Atherosclerosis 243 (2015) 93e98

Contents lists available at ScienceDirect

Atherosclerosis
journal homepage: www.elsevier.com/locate/atherosclerosis

A Mediterranean diet and risk of myocardial infarction, heart failure

and stroke: A population-based cohort study
Thanasis G. Tektonidis a, Agneta Åkesson a, Bruna Gigante b, c, Alicja Wolk a,
Susanna C. Larsson a, *
a €g 13, Box 210, SE-171 77, Stockholm, Sweden
Unit of Nutritional Epidemiology, Institute of Environmental Medicine (IMM), Karolinska Institutet, Nobels va
b
Division of Cardiovascular Medicine, Department of Clinical Sciences, Danderyd University Hospital, Stockholm, Sweden
c
Unit of Cardiovascular Epidemiology, Institute of Environmental Medicine (IMM), Karolinska Institutet, Nobels va €g 13, Box 210, SE-171 77, Stockholm,
Sweden

a r t i c l e i n f o a b s t r a c t

Article history: Background and aims: The Mediterranean diet, which is palatable and easily achievable, has been
Received 11 June 2015 associated with lower all-cause and cardiovascular disease (CVD) incidence and mortality. Data on heart
Received in revised form failure (HF) and stroke types are lacking. The aim was to examine a Mediterranean diet in relation to
21 August 2015
incidence of myocardial infarction (MI), HF and stroke types in a Swedish prospective cohort.
Accepted 28 August 2015
Methods: In a population-based cohort of 32,921 women, diet was assessed through a self-administered
Available online 3 September 2015
questionnaire. The modified Mediterranean diet (mMED) score was created based on high consumption
of vegetables, fruits, legumes, nuts, whole grains, fermented dairy products, fish and monounsaturated
Keywords:
Mediterranean diet
fat, moderate intakes of alcohol and low consumption of red meat, on a 0e8 scale. Relative risks (RR)
Cardiovascular disease with 95% confidence intervals (CI), adjusted for potential confounders, were estimated by Cox propor-
Myocardial infarction tional hazards regression models.
Heart failure Results: During 10 y of follow-up (1998e2008), 1109 MIs, 1648 HFs, 1270 ischemic strokes and 262 total
Stroke hemorrhagic strokes were ascertained. A high adherence to the mMED score (6e8), compared to low,
Primary prevention was associated with a lower risk of MI (RR: 0.74, 95% CI: 0.61e0.90, p ¼ 0.003), HF (RR: 0.79, 95% CI: 0.68
Prospective studies e0.93, p ¼ 0.004) and ischemic stroke (RR: 0.78, 95% CI: 0.65e0.93, p ¼ 0.007), but not hemorrhagic
stroke (RR: 0.88, 95% CI: 0.61e1.29, p ¼ 0.53).
Conclusions: Better adherence to a Mediterranean diet was associated with lower risk of MI, HF and
ischemic stroke. The Mediterranean diet is most likely to be beneficial in primary prevention of all major
types of atherosclerosis-related CVD.
© 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction grains, monounsaturated fat, moderate intakes of dairy products

Cardiovascular disease (CVD) is the leading cause of morbidity
and death worldwide [1,2]. The costs for health care and medica-
tions related to CVD surpass V200 and $300 billion in Europe and
the U.S., respectively [2]. Lifestyle including a healthy diet can
reduce CVD mortality by 50% [3,4]. The Mediterranean diet is
palatable and easily achievable. Therefore, this diet has gained
popularity also in non-Mediterranean countries [5]. It is charac-
terized by high intakes of vegetables, fruits, legumes, nuts, whole
* Corresponding author.
E-mail addresses: thanasis.tektonidis@ki.se (T.G. Tektonidis), Agneta.Akesson@
ki.se (A. Åkesson), Bruna.Gigante@ki.se (B. Gigante), Alicja.Wolk@ki.se (A. Wolk),
Susanna.Larsson@ki.se (S.C. Larsson).
(mainly cheese and yogurt), fish and alcohol and low intakes of red
and processed meat [6]. It has been consistently observed that the
Mediterranean diet is associated with lower all-cause and CVD
mortality in prospective studies [7e14], as well as with lower
incidence of myocardial infarction (MI)/coronary heart disease and
stroke [7,14e19]. Furthermore, high adherence to the Mediterra-
nean diet has been associated with improvement in clinical risk
factors of CVD in clinical trials [20e23]. However, data is scarce on
the relationship between the Mediterranean diet and incidence of
heart failure (HF) [24] and stroke types [14,17,19,25]. Moreover,
most previous studies have assessed the association between a
Mediterranean diet and risk of either total CVD or only MI or stroke.
Hence, the aim of this study was to determine the association be-
tween the adherence to a Mediterranean diet, as indicated by a

http://dx.doi.org/10.1016/j.atherosclerosis.2015.08.039
0021-9150/© 2015 Elsevier Ireland Ltd. All rights reserved.

Descargado para Anonymous User (n/a) en New Granada Military University de ClinicalKey.es por Elsevier en agosto 24, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
94 T.G. Tektonidis et al. / Atherosclerosis 243 (2015) 93e98
modified Mediterranean diet (mMED) score, and incidence of MI,
HF and stroke types, during a 10-year follow up of a large Swedish
population.
2. Subjects and methods
2.1. Study population
The population-based, prospective Swedish Mammography
Cohort provided data for the present study. In the late autumn of
1997, 39,227 women who were born between 1914 and 1948, were
residents of Uppsala and V€astmanland counties of central Sweden,
and had already participated in the Swedish Mammography
Cohort-study 10 years earlier, completed a 350-item questionnaire
about diet and lifestyle (participation rate 70%). Women with a
missing or an erroneous National Registration Number were
omitted. We also excluded those with cancer (n ¼ 1803) and car-
diovascular disease (ischemic heart disease, HF and stroke,
n ¼ 2495) at baseline and those with implausible energy intakes
(i.e., 3 SDs from the loge-transformed mean total energy intake;
n ¼ 404). Furthermore, women with missing values of the com-
ponents of the mMED score were excluded (n ¼ 1319). After these
exclusions, 32,921 women (aged 48e83 years) remained for the
analysis. Informed consent was obtained from each participant and
the study protocol conformed to the ethical guidelines of the 1975
Declaration of Helsinki as reflected in a priori approval by the
Regional Ethical Review Board at Karolinska Institutet in Stock-
holm, Sweden.
2.2. Baseline assessment of covariates and diet
We obtained information on education, body weight and height,
tobacco smoking, aspirin use, prevalence of hypertension, hyper-
cholesterolemia, family history of MI before 60 years of age, alcohol
consumption, physical activity and diet through a self-administered
questionnaire. We calculated body mass index by dividing the
weight in kilograms by the square of height in meter. Data on
diabetes were available through the National Diabetes Register, the
National Patient Register at the National Board of Health and
Welfare, and was also self-reported in the questionnaire.
The food frequency questionnaire (FFQ) reflected the habitual
average consumption of 96 different foods/food items and bever-
ages during the previous year. The FFQ included open-ended
questions with predefined serving sizes (1 glass, 1 cup, teaspoon,
tablespoon, slice) for commonly consumed foods (e.g. dairy
products and bread) and 8 predetermined categories of food fre-
quency (range from never to >3 times/day) for other food items,
including vegetables, fruits, legumes, nuts, grains, fish, red and
processed meat and alcoholic beverages. Reported frequencies of
consumption of specific alcoholic beverages (wine, beer, and li-
quor) were multiplied by the reported amount, consumed at each
occasion, resulting in an average total alcohol intake [26]. The
question about the quality of fat intake was “what type of fat do
you usually use on sandwiches and in dressing and cooking?” We
calculated totally energy intake by multiplying the frequency of
intake of each food item by the energy content of age-specific
portion sizes, using composition values from the Swedish Food
Administration Database [27]. The FFQ used in the present study
has been validated for foods and nutrients by comparison with the
mean of multiple 24-h recall interviews distributed over a year
[28]. The Spearman correlation coefficients between estimates
from the FFQ and the mean of fourteen 24-h recall interviews
ranged from 0.44 (protein) to 0.81 (alcohol) for macronutrients,
with a mean value of 0.65 [28].
Descargado para Anonymous User (n/a) en New Granada Military University de ClinicalKey.es por Elsevier en agosto 24, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
2.3. Modified Mediterranean diet (mMED) score
The mMED score indicating the degree of relative adherence to
the traditional Mediterranean diet was adapted from the Mediter-
ranean diet scale by Trichopoulou et al. [8,9]. The mMED score
included: 1) vegetables and fruits (excluding fruit juices and po-
tatoes), 2) legumes and nuts, 3) non-refined/high fiber grains
(whole meal bread, crisp bread, oatmeal and bran of wheat), 4)
fermented dairy products (cultured milk, yogurt, and cheese), 5)
fish, 6) red and processed meat, 7) use of olive oil and/or rapeseed
oil and 8) alcohol. Participants with an intake above the median
intake received 1 point for the presumed beneficial components:
vegetables and fruits, legumes and nuts, non-refined/high fiber
grains, fermented dairy products, and fish; for intake below the
median intake they received 0 points. For red and processed meat,
the reverse was applied. For alcohol, a value of 1 was assigned to
women who consumed on average between 5 and 15 g of ethanol
per day, otherwise 0 points. For the quality of fat, 1 point was
assigned to women who used olive oil and/or rapeseed oil as a main
source of fat for cooking or as dressing, otherwise 0 points. The total
mMED score ranged from 0 (low adherence) to 8 (high adherence to
Mediterranean diet).
2.4. Case ascertainment and follow-up
Primary incident cases of MI, HF (including HF event listed
either as the primary diagnosis or at any diagnosis position) and
stroke that occurred in the cohort during follow-up were ascer-
tained by linkage of the study cohort to the Swedish Inpatient
Register and the Swedish Cause of Death Register. We used the
International Classification of Diseases, 10th revision, codes to
classify the cardiovascular events as MI (I21), HF (I11 and I50),
ischemic stroke (I63) and hemorrhagic stroke (I60 [subarachnoid
hemorrhage] and I61 [intracerebral hemorrhage]).
2.5. Statistical analysis
Participants accumulated follow-up time from 1 January 1998
until the date of diagnosis of MI, HF, or stroke, date of death (the
Swedish Cause of Death Register), or end of follow-up (31
December 2008), whichever came first. We categorized women
into quartiles of the mMED score (slightly larger 1st quartile). We
also analyzed mMED score as a continuous variable per 1-point
increment. Relative risks (RR) with corresponding 95% confidence
intervals (CIs) of MI, HF, ischemic stroke and hemorrhagic stroke
were estimated by using Cox proportional hazards regression
models with age as the time scale. All multivariable models were
also adjusted for education level (12 years, >12 years) and
established risk factors for cardiovascular disease, including family
history of MI before 60 year of age (yes or no), cigarette smoking
(current, former, never), physical activity as following: >40 min
walking and/or cycling per day (yes or no), >1 h of exercise per
week (yes or no), body mass index (<20, 20e24.9, 25e29.9, 30 kg/
m2), history of hypertension (yes or no), history of hypercholes-
terolemia (yes or no), history of diabetes (yes or no), aspirin use
(never, 1e6 tablets/week, 7 tablets/week) and total energy intake
(continuous, kcal/day). Missing values were treated as a separate
“missing category” in the models and were few (<2%), with the
exception of physical activity (<10%).
In order to determine whether associations could be described
by a single component, we also tested single components of the
mMED score separately as variables above/below the median
(while mutually adjusted for all other components and risk factors).
We also ran a sensitivity analysis after excluding alcohol from the
mMED score. For HF, we performed a sensitivity analysis including
 T.G. Tektonidis et al. / Atherosclerosis 243 (2015) 93e98
HF events recorded as the primary diagnosis only. In addition, since
subjects who suffered an MI have a higher risk of developing HF
[29], we conducted a sub-analysis for HF incidence, by excluding HF
cases occurring after an MI.
A test for trend across quartiles of the mMED score was per-
formed by assigning the median value for each quartile and
modeling this variable as a continuous variable. We checked
whether the proportional hazard assumption was reasonable by
means of scaled Schoenfeld's residuals, which were regressed
against the survival time. There was no evidence of departure from
this assumption. All the statistical analyses were performed with
STATA software, version 12.1 (StataCorp, LP, College Station, Texas,
USA). All P values were 2-sided and the level of statistical signifi-
cance was set to 5%.
3. Results
Over a mean follow-up of 10.4 years (343,685 person-years) of
32,921 women, we ascertained the following number of incident
cardiovascular events: 1109 MIs, 1648 HFs, 1270 ischemic strokes
and 262 hemorrhagic strokes (78 subarachnoid hemorrhages and
184 intracerebral hemorrhages). Daily intakes of each component
of the mMED score are presented in Table 1. Baseline characteristics
of the study population according to the mMED score are presented
in Table 2. Compared with women with the lowest adherence to the
Mediterranean diet (mMED score: 0e3), those with the highest
adherence (mMED score: 6e8) were more likely to have a post-
secondary education and to be more physically active, as indicated
by walking or cycling for >40 min/day and by leisure-time exercise
for >1 h/week and were less likely to be current smokers, over-
weight or obese and frequent users of aspirin. As expected, they
consumed more of the presumed beneficial components of the
mMED score and less red and processed meat. In addition, they
were more likely to use olive or rapeseed oil and had better
compliance with moderate alcohol consumption.
A high adherence to the Mediterranean diet, as indicated by a
high mMED score (6e8), was associated with a statistically signif-
icant lower risk of MI, HF and ischemic stroke but not of total
hemorrhagic stroke, in the age-adjusted model (Table 3). After
further adjustment for known risk factors for cardiovascular dis-
ease, women in the highest quartile of the mMED score had a
Table 1
Consumption of each food component of the modified Mediterranean diet (mMED)
score in the Swedish Mammography Cohort, n ¼ 32,921.
mMED score componentc Meana 50th percentileb
(Servings/day)
4.0 4.0
Vegetables & fruits 4.9 4.4
Legumes & nuts 0.2 0.2
Non-refined/high fiber grains 3.2 3.0
Fermented dairy 3.6 3.0
Fish 0.3 0.2
Red & processed meat 1.0 0.9
% and results did not reach statistical significance [7,17]. The consis-
Use of olive oil and/or rapeseed oil 47.9
Moderate alcohol intake, 5e15 g/day 29.4
For red & processed meat: 1 point if below the median, 0 point if above the median.
For olive oil and/or rapeseed oil: 1 point if use of olive oil or rapeseed oil, (spread or/
and cooking), 0 if no use.
For alcohol intake: 1 point if alcohol intake is 5e15 g/day, 0 point if alcohol intake <5
or >15 g/day.
a
Mean values are age-standardized.
b
The 50th percentile (median value) for each component was used as cut-off
point to create the mMED score, unless otherwise stated.
c
For vegetables & fruits, legume & nuts, non-refined/high fiber grains, fermented
dairy and fish: 1 point if above the median, 0 point if below the median.
Descargado para Anonymous User (n/a) en New Granada Military University de ClinicalKey.es por Elsevier en agosto 24, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
95
statistically significant 26%, 21%, and 22% lower risk of MI, HF and
ischemic stroke, respectively, and a statistically non-significant 12%
lower risk of hemorrhagic stroke, compared with those in the
lowest mMED score quartile (Table 3). The mMED score was not
significantly associated with risk of subarachnoid hemorrhage or
intracerebral hemorrhage separately. For HF, we performed an
additional analysis including only HF events recorded as the pri-
mary diagnosis (n ¼ 832), which resulted in 16% lower risk (RR:
0.84, 95% CI: 0.67e1.04). Furthermore, exclusion of HF incident
cases occurring after an MI did not modify the results majorly (RR:
0.81, 95% CI: 0.69e0.97). Removing alcohol from the mMED score
did not appreciably affect the results for MI, HF or stroke
(Supplemental Table 1). Analyses of the association of single mMED
score components (low or high intake according to median values)
with MI, HF, and ischemic and hemorrhagic stroke showed only a
few significant results (Supplemental Table 2).
4. Discussion
In this prospective cohort of CVD-free women, we observed that
high adherence to the Mediterranean diet was associated with a
21e26% risk reduction of MI, HF and ischemic stroke incidence,
compared with women with low adherence. Associations did not
alter significantly after adjustment for major risk factors for CVD. A
12% lower risk of incident hemorrhagic stroke was also observed,
but the result was not statistically significant. Supplemental anal-
ysis showed that these results are most likely due to the combi-
nation of all 8 components of the mMED score, rather than single
components.
The present findings are consistent with the latest large meta-
analysis that aggregated evidence from major studies evaluating
the association between the Mediterranean diet and risk of total
CVD [30]. After including 14 studies and a total of about 4,000,000
subjects from non- and Mediterranean populations, Sofi et al. [30]
observed a 10% lower risk of CVD (including coronary heart dis-
ease and stroke) for a 2-point increase in a 18-point adherence
score to the Mediterranean diet. Trials on primary CVD prevention
using a Mediterranean diet intervention are lacking, while large
reduction in rates of CVD were observed when components of the
Mediterranean diet (extra virgin olive oil or mixed nuts) were
tested in a primary prevention (the PREDIMED trial) in a Spanish
population at high CVD risk [31]. Moreover, a secondary CVD pre-
vention intervention study (the “Lyon Diet Heart Study”) observed
a protective association among subjects who followed a Mediter-
ranean diet against recurrence after a first MI [21]).
Most previous studies focused on total coronary heart disease,
Range rather than MI specifically as in our study, and observed inverse
associations with the Mediterranean diet [15,18,21,30]. The risk
0e8
reduction in MI incidence in our study was of similar magnitude
0e41.5 with those from two prospective cohorts of non-Mediterranean
0e7.0
populations in the U.S. and the Netherlands (29e30% risk reduc-
0e25.2
0e25.0 tion) [14,16]. Two other studies also found a trend towards an in-
0e6.0 verse association between a high MED score and risk of MI;
0e11.6 however the number of MI cases was limited (133 and 161 cases)
tent results across different populations strongly support that a
Mediterranean diet may be very beneficial in primary prevention of
MI, by lowering the risk by one fourth.
To our knowledge, our study is the first assessing the relation
between adherence to the Mediterranean diet and incidence of HF.
Previous studies have focused on the association between a Med-
iterranean diet and either improvement of biomarkers of HF or
survival after HF diagnosis [20,24]. Fito et al. conducted a ran-
domized controlled trial as part of the PREDIMED trial on the as-
sociation of a Mediterranean diet with HF biomarkers among 930
96 T.G. Tektonidis et al. / Atherosclerosis 243 (2015) 93e98

Table 2
Age-standardizeda baseline characteristics by quartiles of the modified Mediterranean diet (mMED) score in the study population of the Swedish Mammography Cohort, 1997,
n ¼ 32,921.

Median mMED score (0e8) Quartiles of mMED score

Quartile 1 Quartile 2 Quartile 3 Quartile 4

(n ¼ 12,321) (n ¼ 7421) (n ¼ 6752) (n ¼ 6427)

3.0 4.0 5.0 6.0

Characteristicsb
Age, years 61.7 61.0 60.6 60.2
Education >12 years, yes (%) 13.3 18.3 22.8 28.8
Current smoker, yes (%) 26.9 23.4 21.0 18.9
Body mass indexc, kg/m2 25.2 25.0 24.7 24.4
Overweight and obese, >25 kg/m2 (%) 47.3 44.3 40.7 37.6
Physical activity
Walking or/and cycling, >40 min/day (%) 29.3 33.3 36.4 38.9
Exercise, >1 h/week (%) 66.6 72.6 77.4 81.4
Aspirin use, >7 tablets per week (%) 6.0 5.7 5.8 5.4
Family history of MI (%) 12.6 13.5 13.9 13.3
Prevalence of hypertension (%) 20.2 19.1 19.4 19.3
Prevalence of hypercholesterolemia (%) 7.2 7.3 8.1 8.0
Prevalence of diabetes (%) 3.9 4.0 4.0 3.9
Energy intake, kcal/day 1529 1764 1897 2037
Alcohol intake, g/day 5.0 5.4 6.0 6.7
Never & former alcohol users (%) 19.2 16.2 13.6 8.5
Components of mMED score (servings/day or %)
mMED score (0e8) 2.3 4 .0 5.0 6.3
Vegetables & fruits 3.5 4.8 5.7 6.7
Legumes & nuts 0.2 0.2 0.3 0.4
Non-refined/high fiber grains 2.5 3.3 3.7 4.2
Fermented dairy 2.7 3.6 4.1 4.8
Fish 0.2 0.3 0.3 0.4
Red & processed meat 1.1 1.1 1.1 0.9
Olive oil and/or rapeseed oil use, yes (%) 22.5 45.7 61.1 85.2
Moderate alcohol intake, 5e15 g/day (%) 15.4 26.6 34.1 53.7
a
Mean values of age and mMED score are not age-standardized.
b
Continuous variables are presented as means and categorical as proportions.
c
Calculated by dividing weight in kilograms by the square of height in meters.

subjects at high cardiovascular risk [20]. Women who followed a diet lower in fat. In a prospective study of 3215 women (50e80
Mediterranean diet with extra virgin olive oil or nuts for 1 year years old) who suffered from non-fatal HF, the Mediterranean diet
improved risk factors of HF (oxidized low-density lipoprotein, li- score was inversely associated with CVD mortality (HR: 0.85, 95%
poprotein (a) and NT-proBNO) compared to those who followed a CI: 0.70e1.02) [24].

Table 3
Relative risks of myocardial infarction, heart failure and stroke types by quartiles of the modified Mediterranean diet (mMED) score in the Swedish Mammography Cohort,
1998e2008, n ¼ 32,921.

Quartiles of mMED score (0e8) Quartile 1 Quartile 2 Quartile 3 Quartile 4 P-trend 1-point increment

Average mean mMED score (range) 2.3 (0e3) 4.0 (4e4) 5.0 (5e5) 6.3 (6e8) 4.0 (0e8)
Median mMED score 3 4 5 6 4
Myocardial infarction
Cases (n ¼ 1109) 514 256 181 158 1109
Person-years 126,843 77,480 71,131 68,230 343,685
Age-adjusted RR (95% CI) 1 0.87 (0.74e1.01) 0.69 (0.58e0.82) 0.66 (0.55-0-79) <0.0001 0.90 (0.86e0.93)
Multivariate adjusted RRa (95%CI) 1 0.89 (0.76e1.04) 0.75 (0.63e0.89) 0.74 (0.61e0.90) <0.0001 0.92 (0.89e0.96)
Heart failure
Cases (n ¼ 1648) 767 347 290 244 1648
Person-years 126,606 77,561 71,069 68,144 343,381
Age-adjusted RR (95% CI) 1 0.81 (0.71e0.92) 0.78 (0.68e0.89) 0.74 (0.64e0.86) <0.0001 0.92 (0.89e0.95)
Multivariate adjusted RRa (95% CI) 1 0.81 (0.71e0.93) 0.82 (0.71-0-95) 0.79 (0.68e0.93) 0.001 0.94 (0.91e0.97)
Ischemic stroke
Cases (n ¼ 1270) 544 294 243 189 1270
Person-years 126,803 77,511 71,025 68,068 343,408
Age-adjusted RR (95% CI) 1 0.95 (0.82e1.09) 0.89 (0.76e1.04) 0.77 (0.65e0.91) <0.01 0.94 (0.91e0.97)
Multivariate adjusted RRa (95% CI) 1 0.94 (0.81e1.08) 0.89 (0.76e1.05) 0.78 (0.65e0.93) <0.01 0.94 (0.90e0.98)
Hemorrhagic stroke
Cases (n ¼ 262) 106 60 49 47 262
Person-years 128,428 78,300 71,679 68,591 347,000
Age-adjusted RR (95% CI) 1 0.96 (0.70e1.32) 0.87 (0.62e1.23) 0.91 (0.64e1.28) 0.47 0.94 (0.87e1.02)
Multivariate adjusted RRa (95%CI) 1 0.94 (0.68e1.30) 0.85 (0.59e1.21) 0.88 (0.61e1.29) 0.42 0.93 (0.85e1.01)

RR: Relative Risk, CI: Confidence Interval.

a
Adjusted for education level (up to 12 years, >years), family history of myocardial infarction (yes or no), cigarette smoking (current, former, never), >40 min of walking or/
and cycling per day (yes or no), >1 h of exercise per week (yes or no), BMI (<20, 20e24.9, 25e29.9, 30 kg/m2), history of hypertension (yes or no), history of hypercho-
lesterolemia (yes or no), history 12 of diabetes (yes or no), aspirin use (never, 1e6 tablets/week, 7 tablets/week) and total energy intake (continuous, kcal/day).

Descargado para Anonymous User (n/a) en New Granada Military University de ClinicalKey.es por Elsevier en agosto 24, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
 T.G. Tektonidis et al. / Atherosclerosis 243 (2015) 93e98
A recent meta-analysis of dietary patterns in relation to stroke
risk showed that adherence to a Mediterranean diet was associated
with a 32% reduced risk of total stroke [32]. Among previous studies
of the relation between the Mediterranean diet and risk of stroke,
few examined ischemic and hemorrhagic stroke separately
[14,17,19,25]. In our study, we found that a Mediterranean diet was
significantly inversely associated with risk of ischemic stroke but
not hemorrhagic stroke. These findings are consistent with those
from a prospective cohort study of 30,000 individuals in the U.S.
that showed a significant reduction in the risk of ischemic stroke
(RR: 0.79, 95% CI: 0.65e0.96), but not of hemorrhagic stroke, among
those with high adherence to a Mediterranean diet [19].
Our findings are in line with previous evidence showing that the
adherence to a Mediterranean diet is likely to be beneficial in pri-
mary prevention of CVD, also in non-Mediterranean populations
[7,14,16,17,19]. To our knowledge, this is the first large prospective
cohort study to assess the relationship of the Mediterranean diet
with the incidence of several separate CVDs, including MI, HF,
ischemic stroke and hemorrhagic stroke, which together account
for 70% of the deaths attributable to CVD [4]. Moreover, the large
sample size, the long follow-up period in our study, the prospective
assessment of dietary exposure and lifestyle information, and
almost complete follow-up added strength to our study. The mMED
score was based on previous fundamental literature, but was
modified to match the cultural food differences in the study pop-
ulation [8,9]. Compared to most scores used in other study coun-
tries, we pooled legumes and nuts in one category, due to low nut
consumption in the population and we included fermented dairy
because of the latest evidence on survival [33]. Compared with low
or no consumption, high consumption of non-refined grains (whole
grains) and fiber from cereals, in particular from non-refined grains,
has been consistently reported to be inversely associated with
cardiovascular endpoints (7e30% risk reduction) and risk of coro-
nary heart disease (20e40% risk reduction) [34e36]; yet fiber is not
the sole component in the non-refined grains that may reduce CVD
risk, as evidence suggests a synergistic effect of fiber with phyto-
chemicals or other food components in non-refined grains, which
might be called as the “fiber complex” [37]. Hence, we included
both non-refined and high-fiber grain products. Rapeseed oil is a
presumed beneficial source of fat because of its low content of
saturated fatty acids, high content of monounsaturated fatty acids
(as in olive oil) and polyunsaturated fatty acids (including a-lino-
lenic acid), and because it has the lowest omega-6 to omega-3 ratio
of the commonly used vegetable oils [38,39]. In a recent study,
rapeseed oil as part of a hypo-caloric diet was associated with
weight reduction and an improvement in the overall cardiovascular
profile among individuals with the metabolic syndrome [40].
Moreover, in Sweden, the yearly per capita use of olive oil is quite
low (0.8 versus 2.6 kg in Europe), whereas the use of rapeseed oil is
high (7.0 vs 2.3 kg in Europe) [38]. Therefore, in the mMED score,
we included rapeseed oil along with olive oil as a representative
indicator of fat quality in the Swedish population. In general,
regardless of the design of the different scores, evidence consis-
tently indicates more beneficial health outcomes in individuals
with high Mediterranean-like diet scores.
Dietary intake was measured through a food frequency ques-
tionnaire and at baseline only; thus, we were not able to capture
potential changes in diet during the follow-up. Confounding is a
potential issue in observational studies as individuals who have a
healthy lifestyle, including healthier food choices, tend to be more
health conscious. As a result, women with high adherence to the
Mediterranean diet may have been more likely to receive medical
treatment for diabetes, hypertension, hypercholesterolemia or
other CVD risk factors, and this may have prevented a major CVD
event. We partially controlled for this possibility by excluding
Descargado para Anonymous User (n/a) en New Granada Military University de ClinicalKey.es por Elsevier en agosto 24, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
97
women with prior CVD before baseline. In addition, women in the
highest quartile of the mMED score smoked less and were more
physically active compared with those in the lowest quartile. We
did not have information on sociodemographic/socioeconomic
variables, but we adjusted for education level, which partially
controlled for socioeconomic status.
In conclusion, greater adherence to a Mediterranean diet, as
indicated by the mMED score, was associated with lower risk of MI,
HF and of ischemic stroke, but not hemorrhagic stroke in this non-
Mediterranean population. These results are of great interest since
they suggest that a broad, easily achievable and palatable dietary
concept, the Mediterranean diet, is most likely to be beneficial in
primary prevention of all atherosclerosis related major types of
CVD. Since this is the first study on the association between a
Mediterranean diet and risk of HF, our results for HF need to be
replicated in other populations. Future studies should also focus on
the association of Mediterranean diet on different stroke types.
Sources of support
This study was supported by research grants from the Swedish
Research Council, the Swedish Stroke Association, the Swedish
Research Council for Health, Working Life and Welfare (Forte), and
by a Strategic Young Scholar Grant in Epidemiology (SfoEpi) at
Karolinska Institutet. The funders had no role in the study design,
data collection, analysis, decision to publish, or preparation of the
manuscript.
Acknowledgments
All authors have contributed significantly to the work, attested
to the validity and legitimacy of the data and its interpretation, and
agreed to its submission to the Atherosclerosis; as principal inves-
tigator, Alicja Wolk was responsible for data collection; Thanasis G.
Tektonidis, Agneta Åkesson and Susanna C. Larsson conducted the
study and analyzed the data; Thanasis G. Tektonidis drafted the
paper and all authors have reviewed and approved the manuscript
being submitted. None of the authors reported any personal or
financial conflict of interest.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.atherosclerosis.2015.08.039.
References
[1] S.L. Murphy, J. Xu, K.D. Kochanek, Deaths: final data for 2010, Natl. Vital Stat.
Rep. 61 (4) (2013) 111e117.
[2] M. Nichols, N. Townsend, R. Luengo-Fernandez, et al., European Cardiovas-
cular Disease Statistics, European Heart Network and European Society of
Cardiology, Sophia Antipolis, Brussels, 2012.
[3] S.A. Schroeder, Shattuck Lecture. We can do bettereimproving the health of
the American people, N. Engl. J. Med. 357 (12) (2007) 1221e1228.
[4] D. Mozaffarian, E.J. Benjamin, A.S. Go, et al., Heart disease and stroke statistics-
2015 update: a report from the American heart association, Circulation 131
(4) (2015) p. e29-e322.
[5] R. da Silva, A. Bach-Faig, B. Raido Quintana, et al., Worldwide variation of
adherence to the Mediterranean diet, in 1961-1965 and 2000-2003, Public
Health Nutr. 12 (9a) (2009) 1676e1684.
[6] W.C. Willett, F. Sacks, A. Trichopoulou, et al., Mediterranean diet pyramid: a
cultural model for healthy eating, Am. J. Clin. Nutr. 61 (6 Suppl. l) (1995)
1402se1406s.
[7] G. Tognon, L. Lissner, D. Saebye, et al., The Mediterranean diet in relation to
mortality and CVD: a Danish cohort study, Br. J. Nutr. 111 (1) (2014) 151e159.
[8] A. Trichopoulou, T. Costacou, C. Bamia, et al., Adherence to a Mediterranean
diet and survival in a Greek population, N. Engl. J. Med. 348 (26) (2003)
2599e2608.
[9] A. Trichopoulou, A. Kouris-Blazos, M.L. Wahlqvist, et al., Diet and overall
survival in elderly people, BMJ 311 (7018) (1995) 1457e1460.
98 T.G. Tektonidis et al. / Atherosclerosis 243 (2015) 93e98
[10] A. Trichopoulou, P. Orfanos, T. Norat, et al., Modified Mediterranean diet and
survival: EPIC-elderly prospective cohort study, BMJ 330 (7498) (2005) 991.
[11] K.T. Knoops, L.C. de Groot, D. Kromhout, et al., Mediterranean diet, lifestyle
factors, and 10-year mortality in elderly European men and women: the HALE
project, JAMA 292 (12) (2004) 1433e1439.
[12] P.N. Mitrou, V. Kipnis, A.C. Thiebaut, et al., Mediterranean dietary pattern and
prediction of all-cause mortality in a US population: results from the NIH-
AARP diet and health study, Arch. Intern Med. 167 (22) (2007) 2461e2468.
[13] E. Lopez-Garcia, F. Rodriguez-Artalejo, T.Y. Li, et al., The Mediterranean-style
dietary pattern and mortality among men and women with cardiovascular
disease, Am. J. Clin. Nutr. 99 (1) (2014) 172e180.
[14] T.T. Fung, K.M. Rexrode, C.S. Mantzoros, et al., Mediterranean diet and inci-
dence of and mortality from coronary heart disease and stroke in women,
Circulation 119 (8) (2009) 1093e1100.
[15] G. Buckland, C.A. Gonzalez, A. Agudo, et al., Adherence to the Mediterranean
diet and risk of coronary heart disease in the Spanish EPIC cohort study, Am. J.
Epidemiol. 170 (12) (2009) 1518e1529.
[16] M.P. Hoevenaar-Blom, A.C. Nooyens, D. Kromhout, et al., Mediterranean style
diet and 12-year incidence of cardiovascular diseases: the EPIC-NL cohort
study, PLoS One 7 (9) (2012) p. e45458.
[17] H. Gardener, C.B. Wright, Y. Gu, et al., Mediterranean-style diet and risk of
ischemic stroke, myocardial infarction, and vascular death: the Northern
Manhattan Study, Am. J. Clin. Nutr. 94 (6) (2011) 1458e1464.
[18] D.B. Panagiotakos, E.N. Georgousopoulou, C. Pitsavos, et al., Exploring the path
of Mediterranean diet on 10-year incidence of cardiovascular disease: the
ATTICA study, Nutr. Metab. Cardiovasc. Dis. 25 (3) (2015) 327e335.
[19] G. Tsivgoulis, T. Psaltopoulou, V.G. Wadley, et al., Adherence to a mediterra-
nean diet and prediction of incident stroke, Stroke 46 (3) (2015) 780e785.
[20] M. Fito, R. Estruch, J. Salas-Salvado, et al., Effect of the Mediterranean diet on
heart failure biomarkers: a randomized sample from the PREDIMED trial, Eur.
J. Heart Fail 16 (5) (2014) 543e550.
[21] M. de Lorgeril, P. Salen, J.L. Martin, et al., Mediterranean diet, traditional risk
factors, and the rate of cardiovascular complications after myocardial infarc-
tion: final report of the Lyon Diet heart Study, Circulation 99 (6) (1999)
779e785.
[22] K. Esposito, R. Marfella, M. Ciotola, et al., Effect of a mediterranean-style diet
on endothelial dysfunction and markers of vascular inflammation in the
metabolic syndrome: a randomized trial, Jama 292 (12) (2004) 1440e1446.
[23] S. Vincent-Baudry, C. Defoort, M. Gerber, et al., The Medi-RIVAGE study:
reduction of cardiovascular disease risk factors after a 3-mo intervention with
a Mediterranean-type diet or a low-fat diet, Am. J. Clin. Nutr. 82 (5) (2005)
964e971.
[24] E.B. Levitan, C.E. Lewis, L.F. Tinker, et al., Mediterranean and DASH diet scores
and mortality in women with heart failure: the Women's health Initiative,
Circ. Heart Fail 6 (6) (2013) 1116e1123.
Descargado para Anonymous User (n/a) en New Granada Military University de ClinicalKey.es por Elsevier en agosto 24, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
[25] G. Misirli, V. Benetou, P. Lagiou, et al., Relation of the traditional Mediterra-
nean diet to cerebrovascular disease in a Mediterranean population, Am. J.
Epidemiol. 176 (12) (2012) 1185e1192.
[26] O. Stackelberg, M. Bjorck, S.C. Larsson, et al., Alcohol consumption, specific
alcoholic beverages, and abdominal aortic aneurysm, Circulation 130 (8)
(2014) 646e652.
[27] L. Bergstro€m, E. Kylberg, U. Hagman, et al., The food composition database
KOST: the national food administration's information system for nutritive
values of food, Vår Fo € (1991) 43439e43447.
[28] M. Messerer, S.E. Johansson, A. Wolk, The validity of questionnaire-based
micronutrient intake estimates is increased by including dietary supplement
use in Swedish men, J. Nutr. 134 (7) (2004) 1800e1805.
[29] M.G. Sutton, N. Sharpe, Left ventricular remodeling after myocardial infarc-
tion: pathophysiology and therapy, Circulation 101 (25) (2000) 2981e2988.
[30] F. Sofi, C. Macchi, R. Abbate, et al., Mediterranean diet and health status: an
updated meta-analysis and a proposal for a literature-based adherence score,
Public Health Nutr. 17 (12) (2014) 2769e2782.
[31] R. Estruch, E. Ros, J. Salas-Salvado, et al., Primary prevention of cardiovascular
disease with a Mediterranean diet, N. Engl. J. Med. 368 (14) (2013)
1279e1290.
[32] M.D. Kontogianni, D.B. Panagiotakos, Dietary patterns and stroke: a systematic
review and re-meta-analysis, Maturitas 79 (1) (2014) 41e47.
[33] K. Michaelsson, A. Wolk, S. Langenskiold, et al., Milk intake and risk of mor-
tality and fractures in women and men: cohort studies, Bmj (2014) 349 p.
g6015.
[34] M.A. Pereira, E. O'Reilly, K. Augustsson, et al., Dietary fiber and risk of coronary
heart disease: a pooled analysis of cohort studies, Arch. Intern Med. 164 (4)
(2004) 370e376.
[35] A. Satija, F.B. Hu, Cardiovascular benefits of dietary fiber, Curr. Atheroscler.
Rep. 14 (6) (2012) 505e514.
[36] G. Tang, D. Wang, J. Long, et al., Meta-analysis of the association between
whole grain intake and coronary heart disease risk, Am. J. Cardiol. 115 (5)
(2015) 625e629.
[37] D.R. Jacobs Jr., L.M. Steffen, Nutrients, foods, and dietary patterns as exposures
in research: a framework for food synergy, Am. J. Clin. Nutr. 78 (3 Suppl. l)
(2003) 508Se513S.
[38] B. Strandvik, Y. Chen, F. Dangardt, et al., From the Swedish to the Mediter-
ranean diet and the omega-6/omega-3 balance, World Rev. Nutr. Diet. 102
(2011) 73e80.
[39] R. Przybylski, Canola/Rapeseed Oil, in Vegetable Oils in Food Technology,
Wiley-Blackwell, 2011, pp. 107e136.
[40] A. Baxheinrich, B. Stratmann, Y.H. Lee-Barkey, et al., Effects of a rapeseed oil-
enriched hypoenergetic diet with a high content of alpha-linolenic acid on
body weight and cardiovascular risk profile in patients with the metabolic
syndrome, Br. J. Nutr. 108 (4) (2012) 682e691.