1 INTRODUCTION
Numerous reports1–5) have made it clear that the toxicity
of oxidized edible oil is due mainly to oxidation products
from unsaturated fatty acids. The ingredient standard for
commercially processed foods, Japan, established the
chemical properties of oil and fat at AV ≦3 and POV ≦
306). The legislation seems to be based on the understand-
ing that the safety of the edible oil and fat is ensured by
setting limits on these two properties. In some other coun-
tries, only an AV cutoff point is set for oil and fat. Gotoh et
al.7) have argued that AV does not reflect the risks due to
oxidation of oil and fat at all. It is obvious that AV is rela-
tively stable and that the cutoff point is hardly exceeded
during usual food processing. We8) screened the AV of oils
contained in batter coatings of various commercially deep-
fried foods, such as cutlets, tempura, and croquettes, and
found that only 4% of screened foods contained oil with
AV>3 in spite of unpleasant appearance and odor.
Soriguer et al.9) found after extensive epidemiological
studies that the risk of hypertension was positively and
independently associated with the intake of cooking oil
polar compounds. Sanchez-Muniz et al.10) reported that the
triacylglycerol oligomer content in used frying oil gave
more precise information about the alteration of the oil and
its potential toxicity than PC did.
The present legislation on oil and fat described above
may not be a problem, as no food poisoning due to deterio-
rated oil or fat has been reported officially in Japan in
years. But consumption of deep-fried foods is huge nowa-
days11,12), and metabolic syndrome is common in the
younger generation as well as in middle-aged people. Thus,
it is essential to thoroughly address the effects of oxidized
oil on human health from the point of lifestyle-related dis-
eases. For several years, we have been studying13–16) the
effects of oil thermally oxidized to such a degree that it
causes neither diarrhea nor stomachache. Wistar rats
were fed ad libitum a standard diet containing 7 wt% of
the oil for 8-12 weeks. The animals developed histological
damage in the liver and kidneys and hematological changes
without gross symptoms attributable to the experimental
oils. In the present study, the relation between chemical
properties of thermally oxidized frying oil and cytotoxicity
153
＊Correspondence to: Nagao Totani, Department of Nutritional Physiology, Faculty of Nutrition, Kobe-Gakuin University, 518 Arise,
Ikawadani-cho, Nishi-ku, Kobe, 651-2180, JAPAN
E-mail: totani@nutr.kobegakuin.ac.jp
Accepted December 4, 2007 (received for review November 26, 2007)
Journal of Oleo Science
Copyright ©2008 by Japan Oil Chemists’ Society
J. Oleo Sci. 57, (3) 153-160 (2008)
Chemical Properties and Cytotoxicity of Thermally
Oxidized Oil
Nagao Totani1＊
, Munkhjargal Burenjargal2, Miho Yawata1and Yuko Ojiri1
1Faculty of Nutrition, Kobe-Gakuin University (518 Arise, Ikawadani-cho, Nishi-ku, Kobe 651-2180, JAPAN)
2Faculty of Chemistry, National University of Mongolia (P.O. Box 46A-442, Ikh Surguuliin Gudamj-1, Sukhbaatar District, Ulaanbaatar-
210646, MONGOLIA)
Abstract: Heated frying oils with different chemical properties in terms of AV (acid value), POV (peroxide
value), COV (carbonyl value), and contents of polar compounds (PC) and triacylglycerol (TG), as well as
color and odor, were obtained. Male Wistar rats were fed ad libitum for 12 weeks a powdered diet
(AIN93G; no fat) containing 7 wt% of fresh oil (control) or one of the frying oils described above. The rats
were subjected to anthropometric measurements, hematological analyses, and observations of the liver and
kidneys. All of the rats grew well, and no gross symptoms attributable to the experimental oils were
observed. However, the rats fed a diet containing the heated oil developed apparent liver damage to
different degrees regardless of the chemical properties of the ingested oils. Thus, it was suggested that the
chemical properties evaluated here had little to do with the cytotoxicity of heated oil, although the
properties express quality of oil. Volatile compounds seem to be major candidates for the toxic agents in
heated oil because oils with rancid and deteriorated odor show strong toxicity.
Key words: chemical properties, acid value, peroxide value, polar compounds, thermally oxidized oil, volatile compounds,
cytotoxicity, cell damage
Journal of Oleo Science ISSN 1345-8957 print / ISSN 1347-3352 online
http://www.jstage.jst.go.jp/browse/jos/
N. Totani, M. Burenjargal , M. Yawata and Y. Ojiri
was investigated when Wistar rats were fed a diet contain-
ing the oil for a long period.
2 EXPERIMENTAL
2.1Materials
2.1.1 Oil
The fresh oil 1 (used for animal experiment 1) was made
of fresh soybean and rapeseed oils. One liter of this oil was
heated with and without 1% gluten (Nacalai Tesque, Inc.,
Kyoto) at 180℃for 10 h in a 2-L four-necked flask, respec-
tively. Oil heated with gluten was filtered over filter paper
under a reduced pressure.
The fresh oil 2 (used for animal experiment 2) was also
made of fresh soybean and rapeseed oils. To 1 L of this oil
each of the following four different nutrient groups was
added, and the mixture was heated at 180℃for 20 h in a 2-
L four-necked flask under a stream of air at 3 L/m. The
four groups were a mixture of amino acids17), Gln, Gly, Ala,
Tyr, Arg, Pro, Thr, Asp (Nacalai Tesque, Inc., Kyoto), 200
ppm each; 1% gluten (Nacalai Tesque, Inc., Kyoto) from
wheat; 1% sucrose (Wako Pure Chemical Industries Ltd.,
Osaka); and 1% wheat flour (Nacalai Tesque, Inc., Kyoto).
Experimental oils, thus obtained, were allowed to stand at
a room temperature for a day to precipitate solid materials,
and each supernatant was employed for an animal experi-
ment. “B grade recovered vegetable oil” (recovered oil) was
obtained from Miyoshi Oil & Fat Co., Ltd. The oil is com-
posed mainly of soybean and rapeseed oils.
2.1.2 Diets
A commercial pelleted diet (Labo MR Stock, Nihon
Nosan Kohgyo, Japan) and a powdered AIN93G diet with-
out fat (Japan Clea, Tokyo) were purchased. Using a
blender, the latter was mixed uniformly with 7 wt% experi-
mental oils and fresh oil, respectively. The 9 kinds of diets,
thus prepared, were handled and provided as described in
our paper13–16).
2.1.3 Chemical analyses
Methods for chemical analyses of oil were the same as in
our previous papers18,19). The fatty acid compositions of the
fresh oils 1 and 2 analyzed as previously18,19) were as follows:
myristic acid 0.1% and 0.1%; palmitic acid 9.4% and 8.4%;
palmitoleic acid 0.1% and 0.1%; stearic acid 3.6% and 3.3%;
oleic acid 32.6% and 38.2%; linoleic acid 44.1% and 39.2%;
and a-linolenic acid 6.3% and 6.5%; eicosenoic acid 0.5%
and 0.6%; others 3.3% and 3.6%, respectively.
As the oil heated with gluten showed almost no toxicity
in the experiment 2, low boiling compounds in three kinds
of oils employed in the animal experiment 1 were analyzed
preliminary by head-space GC (Head-space sampler, 7694
Hewlett-Packard Company; oven temp., 80℃; vial heating
time, 30 min; loop temp., 150℃; transfer line temp., 200℃)
and GC-MS (6890/5973 Hewlett-Packard Company; Col-
umn, DB-WAX, (Agilent Technologies) F0.25 mm×60 m;
split (1:40); Column temp. 40℃for 1 min, then raised at 5℃
/min to 200℃; carrier gas He 1 mL/min; electron impact
ionization; ionization voltage 70 eV).
2.2Animals
Male Wistar rats aged 9 weeks were obtained from Japan
SLC, Inc., Shizuoka, Japan, and were housed separately in
wire cages at 24±2℃and humidity 50±10%, with light
from 7:00 to 19:00 at Japan SLC, Inc., Animal Experiment
Center, Shizuoka, Japan. Animal care and handling were in
accordance with the Ethical Agreement Concerning Care
and Use of Laboratory Animals for Research and Educa-
tion, Kobe-Gakuin University.
2.3Procedure
2.3.1 Animal experiment 1
Twenty-four animals were maintained on radio-sterilized
Labo MR Stock for 1 week of adaptation; animals were
then divided into three groups (8 rats/group). Two groups
were fed a diet containing respectively 7-wt% oils heated
with/without gluten, and the third group was fed a diet
containing 7-wt% fresh oil 1. All animals were allowed feed
and water ad libitum throughout the experiment. Autoxi-
dation of oil in the diet was avoided by supplying a fresh
diet daily as described in our previous paper15,16). After 12
weeks, a fasting period of 18 h was imposed prior to the
administration of anesthesia. Serum was obtained from
blood drawn from the abdominal aorta. Livers and kidneys
were excised, weighed, and examined.
2.3.2 Animal experiment 2
Forty-eight animals (6 groups) were subjected to the ani-
mal experiment as in 2.3.1; Five groups (amino acid group,
gluten group, sugar group, wheat starch group and recov-
ered oil group) were fed a diet containing 7-wt% experi-
mental oils, respectively. The last group was fed a diet con-
taining 7-wt% fresh oil 2.
2.4Hematological analyses
Activities of aspartate aminotransferase (AST) and ala-
nine aminotransferase (ALT) were determined as in our
previous papers15,16).
2.5Statistical analysis
All the values obtained from animals are revealed as
mean±SD. Data from 8 animals each for experimental and
control groups were analyzed using Student’s t-test for
unpaired observations and results were considered signifi-
cant at p<0.05.
3 RESULTS
3.1Chemical properties of the oils used
154
J. Oleo Sci.57, (3) 153-160 (2008)
Toxicity of Oxidized Oil
As shown in Table 1, the six laboratory-heated oils had
about 20% polar compounds, 83-91% TG, COV values of
about 30, and AV values of 0.1-0.3. Oil heated with/without
gluten had low POV, probably because the decomposition of
peroxides to carbonyl compounds was active in the first 10
h of heating. Recovered oil showed PC and TG concentra-
tions similar to those of other experimental oils, but with
low COV and high AV. The odor of the recovered oil was
the worst of all and that of oil heated with gluten was mild
and acceptable.
3.2Volatile compounds
As shown in Fig. 1, almost no peak was detected besides
that of air in fresh oil (C). Oil that was heated without
gluten (B) showed a large peak at retention time (RT) 4.4
min (peak 3) and small peaks at RT 5.1, 5.8, 6.5, 8.0, 9.0,
11.6, and 18.5 min; volatile compounds probably accounted
for the rancid smell. The oil heated with gluten (A) showed
a large peak at RT 8.0 min (peak 2), which was more than
ten times larger than that of oil heated without gluten.
Peak 1 was confirmed to be identical with peak 3 by the
fragment pattern of mass spectrogram. The other peaks
were less than a half of the corresponding peaks in the oil
heated without gluten.
3.3Growth on sample oils
All the rats in animal experiments 1 and 2 did not exhibit
diarrhea, seborrhea, dermatitis, or excessive hair loss after
the administration of any diet, as in our previous study13–16).
The weights of organs excised are shown in Fig. 2. No dif-
ference in the weight of kidneys was detected between
groups. However, the recovered oil and sugar groups had
significantly heavier livers than those of the control group.
3.4Hematological analyses
In animal experiment 1, the heated oil group had the
highest AST, followed by the group that received oil heated
with gluten and the control group, with no significant dif-
ference from each other. ALT measurement did not show
any difference among the three. In animal experiment 2,
the magnitude of AST and ALT was in the order of recov-
ered oil group, wheat starch group, amino acid group,
gluten group, sugar group, and control.
There was a significant difference in AST and ALT
between the recovered oil group and the control (Table 2).
The occurrence of AST and ALT, both higher than the
maximum AST (101 IU/L for Experiment 1 and 118 IU/L
for Experiment 2) and ALT (69 IU/L for Experiment 1 and
97 IU/L for Experiment 2) of the control, was assessed for
each group and listed in Table 3.
3.5Visible changes in livers and kidneys
Examination of the livers and kidneys revealed dark-red
patches due to dotted bleeding on the surface of the liv-
ers15), particularly from the recovered oil, amino acids,
sugar, and wheat starch groups, suggesting degeneration
of inner tissues (Table 3). A control rat and a rat in the
gluten group also had patches, but their numbers were
155
J. Oleo Sci.57, (3) 153-160 (2008)
Table 1 Chemical Properties of Frying Oil.
Oil heated at 180℃for 10 h with Recovered oil
Gluten None Fresh oil 1
PC (%)
TG (%)
COV
POV (mEq/kg)
AV
Color (Gardner)
Smell
21.6
83.5
30.6
1.4
0.1
8
mild
23.9
85.0
35.4
3.6
0.1
6
rancid
4.2
98.6
2.7
0.9
0.1
2
fresh
24.7
90.9
10.9
26.2
2
11
deteriorated
Oil heated at 180℃for 20 h with
Amino acids Gluten Sucrose Wheat starch Fresh oil 2
PC (%)
TG (%)
COV
POV (mEq/kg)
AV
Color (Gardner)
Smell
21.6
83.8
31.9
62.7
0.3
11
rancid
21.6
82.6
34.1
63.4
0.1
9
mild
17.8
90.0
39.1
76.9
0.1
6
rancid
24.7
84.7
34.2
80.0
0.1
5.5
rancid
5.7
99.4
2
6.6
0.1
2
fresh
N. Totani, M. Burenjargal , M. Yawata and Y. Ojiri
small. The number of dark-red patches corresponded well
with the high AST levels of individual rats, so that the per-
centage of rats with dark-red patches and that of rats hav-
ing high AST showed similar tendencies. But the former
represented the cytotoxicity more critically than the latter
did.
No differences in color and size were observed in the
kidneys of experimental vs. control rats, and no patches
were observed in any group.
3.6Correlation of the chemical properties and toxicity of
heated oil
Chemical properties (PC, TG, COV, POV) of the oil and
damage, based on the number of rats with dark-red patch-
es on the liver/8 rats, were plotted, respectively, to deter-
mine the relation between them (Fig. 3), but the chemical
properties proved unrelated to the occurrence of damage.
4 DISCUSSION
All the experimental oils had high PC and COV and low
TG (Table 1). POVs of the oils used for animal experiments
1 and 2 were low and high, respectively. The reason for
high POV seems to be attributable to the long heating time
under air supply and leaving the oil overnight for sedimen-
tation of solid materials. AVs of all the oils except the
recovered oil were low. The color of the experimental oils
ranged from yellow to brown. The experimental oils
smelled rancid except for the oil heated with gluten, which
had PC, COV, TG, POV, and AV values often found in heat-
ed oil, but had a mild odor with very slight rancidity. In
order to investigate the effect of gluten on the alteration of
oil odor during heating, the odor of three oils used in ani-
mal experiment 1 was analyzed by head-space GC (Fig. 1).
Several peaks probably attributable to the rancid odor
became obviously smaller by the addition of gluten, while
the peak at RT 8.0 min increased. Because peak size is
determined by the summed intensity of fragments generat-
ed from the compound by electron impact, quantification of
each compound is not possible; peaks with the same RT
can be compared because all the GC-MS conditions for
each run were identical. In addition, odor is not related to
the size of the peak, so a small peak can be attributable to a
rancid odor. The mechanism of the reactions between oil
and gluten under heating is not currently known.
In both animal experiments, all the rats grew well and
appeared normal. As shown in Fig. 2, big differences were
not found in the weights of the liver and kidneys. It was
confirmed that frying oil heated for 10-20 h and recovered
oil apparently did not impair the health condition of the
animals. The AST value reveals cell damage in the liver
and kidneys, but the average AST of each group did not
show a significant difference from that of the control
except in the recovered oil group. As described in 3･5, the
percentage of rats with dark-red patches was employed as
the index of cytotoxicity and used for the evaluation of
chemical properties.
Studies of the toxicity of oxidized oil have been focused
on its acute toxicity, such as in food poisoning. The present
study on chronic symptoms caused by the ingestion of oxi-
dized oil indicated that the chemical properties evaluated
here did not correspond to the degree of cytotoxicity (Fig.
3), and that low-molecular-weight volatile compounds
156
J. Oleo Sci.57, (3) 153-160 (2008)
Fig. 1 Head-space GC-MS of Oils heated with/without
Gluten (A)/(B), and Fresh Oil (C).
Toxicity of Oxidized Oil
157
J. Oleo Sci.57, (3) 153-160 (2008)
Fig. 2 Organ Weights of Rats Fed a Diet Containing Oil Heated with a Nutrient at 180℃
for 10 h* or 20 h**.
†p< 0.05, significantly different from the value of control 2 (unpaired t-test).
Fig. 3 Relation between Chemical Properties and Liver Damage in Rats Fed a Diet Containing
the Oil.
N. Totani, M. Burenjargal , M. Yawata and Y. Ojiri
158
J. Oleo Sci.57, (3) 153-160 (2008)
Table 2 AST/ALT of Rats Fed a Diet Containing Oil Heated with a Nutrient at 180℃for 10 h* or 20 h**.
Table 3 Liver Damage of Rats Fed a Diet Containing Oil Heated with a Nutrient at 180℃for 10 h* or 20 h**.
Values are means ±SD for eight animals.
†p< 0.05, significantly different from the value of control 2 (unpaired t-test).
Gluten* None* Fresh oil 1 Recovered oil Amino acids** Gluten** Sucrose** Wheat starch** Fresh oil 2
AST (IU/L)
ALT (IU/L)
95.4±34.1
52.2±11.1
100.1±33.0
53.3±9.1
87.6±17.9
54.4±23.9
181.6±161.9†
114.8±104.2†
105.4±50.6
71.4±46.8
100.8±60.4
63.0±29.2
96.1±38.6
55.3±18.5
136.1±121.4
100.9±105.1
73.1±16.5
48.9±11.4
Gluten* None* Fresh oil 1 Recovered oil Amino acids** Gluten** Sucrose** Wheat starch** Fresh oil 2
Occurrence of AST
higher than maximum
AST of control/8 rats
1/8 2/8 −4/8 4/8 1/8 3/8 3/8 −
Occurrence of ALT
higher than maximum
ALT of control/8 rats
0/8 0/8 −4/8 3/8 2/8 2/8 3/8 −
No. of rats with
dark-red patches on
the liver/8 rats
1/8 3/8 1/8 6/8 6/8 2/8 5/8 5/8 1/8
Toxicity of Oxidized Oil
could be candidates for the cause of cytotoxicity. Our pre-
vious paper20) reported that most of the generated carbonyl
compounds vaporized from oil during frying: frying opera-
tors could be exposed to a large quantity of carbonyl com-
pounds vaporizing with steam generated from water dur-
ing the frying of foodstuffs.
When COV started to drastically increase after reaching
the maximum POV in the autoxidation of methyl linoleate,
low-molecular-weight compounds, secondary decomposi-
tion products of peroxides, contributed to the toxicity of
the oxidized oil21). Among the low-molecular-weight com-
pounds, 4-hydroperoxy-2-en-1-al with 5-9 carbons was
reported to have the strongest toxicity21). Gabriel et al.22)
gave Wistar rats fresh olive oil, and distillable fractions of
fresh and thermally oxidized olive oil, respectively, and
found out that only the rats that received the distillable
fraction of thermally oxidized olive oil showed overt symp-
toms of heated fat toxicity. This was reflected in the histo-
logical scores of these animals, with the liver sustaining
the most numerous and severe lesions. These reports sup-
port our speculation that cytotoxicity of thermally oxidized
oil was attributable to low-molecular-weight decomposition
products.
Leung et al.23) tested food samples for acrylamide by an
LC-MS method and found high levels in all kinds of crisps.
But we did not detect it in used frying oil24). Velasco et al.25)
investigated the formation of monoepoxy fatty acids aris-
ing from oleic and linoleic acids in olive oil and sunflower
oil. Their results showed that the monoepoxides constitut-
ed a major group among the oxidized fatty acid monomers
formed at a high temperature. The content of monoepox-
ides in used frying oils from restaurants and fried-food out-
lets in Spain was found to range from 3.37 to 14.42 mg/g of
oil. These authors, however, did not evaluate the charac-
teristic cytotoxicity of monoepoxides.
ACKNOWLEDGMENTS
This work was partly supported by a grant from the Life
Science Center of Kobe-Gakuin University.
References
1. Sanz, N.; Dienz-Fernandez, C.; Alvarez, A.; Cascales,
M. Age-dependent modification in rat hepatocyte
antioxidant defence system. J. Hepatology 27, 525-534
(1997).
2. Cortesi, R.; Privett, O.S. Toxicity of fatty ozonides and
peroxides. Lipids 7, 715-721 (1971).
3. Nakatsugawa, K.; Kaneda T. Absorption and
metabolism of methyl linoleate hydroperoxides in rats.
Yukagaku 32, 361-366 (1983).
4. Itakura, S. Shishitu no kagaku, Asakura-shoten,
Tokyo, pp164-167 (2001).
5. Ohyanagi, Y.; Inoue, M. Kassei-sanso to Rouka-gensyo,
Kyoritu-shuppan, Tokyo, pp48-52 (2001).
6. Totani, N.: Thermal deterioration of edible oil used in
food manufacturing companies. Oleoscience 6, 493-500
(2006).
7. Gotoh, N.; Wada, S. The importance of peroxide value
in assessing food quality and food safety. J. Am.Oil
Chem. Soc. 83, 473-474 (2006).
8. Totani, N.; Ohno C.; Yamaguchi, A. Is the frying oil in
deep-fried foods safe? J. Oleo Sci.55, 449-456 (2006).
9. Soriguer, F.; Rojo-Martinez, G.; Dobarganes, M.C.; Gar-
cia- Almeida J.M.; Esteva, I.; Beltran, M.; Ruiz De
Adana, M.S.; Tinahones, F.; Gomez-Zumaquero, J.M.;
Garcia-Fuentes, E.; Gonzalez-Romero, S. Hypertension
is related to the degradation of dietary frying oils. Am.
J. Clin. Nutr. 78, 1092-1097 (2003).
10. Sanchez-Muniz, F.J.; Bastida, S. Frying oil discarding:
polar content vs. oligomer content determinations.
Forum Nutr. 56, 345-347 (2003).
11. National Survey of Health and Nutrition, Japan
(2000).
12. National Survey of Health and Nutrition, Japan
(1999).
13. Totani, N.; Satoh, K.; Tsuji, S.; Yamaguchi, A. Effects of
deteriorated frying oil in SHR/NDmc-cp rats. J. Oleo
Sci. 55, 283-289 (2006).
14. Totani, N.; Satoh, K.; Tsuji S.; Yamaguchi, A. Effects of
deteriorated frying oil in Wistar rats. J. Oleo Sci. 55,
291-297 (2006).
15. Totani, N.; Ojiri, Y. Mild ingestion of used frying oil
damages hepatic and renal cells in Wistar rats. J. Oleo
Sci. 56, 261-267 (2007).
16. Totani, N.; Ojiri, Y. Thermal deterioration of oil and
frying foodstuffs. J. Oleo Sci. 56, 543-551 (2007).
17. Totani, N.; Kuzume, T.; Yamaguchi, A.; Takada, M.;
Moriya, M. Amino acids brown oil during frying. J.
Oleo Sci. 55, 441-447 (2006).
18. Totani, N. A small reduction in atmospheric oxygen
decreases thermal deterioration of oil during frying. J.
Oleo Sci. 55, 135-141 (2006).
19. Totani, N.; Yamaguchi, A.; Takada, M.; Moriya, M.
Color deterioration of oil during frying. J. Oleo Sci. 55,
51-57 (2006).
20. Totani, N.; Ono, M.; Munkhjargal, B.; Ojiri, Y. Carbonyl
compounds vaporize from oil with steam during deep-
frying. J. Oleo Sci. 56, 449-456 (2007).
21. Yoshioka, M.; Kaneda, T. Studies on the toxicity of the
autoxidized oils. I. The fractionation of the toxic com-
pound and its identification. Yukagaku 21, 316-321
(1972).
22. Gabriel, H.G.; Alexander, J.C.; Valli, V.E. Nutritional and
metabolic studies of distillable fractions from fresh and
159
J. Oleo Sci.57, (3) 153-160 (2008)
N. Totani, M. Burenjargal , M. Yawata and Y. Ojiri
thermally oxidized corn oil and olive oil. Lipids 13, 49-
55 (1978).
23. Leung, K.S.; Lin, A.; Tsang, C.K.; Yeung, S.T. Acry-
lamide in asian foods in Hong Kong. Food Addit. Con-
tam. 20, 1105-1113 (2003).
24. Totani, N.; Yawata, M.; Takada, M.; Moriya, M. Acry-
lamide content of commercial frying oil. J. Oleo Sci.
56, 103-106 (2007).
25. Velasco, J.; Marmesat, S.; Bordeaux, O.; Marquez-Ruiz,
G.; Dobarganes, C. Formation and evolution of monoe-
poxy fatty acids in thermoxidized olive and sunflower
oils and quantification in used frying oils from restau-
rants and fried-food outlets. Food Chem. 52, 4438-4443
(2004).
160
J. Oleo Sci.57, (3) 153-160 (2008)
