Eur J Nutr (2016) 55:227–236
DOI 10.1007/s00394-015-0840-2
ORIGINAL CONTRIBUTION
Dietary total antioxidant capacity and mortality
in the PREDIMED study
P. Henríquez‑Sánchez · A. Sánchez‑Villegas · C. Ruano‑Rodríguez · A. Gea · R. M. Lamuela‑Raventós · R. Estruch ·
J. Salas‑Salvadó · M. I. Covas · D. Corella · H. Schröder · M. Gutiérrez‑Bedmar · J. M. Santos‑Lozano · X. Pintó ·
F. Arós · M. Fiol · A. Tresserra‑Rimbau · E. Ros · M. A. Martínez‑González · L. Serra‑Majem
Received: 4 November 2014 / Accepted: 18 January 2015 / Published online: 7 February 2015
© Springer-Verlag Berlin Heidelberg 2015
Abstract
Purpose The aim of the present study was to assess the
association between the dietary total antioxidant capacity,
the dietary intake of different antioxidants and mortality in
a Mediterranean population at high cardiovascular disease
risk.
Methods A total of 7,447 subjects from the PREDIMED
study (multicenter, parallel group, randomized controlled
clinical trial), were analyzed treating data as an observa-
tional cohort. Different antioxidant vitamin intake and
total dietary antioxidant capacity were calculated from a
On behalf of the PREDIMED Study Investigators.
P. Henríquez‑Sánchez (*) · A. Sánchez‑Villegas ·
C. Ruano‑Rodríguez · A. Gea · R. M. Lamuela‑Raventós ·
R. Estruch · J. Salas‑Salvadó · M. I. Covas · D. Corella ·
M. Gutiérrez‑Bedmar · J. M. Santos‑Lozano · X. Pintó ·
F. Arós · M. Fiol · A. Tresserra‑Rimbau · E. Ros ·
M. A. Martínez‑González · L. Serra‑Majem
Centro de Investigación Biomédica en Red de Fisiopatología
de la Obesidad y Nutrición, Instituto de Salud Carlos III,
Madrid, Spain
e-mail: patricia.henriquez@ulpgc.es
P. Henríquez‑Sánchez · A. Sánchez‑Villegas ·
C. Ruano‑Rodríguez · L. Serra‑Majem
Research Institute of Biomedical and Health Sciences,
University of Las Palmas de Gran Canaria,
Las Palmas de Gran Canaria, Spain
A. Gea · M. A. Martínez‑González
Department of Preventive Medicine and Public Health,
University of Navarra, Pamplona, Navarra, Spain
R. M. Lamuela‑Raventós · A. Tresserra‑Rimbau
Nutrition and Food Science Department, XaRTA,
INSA, School of Pharmacy, University of Barcelona,
Barcelona, Spain
validated 137-item food frequency questionnaire at base-
line and updated yearly. Deaths were ascertained through
contact with families and general practitioners, review of
medical records and consultation of the National Death
Index. Cox regression models were fitted to assess the
relationship between dietary total antioxidant capacity and
mortality. Dietary total antioxidant capacity was estimated
using ferric-reducing antioxidant power assays.
Results A total of 319 deaths were recorded after a
median follow-up of 4.3 years. Subjects belonging to the
upper quintile of antioxidant capacity were younger, ex-
smokers, with high educational level, and more active and
had higher alcohol intake. Multivariable-adjusted models
R. Estruch
Department of Internal Medicine, Institut d’Investigacions
Biomèdiques August Pi i Sunyer (IDIBAPS), University
of Barcelona, Barcelona, Spain
J. Salas‑Salvadó
Human Nutrition Unit, School of Medicine, University Rovira i
Virgili, Reus, Spain
M. I. Covas
Centro de Investigación Biomédica en Red de Fisiopatología de
la Obesidad y Nutrición, Instituto de Salud Carlos II, Madrid,
Spain
M. I. Covas
Cardiovascular Risk and Nutrition Research Unit, Institut
Municipal d’Investigació Mèdica (IMIM), Barcelona, Spain
D. Corella
Department of Preventive Medicine and Public Health, University
of Valencia, Valencia, Spain
H. Schröder
CIBER Epidemiología y Salud Pública (CIBERESP), Instituto de
Salud Carlos III, Madrid, Spain
13
228
revealed no statistically significant difference between total
dietary antioxidant capacity and mortality (Q5 vs. Q1 ref
HR 0.85; 95 % CI 0.60–1.20) neither for the intake of all
the vitamins studied.
Conclusions No statistically significant association was
found between antioxidant capacity and total mortality in
elderly subjects at high cardiovascular risk.
Keywords Dietary antioxidant capacity · Antioxidant
intake · Mortality · PREDIMED
Introduction
It has been widely reported that high consumption of fruits
and vegetables might positively impact human health [1–4].
The antioxidants present in these food items reduce the
adverse effects of both oxygenated and nitrogenous free
radicals that are produced in the normal physiological func-
tioning of the organism. But fruits and vegetables are not
the only sources of antioxidants in the diet. Other foods
like chocolate, coffee, tea, wine, beer and fresh herbs and
apices [5, 6] also have an important contribution in antioxi-
dant levels.
For years, multiple studies conducted predominantly
in middle-aged populations have showed associations
between antioxidant substances and all-cause or cardio-
vascular disease mortality. Evidence that mortality may be
associated with vitamin C, vitamin E, carotenoids and sele-
nium intake or status has been obtained in several previous
studies [7–11].
Given the diversity in the sources of antioxidants and
the interaction that exists between the different nutrients of
different food items that constitute our diet, it is important
to use an indicator that reflects adequately the daily expo-
sure to antioxidants and that takes into account the synergic
effect of all the antioxidant substances present in the food
H. Schröder
Cardiovascular Risk and Nutrition Research Group, Institut
Municipal d’Investigació Medica (IMIM)-Institut de Recerca del
Hospital del Mar, Barcelona, Spain
M. Gutiérrez‑Bedmar
Department of Preventive Medicine, University of Malaga,
Málaga, Spain
J. M. Santos‑Lozano
Department of Family Medicine, Primary Care Division
of Sevilla, San Pablo Health Center, Seville, Spain
X. Pintó
Lipids and Vascular Risk Unit, Internal Medicine, Hospital
Universitario de Bellvitge, Hospitalet de Llobregat, Barcelona,
Spain
13
Eur J Nutr (2016) 55:227–236
[12]. Dietary total antioxidant capacity measures the over-
all antioxidant capacity of diet and has been shown to be an
indicator of diet quality [13, 14]. Similar to individual anti-
oxidants, total dietary antioxidant capacity has also been
inversely associated with stroke [13, 15], cancer [16] and
heart failure [17].
The antioxidant capacity of the diet determines chronic
inflammatory states [18, 19]. Given the relationship
between the inflammation and some chronic diseases and
death, the aim of the present study was to assess the associ-
ation between the dietary total antioxidant capacity, the die-
tary intake of different antioxidant vitamins and mortality
in a Mediterranean population at high cardiovascular risk.
Materials and methods
Study design
This study was conducted within the frame of the PRED-
IMED study (PREvención con DIeta MEDiterránea): a
randomized, multicenter, parallel group, single-blinded
dietary intervention trial conducted in Spain with the aim
to analyze the effect of the Mediterranean Diet (MeDiet)
on the prevention of major cardiovascular disease. Partici-
pants were randomly assigned to three groups: MeDiet sup-
plemented with extra virgin olive oil (MeDiet + EVOO),
MeDiet with nuts (MeDiet + nuts) and the control group
[20]. The main results of the trial on the primary end point
have been published elsewhere [21].
The PREDIMED study included 7,447 participants, men
aged between 55 and 80 years and women between 60 and
80 years who were at high cardiovascular risk. Subjects,
who had no cardiovascular disease at enrollment, met at
least one of the following criteria: the presence of type 2
diabetes mellitus (T2DM) or the presence of three or more
cardiovascular risk factors (current smoking, hypertension,
F. Arós
Department of Cardiology, University Hospital of Alava, Vitoria,
Spain
M. Fiol
Institute of Health Sciences, University of Balearic Islands,
Palma de Mallorca, Spain
M. Fiol
Hospital Son Espases, Palma de Mallorca, Spain
E. Ros
Lipid Clinic, Endocrinology and Nutrition Service, Institut
d’Investigacions Biomèdiques August Pi Sunyer, Hospital Clinic,
Barcelona, Spain
Eur J Nutr (2016) 55:227–236
dyslipidemia, overweight or obesity, or family history of
premature CVD).
The present analysis was conducted as an observational
prospective cohort study considering baseline data before
the individuals were randomized to the intervention group.
The protocol was approved by the institutional review
boards of each recruitment center and all participants pro-
vided a written informed consent prior to their inclusion in
the study.
For this analysis, we excluded subjects without follow-
up (n = 201), and with values of total energy intake outside
of predefined limits (<800 or >4,000 kcal/day in men and:
<500 or >3,500 kcal/day in women) (n = 231) [22]. Over-
all, 7,015 subjects were analyzed in this study.
Antioxidant capacity intake and dietary assessment
Trained dieticians used a 137-item food frequency ques-
tionnaire (FFQ) to assess dietary habits by face-to-face
interviews. This FFQ was repeatedly administered at base-
line and each year during follow-up. The FFQ has been pre-
viously validated in a sample of participants of the PRED-
IMED study [23, 24]. Energy (kcal/day) and nutrient intake
(g/day) were calculated as frequency multiplied by nutrient
concentration in a specified portion size, where frequencies
were measured in nine categories for each food item. Nutri-
ent data bank was updated using the latest available infor-
mation included in food composition tables for Spain [25].
Alcohol intake was also ascertained through the use of this
questionnaire.
Dietary total antioxidant capacity was estimated using
published databases that provided the antioxidant capacity
measured in foods by assays FRAP (ferric-reducing anti-
oxidant power) [6].
Outcome ascertainment
All-cause mortality was determined by review of the end-
point adjudication committee whose members were una-
ware of the intervention assignments or any dietary expo-
sure. Information on all-cause mortality was updated
yearly. The sources of information were initially obtained
from the continuous contact with participants and their
families that we had during the trial, contact with family
physicians, the yearly comprehensive review of all medical
records and by yearly consultation of the National Death
Index. The analyses included cases confirmed between
October 1, 2003 and December 1, 2010.
Covariates assessment
At baseline socio-demographic information, medical
history and use of medication and lifestyle habits were
229
collected via specific questionnaires. Trained personnel car-
ried out anthropometric and blood pressure measurements.
Height and weight were measured wearing light clothes,
barefoot, using a wall-mounted stadiometer and calibrated
scales. BMI was estimated as weight (kg) divided by the
height (m2) squared. Physical activity was assessed using
a validated Spanish version of the Minnesota leisure time
physical activity questionnaire [26].
Statistical analysis
Baseline characteristics of the population were summarized
according to quintiles of dietary total antioxidant capacity.
Follow-up time was calculated from the date of recruitment
to the date of either death or end of follow-up (the date of
the last visit or the last recorded clinical event of partici-
pants still alive). To assess the risk of all-cause mortality
by quintiles of antioxidant intake and quintiles of dietary
antioxidant capacity, different Cox regression models were
used to estimate multivariable-adjusted hazard ratios (HR)
and their 95 % confidence intervals.
We used Cox regression models with time-dependent
exposures with updated diet and covariates. For dietary
measures, we used the cumulative average of food intakes
from baseline to the censoring events.
To ascertain the association between the quintiles of
dietary total antioxidant capacity and total mortality, a first
model was adjusted for recruitment center, intervention
group, age (years, continuous), sex, education (low, middle,
high), marital status (married, other), BMI (underweight,
normal, overweight and obese), smoking habit (never, past,
current), leisure time physical activity (METs-min/day
score, continuous), alcohol intake (g/day, continuous), total
energy intake (kcal/day, continuous), self-reported history
of cancer, arterial hypertension, history of dyslipidemia and
cardiovascular disease (yes, no). A second model was addi-
tionally adjusted for energy-adjusted intake of saturated
fatty acids (g/day), polyunsaturated fatty acids (g/day),
monounsaturated fatty acids (g/day) and glycemic index.
Tests of linear trend across increasing quintiles of dietary
total antioxidant capacity were conducted by assigning the
medians to each quintiles, and these variables were treated
as continuous in the multivariable models.
Moreover, to assess the presence of interactions between
the dietary total antioxidant capacity and different variables
such as intervention group, sex, obesity or smoking at base-
line, product terms were built and included in the multivari-
able models.
To estimate multivariable-adjusted HR and their 95 %
CI for all-cause mortality according quintiles of each anti-
oxidant intake, the model was adjusted for the same vari-
ables used to adjust the second model in the dietary anti-
oxidant capacity index analysis.
13
230 Eur J Nutr (2016) 55:227–236

Table 1  Baseline characteristics of participants according to quintiles of dietary total antioxidant capacity
Q1 Q2 Q3 Q4 Q5 p value
<12.7 12.7–16.7 16.7–20.7 20.7–27.9 >27.9

Non-dietary factors
Age (years)a 67.7 (6.2) 67.3 (6.2) 67.5 (6.0) 66.5(6.2) 66.3 (6.1) <0.001‡
Sex (woman) (%) 56.76 60.8 65.9 51.5 52.8 <0.001†
Married (%) 77.3 77.0 73.0 77.9 76.1 0.019†
Educational level (%) <0.001†
  Elementary 79.5 79.3 77.3 74.4 68.9
  Secondary 12.70 13.4 14.5 17.3 18.2
  University 6.1 5.4 5.6 7.0 11.3
Body mass index (kg/m2)a 29.9 (3.8) 30.1 (3.9) 30.0 (3.9) 30.0 (3.8) 30.0 (3.9) 0.611‡
Physical activity during leisure time 216.8 (228.8) 222.6 (231.3) 225.7 (228.6) 244.0 (256.8) 250.2 (251.6) 0.006‡
(MET-min/day)a
Alcohol intake (g/day)a 5.7 (11.0) 6.7 (11.6) 7.9 (13.8) 10.1 (16.2) 11.3 (16.5) <0.001‡
Total energy intake (kcal/day)a 2,299 (598) 2,228 (466) 2,051 (534) 2,370 (540) 2,235 (515) <0.001‡
Smoking status (%) <0.001†
  Current smoker 10.9 11.5 9.9 18.9 18.4
  Past smoker 21.7 23.6 22.0 26.4 29.3
Dyslipidemia (%) 68.3 71.2 76.7 73.3 71.0 0.001†
Diabetes (%) 49.3 50.4 46.8 48.3 49.7 0.342†
Hypertension (%) 85.1 82.8 87.2 80.8 77.2 <0.001†
Coronary heart disease (%) 7.9 8.1 7.4 8.3 8.7 0.784†
Food group intakea
Vegetables (g/day) 321 (139) 320 (138) 344 (153) 340 (158) 345 (145) <0.001‡
Fruits (g/day) 340 (184) 339 (178) 372 (192) 372 (218) 412/222) <0.001‡
Nuts (g/day) 8 (11) 8 (12) 10 (13) 12 (16) 13 (16) <0.001‡
Olive oil (g/day) 41 (19) 40 (18) 37 (17) 40 (18) 37 (17) <0.001‡
Red wine (ml/day) 35 (74) 44 (84) 57 (104) 69 (126) 82 (129) <0.001‡
Fish (g/day) 97 (48) 99 (59) 98 (48) 100 (46) 103 (51) 0.050‡
Meat/meat products (g/day) 136 (61) 133 (54) 122 (51) 135 (59) 129 (56) <0.001‡
Milk/milk products (g/day) 372 (224) 363 (199) 345 (194) 410 (234) 421 (239) <0.001‡
Legumes (g/day) 21 (13) 20 (11) 20 (14) 21 (15) 21 (13) 0.034‡
Cereal foods (g/day) 164 (99) 150 (80) 124 (74) 147 (88) 127 (70) <0.001‡
#
Adjusted for energy
‡
p from ANOVA
†
p from χ2 test
a
Means (SD)

Statistical analyses were performed using SPSS soft-
ware package for Windows version 19.0 (SPSS Inc., Chi-
cago, IL, USA) and Stata 12.0 (StataCorp, College Station,
TX, USA). The significance level was set at p < 0.05.
Results
The subjects in the highest quintile of dietary total antioxi-
dant capacity were more likely to be younger, ex-smokers,
physically more active, with secondary or university level
of education and with higher alcohol intake. In contrast,
subjects in the highest quintile of adherence to the anti-
oxidant capacity index were less likely to have a history
of hypertension. The antioxidant capacity index was also
directly associated with alcohol and red wine consumption,
and fruit and cereal intake (Table 1).
During follow-up (median 4.3 years), 319 deaths were
reported. Table 2 shows the association between the quin-
tiles of antioxidant capacity index and mortality risk. There
was an inverse association with total mortality (multi-
variable HR 0.85; 95 % CI 0.60–1.20) and with mortality
from cardiovascular disease (multivariable HR 0.79; 95 %
CI 0.43–1.43), for participants in the highest quintile of

13
 Table 2  Hazard ratios and 95 % confidence intervals for the association between quintiles of dietary total antioxidant capacity and mortality
Eur J Nutr (2016) 55:227–236

Antioxidant capacity (FRAP) Q1 Q2 Q3 Q4 Q5 p for trend FRAP continuous

(cumulative average) <12.7 12.7–16.7 16.7–20.7 20.7–27.9 >27.9

Person-year of follow-up 6,191 6,358 6,087 6,308 6,131

Death from any cause
Number of cases 69 74 52 63 61
Crude rate/1,000 person-years 11.15 11.64 8.54 9.99 9.95
Adjusted HR (95 % IC)a 1 (ref) 1.05 (0.76–1.45) 0.75 (0.53–1.07) 0.68 (0.47–0.98) 0.85 (0.60–1.20) 0.110 0.98 (0.97–1.00)
Adjusted HR (95 % IC)b 1 (ref) 1.07 (0.78–1.47) 0.79 (0.55–1.13) 0.71 (0.49–1.03) 0.88 (0.62–1.26) 0.189 0.99 (0.97–1.00)
Death from cardiovascular disease
Number of cases 22 21 20 20 19
Crude rate/1,000 person-years 3.55 3.30 3.29 3.31 3.10
Adjusted HR (95 % IC)a 1 (ref) 0.85 (0.47–1.54) 0.70 (0.36–1.36) 0.76 (0.42–1.37) 0.79 (0.43–1.43) 0.374 1.00 (0.97–1.02)
Adjusted HR (95 % IC)b 1 (ref) 0.84 (0.46–1.53) 0.74 (0.38–1.46) 0.80 (0.44–1.45) 0.80 (0.44–1.47) 0.444 1.00 (0.97–1.02)
Death from cancer
Number of cases 27 44 22 31 42
Crude rate/1,000 person-years 4.36 6.92 3.60 4.90 6.85
Adjusted HR (95 % IC)a 1 (ref) 1.08 (0.67–1.75) 0.99 (0.60–1.65) 0.77 (0.45–1.34) 1.33 (0.80–2.22) 0.470 1.01 (0.99–1.03)
Adjusted HR (95 % IC)b 1 (ref) 1.10 (0.68–1.78) 1.02 (0.61–1.71) 0.80 (0.46–1.39) 1.33 (0.79–2.23) 0.490 1.01 (0.98–1.03)

Adjusted for energy

FRAP ferric-reducing antioxidant power
a
Adjusted for recruitment center, intervention group, age (years, continuous), sex, education (low, middle, high), marital status (married/other), BMI (underweight, normal, overweight and
obese), smoking habit (never smoker, ex-smoker, current smoker), alcohol consumption (g/year, continuous), total energy intake (kcal/day, continuous), and medical history of hypertension,
diabetes, dyslipidemia and cancer (yes/no)
b
Additionally adjusted for energy-adjusted intake of saturated fatty acids, polyunsaturated fatty acids, monounsaturated fatty acids and glycemic index

13
231
232 Eur J Nutr (2016) 55:227–236

Table 3  Hazard ratios and 95 % confidence intervals for the association between quintiles single antioxidant intake and mortalitya
Q1 Q2 Q3 Q4 Q5 p for trend

Vitamin C
Person-years of follow-up 6,522 6,418 6,150 6,070 5,915
Death from any cause 87 76 66 46 44
Adjusted HR (95 % IC)a 1 (ref) 0.94 (0.68–1.31) 1.03 (0.73–1.46) 0.82 (0.56–1.21) 0.78 (0.51–1.18) 0.173
Vitamin E
Person-years of follow-up 6,264 6,382 6,161 6,103 6,165
Death from any cause 81 71 54 55 58
Adjusted HR (95 % IC)a 1 (ref) 0.86 (0.61–1.22) 1.01 (0.70–1.45) 0.85 (0.57–1.28) 1.42 (0.95–2.13) 0.113
Vitamin D
Person-years of follow-up 6,302 6,305 6,139 6,066 6,264
Death from any cause 71 75 62 53 58
Adjusted HR (95 % IC)a 1 (ref) 0.94 (0.68–1.30) 0.74 (0.52–1.06) 1.00 (0.70–1.42) 1.01 (0.69–1.47) 0.763
β-Carotene
Person-years of follow-up 6,539 6,416 6,130 6,120 5,871
Death from any cause 80 70 67 52 50
Adjusted HR (95 % IC)a 1 (ref) 1.04 (0.76–1.44) 1.00 (0.70–1.43) 0.74 (0.50–1.10) 0.82 (0.56–1.21) 0.148
α-Carotene
Person-years of follow-up 6,419 6,529 6,224 6,015 5,888
Death from any cause 83 69 53 58 56
Adjusted HR (95 % IC)a 1 (ref) 0.85 (0.62–1.17) 0.65 (0.45–0.93) 0.68 (0.46–1.00) 0.86 (0.60–1.24) 0.441
β-Cryptoxanthin
Person-years of follow-up 6,330 6,363 6,292 6,155 5,936
Death from any cause 81 71 59 60 48
Adjusted HR (95 % IC)a 1 (ref) 0.74 (0.53–1.03) 0.83 (0.60–1.16) 0.72 (0.50–1.04) 0.70 (0.48–1.02) 0.099
Lutein
Person-years of follow-up 6,221 6,449 6,362 6,100 5,943
Death from any cause 78 65 79 50 47
Adjusted HR (95 % IC)a 1 (ref) 0.92 (0.66–1.28) 1.05 (0.74–1.47) 0.72 (0.48–1.07) 0.84 (0.58–1.23) 0.317
Zeaxanthin
Person-years of follow-up 6,397 6,300 6,215 6,106 6,058
Death from any cause 72 79 47 73 48
Adjusted HR (95 % IC)a 1 (ref) 0.79 (0.56–1.12) 0.78 (0.54–1.11) 0.78 (0.54–1.13) 1.03 (0.72–1.47) 0.947
Lycopene
Person-years of follow-up 6,661 6,084 6,212 6,092 6,027
Death from any cause 72 81 61 56 49
Adjusted HR (95 % IC)a 1 (ref) 1.23 (0.90–1.69) 0.76 (0.52–1.11) 1.21 (0.83–1.76) 0.91 (0.61–1.37) 0.727
Phytosterols
Person-years of follow-up 6,455 6,411 6,131 6,105 5,974
Death from any cause 93 64 57 46 59
Adjusted HR (95 % IC)a 1 (ref) 0.73 (0.52–1.03) 0.91 (0.65–1.29) 0.55 (0.37–0.82) 0.73 (0.47–1.11) 0.102
Selenium
Person-years of follow-up 6,070 6,190 6,210 6,333 6,273
Death from any cause 73 73 62 53 58
Adjusted HR (95 % IC)a 1 (ref) 0.75 (0.53–1.05) 0.88 (0.62–1.25) 0.72 (0.49–1.06) 0.74 (0.49–1.10) 0.212
Zinc
Person-years of follow-up 6,299 6,261 6,128 6,139 6,249
Death from any cause 68 61 80 49 61
Adjusted HR (95 % IC)a 1 (ref) 0.70 (0.50–0.99) 0.75 (0.53–1.08) 0.80 (0.56–1.14) 0.93 (0.66–1.32) 0.935

a
Adjusted for recruitment center, intervention group, age (years, continuous), sex, education (low, middle, high), marital status (married/other),
BMI (underweight, normal, overweight and obese), smoking habit (never smoker, ex-smoker, current smoker), alcohol consumption (g/year,
continuous), total energy intake (kcal/day, continuous), energy-adjusted intake of saturated fatty acids (g/day), polyunsaturated fatty acids (g/
day), monounsaturated fatty acids (g/day) and glycemic index and medical history of hypertension, diabetes, dyslipidemia and cancer (yes/no)

13
Eur J Nutr (2016) 55:227–236
antioxidant capacity index compared with participants in
the lowest quintile, although these associations were not
significant. Neither was significant the association for anti-
oxidant capacity index and death from cancer. Additional
adjustment for several dietary factors such as fatty acids
and glycemic index did not change the reported results.
After repeating the analyses including several prod-
uct terms in the multivariable models, only a significant
interaction was found for dietary total antioxidant capac-
ity and smoking at baseline for cancer mortality (p for
interaction = 0.046). Thus, the analysis for this outcome
was re-analyzed after stratifying the sample by smoking
status (never/former vs. current smokers). No significant
association between dietary total antioxidant capacity
and cancer mortality was found among never and former
smokers. However, an important reduction (around 80 %)
in the risk of cancer death was reported for those partici-
pants who were current smokers and belonged to the third
and fourth quintiles of dietary total antioxidant capacity as
compared to those within the first category (reference). HR
and 95 % CI were 0.18 (0.04–0.83) and 0.23 (0.07–0.79),
respectively.
In the analysis of the effect of several single antioxidants
and mortality (Table 3), no significant association was
found between vitamin intake and mortality risk. Although
the estimates did not reach statistical significance, other
antioxidants such as vitamin C, β-cryptoxanthin, phytoster-
ols or selenium were inversely associated with the risk of
death from any cause with reductions of around 20–30 %.
HR and 95 % CI for extreme quintiles of vitamin C intake
were 0.78 (0.51–1.18); HR and 95 % CI for extreme quin-
tiles of β-cryptoxanthin intake were 0.70 (0.48–1.02); HR
and 95 % CI for extreme quintiles of phytosterols intake
were 0.73 (0.47–1.11); and HR and 95 % CI for extreme
quintiles of selenium intake were 0.74 (0.49–1.10).
Discussion
In this large prospective study in a population of older sub-
jects with cardiovascular risk factors, we found no associa-
tion between dietary total antioxidant capacity of the diet
and mortality. Only subjects in the fourth quintile of die-
tary total antioxidant capacity had a lower risk of all-cause
mortality when compared to subjects belonging to the first
quintile. These differences disappeared when addition-
ally the results were adjusted for fatty acids and glycemic
index intake. Also subjects in the fourth quintile had a 32 %
lower risk of cancer mortality.
The results did not change after carrying out sev-
eral sensitivity analyses (data not shown), excluding the
deaths occurring in the first 2 years of follow-up or partici-
pants with less than 2 years of follow-up, including only
233
participants aged with less than 60, 65 or 70 years at base-
line or including only deaths occurring <60 or <70 years.
The beneficial effect that antioxidants have on health is
well known. However, the action of individual antioxidants
cannot be entirely determined due to the complex syner-
gistic interactions present among these substances. Total
antioxidant capacity has been recognized as a new tool to
investigate the relationship between dietary antioxidant
intake and mortality/diseases, providing the antioxidant
potential of the total diet rather than that of single nutri-
ents [12]. The antioxidant capacity of the diet is considered
a good indicator of plasma antioxidant status [27–29] and
also an indicator of nutritional quality [14].
Several cohort studies have investigated the protec-
tive effect of dietary total antioxidant capacity on the risk
of some diseases. However, we have not found any study
that related dietary total antioxidant capacity with mortal-
ity in elderly population. In the EPIC cohort, with a large
number of subjects form different European countries
and aged between 35 and 70 years, results showed that a
high dietary intake of antioxidant capacity was associated
with a reduced risk of gastric cancer [16]. A case–control
study conducted in an Italian population found a consistent
inverse relation between this index and colorectal cancer
[30].
Regarding cardiovascular disease, the results by Rauti-
ainen et al. in the Swedish Mammography Cohort showed
that the total antioxidant capacity of diet was associated
with lower risk myocardial infarction [31] and heart fail-
ure [17]. In this cohort of Swedish women, researches also
found an inverse association between total antioxidant
capacity of diet and hemorrhagic stroke but not with cer-
ebral infarction [13]. On the contrary, Del Rio et al. [15] in
the EPIC cohort suggested that a diet rich in total antioxi-
dant capacity of diet reduces the risk of cerebral infarction
but not the risk of hemorrhagic stroke.
In the Rotterdam study, a prospective cohort that
included 5,395 participants aged 55 years and older, dietary
total antioxidant capacity of the diet did not seem to predict
the risk of major neurologic diseases [32]. Other cohort
studies have shown an inverse association between dietary
total antioxidant capacity, diabetes [33] and metabolic syn-
drome [34].
When the different dietary antioxidants were analyzed
separately, results showed that subjects in the highest
quintile had a lower mortality risk than those belonging to
the lowest intake for most of the antioxidant substances,
but these differences were not significant. This finding is
in concordance with the results of other studies also con-
ducted in elderly people and where the evidence of the
association between mortality and antioxidant vitamin
intake is scanty [8, 35, 36]. However, in a previous analysis
conducted also in the frame of the PREDIMED study [37],
13
234
results showed an inverse relationship between polyphenol
intake and mortality.
More evidence exists on the beneficial effect that the
intake of fruits and vegetables with high antioxidant power
has in the prevention of certain chronic diseases and in all-
cause mortality [1, 2, 38].
Though the epidemiological evidence shows that the
consumption of food items rich in antioxidants is related to
a lower risk of mortality and incidence of chronic diseases,
there is no clear evidence that the supplementation with
high doses of antioxidant vitamins is effective in the reduc-
tion in the risk of this type of pathologies. On the other
hand, it has been suggested that the beneficial effect of
dietary antioxidants is higher in subjects having some risk
factors at baseline [39]. In our population, the possibility
of oxidative stress is high though this is an elderly popula-
tion at high cardiovascular risk. Otherwise, the population
studied in this analysis belongs to a dietary intervention
trial in which three groups of intervention are established:
a Mediterranean diet supplemented with olive oil group, a
Mediterranean diet supplemented with mixed nuts group
and a control group without dietetic supplementation who
received advice to follow a low-fat diet.
Nevertheless, as a possible explanation of our findings,
it is probable that dietary antioxidant levels in this popula-
tion were adequate from the beginning of the study though
participants from the PREDIMED are diabetics or have at
least three risk factors for cardiovascular disease. So, these
conditions preclude the adoption of dietetic changes includ-
ing the increase in the antioxidant intake.
The increasing interest for the antioxidant substances
has risen also the development of methods to estimate the
antioxidant value of the diet. Given the synergistic interac-
tions between the different components of foods and the
different mechanisms of action of each one, it is more ade-
quate to use a combination of them to evaluate with more
accuracy the antioxidant capacity of the food [12]. The use
of only this method to estimate the antioxidant capacity in
our population could be a limitation of our study.
In fact, other specific assays such as ORAC or ABTS
could complete the dietary antioxidant capacity assessment
[40].
Another fact to take into account is the influence that
the dietary antioxidant capacity has on the total antioxidant
capacity. In spite of the strong consistence existing of the
direct role of dietary antioxidants in modulating the plasma
antioxidant network, it is necessary to consider the influ-
ence of multiple factors that can modify the bioavailability
of the antioxidants and therefore their activity in vivo [41].
Moreover, non-differential measurement error in the
assessment of dietary total antioxidant capacity may have
biased our results. This bias is particularly important for
this non-binary exposure [42]. The estimation of dietary
13
Eur J Nutr (2016) 55:227–236
total antioxidant capacity was derived through a food fre-
quency questionnaire that included a limited number of
food items. Thus, although the validity and reproducibility
of the dietary questionnaire have been previously evalu-
ated, the antioxidant capacity of the diet could have been
underestimated. Additionally, multiple factors including
cultivation procedures, growing and harvesting conditions
could influence the antioxidant capacity of the diet.
Some strengths of our study also deserve to be men-
tioned. They included its prospective design, its long-term
follow-up and the large simple size; thus, it is not repre-
sentative of general population. Also it is of great inter-
est the use of repeated measurements for calculating the
dietary intake which could be used to reduce measurement
error due to intra-individual variation.
In conclusion, the results of this study in older people
at high cardiovascular risk did not show any association
between dietary antioxidant capacity and the risk of total
mortality, cardiovascular mortality or cancer mortality. Fur-
ther studies that combined this index with other indicators
of antioxidant capacity are needed to analyze more accu-
rately the relationship between the adherence to a diet rich
in antioxidants and the occurrence of disease or death.
Acknowledgments The authors want to thank the participants of
the study for their collaboration and the PREDIMED personnel for
their excellent assistance with all aspects of the trial. This study was
funded by the Spanish Ministry of Health (ISCIII), CIBEROBN,
PI1001407, G03/140, RD06/0045 and the Autonomous Govern-
ment of Catalonia, and Caixa Tarragona (10-1343). The Fundación
Patrimonio Comunal Olivarero and Hojiblanca SA (Málaga, Spain),
California Walnut Commission (Sacramento, CA), Borges SA (Reus,
Spain) and Morella Nuts SA (Reus, Spain), donated the olive oil,
walnuts, almonds and hazelnuts, respectively. None of the funding
sources played a role in the design, collection, analysis, interpretation
or publication of the data. AG is supported by a FPU fellowship from
the Spanish Government.
Conflict of interest The authors declare that they have no conflict
of interest.
References
1. Leenders M, Sluijs I, Ros MM, Boshuizen HC, Siersema PD,
Ferrari P, Weikert C, Tjønneland A, Olsen A, Boutron-Ruault
MC, Clavel-Chapelon F, Nailler L, Teucher B, Li K, Boeing H,
Bergmann MM, Trichopoulou A, Lagiou P, Trichopoulos D, Palli
D, Pala V, Panico S, Tumino R, Sacerdote C, Peeters PH, van
Gils CH, Lund E, Engeset D, Redondo ML, Agudo A, Sánchez
MJ, Navarro C, Ardanaz E, Sonestedt E, Ericson U, Nilsson LM,
Khaw KT, Wareham NJ, Key TJ, Crowe FL, Romieu I, Gunter
MJ, Gallo V, Overvad K, Riboli E, Bueno-de-Mesquita HB
(2013) Fruit and vegetable consumption and mortality: European
prospective investigation into cancer and nutrition. Am J Epide-
miol 178:590–602
2. Agudo A, Cabrera L, Amiano P, Ardanaz E, Barricarte A, Beren-
guer T, Chirlaque MD, Dorronsoro M, Jakszyn P, Larrañaga N,
Martínez C, Navarro C, Quirós JR, Sánchez MJ, Tormo MJ,
Eur J Nutr (2016) 55:227–236
González CA (2007) Fruit and vegetable intakes, dietary anti-
oxidant nutrients, and total mortality in Spanish adults: findings
from the Spanish cohort of the European Prospective Investiga-
tion into Cancer and Nutrition (EPIC-Spain). Am J Clin Nutr
85:1634–1642
3. Dauchet L, Amouyel P, Hercberg S, Dallongeville J (2006) Fruit
and vegetable consumption and risk of coronary heart disease: a
meta-analysis of cohort studies. J Nutr 136:2588–2593
4. He FJ, Nowson CA, MacGregor GA (2006) Fruit and vegetable
consumption and stroke: meta-analysis of cohort studies. Lancet
367:320–326
5. Saura-Calixto F, Goñi I (2006) Antioxidant capacity of the Span-
ish Mediterranean diet. Food Chem 94:442–447
6. Pellegrini N, Serafini M, Colombi B, Del Rio S, Salvatore S,
Bianchi M, Brighenti F (2003) Total antioxidant capacity of plant
foods, beverages and oils consumed in Italy assessed by three dif-
ferent in vitro assays. J Nutr 133:2812–2819
7. Serafini M, Del Rio D (2004) Understanding the association
between dietary antioxidants, redox status and disease: is the total
antioxidant capacity the right tool? Redox Rep 9:145–152
8. Rautiainen S, Larsson S, Virtamo J, Wolk A (2012) Total antioxi-
dant capacity of diet and risk of stroke: a population-based pro-
spective cohort of women. Stroke 43:335–340
9. Puchau B, Zulet MA, de Echávarri AG, Hermsdorff HH, Mar-
tínez JA (2009) Dietary total antioxidant capacity: a novel indi-
cator of diet quality in healthy young adults. J Am Coll Nutr
28:648–656
10. Bates CJ, Hamer M, Mishra GD (2011) Redox-modulatory vita-
mins and minerals that prospectively predict mortality in older
British people: the national diet and nutrition survey of people
aged 65 years and over. Br J Nutr 105:123–132
11. Fletcher AE, Breeze E, Shetty PS (2003) Antioxidant vitamins
and mortality in older persons: findings from the nutrition add-on
study to the Medical Research Council Trial of Assessment and
Management of Older People in the Community. Am J Clin Nutr
78:999–1010
12. Buijsse B, Feskens EJ, Schlettwein-Gsell D, Ferry M, Kok FJ,
Kromhout D, de Groot LC (2005) Plasma carotene and alpha-
tocopherol in relation to 10-year all-cause and cause-specific
mortality in European elderly: the Survey in Europe on Nutrition
and the Elderly, a Concerted Action (SENECA). Am J Clin Nutr
82:879–886
13. Ray AL, Semba RD, Walston J, Ferrucci L, Cappola AR, Ricks
MO, Xue QL, Fried LP (2006) Low serum selenium and total
carotenoids predict mortality among older women living in
the community: the women’s health and aging studies. J Nutr
136:172–176
14. Khaw KT, Bingham S, Welch A, Luben R, Wareham N, Oakes S,
Day N (2001) Relation between plasma ascorbic acid and mor-
tality in men and women in EPIC-Norfolk prospective study: a
prospective population study. European Prospective Investigation
into Cancer and Nutrition. Lancet 357:657–663
15. Del Rio D, Agnoli C, Pellegrini N, Krogh V, Brighenti F, Mazzeo
T, Masala G, Bendinelli B, Berrino F, Sieri S, Tumino R, Rollo
PC, Gallo V, Sacerdote C, Mattiello A, Chiodini P, Panico S
(2011) Total antioxidant capacity of the diet is associated with
lower risk of ischemic stroke in a large Italian cohort. J Nutr
141:118–123
16. Serafini M, Jakszyn P, Luján-Barroso L, Agudo A, Bas Bueno-
de-Mesquita H, van Duijnhoven FJ, Jenab M, Navarro C, Palli
D, Boeing H, Wallström P, Regnér S, Numans ME, Carneiro F,
Boutron-Ruault MC, Clavel-Chapelon F, Morois S, Grioni S,
Panico S, Tumino R, Sacerdote C, Ramon Quirós J, Molina-
Montes E, Huerta Castaño JM, Barricarte A, Amiano P, Khaw
KT, Wareham N, Allen NE, Key TJ, Jeurnink SM, Peeters PH,
Bamia C, Valanou E, Trichopoulou A, Kaaks R, Lukanova A,
235
Bergmann MM, Lindkvist B, Stenling R, Johansson I, Dahm CC,
Overvad K, Jensen M, Olsen A, Tjonneland A, Lund E, Rinaldi
S, Michaud D, Mouw T, Riboli E, González CA (2012) Dietary
total antioxidant capacity and gastric cancer risk in the European
prospective investigation into cancer and nutrition study. Int J
Cancer 131:E544–E554
17. Rautiainen S, Levitan EB, Mittlema MA, Wolk A (2013) Total
antioxidant capacity of diet and risk of heart failure: a popula-
tion-based prospective cohort of women. Am J Med 126:494–500
18. Brighenti F, Valtueña S, Pellegrini N, Ardigò D, Del Rio D, Sal-
vatore S, Piatti P, Serafini M, Zavaroni I (2005) Total antioxi-
dant capacity of the diet is inversely and independently related
to plasma concentration of high-sensitivity C-reactive protein in
adult Italian subjects. Br J Nutr 93:619–625
19. Valtueña S, Pellegrini N, Franzini L, Bianchi MA, Ardido D,
del Rio D, Piatti P, Scazzina F, Zavaroni I, Brighenti F (2008)
Food selection based on total antioxidant capacity can modify
antioxidant intake, systemic inflammation, and liver function
without altering markers of oxidative stress. Am J Clin Nutr
87:1290–1297
20. Martínez-González MÁ, Corella D, Salas-Salvadó J, Ros E,
Covas MI, Fiol M, Wärnberg J, Arós F, Ruíz-Gutiérrez V, Lam-
uela-Raventós RM, Lapetra J, Muñoz MÁ, Martínez JA, Sáez
G, Serra-Majem L, Pintó X, Mitjavila MT, Tur JA, Portillo MP,
Estruch R, PREDIMED Study Investigators (2012) Cohort pro-
file: design and methods of the PREDIMED study. Int J Epide-
miol 41:377–385
21. Estruch R, Ros E, Salas-Salvadó J, Covas MI, Corella D, Arós F,
Gómez-Gracia E, Ruiz-Gutiérrez V, Fiol M, Lapetra J, Lamuela-
Raventos RM, Serra-Majem L, Pintó X, Basora J, Muñoz MA,
Sorlí JV, Martínez JA, Martínez-González MA, PREDIMED
Study Investigators (2013) Primary prevention of cardiovascular
disease with a Mediterranean diet. N Engl J Med 368:1279–1290
22. Willet W (1998) Issues in analysis and presentation of dietary
data. In: Willet W (ed) Nutritional epidemiology, 2nd edn. Oxford
University Press, Oxford, pp 321–346
23. Fernández-Ballart JD, Piñol JL, Zazpe I, Corella D, Carrasco P,
Toledo E, Perez-Bauer M, Martínez-González MA, Salas-Sal-
vadó J, Martín-Moreno JM (2010) Relative validity of a semi-
quantitative food-frequency questionnaire in an elderly Mediter-
ranean population of Spain. Br J Nutr 103:1808–1816
24. de la Fuente-Arrillaga C, Vázquez Z, Bes-Rastrollo M, Sampson
L, Martínez-González MA (2010) Reproducibility of an FFQ val-
idated in Spain. Public Health Nutr 13:1364–1372
25. Mataix L (2003) Tabla de composición de Alimentos Españoles.
Granada, Spain
26. Elosua R, Marrugat J, Molina P, Pons S, Pujol E (1994) Valida-
tion of the Minnesota leisure time physical activity questionnaire
in Spanish men. The MARATHOM investigators. Am J Epide-
miol 139:1197–1209
27. Wang Y, Yang M, Lee SG, Davis CG, Koo SI, Chun OK (2012)
Dietary total antioxidant capacity is associated with diet and
plasma antioxidant status in healthy young adults. J Acad Nutr
Diet 112:1626–1635
28. Yang M, Chung SJ, Floegel A, Song WO, Koo SI, Chun OK
(2013) Dietary antioxidant capacity is associated with improved
serum antioxidant status and decreased serum C-reactive pro-
tein and plasma homocysteine concentrations. Eur J Nutr
52:1901–1911
29. Valtueña S, Del Rio D, Pellegrini N, Ardigò D, Franzini L, Sal-
vatore S, Piatti PM, Riso P, Zavaroni I, Brighenti F (2007) The
total antioxidant capacity of the diet is an independent predictor
of plasma beta-carotene. Eur J Clin Nutr 61:69–76
30. La Vecchia C, Decarli A, Serafini M, Parpinel M, Bellocco
R, Galeone C, Bosetti C, Zucchetto A, Polesel J, Lagiou P,
Negri E, Rossi M (2013) Dietary total antioxidant capacity and
13
236
colorectal cancer: a large case–control study in Italy. Int J Cancer
133:1447–1451
31. Rautiainen S, Levitan EB, Orsini N, Åkesson A, Morgenstern R,
Mittleman MA, Wolk A (2012) Total antioxidant capacity from
diet and risk of myocardial infarction: a prospective cohort of
women. Am J Med 125:974–980
32. Devore EE, Feskens E, Ikram MA, den Heijer T, Vernooij M, van
der Lijn F, Hofman A, Niessen WJ, Breteler MM (2013) Total
antioxidant capacity of the diet and major neurologic outcomes in
older adults. Neurology 80:904–910
33. Okubo H, Syddall HE, Phillips DI, Sayer AA, Dennison EM,
Cooper C, Robinson SM, Hertfordshire Cohort Study Group
(2014) Dietary total antioxidant capacity is related to glucose
tolerance in older people: the Hertfordshire Cohort Study. Nutr
Metab Cardiovasc Dis 24:301–308
34. Bahadoran Z, Golzarand M, Mirmiran P, Shiva N, Azizi F (2012)
Dietary total antioxidant capacity and the occurrence of meta-
bolic syndrome and its components after a 3-year follow-up in
adults: Tehran lipid and glucose study. Nutr Metab (Lond) 9:70
35. Genkinger JM, Platz EA, Hoffman SC, Comstock GW, Helzl-
souer KJ (2004) Fruit, vegetable, and antioxidant intake and all-
cause, cancer, and cardiovascular disease mortality in a commu-
nity-dwelling population in Washington County, Maryland . Am J
Epidemiol 160:1223–1233
36. Roswall N, Olsen A, Christensen J, Hansen L, Dragsted LO,
Overvad K, Tjønneland A (2012) Micronutrient intake in relation
to all-cause mortality in a prospective Danish cohort. Food Nutr
Res 56:5466
13
Eur J Nutr (2016) 55:227–236
37. Tresserra-Rimbau A, Rimm EB, Medina-Remón A, Martínez-
González MA, López-Sabater MC, Covas MI, Corella D, Salas-
Salvadó J, Gómez-Gracia E, Lapetra J, Arós F, Fiol M, Ros E,
Serra-Majem L, Pintó X, Muñoz MA, Gea A, Ruiz-Gutiérrez V,
Estruch R, Lamuela-Raventós RM, PREDIMED Study Investiga-
tors (2014) Polyphenol intake and mortality risk: a re-analysis of
the PREDIMED trial. BMC Med 12:77
38. Hjartåker A1, Knudsen MD, Tretli S, Weiderpass E (2014) Con-
sumption of berries, fruits and vegetables and mortality among
10,000 Norwegian men followed for four decades. Eur J Nutr
[epub ahead of print]
39. Serafini M, Miglio C, Peluso I, Petrosino T (2011) Modulation
of plasma non enzimatic antioxidant capacity (NEAC) by plant
foods: the role of polyphenol. Curr Top Med Chem 11:1821–1846
40. Takashima M, Horie M, Shichiri M, Hagihara Y, Yoshida Y, Niki
E (2012) Assessment of antioxidant capacity for scavenging free
radicals in vitro: a rational basis and practical application. Free
Radic Biol Med 52(7):1242–1252
41. Frankel EN, Meyer AS (2000) Review: the problems of using
one-dimensional methods to evaluate multifunctional food and
biological antioxidants. J Sci Food Agric 80:1925–1941
42. Wacholder S, Hartge P, Lubin JH, Dosemeci M (1995) Non-dif-
ferential misclassification and bias towards the null: a clarifica-
tion. Occup Environ Med 52(8):557–558