LWT - Food Science and Technology 63 (2015) 972e977

Contents lists available at ScienceDirect

LWT - Food Science and Technology

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

Comparison of fat determination methods depending on fat definition

in bakery products
Jae-Min Shin a, b, Seung-Kook Park b, *
a
Seoul Metropolitan Government Research Institute of Public Health and Environment, #202-3, Yangjae-dong, Seocho-gu, Seoul, 137-130, Republic of Korea
b
Department of Food Science and Biotechnology, Kyung Hee University, Seocheon-dong, Giheung-gu, Yongin-si, Gyeonggi-do, 446-701, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: This study aimed to evaluate fat determination depending on fat definition. Three methods were used for
Received 23 December 2014 the determination of fat classes and content in bread products: the Folch method, the automated Soxhlet
Received in revised form method, and the AOAC 996.06 method. The results using these methods were compared. Fat (crude)
30 March 2015
extracted by the Folch and Soxhlet methods was gravimetrically determined and assessed by fat classes
Accepted 5 April 2015
Available online 14 April 2015
using capillary gas chromatography. In most samples, fat (total) content determined by the AOAC 996.06
method was lower than the fat (crude) content determined by the Folch or automated Soxhlet methods.
In addition, the contents of saturated, monounsaturated, polyunsaturated and trans fat determined by
Keywords:
Fat definition
the AOAC 996.06 method were lower than those obtained by the Folch and automated Soxhlet methods.
Fat analysis The analytical fat determination method in food depends on how fat is defined, which produced dif-
Nutrition labeling ference in fat content and content of fat classes.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction expected, due to both the reformulation of products and increased

consumer awareness (Craig-Schmidt, 2006).
Nutritional labeling of food impacts not only on consumers and There are significant differences in nutritional facts labels
the food industry, but on international trade as well. There is some among countries. Because quantitative methods to measure nu-
variation among different countries in the criteria for nutrition facts trients are differently developed for each country. Some countries
labeling. In general, calories, carbohydrates, protein, fat and sodium (Korea, Japan, and Europe, etc) list crude fat content in the ‘Fat’
are listed on nutrition facts labels. In the recent trend, food labels section on food labels (EEC 1990; KFDA 2009; Phamaceutical
also include sugars, saturated fat, trans fat, and cholesterol. Here, Society of Japan, 2010), whereas other countries (United States,
attention needs to be paid to fat. Nutritional facts labeling initially Canada, and Australia etc.) state total fat on a food label (CFIA, 2011;
listed only the fat content, but now, different types of fats are listed FSANZ, 2013; Hawkes, 2004). However, ‘fat’ listed on a food label is
on the label. This is because the high consumption of saturated fatty not the same as ‘total fat’. The definition of fat used in Europe,
acids and cholesterol is mainly responsible for hypercholesterole- Korea, and Japan is that of crude fat, so fat stated on a food label is
mia, which is in turn responsible for the increase in cardiovascular crude fat (interpreted to be food fat). Crude fat is a term used to
morbidity and mortality of ischemic origin (Neaton & Wentworth, refer to the crude mixture of fat-soluble material present in foods.
1992). Consumption of trans fatty acid may increase the risk of The crude fat content of food has traditionally been determined by
coronary heart disease (Ascherio, Katan, Zock, Stampfer, & Willett, methodologies that involve extraction with organic solvents, dry-
1999; Vijver et al., 2000). Therefore, FDA requires the declaration of ing of the extract, and a gravimetric determination of fat. These
the amount of trans fatty acid present in foods, including dietary classic methods include the Folch method, the Soxhlet method, as
supplements, on the nutrition label (FDA, 2003). This rule took well as the Weibull-Berntrop, and the Rose-Gottlieb methods. The
effect on January 2006, and a sharper decline in trans fat intake was definition of ‘total fat’, as established by the US Food and Drug
Determination (FDA) in 1990 through the Nutritional Labeling and
Education Act (NLEA), is ‘the sum of all fatty acids obtained in a lipid
extract, expressed as triglycerides’ (Federal Register, 1993), this
change in definition for nutritional labeling purposes requires a
* Corresponding author. Tel.: þ82 31 201 2655; fax: þ82 31 204 8116.
methodology change, from extraction and gravimetric analysis to
E-mail address: skpark@khu.ac.kr (S.-K. Park).

http://dx.doi.org/10.1016/j.lwt.2015.04.011
0023-6438/© 2015 Elsevier Ltd. All rights reserved.
 J.-M. Shin, S.-K. Park / LWT - Food Science and Technology 63 (2015) 972e977 973

GC analysis of fatty acid methyl esters (FAMEs). However, in Hong 0 (for some products the trans fat value on the nutrition facts was 0)
Kong and Brazil, the definition of total fat corresponds to that of in Korea (KFDA, 2011). Products were ground with a grinder (OMNI
crude fat, so gravimetric methods are accepted to determine total International, Kennesaw, GA, USA) to achieve a homogenate. Ho-
fat in those nations (Brazil, 2003; CFS, 2011). mogenized samples were subjected to fat extraction. The ground
The Folch (Folch, Lees, & Sloane Stanley, 1957) and Bligh-Dyer samples were then stored at 20  C.
(Bligh & Dyer, 1959) methods are considered the classical and
most reliable means for quantitatively extracting lipids from 2.2. Reagents and standards
various types of animal tissues and bakery products. Due to its mild
working conditions (neither high temperature nor pressure Chloroform and methanol, iso-octane, and diethyl ether were
required), the Folch method was used as a reference method in HPLC grade. All other reagents were ACS grade. The internal stan-
studies in relation to the extraction of fats (Pe
rez-Palacios, Ruiz, dard for GC analysis was triundecanoin (C11:0) purchased from
Martín, Muriel, & Antequera, 2008; Ruiz-Mime
nez, Priego-Capote, SigmaeAldrich (St. Louis, MO, USA), and the fatty acid standard was
& Luque de Castro, 2004). the Supelco 37 Component FAME mix (Bellefonte, PA, USA).
Lipid content has been estimated using Soxhlet, which is the
official recommended method. However, the relatively long 2.3. Folch method
extraction time (16e24 h) and the high temperatures needed for
Soxhlet extraction are its main shortcomings. Automated Soxhlet Fifteen grams of sample were mixed with 300 mL chloroform/
extractors have several advantages, including shorter extraction methanol (2:1, v/v) and were shaken for 20 min, and 60 mL of
time, decreased extractant volume, and simultaneous extraction of distilled water was added and shaken for 5 min. After phase sep-
several samples (Luque de Castro & Priego-Capote, 2010). aration, the chloroform phase was filtered through sodium sulfate
AOAC 996.06 is a universally accepted method for determining then collected, evaporated in a rotary evaporator at 45  C, and the
total, saturated, polyunsaturated, and monounsaturated fats in residue was further dried in an oven at 104  C for 1 h and then
food, and has sufficient accuracy and repeatability to satisfy current cooled in a desiccator. The extracted crude fat was recorded
USA nutrition labeling regulations (Ngeh-Ngwainbi, Lin, & gravimetrically. Approximately 25 mg of fat was transferred to a
Chandler, 1997; Ratnayake, 2004). In this method, the tri- test tube to determine saturated, polyunsaturated, mono-
acylglycerol (fat) and fatty acids are extracted from food, and are unsaturated, and trans fat. The fat was methylated and then
then methylated to fatty acid methyl ester using BF3 in methanol, analyzed by capillary GC-flame ionization detection (FID).
and FAMEs are quantitatively measured by capillary gas chroma-
tography. Total fat is calculated as the sum of individual fatty acids 2.4. Automated Soxhlet method
expressed as triglyceride equivalents (AOAC, 2005a, b).
Fat extractions have been carried out in many studies. However, One gram of homogenized sample was placed in a thimble
only few studies (Ali, Angyal, Weaver & Rader, 1997; Rader et al., (22 mm  28 mm, i. d.; Foss North America, Eden Prairie, MN, USA),
1995; Zou, Lusk, Messer, & Lane, 1999) have compared different fitted with metal adaptors, which was then loaded into an auto-
definitions of extracted fats, which had been performed in the mated Soxtherm fat extraction system (Gerhardt, Germany). Briefly,
1990s when the definition of “total fat” was emerging. However, beakers that had been dried in an oven at 104  C and weighed were
these studies did not include comparison of other fats such as placed beneath each extraction thimble in the extraction unit, and
saturated, polyunsaturated, or trans fat. Our previous study (Shin diethyl ether was added to each of the six extraction chambers.
et al., 2013) was performed with cookies and biscuits, and this After extraction, the extract was dried in an oven at 104  C and was
study is a continuation with bread products. These bakery products then cooled in a desiccator prior to gravimetric fat determination.
usually contain butter, cream, shortening and other processed fat, The beakers were weighed, and the percent crude fat was calcu-
thus determination of fat content and fat classes in these bakery lated. To determine saturated, polyunsaturated, monounsaturated
products are important. fat and trans fat content, approximately 25 mg fat from the beakers
Therefore, the aim of this study was to investigate the difference was transferred to a test tube. The fat was methylated and then
between these parameters by obtaining crude fat content (by analyzed by capillary GC-FID.
gravimetric methods) and total fat content (by GC analysis) present
in bread products, and by comparing these results. The crude fat 2.5. AOAC 996.06 method
was extracted from bread products using both the Folch method
and automated Soxhlet method, and the total fat was extracted A ground and homogenized sample (containing 100e200 mg
using the AOAC 996.06 method. The fatty acid profile of the fat) was accurately weighed and placed in a Mojonnier flask.
extracted crude fat was also determined by gas chromatography. Approximately 100 mg of pyrogallic acid was added, followed by
The results were compared to saturated, monounsaturated, poly- 2 mL of triundecanoin (5 mg/mL, C11:0) internal standard solution,
unsaturated and trans fat values obtained by the AOAC 996.06 2 ml of ethanol, and 10 mL of 8.3 M HCl. The flask was placed in a
method. shaking water bath at 70e80  C, set at moderate agitation speed,
and maintained for 40 min. After cooling to room temperature
2. Materials and methods (20e25  C), diethyl ether and petroleum ether were added and
shaken. The solutions were allowed to stand for more than 1 h until
2.1. Samples and the top and bottom layers were clearly separated. The ether
(top) layer was collected and evaporated. The remaining residue
Twelve bread products purchased from four bakeries were used. contained the extracted fat. To determine total, saturated, poly-
The products were categorized as croissants, pastries, pies, others. unsaturated, monounsaturated fat and trans fat content, extracted
Reference values for these bakery products came from the food fat from the beakers was transferred to a test tube. The fat was
labels provided by the manufacturers. All food labels contained fat, methylated and then analyzed by capillary GC-FID. Each fatty acids
saturated fat, and trans fat content, but only fat content and satu- was converted to its triglyceride equivalents summed to obtain
rated fat content were used as reference values in this study, total fat. Saturated, polyunsaturated, monounsaturated fat and
because trans fat content <0.2 g per serving size can be expressed as trans fat were calculated as the sum of respective fatty acids.
974 J.-M. Shin, S.-K. Park / LWT - Food Science and Technology 63 (2015) 972e977

2.6. Methylation of fatty acids (Folch and automated Soxhlet When fat (crude) contents, as extracted by the Folch and auto-
methods) mated Soxhlet methods, were compared to fat (total) content, as
determined by AOAC 996.06, the fat (crude) contents were higher
Approximately 25 mg of the fat extracted using each method than fat (total) content for each product (Table 1). The Folch method
was added to a test tube, and 1 mL of triundecanoin (1 mg/mL, determined fat content values which were 2.78%e5.40% higher, and
C11:0) internal standard solution was added, followed by 1.5 mL of the automated Soxhlet method determined values which were
0.5 NeNaOH/MeOH. The test tube was then sealed with a screw cap 2.18%e5.69% higher than those of the AOAC 996.06. In particular,
and shaken. The tube was heated in a heating block at 100  C for products containing cream (pie with custard cream, and morning
5 min. After it was cooled to room temperature, 2 mL of 14% BF3 was roll with cream) showed a greater difference between fat (crude)
added. The test tube was sealed once again and shaken. It was contents determined by the Folch method and the automated
heated on a heating block at 100  C for 30 min. After it cooled to Soxhlet method, and fat (total) content determined by the AOAC
30e40  C, 1 mL of iso-octane was added. The tube was sealed, and 996.06 method. However, there was no observed difference be-
the contents were mixed on a Vortex mixer. Then, 5 mL of saturated tween crude and total fat for two samples (apple pie, stick pie)
NaCl solution was added. After layer separation, the upper iso- (p < 0.05).
octane layer was transferred to an auto sampler vial for GC analysis. The study of Rader et al. (1995), in accordance with this study,
showed that values for crude fat determined gravimetrically were
2.7. Methylation of fatty acids (AOAC 996.06 method) higher than values for total fat determined chromatographically,
noting that higher gravimetric crude fat values were attributed to
Extracted fat residue was dissolved in 2e3 mL chloroform and the weight of material other than fat in the extraction residues. In
2e3 mL diethyl ether. The mixture was transferred to a test tube addition, Food Analysis (2010) shows that GC analysis of fat content
and then evaporated to dryness in a 40  C water bath under a ni- by the Soxhlet method is faster and cheaper than that by the AOAC
trogen stream. BF3 reagent (2.0 mL of 7%) and 1.0 mL toluene were Method 996.06, but results in a higher estimation of fat content.
added. The test tube was sealed with a screw cap and heated in an For all samples, except for the apple pie sample, the AOAC
oven at 100  C for 45 min, with shaking every 10 min. The test tube 996.06 method resulted in a lower amount of monounsaturated fat
was allowed to cool to room temperature, and 5.0 mL of H2O, 1.0 mL than that of the other two methods (Table 2). In three samples
of hexane, and approximately 1.0 g Na2SO4 were added. The test (croissant-1, pie with custard cream, morning roll with cream), the
tube was capped and shaken for 1 min. The layers were then automated Soxhlet method resulted in a higher amount of mono-
allowed to separate, and then the upper layer was carefully trans- unsaturated fat than that determined by the Folch method. In two
ferred to an auto sampler vial for GC analysis. samples (croissant-2, pastry doughnut), the Folch method resulted
in a higher amount of monounsaturated fat than that of the auto-
2.8. GC analysis mated Soxhlet method. In the remaining seven samples, the
amounts of monounsaturated fat obtained by the automated
The FAME vial was immediately analyzed for total fat (AOAC Soxhlet and Folch methods were not significantly different
996.06 method) and fatty acids (Folch, automated Soxhlet, and (p < 0.05).
AOAC 996.06 methods) using an Agilent Technologies 6890N GC- The AOAC 996.06 method showed the lowest amount of satu-
FID (Santa Clara, CA, USA) and a Supelco 2560 fused-silica capil- rated fat for most samples (Table 3). For apple pie and muffin, there
lary column (100 m  0.25 mm i. d., 0.2 mm film thickness). Helium was no difference in saturated fat values obtained by the three
was the carrier gas, with a flow velocity of 0.75 mL/min. The split methods (p < 0.05). The Folch and automated Soxhlet methods
ratio was 200:1. A single injection of 1 mL was made per sample obtained higher amounts of saturated fat than that of AOAC 996.06
replicate. The injector temperature was 225  C, and the detector method (0.91%e2.37% higher, and 0.3%e3.81% higher, respectively).
temperature was 285  C. Hydrogen and air flows were set to 40 and In the samples containing cream (pie with custard cream, and
400 mL/min, respectively. Oven temperature programming con- morning roll with cream), both the Folch method and the auto-
sisted of an initial temperature of 100  C which was held for 4 min, mated Soxhlet method determined much higher amounts of satu-
and was then increased at a rate of 3  C/min to 240  C, and held for rated fat than that determined by the AOAC 996.06 method. The
15 min at this temperature. Nitrogen was the make-up gas with a saturated fat content obtained by the Folch method was higher
flow rate of 30 mL/min. FAMEs were measured against the C11:0 than that by the automated Soxhlet method for two samples
internal standard; the Supelco 37 component FAME mix was used (croissant-2, and pastry doughnut). However, in the remaining 10
to identify and quantify individual fatty acids. samples, the two methods showed no difference in terms of satu-
rated fat content.
2.9. Statistical methods The AOAC 996.06 method resulted in a lower amount of fat,
saturated fat, and monounsaturated fat than the other two
Values for crude, total, saturated, polyunsaturated, mono- methods, but the AOAC 996.06 and Soxhlet methods jointly
unsaturated, and trans fat are expressed as means ± standard de- determined the lowest levels of polyunsaturated fat content
viations of three replicates. Differences between fat and fat class (Table 4). The automated Soxhlet method did not extract a higher
content within each sample, by the three methods, were compared level of polyunsaturated fat content than that of the Folch method
by one-way analysis of variance using the General Linear Model of for most samples. In the samples containing cream (pie with cus-
SPSS v. 12.0 (Chicago, IL, USA). When a significant difference tard cream, and morning roll with cream), however, the Soxhlet
(p < 0.05) was detected, comparative analyses were conducted method resulted in a higher amount of polyunsaturated fat than
using Duncan's test. that of the Folch method. In addition, the Soxhlet method produced
a higher amount of monounsaturated, saturated, and trans fat than
3. Results and discussion that of the Folch method in the two cream-containing samples.
Similar trans fat contents were observed in one analyzed bakery
The fat contents of bread products were determined by the product (apple pie) among the three methods. However significant
Folch, automated Soxhlet and AOAC 996.06 methods, and the re- differences were observed for the remaining eleven samples
sults were compared. (Table 5). In most samples, the AOAC 996.06 resulted in smaller
 J.-M. Shin, S.-K. Park / LWT - Food Science and Technology 63 (2015) 972e977 975

Table 1
Fat content (%) in 12 different samples, as obtained by three different lipid extraction methods.

Sample group Sample Methods Reference value(%)

Folch Soxhlet AOAC 996.06

Croissants Croissant-1 22.34a ± 0.52 22.53a ± 0.21 18.61b ± 0.20 31.1

Croissant-2 18.30a ± 0.20 17.41b ± 0.03 14.29c ± 0.11 12.0
Pastries Streusel topping pastry 20.00a ± 0.91 18.93a ± 0.06 15.40b ± 0.04 24.7
Danish pastry 25.45a ± 0.87 25.70a ± 0.51 20.67b ± 0.08 20.0
Pastry with chestnut 15.28b ± 0.55 16.24a ± 0.04 11.18c ± 0.09 16.9
Pastry doughnut 30.72a ± 0.44 29.68b ± 0.25 26.41c ± 0.49 30.4
Pies Pie with custard cream 32.75a ± 1.09 32.96a ± 0.32 27.57b ± 0.21 29.5
Apple pie 16.65a ± 0.22 16.81a ± 0.22 16.44a ± 0.17 10.0
Stick pie 32.22a ± 0.66 32.62a ± 0.20 30.94a ± 0.41 17.5
Others Muffin 18.12b ± 0.22 18.91a ± 0.11 18.76ab ± 0.51 18.0
Morning roll with cream 20.53a ± 0.75 20.82a ± 0.25 15.13b ± 0.17 20.1
Mocha streusel bread 8.83a ± 0.07 8.23b ± 0.03 6.05c ± 0.25 5.5

Mean values ± SD, n ¼ 3 replicates. Means with different superscripts differ significantly (p < 0.05).
Reference values provided by the product label.

Table 2
Monounsaturated fat content (%) in 12 different samples, as obtained by three different lipid extraction methods.

Sample group Sample Methods

Folch Soxhlet AOAC 996.06

Croissants Croissant-1 5.38b ± 0.07 5.68a ± 0.08 4.52c ± 0.05

Croissant-2 3.61a ± 0.04 3.49b ± 0.01 3.02c ± 0.02
Pastries Streusel topping pastry 4.92a ± 0.05 5.02a ± 0.09 3.90b ± 0.01
Danish pastry 6.63a ± 0.24 6.65a ± 0.12 5.48b ± 0.07
Pastry with chestnut 4.04a ± 0.07 4.11a ± 0.03 2.86b ± 0.03
Pastry doughnut 7.72a ± 0.20 7.31b ± 0.22 6.41c ± 0.11
Pies Pie with custard cream 7.82b ± 0.13 8.60a ± 0.15 6.53c ± 0.06
Apple pie 6.83a ± 0.14 7.05a ± 0.07 6.60a ± 0.06
Stick pie 9.09a ± 0.09 9.02a ± 0.02 8.04b ± 0.09
Others Muffin 4.57a ± 0.04 4.66a ± 0.11 4.31b ± 0.15
Morning roll with cream 5.61b ± 0.06 6.23a ± 0.11 4.21c ± 0.07
Mocha streusel bread 1.97a ± 0.01 1.96a ± 0.01 1.34b ± 0.06

Mean values ± SD, n ¼ 3 replicates. Means with different superscripts differ significantly (p < 0.05).

Table 3
Saturated fat content (%) in 12 different samples, as obtained by three different lipid extraction methods.

Sample group Sample Methods Reference value(%)

Folch Soxhlet AOAC 996.06

Croissants Croissant-1 12.94b ± 0.04 13.72a ± 0.32 11.43c ± 0.12 18.3

Croissant-2 10.48a ± 0.10 10.23b ± 0.06 9.28c ± 0.07 8.0
Pastries Streusel topping pastry 10.89b ± 0.13 11.31a ± 0.21 9.14c ± 0.02 14.7
Danish pastry 13.56b ± 0.34 14.15a ± 0.25 12.10c ± 0.07 10.2
Pastry with chestnut 8.58b ± 0.16 9.01a ± 0.03 6.58c ± 0.05 11.9
Pastry doughnut 7.98a ± 0.19 7.37b ± 0.03 7.07c ± 0.10 10.7
Pies Pie with custard cream 17.28b ± 0.26 18.80a ± 0.31 14.99c ± 0.08 13.6
Apple pie 7.50a ± 0.16 7.74a ± 0.09 7.57a ± 0.45 5.0
Stick pie 17.18a ± 1.04 17.28a ± 0.10 16.17b ± 0.23 7.8
Others Muffin 3.83a ± 0.09 3.83a ± 0.10 3.82a ± 0.15 7.0
Morning roll with cream 10.74b ± 0.09 11.75a ± 0.24 8.37c ± 0.06 11.4
Mocha streusel bread 4.67a ± 0.06 4.75a ± 0.04 3.40b ± 0.13 2.7

Mean values ± SD, n ¼ 3 replicates. Means with different superscripts differ significantly (p < 0.05).
Reference values provided by the product label.

amounts of trans fats than the amounts determined by the other resulting difference in fat content determined. The analytical fat
two methods, but it obtained a higher amount of trans fat than that determination method which should be applied to food depends on
of the other two methods for two products (stick pie and muffin). how fat is defined, as differences in fat content and the contents of
When the Folch and automated Soxhlet methods were compared, each fat classes (saturated, monounsaturated, polyunsaturated and
the Folch method resulted in a higher level of trans fat than that of trans fat) arise due to the application of various analytical methods.
the Soxhlet method for three samples, whereas for the remaining 9 In Korea, ‘nutrition facts’ on the label of imported products are
samples, the trans fat content determined by the Soxhlet method translated into Korean and printed on a sticker, which is attached to
was higher than or equal to that of the Folch method. the product. When the nutrition facts label says total fat is 3.5 g, it is
This study focused on different experimental methods written as ‘fat’ 3.5 g in Korean. In Korea, it is originally required to
depending on different fat definitions (crude or total fat) and the state crude fat content on the nutrition facts label. In the case of
976 J.-M. Shin, S.-K. Park / LWT - Food Science and Technology 63 (2015) 972e977

Table 4
Polyunsaturated fat content (%) in 12 different samples, as obtained by three different lipid extraction methods.

Sample group Sample Methods

Folch Soxhlet AOAC 996.06

a a
Croissants Croissant-1 1.11 ± 0.02 1.10 ± 0.01 1.10a ± 0.01
Croissant-2 0.62a ± 0.01 0.44c ± 0.01 0.58b ± 0.00
Pastries Streusel topping pastry 1.26a ± 0.07 1.17a ± 0.01 1.12a ± 0.00
Danish pastry 1.52a ± 0.09 1.34b ± 0.02 1.42ab ± 0.03
Pastry with chestnut 0.94a ± 0.02 0.85b ± 0.00 0.75c ± 0.00
Pastry doughnut 13.66a ± 0.35 13.12b ± 0.10 11.34c ± 0.25
Pies Pie with custard cream 5.09b ± 0.09 5.44a ± 0.11 4.29c ± 0.06
Apple pie 2.36a ± 0.04 2.37a ± 0.02 2.34a ± 0.02
Stick pie 5.40a ± 0.02 4.97b ± 0.13 4.96b ± 0.06
Others Muffin 10.17a ± 0.09 10.28a ± 0.34 9.58b ± 0.15
Morning roll with cream 1.60b ± 0.07 1.71a ± 0.03 1.32c ± 0.03
Mocha streusel bread 1.13a ± 0.02 1.06b ± 0.00 0.85c ± 0.04

Mean values ± SD, n ¼ 3 replicates. Means with different superscripts differ significantly (p < 0.05).

Table 5
Trans-fat content (%) in 12 different samples, as obtained by three different lipid extraction methods.

Sample group Sample Methods

Folch Soxhlet AOAC 996.06

b a
Croissants Croissant-1 0.70 ± 0.02 0.75 ± 0.01 0.61c ± 0.01
Croissant-2 0.80a ± 0.01 0.76b ± 0.00 0.66c ± 0.00
Pastries Streusel topping pastry 0.57b ± 0.00 0.61a ± 0.00 0.47c ± 0.00
Danish pastry 0.66b ± 0.01 0.70a ± 0.00 0.59c ± 0.00
Pastry with chestnut 0.58a ± 0.01 0.61a ± 0.02 0.43b ± 0.00
Pastry doughnut 0.43a ± 0.01 0.39b ± 0.00 0.39b ± 0.00
Pies Pie with custard cream 0.44b ± 0.01 0.49a ± 0.01 0.43b ± 0.00
Apple pie 0.03a ± 0.00 0.04a ± 0.00 0.05a ± 0.00
Stick pie 0.32a ± 0.00 0.30b ± 0.00 0.32a ± 0.00
Others Muffin 0.20b ± 0.00 0.21b ± 0.00 0.23a ± 0.02
Morning roll with cream 0.63b ± 0.01 0.70a ± 0.01 0.48c ± 0.00
Mocha streusel bread 0.22b ± 0.00 0.23a ± 0.00 0.16c ± 0.00

Mean values ± SD, n ¼ 3 replicates. Means with different superscripts differ significantly (p < 0.05).

imported food, its total fat content is essentially listed as crude fat identification of nutrients for the nutrition label, different defini-
on the label when it is sold in Korea. At present, when Korea ex- tions for the same fat can be used and may cause differences in the
ports processed foods to the United States, crude fat content is stated amount of fat for the same food. These differences can be the
obtained using the gravimetric method and the obtained value is result of varying interpretations of the role or relevance of the
listed as total fat content on the nutrition facts label, not the total fat nutrient within the diet, and thus, are a potential stumbling block
value obtained with the AOAC 996.06 method, since total fat and to international harmonization of such nutritional regulations. A
crude fat are internationally accepted as the same in their meaning uniform definition and presentation of fat on labels is required in
and content. In Korea, it is required to follow the Food Code of KFDA order for consumers to have a clear and unbiased understanding of
in analyzing fat and fatty acids. The Food code of KFDA requires that their dietary fat intake, as well as to facilitate international trade.
crude fat be extracted first and a little amount of the fat be taken to
be used for fatty acid analysis with GC (KFDA, 2009). It is because
Korea uses the crude fat definition and so crude fat content should 4. Conclusion
be obtained prior to the analysis of fatty acids. However, if fatty acid
analysis is performed according to this KFDA method, each fat The content of fat (total) extracted by the AOAC 996.06 method
content (saturated, polyunsaturated, and trans fat) depends on the (acid hydrolysis-GC method) was lower than that of fat (crude)
crude fat content. Therefore, the Food Code was amended in 2011 in extracted by the automated Soxhlet and Folch methods (gravi-
order to improve inhomogeneity probably occurring in the fat metric methods) for most samples. In addition, the contents of
taking procedure because the lipid composition of the crude fat is saturated fat, monounsaturated fat, and polyunsaturated fat
inhomogeneous and dependent on the food source and processing. determined by the AOAC 996.06 method were lower than those
The revised Food Code requires that AOAC 996.06 be used for fatty obtained by the Folch and automated Soxhlet methods. The auto-
acid analysis. According to the revised Food code, the total fat mated Soxhlet method produced a higher amount of saturated fat,
content obtained by this method is not used and each fat content monounsaturated fat, polyunsaturated fat and trans fat than that of
(saturated, polyunsaturated, and trans fat) is only used. Because the Folch method for cream-containing samples (pie with custard
Korea uses the crude fat definition, gravimetric method is used to cream, morning roll with cream). No difference was observed for fat
obtain fat content and AOAC 996.06 is used only for fatty acid content and contents of fat classes (saturated, monounsaturated,
analysis. The fatty acid analysis method of the revised law is better polyunsaturated and trans fat) obtained by each three methods in
than the previous one, but there is still a contradiction in Korean apple pie. There is no single standard method for the determination
nutrition labeling. In other words, the total amount of individual of fats in different foods. But it is apparent that different fat defi-
fatty acids is less than the fat (crude) content. It is because the fat nitions require different analysis methods, thus resulting in
definition and analysis method are inconsistent. In the different fat content, as well as different contents of fat classes.
 J.-M. Shin, S.-K. Park / LWT - Food Science and Technology 63 (2015) 972e977 977

References [KFDA] Korea Food and Drug Administration. (2009). Food code (pp. 104e105).
Seoul: Korea Foods Industry Assoc. Co., Ltd.
[KFDA] Korea Food and Drug Administration. (2011). Department of food standard-
Ali, L. H., Angyal, G., Weaver, C. M., & Rader, J. I. (1997). Comparison of capillary
ization. Labeling of trans fats. Seoul, Korea.
column gas chromatographic and AOAC gravimetric procedures for total fat and
Luque de Castro, M. D., & Priego-Capote, F. (2010). Soxhlet extraction: past and
distribution of fatty acids in foods. Food Chemistry, 58, 149e160.
present panacea. Journal of Chromatography, 1217, 2383e2389.
Ascherio, A., Katan, M. B., Zock, P. L., Stampfer, M. J., & Willett, W. C. (1999). Trans
Neaton, J. D., Wentworth, D., &, Multiple Risk Factor Intervention Trial Research
fatty acids and coronary heart disease. England Journal of Medicine, 340,
Group. (1992). Serum cholesterol, blood pressure, cigarette smoking, and death
1994e1998.
from coronary heart disease. Overall findings and differences by age for 316,099
Assn of Official Analytical Chemists Intl (AOAC). (2005a). In W. Horwitz (Ed.), Official
white men. Archives of Internal Medicine, 152, 56.
method 996.01. Official methods of analysis of AOAC international (18th ed.).
Ngeh-Ngwainbi, J., Lin, J., & Chandler, A. (1997). Determination of total, saturated,
Gaithersberg: Assn. of Official Analytical Chemists.
unsaturated, and monounsaturated fats in cereal products by acid-hydrolysis
Assn of Official Analytical Chemists Intl (AOAC). (2005b). In W. Horwitz (Ed.), Official
and capillary gas-chromatography tcollaborative study. Journal of AOAC Inter-
method 996.06. Official methods of analysis of AOAC international (18th ed.).
national, 80, 359e372.
Gaithersberg: Assn. of Official Analytical Chemists.
Nielsen, S. S. (2010). Food analysis. 2010 (4th ed.). In Food science text series. London:
Bligh, E. G., & Dyer, W. J. (1959). A rapid method of total lipid extraction and pu-
Springer.
rification. Canadian Journal of Biochemistry and Physiology, 37, 911e917.

rez-Palacios, T., Ruiz, J., Martín, D., Muriel, E., & Antequera, T. (2008). Comparison
Pe
Brazil. (2003). Ministe
rio da Saúde, Secretaria de Vigil^ ancia Sanit

aria; Resoluç~ao-
~e sobre o Regulamento Te
cnico of different methods for total lipid quantification in meat and meat products.
RDC n 360, de 23 de dezembro de 2003, Dispo

rio Oficial da Unia
~o de Food Chemistry, 110, 1025e1029.
sobre Rotulagem Nutricional de Alimentos Embalados, Dia
~o, 1, 33e34. Phamaceutical Society of Japan. (2010). Methods of analysis in health science. Tokyo:
26 de dezembro de 2003. Seça
Kane-hara Shuppan.
[CFIA] Canadian Food Inspection Agency. (2011). Food, labelling, nutritional labeling.
Rader, J. I., Angyal, G., O'Dell, R. G., Weaver, C. M., Sheppard, A. J., & Bueno, M. P.
Available from http://www.inspection.gc.ca/english/fssa/labeti/nutricon/
(1995). Determination of total fat and saturated fat in foods by packed column
nutriconapp4e.shtml Accessed 11.03.12.
gas-liquid chromatography after acid hydrolysis. Food Chemistry, 54, 419e427.
[CFS] Centre for Food Safety. (2011). Food legislation, guidelines. Available from
Ratnayake, W. M. N. (2004). Overview of methods for the determination of trans
http://www.cfs.gov.hk/english/food_leg/files/nl_method_guidance_e.pdf
fatty acids by gas chromatography, silver-ion thin-layer chromatography, silver-
Accessed 09.03.12.
ion liquid chromatography, and gas chromatography/mass spectrometry. Jour-
Craig-Schmidt, M. C. (2006). World-wide consumption of trans fatty acids.
nal of AOAC International, 87, 523e539.
Atherosclerosis Supplements, 7, 1e4.
Ruiz-Mime
nez, J., Priego-Capote, F., & Luque de Castro, M. D. (2004). Identification
EEC. (1990). Council directive on nutrition labelling for foodstuffs (90/496). Official
and quantification of trans fatty acids in bakery products by gas
Journal of the European Communities, L276, 40e44.
chromatography-mass spectrometry after dynamic ultrasound-assisted
[FDA] Food and Drug Administration, Department of Health and Human Services.
extraction. Journal of Chromatography A, 1045, 203e210.
2003. Food labeling: trans fatty acids in nutrition labeling, nutrient content
Shin, J. M., Hwang, Y. O., Tu, O. J., Jo, H. B., Kim, J. H., Chae, Y. Z., et al. (2013).
claims, and health claims. 68:41433e41506. CFR Part 101.
Comparison of different methods to quantify fat classes in bakery products.
Federal Register. 1993. 58:631.
Food Chemistry, 136, 703e709.
Folch, J., Lees, M., & Sloane Stanley, G. H. (1957). A simple method for the isolation
Vijver, L. P. L., Kardinaal, A. F. M., Couet, C., Aro, A., Kafatos, A., & Steingrimsdottir, L.
and purification of total lipids from animal tissues. The Journal of Biological
(2000). Association between trans fatty aicd intake and cardiovascular risk
Chemistry, 226, 497e509.
factors in Europe: the TRANSFAIR study. European Journal of Clinical Nutrition,
[FSANZ] Food Standards Australia New Zeland. (2013). Food standards code, user
54, 126e135.
guides to the food standards code. Available from http://www.foodstandards.gov.
Zou, W., Lusk, C., Messer, D., & Lane, R. (1999). Fat contents of cereal foods: com-
au/code/userguide/Documents/Userguide_Prescribed%20Nutrition%
parison of classical with recently developed extraction techniques. Journal of
20Information_PartA_March12.pdf Accessed 24.05.13.
AOAC International, 82, 141e149.
Hawkes, C. (2004). Nutrition labels and health claims: The global regulatory envi-
ronment. Geneva: World Health Organization. Available from http://libdoc.who.
int/publications/2004/9241591714.pdf Accessed 12.05.12.