Association-between-coronary-atherosclerosis-and-v-1.pdf

hbjnbkmnl

1. Association between coronary atherosclerosis and visceral adiposity
index
Zsolt Bagyura*, Loretta Kiss, Árpád Lux, Csaba Csobay-Novák, Ádám L. Jermendy,
Lívia Polgár, Zsolt Szelid, Pál Soós, Béla Merkely
Heart and Vascular Center, Semmelweis University, Városmajor utca 68, Budapest, H-1122, Hungary
Received 29 July 2019; received in revised form 29 December 2019; accepted 29 January 2020
Handling Editor: M Averna
Available online 12 February 2020
KEYWORDS
Visceral adiposity
index;
Central obesity;
Coronary
calcification;
Calcium score;
Cardiovascular risk;
Cardiology
AbstractBackground and aims:Visceral obesity is a marker of dysfunctional adipose tissue and
ectopic fat infiltration. Many studies have shown that visceral fat dysfunction has a close rela-
tionship with cardiovascular disease. For a better identification of visceral adiposity dysfunction,
the visceral adiposity index (VAI) is used. Coronary artery calcium score (CACS) is known to have
a strong correlation with the total plaque burden therefore provides information about the
severity of the coronary atherosclerosis. CACS is a strong predictor of cardiac events and it refines
cardiovascular risk assessment beyond conventional risk factors. Our aim was to evaluate the as-
sociation between VAI and CACS in an asymptomatic Caucasian population.
Methods and results:Computed tomography scans of 460 participants were analyzed in a cross-
sectional, voluntary screening program. A health questionnaire, physical examination and labo-
ratory tests were also performed. Participants with a history of cardiovascular disease were
excluded from the analysis. Mean VAI was 1.41fi0.07 in men and 2.00fi0.15 in women. VAI
showed a positive correlation with total coronary calcium score (rZ0.242) in males but not
in females. VAI was stratified into tertiles by gender. In males, third VAI tertile was independently
associated with CACS>100 (OR: 3.21, pZ0.02) but not with CACS>0 after the effects of conven-
tional risk factors were eliminated.
Conclusion:VAI tertiles were associated with calcium scores and the highest VAI tertile was an
independent predictor for the presence of CACS>100 in males but not in females.
ª2020 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, Feder-
ico II University. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction
Atherosclerosis is among the leading causes of death in the
Western world [1]. Obesity is a well-known risk factor of
cardiovascular disease (CVD) [2]. However, central or
visceral obesity appear to be more strongly associated with
cardiovascular risk [3]. Many studies have shown proin-
flammatory cytokines, adipocytokines [4] tend to increase
and insulin sensitivity decrease in patients with central
obesity [5]. Estimating the extent of visceral adiposity with
waist circumference (WC) measurement is widely used
CVD risk assessment. However, this method cannot
Acronyms:ACE, angiotensin-converting-enzyme; ACS, acute cor-
onary syndrome; ASA, acetylsalicylic acid; BMI, body mass index;
CA, Cochran-Armitage; CACS, Coronary artery calcium score; CI,
confidence interval; CT, computer tomography; CVD, cardiovas-
cular disease; DBP, diastolic blood pressure; DM, diabetes melli-
tus; HBa1c, haemoglobin-A1c; HDL-C, high-density lipoprotein
cholesterol; HU, Hounsfield unit; IQR, interquartile range; LDL-C,
low-density lipoprotein cholesterol; OR, odds ratio; VAI, visceral
adiposity index; SBP, systolic blood pressure; SD, standard
deviation.
* Corresponding author.
E-mail address:[email protected](Z. Bagyura).
https://doi.org/10.1016/j.numecd.2020.01.013
0939-4753/ª2020 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. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
Nutrition, Metabolism & Cardiovascular Diseases (2020)30, 796e803
Available online atwww.sciencedirect.com
Nutrition, Metabolism & Cardiovascular Diseases
journal homepage:www.elsevier.com/locate/nmcd

2. distinguish between subcutaneous and visceral fat accu-
mulation. Recently, the visceral adiposity index (VAI) was
developed [6] for identification of visceral adiposity
dysfunction. VAI is a simple, gender specific marker
combining anthropometric data and lipid profiles and are
reliable indicator of visceral at dysfunction. Several studies
found a strong association of VAI with cardiometabolic risk
[7]. Recently, Barazzoni et al. found that compared to WC
or BMI VAI had the highest 5-year predictive value for
metabolic syndrome in overweight or obese patients [8].
Several studies showed that VAI is a predictor of incident
hypertension [9]. Ji et al. reported that VAI has a strong
correlation with insulin resistance even in a population
without central obesity [10].
Calcification of the coronary plaques shows a strong
correlation with the total plaque burden [11]. Scanning of
the heart with low-dose non-contrast computer tomog-
raphy (CT) provides quantitative information of the extent
of the coronary calcification. Coronary artery calcium score
(CACS) or Agatston score is a reliable surrogate marker of
atherosclerosis and it is a strong predictor of cardiac
events [12], therefore could be used for assess the car-
diovascular risk beyond conventional risk factors [13]. For
ruling out obstructive coronary artery disease CAD, the
negative predictive value of a 0 CACS is almost 100%,
therefore it indicates a very low cardiovascular event rate
[14] CACS above 100 is associated with moderate risk for
coronary heart disease [15].
Only limited number of studies have examined the
relationship between VAI and CACS [16]. Therefore, our
aim was to evaluate the association between VAI and CACS
in an asymptomatic Caucasian population.
Methods
The Budakalász Health Examination Survey, a cross-
sectional voluntary cardiovascular screening program tar-
geting the adult population (>20 years, ~8000 inhabitants)
of a Central-Hungarian town (Budakalász) was performed
in 2011e2013 [17]. Medical history with special attention
to cardiovascular disease, related signs and symptoms,
lifestyle (alcohol consumption, sport activities, smoking
habits) and family history were recorded by an experi-
enced physician. Anthropometric parameters rounded to
the nearest 0.1 cm and 0.1 kg were measured in standing
position while participants were wearing light indoor
clothing without shoes.
All laboratory tests were performed in our institu-
tion's central laboratory with rigorous quality control.
Concentration of lipid fractions was measured by using a
colorimetric assay (Roche Diagnostics Ltd, Mannheim,
Germany). Hypertension, dyslipidaemia and diabetes
mellitus in medical history were regarded positive if
they have been was formerly diagnosed or the patient
had received treatment. Body mass index was calculated
by using Quetelet's form. Blood pressure measurement
was performed on the arms after 20 min rest in aflat
lying position. Blood pressure above 140 mmHg systolic
and/or 90 mmHg diastolic was defined as pathologically
high.
Prospectively ECG-triggered low dose cardiac CT scans
(Brilliance iCT, Philips Healthcare, Best, The Netherlands) for
calcium scoring were acquired of the heart with a narrow
field-of-view. Effective radiation dose was 0.5 mSv or less.
Axial images were used for the quantitative analysis of
coronary calcification using a commercially available soft-
ware application (Calcium scoring, Heartbeat-CS, Philips
Healthcare). Coronary artery plaques were identified by the
software automatically, followed by selection of real coro-
nary plaques by an expert observer manually. Based on
these results the software calculated the CACS, the calcifi-
cation area and the volume in a semi-automatic way.
The participation rate in the Budakalász Health Exami-
nation Survey was around 30% (nZ2420) of the eligible total
population. From this 2420 persons a low dose cardiac CT
scan was offered for males older than 35 years and in females
above 40 years of age (nZ2029). Total number of 511 par-
ticipants volunteered for CT scan. This group was not signif-
icantly different from those who were offered the CT in
gender, LDL and triglyceride levels, blood pressure, BMI and
VAI but the CT subgroup was significantly older (62.12þ-
10.03 vs. 58.17þ-11.91, p<0.001), had slightly higher total
cholesterol levels (5.57þ1.17 vs 5.56þ-1.13, pZ0.04).
Moreover hypertension (27.8% vs. 21.9%, pZ0.002), dysli-
pidaemia (30.3% vs. 21.7%, p<0.001) and diabetes mellitus
(32.3% vs. 24%, pZ0.004) were more frequent in this group.
Participants with the following cardiovascular history were
excluded from the further analysis: 24 patients with previous
myocardial infarction (4.7%), 20 patients with stroke (3.9%), 4
patients with transient ischemic attack (0.8%). Also, 3 persons
(0.6%) were excluded due to the lack of the laboratory results.
As a participant could have more than one exclusion criteria,
overall four hundred sixty participants were included in the
study. The studyflowchart is presented inFig. 1. The VAI was
calculated using the following sex-specific formula [18]:
flMales: VAI Z [WC/{39.68þ(1.88*BMI)}])(TG/
1.03))(1.31/HDL-C)
flFemales: VAI Z[WC/{36.58þ(1.89)BMI)}])(TG/
0.81))(1.52/HDL-C)
A post hoc power analysis was conducted using online
calculator developed by MGH Biostatistics Center. The
sample size of 460 was used for the statistical power an-
alyses. The significance level used for this analysis was
p<0.05. The standard deviation of the dependent variable
(CACS) was 500.6, the standard deviation of the indepen-
dent variable (VAI) was 0.32, the minimal detectable dif-
ference entered was 244.8. The post hoc analyses revealed
that the probability is 91 percent that the study detects a
relationship between the independent and the dependent
variables.
Statistical methods
SPSS for Windows, version 18.0 (IBM, Armonk, NY) was
used for statistical analysis. All continuous variables were
Calcium score and visceral adiposity 797

3. expressed as mean with standard deviation (SD) or as
medians with interquartile range as appropriate depend-
ing on the distribution of the values, whereas categorical
variables were expressed as percentage. Comparisons of
means, medians and proportions were performed with
variance analysis, Kruskal eWallis test,
JonckheereeTerpstra test, Cochran-Armitage (CA) test and
Chi-square tests, respectively. Spearman's correlation was
used to test the association between VAI and CACS.
Multivariate regression analysis was performed adjusted
for age, gender and risk factors with the 1st VAI tertile as
reference category. All analyses were performed two-
tailed and p<0.05 was considered significant.
Ethics approval and consent to participate
The research had ethics approval from the Medical
Research Council Scientific and Ethics Committee
(permission number: 8224e0/2011/EKU (265/PI/11)). All
procedures followed were in accordance with the ethical
standards of the responsible committee on human exper-
imentation (institutional and national) and with the Hel-
sinki Declaration of 1975, as revised in 2000 (5). Written
informed consent was obtained from all patients for being
included in the study.
Results
The basic characteristics of the male and female
participants
The mean age was 61.6 (fi10.2) years; 41.1% of the par-
ticipants were men and all patients belonged to the
Caucasian race. The mean BMI was 28.6 kg/m
2
(fi5.1), the
mean VAI was 1.76 (fi0.32) and the median CACS score
was 19.65 (fi161.2). Clinical baseline characteristics of the
189 male and 271 female patients are described inTable 1.
Waist circumference, smoking status, mean systolic blood
pressure, HDL-C, total cholesterol and serum creatinine
were significantly different between genders. Mean VAI
was 1.41fi0.07 in men and 2.00fi0.15 in women
(p<0.001), butt BMI was not significantly different be-
tween the two genders. Drug therapies could have effect
on atherosclerotic plaques, but we have found no differ-
ence in the lipid lowering angiotensin-converting
enzyme inhibitor and acetylsalicylic acid drug usage be-
tween the two genders.
VAI differs substantially between males and females,
therefore VAI was stratified into tertiles by gender (Table
2.) for further analysis. Waist circumference was signifi-
cantly different in the three tertile groups in both gen-
ders. Age, BMI, low-density lipoprotein (LDL-C), total
cholesterol and haemoglobin-A1c (HBa1c) levels were
significantly different in the three tertile groups in males
but not in females. Also, the frequency of hypertension
and diabetes increased gradually in males in the groups
and was highest in the third tertile group. In contrast,
there were no significant differences in the age, BMI or
the other above-mentioned factors’distribution among
the three groups in females. Menopause can affect
adiposity; therefore, we analysed the menstruation sta-
tus in females. As there was no difference in this
parameter across the three tertiles (tertile 1: 64 (71.1%),
tertile 2: 63 (69.2%) and tertile 3: 63 (70%), pZ0.589)
the effect of menopause was not further analysed. Clin-
ical characteristics in the VAI tertiles by gender are
described inTable 3.
Targeted popula?on (adult
inhabitants of Budakalász
n~8000)
Total study popula?on
n=2420
volunteering
Par?cipants of the CT scan
(Ca-scoring) n=511
Not par?cipa?ng in CT substudy •Not mee?ng the incluson citeria (n= 388)
•Declined to par?cipate (n= 1521)
Involved in analysis n=460
Excluded for analysis
•previous myocardial infarc?on (n=24, 4.7%)
•stroke (n=20, 3.9%)
•transient ischemic a?ack (n=4, 0.8%).
•missing data (n=3, 0.6%)
Figure 1Studyflowchart.
798 Z. Bagyura et al.

4. CT scan results
Median total calcium score was 19.4 (IQR: 161.6) in the
study group. VAI only showed a positive correlation with
total coronary calcium score in males (r Z0.242,
p<0.001). The JonckheereeTerpstra test for total CACS
confirmed a trend across VAI groups in males (J-T statistic:
7419, p<0.001) but not in females. Total CACS was 0 in 41
cases in males (21.7%) and in 108 cases (39.9%) in females.
It was less or equal to 100 in 112 cases (59.3%) in males and
209 cases (77.1%) in females. There was no statistically
significant difference in the frequency of CACS>0in
neither gender. Frequency of CACS>100 was significantly
different in the three VAI groups in males but not in fe-
males (Table 3) and there was a trend across VAI groups
(CA statistic: 14.476, p<0.001). The highest tertile group
had significantly higher odds ratio for CACS>100 only in
males (OR 5.37, 95% CI: 2.43e11.83, p<0.01) compared to
the lowest tertile (Table 4).
Multivariate analysis
We performed multivariate analysis with presence of
CACS>100 adjusted for age. In this model the highest
tertile group had significantly higher odds ratio for CACS
>100 in males (Model 1: OR 3.75, 95% CI:1.55e9.08,
pZ0.003) compared to the lowest tertile, but not in fe-
males (Table 4). Similarly, after further adjustments for
BMI, hypertension, dyslipidaemia, smoking status, total
cholesterol and uric acid level, diabetes mellitus and
HbA1c levels in model 2 the highest tertile group was
independently associated with CACS>100 in males (Model
2: OR 3.41, 95% CI: 1.27e9.41, pZ0.018) compared to the
lowest tertile. BMI was not independently associated with
CACS>100 in either gender.
Discussion
In this study we presented an independent association of
VAI with coronary calcification in an asymptomatic
Caucasian population. We found that VAI showed a posi-
tive correlation with CACS (rZ0.242). In logistic regres-
sion analyses, compared to the lowest VAI tertile, the
highest tertile was an independent predictor and showed
significantly increased odds for the presence of moderate-
risk coronary calcification (CACS>100) in males but not in
Table 1Clinical baseline characteristics.
Males (NZ189) Females (N Z271) p
Age, mean (SD) 60.55 (11.16) 62.11 (9.41) 0.107
Waist Circumference (cm), mean (SD) 103.66 (11.66) 96.54 (12.81) <0.001
BMI, mean (SD) 28.61 (4.31) 28.56 (5.53) 0.924
Hypertension, n (%) 153 (81.0%) 221 (81.5%) 0.873
Dyslipidaemia, n (%) 79 (41.8%) 132 (48.7%) 0.143
Statin therapy, n (%) 35 (18.5%) 54 (19.9%) 0.721
Other lipid lowering therapy, n (%) 6 (3.2%) 3 (1.1%) 0.115
ACE inhibitor therapy, n (%) 54 (28.6%) 77 (28.4%) 0.971
ASA therapy, n (%) 27 (14.3%) 31 (11.4%) 0.366
DM, n (%) 30 (15.9%) 33 (12.2%) 0.257
Active smoker, n (%) 24 (12.7%) 28 (10.3%) 0.430
Smoker (currentþformer), n (%) 92 (48.7%) 91 (33.6%) 0.010
SBP (mmHg), mean (SD) 135.05 (16.16) 138.52 (19.99) 0.040
DBP (mmHg), mean (SD) 80.50 (8.82) 80.26 (9.85) 0.789
LDL-C (mmol/l), mean (SD) 3.45 (0.97) 3.53 (1.05) 0.400
HDL-C (mmol/l), mean (SD) 1.33 (0.26) 1.65 (0.49) <0.001
Triglyceride (mmol/l), mean (SD) 2.47 (1.50) 2.31 (1.52) 0.265
Total cholesterol (mmol/l), mean (SD) 5.54 (1.11) 5.86 (1.15) 0.003
Serum creatinine (mmol/l), mean (SD) 85.77 (15.71) 80.87 (15.11) 0.010
Serum uric acid (mmol/l), mean (SD) 350.47 (79.11) 299.66 (68.98) <0.001
HbA1c (%), mean (SD) 5.80 (0.74) 5.91 (0.75) 0.112
VAI, mean (SD) 1.41 (0.07) 2.00 (0.15) <0.001
LogCACS score, mean (SD) 1.57 (1.13) 1.03 (1.03) <0.001
CACS, median (IQR) 55.7 (349) 6.66 (86.8) <0.001
CACS>0, n (%) 148 (78.3%) 163 (60.1%) <0.001
CACS>100, n (%) 77 (40.7%) 62 (22.9%) <0.001
Agatston score>400, n (%) 42 (22.2%) 21 (7.7%) <0.001
MeansfiSDs (or medianfiIQR in case of CACS) represented the continuous variables, and proportions the categorical variables. CACSecoronary artery calcium score, DBP-diastolic blood
pressure, DM-diabetes mellitus, HDLC-high-density lipoprotein cholesterol, IQReinterquartile range, LDLC-low-density lipoprotein cholesterol, SBP-systolic blood pressure, SDestandard
deviation
.
Table 2Tertiles of VAI by gender in the study population.
Tertile Males (min emax, N) Females (min emax, N)
1 1.03 e1.37, 62 1.35 e1.94, 90
2 1.38 e1.44, 64 1.95 e2.05, 91
3 1.45 e1.59, 63 2.06 e2.36, 90
Calcium score and visceral adiposity 799

5. females. The association remained significant even after
adjusting for confounding variables including BMI. We
found no statistically significant association of VAI with
the presence of any coronary calcification (CACS>0).
Visceral obesity is a marker of dysfunctional adipose
tissue and a well-known risk factor of CVD [2] and the
most prevalent manifestation of metabolic syndrome
[19e22]. Waist circumference measurement [23] is known
Table 3Clinical baseline characteristics in the VAI tertiles by gender.
Males Females
VAI tertile 1
(NZ62)
VAI tertile 2
(NZ64)
VAI tertile 3
(NZ63)
p VAI tertile] 1
(NZ90)
VAI tertile 2
(NZ91)
VAI tertile 3
(NZ90)
p
Age, mean (SD) 55.92 (11.55) 62.4 (10.95) 62.98 (9.98) <0.001 60.58 (9.54) 61.91 (9.48) 63.84 (9.01) 0.064
WC (cm), mean (SD) 94.1 (7.94) 104.2 (7.15) 111.67 (11.69) <0.001 89.27 (12.55) 98.30 (11.55) 102.05 (10.87)<0.001
BMI, mean (SD) 26.89 (3.32) 28.97 (3.48) 29.84 (5.25) <0.001 28.73 (6.01) 29.34 (5.68) 27.62 (4.75) 0.104
Hypertension, n (%) 41 (66.1%) 53 (82.8%) 58 (92.1%) 0.001 76 (84.4%) 75 (82.4%) 70 (77.8%) 0.497
Dyslipidaemia, n (%) 21 (33.9%) 30 (46.9%) 27 (42.9%) 0.317 40 (44.4%) 45 (49.5%) 48 (52.2%) 0.571
DM, n (%) 2 (3.2%) 10 (15.6%) 17 (27.0%) 0.001 11 (12.2%) 12 (13.2%) 10 (11.1%) 0.913
Active smoker, n (%) 11 (17.7%) 6 (9.4%) 7 (11.1%) 0.332 13 (14.4%) 6 (6.6%) 9 (10.0%) 0.220
Smoker (currentþformer),
n (%)
22 (35.5%) 35 (54.7%) 35 (56.6%) 0.04 28 (31.1%) 36 (39.6%) 27 (33.6%) 0.329
SBP (mmHg), mean (SD) 133.3 (16.42) 135.32 (17.34) 137.8 (14.53) 0.601 139.84 (22.02) 138.15 (19.10) 137.58 (18.88) 0.733
DBP (mmHg), mean (SD) 80.42 (8.71) 80.43 (9.72) 81.87 (8.16) 0.728 81.08 (8.58) 81.01 (11.55) 78.7 (9.03) 0.183
LDL-C (mmol/l), mean (SD) 3.79 (1.11) 3.37 (0.79) 3.25 (0.92) 0.005 3.60 (1.13) 3.49 (1.01) 3.51 (1.02) 0.772
HDL-C (mmol/l), mean (SD) 1.36 (0.31) 1.28 (0.26) 1.37 (0.46) 0.263 1.73 (0.56) 1.60 (0.42) 1.63 (0.49) 0.219
Triglyceride (mmol/l),
mean (SD)
2.37 (1.47) 2.32 (1.36) 2.69 (1.66) 0.311 2.16 (1.23) 2.26 (134) 2.50 (1.91) 0.321
Total cholesterol (mmol/l),
mean (SD)
5.87 (1.32) 5.40 (0.9) 5.44 (0.98) 0.019 5.95 (1.19) 5.76 (1.21) 5.86 (1.06) 0.564
Statin therapy, n (%) 5 (8.1) 17 (26.6) 13 (20.6) 0.178 19 (21.1) 20 (22.0) 15 (16.7) 0.672
ACE inhibitor therapy, n (%) 12 (19.4) 22 (34.4) 20 (31) 0.817 30 (33.3) 25 (27.5) 22 (24.4) 0.405
Other lipid lowering
therapy, n (%)
2 (3.2) 2 (3.2) 2 (3.2) 0.999 0 (0) 1 (1.1) 2 (2.2) 0.362
ASA therapy, n (%) 5 (8.1) 10 (14.5) 12 (20.7) (0.142) 9 (10) 11 (12.1) 11 (12.2) 0.871
Serum creatinine (mmol/l),
mean (SD)
84.53 (15.7) 87.1 (14.21) 58.2 (15.34) 0.322 69.69 (14.60) 72.69 (15.31) 70.32 (15.40) 0.353
Serum uric acid (mmol/l),
mean (SD)
344.18 (80.75) 349.38 (85.83) 357.67 (70.66) 0.634 300.34 (70.49) 297.20 (68.98) 301.48 (68.17) 0.911
HbA1c (%), mean (SD) 5.53 (0.37) 5.81 (0.59) 5.97 (0.72) <0.001 5.84 (0.74) 5.95 (0.72) 5.96 (0.81) 0.500
VAI, mean (SD) 1.33 (0.05) 1.41 (0.02) 1.48 (0.03) <0.001 1.84 (0.10) 2.00 (0.32) 2.16 (0.76) <0.001
LogCACS score, mean (SD) 1.15 (1.05) 1.68 (1.10) 1.85 (1.13) 0.001 1.05 (1.05) 1.02 (0.97) 1.02 (1.01) 0.970
CACS, median (IQR) 14.3 (62.4) 56.82 (342.1) 147.54 (530.1) 0.001 11.57 (79.9) 6.42 (89.7) 5.46 (105.5) 0.963
CACS>0, n (%) 44 (71.0%) 52 (81.3%) 52 (82.5%) 0.228 52 (57.8%) 59 (64.8%) 52 (57.8%) 0.534
CACS>100, n (%) 13 (21.0%) 26 (40.6%) 37 (58.7%) <0.001 19 (21.1%) 20 (22.0%) 23 (25.6%) 0.753
Agatston score>400, n (%) 9 (14.5%) 14 (21.9%) 18 (28.6%) 0.162 8 (8.9%) 3 (3.3%) 10 (11.1%) 0.128
MeansfiSDs (or medianfiIQR in case of CACS) represented the continuous variables, and proportions the categorical variables. Continuous
variables were analysed via One-way ANOVA or KruskaleWallis test in case of CACS, categorical variables were analysed via the Chi-square test.
ACEeangiotensin-converting-enzyme, ASAeacetylsalicylic acid, CACSecoronary artery calcium score, DBP - diastolic blood pressure, DM -
diabetes mellitus, HDL-C - high-density lipoprotein cholesterol, IQReinterquartile range LDL-C - low-density lipoprotein cholesterol, SBP -
systolic blood pressure, SD - standard deviation.
Table 4Odds ratios for coronary artery calcification (CACS>100) by gender according to tertiles of VAI.
VAI tertile 1 VAI tertile 2 VAI tertile 3
Unadjusted Female OR (CI) 1 (reference) 1.05 (0.52 e2.14) 1.28 (0.64 e2.57)
p 0.754 0.887 0.481
Male OR (CI) 1 (reference) 2.58 (1.17 e5.68) 5.37 (2.43 e11.83)
p <0.01 0.02 <0.01
Model 1 Female OR (CI) 1 (reference) 0.915 (0.43 e1.94) 0.936 (0.44 e1.99)
p 0.972 0.82 0.86
Male OR (CI) 1 (reference) 1.59 (0.65 e3.86) 3.75 (1.55 e9.08)
p 0.09 0.308 0.003
Model 2 Female OR (CI) 1 (reference) 0.833 (0.37 e1.88) 0.95 (0.42 e2.12)
p 0.903 0.66 0.894
Male OR (CI) 1 (reference) 1.27 (0.47 e3.42) 3.41 (1.27 e9.41)
p 0.03 0.642 0.018
Model 1: adjusted for age. Model 2: adjusted for age, hypertension, dyslipidaemia, smoking status, total cholesterol, uric acid, diabetes mellitus
and HbA1c. CACS: coronary artery calcium score, OR: odds ratio, CI: 95% confidence interval; VAI: visceral adiposity index.
800 Z. Bagyura et al.

6. to be more informative in connection with visceral
adiposity than BMI, although it cannot distinguish well
between subcutaneous fat and visceral obesity. To over-
come this VAI was developed for more precise assessment
of cardiometabolic risk as it is based on sex, WC, BMI and
lipid parameters [6]. VAI is known to be increased signif-
icantly in metabolically healthy obese individuals
compared to metabolically healthy normal-weight in-
dividuals and is a novel risk factor of CVD [7].
There is no ideal cut-off value for VAI that helps to
distinguish normal and dysfunctional visceral adiposity.
Some researchers use tertiles [24], others use quartiles as a
cut-off value for the analysis [25]. We used tertiles and
found that the third tertile of VAI is correlated positively
with CACS in both regression models.
Our results showed that there is an association between
VAI and the prevalence of classic markers of car-
diometabolic risk (BMI, hypertension, as well as diabetes
mellitus and HBa1c) in males, but not in females. We
found no significant differences in the VAI terciles
regarding dyslipidaemia, triglyceride LDL-C and HDL-C
levels. These results are partially in line with thefindings
of Amaro et al. [6] who demonstrated that VAI has strong
correlation with metabolic markers independently of
genders.
According to our results the three tertiles in males are
more heterogeneous in the frequency of having cardio-
vascular risk factors, and higher tertile means higher fre-
quency of cardiovascular risk factors. In contrast there is
no significant difference in females among the tertiles. In
our study the range of VAI differs basically in the two
genders. In males in thefirst tertile group the VAI is be-
tween 1.03 and 1.37, but in females thefirst tertile is be-
tween 1.35 and 1.94, so the average VAI is significantly
higher even in thefirst tertile. Moreover, the maximal
value in thefirst tertile of females is higher than the
overall maximal VAI value in males. Distribution of VAI is
different between the two genders in our population
mainly because females have relatively worse waist
circumference. In our study group according to our results
the metabolically active central obesity affects almost all
females (high average VAI) in contrast to males were there
is a gradual increase from“normal”VAI to the worse
values.
In our study we included males over 35 years and fe-
males over 40 years only. If we compare the average age in
the tertile groups in males age is increasing significantly in
contrast to females where thefirst tertiles is almostfive
years older in average and there is no significant difference
among the tertile groups. Moreover, the menopause of
women could affect the results, but we found no difference
in this parameter across the three tertiles in females.
Males with higher CACS had a significantly higher VAI
compared to those without CACS in our study. The risk of
CACS>100 was significantly higher in the upper VAI ter-
tiles compared to the lowest tertile (OR 3.41, 95%
CI:1.4e8.31, pZ0.007), even after adjusting for several
factors (OR 3.21, 95% CI:1.16e8.85, pZ0.024). In our study
BMI was not independently associated with CACS in
contrast to VAI. The association between VAI and CACS but
not BMI could be explained by the metabolically active
central fat accumulation, that results in increased circu-
lating free fatty acid levels, systemic inflammation, insulin
sensitivity and vasculopathy [26,27].
In contrast to our results Park et al. [16] have found
association between VAI and CACS>100 in both genders.
Moreover, we did not
find association between CACS>0
and VAI, in contrast to Parks' group, who found that the
risk of CACS>0 was higher in the upper VAI tertiles
compared to the lowest tertile in both genders. However,
there are important differences between our and Park's
study. Park et al. investigated more than 33,000 patients
from the Korean population, 80.2% of total participants
were men. This gender imbalance could have affected the
results as Park et al., who created three tertile groups
based only on VAI but not gender. In contrast, in our study
41.1% of the participants were men and all participants
were from the Caucasian race. It is noteworthy that VAI
was modelled on a Caucasian population, and this is an
important limitation to consider when it is applied in non-
Caucasian populations. According to our results, the dis-
tribution of VAI was significantly different between gen-
ders, therefore we created gender-based tertiles, and this
could explain the difference that we found between males
and females. It is probable that the association between
VAI and CACS is gender-dependent and different in the
Central European and Korean population. Our result are in
line with the results of Randrianarisoa et al. [28] who
found significant association between VAI and carotid in-
tima media thickness, a marker of subclinical atheroscle-
rosis, along with age, smoking and male sex and with the
results of Kouli et al. [29] who found that VAI is inde-
pendently associated with elevated 10-year CVD risk,
particularly in men. Finally, in Park's study, the mean
participant age was 41.5 years, and the mean BMI was
24.3 kg/m
2
. Our study group was more than 20 years older
and more obese in average. It is possible that these are the
main reasons behind that the mean CACS was significantly
higher (11.2fi72.0. vs. 211.8fi540.1). According to their
results, they have found 1.8 to 1.2 OR for CACS>100 in the
upper VAI tertiles compared to the lowest tertile. In
contrast, we have found an approximately two times
higher risk, but only in males.
This study has some limitations. Firstly, as cross-
sectional study we cannot imply causality between VAI
and CACS based on these results. Secondly, we indicated
low-dose CT only for men above 35 and women above 40,
therefore patients with lowest cardiovascular risk were
excluded from our study. Third, there is no appropriate VAI
cut-off value to estimate cardiovascular risk; we used
tertile values and compared to CACS or to the proportion
of subjects with CACS>100.
VAI is a simple, gender specific indicator of visceral fat
dysfunction and known to be associated with increased
risk of CVD. CACS is a reliable surrogate marker of
atherosclerosis, however it is not widely used in primary
care settings. In conclusion, in an asymptomatic popula-
tion VAI tertiles were associated with CACS and the third
Calcium score and visceral adiposity 801

7. VAI tertile was an independent predictor for the presence
of moderate-risk coronary calcification in males but not in
females. Our results suggest that VAI could be a useful
clinical marker of cardiometabolic risk and effect of central
obesity as a risk factor in more pronounced in males and
could have ethnic difference. The association between VAI
and CACS may have significant implications for identifying
patients in risk for atherosclerotic coronary artery disease
in primary prevention.
Funding
This work was supported by the National Research,
Development and Innovation Office of Hungary (NKFIA;
NVKP-16-1-2016-0017). The research wasfinanced by the
Higher Education Institutional Excellence Programme of
the Ministry of Human Capacities in Hungary, within the
framework of the Therapeutic Development thematic
programme of the Semmelweis University.
Authors' contributions
Zsolt Bagyura, Loretta Kiss carried out the statistical
analysis and drafted the manuscript. Zsolt Bagyura, Loretta
Kiss, Pál Soós, Zsolt Szelid and Béla Merkely participated in
the design of the study Zsolt Bagyura, Loretta Kiss, Lívia
Polgár, Árpád Lux, Pál Soós, Zsolt Szelid performed the
data collection from the patients. Csaba Csobay-Novák,
Ádám L. Jermendy, Árpád Lux performed the calcium
scoring. All authors read and approved the final
manuscript.
Declaration of Competing Interest
The authors declare that they have no conflict interests.
References
[1] Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R, et al.
Heart disease and stroke statistics-2017 update: a report from the
American heart association. Circulation 2017;135(10):e146e603.
https://doi.org/10.1161/cir.0000000000000485.
[2] Hubert HB, Feinleib M, McNamara PM, Castelli WP. Obesity as an
independent risk factor for cardiovascular disease: a 26-year
follow-up of participants in the Framingham Heart Study. Circula-
tion 1983;67(5):968e77.https://doi.org/10.1161/01.CIR.67.5.968.
[3] Despres JP. Body fat distribution and risk of cardiovascular disease:
an update. Circulation 2012;126(10):1301e13.https://doi.org/10.
1161/circulationaha.111.067264.
[4] Oh JY. Regional adiposity, adipokines, and insulin resistance in type
2 diabetes. Diabetes Metabol J 2012;36(6):412 e4.https:
//doi.org/10.4093/dmj.2012.36.6.412.
[5]Lebovitz HE, Banerji MA. Point: visceral adiposity is causally related
to insulin resistance. Diabetes Care 2005;28(9):2322e5.
[6] Amato MC, Giordano C, Galia M, Criscimanna A, Vitabile S, Midiri M,
et al. Visceral Adiposity Index: a reliable indicator of visceral fat
function associated with cardiometabolic risk. Diabetes Care 2010;
33(4):920e2.https://doi.org/10.2337/dc09-1825.
[7] Kang YM, Jung CH, Cho YK, Jang JE, Hwang JY, Kim EH, et al. Visceral
adiposity index predicts the conversion of metabolically healthy
obesity to an unhealthy phenotype. PloS One 2017;12(6):e0179635.
https://doi.org/10.1371/journal.pone.0179635.
[8] Barazzoni R, Gortan Cappellari G, Semolic A, Ius M, Zanetti M,
Gabrielli A, et al. Central adiposity markers, plasma lipid profile and
cardiometabolic risk prediction in overweight-obese individuals. Clin
Nutr 2019;38(3):1171e9.https://doi.org/10.1016/j.clnu.2018.04.014.
[9] Zhang Z, Shi D, Zhang Q, Wang S, Liu K, Meng Q, et al. Visceral
adiposity index (VAI), a powerful predictor of incident hypertension
in prehypertensives. Intern Emerg Med 2018;13(4):509e16.https:
//doi.org/10.1007/s11739-018-1836-8.
[10] Ji B, Qu H, Wang H, Wei H, Deng H. Association between the visceral
adiposity index and homeostatic model assessment of insulin resis-
tance in participants with normal waist circumference. Angiology
2017;68(8):716e21.https://doi.org/10.1177/0003319716682120.
[11]Sangiorgi G, Rumberger JA, Severson A, Edwards WD, Gregoire J,
Fitzpatrick LA, et al. Arterial calcification and not lumen stenosis is
highly correlated with atherosclerotic plaque burden in humans: a
histologic study of 723 coronary artery segments using non-
decalcifying methodology. J Am Coll Cardiol 1998;31(1):126e33.
[12] Detrano R, Guerci AD, Carr JJ, Bild DE, Burke G, Folsom AR, et al.
Coronary calcium as a predictor of coronary events in four racial or
ethnic groups. N Engl J Med 2008;358(13):1336e45.https:
//doi.org/10.1056/NEJMoa072100.
[13] Alluri K, Joshi PH, Henry TS, Blumenthal RS, Nasir K, Blaha MJ.
Scoring of coronary artery calcium scans: history, assumptions,
current limitations, and future directions. Atherosclerosis 2015;
239(1):109e17. https:
//doi.org/10.1016/j.atherosclerosis.2014.12.040.
[14]Haberl R, Becker A, Leber A, Knez A, Becker C, Lang C, et al. Cor-
relation of coronary calcification and angiographically docu-
mented stenoses in patients with suspected coronary artery
disease: results of 1,764 patients. J Am Coll Cardiol 2001;37(2):
451e7.
[15] Okwuosa TM, Greenland P, Ning H, Liu K, Bild DE, Burke GL, et al.
Distribution of coronary artery calcium scores by Framingham 10-
year risk strata in the MESA (Multi-Ethnic Study of Atherosclerosis)
potential implications for coronary risk assessment. J Am Coll Cardiol
2011;57(18):1838e45.https://doi.org/10.1016/j.jacc.2010.11.053.
[16] Park HJ, Kim J, Park SE, Park CY, Lee WY, Oh KW, et al. Increased
risk of subclinical atherosclerosis associated with high visceral
adiposity index in apparently healthy Korean adults: the Kangbuk
Samsung Health Study. Ann Med 2016;48(6):410 e6.https:
//doi.org/10.1080/07853890.2016.1183258.
[17] Kiss LZ, Bagyura Z, Csobay-Novák C, Lux Á, Polgár L, Jermendy Á,
et al. Serum uric acid is independently associated with coronary
calcification in an asymptomatic population. J Cardiovasc Transl
Res 2018.https://doi.org/10.1007/s12265-018-9843-8.
[18] Salazar MR, Carbajal HA, Espeche WG, Aizpurua M, Maciel PM,
Reaven GM. Identification of cardiometabolic risk: visceral adiposity
index versus triglyceride/HDL cholesterol ratio. Am J Med 2014;
127(2):152e7.https://doi.org/10.1016/j.amjmed.2013.10.012.
[19] Despres JP, Lemieux I. Abdominal obesity and metabolic syndrome.
Nature 2006;444(7121):881e7.https://doi.org/10.1038/nature05488.
[20]DeNino WF, Tchernof A, Dionne IJ, Toth MJ, Ades PA, Sites CK, et al.
Contribution of abdominal adiposity to age-related differences in
insulin sensitivity and plasma lipids in healthy nonobese women.
Diabetes Care 2001;24(5):925e32.
[21] Despres JP. Is visceral obesity the cause of the metabolic syn-
drome? Ann Med 2006;38(1):52 e63.https://doi.org/10.
1080/07853890500383895.
[22]Bjorntorp P. Metabolic implications of body fat distribution. Dia-
betes Care 1991;14(12):1132e43.
[23] Klein S, Allison DB, Heymsfield SB, Kelley DE, Leibel RL, Nonas C,
et al. Waist circumference and cardiometabolic risk: a consensus
statement from shaping America's health: association for weight
management and obesity prevention; NAASO, the obesity society;
the American society for nutrition; and the American diabetes
association. Am J Clin Nutr 2007;85(5):1197 e202.https:
//doi.org/10.1093/ajcn/85.5.1197.
[24] Amato MC, Giordano C, Pitrone M, Galluzzo A. Cut-off points of the
visceral adiposity index (VAI) identifying a visceral adipose
dysfunction associated with cardiometabolic risk in a Caucasian
Sicilian population. Lipids Health Dis 2011;10:183.https:
//doi.org/10.1186/1476-511x-10-183.
[25] Ding Y, Gu D, Zhang Y, Han W, Liu H, Qu Q. Significantly increased
visceral adiposity index in prehypertension. PloS One 2015;10(4):
e0123414.https://doi.org/10.1371/journal.pone.0123414.
802 Z. Bagyura et al.

8. [26] Bays HE. "Sick fat," metabolic disease, and atherosclerosis. Am J
Med 2009;122(1 Suppl):S26 e37.https://doi.org/10.1016/
j.amjmed.2008.10.015.
[27] Steinberg HO, Tarshoby M, Monestel R, Hook G, Cronin J,
Johnson A, et al. Elevated circulating free fatty acid levels impair
endothelium-dependent vasodilation. J Clin Invest 1997;100(5):
1230e9.https://doi.org/10.1172/jci119636.
[28] Randrianarisoa E, Lehn-Stefan A, Hieronimus A, Rietig R,
Fritsche A, Machann J, et al. Visceral adiposity index as an
independent marker of subclinical atherosclerosis in individuals
prone to diabetes mellitus. J Atherosclerosis Thromb 2019.https:
//doi.org/10.5551/jat.47274.
[29] Kouli GM, Panagiotakos DB, Kyrou I, Georgousopoulou EN,
Chrysohoou C, Tsigos C, et al. Visceral adiposity index and 10-year
cardiovascular disease incidence: the ATTICA study. Nutr Metabol
Cardiovasc Dis : Nutr Metabol Cardiovasc Dis 2017;27(10):881e9.
https://doi.org/10.1016/j.numecd.2017.06.015.
Calcium score and visceral adiposity 803