Food and Chemical Toxicology 48 (2010) 29722979

Contents lists available at ScienceDirect

Food and Chemical Toxicology

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

Olive oil stability under deep-frying conditions

Susana Casal a, Ricardo Malheiro b, Artur Sendas a,b, Beatriz P.P. Oliveira a, Jos Alberto Pereira b,*
a
b

REQUIMTE/Servio de Bromatologia, Faculdade de Farmcia da Universidade do Porto, Rua Anbal Cunha 164, 4050-047 Porto, Portugal
CIMO/School of Agriculture, Polytechnic Institute of Bragana Apartado 1172, 5301-854 Bragana, Portugal

a r t i c l e

i n f o

Article history:
Received 4 May 2010
Accepted 26 July 2010

Keywords:
Olive oil
Oxidative stability
Frying
Vitamin E
Fatty acids
Total phenols

a b s t r a c t
The suitability of different commercial olive oil categories for domestic frying was investigated. Oil samples were taken every 3 h of frying and evaluated for free acidity, peroxide and p-anisidine values, specic
extinction coefcients, oxidative stability, fatty acids, vitamin E, b-carotene and total phenols, until the
total polar compounds achieved the maximum legal value (25%). All olive oils were fried during more
time than the commercial vegetable oil blend taken for comparison (from 24 to 27 h, against 15 h).
The extra-virgin Protected Designation of Origin (PDO) olive oil was characterized by reduced levels of
oxidation and hydrolysis, and superior amounts of minor antioxidant compounds. The olive oil commercial category behaves similarly, but Cobranosa olive oils performance was slightly worse, and
clearly different between years, highlighting the importance of blending different cultivars. The vegetable
oil, despite containing signicantly higher amounts of vitamin E, was highly susceptible to oxidation
under frying conditions when compared to all olive oils.
The results also show that the chemical composition of olive oils, particularly the amount of natural
antioxidants, are important parameters in their predictive behavior along the frying process, but mostly
that olive oil is clearly resistant to frying conditions, independently to the commercial category chosen.
2010 Elsevier Ltd. All rights reserved.

1. Introduction
Frying is one of the most popular culinary processes worldwide,
both for industrial and domestic food preparation procedures.
Fried products have unique organoleptic and sensorial properties,
including avor, texture, and appearance, which turn them largely
enjoyed by consumers. Also, this procedure considerably reduces
cooking time (Snchez-Muniz and Bastida, 2006) and is regarded
as inducing equal or even smaller nutrient losses when compared
with other common culinary processes (Fillion and Henry, 1998).
Frying can enhance the food nutritive value due to the simultaneous incorporation of important lipid components, namely vitamin E and essential fatty acids, providing that it is consumed
under a balanced diet, because the lipid uptake might increase signicantly the total daily energy intake.
The high temperatures used during frying, in the presence of
oxygen and water, induce important chemical changes of the oils,
namely by oxidation, polymerization, cyclization, and hydrolysis
(Paul and Mittal, 1997; Saguy and Dana, 2003), inevitably reducing their shelf life and affecting directly the quality of the nal
fried food (Kochhar, 2001). These chemical reactions are inuenced by the type and quality of the oil, the food properties,
and the food/oil ratio, among other parameters (Saguy and Dana,
* Corresponding author. Tel.: +351 273303277; fax: +351 273325405.
E-mail address: jpereira@ipb.pt (J.A. Pereira).
0278-6915/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.fct.2010.07.036

2003), altogether determining the frying oil performance (Andrikopoulos et al., 2002). Each vegetable oil is characterized by typical
stabilities against oxidation, dependent on the fatty acids composition, particularly the insaturation degree, and the content and
composition of minor compounds such as tocopherols (particularly c-tocopherol), certain sterols, hydrocarbons (squalene),
carotenoids, polyphenols, and trace metals (Hoffman, 1989; Kochhar, 2001).
Several studies support a relationship between the Mediterranean diet and a lower incidence of some important diseases of
our century, including cancer (Assmann et al., 1997) and cardiovascular diseases (Covas, 2007). Olive oil is typically the main lipid
source in the Mediterranean diet, being used as salad dressing
and for frying purposes. Its benecial properties are associated
with the richness in monounsaturated fatty acids, but other minor
compounds also take an important part (Varela and Ruiz-Roso,
2000). The differences to other common vegetable oils are enhanced by the fact that olive oil is mostly obtained by cold pressure
procedures, without being subjected to rening virgin olive oil
retaining therefore higher amounts of important bioactive components of the olive fruit. More recently, however, mostly supported
by economical reasons and reduced consumer information, the
consumption of olive oil, a blend of rened with virgin olive oils
in undeclared proportions, is increasing, together with other rened vegetable oils blends, including sunower, soybean, and high
oleic acid seed oils, particularly for frying purposes.

S. Casal et al. / Food and Chemical Toxicology 48 (2010) 29722979

The objective of the present work is to compare the stability under real domestic frying conditions of several olive oils categories.
The vegetable oil blend most frequently consumed in Portugal,
sunower based, was also included in the study. Aware that predictions based on the oxidative stability alone generally do not correlate with the real oxidation kinetics under frying conditions, and
in the presence of food (Velasco and Dobarganes, 2002), this study
was performed under real domestic frying conditions. The frying
media degradation was evaluated taking the actual law as rejection
point, particularly the total polar compounds, while complemented by several chemical evaluations in order to support and
understand the degradation patterns and potential nutritional
losses.

2. Materials and methods

All reagents were of analytical, chromatographic or spectroscopic grade and
were supplied from Merck (Darmstadt, Germany) or SigmaAldrich (St. Louis,
USA). All experiments and analytical determinations were performed at least in
duplicate.

2.1. Samples
Five oil samples were used throughout the frying studies. One commercial extra-virgin PDO olive oil Azeite de Trs-os-Montes from the northeast of Portugal
(EVOO), two monovarietal virgin olive oils from the same region (Cv. Cobranosa
from two consecutive years 2004 and 2005) (COB2004 and COB2005), a commercial blend of rened olive oil and virgin olive oil labelled as olive oil (ROO), and a
rened vegetable oil blend with large commercial signicance importance in Portugal, based mostly on sunower oil (SO). Pure rened olive oil, despite being mentioned as a commercial category in the EEC Regulation (2568/91), was not
included in the present study because its commercialisation is not legally recognized in Portugal (Dec-Lei 16/2004).
Oil samples were taken every 3 h of frying from each fryer (about 30 mL), with
immediate reposition with equal amounts of fresh oil, in order to avoid an early end
of the study due to low oil level. Each sample was kept under nitrogen atmosphere
until analysis.

2.2. Frying conditions

Domestic deep-fat electric fryers (Kenwood, DF 150) with 1.5 L capacity, designed for a maximum load of 400 g, were used for frying. Fresh potato slices
(300 g), shelled shortly before frying and cut into toothpicks, were used throughout
all frying sections. The fryers were regulated for 170 C, and a batch of potato slices
was fried every hour, during 9 h/day, until oil discard. The end of frying assays was
determined by the value of total polar compounds (max. 25%), in accordance to the
Portuguese law (Portaria No. 1135/95).

2.3. Physical and chemical parameters evaluated

The free acidity (FA), peroxide value (PV), coefcients of specic extinction at
232 nm and 270 nm (K232 and K270), and fatty acid composition were determined
according to ofcial methods describe in EEC Reg. No. 2568/91. The other parameters were either based on international standards or literature reports as briey described below.

2.3.1. Total polar compounds

Total polar compounds estimation was based on the dielectric constant
changes, measured directly on the hot oil, with a Test 265 cooking oil tester (Testo,
Germany), according to the manufacturer operation guide.

2.3.2. Free fatty acids

The free fatty acids (FFA), expressed as free oleic acid percentage, was determined by titration of an accurate sample solution, dissolved in ethanol/ether (1:1,
v/v), with 0.1 M sodium hydroxide solution, using phenolphthalein as indicator.

2.3.3. Peroxide value

The PV, expressed in milliequivalents of active oxygen per kilogram (mEq O2/
kg), was determined as follows: a mixture of oil and chloroform/acetic acid 2:3
(v/v) was left to react in the dark with saturated potassium iodine solution, and
the free iodine titrated with a sodium thiosulfate standard solution.

2973

2.3.4. Anisidine value

Unsaturated aldehydes, secondary oxidation products, were estimated by the
anisidine value (AV), as detailed in ISO 6885:2006. This measurement is based on
the absorbance increase per g of oil, measured at 350 nm (Shimadzu UV 160A spectrophotometer, Japan), of an olive oil solution in iso-octane, before and after reaction with p-anisidine reagent in the dark.
2.3.5. K232 and K270 extinction coefcients
The extinction coefcients K232 and K270 (absorption of 1% solution (m/v) in isooctane at 232 and 270 nm, respectively, with 1 cm of path length) were measured
using a UV spectrophotometer (Shimadzu UV 160A spectrophotometer, Japan).
2.3.6. Oxidative stability (Rancimat)
The oxidative stability was estimated by measuring the oxidation induction
time, on a Rancimat apparatus (Metrohm CH series 679). Air (20 L/h) was bubbled
through the oil (3.0 g) heated at 110 0.2 C, with the volatile compounds being
collected in water, and the increasing water conductivity continually measured.
The time taken to reach the conductivity inection was recorded.
2.3.7. Fatty acids composition
Fatty acids were evaluated as their methyl esters after alkaline trans-esterication with cold methanolic solution of potassium hydroxide, according with Regulation EEC 2568/91. The methyl esters were analyzed on a Chrompack CP 9001 gas
chromatograph equipped with a split-splitless injector, a FID detector, an autosampler Chompack CP-9050 and a 50 m  0.25 lm i.d. fused silica capillary column
coated with a 0.19 l lm of CP-Sil 88 (Chrompack). Helium was used as carrier
gas at an internal pressure of 120 kPa. The temperatures of the detector and injector
were 250 C and 230 C, respectively. The split ratio was 1:50 and the injected volume 1 lL. The results are expressed in relative percentage of each fatty acid methyl
ester, calculated by internal normalization of the chromatographic peak areas. A
mixture of fatty acid methyl esters (Supelco 37 FAME Mix) was used for identication purposes (Sigma, Spain).
2.3.8. Tocopherol and tocotrienol compositions
Tocopherols and tocotrienol measurements were based on the ISO 9936:2006
standard with the addition of tocol as internal standard. Briey, an accurate solution of oil in hexane, with tocol, was analyzed by HPLC on a silica gel column (Varian, Inertsil 5 Si, 250 mm, 3 mm i.d.; Middelburg, The Netherlands), eluting with
hexane/dioxane (97:3) (v/v) at a ow rate of 0.7 mL/min. A uorescence detector
(Jasco 821-FP) was used, with excitation wavelength at 290 nm and emission wavelength at 330 nm, gain 10. The concentrations were expressed as mg/kg of oil.
2.3.9. b-Carotene
Carotene concentration was calculated from the UV measurements of the hexane extracts obtained in accordance with Maia et al. (2008). For the purpose, oil
samples were dissolved in dimethylformamide and extracted with several portions
of n-hexane. The combined solutions were concentrated under a gentle nitrogen
ow, and their absorbance evaluated at 454 nm. Quantication was achieved by
external standards, based on calibration curves obtained with standard solutions
extracted in accordance with the samples protocol.
2.3.10. Total phenols content
Total phenols were isolated from a solution of oil in hexane by triple-extraction
with watermethanol (60:40 v/v), and estimated with FolinCiocalteu reagent at
725 nm. Results were expressed as mg of caffeic acid per kg of oil.

3. Results and discussion

3.1. Changes in the total polar compounds (TPC)
In accordance with the Portuguese law and as mandatory in
several countries, the legal rejection point for frying oils is 25% of
TPC, including under this label the hydrolysis products (diglycerides, monoglycerides and free fatty acids), the oxidation and polymeric derivatives, all formed at temperatures below 180 C
(Portaria No. 1135/95). Under real frying facilities, mostly restaurants and industries, this parameter is usually evaluated by commercial rapid tests, mostly based on colorimetric readings, which
have proven to correlate well with the values obtained by ofcial
standards, and allow rapid on-line measurements.
All samples were evaluated before the rst frying session and at
every 3 h of frying, until the 25% were achieved. All olive oils presented similar TPC amounts before frying, with 5.5% (EVOO), 7.0%
(COB2004), 6.5% (COB2005), and 6.0% (ROO) (Table 1). The TPC of

2974

S. Casal et al. / Food and Chemical Toxicology 48 (2010) 29722979

Table 1
Physical and chemical characteristics of the oils during frying.
Frying time (h)

TPC (%)

FFA

PV

p-AV

K232

K270

DK

ROS (h:min)

EVOO
0
3
6
9
12
15
18
21
24
27

5.5
7.5
8.5
11.5
13.5
15.5
18.0
20.5
22.5
25.0

0.2
0.3
0.3
0.3
0.3
0.4
0.4
0.5
0.5
0.6

11
14
15
15
15
12
11
11
10
11

6
32
41
50
52
66
58
59
59
60

1.66
1.80
1.92
2.16
2.36
2.08
2.14
2.13
2.24
2.05

0.20
0.90
1.01
1.03
0.99
0.97
0.95
1.02
1.01
1.02

0.00
0.07
0.07
0.06
0.05
0.04
0.02
0.01
0.02
0.01

16:24
16:00
4:44
2:56
2:27
2:31
1:59
1:54
1:59
1:59

COB2004
0
3
6
9
12
15
18
21
24

7.0
8.5
11.0
14.5
16.0
19.0
21.5
23.0
25.5

0.2
0.3
0.3
0.3
0.4
0.4
0.4
0.5
0.5

11
11
9
10
10
11
10
11
10

3
32
42
49
55
58
62
58
64

1.40
1.98
2.15
1.92
2.06
2.02
1.93
2.24
2.27

0.15
0.64
0.88
0.95
0.89
0.89
0.97
0.91
1.04

0.01
0.05
0.04
0.03
0.02
0.01
0.01
0.02
0.02

8:51
7:06
3:23
1:48
1:45
1:42
1:42
1:16
1:14

COB2005
0
3
6
9
12
15
18
21
24

6.5
8.5
11.0
14.5
16.0
19.0
21.5
22.0
25.5

0.2
0.3
0.3
0.4
0.4
0.4
0.5
0.5
0.6

11
9
12
10
11
13
14
14
14

4
32
43
50
56
61
61
62
65

1.72
2.16
2.18
2.15
2.13
2.39
2.39
1.97
2.24

0.14
0.82
0.90
0.96
1.00
1.05
1.05
1.08
1.13

0.00
0.08
0.03
0.01
0.04
0.02
0.02
0.01
0.01

6:01
1:49
1:47
1:09
1:11
0:54
0:51
0:45
0:42

ROO
0
3
6
9
12
15
18
21
24
27

6.0
8.0
9.5
12.5
14.5
16.5
19.0
21.5
23.0
26.0

0.3
0.3
0.3
0.5
0.5
0.5
0.5
0.6
0.7
0.7

12
11
10
10
11
10
11
10
9
9

6
31
44
49
50
55
56
55
61
58

1.84
2.24
2.11
2.10
1.94
1.88
1.98
2.18
2.36
2.09

0.33
1.06
1.10
1.10
1.05
1.02
1.04
1.06
1.10
1.06

0.02
0.08
0.08
0.04
0.03
0.02
0.02
0.01
0.01
0.01

15:18
14:30
3:19
2:22
2:08
2:06
1:38
1:53
1:56
1:51

SO
0
3
6
9
12
15

13.5
17.5
18.0
20.5
23.5
27.0

0.1
0.1
0.1
0.1
0.1
0.2

7
28
23
21
21
15

10
58
92
116
144
167

2.11
2.13
2.02
2.07
2.43
2.28

1.10
1.99
2.11
2.24
2.63
2.51

0.14
0.08
0.04
0.03
0.02
0.01

3:55
2:57
2:39
2:42
2:20
2:19

the vegetable oil (SO) was signicantly higher (13.5%), but in accordance with similar vegetable oils commonly evaluated in the laboratory where this study was developed.
The total frying time was similar for all the olive oils, ranging
from 24 to 27 h (3 days), with steady increases up to the mandatory rejection point (25%) (Fig. 1). Both monovarietal extra-virgin
olive oils achieved the rejection point after 24 h, followed by the
ROO mixture with around 26 h (26% at the 27th h), and nally
the EVOO with 27 h of frying. The vegetable oil (SO) reached the
TPC limit sooner, with 27% already after the 15th frying hour.
If we are detained on the formation of TPC during the consecutive frying sessions, the degradation rates are similar between all
olive samples, with a 0.7% increase per hour in the EVOO, and
0.8% in all the other olive samples, without clear differences. This
degradation rate was, however, slightly higher in the vegetable
oil used for comparison, with 0.9% per hour.
The reduced differences between all ve samples, independently of their identity, highlight that the formation of TPC must
be highly dependent upon the frying conditions of the study,

namely frying temperature, type and amount of food, etc., situations that were kept controlled during this study. In order to
understand the chemical factors supporting the differences ob-

Fig. 1. Changes occurred in the total polar compounds (% TPC) during the frying
sessions.

S. Casal et al. / Food and Chemical Toxicology 48 (2010) 29722979

served, several parameters were evaluated and are discussed

below.
3.2. Changes in free fatty acids
The FFA, resulting mainly from hydrolysis of triglycerides, was
evaluated by conventional acidity measurement. All the virgin olive oils presented acidity levels below 0.3% before heating, in
accordance with the labelling category extra-virgin olive oils
(60.8%). The vegetable oil (SO) acidity was comparatively smaller
(0.1%), due to the neutralization step during rening, and entirely
within the legal limits for vegetable oils, as dened in Codex Standard 210 (2009).
A steady increase in the acidity was observed for all olive oils
(Fig. 2), with nal values between 0.5% and 0.7%. For the vegetable
oil the hydrolysis was reduced until rejection, with only 0.2% after
15 h of frying.
The commercial olive oil (ROO) presented a slightly higher acidity than all the other virgin olive oils. Knowing that rening removes the free fatty acids by saponication, smaller amounts are
expected for rened olive oil. Therefore, one can suppose that the
virgin oil used in the blend presented a superior acidity.
Despite the inexistence of mandatory rejection points for used
oil regarding hydrolysis (Paul and Mittal, 1997), this parameter is
some timed used in rapid commercial tests as an estimation of
the rejection point. Based on the information obtained from these
frying experiments, the hydrolysis was reduced at the rejection
point, being a parameter with minor capacity to evaluate the true
thermal degradation of frying oils.
3.3. Changes in peroxide and p-anisidine values
The hydroperoxides formation is correlated with the fatty acids
oxidation susceptibility and the antioxidant levels. However,
hydroperoxides are transient chemical compounds and, despite
being a mandatory parameter for both bottled olive and vegetable
oils oxidative status, they are not always directly correlated with
sample oxidation, mostly on heated samples. The p-anisidine value, despite being a more empiric determination, is well correlated
with the level of secondary oxidation products, namely aldehydes,
far more stable than hydroperoxides. Therefore, for an accurate
estimation of the oxidation status, both parameters should be
interpreted simultaneously.
The results for peroxide (PV) and p-anisidine (p-AV) levels are
detailed in Table 1. All unheated samples presented peroxide values within the legal limits: below 20 mEq O2/kg for virgin oils,
10 mEq O2/kg for the vegetable oil, and 15 mEq O2/kg for the blend
of rened with virgin olive oil. The p-anisidine values before frying

Fig. 2. Changes occurred in free acidity values (% of free fatty acids) during frying
sessions.

2975

were below 10, indicating the almost absence of secondary oxidation products in all samples under study.
All olive oils presented similar peroxide levels during frying,
with slight variations around the initial amounts (1011 mEq O2/
kg) due to the volatile nature of peroxides. The EVOO sample
was characterized by a peroxide increase during the rst 12 h
(15 mEq O2/kg), followed by a return to the initial levels. Both
monovarietal olive oils extracted from Cv. Cobranosa, for 2004
and 2005, had similar behavior during the rst 12 h, but the
COB2005 PV increased afterward up to the rejection point, with a
nal value of 14 mEq O2/kg, in opposition to only 10 mEq O2/kg
in the COB2004 sample. This observation is interesting from the
chemical point of view because both samples are similar concerning their origin, but the older one (COB2004) appears to be more
resistant to oxidation. The olive oil blend (ROO) presented comparatively smaller peroxide amounts, as expected by the presence of
rened oil where the peroxides are eliminated, and the behavior
during frying was also quite constant, being the sample with less
peroxide amounts, in opposition to the hydrolysis results.
The vegetable oil assumed a clearly different behavior, starting
with only 7 mEq O2/kg and achieving, after 3 h, a peroxide value of
28 mEq O2/kg. From this moment, this parameter decreased until
the rejection, reporting a nal peroxide value of 15 mEq O2/kg.
The increased levels might be explained by the high polyunsaturated fatty acids level characteristic of sunower based formulas,
mostly linoleic acid, highly susceptible to oxidation when compared to monounsaturated oils, as olive oil.
As regards to p-anisidine values, all olive oil samples presented
an identical evolution during the entire study, independently of the
initial peroxide values and variations, with nal values from 58 to
65. COB2005 sample in particular, although presenting a higher PV,
was characterized by similar p-anisidine values, indicating that the
formation of secondary stable oxidation products occurs at similar
rates in all olive oil samples (Fig. 3).
The vegetable oil blend (SO) presented a steady increase in the
p-anisidine value, from 10 to 167, representing a clear formation of
secondary oxidation products when compared with the olive oil
samples. It is noticed that after 3 h of frying the value of p-anisidine is identical to the values obtained by the olive oils after 15
18 h. This result attests that this vegetable oil presents a minor
resistance to oxidation under frying conditions when compared
to the olive oils.
The higher p-anisidine and peroxide levels in the vegetable oil
(SO), being related with the formation of oxidation products, can
also give some insight into the higher TPC observed in this sample.
The similarity in the olive samples behaviors and its opposition to
the SO blend indicates that the fatty acid composition might be the
main responsible.

Fig. 3. Changes in p-anisidine value during the frying sessions.

2976

S. Casal et al. / Food and Chemical Toxicology 48 (2010) 29722979

3.4. Changes in the specic extinction coefcient at 232 and 270 nm

The ultraviolet spectrophotometric analysis, expressed as specic extinction coefcients, indicates the olive oil oxidation status
by providing information on the quality, preservation state and/or
changes brought about by technological processes, being mandatory for olive oils categories (Annexes II and IX in European Community Regulation EEC/2568/91). The K232 is correlated with the
formation of conjugated dienes of polyunsaturated fatty acids,
while the K270 values are indicative of the presence of primary
and secondary oxidation products, including conjugated trienes
and carbonyl compounds. The maximum values allowed for K232
and K270 are, respectively, 2.50 and 0.20 for extra-virgin olive oils,
and 2.60 and 0.25 for virgin olive oil. For oils partially (olive oil) or
totally rened (common vegetable oils), the K270 limit is comparatively higher (0.90) and the K232 is not evaluated. The DK value,
corresponding to the variation within the 270 nm region, is also
regulated for olive oils, with maximum amounts of 0.01 for virgin
olive oils and 0.15 for blends.
All unheated olive oil samples were within the legislation limits
(Table 1). Despite the inexistence of limits regarding the rened
vegetable oil (SO), the values were within those for regulated for
100% rened olive oil (K270 6 1.10 and DK 6 0.16).
The specic coefcient K232 increased in the analyzed samples,
experiencing slight variations around a plateau level with none
exceeding the maximum legal value for this parameter, despite
being regulated only for unheated samples. The K270 measurements increased above the legislation limit soon after the rst frying hours and kept constant thereafter until the end of the study,
with no clear differences between all the olive oils under study.
The SO also increased proportionally, but with a high reading already before frying, due to the severe conditions of the rening
process. The DK values presented also similar behaviors in all oils,
with initial increases followed by decreases until the end of the frying process, but surpassing the limits at the rst heating sessions.
Only with the vegetable oil (SO) a continuous decline along the
15 h of frying was observed.
For these spectrophotometric parameters, no relevant correlations were observed with the frying time (Table 4), except in the
DK value of the vegetable oil (SO) where a very signicant negative
correlation was established.
3.5. Changes in the oxidative stability
The Rancimat method is frequently used for evaluating or predicting oxidative stabilities under heating conditions, known as
induction time. Despite being an important tool for comparison
of different vegetable oils, the results are not directly transposable
to real frying conditions.
Clearly distinct results were observed for all samples (Table 1).
The highest oxidative stability was depicted by the EVOO, with
16 h, followed by the ROO with 15 h. Indeed, during the frying
studies these samples presented similar results, with 27 h and
26 h, respectively, as previously discussed. The monovarietal samples presented a comparatively reduced stability, with only near
9 h for the COB2004 and 6 h for COB2005, but in line with the previous discussed results, with a worse performance for the fresher
oil. Nevertheless, these values are smaller than expected because
under real frying conditions only a slight reduction was observed,
with 24 h for both samples. The vegetable oil (SO) presented a
clearly reduced stability, with only 4 h, in line with the frying
performance.
In order to understand the changes in the oxidative stability
during the frying sessions, samples were taken from each frying
session and also evaluated with the Rancimat test. The oxidative
stabilities of both EVOO and ROO kept almost constant after the

rst 3 h of frying, but a clear reduction was observed at the 6th h

(4:44 h and 3:19 h, respectively), and after 18 h of frying their values remained constant until the end of the process (2 h). The
monovarietal COB2004 presented a similar reduction trend but
the COB2005 stability was reduced already in the rst sampling,
after 3 h of heating, supporting its inferior quality, and in line with
the previous observations. The SO, despite presenting reduced initial levels, was characterized by a smother reduction, varying between 3.6 and 2.2 h.
3.6. Changes in the fatty acids composition
In order to understand the chemical issues behind the frying
susceptibility observed for the samples under analysis, several
parameters were evaluated. Being the fatty acids the main oils constituent and their nature, particularly the unsaturation degree,
among the most important factors determining the oxidative stability, the fatty acid composition was evaluated in detail.
Olive oil is, as expected, was characterized by a high degree of
monounsaturated fatty acids (MUFA), particularly oleic acid, an
important issue contributing to the cardiovascular health effects
(Covas, 2007). All olive oil samples presented oleic acid amounts
within those expected (55.083.0%, Reg. EEC 2568/91), varying
from 73.4% in the COB2005 sample to 77.0% in the ROO (Table 2).
The sunower based vegetable oil (SO) was characterized by a
dominant polyunsaturated fraction, particularly linoleic acid
(62.8%), and a comparatively low amount of oleic acid (24.8%)
and linolenic acid (0.1%). All olive oil samples were also within
the regulated levels regarding trans fatty acids, with reduced
amounts in comparison with the olive oil mixture ROO and the
SO sample.
When comparing the relative percentage of fatty acids during
the frying sessions, a clear reduction of unsaturated fatty acids
was observed, with the consequent increase in the saturated fatty
acids amount (Table 2). The decrease was particularly noticeable in
the polyunsaturated fatty acids, presumably by oxidation, with the
linolenic acid reduced to about half the initial amounts, despite
being a minor fatty acid. The SO alterations followed the same
trend, with equivalent losses for the same sampling times, despite
the increased polyunsaturated amounts. A linear trans fatty acids
increase with frying time was observed for all samples under
study, and despite the initial higher amounts for the rened (SO)
or partially rened (RO) samples, the nal values were equivalent
in all samples, but still clearly within the requirement for a zero
trans content per serving. Based on these observations, and despite the clear differences on the fatty acid composition of both
matrices, the degradation rated appears to be similar.
3.7. Changes in the tocopherol and tocotrienol content
Vitamin E, especially a-tocopherol, is an important vegetable oil
component contributing to the nutritional value, both by its vitaminic action and antioxidant properties (Kiritsakis, 1998). Their presence in the oils is also important from the technological point of
view due to its antioxidant action, being generally oxidized to
quinidine and dimmers while protecting the fat (Hoffman, 1989).
Table 3 details the changes occurred in the tocopherol and tocotrienol contents, with an increased degradation as function of the
frying time. Prior frying all olive oils presented variable amounts
of total vitamin E, from 139 mg/kg to 189 mg/kg, mainly a-tocopherol. These values are within the expected for olive oils. In line with
the previous observations, the monovarietal oils presented lower
vitamin E amounts, particularly the COB2005 sample, an important
parameter for the antioxidant capability and thus the oil performance. The vegetable oil (SO) presented signicantly higher

2977

S. Casal et al. / Food and Chemical Toxicology 48 (2010) 29722979

Table 2
Changes in the fatty acid composition during frying at 170 C (g/100 g fatty acids).
Frying time (h)

SFA

18:1

EVOO
0
3
6
9
12
15
18
21
24
27

14.50 0.27
14.39 0.52
14.95 0.01
14.80 0.02
14.60 0.45
14.43 0.02
14.91 0.46
14.85 0.15
15.38 0.09
15.38 0.09

76.80 0.25
76.83 0.52
76.85 0.05
76.85 0.23
77.40 0.37
77.16 0.11
77.96 0.50
77.29 0.10
76.75 0.10
75.94 0.59

COB2004
0
3
6
9
12
15
18
21
24

14.69 0.11
14.97 0.22
15.05 0.12
15.48 0.05
15.45 0.14
15.50 0.05
15.71 0.00
16.47 0.21
16.09 0.04

COB2005
0
3
6
9
12
15
18
21
24

18:3

Trans

5.64 0.04
5.60 0.12
5.13 0.01
5.04 0.00
4.86 0.00
4.68 0.03
4.46 0.04
4.32 0.02
4.18 0.02
4.31 0.05

0.76 0.0
0.71 0.0
0.62 0.0
0.60 0.0
0.53 0.0
0.49 0.0
0.45 0.0
0.41 0.0
0.40 0.0
0.41 0.0

0.03 0.00
0.06 0.01
0.17 0.02
0.15 0.01
0.19 0.01
0.25 0.03
0.28 0.01
0.34 0.02
0.37 0.01
0.41 0.02

75.41 0.00
75.33 0.22
74.95 0.22
75.05 0.07
75.36 0.33
75.40 0.11
75.17 0.08
74.51 0.24
74.99 0.02

7.00 0.06
6.73 0.06
6.47 0.02
6.30 0.01
6.06 0.04
5.88 0.01
5.67 0.01
5.40 0.10
5.34 0.03

0.74 0.00
0.67 0.00
0.62 0.01
0.57 0.00
0.52 0.00
0.47 0.00
0.45 0.00
0.44 0.01
0.39 0.00

0.03 0.00
0.09 0.00
0.14 0.03
0.21 0.01
0.25 0.02
0.29 0.01
0.35 0.02
0.45 0.05
0.45 0.00

15.27 0.09
15.59 0.20
15.75 0.07
15.86 0.13
16.01 0.20
16.18 0.08
16.33 0.02
16.51 0.11
16.72 0.02

73.38 0.18
74.04 0.10
74.13 0.12
73.60 0.05
74.03 0.21
74.06 0.09
73.64 0.41
74.17 0.39
73.76 0.04

7.87 0.06
7.55 0.04
7.26 0.05
7.18 0.03
6.77 0.02
6.50 0.01
6.29 0.01
6.18 0.04
5.91 0.05

0.64 0.00
0.60 0.00
0.55 0.00
0.52 0.02
0.47 0.00
0.42 0.00
0.41 0.00
0.39 0.00
0.36 0.01

0.03 0.01
0.09 0.01
0.14 0.01
0.18 0.02
0.24 0.00
0.29 0.01
0.35 0.01
0.42 0.00
0.44 0.01

ROO
0
3
6
9
12
15
18
21
24
27

14.43 0.06
15.11 0.09
15.41 0.62
15.03 0.04
15.27 0.31
15.58 0.03
15.80 0.02
15.87 0.10
15.72 0.00
16.07 0.14

77.05 0.08
76.09 0.17
75.35 0.17
76.34 0.11
76.04 0.07
75.93 0.12
76.11 0.03
75.94 0.29
75.92 0.22
75.66 0.04

5.59 0.00
5.71 0.01
5.42 0.16
5.24 0.04
5.10 0.04
4.86 0.00
4.66 0.00
4.49 0.05
4.32 0.03
4.21 0.01

0.74 0.00
0.57 0.00
0.54 0.02
0.45 0.01
0.43 0.01
0.40 0.00
0.37 0.00
0.34 0.01
0.32 0.00
0.31 0.01

0.10 0.00
0.22 0.00
0.38 0.03
0.33 0.02
0.39 0.00
0.41 0.00
0.46 0.00
0.52 0.01
0.54 0.01
0.58 0.02

SO
0
3
6
9
12
15

11.28 0.08
11.38 0.01
11.35 0.00
11.43 0.03
11.66 0.01
11.77 0.02

24.82 0.90
24.63 0.06
24.49 0.06
24.72 0.01
25.35 0.20
25.21 0.00

62.78 0.93
62.73 0.03
62.84 0.04
62.45 0.01
61.57 0.11
61.31 0.07

0.11 0.00
0.10 0.00
0.10 0.00
0.10 0.02
0.09 0.00
0.09 0.00

0.27 0.01
0.29 0.00
0.31 0.00
0.34 0.01
0.38 0.05
0.39 0.02

amounts of total vitamin E (474 mg/kg), again with a-tocopherol

as the main vitamer, as expected from a sunower based formula.
Vitamin E degraded sharply in all olive oils under study (Table 3), being practically inexistent after 36 h of frying. In opposition, the vegetable oil levels were only slightly reduced, with still
about 65% of the initial amounts at the rejection point. The atocopherol degraded at a faster rate in all olive oils (1548 mg/
kg/h) when compared to the vegetable oil (9 mg/kg/h). Therefore,
and despite being the most important lipid antioxidant component
of the oils, vitamin E seems to play a reduced part in the oil frying
resistance.
3.8. Changes in the b-carotene
Carotenoids are also naturally present in virgin olive oils, because only physical processes are used during extraction. Table 3
reports the changes occurred in the b-carotene contents of the oils
subjected to the frying process. All samples presented distinct bcarotene amounts, within the expected for olive oil (Velasco and

18:2

Dobarganes, 2002), clearly higher in the EVOO. In accordance with

the previous observations, the COB2005 sample presented lower
amounts than the COB2004 sample, and even lower that the ROO
sample. During frying a similar pattern was observed, with a
reduction up to the 912th h, followed by a slight increase in all
samples, probably from the potatoes. Only diminutive amounts
of b-carotene were determined in the sunower based sample.
3.9. Total phenols content
In comparison with other vegetable oils, olives and olive oils are
receiving particular attention for their phenolic content given that
these compounds have antioxidant and antimicrobial effects
(Ocakoglu et al., 2009). Indeed, phenols are frequently described
as key factors in the olive oils oxidative resistance (Gmez-Alonso
et al., 2003).
The amount of total phenols, expressed as mg of caffeic acid
equivalents, was higher in the EVOO sample, with 204 mg/kg, followed by the ROO (154 mg/kg) and by the COB2004 and COB2005,

2978

S. Casal et al. / Food and Chemical Toxicology 48 (2010) 29722979

Table 3
Vitamin E, b-carotene and phenols oil contents during frying (mg/kg).

a-T

a-Ttr

b-Ttr
(mg/kg)

c-Ttr

(mg/kg)

b-T
(mg/kg)

c-T

(mg/kg)

(mg/kg)

(mg/kg)

d-T
(mg/kg)

Total vitamin E
(mg/kg)

b-Carotene
(mg/kg)

Phenolsa
(mg/kg)

EVOO
0
3
6
9
12
15
18
21
24
27

179.9 2.3
101.4 1.1
4.7 0.2
n.d.
n.d.
n.d.

1.1 0.2
0.6 0.1
n.d.
n.d.
n.d.
n.d.

1.9 0.1
0.9 0.1
n.d.
n.d.
n.d.
n.d.

5.7 0.2
2.0 0.1
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

189
105
5
n.d.
n.d.
n.d.

5.2 0.1
3.7 0.1
3.0 0.1
2.0 0.0
1.4 0.0
1.7 0.1
2.5 0.2
1.7 0.1
2.2 0.2
3.0 0.1

204 17
119 29
53 3
n.d.

COB2004
0
3
6
9
12
15
18
21
24

151.9 3.2
8.4 0.0.
5.6 5.1
1.7 0.1
n.d.
n.d.

1.6 1.0
n.d.
n.d.
n.d.
n.d.
n.d.

1.7 0.0
n.d.
n.d.
n.d.
n.d.
n.d.

1.6 0.0
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

157
8
6
2
n.d.
n.d.

2.04 0.06
0.65 0.08
0.57 0.02
0.74 0.07
0.88 0.07
0.99 0.02
1.04 0.08
1.23 0.03
1.07 0.03

98 15
59 1
n.d.

COB2005
0
3
6
9
12
15
18
21
24

136.8 2.5
2.1 0.0
0.1 0.0
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

1.2 0.2
n.d.
n.d.
n.d.
n.d.
n.d.

1.1 0.2
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

139
2
n.d.
n.d.
n.d.
n.d.

0.97 0.11
0.49 0.05
0.49 0.03
0.63 0.08
0.91 0.06
1.08 0.11
1.11 0.04
1.20 0.03
1.38 0.04

63 2
50 11
n.d.

ROO
0
3
6
9
12
15
18
21
24
27

139.1 0.4
94.1 0.2
2.6 0.2
n.d.
n.d.
n.d.

1.1 0.2
n.d.
n.d.
n.d.
n.d.
n.d.

3.2 0.3
1.3 0.0
0.1 0.0
n.d.
n.d.
n.d.

21.2 1.6
4.7 0.0
0.2 0.0
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

n.d.
n.d.
n.d.
n.d.
n.d.
n.d.

0.5 0.1
0.1 0.1
n.d.
n.d.
n.d.
n.d.

165
100
3
n.d.
n.d.
n.d.

1.6 0.0
1.3 0.0
1.6 0.2
1.5 0.1
2.0 0.0
2.0 0.0
2.1 0.1
2.1 0.2
2.6 0.1
3.7 0.1

154 5
72 6
n.d.

SO
0
3
6
9
12
15

432.2 9.0
413.1 0.0
372.1 0.7
350.4 0.7
322.3 2.5
291.3 0.7

2.9 0.4
2.3 0.2
1.7 0.1
1.7 0.0
1.4 0.0
1.1 0.0

21.1 0.2
19.7 0.2
18.2 0.2
17.5 0.0
16.1 0.1
15.1 0.2

5.6 0.1
4.5 0.0
4.3 0.0
3.5 0.0
2.3 0.4
1.6 0.2

8.0 0.2
7.3 0.1
6.4 0.0
5.6 0.0
4.5 0.1
3.5 0.0

3.2 0.1
2.6 0.3
3.9 0.0
3.5 0.1
2.3 0.6
1.8 0.6

1.2 0.1
1.0 0.1
1.0 0.0
1.0 0.1
0.8 0.1
0.7 0.1

474
451
408
383
350
315

0.23 0.05
n.d.
n.d.
n.d.
n.d.
n.d.

31 1
n.d.

Frying
time (h)

n.d. not detectable.

Not determined.
a
Phenols: mg caffeic acid equiv./kg.

with only 98 and 63 mg/kg, respectively. These amounts are in

accordance with the frying performances. Nevertheless, after 6 h
of frying the total phenols were only quantied in the EVOO sample, being absent in all samples at the 9th h, in accordance with the
data published by Gmez-Alonso et al. (2003). These results are totally comparable with the vitamin E performance, indicating that
both compounds are fast degraded during the rst frying hours.
Therefore, if these compounds take a part in the frying resistance
towards oxidation, it must be restricted to the rst heating hours.
If the Rancimat results are also compared, one can observe that
while both vitamin E and phenols are present the oxidative resistance is shortly reduced, but once these compounds are degraded
the oxidative resistance is signicantly reduced. The olive oil fatty
acid composition, highly monounsaturated, should be the rate limiting factor determining its oxidation. The almost absence of pro-

tective phenols and the polyunsaturated nature are certainly

important factors contributing to the precocious oxidation of the
vegetable oil blend (SO).
3.10. Correlations
The correlation between the parameters evaluated and the frying time are depicted in Table 4. Signicant correlations were observed for the TPC, as expected, for all the samples. The acidity,
in opposition, was only signicant in the olive oil samples but
not in the sunower blend (SO) with a reduced hydrolysis rate
up to the last frying session.
Despite being formed in reduced amounts, the trans fatty acids
correlation with the frying time was highly signicant, independently of the samples identity. The PUFA reduction was also statis-

2979

S. Casal et al. / Food and Chemical Toxicology 48 (2010) 29722979

Table 4
Correlation between the evaluated parameters and frying time.
EVOO
r
TPC
FA
PV
p-AV
K232
K270
DK
ROS
SFA
MUFA
PUFA
Trans FA
Vitamin E
b-Carotene

0.996
0.913
0.318
0.673
0.412
0.328
0.272
0.601
0.559
0.087
0.967
0.977
0.997
0.297

COB2004
p
***
***

n.s.
**
*

n.s.
n.s.
**
*

n.s.
***
***
**

n.s.

COB2005

0.995
0.919
0.283
0.766
0.483
0.568
0.198
0.680
0.881
0.309
0.993
0.988
0.635
0.010

***
***

n.s.
**
*
*

n.s.
**
***

n.s.
***
***

n.s.
n.s.

ROO
p

0.991
0.939
0.061
0.782
0.228
0.601
0.115
0.525
0.990
0.052
0.988
0.996
0.761
0.615

***
***

n.s.
**

n.s.
*

n.s.
*
***

n.s.
***
***

n.s.
*

SO

r2

0.998
0.905
0.411
0.639
0.119
0.256
0.412
0.569
0.814
0.030
0.992
0.860
0.987
0.717

***

0.970
0.428
0.013
0.981
0.393
0.801
0.878
0.685
0.873
0.556
0.802
0.981
0.996

***

***
*
**

n.s.
n.s.
*
*
***

n.s.
***
***

n.s.
**

n.s.
n.s.
***

n.s.
*
**
*
**

n.s.
*
***
***

n.s. not signicant.

p 6 0.05.
**
p 6 0.01.
***
p 6 0.001.
*

tically noticeable, but with higher signicance in the olive oils

when compared with the vegetable oil. Among the oxidative
parameters evaluated, only the p-anisidine was well correlated
with the frying hours, but with higher signicance in the vegetable
oil. The vitamin E reduction was also correlated with the frying
hours but only with statistically signicance in the EVOO and SO
samples, mostly because of their early loss.
4. Conclusions
In the present work is proved that olive oil, independently of its
category label, is clearly resistant to degradation under domestic
frying conditions (170 C). Among the olive oil categories, the extra-virgin PDO olive oil was more resistant under the conditions
of this study. Being from the Trs-os-Montes region, this olive
oil is particularly rich in Cobranosa cultivar. Nevertheless, when
this cultivar was evaluated individually, the monovarietal olive oils
were clearly less resistant, probably due to the lower oleic acid,
vitamin E and phenol amounts. Also, the monovarietal olive oils
from consecutive years presented different oxidative susceptibilities, strengthen the importance of avoiding single cultivar olive
oils, a practice that has increased in the last years because of their
natural chemical and sensory characteristics. The olive oil, a
commercial blend of rened with virgin olive oil, traditionally
cheaper, behaved similarly to the EVOO sample. This highlights
that, despite the presence of undeclared portions of rened olive
oil, the most important compounds dening its oxidative resistance are still present, namely vitamin E, phenols and mostly the
monounsaturated triglycerides. The vegetable oil taking as comparison, despite being less expensive and enclose higher amounts
of vitamin E, is decient in important bioactive compounds, like
phenols, and highly susceptible to early oxidation, contributing
to the ingestion of higher amounts of oxidized fatty acids.
Vitamin E, phenols, and b-carotene seem to be key compounds
determining the reduced loss of oxidative stability during the rst
frying hours. The highly monounsaturated nature of the olive oils is
a second decisive parameter for the reduced thermal degradation
under real domestic frying conditions, as observed by comparison
with the commercial vegetable oil blend.
Conict of Interest
The authors declare that there are no conicts of interest.

References
Andrikopoulos, N.K., Kalogeropoulos, N., Falirea, A., Barbagianni, M.N., 2002.
Performance of virgin olive oil and vegetable shortening during domestic
deep-frying and pan-frying of potatoes. Int. J. Food Sci. Technol. 37, 177190.
Assmann, G., de Backer, G., Bagnara, S., Betteridge, J., Crepaldi, G., Fernandez-Cruz,
A., Godtfredsen, J., Jacotot, B., Paoletti, R., Renaud, S., Ricci, G., Rocha, E.,
Trautwein, E., Urbinati, G.C., Varela, G., Williams, C., 1997. International
consensus statement on olive oil and the Mediterranean diet: implications for
health in Europe. The Olive Oil and the Mediterranean Diet Panel. Eur. J. Cancer
Prev. 6, 418421.
Codex Standard 210, 2009. Codex Standard for Named Vegetable Oils, Codex
Alimentarius.
Commission Regulation (EEC) No. 2568/91 on the Characteristics of Olive Oil and
Olive Residue Oil and on the Relevant Methods of Analysis, OJ L 248, 5.9.1991, p.
1.
Covas, M.-I., 2007. Olive oil and the cardiovascular system. Pharmacol. Res. 55, 175
186.
Decreto-lei 16/2004 Dirio da Repblica. I Srie A, No. 11, pp. 260264.
Fillion, L., Henry, C.J.K., 1998. Nutrient losses and gains during frying: a review. Int. J.
Food Sci. Nutr. 49, 157168.
Gmez-Alonso, S., Fregapane, G., Salvador, M.D., Gordon, M.H., 2003. Changes in
phenolic composition and antioxidant activity of virgin olive oil during frying. J.
Agric. Food Chem. 51, 667672.
Hoffman, G., 1989. The chemistry of edible fats. In: The Chemistry and Technology
of Edible Oils and Fats and their High Fat Products. Academic Press, London, pp.
128.
ISO 6885:2006. Animal and Vegetable Fats and Oils Determination of Anisidine
Value.
ISO 9936:2006. Animal and Vegetable Fats and Oils Determination of Tocopherol
and Tocotrienol Contents by High-Performance Liquid Chromatography.
Kiritsakis, A., 1998. Olive Oil: From the Tree to the Table. Food and Nutrition Press,
Connecticut. p. 348.
Kochhar, S.P., 2001. The composition of frying oils. In: Rossel, J.B. (Ed.), Frying.
Improving Quality. Woodhead Publishing Ltd., Cambridge, pp. 87114.
Maia, L., Casal, S., Oliveira, M.B.P.P., 2008. Validation of a micromethod for
quantication of lutein and b-carotene in olive oil. J. Liq. Chromatogr. Relat.
Technol. 31, 733742.
Ocakoglu, D., Tokatli, F., Ozen, B., Korel, F., 2009. Distribution of simple phenols,
phenolic acids and avonoids in Turkish monovarietal extra virgin olive oils for
two harvest years. Food Chem. 113, 401410.
Portaria No. 1135/95, 1995. Dirio da Repblica. I Srie-B, No. 214, p. 5836.
Paul, S., Mittal, G.S., 1997. Regulating the use of degraded oil/fat in deep-fat/oil food
frying. Crit. Rev. Food Sci. Nutr. 37, 636662.
Saguy, I.S., Dana, D., 2003. Integrated approach to deep fat frying: engineering,
nutrition, health and consumers aspects. J. Food Eng. 56, 143152.
Snchez-Muniz, F.J., Bastida, S., 2006. Effect of frying and thermal oxidation on olive
oil and food quality. In: Quiles, J.L., Ramrez-Tortosa, M.C., Yaqood, P. (Eds.),
Olive Oil and Human Health. CAB International Publishing, Oxford Shire, UK, pp.
74108.
Varela, G., Ruiz-Roso, B., 2000. Some nutritional aspects of olive oil. In: Harwood, J.,
Aparicio, R. (Eds.), Handbook of Olive Oil. Analysis and Properties. Aspen
Publisher, Inc., Gaithersburg, Maryland, pp. 565580.
Velasco, J., Dobarganes, C., 2002. Oxidative stability of virgin olive oil. Eur. J. Lipid
Sci. Technol. 104, 661676.