Oxidative stability of olive oil–lemon juice salad dressings
stabilized with polysaccharides
D. Paraskevopoulou, D. Boskou, A. Paraskevopoulou
Laboratory of Food Chemistry and Technology, School of Chemistry, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece
Received 26 September 2005; received in revised form 16 January 2006; accepted 13 March 2006
Lipid oxidation is a major cause of quality deterioration in food emulsions. Polysaccharides used to improve emulsion stability and
texture may also aﬀect lipid oxidation. In the present study, the oxidative stability of olive oil–lemon juice salad dressings, stabilized with
gum arabic or propylene glycol alginate in admixture with xanthan, was investigated. Oil-in-water emulsions (50:50, v/v) were prepared
with lemon juice and extra virgin olive oil and then homogenized at various homogenization rates to form diﬀerent particle sizes. Kee-
pability was followed by storing at room temperature for 6–8 months and measuring the formation of primary and secondary oxidation
products. The shelf life was compared to that of the bulk olive oil. It was shown that the polysaccharides had the ability to inhibit lipid
oxidation, probably due to their amphiphilic character (gum arabic and propylene glycol alginate) as well as their ability to induce vis-
cosity increase. Olive oil–lemon juice emulsions were also assessed for consumer acceptance. The panellists were asked to smell the sam-
ples and rate them according to rancidity using a four-point (1 = no perception, 4 = extreme) intensity scale. The results were in
accordance to those of chemical analysis. Lipid oxidation was not aﬀected by the oil droplet size, as demonstrated by peroxide value
measurements and sensory evaluation.
Ó2006 Elsevier Ltd. All rights reserved.
Keywords: Salad dressing; Olive oil; Lemon juice; Gum arabic; Propylene glycol alginate; Xanthan; Oxidation; Keepability; Droplet size
Emulsions are thermodynamically unstable systems
and they tend to separate into two layers over time
through a number of mechanisms (Dickinson, 1992;
McClements, 1999). In order to make emulsions kineti-
cally stable for a reasonable period of time, emulsiﬁers
and/or stabilizers (e.g. proteins, phospholipids, polysac-
charides) must be added. Polysaccharides are usually
added to o/w emulsions to enhance viscosity of the aque-
ous phase, thus preserving desirable textural characteris-
tics and stabilizing oil droplets against creaming
(Dickinson, 1992; Paraskevopoulou et al., 2003; Paraskev-
opoulou, Boskou, & Kiosseoglou, 2005; Paraskevopou-
lou, Kiosseoglou, Alevisopoulos, & Kasapis, 1997).
Amphiphilic polysaccharides, such as gum arabic and
propylene glycol alginate, act both as emulsiﬁers and sta-
bilizers. Because of their surface-active properties they can
adsorb onto the surface of freshly formed droplets during
homogenization and prevent them from aggregation by
steric and/or electrostatic forces. Studies have shown that
polysaccharides are also capable of retarding lipid oxida-
tion in o/w emulsions (Matsumura et al., 2003; Shimada,
Fujikawa, Yahara, & Nakamura, 1992; Shimada et al.,
1994; Shimada, Okada, Matsuo, & Yoshioka, 1996).
The polysaccharide-induced viscosity increase of the aque-
ous phase inhibits oxygen diﬀusion and slows down the
movement of oil droplets, thus reducing their collision
probability. Additionally, the metal ion-chelating ability
of polysaccharides has been proposed to account for their
ability to inhibit lipid oxidation, while tragacanth gum
has been found to act as a radical chain-breaker because
of its ability to donate hydrogen (Shimada et al., 1992).
0308-8146/$ - see front matter Ó2006 Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +30 231 997832; fax: +30 231 997779.
E-mail address: email@example.com (A. Paraskevopoulou).
Food Chemistry 101 (2007) 1197–1204
Lipid oxidation is considered to be the predominant
cause of quality deterioration of oils, fats and various
fat-containing foods during storage. It involves the interac-
tion between unsaturated lipids and oxygen-active species.
It can cause alterations in the quality characteristics of
foods, such as appearance, taste, texture, shelf life and
nutritional proﬁle and lead to the development of undesir-
able oﬀ-ﬂavour (Min & Boﬀ, 2002). Numerous studies have
been carried out dealing with oxidation mechanisms in
bulk oils although, in most processed foods, lipids are
found as o/w or w/o emulsions. The oxidation of emulsiﬁed
lipid is mechanistically diﬀerent from that of bulk oils. In
such products, the organization of the lipid molecules
within the system and their interactions with other food
components inﬂuences their susceptibility to lipid oxida-
tion. A great number of studies have highlighted the impact
of various factors in determining the oxidative stability of
an oil-in-water emulsion, including chemical structure of
lipids, oxygen concentration, antioxidants, interfacial char-
acteristics, droplet characteristics (concentration, size),
interactions with aqueous phase components (salts, sugars,
polysaccharides, proteins.) and the presence of prooxidants
(e.g. transition metal impurities) (McClements & Decker,
Salad dressings are very popular oil-in-water emulsion
products, which may vary in fat content (20–65%) and vis-
cosity (Dickinson & Stainsby, 1982). ‘‘Greek salad dress-
ings’’ are, by deﬁnition, mixtures of virgin olive oil and
lemon juice instantly prepared before use (Paraskevopou-
lou et al., 2005). Due to their composition, they are a good
source of biophenols (Boskou & Visioli, 2003) as well as
lipid-soluble and water-soluble vitamins (tocopherols, b-
carotene, ascorbic acid). Besides, virgin olive oil, thanks
to its balanced fatty acid composition, has highly appreci-
ated nutritional characteristics, known for a long time to
the people of the Mediterranean region, who use it daily
for a variety of culinary purposes (Garcia Mesa, Jimenez-
Marquez, Beltran-Maza, & Friaz-Ruiz, 1998; Romero,
Guesta, & Sanchez-Muniz, 1998). Its consumption has also
increased in non-Mediterranean areas because of the grow-
ing interest in the Mediterranean diet and the belief that it
prevents certain diseases (Boskou & Visioli, 2003; Tricho-
The development of a ‘‘Greek salad dressing’’ contain-
ing olive oil and lemon juice that would exhibit reasonable
physicochemical stability over prolonged storage has been
the subject of a previous work (Paraskevopoulou et al.,
2005). Diﬀerent combinations of xanthan gum with gum
arabic or propylene glycol alginate exhibited a positive
eﬀect on the creaming and rheological behaviour, rate of
oil droplet coalescence and sensory properties of the salad
dressings. Gum arabic (gum Acacia), a hydrocolloid pro-
duced by the natural exudation of acacia trees, is a complex
mixture of six main carbohydrate components and a small
but functionally important amount of protein (2% w/w),
which is found as an integral part of the structure (Ware-
ing, 1999). Propylene glycol alginate, an esteriﬁed form of
alginic acid, is appreciated as a particularly eﬀective thick-
ener in several acidic food applications, such as salad dress-
ings and lactic drinks (Onsøyen, 1999). Xanthan gum is an
extracellular polysaccharide made by the bacterium Xan-
thomonas campestris (Urlacher & Noble, 1999).
The favourable eﬀect of the above-mentioned polysac-
charides, as well as the role of the oil droplet size on the
keepability of ‘‘Greek salad dressings’’, was the objective
of this work. By modifying the emulsiﬁcation energy, emul-
sions with diﬀerent droplet sizes were prepared. Keepabil-
ity was measured by periodically measuring peroxide
value, acidity and speciﬁc extinction coeﬃcient K
thermore, the sensory perception of lipid oxidation prod-
ucts of the emulsions by the consumers was assessed.
2. Materials and methods
Gum arabic (GA) and xanthan gum (X) were purchased
from Sigma Chemical Co. (USA). Propylene glycol algi-
nate (PGA) (degree of esteriﬁcation 40–60%) was kindly
supplied by FMC Biopolymer (Brussels, Belgium). Extra
virgin olive oil was bought from the local market. Lemon
juice was a commercial sample purchased in a local super-
market. Egg yolk (EY) was obtained by ﬁrst breaking fresh
hen’s eggs and, following complete removal of the adhering
white by rolling the intact yolks on tissue paper, the vitel-
line membrane was punctured and the liquid yolks of a
number of eggs were collected. Benzoic acid was obtained
from Riedel de Hae
2.2. Preparation of salad dressings
Oil-in-water emulsions (50% v/v) were prepared as fol-
lows: a lemon juice polysaccharide solution was ﬁrst
prepared by slowly dispersing 1% w/v gum arabic (or
1% w/v PGA) and 0.5% w/v xanthan gum, with stirring
for at least 6 h, to ensure complete dissolution. The
emulsions were prepared by adding dropwise, while mix-
ing with a propeller-type mechanical stirrer, 50 ml of
virgin olive oil to 50 ml of the polysaccharide solution.
The droplet size of the resulting crude emulsion was then
reduced further using an Ultra-Turrax T25 homogenizer
(IKA Instruments, Germany), equipped with an
S25KG-25F dispersing tool. By modiﬁcation of the
energy of emulsiﬁcation three gum arabic/xanthan-stabi-
lized (GA/X 9500 rpm, GA/X 13,500 rpm, GA/X
24,000 rpm) and three propylene glycol alginate/xan-
than-stabilized (PGA/X 8000, PGA/X 9500, PGA/X
20,500 rpm) emulsions with diﬀerent oil droplet sizes,
were produced. Thus, it was possible to test the eﬀect
of droplet size and the eﬀect of stabilizing agent on the
oxidative stability of the produced Greek-type salad
dressings. Benzoic acid was added at a concentration of
1&w/v in the continuous phase. The pH values of the
ﬁnal samples ranged between 3.4 and 3.6.
1198 D. Paraskevopoulou et al. / Food Chemistry 101 (2007) 1197–1204
For comparison, another olive oil-in-lemon juice salad
dressing (50:50 v/v) containing 1% w/v egg yolk solids, in
the place of GA or PGA, was prepared by the same emul-
siﬁcation procedure, following homogenization at
13,500 rpm for 2 min.
Olive oil samples and salad dressings (100 ml) were
stored in screw-capped glass containers (100 ml) prior to
2.3. Droplet size determination
The droplet size distribution of the salad dressings over
the storage period was measured using a laser light scatter-
ing instrument (Malvern Instruments, UK). Measurements
are either reported as the full particle size distribution or as
the surface-volume mean diameter d3=2¼Pnid3
is the number of droplets of radius d
. To prevent
multiple scattering eﬀects, the emulsions were diluted with
deionized water prior to analysis so that the droplet con-
centration was less than about 0.02% v/v. The dilute emul-
sions were placed directly into the measurement cell of the
instrument and stirred slowly during the measurement.
Each sample was analyzed 4 times and the data are pre-
sented as averages. Droplet sizes were checked periodically
to monitor emulsion stability.
2.4. Viscosity measurements
The viscosities of the salad dressings were determined at
25 °C at various shear rates (from 0.1 to 20.4 s
) with the
aid of a Brookﬁeld DV-II, LV viscometer (USA), equipped
with concentric cylinder geometry. The ﬂow curves giving
viscosity g(MPa s) as a function of shear rate _
characteristic of shear-thinning behaviour.
2.5. Oxidative stability evaluation
2.5.1. Experimental design
Twenty containers of each emulsion were stored in
conditions similar to those in consumers’ sales points:
inside a chamber at 23 °C and placed in shelf units,
2.5 m high with illumination for 12 h/day. Containers
were sampled every three or four weeks for complete
3.5–6 months storage.
Furthermore, 100 ml of virgin olive oil were stored
under the same conditions and sampled in a similar way.
2.5.2. Analytical determinations
Free acidity, peroxide value (PV) and speciﬁc extinction
were determined according to the analytical
methods described in Regulation EC/2568/91 of the Com-
mission of the European Union (EU, 1991). Oil was sepa-
rated from the salad dressings by repeated freeze–thaw
cycles, followed by centrifugation (Jacobsen et al., 1999).
Two containers of each sample were independently ana-
lyzed in each sampling and each parameter was measured
twice. The results are expressed as mean values.
Free acidity, given as % oleic acid, was determined by
titration of a solution of oil in ethanol/ether (1:1) with eth-
Peroxide value, expressed as milliequivalents of active
oxygen per kilogramme of oil (meq/kg), was determined
as follows: a mixture of oil and chloroform/acetic acid
was left to react with a solution of potassium iodide in
darkness; the free iodine was then titrated with a sodium
extinction coeﬃcient was calculated from the
adsorption of a solution of the oil in iso-octane at
232 nm, using a UV spectrophotometer (Hitachi, Japan)
and a path length of 1 cm.
2.6. Organoleptic assessment
Organoleptic evaluation was performed for salad dress-
ings and extra virgin olive oil, following their storage time,
simultaneously with oxidative stability measurements. The
analytical panel, consisting of 20 untrained members, was
asked to make its evaluation on the basis of rancidity.
The panellists were students and members of the staﬀ of
the Laboratory and were asked to score the samples by
smelling them. Rancidity grading was based upon a four-
point intensity scale: 1 = no perception; 2 = weak; 3 =
medium; 4 = extreme (Jellinek, 1984). The dressings and
olive oil samples were served in glass beakers (20 ml in
each beaker) at room temperature.
2.7. Statistical analysis
All experiments were performed on quadruplicate sam-
ples. Statistical analyses were conducted with the one-way
ANOVA software package. Signiﬁcant diﬀerences among
means (p< 0.05) were determined by LSD test.
3. Results and discussion
3.1. Properties of the salad dressings
Prepared salad dressings were monomodal with the drop-
let size distributions overlapping one another, as shown in
Fig. 1. The emulsions prepared with a mixture of GA/X
were physicochemically stable over the storage period as
no coalescence and no droplet size or viscosity variations
were observed over the time of experiments (Table 1,
Fig. 2). On the other hand, the droplet size of the rest of
the emulsions (prepared with a mixture of PGA/X or EY/
X) increased signiﬁcantly (p< 0.05) during storage. From
the droplet size data obtained by the laser diﬀraction, it
was observed that GA/X-stabilized emulsions had the low-
est droplet sizes, likely due to the fact that gum arabic is
more surface-active than PGA. The larger droplet size of
EY/X-stabilized emulsions could be attributed to their rela-
tively low egg yolk content (1% w/v egg yolk solids).
Finally, their viscosity remained stable throughout the
storage period (Fig. 2).
D. Paraskevopoulou et al. / Food Chemistry 101 (2007) 1197–1204 1199
3.2. Acidity of the lipid phase
The salad dressings showed an increase of acidity
with the storage time that became signiﬁcant (p< 0.05)
at the end of the storage period (Fig. 3). This increase
was attributed to the migration of acids/acidic com-
pounds from lemon juice to olive oil droplets during their
3.3. Oxidative stability of salad dressings
3.3.1. Quality characteristics of olive oil sample
The quality characteristics that inﬂuence the storage
stability of olive oil were determined before the preparation
of salad dressings and are shown in Table 2. Acidity, per-
oxide value and coeﬃcients of speciﬁc extinction (K
) were determined. All values were lower than the lim-
its set by the EU Regulation 2568/91 for extra virgin olive
0.1 1 10 100 1000 10000
Fig. 1. Droplet size distribution of olive oil–lemon juice emulsions
prepared at various energies of emulsiﬁcation and stabilized with GA/X
(j,m,d), PGA/X (h,n,s) and EY/X ( ) mixtures 24 h after prepara-
tion. Emulsions prepared with: low energy input (j,h); medium energy
input (m,n); high energy input (d,s).
Eﬀect of energy of emulsiﬁcation on the mean droplet diameter (d
) increase with time of olive oil–lemon juice salad dressings stabilized with mixtures of
PGA/X, GA/X or EY/X
t(days) Propylene glycol alginate/xanthan Gum arabic/xanthan Egg yolk/xanthan
8000 rpm 9500 rpm 20,500 rpm 9500 rpm 13,500 rpm 24,000 rpm 13,500 rpm
160 – – – 6.1
For each column, diﬀerent superscripts indicate signiﬁcant diﬀerences (p< 0.05) among mean droplet diameters.
0 50 100 150 200
Fig. 2. Shear viscosity (at 2 s
,25°C) of olive oil–lemon juice emulsions
prepared with GA (j,m,d), PGA (h,n,s)orEY( ) 1% (w/v) (on a dry
basis) and xanthan 0.45% (w/v) as a function of storage time. Emulsions
prepared with: low energy input (j,h); medium energy input (m,n); high
energy input (d,s).
0 50 100 150 200
(% oleic acid)
Fig. 3. Changes in acidity over storage time at 23 °C in virgin olive oil
sample and olive oil–lemon juice emulsions prepared with GA (j,m,d),
PGA (h,n,s)orEY( ) 1% (w/v) (on a dry basis) and xanthan 0.45%
(w/v) as a function of storage time. Emulsions prepared with: low energy
input (j,h); medium energy input (m,n); high energy input (d,s).
1200 D. Paraskevopoulou et al. / Food Chemistry 101 (2007) 1197–1204
3.3.2. Primary oxidation
Lipid oxidation in oil-in-water emulsions is taking place
at the surface of the oil droplets. Many studies suggest that
the interaction between lipid hydroperoxides, the ﬁrst
products formed by oxidation, located at the droplet sur-
face and transition metals, from the aqueous phase, is the
most common cause of oxidative instability (Mei, McCle-
ments, & Decker, 1998a; Mei, McClements, & Decker,
1998b). The most likely mechanism is the decomposition
of lipid hydroperoxides (ROOH) by the pro-oxidants into
highly reactive peroxyl (ROO
) and alkoxyl (RO
which react with unsaturated lipids within the droplets or
at the oil–water interface, leading to the formation of lipid
radicals. The lipid oxidation chain reaction propagates as
these lipid radicals react with other lipids in their immedi-
Fig. 4 shows the storage stability of virgin olive oil and
GA/X-, PGA/X- and EY/X-stabilized olive oil/lemon juice
salad dressings. Oxidation was more rapid in olive oil sam-
ple than in salad dressings. The rate of hydroperoxide for-
mation increased sharply in olive oil after an induction
period of thirty (30) days. The polysaccharide- and egg
yolk-stabilized emulsions exhibited a signiﬁcantly lower
peroxide value (PV) (p< 0.05) than that of the oil through-
out the storage period, which indicated the eﬀectiveness of
the above emulsiﬁers/stabilizers in the keepability of the
salad dressings. At the time (100 days) olive oil reached a
peroxide value of 75 meq O
/Kg oil, GA/X- and EY/X-sta-
bilized salad dressings had seven times lower peroxide val-
ues while PGA/X-stabilized emulsions four times.
The viscosity increase of the continuous phase (lemon
juice) and hence the slowing down of the movement of the
reactants present in it may account for their ability to retard
lipid oxidation. The eﬀect of polysaccharide addition on the
viscosity of lemon juice/olive oil emulsions has already been
examined (Paraskevopoulou et al., 2005). Measurement of
viscosity of the salad dressings at various stages of storage
indicates that increased viscosity contributes to an increased
stability to lipid oxidation (Fig. 2).
3.3.4. Droplet surface
Additionally, the presence of GA and PGA at the oil
droplet surface, due to their surface-active parts in their
molecules, may also contribute to the oxidative stability
of the salad dressings. Lipid hydroperoxides are surface-
active and therefore migrate to and concentrate at the
surface of the emulsion droplets, being susceptible to inter-
actions with aqueous phase oxidation catalysts, such as
iron (Nuchi, Hernandez, McClements, & Decker, 2002).
Proteins, as well as polysaccharides, are often used in food
emulsions to stabilize droplets against ﬂocculation or coa-
lescence and to improve their textural properties (Dickin-
son, 1992; McClements, 1999). According to McClements
and Decker (2000), adsorbed emulsiﬁers are likely to be
particularly eﬀective at retarding lipid oxidation because
of the ability of the interfacial membrane to act as a phys-
ical barrier that separates lipid substrates from pro-oxi-
dants in the aqueous phase. Additionally, certain types of
emulsiﬁer molecules, containing either sugar or amino acid
moieties (e.g. gum arabic, soluble soybean polysaccharide
and proteins), may act as a chemical barrier to lipid oxida-
tion by scavenging free radicals (Matsumura, Satake,
Egami, & Mori, 2000; McClements & Decker, 2000). The
inhibiting eﬀect of peptide-bound polysaccharide GA
against lipid oxidation in methyl oleate or methyl linoleate
emulsions emulsiﬁed by b-casein or other surfactants
(sugar ester S-1670, Tween 20) has been reported by Mat-
sumura et al. (2000, 2003). Another mechanism that may
account for the capability of the three polysaccharides to
inhibit lipid oxidation is metal ion chelation. Xanthan
has been shown to suppress oil peroxidation by inactiva-
tion of Fe
ions, present in the system (Karadjova,
Zachariadis, Boskou, & Stratis, 1998), due to its ability
to chelate metal ions at negatively charged pyruvate sites
(Shimada et al., 1992, 1994).
3.3.5. Emulsiﬁer/stabilizer type
As can be seen (Fig. 4), hydroperoxide formation in the
salad dressings was signiﬁcantly aﬀected by the emulsiﬁer/
stabilizer type used for their preparation. During the ﬁrst
days of storage all the emulsions appeared to be equally
Quality characteristics of the virgin olive oil used in the experiments
Quality characteristics Value Quality limits
Acidity (percentage oleic acid) 0.49 61%
Peroxide value (meq O
/kg of oil) 9.65 620
[where K = absorbance/C (g/100 ml oil)] 1.72 62.50
[where K = absorbance/C (g/100 ml oil)] 0.20 60.20
Set by the EU Regulation 2568/91.
0 50 100 150 200
(meq O2/kg oil)
Fig. 4. Changes in peroxide values (PVs) over storage time at 23 °Cin
virgin olive oil sample (r) and olive oil–lemon juice emulsions prepared
with GA (j,m,d), PGA (h,n,s)orEY( ) 1% (w/v) (on a dry basis)
and xanthan 0.45% (w/v) as a function of storage time. Emulsions
prepared with: low energy input (j,h); medium energy input (m,n); high
energy input (d,s).
D. Paraskevopoulou et al. / Food Chemistry 101 (2007) 1197–1204 1201
eﬀective at stabilizing olive oil droplets. After 20 days of
storage, GA/X and EY/X were found to be more eﬀective
than PGA/X since both are better emulsiﬁers than PGA.
Their ability to protect lipid hydroperoxides is likely due
to the ability of these molecules to adsorb at the oil droplet
surface, forming a ﬁlm of high surface shear viscosity.
3.3.6. Oil droplet size
Oil droplet size inﬂuence on hydroperoxide formation
rates is also seen in Fig. 4. For a ﬁxed oil droplet concen-
tration (50% in our experiments), an increased rate of oxi-
dation was expected as the droplet size decreased, because
of the increased surface area that was exposed to the aque-
ous phase. However, no dependence of the lipid oxidation
rate on droplet size of olive oil/lemon juice emulsions was
observed at any point of time during their storage. This
was explained by the limited amounts of lipid hydroperox-
ides (10 meq O
/kg oil on preparation day) that were
available in the emulsion systems and might have been
present at the droplet surface. Other researchers have
reported similar droplet eﬀects on lipid oxidation (Osborn
& Akoh, 2004; Roozen, Frankel, & Kinsella, 1994). If there
was a high concentration of hydroperoxides in the systems,
then increasing the surface area by decreasing the oil drop-
let size might increased their concentration at the interface,
leading to the expected increase of lipid oxidation with
decreasing droplet size.
3.3.7. Oxidation products
During lipid oxidation, a number of decomposition reac-
tions occur simultaneously, that in turn result in the gener-
ation of a wide variety of diﬀerent molecules, including
aldehydes, ketones, alcohols and hydrocarbons. These oxi-
dation products are responsible for the characteristic phys-
icochemical and sensory properties of oxidized oils, since
they are likely to be more surface-active than the initial lipid
and some of them are water-soluble. The measurement of
speciﬁc extinction coeﬃcient was used to determine
the level of conjugated dienes present in the emulsiﬁed oil. It
is known that the oxidation products of oils and fats, which
may result from their decomposition, display characteristic
spectra in the ultraviolet region and at about 232 nm.
Therefore, a determination of the absorbance at 232 nm is
an indication of the state of oxidation of a fat.
coeﬃcient of speciﬁc extinction showed a sig-
niﬁcantly (p< 0.05) higher initial value for olive oil–lemon
juice salad dressings stabilized with both GA (3.2) and
PGA (4.0), in comparison to olive oil (1.75). This is
probably due to carbonyl compounds present in lemon
juice being diﬀerent from those due to oil rancidity (Fig. 5).
coeﬃcient in virgin olive oil increased with stor-
age time and the limit of 2.50 was exceeded after 100 days of
storage. In salad dressings, on the other hand, the K
ﬁcient remained practically constant during the storage per-
iod, indicating a delay in the formation of oxidation
products. This was expected because of the delayed devel-
opment of hydroperoxides in the emulsions (Fig. 4). Emul-
sions stabilized with a mixture of PGA/X exhibited K
values (4.0) signiﬁcantly higher than the corresponding
ones for GA/X-stabilized emulsions (3.5). Likewise, EY/
X-stabilized emulsions registered a constant value for K
(3.4) throughout the storage period. This was probably
due to the fact that egg yolk proteins are below their isoelec-
tric point in the pH (3.5) of the emulsion systems, thereby
producing positively charged emulsion droplets. As Osborn
and Akoh (2003) have pointed out, at this pH value, the
positively charged metal ions can not bind to the emulsion
droplets, which explains the decreased oxidation rates in the
EY/X-stabilized salad dressings.
3.4. Sensory studies
Sensory evaluations have been carried out to character-
ize the salad dressings and to ﬁnd a correlation with chem-
ical analysis. The sensory evaluation of the salad dressings
appears to be necessary since the increased shelf life of
these products by the addition of polysaccharides, such
as GA and PGA, is futile if sensory properties are unac-
ceptable to consumers. For this reason, a questionnaire
was conducted to detect the acceptability of lemon juice–
olive oil emulsions (4-point intensity scale: 1 = no percep-
tion; 2 = weak; 3 = medium; 4 = extreme). The results
are presented in Table 3. Storage of virgin olive oil at
23 °C revealed the appearance of rancid odour (2.33 ±
1.07) in almost 9 weeks. This was also indicated by the per-
oxide value and K
measurements. Lower values for ran-
cidity were measured in all emulsions compared to olive oil
even after 33 weeks.
On the other hand, in the emulsions stabilized with
PGA/X, the intensity of rancidity was not changed signif-
icantly during storage (1.33 ± 0.65–1.58 ± 0.90 after 15
0 50 100 150 200
Fig. 5. Changes in speciﬁc extinction coeﬃcient values (K
) over storage
time at 23 °C in virgin olive oil sample (r) and olive oil–lemon juice
emulsions prepared with GA (j,m,d), PGA (h,n,s)orEY( ) 1% (w/
v) (on a dry basis) and xanthan 0.45% (w/v) as a function of storage time.
Emulsions prepared with: low energy input (j,h); medium energy input
(m,n); high energy input (d,s).
1202 D. Paraskevopoulou et al. / Food Chemistry 101 (2007) 1197–1204
weeks), whereas in the EY/X or GA/X-stabilized emulsions
the panellists reported a weak perception of rancid ﬂavour
only after 21 weeks (2.00 ± 0.93 and 1.63 ± 0.75–
2.25 ± 1.28, respectively). Finally, the statistical analysis
between salad dressings of similar initial droplet size, stabi-
lized with mixtures of GA/X or PGA/X and homogenized
at 20,500 rpm and 9500 rpm, respectively, revealed no sig-
niﬁcant diﬀerences among them.
It can be concluded that the polysaccharides, gum ara-
bic and propylene glycol alginate, have the ability to inhibit
lipid oxidation in addition to their stabilizing/emulsifying
capacity. This is probably due to their amphiphilic charac-
ter and partly due to the presence of xanthan. Gum arabic,
in particular, more eﬀectively suppressed the oxidation dur-
ing the storage period. This is attributed to its better sur-
face-active properties in comparison to propylene glycol
alginate. Lipid oxidation was not aﬀected by the oil droplet
size, as demonstrated by peroxide value measurements and
Boskou, D., & Visioli, F. (2003). Biophenols in olive oil and olives. In M.
Pilar Vaquero, T. Garcia-Arias, A. Carbajal, & F. J. Sanchez-Muniz
(Eds.), Bioavailability of micronutrients and minor dietary compounds.
Metabolic and technological aspects (pp. 161–169). Kerala: Research
Dickinson, E. (1992). An introduction to food colloids. Oxford: Oxford
Dickinson, E., & Stainsby, G. (1982). Colloids in food. London: Applied
EU. (1991). Regulation 2568. Characteristics of olive oil and olive residue
oil and the relevant methods of analysis. Oﬃcial J. L. 248.
Garcia Mesa, J. A., Jimenez-Marquez, A. J., Beltran-Maza, G., & Friaz-
Ruiz, L. (1998). Thermooxidation of diﬀerent vegetable oils used in
deep frying on dietary fat intake. Nutrition Reviews, 49, 375–376.
Jacobsen, C., Hartvigsen, K., Lund, P., Meyer, A. S., Adler-Nissen, J.,
Holstborg, J., et al. (1999). Oxidation in ﬁsh-oil-enriched mayonnaise.
1. Assessment of propyl gallate as an antioxidant by discriminant
partial least squares regression analysis. European Food Research
Technology, 210, 13–30.
Jellinek, G. (1984). Sensory evaluation of food. Chichester: Ellis Horwood.
Karadjova, I., Zachariadis, G., Boskou, G., & Stratis, J. (1998).
Electrothermal atomic adsorption spectrometric determination of
aluminium, cadmium, chromium, copper, iron, manganese, nickel
and lead in olive oil. Journal of Analytical Atomic Spectrometry, 13(3),
Matsumura, Y., Egami, M., Satake, C., Maeda, Y., Takahashi, T.,
Nakamura, A., et al. (2003). Inhibitory eﬀects of peptide-bound
polysaccharides on lipid oxidation in emulsions. Food Chemistry,
Matsumura, Y., Satake, C., Egami, M., & Mori, T. (2000). Interaction of
gum arabic, maltodextrin and pullulan with lipids in emulsions.
Bioscience Biotechnology and Biochemistry, 64(9), 1827–1835.
McClements, D. J., & Decker, E. A. (2000). Lipid oxidation in oil-in-water
emulsions: impact of molecular environment on chemical reactions in
heterogeneous food systems. Journal of Food Science, 65(8),
McClements, D. J. (1999). Food emulsions: principles, practice and
techniques. Boca Raton: CRC Press.
Sensory scores for rancidity for salad dressings and virgin olive oil during storage at 23 °C (sensory scale 1–4; mean ± SD)
Virgin olive oil Egg yolk/xanthan
Storage period (weeks) Storage period (weeks)
1.42 ± 0.51
2.33 ± 1.07
13,500 rpm 1.30 ± 0.47
1.47 ± 0.61
1.88 ± 0.89
2.00 ± 0.93
2.09 ± 0.94
Propylene glycol alginate/xanthan Gum arabic/xanthan
Storage period (weeks) Storage period (weeks)
06915 0 4 14 2133
8000 rpm 1.18 ± 0.39
1.33 ± 0.62
1.50 ± 0.63
1.58 ± 0.90
9500 rpm 1.30 ± 0.47
1.21 ± 0.42
1.50 ± 0.73
1.63 ± 0.75
2.09 ± 0.94
9500 rpm 1.35 ± 0.49
1.40 ± 0.63
1.44 ± 0.73
1.53 ± 0.72
13,500 rpm 1.48 ± 0.85
1.58 ± 0.77
2.25 ± 0.93
2.25 ± 1.28
2.45 ± 1.21
20,500 rpm 1.24 ± 0.44
1.47 ± 0.74
1.56 ± 0.96
1.33 ± 0.65
24,000 rpm 1.48 ± 0.67
1.11 ± 0.32
1.56 ± 0.63
1.75 ± 0.46
2.00 ± 0.89
a–c: Diﬀerent superscripts mean that the results in each row for each polysaccharide are signiﬁcantly diﬀerent (p< 0.05).
D. Paraskevopoulou et al. / Food Chemistry 101 (2007) 1197–1204 1203
Mei, L. Y., McClements, D. J., & Decker, E. A. (1998a). Iron-catalyzed
lipid oxidation in emulsion as aﬀected by surfactant, pH and NaCl.
Food Chemistry, 61(3), 307–312.
Mei, L. Y., McClements, D. J., & Decker, E. A. (1998b). Evidence of iron
association with emulsion droplets and its impact on lipid oxidation.
Journal of Agricultural and Food Chemistry, 46(12), 5072–5077.
Min, D. B., & Boﬀ, J. M. (2002). Lipid oxidation of edible oil. In C. C.
Akoh & D. B. Min (Eds.), Food lipids – chemistry, nutrition and
biotechnology (pp. 335–363). New York: Springer.
Nuchi, C. D., Hernandez, P., McClements, D. J., & Decker, E. A. (2002).
Ability of lipid hydroperoxides to partition into surfactant micelles and
alter lipid oxidation rates in emulsions. Journal of Agricultural and
Food Chemistry, 50(19), 5445–5449.
Onsøyen, E. (1999). Alginates. In A. Imeson (Ed.). Thickening and gelling
agents for food (Vol. 2, pp. 22–44). Gaithersburg: Aspen Publishers
Osborn, H. T., & Akoh, C. C. (2003). Copper-catalyzed oxidation of a
structured lipid-based emulsion containing a-tocopherol and citric
acid: inﬂuence of pH and NaCl. Journal of Agricultural and Food
Chemistry, 51(23), 6851–6855.
Osborn, H. T., & Akoh, C. C. (2004). Eﬀect of emulsiﬁer type, droplet
size, and oil concentration on lipid oxidation in structured lipid-based
oil-in-water emulsions. Food Chemistry, 84(3), 451–456.
Paraskevopoulou, A., Athanasiadis, I., Blekas, G., Koutinas, A. A.,
Kanellaki, M., & Kiosseoglou, V. (2003). Inﬂuence of polysaccharide
addition on stability of a cheese whey keﬁr–milk mixture. Food
Hydrocolloids, 17(5), 615–620.
Paraskevopoulou, A., Boskou, D., & Kiosseoglou, V. (2005). Stabilization
of olive oil–lemon juice emulsion with polysaccharides. Food Chem-
istry, 90(4), 627–634.
Paraskevopoulou, A., Kiosseoglou, V., Alevisopoulos, S., & Kasapis, S.
(1997). Small deformation properties of model salad dressings
prepared with reduced cholesterol egg yolk. Journal of Texture Studies,
Romero, A., Guesta, C., & Sanchez-Muniz, F. J. (1998). Behaviour of
extra virgin olive oil in potato frying: thermooxidative alteration of the
fat content in the fried food. Grasas Aceites, 49, 370–378.
Roozen, J. P., Frankel, E. N., & Kinsella, J. E. (1994). Enzymic and
autoxidation of lipids in low fat foods: model of linoleic acid in
emulsiﬁed hexadecane. Food Chemistry, 50(1), 33–38.
Shimada, K., Fujikawa, K., Yahara, K., & Nakamura, T. (1992).
Antioxidative properties of xanthan on the autoxidation of soybean
oil in cyclodextrin emulsion. Journal of Agricultural and Food
Chemistry, 40(6), 945–948.
Shimada, K., Muta, H., Nakamura, Y., Okada, H., Matsuo, K.,
Yoshioka, S., et al. (1994). Iron-binding property and antioxidative
activity of xanthan on the autoxidation of soybean oil in emulsion.
Journal of Agricultural and Food Chemistry, 42(8), 1607–1611.
Shimada, K., Okada, H., Matsuo, K., & Yoshioka, S. (1996). Involvement
of chelating action and viscosity in the antioxidative eﬀect of xanthan
in an oil/water emulsion. Bioscience, Biotechnology and Biochemistry,
Trichopoulou, A. (1995). Olive oil and breast cancer. Cancer Causes
Control, 6, 475–476.
Urlacher, B., & Noble, O. (1999). Xanthan gum. In A. Imeson (Ed.).
Thickening and gelling agents for food (Vol. 13, pp. 284–311).
Gaithersburg: Aspen Publishers Inc.
Wareing, V. M. (1999). Exudate gums. In A. Imeson (Ed.). Thickening
and gelling agents for food (Vol. 5, pp. 86–99). Gaithersburg: Aspen
1204 D. Paraskevopoulou et al. / Food Chemistry 101 (2007) 1197–1204