Frequency of early vascular aging and associated risk factors among an adult population in Latin America: the OPTIMO study

Abstract

The main objective was to estimate the frequency of early vascular aging (EVA) in a sample of subjects from Latin America, with emphasis in young adults. We included 1416 subjects from 12 countries in Latin America who provided information about lifestyle, cardiovascular risk factors (CVRF), and anthropometrics. We measured pulse wave velocity (PWV) as a marker of arterial stiffness, and blood pressure (BP) using an oscillometric device (Mobil-O-Graph). To determine the frequency of EVA, we used multiple linear regression to estimate each subject’s PWV expected for his/her age and systolic BP, and compared with observed values to obtain standardized residuals (z-scores). We defined EVA when z-score was ≥1.96. Finally, a multivariable logistic regression analysis was performed to determine baseline characteristics associated with EVA. Mean age was 49.9 ± 15.5 years, male gender was 50.3%. Mean PWV was 7.52 m/s (SD 1.97), mean systolic BP was 125.3 mmHg (SD 16.7) and mean diastolic BP was 78.9 mmHg (SD 12.2). The frequency of EVA was 5.7% in the total population, 9.8% in adults of 40 years or less and 18.7% in those 30 years or less. In these young adults, multiple logistic regression analyses demonstrated that dyslipidemia and hypertension showed an independent association with EVA, and smoking a borderline association (p = 0.07). In conclusion, the frequency of EVA in a sample from Latin America was around 6%, with higher rates in young adults. These results would support the search of CVRF and EVA during early adulthood.

Full text

Journal of Human Hypertension
https://doi.org/10.1038/s41371-018-0038-1
ARTICLE
Frequency of early vascular aging and associated risk factors among
an adult population in Latin America: the OPTIMO study
Fernando Botto1●Sebastian Obregon1●Fernando Rubinstein2●Angelo Scuteri3●Peter M. Nilsson4●Carol Kotliar1
Received: 24 September 2017 / Revised: 20 December 2017 / Accepted: 17 January 2018
© Macmillan Publishers Ltd., part of Springer Nature 2018
Abstract
The main objective was to estimate the frequency of early vascular aging (EVA) in a sample of subjects from Latin America,
with emphasis in young adults. We included 1416 subjects from 12 countries in Latin America who provided information
about lifestyle, cardiovascular risk factors (CVRF), and anthropometrics. We measured pulse wave velocity (PWV) as a
marker of arterial stiffness, and blood pressure (BP) using an oscillometric device (Mobil-O-Graph). To determine the
frequency of EVA, we used multiple linear regression to estimate each subject’s PWV expected for his/her age and systolic
BP, and compared with observed values to obtain standardized residuals (z-scores). We defined EVA when z-score was
≥1.96. Finally, a multivariable logistic regression analysis was performed to determine baseline characteristics associated
with EVA. Mean age was 49.9 ± 15.5 years, male gender was 50.3%. Mean PWV was 7.52 m/s (SD 1.97), mean systolic BP
was 125.3 mmHg (SD 16.7) and mean diastolic BP was 78.9 mmHg (SD 12.2). The frequency of EVA was 5.7% in the total
population, 9.8% in adults of 40 years or less and 18.7% in those 30 years or less. In these young adults, multiple logistic
regression analyses demonstrated that dyslipidemia and hypertension showed an independent association with EVA, and
smoking a borderline association (p=0.07). In conclusion, the frequency of EVA in a sample from Latin America was
around 6%, with higher rates in young adults. These results would support the search of CVRF and EVA during early
adulthood.
Introduction
Arteriosclerosis develops throughout the life-course starting
at very early stages (i.e., in utero and during childhood) and
is influenced by genetic and environmental cardiovascular
risk factors (CVRF), even though the clinical expression
appears decades later [1–3]. This phenomenon has fostered
the adoption of a life-course approach to reduce CVRF and
cardiometabolic disease [4].
First described in 2008 by Nilsson PM et al., the “early
(or accelerated) vascular aging”(EVA) is a growing clinical
concept that mainly refers to the observation of an increased
arterial stiffness (arteriosclerosis) in either susceptible
individuals under the influence of CVRF, when compared
with the expected arterial stiffness according to their
chronological age [1,5].
Carotid to femoral pulse wave velocity (c-f PWV)
measurement represents the propagation velocity of the
pulse wave and is currently regarded as the gold standard
for the assessment of arterial stiffness [6,7]. Despite being
strongly correlated with age, blood pressure (BP), and
metabolic factors [8–14], c-f PWV represents an indepen-
dent predictor of coronary heart disease, stroke, cognitive
decline, and cardiovascular death, after adjusting for the
established CVRF [15,16]. Therefore, c-f PWV plays a
central role in the EVA definition, but given the lack of an
operational definition (threshold values), it has been
*Fernando Botto
ferbotto@icloud.com
1Center of Hypertension and Vascular Aging, Cardiology Institute
and Cardiovascular Therapeutics, Hospital Universitario Austral,
Pilar, Buenos Aires, Argentina
2Instituto de Efectividad Clínica y Sanitaria, Buenos Aires,
Argentina
3San Raffaele Pisana - Istituto Ricovero e Cura a Carattere
Scientifico (IRCCS), Rome, Italy
4Department of Clinical Sciences, Lund University, Skane
University Hospital, Malmö, Sweden
Electronic supplementary material The online version of this article
(https://doi.org/10.1038/s41371-018-0038-1) contains supplementary
material, which is available to authorized users.
1234567890();,:

proposed to define EVA when PWV values are above the
97.5th percentile of the age-adjusted z-score, using as a
normal reference cohort the Reference Values for Arterial
Stiffness Collaboration [7,17]
The OPTIMO study was designed to evaluate lifestyle
predictors of healthy arteries in a population sample from
Latin America. In the present analysis, our primary objec-
tive was to determine the frequency of EVA focused on the
detection of subjects with a true increase in the arterial
stiffness (i.e., elevation of estimated PWV) independently
of the effect of age and BP. Since higher PWV values
characterize older people, we sought to investigate the
existence of EVA with emphasis on young adults (age range
between 20 and 40 years), when both EVA and subclinical
arteriosclerosis reach a high prevalence [17,18]. Second,
we evaluated the relationship of baseline variables with
EVA.
Materials and methods
The OPTIMO study design is based on an international
prospective cohort of adults aged 20 years or more from 12
Latin American countries (Argentina, Brazil, Chile,
Colombia, Costa Rica, Guatemala, Honduras, México,
Nicaragua, Panamá, Dominican Republic, and El Salvador).
The present report is based on a cross-sectional analysis of
the baseline information collected over 1 year, from October
2014 to October 2015.
Individuals from public and private general and cardio-
vascular health facilities, and also from public places (i.e.,
shopping malls) and working areas (i.e., factory or labora-
tory employees) were asked to voluntarily participate, after
signing an informed consent. The protocol required a con-
secutive recruitment during the time of collaboration. There
was no formal invitation, therefore each investigator offered
participation to subjects following local strategies and ethics
rules.
All participants provided information about their lifestyle
(dietary habits, alcohol intake, smoking, and physical
activity), demographic (age and gender), anthropometric
and socio-economic variables (education, occupation, and
marital status), and medical history, by completing the
WHO STEPS surveillance questionnaire [19]. Although this
is a self-administered survey, research staff helped indivi-
duals to complete it adequately. As blood samples were
drawn only in one-fourth of the cases, such results are not
included in this analysis.
Nutrition was tabulated using the unit “days per week”
(d/w) as a continuous variable for consumption of fruits,
vegetables, fish, seeds/nuts, and alcohol. Regular alcohol
intake was also defined as 3 or more d/w, regular fish
consumption as 2 or more d/w, and regular exercise 3 or
more d/w. Regarding medical history, diagnosis of hyper-
tension (HTN), dyslipidemia, or diabetes were considered
present if the subject reported use of any drug medication
for the condition or if he/she referred that it was previously
diagnosed by a physician. Overweight and obesity were
defined as a body mass index of ≥25 kg/m2and ≥30 kg/m2,
respectively. Prior atherosclerotic cardiovascular disease
(ASCVD) was defined as coronary heart disease, myo-
cardial infarction, stroke, or peripheral arterial disease.
Noninvasive measurements of systolic BP (SBP) and
diastolic BP (DBP), heart rate, and an estimated value of
PWV were determined using the Mobil-O-Graph®(IEM,
Stolberg, Germany), a commercially available brachial
oscillometric BP monitor validated by the European Society
of Hypertension [20]. The device includes the ARCSolver®
method (Austrian Institute of Technology, Vienna) that
generates the aortic pulse wave using a proprietary transfer
function after checking for signal quality. Estimated PWV
using this technology has been successfully compared with
tonometric reference devices [13,21–24]. A regular bra-
chial cuff adjusted to the circumference of the left arm in
each individual was applied and measurements were per-
formed after 5 min of rest in a sitting position. During the
assessment, speaking was not allowed. The monitor proto-
col includes a first brachial BP measurement, then a 30-s
pause, followed by a second measurement. The first is
automatically discarded and the second one is reported.
Measurement was repeated only in the case of a poor or
regular quality according to the quality checking. Monitor
application and evaluation was performed by physicians
previously trained with the method for use in clinical
practice. The OPTIMO protocol did not instruct about the
presence or not of the person who operated the device.
Actually, the operator was usually present. Patients with
atrial fibrillation or frequent premature cardiac beat con-
tractions and those who consumed food or smoked in the
prior 2 h were excluded.
The study protocol was approved by the research ethics
board at every screening site. Informed consent was
obtained from each participating individual.
Statistical analysis
Continuous variables are expressed as mean ± standard
deviation unless otherwise specified, and categorical vari-
ables are expressed as percentages. Estimated PWV and BP
results according to the age are described in decades of life.
We developed a multiple linear regression model among
the whole population to estimate each subject’s PWV
expected for his/her age and SBP registered during the
examination. We checked for linearity and model fitting by
comparing the observed PWV and the PWV predicted by
the model. Then, we performed a standardized residuals
F. Botto et al.

analysis (z-scores) to determine EVA frequency using the
function “Observed PWV–Predicted PWV/SD Predicted
PWV”. EVA was determined when z-score exceeded +1.96
(equivalent to 97.5th percentile). Further, low PWV was
defined by a z-score lower than −1.96 and normal PWV
when z-score was in between. We determined the frequency
of EVA in the general population, as well as in young adults
under 40 years and under 30 years.
As the primary objective of the present analysis was
to describe the distribution of the predicted value of
PWV adjusted for age and SBP in a general population
sample, based on prior information [17] we expected a
frequency of EVA around 10% (95% confidence interval
(CI): 8–12%). Taking into account, an alpha error of
0.05 and a power of 90% we estimated a target sample size
of 1100 individuals.
We also evaluated independent baseline characteristics
associated with EVA using multiple logistic regression
models based on the whole population and further in sub-
jects under 40 years, after excluding those with prior
ASCVD events. We reported adjusted odds ratios (OR)
95% CIs and associated p-values to three decimals. For all
tests, a p-value < 0.05 was considered significant. All ana-
lyses were performed using STATA version 14 (StataCorp,
USA).
Results
We recruited a total of 1511 subjects, of whom 1416
were included in the present report after excluding
those with incomplete data (3.3%) or measurement failure
(3%). Distribution per country was mostly in Argentina
(76%), followed by México (7%), Brazil (5%), Colombia
(5%), and the remaining countries (7%). Mean age was
49.9 (SD 15.5) years and male gender was 50.3%. Ninety-
four percent reported completion of secondary school
or had reached a higher educational level. The frequency
of classic CVRF in the overall population was: current
smokers 12.2%, dyslipidemia 58.3%, HTN 41.2%,
diabetes 11.6%, overweight 34.9%, and obesity 24.4%.
Regular exercise was reported by 60% of participants.
Table 1shows intake of alcohol, fruits, vegetables, fish,
and nuts.
During the evaluation, 25% of the total population had
HTN (of these 65% had prior HTN). Among those with
prior HTN, 40% showed values >140 and/or >90 mmHg.
Among those without prior HTN, 15% showed BP elevated
values. Table 2describes the distribution of age, PWV,
SBP, and DBP in the total population. Mean PWV was 7.5
m/s (SD 1.9), with a 90th percentile of 10.1 m/s. We found
159 (11.2%) cases with PWV >10 m/s. Mean SBP was
125.3 mmHg (SD 16.7) and mean DBP was 78.9 mmHg
Table 1 Baseline characteristics of the total OPTIMO study
population (n=1416)
Variable Result
Age, mean (SD) 49.9 (15.5)
Male gender (%) 50.3
Height, cm (SD) 168.4 (9.8)
Weight, kg (SD) 75.9 (17.3)
Education level (%)
<7 years (no or incomplete primary) 1.8
7 years (complete primary) 4.2
12 years (complete secondary) 23.3
>12 years (tertiary, no university) 22.7
>12 years (university or post-degree) 48
Cardiovascular risk factors (%)
Current smoking 12.2
Dyslipidemia 58.3
Hypertension 41.2
Diabetes 11.6
Overweight 34.9
Obesity 24.4
Prior ASCVD event (%) 6.3
Aspirin treatment (%) 15.6
Antihypertensive drugs use (%) 39.2
Lipid-lowering drugs use (%) 48
Any statin 41
Regular alcohol intake (%) 61.3
Alcohol intake, days per week (SD) 3.7 (1.9)
Fruits, days per week (SD) 4.8 (2.2)
Vegetables, days per week (SD) 5 (2)
Fish, days per week (SD) 1.1 (0.9)
Seeds/nuts, days per week (SD) 1.4 (1.9)
Regular exercise (%) 60.2
Exercise, days per week (SD) 2.2 (1.9)
ASCVD atherosclerotic cardiovascular disease
Table 2 Age, PWV, SBP, and DBP values (n=1416)
Mean
(SD)
Min-max Percentiles
10% 25% 50% 75% 90%
Age, years 49.9
(15.5)
20–91 27 38 51 61 70
PWV, m/s 7.5
(1.9)
2.1–15 5.1 6.0 7.3 8.7 10.1
SBP,
mmHg
125.3
(16.7)
87–198 105 114 123.5 135.5 145
DBP,
mmHg
78.9
(12.2)
49–118 65 72 79 86 92
PWV pulse wave velocity, SBP systolic blood pressure, DBP diastolic
blood pressure
Early vascular aging in Latin America...

(SD 12.2). Regarding young adults (<40 years), actual BP
values showed systolic and/or diastolic HTN in 43 of 376
(11.4%), categorized as follows: isolated systolic HTN in 24
(6.4%), isolated diastolic HTN in 11 (2.9%), and combined
systolic and diastolic HTN in 8 (2.1%).
Table 3shows mean PWV per decade of life in the total
population, and in the subgroup of healthy subjects with
neither CVRF nor prior ASCVD events (n=455). As
expected, the observed PWV was lower in the healthy
subgroup of each age category compared with the whole
population.
In Supplemental Table 1, we present PWV means and
SD according to decades and BP categories. Individuals
<30 years with BP <120/80 mmHg showed the lowest PWV
with a mean value of 4.8 m/s and individuals >80 years with
BP >140/90 mmHg showed a mean PWV almost three
times higher, with a value of 13.2 m/s.
Frequency of EVA
Multiple linear regression analyses with PWV as dependent
variable was applied in a final model that included age,
quadratic age (age2), and SBP, showing for each variable a
significant association with PWV. Supplemental Fig. 1
shows model development and the final model with, as
expected, a very good linear prediction of PWV. Supple-
mental Fig. 2 shows the linear relationship of predicted and
observed PWV values by age as a continuous variable, and
Supplemental Figure 3 shows boxplots of the observed
PWV across age categories, where some outliers stand out,
such as high PWV values in young subjects and low PWV
in elders. Boxplots in Fig. 1indicate a very good correlation
between PWV values predicted by the model and the
observed PWV values across age categories, after adjusting
for age and SBP. The PWV outliers described before remain
after including SBP in the model.
Finally, Fig. 2shows a scatter-plot of PWV z-scores
distribution according to age and SBP (standardized ana-
lysis). PWV z-score values higher than those predicted by
the model (z-score higher than +1.96) persist in younger
subjects. They complied with the EVA definition and
represent 5.7% (81/1416) of the total sample. Additionally,
PWV z-score values lower than those predicted by the
model (z-score lower than −1.96) depict older subjects with
the healthier arteries, who represent 4.5% (63/1416) of total
population.
EVA frequency was also determined in the subgroup of
young adults: we found 9.8% (37/376) in subjects <40 years
and 18.7% (30/160) in those <30 years.
Variables associated with EVA
Multiple logistic regression analysis performed in the whole
population, after exclusion of 90 individuals with prior
ASCVD, and adjusted for alcohol intake and smoking,
showed that variables significantly associated with EVA
were age, OR 0.92 (95% CI: 0.90–0.94, p< 0.001), dysli-
pidemia, OR 2.36 (95% CI: 1.31–4.23, p=0.004), and
aspirin intake, OR 4.28 (95% CI: 1.84–9.97, p=0.001)
(Supplemental Figure 4a).
A similar analysis restricted to subjects aged 40 years or
less determined that baseline characteristics significantly
associated with EVA were age, OR 0.78 (95% CI:
0.72–0.85, p< 0.001), dyslipidemia, OR 6.88 (95% CI:
2.79–16.99, p< 0.001), history of HTN, OR 3.29 (95% CI:
1.02–10.55, p=0.045), and regular alcohol intake, OR 0.27
(95% CI: 0.10–0.73, p=0.011). Smoking showed a bor-
derline significance, OR 2.32 (95% CI: 0.91–5.85, p=
0.075) (Supplemental Figure 4b).
Table 3 Mean PWV according to age categories in the total
population and in healthy subjects with neither cardiovascular risk
factors nor prior ASCVD events
Age (years) Total population
(n=1416)
Healthy subjects
(n=455)
nMean PWV (SD),
m/s
nMean PWV (SD),
m/s
<30 160 5.20 (1.18) 105 4.98 (0.91)
30–39 216 5.77 (0.79) 110 5.65 (0.73)
40–49 261 6.54 (0.70) 87 6.35 (0.78)
50–59 272 7.79 (0.81) 75 7.55 (0.77)
60–69 264 9.14 (0.86) 49 8.93 (0.99)
70–79 102 10.37 (1.36) 19 9.93 (1.86)
>79 41 12.35 (1.86) --- ---
PWV pulse wave velocity, ASCVD atherosclerotic cardiovascular
disease
PWV: pulse wave velocity
SBP: systolic blood pressure
0 5 10 15
20-29 30-39 40-49 50-59 60-69 >70
AGE (years)
PWV Fitte d values
PWV, m/s
Fig. 1 Boxplots of PWV values predicted by the model and the
observed PWV values adjusted for age and SBP. (Colour figure online)
F. Botto et al.

We further explored the association between alcohol
intake and EVA after increasing the sample by including
subjects <50 years but we did not find a significant asso-
ciation, OR 0.57 (95% CI: 0.24–1.30, p=0.18).
Discussion
We found an overall frequency of EVA of around 6% in a
mixed population sample from 12 countries in Latin
America using a simple, practical, and affordable oscillo-
metric device on the brachial artery, which estimates PWV
as a surrogate of arterial stiffness. Interestingly, the EVA
frequency was particularly elevated in young adults under
40 years (9.8%), and even higher in subjects under 30 years
(18.7%).
According to our standardized analysis, including age
and SBP determined simultaneously with PWV, z-scores
suggested that EVA has a low frequency after the age of 60
years because PWV values are mostly predicted by older
age and SBP. Interestingly, in this age subgroup we also
recorded a frequency of 4.5% of individuals with a very low
PWV that represent elderly subjects with the healthiest
arteries.
We further described PWV means (and SD) in the total
population stratified by age decades and BP categories. To
separate healthy arteries and EVA in our sample, the per-
centiles 2.5th and 97.5th can be calculated by applying the
formula PWV ± 1.96 SD to the values showed in Supple-
mental Table 1.
Finally, our study suggests that in young adults between
20 and 40 years the presence of dyslipidemia and HTN, and
probably smoking, may contribute to the early development
of arterial stiffness, or EVA, starting even in the early
twenties when medical care is less often requested.
We also wanted to put our study results in perspective of
previous studies with similar aims. The Guimaràes/Vizela
Study [17] included 2542 randomly sampled subjects over
18 years from Northern Portugal. The authors applied an
EVA definition based upon the age-adjusted normal Eur-
opean population of the Reference Values for Arterial
Stiffness Collaboration, after measurements with a tono-
metric device (Sphygmocor) [7]. Therefore, a PWV ≥97.5th
percentile of z-score for mean PWV values adjusted for age
was considered as a practical definition of EVA. Conse-
quently, they reported a 12.5% overall prevalence of EVA,
with a 19.3% in subjects <40 years and 26.1% in those <30
years. Z-scores analyses in OPTIMO study, using its own
sample z-scores as a normal reference, demonstrated a
lower frequency of EVA (overall 5.7%, but 9.8% in sub-
jects <40 years and 18.7% in <30 years).
The OPTIMO study design included a convenience
sampling that probably reflects a more selected population if
we take into account, for example, a high average educa-
tional level, which allows better lifestyle, nutrition, and
healthcare, and also a higher use of lipid-lowering drugs,
compared with the Guimaràes/Vizela Study (48% vs.
17.7%). Anyway, both studies found an elevated frequency
of EVA in young adults.
We believe that the observed low frequency of EVA after
60 years reflects the fact that PWV is supposed to be a
surrogate marker of arterial stiffness, which attempts to
identify vascular damage at earlier stages of life than
expected, but loses diagnostic precision with advanced age
and higher SBP values [15]. Accordingly, multiple logistic
regression analysis determined that age had an inverse
independent relationship with EVA (OR 0.93) implying a
stronger association in younger subjects than in older. This
evidence does not argue against the elevated prevalence rate
of PWV in the elderly and its independent prognostic value
in this subgroup. However, to the best of our knowledge,
the OPTIMO and the Guimaràes/Vizela studies [17] are the
only screening studies that have reported a high frequency
of EVA in younger adults.
Existing evidence reinforces the importance of EVA
diagnosis in youth. Among 2849 elderly individuals, the
Rotterdam study established that PWV added to the Fra-
mingham score allowed for a limited risk reclassification
and did not provide a significant clinical utility to predict
cardiovascular events during 8 years of follow-up [25].
Furthermore, an individual participant meta-analysis that
included 17,635 subjects, determined that the predictive
power of PWV for future ASCVD events and mortality was
stronger in younger and middle-aged subjects than in older
people [15].
Regarding the PWV cut-off value of 10 m/s proposed by
a consensus document [26], we believe that using fix
thresholds regardless of age and BP, whichever the
PWV: pulse wave velocity
SBP: systolic blood pressure
Fig. 2 Scatter plots of the distribution of PWV z-scores according to
age and SBP. (Colour figure online)
Early vascular aging in Latin America...

measurement technique applied, does not work for our
operational definition of EVA, and probably it also repre-
sents an imperfect marker of subclinical organ damage,
particularly in the older. In OPTIMO study, we found
11.2% individuals with PWV over the proposed limit that
does not necessarily match with cases of EVA. Actually,
EVA cases in young adults are mostly below PWV of 10 m/
s (see outliers in Fig. 1).
Regarding the use of a similar oscillometric device in a
general population, Nunan et al. [27] performed an inter-
esting study estimating PWV with Mobil-O-Graph in
1794 subjects from a community setting (“real world”)in
Vienna, Austria. As we did in OPTIMO study, they
adjusted PWV for age and BP levels, but also for gender.
PWV results were quite similar in both studies, with mild
differences around 0.65 m/s on average in subgroups <70
years, probably due to population samples features.
Between subgroups 70 and 79 years old, there was a higher
difference of 1 m/s (11.4 m/s in Nunan et al. vs. 10.37 m/s in
OPTIMO study). However, in those elderly subjects with
BP <140/90 mmHg, both studies reported a similar PWV
(10.6 and 10.5 m/s, respectively). A possible explanation is
that in the first study the rate of elevated BP >140/90 mmHg
was higher compared with OPTIMO study (60% vs. 30%,
respectively), probably due to a higher proportion of phar-
macologically treated subjects in our sample (95% of those
with prior HTN).
To the best of our knowledge, this is the first study
determining EVA frequency in a sample from Latin
America. There exists some regional reports from
population-based studies on normal/reference PWV values
according to age and BP categories in healthy individuals,
mostly performed with tonometric methods [22,28–30].
They provide a basis for diagnosis of vascular aging but,
however, they do not report EVA prevalence, neither in the
general nor in the young adult population.
Analyses of the baseline characteristics of young adults
under 40 years in our OPTIMO study suggest that EVA is
associated with biological determinants, such as history of
HTN, dyslipidemia, and smoking. Similarly, the Amster-
dam Growth and Health Longitudinal Study (AGAHLS)
described that the same CVRF during adolescence and
young adulthood anticipated the development of arterial
stiffness at the age of 36 years [31]. Both arteriosclerosis
and EVA reach a high prevalence in youth, relatively
speaking [17,18]. Therefore, in spite of the lack of
data related to ASCVD outcomes in young populations,
these findings support the aim of promoting cardiovascular
prevention during childhood or adolescence, that is,
through healthy lifestyle in order to prevent arterial
stiffening during the following years and thereby potentially
reducing the risk of future ASCVD events and increasing
survival [32,33].
A potential biased finding in the OPTIMO study was the
observation of a beneficial relationship between regular
alcohol intake and EVA in young adults. The INTER-
HEART Study demonstrated that regular alcohol intake,
defined as three or more times per week, was an indepen-
dent “protector”for myocardial infarction adjusted by age,
sex, and smoking (OR 0.79, 95% CI: 0.73–0.86), but not
when other CVRFs were added to the model [34]. Fur-
thermore, a subgroup analysis demonstrated that regular
alcohol intake was not significantly associated to risk of
myocardial infarction in subjects <45 years (OR 0.94, 95%
CI: 0.81–1.11) [35]. Our data showed a benefit of regular
alcohol intake on PWV in young adults <40 years, however,
the effect disappeared by expanding the subgroup sample
size to subjects <50 years. The risk of EVA for aspirin users
could reflect another observational bias, because subjects
with elevated cardiovascular risk are more expected to be
treated or self-medicated with aspirin.
Strengths and limitations
The prospective cohort design included a diverse ethnic
population from many countries of Latin America. How-
ever, a high proportion of participants were included in one
country. Therefore, the lack of a representative sample
limits us to refer to frequency of EVA instead of prevalence.
Importantly, data collection was performed using a
standardized questionnaire with predetermined definitions,
and PWV, as well as BP were determined in all subjects
using the same type of device and measurement protocol. In
spite of the lack of a direct measurement of PWV, the
oscillometric device Mobil-O-Graph has been satisfactorily
compared with the applanation tonometry (SphygmoCor)
and other devices [13,21–24]. Furthermore, it is easy to use
with a simple training and is relatively inexpensive.
Luzardo et al. [22] compared Mobil-O-Graph measure-
ments with a tonometric device (SphygmoCor) in a volun-
teers sample from Uruguay. In the substudy performed at
rest in the laboratory, they found no significant differences
in observed results, reporting a mean PWV of 7.3 (SD 1.9)
m/s with tonometry and 7.0 (SD 2.2) m/s (p=0.11) with
oscillometry. They found a statistically significant differ-
ence in PWV (7.9 m/s, SD 2.1, vs. 7.4 m/s, SD 1.6) but only
in the substudy performed with Mobil-O-Graph “ambula-
tory”monitoring during 24 h, a different condition com-
pared with our OPTIMO evaluation at rest. Anyway,
tonometric methods represent the gold standard for PWV
and large prospective studies are still needed to validate
oscillometric devices.
Increasing age is a strong marker of arterial stiffness,
and SBP is a surrogate of it, particularly in older people
[8,9,11–13]. Therefore, we believe that our proposed
methodology to predict PWV based on a multiple linear
F. Botto et al.

regression model adjusted for age and SBP, followed by a
standardized residuals analysis, allowed us to determine a
realistic frequency of EVA in our sample. We also adjusted
the analyses for treatment with lipid-lowering and anti-
hypertensive drugs.
Our limited study population sample size is insufficient
to generate firm conclusions on frequencies, particularly in
some categories of age and BP distribution including a
small number of participants, as well as some associations
between baseline characteristics and EVA. Furthermore, our
sample is biased due to the convenience sampling design
performed at some private medical centers or shopping
areas where educated and health-conscious people pre-
dominate, instead of recruiting a random population sample.
As previously mentioned, an increased use of lipid-lowering
(48%) and antihypertensive drugs (40%), a high rate of self-
reported regular exercise (60%), and dominance of partici-
pants with high educational levels, support the existence of
this potential bias. The cross-sectional design only allows us
to determine associations, but not causality. Finally, the lack
of an independent comparative cohort in Latin America for
derivation of cut-off levels for EVA to be used in OPTIMO
study is regretful, but as the OPTIMO study is the first on
the continent other screening studies will probably follow.
In recent years, arterial stiffness has emerged as an
independent predictor of cardiovascular risk and represent a
core component of the novel EVA syndrome along with
other changes of the arterial wall [1,5]. Measurement of
PWV is currently accepted as the most simple, noninvasive,
and reproducible method to determine arterial stiffness [6].
An individual participant meta-analysis of 17,635 subjects
has demonstrated its independent prognostic value to pre-
dict future ASCVD events and mortality, even after
adjusting for classic CVRF, and allows reclassification of
risk categories, particularly in young people [15].
In spite of attempts to standardize reference values of
PWV obtained by different techniques (i.e., tonometric,
oscillometric, ultrasound, magnetic resonance imaging) [7],
the EVA syndrome represents a “proof of concept”still
without an exact established definition [36]. However, we
believe that searching for EVA based on PWV values from
any validated method at this stage represents a step forward
in cardiovascular prevention. In this direction, the OPTIMO
study represents a multicenter experience from Latin
America setting the basis of EVA prevalence and
encourages future clinical work and research in the field.
Young adulthood (age between 20 and 40 years) repre-
sents the healthiest period of life, and therefore cardiovas-
cular health promotion is usually scarce in this age group.
Accumulated evidence calls for attention and debate
regarding the age of initiation of cardiovascular screening
for subclinical arteriosclerosis and atherosclerosis. The
“Progression of Early Subclinical Atherosclerosis”(PESA)
study [18] included 40–54 years asymptomatic participants
and determined the existence of 63% of subclinical ather-
osclerosis in any of the carotid, abdominal aortic, ilio-
femoral, or coronary territories. The authors of PESA
reported that among subjects with a low 10-year risk cal-
culated with the Framingham risk score, 58% had sub-
clinical disease. If we consider, for example, that in the
United States [37], the mean age of ST-elevation myo-
cardial infarction is 64 (SD 13) years, therefore 34% (1 SD)
occur between 51 and 64 years and 14% (2 SD) between 38
and 51 years. Consequently, strategies to prevent these
ASCVD events in young people should start 10–15 years
earlier. We believe that detecting an EVA syndrome before
the age of 40 and even 30 years represents a con-
temporaneous challenge for cardiovascular prevention.
Conclusion
The OPTIMO study shows a frequency of EVA around 6%
among a population sample from Latin America, with a
higher rate in young adults. Baseline variables associated
with EVA in the former subgroup were well-known CVRF,
such as HTN, dyslipidemia, and smoking. Our results are
consistent with those from others, and call for a debate
about the search for CVRF and subclinical atherosclerosis
during early adulthood. In this regard, determination of
arterial stiffness (EVA) using measurements of PWV with
simple and inexpensive devices might help to select sub-
jects from these age groups at increased cardiovascular risk
who deserve a more intense preventive intervention and
follow-up.
Summary Table
What is known about this topic?
●EVA is a growing clinical concept that refers to an
increased arterial stiffness (arteriosclerosis) when com-
pared with the expected level of arterial stiffness
according to the chronological age.
●There is no data about EVA prevalence and its
characteristics obtained from a population sample in
Latin America.
●Little is known about determinants of EVA in young
adults.
What this study adds?
●We found an overall frequency of EVA of around 6%
using a simple oscillometric device on the brachial
artery, which calculates the PWV as a marker of arterial
stiffness.
Early vascular aging in Latin America...

●EVA frequency was higher in young adults under 40
years (9.8%), and even higher in subjects under 30 years
(18.7%).
●In young adults, the presence of dyslipidemia and
hypertension, and probably smoking, may contribute to
the early development of EVA.
Acknowledgements Investigators: (1) Argentina: Ana Di Leva, Martín
Koretzky, Pedro Forcada, Gabriel Waisman, Laura Brandani, Gabriela
Fischer Sohn, Ezequiel Huguet, Mariana Haehnel, Patricia Carrizo,
Patricia Pardini, Gustavo Maccallini; (2) Brazil: Marco Mota, Nelson
Dinamarco, Martin Vilela; (3) Chile: Enrique Lorca; (4) Colombia:
Jannes Buelvas, Gabriel Robledo Káiser; (5) Costa Rica: Francisco
Rivera Valvidia; (6) El Salvador: Freddis E. Molina, Jaime Ventura,
José A. Velasquez; (7) Guatemala: Laura Voguel, Julio Arriola; (8)
Honduras: Gerardo Sosa, Dora Arévalo, Jaqueline Gonzalez, Mauricio
Varela, Marcelino Abadie, José R.Vasquez; (9) Mexico: Ernesto
Cardonna Muñoz; (10) Nicaragua: José D. Meneses, José A. Montiel;
(11) Panama: José L. Donato; (12) Republica Dominicana: Nelson
Baez, Luis Ney Novas, Solange R. Ureña.
Funding OPTIMO study was basically a research program performed
with the collaboration of the aforementioned investigators who
received no honoraria for their participation.
Compliance with ethical standards
Conflicts of interest The authors declare that they have no conflict of
interest.
References
1. Nilsson PM, Lurbe E, Laurent S. The early life origins of vascular
ageing and cardiovascular risk: the EVA syndrome. J Hypertens.
2008;26:1049–57.
2. Aatola H, Hutri-Kähönen N, Juonala M, Viikari JSA, Hulkkonen
J, Laitinen T, et al. Lifetime risk factors and arterial pulse wave
velocity in adulthood: the cardiovascular risk in young finns
study. Hypertension. 2010;55:806–11.
3. Juonala M, Magnussen CG, Venn A, Dwyer T, Burns TL, Davis
PH, et al. Influence of age on associations between childhood risk
factors and carotid intima-media thickness in adulthood: the car-
diovascular risk in young finns study, the childhood determinants
of adult health study, the bogalusa heart study, and the muscatine
st. Circulation. 2010;122:2514–20.
4. Olsen MH, Angell SY, Asma S, Boutouyrie P, Burger D, Chirinos
JA, et al. A call to action and a lifecourse strategy to address the
global burden of raised blood pressure on current and future
generations: the Lancet Commission on hypertension. Lancet.
2016;388:2665–712.
5. Nilsson PM, Boutouyrie P, Laurent S. Vascular aging: a tale of
eva and ADAM in cardiovascular risk assessment and prevention.
Hypertension. 2009;54:3–10.
6. Laurent Stephane, Cockcroft John, Bortel LucVan,
Boutouyrie Pierre, Giannattasio Cristina, Hayoz Daniel,
Pannier Bruno, et al. Expert consensus document on arterial
stiffness: methodological issues and clinical applications. Eur
Heart J. 2006;27:2588–605.
7. The Reference Values for Arterial Stiffness Collaboration.
Determinants of pulse wave velocity in healthy people and in the
presence of cardiovascular risk factors: “establishing normal and
reference values. Eur Hear J. 2010;31:2338–50.
8. McEniery CM, Yasmin, Hall IR, Qasem A, Wilkinson IB,
Cockcroft JR. Normal vascular aging: differential effects on wave
reflection and aortic pulse wave velocity - the Anglo-Cardiff
Collaborative Trial (ACCT). J Am Coll Cardiol.
2005;46:1753–60.
9. Rogers WJ, Hu YL, Coast D, Vido DA, Kramer CM, Pyeritz RE,
et al. Age-associated changes in regional aortic pulse wave
velocity. J Am Coll Cardiol. 2001;38:1123–9.
10. Mitchell GF, Parise H, Benjamin EJ, Larson MG, Keyes MJ, Vita
JA, et al. Changes in arterial stiffness and wave reflection with
advancing age in healthy men and women: the Framingham Heart
Study. Hypertension . 2004;43:1239–45.
11. Avolio AP, Ph D, Fa-quan D, Wei-qiang LI, Yao-fei L, Zhen-
dong H, et al. Effects of aging on arterial distensibility in popu-
lations with high and low prevalence of hypertension: comparison
between urban and rural communities in China. Circulation .
1985;71:202–10.
12. Cecelja M, Chowienczyk P. Dissociation of aortic pulse wave
velocity with risk factors for cardiovascular disease other than
hypertension. Hypertension. 2009;54:1328. LP-1336
13. Feistritzer H-J, Reinstadler SJ, Klug G, Kremser C, Seidner B,
Esterhammer R, et al. Comparison of an oscillometric method
with cardiac magnetic resonance for the analysis of aortic pulse
wave velocity. PLoS ONE. 2015;10:e0116862.
14. Scuteri A, Najjar SS, Orru M, Usala G, Piras MG, Ferrucci L,
et al. The central arterial burden of the metabolic syndrome is
similar in men and women: the SardiNIA Study. Eur Heart J.
2010;31:602–13.
15. Ben-Shlomo Y, Spears M, Boustred C, May M, Anderson SG,
Benjamin EJ, et al. Aortic pulse wave velocity improves cardio-
vascular event prediction: an individual participant meta-analysis
of prospective observational data from 17,635 subjects. J Am Coll
Cardiol. 2014;63:636–46.
16. Scuteri A, Wang H. Pulse wave velocity as a marker of cognitive
impairment in the elderly. J Alzheimer’s Dis. 2014;42:s401–10.
17. Cunha PG, Cotter J, Oliveira P, Vila I, Boutouyrie P, Laurent S,
et al. Pulse wave velocity distribution in a cohort study: from
arterial stiffness to early vascular aging. J Hypertens.
2015;33:1438–45.
18. Fernandez-Friera L, Peñalvo JL, Fernandez-Ortiz A, Ibañez B,
Lopez-Melgar B, Laclaustra M, et al. Prevalence, vascular dis-
tribution, and multiterritorial extent of subclinical atherosclerosis
in a middle-aged cohort the PESA (Progression of Early Sub-
clinical Atherosclerosis) study. Circulation. 2015;131:2104–13.
19. WHO | STEPwise approach to surveillance (STEPS). WHO
[Internet]. 2015 [cited 2017 Jan 10]; Available from: http://www.
who.int/chp/steps/en/
20. Franssen PML, Imholz BPM. Evaluation of the Mobil-O-Graph
new generation ABPM device using the ESH criteria. Blood Press
Monit. 2010;15:229–31.
21. Weiss W, Gohlisch C, Harsch-Gladisch C, Tö M, Zidek W, Van
Der Giet M, et al. Oscillometric estimation of central blood
pressure: validation of the Mobil-O-Graph in comparison with the
SphygmoCor device. Blood Press Monit 2012;17:128-131
22. Luzardo L, Lujambio I, Sottolano M, da Rosa A, Thijs L, Noboa
O, et al. 24-H ambulatory recording of aortic pulse wave velocity
and central systolic augmentation: a Feasibility study. Hypertens
Res. 2012;35:980–7.
23. Hametner B, Wassertheurer S, Kropf J, Mayer C, Eber B, Weber
T. Oscillometric estimation of aortic pulse wave velocity: com-
parison with intra-aortic catheter measurements. Blood Press
Monit. 2013;18:173–6.
24. Reshetnik A, Gohlisch C, Tölle M, Zidek W, Van Der Giet M.
Oscillometric assessment of arterial stiffness in everyday clinical
practice. Hypertens Res. 2017;40(2):140-145.
F. Botto et al.

25. Verwoert G, Elias-Smale S, Rizopoulos D, Koller M, Steyerberg
E, Hofman A, et al. Does aortic stiffness improve the prediction of
coronary heart disease in elderly? The Rotterdam Study. J Hum
Hypertens. 2012;26:28–34.
26. Van Bortel L, Laurent S, Boutouyrie P, Chowienczyk P,
Cruickshank JK, Backer T, et al. Expert consensus document on
the measurement of aortic stiffness in daily practice using carotid-
femoral pulse wave velocity. J Hypertens. 2012;30(3):445-8.
27. Nunan D, Fleming S, Hametner B, Wassertheurer S. Performance
of pulse wave velocity measured using a brachial cuff in a com-
munity setting. Blood Press Monit. 2014;19:315–9.
28. Farro I, Bia D, Zocalo Y, Torrado J, Farro F, Florio L, et al. Pulse
wave velocity as marker of preclinical arterial disease: reference
levels in a Uruguayan population considering wave detection
algorithms, path lengths, aging, and blood pressure. Int. J
Hypertens. 2012;2012:1–10.
29. Díaz A, Galli C, Tringler M, Ramírez A, Cabrera Fischer EI.
Reference values of pulse wave velocity in healthy people from an
urban and rural argentinean population. Int J Hypertens.
2014;2014:653239.
30. Christen AI, Miranda AP, Caride SG, Armentano RL. Pulse wave
velocity: relevance of age in normotensive, essential hypertensive
and borderline hypertensive patients. Rev Argent Cardiol.
2015;83:124–9.
31. Ferreira I, Van De Laar RJ, Prins MH, Twisk JW, Stehouwer CD.
Carotid stiffness in young adults: a life-course analysis of its early
determinants: the Amsterdam growth and health longitudinal
study. Hypertension. 2012;59:54–61.
32. Laitinen T, Laitinen TT, Pahkala K, Magnussen CG, Viikari JSA,
Oikonen M, et al. Ideal cardiovascular health in childhood and
cardiometabolic outcomes in adulthood: the cardiovascular risk in
young finns study. Circulation. 2012;125:1971–8.
33. Kaikkonen JE, Mikkilä V, Magnussen CG, Juonala M, Viikari
JSA, Raitakari OT. Does childhood nutrition influence adult car-
diovascular disease risk?--Insights from the Young Finns Study.
Ann Med. 2013;45:120–8.
34. Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F,
et al. Effect of potentially modifiable risk factors associated with
myocardial infarction in 52 countries (the INTERHEART study):
case control study. Lancet. 2004;364:937–52.
35. Leong DP, Smyth A, Teo KK, McKee M, Rangarajan S, Pais P,
et al. Patterns of alcohol consumption and myocardial infarction
risk: observations from 52 countries in the INTERHEART case-
control study. Circulation. 2014;130:390–8.
36. Nilsson P, Olsen M, Laurent S. Early vascular ageing (EVA).
New directions in cardiovascular protection. 1st edition. Amster-
dam: Elsevier Inc.: Academic Press; 2015. xi-xii.
37. Pride YB, Canto JG, Frederick PD, Gibson CM. Outcomes among
patients with ST-segment-elevation myocardial infarction pre-
senting to interventional hospitals with and without on-site cardiac
surgery. Circ Cardiovasc Qual Outcomes. 2009;2:574–82.
Early vascular aging in Latin America...