistical
calculations.

C.-M. Kastorini et al.

<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001

226

NCEP denition

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

Age (per 1 year)

Men vs. Women
History of metabolic syndrome (y/n)
Family history of CVD (y/n)
Physically active vs. sedentary
MedDietScore (per 1/55 unit)
Smokers vs. non-smokers
C-Reactive Protein (per 1 mg/L)
Interleukin-6 (per 1 pg/mL)
Uric acid (per 1 mg/dL)
eGFR (per 1 unit)

1.09 (1.07e1.10)**
1.89 (1.44e2.50)**
1.76 (1.31e2.32)**
e
e
e
e
e
e
e
e

1.08 (1.06e1.10)**
1.57 (1.06e2.32)*
1.83 (1.24e2.72)*
1.37 (0.92e2.03)
1.26 (0.87e1.84)
0.98 (0.94e1.01)
1.65 (1.11e2.45)*
e
e
e
e

1.09 (1.07e1.11)**
1.50 (0.99e2.26)
1.72 (1.13e2.61)*
1.26 (0.83e1.92)
1.29 (0.87e1.92)
0.97 (0.94e1.01)
1.60 (1.06e2.42)*
1.06 (0.99e1.14)
e
e
e

1.08 (1.06e1.10)**
1.49 (0.99e2.25)
1.78 (1.18e2.68)*
1.25 (0.83e1.90)
1.24 (0.84e1.84)
0.97 (0.94e1.00)
1.61 (1.07e2.42)*
e
1.43 (0.81e2.52)
e
e

1.08 (1.06e1.10)**
1.54 (0.98e2.41)
1.81 (1.20e2.73)*
1.37 (0.91e2.04)
1.33 (0.91e1.96)
0.98 (0.94e1.01)
1.59 (1.06e2.38)*
e
e
1.05 (0.90e1.22)
e

1.07 (1.05e1.09)**
1.85 (1.21e2.82)*
1.85 (1.24e2.76)*
1.33 (0.89e1.99)
1.24 (0.84e1.82)
0.98 (0.95e1.01)
1.70 (1.14e2.54)*
e
e
e
0.99 (0.97e1.00)*

Results are presented as OR (95%CI) obtained from logistic regression. *: p  0.05, **: p  0.01. Bold values highlight the main result (i.e., History of metabolic syndrome).
CVD: cardiovascular disease, eGFR: estimated Glomerular ltration rate.
No signicant mediation effect of the additional variables in the model that contained MetS was observed based on Sobels test.

Metabolic syndrome and 10-year cardiovascular disease incidence

Table 2 Results from the multiple logistic regression analysis that was developed to evaluate the likelihood of having a cardiovascular event (outcome) according to the metabolic syndrome presence
according to the National Cholesterol Education Program (NCEP-2005) denition (main effect).

Table 3 Results from the multiple logistic regression analysis that was developed to evaluate the likelihood of having a cardiovascular event (outcome) according to the metabolic syndrome presence
according to the International Diabetes Federation (IDF) denition (main effect).
IDF denition

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

Age (per 1 year)

Men vs. Women
History of metabolic syndrome (y/n)
Family history of CVD (y/n)
Physically active vs. sedentary
MedDietScore (per 1/55 unit)
Smokers vs. non-smokers
C-Reactive Protein (per 1 mg/L)
Interleukin-6 (per 1 pg/mL)
Uric acid (per 1 mg/dL)
eGFR (per 1 unit)

1.09 (1.08e1.10)**
2.01 (1.53e2.64)**
1.04 (0.78e1.39)
e
e
e
e
e
e
e
e

1.09 (1.07e1.10)**
1.75 (1.19e2.57)*
0.91 (0.62e1.35)
1.37 (0.93e2.03)
1.17 (0.81e1.70)
0.97 (0.94e1.00)
1.62 (1.09e2.39)*
e
e
e
e

1.09 (1.07e1.11)**
1.66 (1.11e2.49)*
0.77 (0.50e1.18)
1.26 (0.83e1.92)
1.22 (0.82e1.80)
0.97 (0.93e1.00)*
1.56 (1.03e2.34)*
1.09 (1.01e1.18)*
e
e
e

1.08 (1.06e1.10)**
1.65 (1.10e2.46)*
0.88 (0.58e1.33)
1.26 (0.84e1.90)
1.17 (0.79e1.72)
0.97 (0.93e1.00)*
1.58 (1.06e2.37)*
e
1.64 (0.94e2.87)
e
e

1.08
1.60
0.92
1.38
1.24
0.97
1.57
e
e
1.10
e

1.08 (1.06e1.10)**
2.08 (1.38e3.15)**
0.94 (0.63e1.40)
1.33 (0.89e1.99)
1.15 (0.79e1.69)
0.97 (0.94e1.01)
1.68 (1.12e2.50)*
e
e
e
0.98 (0.97e1.00)*

(1.07e1.10)**
(1.02e2.50)*
(0.61e1.39)
(0.92e2.06)
(0.86e1.82)
(0.94e1.01)
(1.05e2.34)*

(0.95e1.28)

Results are presented as OR (95%CI) obtained from logistic regression. *: p  0.05, **: p  0.01. Bold values highlight the main result (i.e., History of metabolic syndrome).
CVD: cardiovascular disease, eGFR: estimated Glomerular ltration rate.
No signicant mediation effect of the additional variables in the model that contained MetS was observed based on Sobels test.

227

1.08 (1.06e1.10)**
2.08 (1.37e3.15)**
1.01 (0.67e1.52)
1.33 (0.89e1.99)
1.16 (0.79e1.70)
0.97 (0.94e1.01)
1.67 (1.12e2.50)*
e
e
e
0.98 (0.97e1.00)*
e
e

denition, the association of MetS history was no longer

signicant.
Effect of inammatory and other parameters

Results are presented as OR (95%CI) obtained from logistic regression. *: p  0.05, **: p  0.01. Bold values highlight the main result (i.e., History of metabolic syndrome).
CVD: cardiovascular disease, eGFR: estimated Glomerular ltration rate.
No signicant mediation effect of the additional variables in the model that contained MetS was observed based on Sobels test.

Model 6

1.08 (1.07e1.10)**
1.61 (1.03e2.52)*
0.99 (0.65e1.52)
1.38 (0.93e2.06)
1.25 (0.85e1.82)
0.97 (0.94e1.01)
1.57 (1.05e2.34)*
e
e
1.10 (0.94e1.27)
e
1.08 (1.06e1.10)**
1.66 (1.11e2.47)*
0.93 (0.61e1.42)
1.26 (0.84e1.91)
1.17 (0.79e1.73)
0.97 (0.93e1.00)
1.58 (1.06e2.37)*
e
1.62 (0.93e2.83)
e
e
1.09 (1.07e1.11)**
1.67 (1.11e2.51)*
0.84 (0.54e1.30)
1.27 (0.84e1.93)
1.22 (0.83e1.82)
0.97 (0.93e1.00)
1.56 (1.03e2.34)*
1.09 (1.01e1.17)*
e
e
e

Model 3

1.09 (1.08e1.10)**
1.98 (1.51e2.61)**
1.23 (0.91e1.66)
e
e
e
e
e

Age (per 1 year)

Men vs. Women
History of metabolic syndrome (y/n)
Family history of CVD (y/n)
Physically active vs. sedentary
MedDietScore (per 1/55 unit)
Smokers vs. non-smokers
C-Reactive Protein (per 1 mg/L)
Interleukin-6 (per 1 pg/mL)
Uric acid (per 1 mg/dL)
eGFR (per 1 unit)

Model 2
Model 1
Harmonized denition

1.09 (1.07e1.10)**
1.75 (1.19e2.57)*
0.98 (0.66e1.47)
1.38 (0.93e2.04)
1.18 (0.81e1.71)
0.97 (0.94e1.01)
1.62 (1.09e2.39)*
e
e
e
e

Model 4

Model 5

C.-M. Kastorini et al.

Table 4 Results from the multiple logistic regression analysis that was developed to evaluate the likelihood of having a cardiovascular event (outcome) according to the metabolic syndrome presence
according to the Harmonized Metabolic Syndrome denition (main effect).

228

When examining the potential inuence of inammatory

and renal function markers (i.e., CRP, Il-6, uric acid and
eGFR) on the aforementioned association, none of the
added markers mediated the inuence of MetS on CVD
incidence (all ps derived from Sobel test were >0.40).
Further analyses revealed that all C-statistic values
exceeded 0.789, indicating very good predictive probability of the estimated models (Table 5).
Discussion
As it was observed, a considerable proportion of the
ATTICA study participants had the MetS, based on all
three denitions. According to the newest Harmonized
denition approximately 50% of the sample was classied
as having the MetS, similar to the prevalence estimated
by the IDF denition. Few studies have evaluated the
prevalence of MetS according to the Harmonized criteria,
in Greece. Based on the ndings of a recent study [24],
prevalence of MetS was quite similar to the ATTICA study
results, as 45.7%, 43.4% and 24.5% were classied as
having the MetS when the Harmonized, IDF or NCEP
denitions were used, respectively. Furthermore, the
NCEP denition was more predictive of CVD risk [25], as
also shown by the ndings of the present work. In
particular, according to the present ndings, MetS was
associated with 83% higher likelihood of developing CVD,
after 10 years of follow-up of the ATTICA study participants, based on the revised NCEP denition. However,
not statistically signicant ndings were observed when
the IDF or the Harmonized denitions were used. Then
again, when examining the C-statistic values very good
predictive probability was observed for all the estimated
models. This observation however could be attributed to
the fact that all models were adjusted for the most
common CVD risk factors.
Aiming to evaluate the reasons for these discrepancies
in the results when using the different MetS denitions,
the parameters and/or cut-off values that mostly affect
these ndings were examined. The main factor inuencing these differences were the cut-off values of waist
circumference, as cut-off values for the other MetS
criteria were similar. Indeed, the IDF and the Harmonized
denitions share many common characteristics, the most
important being waist circumference cut-offs. Furthermore, as already mentioned, approximately half of the
participants were classied as having MetS when these
denitions were used. It should be also noted that 30% of
the sample had waist circumference over 102 cm for men
and 88 cm for women, while almost 50% of the sample
had waist circumference over 94 cm for males and 80 cm
for females [26]. It is important to highlight that the
revised NCEP and the Harmonized denitions differ only
as far as waist circumference cut points are concerned. In

0.799 (0.761e0.837)*
0.793 (0.755e0.831)*
0.793 (0.754e0.831)*
Adjusted model: Adjusted for age, sex, family history of cardiovascular disease, physical activity, MedDietScore, smoking habits.
eGFR: estimated Glomerular ltration rate, MetS: metabolic syndrome, CI: condence interval.
C-statistic for the model including only the confounding factors (not including the metabolic syndrome): 0.789 (0.751e0.827).
*p < 0.001.

0.799 (0.760e0.837)*
0.792 (0.754e0.831)*
0.792 (0.754e0.831)*
0.807 (0.769e0.845)*
0.802 (0.764e0.840)*
0.801 (0.763e0.839)*
0.795 (0.757e0.834)*
0.789 (0.751e0.828)*
0.789 (0.751e0.827)*
NCEP ATP III
IDF
Harmonized

0.809 (0.771e0.847)*
0.804 (0.766e0.842)*
0.804 (0.765e0.842)*

Adjusted model uric acid

Adjusted model eGFR
Adjusted model Interleukin-6
Adjusted model C-Reactive protein
Adjusted model
MetS denition

Table 5 Values of the C-statistic (CI) derived from the multiple logistic regression analysis that were used to model metabolic syndrome presence, according to different denitions in relation to 10year cardiovascular events.

Metabolic syndrome and 10-year cardiovascular disease incidence

229

particular, according to the Harmonized MetS criteria it is

suggested that for subjects of European origin either the
IDF or the AHA/NHLBI cut points can be used [2]; if the
latter are used, the Harmonized criteria denition coincides with the revised NCEP denition. Results of this
work contribute to the better understanding of the
inherent pathophysiological characteristics that differentiate the metabolic syndrome denitions regarding
CVD incidence.
The choice of the ideal MetS denition is a quite a
difcult task since many factors might inuence this
decision [24,27,28]. Taking into consideration the similarities of the most recent MetS denitions, the most
important factor is the proper selection of waist
circumference cut-off values [2]. Based on the ndings of
the present work more subjects were classied as having
the MetS, using the IDF and Harmonized denitions
compared with the NCEP criteria. As expected these
subjects had better metabolic proles compared with the
ones characterized as having the MetS according to the
revised NCEP ATP III denition (data not shown). Thus on
the one hand when using the NCEP denition subjects
with MetS are classied as having higher risk of CVD, yet
when using the other denitions more people might
receive health advice and instructions to modify their
lifestyle habits and possibly pharmaceutical treatment, as
well.
As already mentioned, when evaluated with the NCEP
denition, history of MetS was associated with 83%
higher likelihood of CVD (Table 2, Models 1 & 2). These
results are in accordance with previous ndings
regarding the role of MetS on CVD risk. In more detail,
based on a recent meta-analysis, MetS dened by the
NCEP or the revised NCEP criteria, was associated with 2times higher risk of cardiovascular outcomes [6]. Similar
ndings were also observed by older meta-analyses, as
well [7,26]. These ndings are not surprising, considering
that the main MetS components; abnormal cholesterol
and blood pressure levels, insulin resistance and obesity,
are associated with or predispose to diabetes mellitus
and CVD [29].
After taking into account various confounders, MetS
dened by the revised NCEP criteria, was associated with
higher CVD risk. Nevertheless, ndings of this work did
not verify the research hypothesis that CVD surrogate
markers inuence the predictive role of MetS regarding
CVD incidence. When inammatory or other factors were
included in the models, although they tended to be
positively associated with the outcome, they did not
confer much to the correct CVD classication, as originally expected. In more detail, when the area under the
ROC curves was calculated, the higher predictive probability for CVD was observed for the models including CRP
or Il-6, followed by the ones including uric acid, despite
the MetS denition used. It is already quite known that
the MetS is associated with low grade inammation [1],
with central obesity being one of its main causes. The
hypoxia caused by the adipocytes enlargement might be
the reason for the overproduction of adipokines such as

230

Il-6 and CRP [1], promoting insulin resistance [30,31].

Furthermore, CRP has been positively associated with the
number of MetS components and is suggested to predict
future CVD [32]. However, ndings of the present work did
not suggest that this marker acted as a mediator inuencing the role of MetS on CVD incidence. Indeed the role
of CRP on CVD risk is worthy of further investigation since
the results from epidemiological and Mendelian randomization studies are controversial [33].
Strengths and limitations
The present study has several strengths since it is the rst
prospective study based on a sample quite representative
of the general Greek population, and with a long (i.e., 10year) follow-up period. Nonetheless, some limitations are
also present. The baseline evaluation was performed once,
and may be prone to measurement error. Thus, the prevalence of various clinical risk factors or the levels of biological factors may have been overestimated. However, the
applied methodology was similar to those of other prospective epidemiological studies in Europe and the US, and
therefore the results are comparable.
Conclusion
The aims of the present work were rstly to examine the
inuence of three MetS denitions and reveal which one
may better predict CVD incidence and secondly to evaluate
whether inammatory and renal markers may have an
additional explanatory value to the 10-year CVD risk and
improve the predictive ability of the MetS. Our ndings
revealed that only the revised NCEP denition was
signicantly associated with CVD development; waist
circumference cut-off values were the main factor differentiating the metabolic syndrome denitions ability to
predict CVD incidence. Furthermore, the inclusion of CRP,
Il-6, uric acid or eGFR in the models did not mediate the
inuence of MetS on CVD incidence, as initially expected.
Future studies are needed in order to better clarify and
understand the role of these markers and their interaction
with the MetS components, regarding CVD development.
Funding
Demosthenes Panagiotakos and Ekavi Georgousopoulou
received research grants by Coca-Cola SA.
Conict of interest
None to declare.
Acknowledgments
The authors would like to thank the ATTICA study group of
investigators: Yannis Skoumas, Natassa Katinioti, Labros
Papadimitriou, Constantina Masoura, Spiros Vellas, Yannis
Lentzas, Manolis Kambaxis, Konstadina Palliou, Vassiliki

C.-M. Kastorini et al.

Metaxa, Agathi Ntzouvani, Dimitris Mpougatsas, Nikolaos

Skourlis, Christina Papanikolaou, Georgia-Maria Kouli,
Aimilia Christou, Adella Zana, Maria Ntertimani, Aikaterini
Kalogeropoulou, Evangelia Pitaraki, Alexandros Laskaris,
Mihail Hatzigeorgiou and Athanasios Grekas, Eleni Kokkou
for their assistance in the initial physical examination and
follow-up evaluation, E Tsetsekou for her assistance in
psychological evaluation, as well as the laboratory team:
Carmen Vassiliadou and George Dedoussis (genetic analysis), Marina Toutouza-Giotsa, Constadina Tselika and Sia
Poulopoulou (biochemical analysis) and Maria Toutouza
for the database management.
References
[1] Kaur J. A comprehensive review on metabolic syndrome. Cardiol
Res Pract 2014;2014:943162.
[2] Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI,
Donato KA, et al. International Diabetes Federation Task Force on
Epidemiology and Prevention; Hational Heart, Lung, and Blood
Institute; American Heart Association; World Heart Federation;
International Atherosclerosis Society; International Association for
the Study of Obesity. Harmonizing the metabolic syndrome: a
joint interim statement of the International Diabetes Federation
Task Force on Epidemiology and Prevention; National Heart, Lung,
and Blood Institute; American Heart Association; World Heart
Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 2009 Oct
20;120(16):1640e5.
[3] Grundy SM. Metabolic syndrome pandemic. Arterioscler Thromb
Vasc Biol 2008 Apr;28(4):629e36.
[4] Haffner SM. The metabolic syndrome: inammation, diabetes
mellitus, and cardiovascular disease. Am J Cardiol 2006 Jan 16;
97(2A):3Ae11A.
[5] Russo I. The prothrombotic tendency in metabolic syndrome:
focus on the potential mechanisms involved in impaired haemostasis and brinolytic balance. Sci Cairo 2012;2012:525374.
[6] Mottillo S, Filion KB, Genest J, Joseph L, Pilote L, Poirier P, et al. The
metabolic syndrome and cardiovascular risk a systematic review
and meta-analysis. J Am Coll Cardiol 2010 Sep 28;56(14):1113e32.
[7] Galassi A, Reynolds K, He J. Metabolic syndrome and risk of cardiovascular disease: a meta-analysis. Am J Med 2006 Oct;119(10):812e9.
[8] Kim SY, Guevara JP, Kim KM, Choi HK, Heitjan DF, Albert DA. Hyperuricemia and coronary heart disease: a systematic review and
meta-analysis. Arthritis Care Res Hob 2010 Feb;62(2):170e80.
[9] Li M, Hou W, Zhang X, Hu L, Tang Z. Hyperuricemia and risk of
stroke: a systematic review and meta-analysis of prospective
studies. Atherosclerosis 2014 Feb;232(2):265e70.
[10] Chronic Kidney Disease Prognosis Consortium, Matsushita K, van
der Velde M, Astor BC, Woodward M, Levey AS, et al. Association of
estimated glomerular ltration rate and albuminuria with allcause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet 2010 Jun 12;
375(9731):2073e81.
[11] Chaudhary K, Malhotra K, Sowers J, Aroor A. Uric acid e key
ingredient in the recipe for cardiorenal metabolic syndrome.
Cardiorenal Med 2013 Oct;3(3):208e20.
[12] Okada R, Yasuda Y, Tsushita K, Wakai K, Hamajima N, Matsuo S.
The number of metabolic syndrome components is a good risk
indicator for both early- and late-stage kidney damage. Nutr
Metab Cardiovasc Dis 2014 Mar;24(3):277e85.
[13] Panagiotakos DB, Georgousopoulou EN, Pitsavos C, Chrysohoou C,
Metaxa V, Georgiopoulos GA, et al., ATTICA Study group. Ten-year
(2002-2012) cardiovascular disease incidence and all-cause mortality, in urban Greek population: the ATTICA Study. Int J Cardiol
2015 Feb;1(180):178e84.
[14] Grundy SM, Brewer Jr HB, Cleeman JI, Smith Jr SC, Lenfant C.
Denition of metabolic syndrome: report of the National Heart,
Lung, and Blood Institute/American Heart Association conference
on scientic issues related to denition. Circulation 2004;109:
433e8.

Metabolic syndrome and 10-year cardiovascular disease incidence

[15] The International Diabetes Federation: The IDF consensus worldwide denition of metabolic syndrome [article online]; 2005. In:
http://www.idf.org/%20webdata/docs/MetS_def_update2006.pdf.
[16] Pitsavos C, Panagiotakos DB, Chrysohoou C, Stefanadis C. Epidemiology of cardiovascular risk factors in Greece; aims, design and
baseline characteristics of the ATTICA study. BMC Public Health
2003;3(32):1e9.
[17] Katsouyanni K, Rimm EB, Gnardellis C, Trichopoulos D,
Polychronopoulos E, Trichopoulou A. Reproducibility and relative
validity of an extensive semi-quantitative food frequency questionnaire using dietary records and biochemical markers among
Greek schoolteachers. Int J Epidemiol 1997;26:S118e27.
[18] Panagiotakos DB, Pitsavos C, Stefanadis C. Dietary patterns: a
Mediterranean diet score and its relation to clinical and biological
markers of cardiovascular disease risk. Nutr Metab Cardiovasc Dis
2006;16:559e68.
[19] Craig CL, Marshall AL, Sjstrm M, Bauman AE, Booth ML,
Ainsworth BE, et al. International physical activity questionnaire:
12-country reliability and validity. Med Sci Sports Exerc 2003;
35(8):1381e95.
[20] Papathanasiou G, Georgoudis G, Papandreou M, Spyropoulos P,
Georgakopoulos D, Kalfakakou V, et al. Reliability measures of the
short international physical activity questionnaire (IPAQ) in Greek
young adults. Hell J Cardiol 2009;50(4):283e94.
[21] Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro 3rd AF,
Feldman HI, et al. A new equation to estimate glomerular ltration
rate. Ann Intern Med 2009;150:604e12.
[22] Metz CE. Basic principles of ROC analysis. Semin Nucl Med 1978;8:
283e98.
[23] Sobel Michael E. Asymptotic condence intervals for indirect effects in structural equation models. Sociol Methodol 1982;13:
290e312.
[24] Athyros VG, Ganotakis ES, Tziomalos K, Papageorgiou AA,
Anagnostis P, Griva T, et al. Comparison of four denitions of the

231

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

metabolic syndrome in a Greek (Mediterranean) population. Curr

Med Res Opin 2010 Mar;26(3):713e9.
Panagiotakos DB, Pitsavos C, Chrysohoou C, Skoumas J,
Tousoulis D, Toutouza M, et al. Impact of lifestyle habits on the
prevalence of the metabolic syndrome among Greek adults from
the ATTICA study. Am Heart J 2004 Jan;147(1):106e12.
Gami AS, Witt BJ, Howard DE, Erwin PJ, Gami LA, Somers VK, et al.
Metabolic syndrome and risk of incident cardiovascular events
and death: a systematic review and meta-analysis of longitudinal
studies. J Am Coll Cardiol 2007 Jan 30;49(4):403e14.
Strazzullo P, Barbato A, Siani A, Cappuccio FP, Versiero M,
Schiattarella P, et al. Diagnostic criteria for metabolic syndrome: a
comparative analysis in an unselected sample of adult male
population. Metabolism 2008 Mar;57(3):355e61.
Mancia G, Bombelli M, Facchetti R, Casati A, Ronchi I, QuartiTrevano F, et al. Impact of different denitions of the metabolic
syndrome on the prevalence of organ damage, cardiometabolic risk
and cardiovascular events. J Hypertens 2010 May;28(5):999e1006.
Eckel RH. Mechanisms of the components of the metabolic syndrome that predispose to diabetes and atherosclerotic CVD. Proc
Nutr Soc 2007 Feb;66(1):82e95.
Ndumele CE, Pradhan AD, Ridker PM. Interrelationships between
inammation, C-reactive protein, and insulin resistance. J Cardiometab Syndr 2006 Summer;1(3):190e6.
Ritchie SA, Connell JM. The link between abdominal obesity,
metabolic syndrome and cardiovascular disease. Nutr Metab Cardiovasc Dis 2007 May;17(4):319e26.
Emerging Risk Factors Collaboration, Kaptoge S, Di Angelantonio E,
Lowe G, Pepys MB, Thompson SG, et al. C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis. Lancet 2010 Jan 9;
375(9709):132e40.
Strang F, Schunkert H. C-reactive protein and coronary heart disease: all saideis not it? Mediat Inamm 2014;2014:757123.