Lipid Oxidation Measurement Method

8Lipid Oxidation:

Measurement Methods

Fereidoon Shahidi and Ying Zhong

Memorial University of Newfoundland,

St. John’s, Newfoundland, Canada

1. INTRODUCTION

Dietary lipids, naturally occurring in raw food materials or added during food

processing, play an important role in food nutrition and flavor. Meanwhile, lipid

oxidation is a major cause of food quality deterioration, and has been a challenge

for manufacturers and food scientists alike. Lipids are susceptible to oxidative

processes in the presence of catalytic systems such as light, heat, enzymes, metals,

metalloproteins, and micro-organisms, giving rise to the development of off-flavors

and loss of essential amino acids, fat-soluble vitamins, and other bioactives. Lipids

may undergo autoxidation, photo-oxidation, thermal oxidation, and enzymatic

oxidation under different conditions, most of which involve some type of free radi-

cal or oxygen species (1, 2). Among these, only autoxidation and thermal oxidation

are discussed here in detail.

Autoxidation is the most common process leading to oxidative deterioration and

is defined as the spontaneous reaction of atmospheric oxygen with lipids (3). The

process can be accelerated at higher temperatures, such as those experienced during

deep-fat frying, which is called thermal oxidation, with increases in free fatty acid

and polar matter contents, foaming, color, and viscosity (4). Unsaturated fatty acids

Bailey’s Industrial Oil and Fat Products, Sixth Edition, Six Volume Set.Edited by Fereidoon Shahidi. Copyright # 2005 John Wiley & Sons, Inc.

357

are generally the reactants affected by such reactions, whether they are present as

free fatty acids, triacylglycerols (as well as diacyglycerols or monoacylglycerols),

or phospholipids (3). It has been accepted that both autoxidation and thermal

oxidation of unsaturated fatty acids occurs via a free radical chain reaction that

proceeds through three steps of initiation, propagation, and termination (5). A

simplified scheme explaining the mechanism of autoxidation is given below:

As oxidation normally proceeds very slowly at the initial stage, the time to reach a

sudden increase in oxidation rate is referred to as the induction period (6). Lipid

hydroperoxides have been identified as primary products of autoxidation; decom-

position of hydroperoxides yields aldehydes, ketones, alcohols, hydrocarbons, vola-

tile organic acids, and epoxy compounds, known as secondary oxidation products.

These compounds, together with free radicals, constitute the bases for measurement

of oxidative deterioration of food lipids. This chapter aims to explore current

methods for measuring lipid oxidation in food lipids.

2. METHODS FOR MEASURING LIPID OXIDATION

Numerous analytical methods are routinely used for measuring lipid oxidation in

foods. However, there is no uniform and standard method for detecting all oxidative

changes in all food systems (7). Therefore, it is necessary to select a proper and

adequate method for a particular application. The available methods to monitor

lipid oxidation in foods can be classified into five groups based on what they mea-

sure: the absorption of oxygen, the loss of initial substrates, the formation of free

radicals, and the formation of primary and secondary oxidation products (8). A

number of physical and chemical tests, including instrumental analyses, have

been employed in laboratories and the industry for measurement of various lipid

oxidation parameters. These include the weight-gain and headspace oxygen uptake

method for oxygen absorption; chromatographic analysis for changes in reactants;

L

LH

+ LOOL

+ LOOH

Initiation:

LHinitiator

Propagation:

Termination:

Nonradical products

+LOO L

O2

+ H

+LOO

+ L

L

2 LOO

L

358 LIPID OXIDATION: MEASUREMENT METHODS

iodometric titration, ferric ion complexes, and Fourier transform infrared (FTIR)

method for peroxide value; spectrometry for conjugated dienes and trienes, 2-thio-

barbituric acid (TBA) value, p-anisidine value (p-AnV), and carbonyl value;

Rancimat and Oxidative Stability Instrument (OSI) method for oil stability index;

and electron spin resonance (ESR) spectrometric assay for free-radical type and

concentration. Other techniques based on different principles, such as differential

scanning calorimetry (DSC) and nuclear magnetic resonance (NMR), have also

been used for measuring lipid oxidation. In addition, sensory tests provide subjec-

tive or objective evaluation of oxidative deterioration, depending on certain details.

3. MEASUREMENT OF OXYGEN ABSORPTION

3.1. Weight Gain

Consumption of oxygen during the initial stage of autoxidation results in an

increase in the weight of fat or oil, which theoretically reflects its oxidation level.

Heating an oil and periodically testing for weight gain is one of the oldest methods

for evaluating oxidative stability (9). This method requires simple equipment and

directly indicates oxygen absorption through mass change. Oil samples are weighed

and stored in an oven at a set temperature with no air circulation. To avoid the influ-

ence of mass change by volatiles, samples can be preheated in an inert atmosphere.

Samples are then taken out of the oven at different time intervals, cooled to ambient

temperature, and reweighed; the weight gain is then recorded. The induction period

can be obtained by plotting weight gain against storage time. In some cases, the

time required to attain a 0.5% weight increase is taken as an index of oil stability

(7, 9, 10).

As a physical method for measuring lipid oxidation, the weight-gain method has

several drawbacks such as discontinuous heating of the sample, which may give rise

to non-reproducible results, and requiring long analysis time and intensive human

participation (7). Nevertheless, this method offers advantages such as low instru-

mentation cost as well as a high capacity and processing speed of samples without

limitation (7). Antolovich et al. (9) suggested that this technique may be extended

to more sophisticated continuous monitoring of mass and energy changes as in ther-

mogravimetry (TG)/differential scanning calorimetry (DSC). The weight-gain

method can also be used for measuring antioxidant activity by comparing the

results in the presence and absence of an antioxidant. Nevertheless, this method

is useful only when highly unsaturated oils, such as marine oils and vegetable

oils containing a high content of polyunsaturated fatty acids, are examined.

3.2. Headspace Oxygen Uptake

In addition to the weight-gain method, oxygen consumption can be measured

directly by monitoring the drop of oxygen pressure. Using headspace oxygen meth-

od, an oil sample is placed in a closed vessel also containing certain amount of oxy-

gen at elevated temperatures, commonly around 100�C. The pressure reduction in

MEASUREMENT OF OXYGEN ABSORPTION 359

the vessel, which is due to the oxygen consumption, is monitored continuously and

recorded automatically. The induction period as the point of maximum change in

rate of oxygen uptake can be calculated (11). A commercial instrument for this

method, known as Oxidograph, is available. In the Oxidograph, the pressure change

in the reaction vessel is measured electronically by means of pressure transducers

(7, 12).

Oxygen consumption can also be measured by electrochemical detection of

changes in oxygen concentration. However, the analysis of the graphical data

obtained has been the bottleneck for this technique. The use of a semiautomatic

polarographic method has been proposed as an improvement for evaluation of lipid

oxidation by determination of oxygen consumption (13). As described by Genot

et al. (13), this method is based on use of two oxygen meters with microcathode

oxygen electrodes, coupled to a computerized data collection and processing unit.

The headspace oxygen method is simple and reproducible and may be the best

analytical method to evaluate the oxidative stability of fats and oils (14). Its appli-

cation in measurement of lipid oxidation in food products other than fats and oils,

however, is limited because protein oxidation also absorbs oxygen (15).

4. MEASUREMENT OF REACTANT CHANGE

Lipid oxidation can also be assessed by quantitatively measuring the loss of initial

substrates. In foods containing fats or oils, unsaturated fatty acids are the main

reactants whose composition changes significantly during oxidation. Changes in

fatty acid composition provide an indirect measure of the extent of lipid oxidation

(15). In this method, lipids are extracted from food, if necessary, and subsequently

converted into derivatives suitable for chromatographic analysis (7). Fatty acid

methyl esters (FAME) are the derivatives frequently used for determination of fatty

acid composition, usually by gas chromatography (GC) (16). Similarly, iodine

value, which reflects the loss of unsaturation, can also be used as an index of lipid

oxidation (17).

Measurement of changes in fatty acid composition is useful for identification of

lipid class and fatty acids that are involved in oxidation reactions (7). However,

because the distribution of unsaturated fatty acids varies in different food systems,

for instance, the highly unsaturated fatty acids being located predominantly in

phospholipids of muscle foods, separation of lipids into neutral, glycolipid, phos-

pholipid, and other classes may be necessary (7, 15). Moreover, it is an insensitive

way of assessing oxidative deterioration. For comparison through calculation, oxi-

dation of 0.4% polyunsaturated fatty acids to monohydroperoxides would represent

a change of 16 meq oxygen/kg oil in peroxide value, whereas a change of less than

1.0 meq oxygen/kg oil could readily be detected by measuring peroxide value (12).

Additionally, the application of this method is limited because of its inability

to serve as an indicator of oxidation of more saturated lipids (7). Nevertheless,

its usefulness for measuring oxidation of highly unsaturated oils cannot be

underestimated.


5. MEASUREMENT OF PRIMARY PRODUCTS OF OXIDATION

5.1. Peroxide Value (PV)

Lipid oxidation involves the continuous formation of hydroperoxides as primary

oxidation products that may break down to a variety of nonvolatile and volatile

secondary products (8, 15). The formation rate of hydroperoxides outweighs their

rate of decomposition during the initial stage of oxidation, and this becomes

reversed at later stages. Therefore, the peroxide value (PV) is an indicator of the

initial stages of oxidative change (18). However, one can assess whether a lipid

is in the growth or decay portion of the hydroperoxide concentration by monitoring

the amount of hydroperoxides as a function of time (7).

Analytical methods for measuring hydroperoxides in fats and oils can be classi-

fied as those determining the total amount of hydroperoxides and those based on

chromatographic techniques giving detailed information on the structure and the

amount of specific hydroperoxides present in a certain oil sample (8). The PV repre-

sents the total hydroperoxide content and is one of the most common quality indi-

cators of fats and oils during production and storage (9, 18). A number of methods

have been developed for determination of PV, among which the iodometric titra-

tion, ferric ion complex measurement spectrophotometry, and infrared spectroscopy

are most frequently used (19).

5.1.1. Iodometric Titration Method Iodometric titration assay, which is based

on the oxidation of the iodide ion (I�) by hydroperoxides (ROOH), is the basis of

current standard methods for determination of PV (9). In this method, a saturated

solution of potassium iodide is added to oil samples to react with hydroperoxides.

The liberated iodine (I2) is then titrated with a standardized solution of sodium thio-

sulfate and starch as an endpoint indicator (7, 9, 20). The PV is obtained by calcu-

lation and reported as milliequivalents of oxygen per kilogram of sample (meq/kg).

The official determination is described by IUPAC (21). Chemical reactions involved

are given below:

ROOH þ 2Hþ þ 2KI ! I2 þ ROH þ H2O þ 2Kþ

I2 þ 2NaS2O3 ! Na2S2O6 þ 2NaI

Although iodometric titration is the most common method for measurement of PV,

it suffers from several disadvantages. The procedure is time-consuming and labor-

intensive (18). As described by Ruiz et al. (18), the assay includes six steps: accu-

rate weighing of the sample, dissolution of lipids in chloroform, acidification with

acetic acid, addition of potassium iodide, incubation for exactly 5 minutes, and

titration with sodium thiosulfate. This technique requires a large amount of sample

and generates a significant amount of waste (18, 22, 23). Furthermore, possible

absorption of iodine across unsaturated bonds and oxidation of iodide by dissolved

oxygen are among potential drawbacks of this method (7, 9). Besides, lack of sen-

sitivity, possible interferences, and difficulties in determining the titration endpoint

MEASUREMENT OF PRIMARY PRODUCTS OF OXIDATION 361

are also the main limitations (8, 23). To overcome these drawbacks, novel methods

based on the same reaction have been developed, in which some other techniques

are adopted as modification of the classical iodometric assay. Techniques such as

colorimetric determination at 560 nm (24), potentiometric endpoint determination

(25), and spectrophotometric determination of the I�3 chromophore at 290 nm or

360 nm (26, 27) have been proposed. In addition, an electrochemical technique

has been used as an alternative to the titration step in order to increase the sensitiv-

ity for determination of low PV by reduction of the released iodine at a platinum

electrode maintained at a constant potential (7).

5.1.2. Ferric Ion Complexes Other chemical methods based on the oxidation of

ferrous ion (Fe2þ) to ferric ion (Fe3þ) in an acidic medium and the formation of

iron complexes have also been widely accepted. These methods spectrophotometri-

cally measure the ability of lipid hydroperoxides to oxidize ferrous ions to ferric

ions, which are complexed by either thiocyanate or xylenol orange (23, 28, 29).

Ferric thiocyanate is a red-violet complex that shows strong absorption at 500–

510 nm (8). The method of determining PV by coloremetric detection of ferric thio-

cyanate is simple, reproducible, and more sensitive than the standard iodometric

assay, and has been used to measure lipid oxidation in milk products, fats, oils,

and liposomes (8, 23).

The ferrous oxidation of xylenol orange (FOX) assay uses dye xylenol orange to

form a blue-purple complex with a maximum absorption at 550–600 nm (8). This

method is rapid, inexpensive, and not sensitive to ambient oxygen or light (30). It

can consistently quantify lower hydroperoxide levels; and good agreement exists

between the FOX assay and the iodometric method (30). The FOX method has

been successfully adapted to a variety of applications. However, because many fac-

tors, such as the amount of sample, solvent used, and source of xylenol orange, may

affect the absorption coefficient, knowledge of the nature of hydroperoxides present

in the sample, and careful control of the conditions used are required for accurate

measurements (8).

5.1.3. Fourier Transform Infrared Spectroscopy (FTIR) It has been recog-

nized that hydroperoxides can quantitatively be determined by IR spectroscopy

via measurement of their characteristic O-H stretching absorption band (31). An

absorption band at about 2.93 mm indicates the generation of hydroperoxides,

whereas the replacement of a hydrogen atom on a double bond or polymerization

accounts for the disappearance of a band at 3.20 mm. The formation of aldehydes,

ketones, or acids gives rise to an extra band at 5.72 mm. Furthermore, cis-, trans-

isomerization and formation of conjugated dienes can be detected through the

changes in the absorption band in the range of 10 mm to 11 mm (7).

A rapid Fourier transform infrared spectroscopy (FTIR) method based on the

stoichiometric reaction of triphenylphosphine (TPP) with hydroperoxides has

been developed and successfully applied to determination of PV of edible oils

(32). The hydroperoxides present in oil samples react stoichiometrically with

TPP to produce triphenylphosphine oxide (TPPO), which has an intense absorption


band at 542 cm�1 in the mid-IR spectrum (8, 18). The band intensity is measured

and converted to peroxide value. The chemical reaction involved is given below:

ROOH þ TPP�!TPP¼O þ ROH

By using tert-butyl hydroperoxide spiked oil standards and evaluation of the band

formed at 542 cm�1, a linear calibration graph covering the range of 1–100 PV was

obtained (18). More recently, disposable polymer IR (PIR) cards have been used as

sample holders where unsaturated oil samples oxidize at a fairly rapid rate (33). In

the FTIR/PIR card method, warm air continuously flows over the sample allowing

oxidation to be monitored at moderate temperatures. At periodic intervals, indivi-

dual cards are removed and the FTIR spectra scanned (33). Another new FTIR

approach uses flow injection analysis (FIA), which offers exact and highly repro-

ducible timing of sample manipulation and reaction as well as a closed environment

with oxygen and light being easily excluded (18).

The FTIR spectroscopy is a simple, rapid, and highly precise method. It shows

excellent correlation with the iodometric method and avoids the solvent and reagent

disposal problems associated with the standard wet chemical method (18, 32). The

FTIR method provides an automated, efficient and low-cost means of evaluating

oxidation in oils undergoing thermal stress and has gained considerable interest

for quality control in the industry (8, 20, 34). However, there is a need to charac-

terize the spectral changes, assign wavelengths to more common molecular species

produced, and access potential spectral cross interferences (20). Recently, an

improved Fourier transform infrared attenuated total reflectance (FTR-ATR) meth-

od using the whole FTIR spectral data instead of particular wavenumbers has been

proposed (34).

In addition to the three major methods discussed above, other techniques have

also been employed in determination of PV, such as chemiluminescence and chro-

matography. Chemiluminescence method is based on detecting the chemilumines-

cent products generated during the reaction of hydroperoxides with substances such

as luminol and dichlorofluorescein (7, 35). This method was reviewd by Jimenez

et al. (36). High correlations have been found between chemiluminescence and

other standard methods, indicating that chemuliminescence could serve as an accu-

rate tool for determination of PV (37). However, this method has low sensitivity to

tert-butyl hydroperoxide, tert-butyl perbenzoate, diacyl peroxides, and dialkyl per-

oxides (35). Chromatographic techniques, mainly gas chromatography (GC) and

high-performance liquid chromatography (HPLC), have also been employed for

evaluation of lipid oxidation. These methods provide information about specific

hydroperoxides, whereas other assays measure their total amount. Chromatographic

methods require small amounts of sample, and interference from minor compounds

other than hydroperoxides can be easily excluded (8). HPLC shows advantages over

GC and has become a popular technique for hydroperoxide analysis. It operates at

room temperature, thus decreases the risk of artifact formation, and no prior deri-

vatization is required (8). A wide range of hydroperoxides can be analyzed using

either normal or reverse-phase HPLC. Thus, hydroperoxides, the primary products

MEASUREMENT OF PRIMARY PRODUCTS OF OXIDATION 363

and intermediates in lipid oxidation reaction, provide an important parameter for

evaluation of oxidation level. In addition, the inhibition of formation or action of

these unstable species by antioxidants can be used as a means of assessing antiox-

idant activity (9). Measurement of hydroperoxides is also carried out in accelerated

tests to establish the oxidative stability of a given oil. A case in point is the active

oxygen method (AOM), in which air is bubbled through fat or oil held at 98–100�Cand PV is determined periodically (7, 38). The time required to reach a PV of 100

meq/kg is the AOM stability of the oil sample (7). This method is now considered

outdated and is replaced by other standard methods in the industry, although

product specifications still routinely give AOM values (38).

5.2. Conjugated Dienes and Trienes

It was discovered in 1933 that the formation of conjugated dienes in fats or oils

gives rise to an absorption peak at 230–235 nm in the ultraviolet (UV) region. In

the 1960s, monitoring diene conjugation emerged as a useful technique for the

study of lipid oxidation (9). During the formation of hydroperoxides from unsatu-

rated fatty acids conjugated dienes are typically produced, due to the rearrangement

of the double bonds. The resulting conjugated dienes exhibit an intense absorption

at 234 nm; similarly conjugated trienes absorb at 268 nm (7). An increase in UV

absorption theoretically reflects the formation of primary oxidation products in fats

and oils. Good correlations between conjugated dienes and peroxide value have

been found (39, 40).

Ultraviolet detection of conjugated dienes is simple, fast, and requires no

chemical reagents and only small amounts of samples are needed. However, this

OH

OOH

O

hydroperoxydiene oxodiene

Reduction

hydroxydiene

conjugated triene

conjugated tetraene

and

Figure 1. Chemical reaction steps in conjugable oxidation products (COP) assay.


TABLE 1. Summary of Methods for Analysis of Primary Oxidation Products.

Method Principle Measurement Sensitivity Applications

Iodometric titration (PV) Reduction of ROOH with KI

and measurement of I2

Titration with Na2S2O3 �0.5-meq/kg fat Fats and oils

Ferric ion complexes (PV) Reduction of ROOH with Fe2þ

and formation of Fe3þ

complexes

Absorption at 500–510 nm

of the red complex with

SCN�

�0.1-meq/kg fat Fats, oils and food lipids

Absorption at 560 nm of the

blue-purple complex with

xylenol orange

�0.5-meq/kg sample All samples

FTIR (PV) Reduction of ROOH with TPP Absorption at 542 cm�1 of

TPPO

�0.2-meq/kg fat Fats and oils

Chemiluminescence (PV) Reaction with luminol in the

presence of heme catalyst

Chemiluminescence

emission of oxidized

luminol

�1 pmol Fats and oils

GC-MS (PV) Reduction of ROOH to ROH

and quantitation of ROH

derivatives

ROH derivatives From ng to fg depending on

technical details, amount

of sample and detection

system

All samples

UV spectrometry

(conjugated dienes and

trienes)

Estimation of conjugated

dienes and trienes

Absorption at 230–234 nm

and 268 nm

�0.2 meq/kg lipid All samples

NOTE: The oxygen absorption measurement and loss of double bonds for fatty acid analysis are not considered as primary changes in this table.

Adapted from (8).

method has less specificity and sensitivity than PV measurement (9, 12). Further-

more, the result may be affected by the presence of compounds absorbing in the

same region, such as carotenoids (7). To avoid these interferences, an alternative

spectroscopic method measuring conjugable oxidation products (COPs) has

been proposed. In this method, hydroperoxides and some decomposition products

are converted to more conjugated chromophores by reduction and subsequent

dehydration (Figure 1). The concentrations of the resultant conjugated trienes

and tetraenes are determined from their respective absorption at 268 nm and

301 nm and expressed as COP values (7, 12).

Table 1 summarizes different methods available for analysis of primary oxida-

tion products. Both chemical and instrumental methods are included in this

table.

6. MEASUREMENT OF SECONDARY PRODUCTS OF OXIDATION

The primary oxidation products (hydroperoxides) are unstable and susceptible to

decomposistion. A complex mixture of volatile, nonvolatile, and polymeric second-

ary oxidation products is formed through decomposition reactions, providing var-

ious indices of lipid oxidation (5). Secondary oxidation products include aldehydes,

ketones, alcohols, hydrocarbons, volatile organic acids, and epoxy compounds,

among others. Methods for assessing lipid oxidation based on their formation are

discussed in this section.

6.1. Thiobarbituric Acid (TBA) Test

The thiobarbituric acid (TBA) test was proposed over 40 years ago and is now

one of the most extensively used methods to detect oxidative deterioration of

fat-containing foods (41). During lipid oxidation, malonaldehyde (MA), a minor

component of fatty acids with 3 or more double bonds, is formed as a result of

the degradation of polyunsaturated fatty acids. It is usually used as an indicator

of the lipid oxidation process, both for the early appearance as oxidation occurs

and for the sensitivity of the analytical method (42). In this assay, the MA is reacted

with thiobarbituric acid (TBA) to form a pink MA-TBA complex that is measured

spectrophotometrically at its absorption maximum at 530–535 nm (Figure 2) (9, 43, 44).

The extent of oxidation is reported as the TBA value and is expressed as milligrams

HN

N OH NH

NH

S

O

S

OH

O

N

NHS OH

OH

H C

O

CH2 C

O

H

TBA MATBA-MA adduct

+

Figure 2. Reaction of 2-thiobarbituric acid (TBA) and malonaldehyde (MA).


of MA equivalents per kilogram sample or as micromoles of MA equivalents per

gram of sample. It must, however, be noted that alkenals and alkadienals also react

with the TBA reagent and produce a pink color. Thus, the term thiobarbituric acid

reactive substances (TBARS) is now used instead of MA.

The TBA test can be performed by various procedures, among which four major

types have frequently been employed. These include test on the whole sample, test

on an aqueous or acid extract of sample, test on a steam distillate, and test on

extracted lipid from a sample (45). The test on a steam distillate (distillation meth-

od) is the most commonly used method for determining TBA value. Tarladgis et al.

(46) found that the distillation of an acidified sample was essential to liberate MA

from precursor or bound forms, to produce maximal color development and, espe-

cially, to separate TBARS from the food matrix (44). Although the distillation

method is the most popular TBA method, it is generally considered less accurate

and reproducible than the method using food extracts (15). However, trends

obtained in comparative studies always provide useful information that correspond

with other measurements. Comparison of different TBA test procedures has

been made by Hoyland et al. (46), Shahidi et al. (47), Pikul et al. (48), and

Wang et al. (49).

The TBA test is used frequently to assess the oxidative state of a variety of food

systems, despite its limitations, such as lack of specificity and sensitivity (44). As

already noted, many other substances may react with the TBA reagent and contri-

bute to absorption, causing an overestimation of the intensity of color complex (44).

Interferences may come from additional absorption of other alkanals, 2-alkenals,

2,4-alkdienals, ketones, ketosteroids, acids, esters, proteins, sucrose, urea, pyri-

dines, and pyrimidines, also referred to as TBARS (43, 50). For instance, the

reaction of TBA with various aldehydes leads to the development of a yellow

chromogen (aldehyde-TBA adduct) with an absorption maximum at 450 nm, which

overlaps with the pink peak at 532 nm resulting in erroneously high TBA values in

certain cases (43, 45, 51). Furthermore, the presence of barbituric acid impurities in

the TBA reagent may produce TBA-MA-barbituric acid and MA-barbituric acid

adducts that absorb at 513 nm and 490 nm, respectively, indicating that thiobarbi-

turic acid should be purified before use (43). In addition, nitrite can interfere in the

TBA test, whereas sulfanilamide could be added to samples to avoid the interfer-

ence when residual nitrite is present (52). In order to improve the specificity and

sensitivity of the TBA test, several modifications to the original TBA procedures

have been proposed, including reduction of the heating temperature to stabilize

the yellow color aldehyde-TBA complex (53), addition of antioxidants to sample

in an attempt to prevent oxidation during the test (54), extraction of the MA prior

to the formation of the chromogen (43), direct FTIR analysis of TBARS, and use of

HPLC to separate the complex before measurement or to characterize the individual

species of TBARS (9, 43).

Despite it limitations, the TBA test provides an excellent means for evaluating

lipid oxidation in foods, especially on a comparative basis. However, its use in bulk

oils is less common than the so-called para-anisidine value (p-AnV) detailed

below.

MEASUREMENT OF SECONDARY PRODUCTS OF OXIDATION 367

6.2. p -Anisidine Value (p -AnV)

The p-anisidine value (p-AnV) method measures the content of aldehydes (princi-

pally 2-alkenals and 2,4-alkadienals) generated during the decomposition of hydro-

peroxides. It is based on the color reaction of p-methoxyaniline (anisidine) and the

aldehydic compounds (55). The reaction of p-anisidine reagent with aldehydes

under acidic conditions affords yellowish products that absorb at 350 nm

(Figure 3) (7, 12). The color is quantified and converted to p-AnV. The p-AnV is

defined as the absorbance of a solution resulting from the reaction of 1 g of fat in

isooctane solution (100 ml) with p-anisidine (0.25% in glacial acetic acid) (12).

This test is more sensitive to unsaturated aldehydes than to saturated aldehydes

because the colored products from unsaturated aldehydes absorb more strongly at

this wavelength (12). However, it correlates well with the amount of total volatile

substances (55). The p-AnV is a reliable indicator of oxidative rancidity in fats and

oils and fatty foods (56). A highly significant correlation between p-AnV and flavor

scores and PV has been found (57). Nevertheless, some authors have indicated that

p-AnV is comparable only within the same oil type because initial AnV varies

among oil sources (58). For instance, oils with high levels of polyunsaturated fatty

acids might have higher AnV even when fresh (59).

This method is used less frequently in North America, but is widely employed in

Europe (38), particularly as a part of the Totox number, as explained below. Caution

CH3O

NH2

CH C

CH

O OH

H

CH3O

N OH

CH3O

NH2

CH3O

N NH

OCH3

+

Malonaldehyde(enolic form)

p-Methoxyaniline(p-anisidine)

Figure 3. Possible reactions between p-anisidine reagent and malonaldehyde.


must be exercised when performing this test because of toxicity of the anisidine

reagent (55).

6.3. Totox Value

The Totox value is a measure of the total oxidation, including primary and second-

ary oxidation products. It is a combination of PV and p-AnV:

Totox value ¼ 2 PV þ p-AnV

During lipid oxidation, it is often observed that PV first rises, then falls as hydro-

peroxides decompose (38). PV and p-AnV reflect the oxidation level at early and

later stages of oxidation reaction, respectively. Totox value measures both hydro-

peroxides and their beakdown products, and provides a better estimation of the

progressive oxidative deterioration of fats and oils (38). However, Totox value

has no scientific basis because it is a combination of two indicators with different

dimensions (7). Recently, Wanasundara and Shahidi used TBA values and defined

TotoxTBA as 2PV þ TBA using the TBA test in place of the p-AnV assay (60).

6.4. Carbonyls

The carbonyl compounds, including aldehydes and ketones, are the secondary oxi-

dation products generated from degradation of hydroperoxides, and are suggested

to be the major contributors to off-flavors associated with the rancidity of many

food products (9). The analysis of total carbonyl compounds, which is based on

the absorbance of the carbonyl derivatives, provides another approach to measure

the extent of lipid oxidation in fats and oils. In this method, the total carbonyl

content is measured by a colorimetric 2,4-dinitrophenylhydrazone procedure. The

carbonyl compounds formed during lipid oxidation are reacted with 2,4-dinitrophe-

nylhydrazine (DNPH) followed by the reaction of the resulting hydrazones with

alkali (Figure 4). The final colored products are then analyzed spectrophotometrically

R C

R

O + NO2N

H

H2N

NO2

NO2N

H

N

NO2

C

OH−

−H2O

R

R

NO2NN

NO2

CR

R−

Figure 4. Reactions between carbonyls and 2,4-dinitrophenylhydrazine.


at a given wavelength (7, 15). Many variations of this method using an alternative

solvent, reagent, wavelength, or workup have been reported. The determination of

total content of carbonyls has been used in different oxidative stability studies.

However, it has been criticized because the determination conditions cause degra-

dation of hydroperoxides into carbonyl derivatives, giving erroneous results (58).

Carbonyls produced from protein oxidation may also give rise to higher values

than those expected from lipid oxidation alone. The addition of triphenylphosphine

(TPP) prior to carbonyl determination has been proposed to avoid the interference

from hydroperoxides. Hydroperoxides are reduced by TPP, and neither TPP nor

TPPO, the oxidation products of TPP, interfere with the measurement of carbonyl

content (61). In quality assessment of used frying fats, where short-chain carbonyls

are already removed by distillation at the high temperature of the deep-frying,

selectivity can be improved by determination of higher carbonyl compounds instead

of the total carbonyls. HPLC is used to separate the DNPH derivatives of higher

carbonyls from those of short-chain carbonyl compounds (62).

Apart from detection of total carbonyl content, the analysis of individual carbo-

nyl compounds has gained popularity for following lipid oxidation. Hexanal, one of

the major secondary products formed during the oxidation of linoleic and other o6

fatty acids, serves as a reliable indicator of lipid oxidatin in foods rich in o6 fatty

acids (7). A strong linear relationship was reported between hexanal content, sen-

sory scores, and TBA values (63). Moreover, measurement of hexanal offers the

advantage of analyzing a single, well-defined end product for antioxidant efficiency

studies (9). Hexanal can be quantified by chromatography (64) or as the intensity of

the carbonyl band by NIR spectroscopy (65). Nevertheless, these methods may

require volatilization of hexanal, whereas hexanal volatilization may be hindered

due to covalent or other types of binding between hexanal and proteins in foods

and, thus, may affect accurate hexanal quantifications (66). More recently, an

indirect enzyme-linked immunosorbanct assay (ELISA) has been developed for

monitoring lipid oxidation through quantification of hexanal-protein adducts, which

are recognized by polyclonal or monoclonal antibodies (66).

Other carbonyl compounds, including propanal, pentanal, decadienal, etc., are

also used for evaluating lipid oxidation in foods. For instance, propanal is a recom-

mended indicator for lipid oxidation in foods that are high in o3 fatty acids, such as

marine oils (67, 68). In general, it is essential to use appropriate indicators when

assessing the oxidative deterioration of different food systems.

6.5. Oil Stability Index (OSI)

During lipid oxidation, volatile organic acids, mainly formic acid and acetic acid,

are produced as secondary volatile oxidation products at high temperatures, simul-

taneously with hydroperoxides (20, 69). In addition, other secondary products,

including alcohols and carbonyl compounds, can be further oxidized to carboxylic

acids (20). The oil stability index (OSI) method measures the formation of volatile

acids by monitoring the change in electrical conductivity when effluent from

oxidizing oils is passed through water (12). The OSI value is defined as the point

of maximal change of the rate of oxidation, attributed to the increase of conductivity


by the formation of volatile organic acids during lipid oxidation (70). However, this

method requires a somewhat higher level of oxidation (PV > 100) to obtain mea-

surable results than other methods in which hydroperoxides are the most important

products formed and detected (71). Therefore, to determine oil stability in the

laboratory, especially for some oils that are stable under normal conditions, the oxi-

dation process is accelerated by exposing oil samples to elevated temperatures in

the presence of an excess amount of air or oxygen (72, 73). The OSI method differs

from ambient storage conditions by using a flow of air and high temperatures

to accelerate oxidation (71). The OSI is an automated development of the active-

oxygen method (AOM), because both employ the principle of accelerated oxida-

tion. Nevertheless, the OSI test measures the changes in conductivity caused by

ionic volatile acids, whereas PV is determined in the AOM (7).

Two pieces of commercially available equipment, the Rancimat (Metrohm Ltd.)

and the Oxidative Stability Instrument (Omnion Inc.), are employed for determin-

ing the OSI value. Rancimat is a rapid automated method, which agrees well with

the AOM (71). In the Rancimat assay, a flow of air is bubbled through a heated oil,

usually at 100�C or above. For marine oils, temperatures as low as 80�C are often

used. Volatile compounds formed during accelerated oxidation are collected in dis-

tilled water, increasing the water conductivity. The change of conductivity is plotted

automatically and the induction period of the oil or the time taken to reach a fixed

level of conductivity is recorded (20, 74). The Rancimat assay enables continuous

monitoring of the oxidation process. As reported by Farooq et al. (75), analysis by

the Rancimat method is four to five times more rapid than that by the AOM. Excel-

lent correlation between Rancimat and conjugated dienes has been found (72).

However, the main shortcoming of this method is that only eight samples can be

included in each batch. Another appatatus, the Oxidative Stability Instrument, oper-

ates on the same principle as the Rancimat, and has the capacity of simultaneously

analyzing up to 24 samples (20). Various modifications have been proposed for

assessing lipid oxidation by the OSI method. These include the use of auxiliary

energies, such as microwaves to shorten the analysis time (72) and a combination

of the OSI method with chromatography to obtain specific information about vola-

tile products (76). The volatiles trapped during measurement by the Rancimat assay

can be analyzed by headspace-GC (HS-GC) with FID and GC-MS for quantifica-

tion of individual volatiles, thus improving the specificity of the assessment (76).

Although the OSI method is useful for quality control of oils, it is not recom-

mended for measurement of antioxidant activity for certain reasons. The high tem-

peratures used do not allow reliable predictions of antioxidant effectiveness at

lower temperatures. Volatile antioxidants may be swept out of the oil by the air

flow under test conditions, and also the oils are severely deteriorated when endpoint

is reached (12).

6.6. Hydrocarbons and Fluorescence Assay

Formation of saturated hydrocarbons, especially short-chain (C1-C5) hydrocarbons

such as ethane, propane, and pentane, can be measured for monitoring lipid oxida-

tion when aldehydes are either absent or undetectable (7, 15). Pentane content,


determined by GC techniques, has been a useful parameter to assess rancidity of

fats and oils as well as freeze-dried muscle foods (7, 15). Significant correlations

existed between pentane levels and rancid odor scores (15).

It has been observed that the content of secondary oxidation products, such as

malonaldehyde (MA), decreases with increased lipid oxidation, which can be

explained by further reaction of MA with proteins. MA reacts with compounds con-

taining primary amino groups (proteins, amino acids, DNA, phospholipids) to form

fluorescent products (Figure 5) (37). A fluorescence assay has been successfully

used to assess lipid oxidation in muscle foods and biological tissues.

In addition to MA, hydroperoxides and other aldehydes also react with amino

compounds generating various fluorescent products with different excitation and

emission maxima (37). Significant correlations existed between this method and

the TBA value as well as oxygen absorption level, and appears to be a reliable

RN CH CH CH NHR

RN CH CH CHOH

Malonaldehyde

Amine-Malonaldehyde Adduct(Non-fluorescent)

Conjugated Schiff Base(Fluorescent)

H C

O

CH CHOHRNH2

H C

O

CH2 C H

O

RNH2

Figure 5. Reaction of lipid oxidation products such as malonaldehyde and amines.

TABLE 2. Summary of Methods for Analysis of Secondary Oxidation Products.

Method Compounds Comments Applications

TBA TBARS, mainly

malonaldehyde

Spectrometry technique

It can be carried out

on whole sample

All samples,

especially fish

oils

p-Anisidine Aldehydes, mainly

alkenals

Absorption at 350-nm

Standard method

Fats and oils

Carbonyls Total carbonyls or

specific carbonyl

compound formed

Spectrometry technique

and HPLC for total

or specific carbonyl

compounds

Fats and oils

OSI method

(Rancimat &

Oxidative Stability

Instrument)

Volatile organic acids Monitoring changes in

conductivity Rapid

and automated

Fats and oils

Gas Chromatography Volatile carbonyls and

hydrocarbons

Direct headspace Rapid

analysis

All samples

Adapted from (8).


indicator of oxidative deterioration in muscle foods, especially in freeze-dried

products (37, 77).

Table 2 exhibits the summary of methods for analysis of secondary oxidation

products.

7. MEASUREMENT OF FREE RADICALS

The initial steps of lipid oxidation involve chain reactions of free radicals as impor-

tant short-lived intermediates. Oxidation level of fats and oils can be measured

directly by detecting the formation of radicals. Methods based on the detection

of radicals or on the tendency for the formation of radicals provide a good indica-

tion of initiation of lipid oxidation (78, 79).

Electron spin resonance (ESR), also referred to as electron paramagnetic reso-

nance (EPR) spectroscopy, relies on the paramagnetic properties of the unpaired

electrons in radicals and has been developed for assessing the formation of free

radicals originating in the early stages of oxidation and the onset of primary oxida-

tion (6, 78). The assay measures the absorption of microwave energy when a

sample is placed in a varied magnetic field (7). Quantification of radical concentra-

tions is complicated by comparison with stable paramagnetic compounds, such as

transition metals and nitroxyl radicals (78). However, the short lifetimes and low

steady-state concentration of the highly reactive lipid-derived radicals make it dif-

ficult to detect these radicals at concentrations lower than the minimum detectable

concentration of 10�9 M (78). To overcome this problem, various approaches have

been used, including pulse radiolysis and UV photolysis, continuous flow systems

and spin trapping, among which spin trapping has been the most widely employed

procedure (9). Spin trapping technique allows the accumulation of detectable con-

centrations of longer-lived radicals by addition to samples of a spin trapping agent,

which reacts with free radicals to form more stable spin adducts, but often at the

expense of the ability to identify the original radical (6, 9, 78). Nitroso compounds

and nitrones are the most common spin traps, both leading to nitroxyl type spin

adducts, such as a-phenyl-tert-butylnitrone (PBN) adducts (Figure 6) (78).

H

PhN+

O−

CMe3

Me3C N O

PBN

MNP

H

RPh

R NO

CMe3

NO

CMe3

+

+

R

R

Figure 6. Formation of nitroxyl radical spin adducts.

MEASUREMENT OF FREE RADICALS 373

ESR spectroscopy is of great value for the study of the early stages of lipid oxi-

dation and prediction of oxidative stability of fats and oils. It has high sensitivity

and allows mild conditions by applying significantly low temperatures and requires

little sample preparation (6, 78, 80). Strong linear correlations were found between

ESR and Rancimat and oxygen consumption analyses (6, 79). ESR has also been

used for evaluation of antioxidant activity (81). Nevertheless, spin traps used in the

ESR assay have been reported to exhibit widely differing trapping efficiencies for

different radicals and show both pro-oxidant and antioxidant effects (9, 82, 83).

Moreover, spin adducts can act as antioxidants, giving erroneous results of oxida-

tive stability of samples (9). However, even with these limitations, the ESR spectro-

scopy is a suitable method for measuring lipid oxidation in foods and in biological

tissues.

8. OTHER METHODS

8.1. Differential Scanning Calorimetry (DSC)

During lipid oxidation, fat or oil materials reveal a number of thermally induced

transitions, such as the transfer of oxygen molecules to unsaturated fatty acids

(exothermic process) (84). Therefore, thermal analysis can be applied in accelerated

oil stability tests. The differential scanning calorimetry (DSC) technique, which is

based on thermal release of oxidation reactions, has the potential as a nonchemical

method for assessing oxidative stability of fats and oils, indicating the onset of

advanced oxidation (termination) (6). It provides unique energy profile information,

which specifically measures the temperature and heat flows associated with lipid

oxidation as a function of time and temperature (85). The method uses isothermal

or nonisothermal conditions and a flow of oxygen as purge gas, with a calorimeter

measuring the heat flow into (endothermic) or out of (exothermic) an oil sample

undergoing oxidation changes (6, 84). The oxidation curves of the sample are

obtained with different heating time, and a dramatic increase for the evolved

heat can be observed with the appearance of a sharp exothermic curve during initia-

tion of oxidation. The endpoint is taken at the time where a rapid exothermic reac-

tion between oil and oxygen occurs and induction period (IP) determined

automatically by intersection of extrapolated baseline and tangent line (leading

edge) of the exotherm (Figure 7) (6, 84). The DSC also measures oxidation onset

temperature, the temperature at maximum reaction, and the ending temperature

(84). The isothermal and nonisothermal DSC show good agreement, suggesting

that both isothermal and nonisothermal DSC are suitable for oxidation studies of

oils (86). The DSC technique has recently been reviewed by Tan et al. (84, 87).

The DSC is a sensitive, effective, and consistent method for characterization of

the quality of oils at different stages of oxidation (20). It is simple and rapid, and it

requires no solvent or chemical reagent. As reported by Hassel et al. (89), oils sam-

ples, which required 14 days via AOM, could be evaluated in less than 4 hours

by DSC. Thus, DSC is a reliable alternative to current methods for monitoring lipid


oxidation (85). The results from DSC show excellent correlations with other

accelerated methods and chemical analyses (6, 73, 85).

8.2. Nuclear Magnetic Resonance (NMR) Spectroscopy

High-resolution 1H NMR spectroscopy, in which hydrogen atoms (proton, 1H) with

various locations in the triacylglycerol (TAG) molecules are determined, has been

used to evaluate oxidative deterioration of fats and oils (7). The principle of NMR is

that hydrogen atoms in a strong magnetic field absorb energy, in the radiofrequency

range, depending on their molecular environment, in which changes occur during

the oxidation process (7). These changes may be monitored by NMR spectroscopy

as a reflection of oxidation level of food lipids. The oil sample is dissolved in

CDCl3 to avoid inference from solvent, and its NMR spectrum recorded, with tetra-

methylsilane (TMS) as an internal standard (7). The spectrum shows several groups

of signals, corresponding to the hydrogen atoms in different locations in the TAG

molecules (Figure 8). The total number of each of these differently located protons

can be calculated, from which ratios of aliphatic to olefinic protons (Rao) and ali-

phatic to diallylmethylene protons (Rad) may be obtained (7). Both ratios increase

steadily during lipid oxidation and may serve as an index of oxidative deterioration

of oil samples. This method was reviewed by Guillen et al. (90). NMR spectroscopy

has been used by many researchers, and the changes in Rao and Rad measured by

NMR correlated well with Totox values (91, 92), conjugated diene values, and TBA

values (93). In addition to 1H NMR, 13C NMR and 31P NMR are also powerful tools

to predict oxidative stability of oils (94–96). 13C NMR enables direct observation

of carbon atoms. The selectivity and dispersion of 13C NMR spectra are very

high (96). 13C NMR assesses lipid oxidation by monitoring the changes of carbon

chains in TAG molecules, revealing the specific sites that oxidative degradation

50 100 150 200 250Time (min)

IP

Exo

ther

mic

Figure 7. Determination of induction period (IP) by DSC.

OTHER METHODS 375

occurs (94). However, because the abundance of the NMR active 13C nucleus iso-

tope is only 1.12% of 12C, the sensitivity of 13C NMR is usually much lower than

that of 1H NMR (96).

NMR spectroscopy is a rapid, nondestructive, and reliable technique for asses-

sing lipid oxidation. It simultaneously measures both the primary and the secondary

oxidative changes in oils, and provides specific information on oxidative regions in

the TAG molecules. Thus, NMR spectroscopy is considered a more suitable means

for estimating lipid oxidation than chemical determinations.

8.3. Sensory Evaluation

For the food industry, the detection of oxidative off-flavors by taste or smell is the

main method of deciding when a lipid-containing food is no longer fit for consump-

tion (12). Terminologies and methodologies have been developed for sensory

evaluation of specific food products such as meats, peanuts, and vegetable oils

(97). In the edible oil industry, the AOCS (American Oil Chemists’ Society) Flavor

Quality Scale (revised) with separate grading and flavor intensity has been

employed for describing lipid oxidation (97), as summarized in Table 3. The

descriptive analysis, including the detection and the description of both the quali-

tative and quantitative sensory aspects of a product, is performed by a trained panel,

as the sensitivity to the off-flavors varies among different individuals (12, 97). The

sensory induction period of the product can be determined.

8 7 6 5 4 3 2 1 0 PPM

a

b

c

d

e

f

h

g

TMS

H C

H

O CO CH2 (CH2)n CH CH

CH O CO CH2 (CH2)n CH3

C

H

H O CO CH2 (CH2)n CH3

d g a

CH2 CH CH CH2 CH2 (CH2)n CH3

a c a a e f g hb

a

b

b

b

Figure 8. 1H NMR spectrum of oxidized canola oil.


Sensory evaluation of lipid oxidation has been conducted by many researchers

(98–100). However, as a subjective method, the reproducibility of sensory analysis

is generally considered worse than that of chemical or instrumental methods. More

recently, use of an electronic nose to monitor the formation of volatile compounds

associated with off-flavors from lipid oxidation has been proposed to supplement

information from human sensory panels (101).

9. MEASUREMENT OF FRYING FAT DETERIORATION

Deep-fat frying is a popular method for food preparation, in which vegetable oils

not only are used as a heat-exchange medium, but also contribute to the quality of

fried products (7). However, lipid oxidation easily occurs at relatively high tem-

peratures, producing a complex series of compounds that exerts undesirable effects

on food flavor and quality (4). The measurement of lipid oxidation, therefore, is

essential to determine its effect on food and oil quality, as well as the useful life

of fats or oils subjected to frying. The oxidative changes in frying fats are charac-

terized by a decrease in the total unsaturation of the fat with increases in the free

fatty acid content, foaming, color, and viscosity as well as the content of polar com-

pounds and polymeric material (4). Quality evaluation of frying fats, may be carried

out in different ways. Physical methods estimate oxidative degradation by monitor-

ing changes in physical properties of frying fats, such as molecular weight, specific

gravity, smoke point, refractive index, chromatic parameter, viscosity, surface

tension, and dielectric constant (4). Generally, rejection point of frying fat is estab-

lished by sensory assessment. Chemical methods include the iodine value, saponi-

fication value, free fatty acid content, peroxide value, TBA value, or p-anisidine

value, among others. PV is less useful because hydroperoxides decompose at about

150�C, and no accumulation of peroxides can be detected.

The extent of oxidation can also be assessed by the analysis of oxidized fatty

acids by spectroscopic means such as IR and NMR techniques (102). Moreover,

GC-MS for volatile profile analysis (103) and HPLC for determination of DNPH

derivatives of nonvolatile higher carbonyl compounds (62) provide qualitative

TABLE 3. A Partial List of Terms Used to Describe

Oxidized Oil.

Flavor-Related Terms Process-Oriented Terms

Buttery Hydrogenated

Nutty Oxidized

Beany Reverted

Grassy Light-struck

Watermelon Rancid

Painty

Fishy

Adapted from (97).

MEASUREMENT OF FRYING FAT DETERIORATION 377

and quantitative evaluation of oxidation in frying fats. Cyclic fatty acids (Figure 9),

which may contain hydroxy and keto groups, are formed during deep frying and can

be measured by chromatography after derivatization (4, 7). Furthermore, determi-

nation of polar material in frying fats is a reliable approach for oil quality evalua-

tion and is an official method in Europe. This method involves separation of fat into

a polar and nonpolar fraction via silica gel chromatography. Nonpolar fat can be

weighed and the total polar material calculated or determined directly by their elu-

tion from the silica gel column (4,7).

Routine analysis for frying fat deterioration has been reviewed by Gertz (104).

Usually, more than two methods are required when using chemical analysis because

no single group of compounds has been identified as a key indicator of oxidative

degradation of frying fats.

10. METHODS FOR MEASURING ANTIOXIDANT ACTIVITY

A variety of natural and synthetic antioxidants are used in fat-containing foods in

order to inhibit lipid oxidation with a wide range of efficiencies, depending on their

properties, concentrations, and processing conditions. The need to measure antiox-

idant activity is well documented. Although numerous methods have been proposed

for measurement of antioxidant activity, the essential features of any test are a sui-

table substrate, an oxidation initiator, and an appropriate measure of endpoint (9).

Therefore, certain aspects should be taken into consideration when selecting a test

for measuring antioxidant activity. These include the model food system used for

(CH2)nH

(CH2)8−nCOOH

(CH2)nH

(CH2)12−nCOOH

n = 1 & 2 n = 3 & 4

COOH

CH3

(CH2)nH

R

(CH2)nH

R

(CH2)nH

R

9 7 5 3

Double bond may present on the positions of C4, C5, C7, C8, or C9.

R = −(CH2)11−nCOOH 1 < n < 11

Figure 9. Chemical structures of cyclic fatty acids formed during deep frying.


the test, and the means by which oxidation is accelerated and monitored (12). Nor-

mally, most assessments of antioxidant activity are performed in oil, or other model

systems, giving sensible prediction for the activity in oil or water-in-oil emulsions,

whereas the results may be misleading for oil-in-water emulsions (12). Further-

more, stripping of oils may be necessary in such evaluations because the endogen-

ous antioxidants in nonstripped oils are found to enhance the oxidative stability of

oils, thus giving rise to erroneous results in the efficiency of antioxidants under

investigation (105–107). In addition to oils and fats, lipid substrates used for testing

antioxidant activity could be fatty acids, fatty acid ethyl esters or triacylglycerols

(9), and b-carotene (108–110). In some cases, such as radical scavenging methods,

no substrate is used. Most test procedures involve initiators to accelerate oxidation.

The combination of increased temperature and oxygen supply, addition of metal

catalysts, and exposure of the reactants to light can reduce the oxidative stability

by a large amount (9, 12). Nevertheless, the elevated temperature may bring about

changes in the oxidation mechanism, thus causing difficulties in the prediction of

TABLE 4. Methods of Expressing Results of Antioxidant Activity Tests.

Method Dimensions

Induction period h, d

Time to reach a set level of oxidation (pre-

induction period)

h, d

Rate of oxidation (pre-induction period) mol kg�1 hr�1, gL�1 d�1

Concentration to produce equivalent effect to

reference antioxidant (pre-induction period)

mol kg�1, gL�1

Concentration of ROOH functional group after

set time period

mequiv. kg�1

Concentration of oxidation product after set

time period

mg kg�1 (ppm w/w)

Scale reading after set time period Absorbance, conductivity, etc.

Free stable radical quenching (DPPH) Percentage inhibition

EC50, concentration to decrease concentration

of test free radical by 50%

TEC50, time to decrease concentration of test

free radical by 50%

Total radical-trapping antioxidant parameter

(TRAP)

mmol peroxy radical deactivated L�1

ABTS assay, phycoerythrin assay TEAC (mM Trolox equivalent to 1-mM test

substance)

Phycoerythrin assay ORAC, oxygen radical absorbance capacity;

mmol of Trolox equivalents

FRAP assay Absorbance of Fe2þ complex at 593 nm

produced by antioxidant reduction of

corresponding tripyridyltriazine Fe3þ

complex

Metal chelating assay Percentage of inhibition of ferrozine-Fe2þ

complex formation

NOTE: Also see Tables 1 and 2 for other tests applicable to antioxidant activity determination.

Adapted from (9).

METHODS FOR MEASURING ANTIOXIDANT ACTIVITY 379

antioxidant activity at low temperatures as compared with those at high tempera-

tures (9, 12). After the substrate is oxidized under standard conditions, the oxidation

is monitored by chemical, instrumental, or sensory methods. An appropriate mea-

sure of endpoint is essential for assessing antioxidant activity. Analytical strategies

for endpoint determination include measurement at a fixed time point, measurement

of reaction rate, lag phase measurement, and integrated rate measurement (9). The

resulting antioxidant activity is expressed using a wide range of parameters (Table 4).

Approaches proposed for testing antioxidant activity include measuring of the

current state of oil samples, as discussed above, and radical scavenging assays, which

are gaining popularity in the evaluation of antioxidant activity. Radical scavenging

methods measure the relative abilities of antioxidants to scavenge synthetic radicals

or natural in comparison with the antioxidant potency of a standard antioxidant

compound (111). Trolox (6-hydroxy-2,5,7,8-tetramethylchroma-2-carboxylic acid),

ascorbic acid, and quercetin are among the standard antioxidants frequently

used. The most commonly used synthetic radicals are DPPH (2,2-diphenyl-1-

picrylhydrazyl) and ABTS (3-ethylbenzthiazoline-sulfonic acid) radicals. DPPH test

(112–116) and ABTS assay (117–122) are simple, rapid, and involve no substrate.

However, it has been suggested that these artificial substrate-free methods do not

always adequately mimic the processes in food systems, which sometimes makes

them less valuable for predicting the effectiveness of the antioxidant in foods (9).

Other measurements of antioxidant activity include FRAP (ferric reducing-

antioxidant power) (123–126), TRAP (total radical-trapping antioxidant parameter)

(123, 127), phycoerythrin assay (128, 129), and test of metal chelating capacity

(130, 131), among others. Reviews on methods for testing antioxidant activity

have been published (9, 12).

11. CONCLUSIONS AND RECOMMENDATIONS

Lipid oxidation may be assessed in many ways, among which changes in the initial

reactants and formation of oxidation products are most commonly assessed. Mean-

while, sensory analysis assesses both the subjective and, in some cases, objective

measurements of oxidative changes in foods. Each method shows both advantages

and disadvantages, thus it is important to select the most adequate method, depend-

ing on the system under investigation and the state of oxidation itself. The use of

two or more methods assessing both primary and secondary oxidation products is

highly recommended.

REFERENCES

1. F. Shahidi, Nahrung., 44, 158–163 (2000).

2. J. R. Vercellotti, A. J. St. Angelo, and A. M. Spanier, in A. J. St. Angelo, ed., Lipid

Oxidation in Food, ACS Symposium Series 500, American Chemical Society,

Washington, D.C., 1992, pp. 1–11.


3. M. Gordon, in J. Pokorny, N. Yanishlieva, and M. Gordon, eds., Antioxidants in Food:

Practical Applications, Woodhead Publishing, Ltd., Cambridge, England, 2001,

pp. 7–21.

4. E. G. Perkins, in A. J. St. Angelo, ed., Lipid Oxidation in Food, American Chemical

Society, Washington, D.C., 1992, pp. 310–321.

5. A. Kamal-Eldin, M. Makinen, and A. M. Lampi, in A. Kamal-Eldin, ed., Lipid Oxidation

Pathways, AOCS Press, Champaign, Illinois, 2003, pp. 1–36.

6. J. Velasco, M. L. Anderson, and L. H. Skibsted, Food Chem., 85, 623–632 (2004).

7. F. Shahidi and U. N. Wanasundara, in C. C. Akoh and D. B. Min, eds., Food Lipids:

Chemistry, Nutrition and Biotechnology, Marcel Dekker, Inc., New York, 2002,

pp. 465–487.

8. M. C. Dobarganes and J. Velasco, Eur. J. Lipid Sci. Technol., 104, 420–428 (2002).

9. M. Antolovich, P. D. Prenzler, E. Patsalides, S. Mcdonald, and K. Robards, Analyst, 127,

183–198 (2002).

10. I. Niklova, S. Schmidt, K. Habalova, and S. Sekretar, Eur. J. Lipid Sci. Technol., 103,

299–306 (2001).

11. J. Velasco and C. Dobarganes, Eur. J. Lipid Sci. Technol., 104, 661–676 (2002).

12. M. Gordon, in J. Pokorny, N. Yanishlieva, and M. Gordon, eds., Antioxidants in

Food: Practical Applications, Woodhead Publishing, Ltd., Cambridge, England,

2001, pp. 71–84.

13. C. Genot, G. Kansci, and M. Laroche, Sciences des Aliments., 14, 673–682 (1994).

14. H. J. Chung, A. S. Colakoglu, and D. B. Min, J. Food Sci., 69, 83–88 (2004).

15. S. L. Melton, Food Technol., 37, 105–111 (1983).

16. G. O. Fruhwirth, T. Wenzl, R. El-Toukhy, F. S. Wagner, and A. Hermetter, Eur. J. Lipid

Sci. Technol., 105, 266–274 (2003).

17. B. J. F. Hudson, in J. C. Allen and J. Hamilton, eds., Rancidity of Foods, Applied Science

Publishers, London, England, 1983, pp. 47–58.

18. A. Riuz, M. J. Ayora-Canada, and B. Lendl, Analyst, 126, 242–246 (2001).

19. G. Yildiz, R. L. Wehling, and S. L. Cuppett, J. Amer. Oil Chem. Soc., 80, 103–107 (2003).

20. A. Kiritsakis, A. Kanavouras, and K. Kiritsakis, Eur. J. Lipid Sci. Technol., 104, 628–638

(2002).

21. Determination of Peroxide Value, in IUPAC: Standard Methods for the Analysis of Oils,

Fats and Derivatives, IUPAC Standard Method 2.501, International Union of Pure and

Applied Chemistry, Pergamon, Oxford, England, 1992.

22. D. L. Berner, INFORM, 1, 884–886 (1996).

23. S. Eymard and C. Genot, Eur. J. Lipid. Sci. Technol., 105, 497–501 (2003).

24. T. Asakawa and S. Matsushita, J. Amer. Oil. Chem. Soc., 55, 691–620 (1978).

25. S. Hara and Y. Totani, J. Amer. Oil Chem. Soc., 65, 1948–1950 (1988).

26. M. Hicks and J. M. Gebicki, Anal. Biochem., 99, 249–253 (1979).

27. J. M. Gebicki and J. Guille, Anal. Bichem., 176, 360–364 (1989).

28. Z. Y. Jiang, A. C. S. Woollard, and S. P. Wolff, Lipids, 26, 853–856 (1991).

29. S. P. Wolff, Methods Enzymol., 233, 182–189 (1994).

30. J. M. DeLong, R. K. Prange, D. M. Hodges, C. F. Forney, M. C. Bishop, and M. Quillian,

J. Agric. Food Chem., 50, 248–254 (2002).

REFERENCES 381

31. J. Sedman, F. R. van de Voort, and A. A. Ismail, in R. E. McDonald and M. M. Mossoba,

eds., New Techniques and Applications in Lipid Analysis, AOCS Press, Champaign,

Illinois, 1997, pp. 283–324.

32. J. Dong, K. Ma, F. R. van de Voort, and A. A. Ismail, J. AOAC Int., 80, 345–352 (1997).

33. T. A. Russin, F. R. van de Voort, and J. Sedman, J. Amer. Oil Chem. Soc., 80, 635–641

(2003).

34. B. Innawong, P. Mallikarjunan, J. Irudayaraj, and J. E. Marcy, Lebensm. Wiss. U.

Technol., 37, 23–28 (2004).

35. T. Miyazawa, K. Fujimoto, and T. Kandeda, Agric. Biol. Chem., 51, 2569–2573 (1987).

36. A. M. Jimenez and M. J. Navas, Grasas y Aceites, 53, 64–75 (2002).

37. J. P. Wold and M. Mielnik, J. Food Sci., 65, 87–95 (2000).

38. C. Stauffer, in Fats & Oils, Eagan Press, St. Paul, Minnesota, 1996, pp. 15–27.

39. F. Shahidi, U. Wanasundara, and N. Brunet, Food Res. Int., 27, 555–562 (1994).

40. U. N. Wanasundara, F. Shahidi, and C. R. Jablonski, Food Chem., 52, 249–253 (1995).

41. E. Kishida, S. Tokumaru, Y. Ishitani, M. Yamamoto, M. Oribe, H. Iguchi, and S. Kojo,

J. Agric. Food Chem., 41, 1598–1600 (1993).

42. S. Cesa, J. Agric. Food Chem., 52, 2119–2122 (2004).

43. D. Jardine, M. Antolovich, P. D. Prenzler, and K. Robards, J. Agric. Food Chem., 50,

1720–1724 (2002).

44. A. de las Heras, A. Schoch, M. Gibis, and A. Fischer, Eur. Food Res. Technol., 217, 180–

184 (2003).

45. D. V. Hoyland and A. J. Taylor, Food Chem., 40, 271–291 (1991).

46. B. G. Tarladgis, B. M. Watts, M. T. Younathan, and L. Dugan, J. Amer. Oil Chem. Soc.,

37, 44–48 (1960).

47. F. Shahidi and C. Hong, J. Food Biochem., 15, 97–105 (1991).

48. J. Pikul, D. E. Leszczynski, and F. A. Kummerow, J. Agric. Food Chem., 37, 1309–1313

(1989).

49. C. Wang, L. Zhu, and M. S. Brewer, in 1996 IFT annual meeting: Book of Abstracts,

pp. 161.

50. R. Guillen-Sans and M. Guzman-Chozas, Crit. Rev. Food Sci. Nutr., 38, 315–330

(1998).

51. Q. Sun, C. Faustman, A. Senecal, A. L. Wilkinson, and H. Furr, Meat Sci., 57, 55–60

(2001).

52. F. Shahidi, L. J. Rubin, L. L. Diosady, and D. F. Wood, J. Food Sci., 55, 274–275 (1985).

53. R. Marcuse and J. Pokorny, Fett. Wiss. Technol., 96, 185–187 (1994).

54. H. de A. Gomes, E. N. da Silva, M. R. Lopes do Nascimento, and H. T. Fukuma, Food

Chem., 80, 433–437 (2003).

55. F. Doleschall, Z. Kemeny, K. Recseg, and K. Kovari, Eur. J. Lipid Sci. Technol., 104, 14–

18 (2002).

56. G. H. van der Merwe, L. M. du Plessis, and J. RN. Taylor, J. Sci. Food Agric., 84, 52–58

(2003).

57. G. R. List, C. D. Evans, W. F. Kwolek, K. Warner, and B. K. Boundy, J. Amer. Oil Chem.

Soc., 51, 17–21 (1974).

58. M. D. Guillen and N. Cabo, Food Chem., 77, 503–510 (2002).


59. P. J. White, in K. Warner and N. A. M. Eskin, eds., Methods to Assess Quality and

Stability of Oils and Fat-containing Foods, AOCS Press, Chamaign, Illinois, 1995,

pp. 159–178.

60. U. N. Wanasundara and F. Shahidi, J. Food Lipids, 2, 73–86 (1995).

61. T. Chiba, M. Takazawa, and K. Fujinoto, J. Amer. Oil Chem. Soc., 66, 1588–1592 (1989).

62. E. Schulte, Anal. Bioanal. Chem., 372, 644–648 (2002).

63. F. Shahidi and R. B. Pegg, J. Food Lipids, 1, 177–186 (1994).

64. C. F. Goodridge, R. M. Beaudry, J. J. Pestka, and D. M. Smith, J. Agric. Food Chem., 51,

4185–4190 (2003).

65. H. Z. Zhang and T. C. Lee, IFT Annual Meeting, 1995, pp. 161.

66. C. F. Goodridge, R. M. Beaudry, J. J. Pestka, and D. M. Smith, J. Agric. Food Chem., 51,

7533–7539 (2003).

67. U. N. Wanasundara and F. Shahidi, Paper presented at the Canadian Section of American

Oil Chemists’ Society’s Annual Meeting, Guelph, Ontario, Canada, November 15–16,

1995.

68. F. Shahidi and S. A. Spurvey, J. Food Lipids, 3, 13–25 (1996).

69. K. Schwarz, G. Bertelsen, L. R. Nissen, P. T. Gardner, M. I. Heinonen, A. Hopia,

T. Huynh-Ba, P. Lambelet, D. McPhail, L. H. Skibsted, and L. Tijburg, Eur. Food Res.

Technol., 212, 319–328 (2001).

70. K. Miura, H. Kikuzaki, and N. Nakatani, J. Agric. Food Chem., 50, 1845–1851 (2002).

71. S. de la Presa-Owens, M. C. Lopez-Sabater, and M. Rivero-Urgell, J. Agric. Food Chem.,

43, 2879–2882 (1995).

72. M. P. Canizares-Macias, J. A. Garcia-Mesa, and M. D. Luque de Castro, Anal. Bioanal.

Chem., 378, 479–483 (2004).

73. C. P. Tan, Y. B. Che-Man, J. Selamat, and M. S. A. Yusoff, Food Chem., 76, 385–389

(2002).

74. R. Aparicio, L. Roda, M. A. Albi, and F. Gutierrez, J. Agric. Food Chem., 47, 4150–4155

(1999).

75. A. Faroop, M. I. Bhanger, and T. G. Kazi, J. Amer. Oil Chem. Soc., 80, 151–155 (2003).

76. L. C. Boyd, V. C. Nwosu, C. L. Young, and L. MacMillian, J. Food Lipids, 5, 269–282

(1998).

77. A. L. Wilkinson, Q. Sun, A. Senecal, and C. Faustman, J. Food Sci., 66, 20–24 (2001).

78. M. L. Andersen and L. H. Skibsted, Eur. J. Lipid Sci. Technol., 104, 65–68 (2002).

79. C. U. Carlsen, M. L. Andersen, and L. H. Skibsted, Eur. Food Res. Technol., 213, 170–

173 (2001).

80. M. K. Thomsen, D. Kristensen, and L. H. Skibsted, J. Amer. Oil Chem. Soc., 77, 725–730

(2000).

81. L. R. Nissen, Tuong-Huynh-Ba, M. A. Petersen, G. Bertelsen, and L. H. Skibsted, Food

Chem., 79, 387–394 (2002).

82. B. Halliwell and J. M. C. Gutteridge, in Free Radicals in Biology and Medicine, Oxford

University Press, New York, 1999.

83. D. M. Lee, Methods Enzymol., 234, 513–523 (1994).

84. C. P. Tan and Y. B. Che Man, Trends Food Sci. Technol., 13, 312–318 (2002).

85. C. P. Tan and Y. B. Che Man, Food Chem., 67, 177–184 (1999).

REFERENCES 383

86. G. Litwinienko, A. Daniluk, and T. Kasprzycka-Guttman, J. Amer. Oil Chem. Soc., 76,

655–657 (1999).

87. C. P. Tan and Y. B. Che Man, Lipid Technol., 14, 112–115 (2002).

88. B. Kowalski, K. Ratusz, D. Kowalska, and W. Bekas, Eur. J. Lipid Sci. Technol., 106,

165–169 (2004).

89. R. L. Hassel, J. Amer. Oil Chem. Soc., 53, 179–181 (1976).

90. M. D. Guillen and A. Ruiz, Trends Food Sci. Technol., 12, 328–338 (2001).

91. F. Shahidi, U. Wanasundara, and N. Brunet, Food Res. Int., 27, 555–562 (1994).

92. U. Wanasundara and F. Shahidi, J. Food Lipids, 1, 15–24 (1993).

93. S. P. J. Namal-Senanayake and F. Shahidi, J. Food Lipids, 6, 195–203 (1999).

94. I. Medina, R. Sacchi, I. Giudicianni, and S. Aubourg, J. Amer. Oil Chem. Soc., 75, 147–

154 (1998).

95. F. J. Hidalgo, G. Gomez, J. L. Navarro, and R. Zamora, J. Agric. Food Chem., 50, 5825–

5831 (2002).

96. B. W. L. Diehl, in R. J. Hamilton, ed., Lipid Analysis in Oils and Fats, Blackie Academic

& Professional, London, 1998, pp. 87–135.

97. G. V. Cinille and C. A. Dus, in A. J. St. Angelo ed., Lipid Oxidation in Food, American

Chemical Society, Washington, D.C., 1992, pp. 279–289.

98. C. J. Broadbent and O. A. Pike, J. Amer. Oil Chem. Soc., 80, 59–63 (2003).

99. E. A. Coppin and O. A. Pike, J. Amer. Oil Chem. Soc., 78, 13–18 (2001).

100. M. Timm-Heinrich, X. Xu, N. S. Nielsen, and C. Jacobsen, Eur. J. Lipid Sci. Technol.,

105, 436–448 (2003).

101. N. Shen, S. Moizuddin, L. Wilson, S. Duvick, P. White, and L. Pollak, J. Amer. Oil Chem.

Soc., 78, 937–940 (2001).

102. S. Khatoon and A. G. G. Krishna, J. Food Lipids, 5, 247–267 (1998).

103. V. Rosales and J. de D, Dissertation Abstracts International, B., 57, 7290 (1997).

104. C. Gertz, Lipid Technol., 13, 44–47 (2001).

105. T. Keceli and M. H. Gordon, J. Sci. Food Agric., 81, 1391–1396 (2001).

106. M. A. Khan and F. Shahidi, J. Amer. Oil Chem. Soc., 77, 963–968 (2000).

107. V. Povilaityte, M. E. Cuvelier, and C. Berset, J. Food Lipids, 8, 45–64 (2001).

108. F. Shahidi, C. Desilva, and R. Amarowicz, J. Amer. Oil Chem. Soc., 80, 443–450

(2003).

109. F. Shahidi and R. Amarowicz, J. Amer. Oil Chem. Soc., 73, 1197–1199 (1996).

110. R. Amarowicz, M. Naczk, and F. Shahidi, J. Amer. Oil Chem. Soc., 77, 957–961 (2000).

111. O. A. Zaporozhets, O. A. Krushynska, N. A. Lipkovska, and V. N. Barvinchenko,

J. Agric. Food Chem., 52, 21–25 (2004).

112. D. Bandoniene, M. Murkovic, W. Pfannhauser, P. R. Venskutonis, and D. Gruzdiene, Eur.

Food Res. Technol., 214, 143–147 (2002).

113. N. Nenadis and M. Tsimidou, J. Amer. Oil Chem. Soc., 79, 1191–1195 (2002).

114. H. Matsuura, H. Chiji, C. Asakawa, M. Amano, T. Yoshihara, and J. Mizutani, Biosci.

Biotechnol. Biochem., 67, 2311–2316 (2003).

115. P. J. Park, F. Shahidi, and Y. J. Jeon, J. Food Lipids, 11, 15–28 (2004).

116. R. Amarowicz, B. Raab, and F. Shahidi, J. Food Lipids, 10, 51–62 (2003).


117. A. Podsedek, D. Sosnowska, and B. Anders, Eur. Food Res. Technol., 217, 296–300

(2003).

118. A. M. Alonso, D. A. Guillen, and C. G. Barroso, Eur. Food Res. Technol., 216, 445–448

(2003).

119. N. J. Miller and C. Rice-Evan, Redox Rep., 2, 161–171 (1996).

120. C. E. Rice-Evans and N. J. Miller, Methods Enzymol., 234, 279 (1994).

121. L. Bramati, F. Aquilano, and P. Pietta, J. Agric. Food Chem., 51, 7472–7474 (2003).

122. N. J. Miller and C. A. Rice-Evans, Free Rad. Res., 26, 195–199 (1997).

123. N. Pellegrini, M. Serafini, B. Colobi, D. del Rio, S. Salvatore, M. Bianchi, and

F. Brighenti, J. Nutr., 133, 2812–2819 (2003).

124. I. E. F. Benzie and J. J. Strain, Anal. Biochem., 239, 70 (1996).

125. I. E. F. Benzie and J. J. Strain, Methods Enzymol., 299, 15 (1999).

126. J. Chen, H. Lindmark-Mansson, L. Gorton, and B. Akesson, Int. Dairy J., 13, 927–935

(2003).

127. A. Ghiselli, M. Serafini, G. Maiani, E. Azzini, and A. Fero-Luzzi, Free Rad. Biol. Med.,

18, 29–36 (1995).

128. O. Boxin, M. Hampsch-Woodill, and R. L. Prior, J. Agric. Food Chem., 49, 4619–4626

(2001).

129. S. McDonald, P. D. Prenzler, M. Antolovich, and K. Robards, Food Chem., 73, 73–84

(2001).

130. S. N. El and S. Karakaya, Int. J. Food Sci. Nutr., 55, 67–74 (2004).

131. M. Oktay, I. Gulcin, and O. I. Kufrevioglu, Lebensm. Wiss. U. Technol., 36, 263–271

(2003).

REFERENCES 385

Lipid Oxidation Measurement Method

Documents