EVALUATION OF GAS CHROMATOGRAPHIC AND FOURIER TRANSFORM
INFRARED SPECTROSCOPIC METHODS FOR THE DETERMINATION OF TRANS-
FAT LEVELS IN THE DIETS OF LAYING HENS AND THEIR EGGS
by
ROBERT LOUIS FUSCO
(Under the Direction of Ronald B. Pegg)
Abstract
Five groups of 10 single comb white leghorn hens were fed different diets (standard corn/soy
laying rations supplemented with edible oils) to ascertain the effect of trans-fat deposition in
eggs. Trans-fat contents of the feeds and eggs were determined by gas chromatography (GC) and
Fourier transform infrared (FTIR) spectroscopy. Diets devoid of hydrogenated oils contained no
detectable levels of trans-fats. Feeds supplemented with tallow, shortening, or their combination
contained 2.78±0.08, 3.25±0.03, and 2.85±0.05 g/100-g dietary fat, and 3.33±0.32, 4.16±0.03,
and 3.52±0.05 g/100-g dietary fat as assessed by GC and FTIR, respectively. Eggs from hens fed
these latter diets contained 0.68±0.05, 0.81±0.05, and 0.76±0.04 g/100 g dietary fat, by GC, but
were below the 1% detection limit of FTIR. Nearly 25% of trans-fats in the diets accumulated in
the eggs. Supplementation of edible oils did not appreciably increase trans-fat levels in the eggs.
Data acquired by these methods correlated well, r=0.98.
INDEX WORDS: Trans-fat, GC-FID, FTIR-ATR, Silver-Ion TLC, SCWL Hen
EVALUATION OF GAS CHROMATOGRAPHIC AND FOURIER TRANSFORM
INFRARED SPECTROSCOPIC METHODS FOR THE DETERMINATION OF TRANS-
FAT LEVELS IN THE DIETS OF LAYING HENS AND THEIR EGGS
by
ROBERT LOUIS FUSCO
B.S., The University of Connecticut, 2007
A Thesis Submitted to the Graduate Faculty of the University of Georgia in Partial
Fulfillment of the Requirements for the Degree
MASTER OF SCIENCE
ATHENS, GEORGIA
2009
©2009
Robert Louis Fusco
All Rights Reserved
EVALUATION OF GAS CHROMATOGRAPHIC AND FOURIER TRANSFORM
INFRARED SPECTROSCOPIC METHODS FOR THE DETERMINATION OF TRANS-
FAT LEVELS IN THE DIETS OF LAYING HENS AND THEIR EGGS
by
ROBERT LOUIS FUSCO
Major Professor: Ronald B. Pegg
Committee: Rakesh K. Singh Michael P. Lacy
Electronic Version Approved: Maureen Grasso Dean of the Graduate School The University of Georgia May 2009
I would like to thank Dr. Ronald Pegg, Dr. Rakesh Singh, Dr. Michael Lacy, and Dr.
Ramesh Avula for their guidance throughout my time at the University of Georgia. I would also
like to thank the American Egg Board for funding this project.
iv
ACKNOWLEDGEMENTS
TABLE OF CONTENTS
Page
LIST OF TABLES....................................................................................................................... viii
LIST OF FIGURES ....................................................................................................................... ix
CHAPTER
1 LITERATURE REVIEW ...........................................................................................1
Origins of Trans-Fats ..............................................................................................1
Health Effects Of Trans Fatty Acids......................................................................6
Recent Government Regulations ...........................................................................8
Trans-Fat Alternatives ............................................................................................9
Fatty acid supplementation of laying hen feeds..................................................13
Trans Fatty Acid Analysis: Gas Chromatography (GC) ...................................14
Trans Fatty Acid Analysis: FTIR.........................................................................16
2 MATERIALS, METHODS, RESULTS, AND DISCUSSION ..............................18
Chemicals ...............................................................................................................18
Feed Trials..............................................................................................................18
Acid Hydrolysis......................................................................................................19
Lipid Extraction ....................................................................................................19
Fatty Acid Methylation .........................................................................................21
Gas Chromatographic Analysis ...........................................................................21
Silver-Ion Thin Layer Chromatography (Ag+-TLC) .........................................23
Linearity: Gas Chromatography ........................................................................23
v
Precision and Accuracy: Gas Chromatography ................................................24
Limit of Detection (LOD) and Limit of Quantification (LOQ): Gas
Chromatography .............................................................................................24
Recovery: Gas Chromatography ........................................................................24
Fourier Transform Infrared Spectroscopy (FTIR) Analysis ............................25
Linearity: Fourier Transform Infrared Spectroscopy (FTIR) Analysis .........26
Precision, Accuracy, and Limit of Detection (LOD): Fourier Transform
Infrared Spectroscopy.....................................................................................26
Statistical Methods ................................................................................................27
Results and Discussion ..........................................................................................27
Gas Chromatography............................................................................................27
Fourier Transform Infrared Spectroscopy .........................................................32
Conclusions ............................................................................................................34
References .....................................................................................................................................36
vi
LIST OF TABLES
Page
Table 1: Egg weight (g) of hens fed diets of varying levels of trans-fat for 21 days ...................42
Table 2: Response factors of FAMEs as determined by GC-FID using Supelco 37 component
FAME mixture ..............................................................................................................43
Table 3: Fatty acid profile of egg yolk lipids from hens consuming diets with varying levels of
trans-fat .........................................................................................................................45
Table 4: Percentage of fatty acids in egg yolk lipids from hens consuming diets with varying
levels of trans-fat...........................................................................................................47
Table 5: Fatty acid composition of feeds with and without added trans-fats ...............................49
Table 6: Recovery of triheptadecanoin and methyl heptadecanoate during GC-FID analysis .....50
Table 7: Precision and accuracy of trans-fat analysis using GC-FID...........................................51
Table 8: Precision and accuracy of trans-fat analysis using FTIR-ATR ......................................52
vii
LIST OF FIGURES
Page
Figure 1: Partial hydrogenation reaction mechanism ...................................................................53
Figure 2: GC-FID Chromatogram of egg yolk lipids from diet I .................................................55
Figure 3: GC-FID Chromatogram of egg yolk lipids from diet II ................................................57
Figure 4: GC-FID Chromatogram of egg yolk lipids from diet III...............................................59
Figure 5: GC-FID Chromatogram of egg yolk lipids from diet IV ..............................................61
Figure 6: GC-FID Chromatogram of egg yolk lipids from diet V................................................63
Figure 7: GC-FID Chromatogram of diet I ...................................................................................65
Figure 8: GC-FID Chromatogram of diet II..................................................................................67
Figure 9: GC-FID Chromatogram of diet III ................................................................................69
Figure 10: GC-FID Chromatogram of diet IV.............................................................................71
Figure 11: GC-FID Chromatogram of diet V ...............................................................................73
Figure 12: Overlay of silver ion-TLC chromatogram and original chromatogram for
shortening ......................................................................................................................75
Figure 13: Chromatogram of silver ion-TLC separation of trans monounsaturated fraction of
shortening ......................................................................................................................77
Figure 14: Interference band of egg lipids during FTIR analysis .................................................79
Figure 15: Second derivative spectra of absorbance of trans band at 967 cm-1 ...........................81
Figure 16: Spectra of absorbance of trans band at 967 cm-1 ........................................................83
viii
Figure 17: TQ Analyst calibrations of prepared standards of 0.5 to 10% (w/w) methyl elaidate in
methyl oleate .................................................................................................................85
Figure 18: Correlation of GC versus FTIR data ...........................................................................87
ix
Chapter 1
LITERATURE REVIEW
Potential health effects of dietary trans fatty acids and the US Food And Drug Administration
(FDA) mandate that as of January 1, 2006, the Nutrition Facts panels on all packaged food labels
must indicate the quantity of trans fatty acids in a serving of the food product, have driven the
development of efficient methods for accurate measurements of isomeric fatty acids in foods.
Since this mandate food manufactures have reformulated many of their products and the general
public has had an increased awareness of the effects of dietary trans fatty acids.
Origins of Trans-Fats
Mono- and poly- unsaturated fatty acid molecules exist in either a cis or trans
configuration. Cis and trans isomers are geometric isomers of on another, i.e., they differ only in
the spatial configuration of the atoms in the compound. A cis-double bond contains hydrogen
atoms on the same side of the double bond, while a trans double bond contains hydrogen atoms
on opposite sides of the double bond. In nature, double bounds occur almost exclusively in the
cis configuration; it has been reasoned that this is because fatty acids are important structural
components of cell membranes, and the cis double bond imparts a bend in the molecule,
reducing the van der Waals forces between fatty acids, and as a result maintains the fluidity of
the membrane (Kodali and List, 2005).
1
The principal source of natural trans fatty acids comes from biohydrogenation by rumen
microorganisms. In the gut of ruminant animals, bacteria convert cis-9,cis-12-octadecadienoic
acid into cis-9,trans-11-octadecadienoic acid, i.e., one of the conjugated linoleic acid (CLA)
isomers. A majority of the cis-9 double bonds in CLA are subsequently hydrogenated resulting
in trans-11-octadecenoic acid. These “natural” trans-fats derived from biohydrogenation
constitute 3-8% of the fat in beef and milk (Kodali and List, 2005).
The trans fatty acids of meat and milk of ruminant animals arising from
biohydrogenation is different from trans fatty acids arising from partial hydrogenation practiced
by the edible oil industry. The hydrogenation process generates many different cis and trans
isomers while biohydrogenation produces mainly trans-11-octadecenoic acid (Ratnayake,
2004). Trans-11-octadecenoic acid has been shown to form conjugated cis-9,trans-11-
octadecadienoic acid during metabolism in humans (Jiang and others 1996). CLA
supplementation in the diet did not change LDL or HDL cholesterol levels during clinical trials,
nor has an association been shown between the consumption of ruminant animal trans fatty acid
(in meat and milk) and cardiovascular disease (CVD) or coronary heart disease (CHD) (Huth,
2007; Jakobsen and others 2008). Therefore, the following discussion of the health related
effects of trans-fats will consider only those resulting from the partial hydrogenation of edible
oils.
The main source of dietary trans fatty acids comes from the partial hydrogenation of
vegetable oils. Hydrogenated vegetable oils can contain 10% to 40% trans fatty acids depending
on the reaction conditions and the degree of hydrogenation to which the starting material was
subjected (Kodali and List, 2005). Wilhelm Normann patented the process of hydrogenating
2
liquid oils in 1902. In 1909 Proctor and Gamble purchased the patent from Wilhelm, for the
production of shortening from cottonseed oil (Crisco) as a replacement for lard.
During the hydrogenation of oils, hydrogen atoms are added across unsaturated carbon-
carbon double bonds, creating saturated carbon-carbon single bonds. This is done at high
temperatures and pressures in the presence of a metal catalyst. In industry, reduced nickel is the
most widely used catalyst because of its good activity, selection, filterability, reusability, and
economical use. For efficient conversion of double bonds to single bonds, the concentration of
hydrogen on the catalyst surface and the contact of triacylglycerols with that surface must be
maximized. Hydrogen is dispersed as small bubbles into the bulk oil; smaller bubbles have a
greater surface to mass ratio than larger bubbles. The catalyst may be fixed onto a support
thereby further increasing the surface area of the catalyst to mass of the catalyst. To promote the
diffusion of triacylglycerols and from the surface of the catalyst gentle agitation, at temperatures
between 250-300oC is employed (Akoh and Min, 2002).
Hydrogen atoms will attach to either of the carbons of the double bond. If there is
another hydrogen atom available, it will attack the other now negatively charged carbon to form
a carbon-carbon single bond. If, however, another hydrogen atom is not available, then the
double bond will reform as either a cis or trans double bond. This can occur on either side of the
carbon anion (Akoh and Min, 2002).
The isomerization of fatty acids will decrease when there is a higher concentration of
hydrogen present on the catalyst surface, so that the saturated single bond is formed before the
double bond reforms. This can be achieved by using low temperatures, high pressures, high
agitation rates, high gassing rates, and low catalyst concentrations. These conditions will
completely saturate the catalyst with hydrogen thereby allowing saturated bonds to form faster.
3
A study by Evans et al. showed that the formation of trans isomers can be reduced to 0.4% trans
bond formation per IV unit reduction at 120oC and 689 kPa, compared to 0.7% trans formation
per IV unit reduction, at 170oC and 34 kPa (Kodali and List, 2005). The mechanism for
hydrogenation is depicted in Figure 1.
Despite the main reaction forming saturated fatty acids, there are also two side reactions
that occur: the isomerization of cis double bonds to trans double bonds and the migration of the
double bond along the carbon chain. The shift of the double bond creates fatty acid isomers that
are not commonly found otherwise. The dominant naturally-occurring isomer of octadecenoic
acid is C18:1 n-9 (oleic acid) where the position of the double bond is located at the 9th carbon
from the terminal methyl end. However during hydrogenation, a distribution of isomers is
generated ranging from the 4th carbon to the 16th carbon with the majority occurring at the 8th
thru 11th positions (Kramer and others 2002). This migration happens for both the cis and trans
isomers.
There are two main reasons for the partial hydrogenation of edible oils: to increase the
oxidative stability as well as the solid fat content (SFC) of the oil (Akoh and Min, 2002).
Autooxidation occurs when oxygen radicals interact with the double bond in a fatty acid
producing hydroperoxides which subsequently breakdown to odiferous secondary oxidation
products leading to rancidity. Trans fatty acids are more stable than cis fatty acids, but much
less stable than saturated fats. As such, they undergo oxidation at a slower rate than their cis-
monounsaturated counterparts, but at a faster rate than saturated fats. For example Oleic acid
undergoes oxidation 10 times faster than stearic acid and 10 times slower than linoleic acid
(Kodali and List, 2005). As a result, trans-monounsaturated fats and the products fabricated
4
from them have a longer shelf life than those made with cis-monounsaturated and
polyunsaturated fatty acids (Gunstone, 2007).
Triacylglycerols can exist in various crystal structures because fatty acid chains pack into
different crystal lattices. Different polymorphic forms like α, β′, and β, impart different physical
attributes to solid fats. The α form is the least thermodynamically stable and has the lowest
melting point, followed by β′ and then β with the latter having the highest melting point. Fats
will initially crystallize into the α form, but it is unstable and will transform to either β′ or β
forms. The β′ form is the most functional for margarines and shortenings, because it forms very
small crystals and instills desirable characteristics into products such as a smooth mouth feel and
sharp melting point. It has been report that the β form contains larger crystals and can give fats a
grainy mouth feel. When comparing the crystaline structures of tristearin, trielaidin, and triolein,
Kodali showed that while tristearin and triolean will pack into a β′ form, trielaidin will only pack
into the β form (Kodali and List, 2005). The melting points of tristearin, trielaidin, and triolein
are 73oC, 42oC, and 5oC respectively.
Solid fat content adds important characteristics to baked goods and confections. A high
solid fat content adds snap and improved textural properties to these goods. For example cocoa
butter illustrates the importance of solid fat content and melting characteristics of fats. Cocoa
butter comprises primarily stearic, palmitic, and oleic acid. This unique set of fatty acids give
cocoa butter a sharp melting point between 20oC and 35oC, allowing the fat to melt at body
temperature (Vaclavik, 2007). When the fat melts in the mouth, it provides a cooling sensation
as it draws energy, in the form of heat, from the mouth. Most fats have more than three primary
fatty acids in them, causing them to have a broader melting point. Fatty acids that have a melting
point higher then body temperature generally result in a waxy mouth feel.
5
Health Effects of Trans Fatty Acids
Heart disease is the leading cause of death for both men and women in the United States
(Kung and others 2008). Several studies have indicated that the consumption of trans-fats, even
at low levels (i.e., 1-3% of total energy intake), can substantially increase the risk for coronary
heart disease (CHD) (Mozaffarian and others 2006). The Nurses Health Study, an
epidemiological study conducted by Hu and colleagues, followed 120,000 female nurses for 14
years and concluded that when a 2 % increment in energy from trans unsaturated fat replaced
carbohydrates in the diet the risk of developing CHD was 1.93 times more likely compared to
only a 1.17 times when a 5 % increment in energy from saturated fat replaced carbohydrates in
the diet. (Hu and others 1997). Hu and coworkers estimated that the replacement of 2 % of
energy from trans-fat with energy from cis-unsaturated fats would reduce the risk of CHD by a
whopping 53% (Hu and others 1997).
Two of the main biomarkers for cardiovascular disease (CVD) are the levels of low
density lipoprotein (LDL) and high density lipoprotein (HDL) cholesterol. Because of its size
LDL cholesterol, “bad cholesterol”, is able to enter the capillaries of endothelial cells where it
can be oxidized. Oxidized LDL cholesterol leads to plaque formation on artery walls, one of the
symptoms of CVD. It has been shown that saturated fatty acids increase total and LDL
cholesterol levels but do not change HDL cholesterol levels, while trans fatty acids not only
increase total and LDL cholesterol but also decrease HDL cholesterol (Katan and others 1995).
The lowering of HDL cholesterol causes trans-fats to have a less favorable effect on the plasma
lipoprotein profile than saturated fats (Kyoko Hayakawa and others 2000). The increased risk of
CHD from increased trans-fat consumption is greater than what would be predicted by
cholesterol levels alone. Several studies point to a correlation between trans-fat consumption
6
and an increase in systemic inflammation. In observational studies, an increased trans fatty acid
intake led to an increased activity of the tumor necrosis factor (TNF) system as well as increased
levels of C-reactive protein (CRP) and interleukin-6 (IL-6) (Lopez-Garcia and others 2005;
Micha and Mozaffarian, 2008; Mozaffarian, 2006; Mozaffarian and others 2004). All of which
are biomarkers of increased systemic inflammation and endothelial dysfunction. During an
experimental study of 50 healthy men fed a diet consisting of 8% of energy from trans fatty
acids, plasma levels of CRP and IL-6 increased when compared to a diet consisting of 8% of
energy of oleic acid (Baer and others 2004). Mozzofarin described a possible mechanism for
these increases: trans fatty acids consumed in the diet incorporate themselves into the cell
membrane of endothelial cells and macrophages, as is seen with n-3 and n-6 monounsaturated
fats. Endothelial cells and macrophages have various cell-specific pathways that activate the
TNF system (Mozaffarian and others 2004).
While most research regarding the adverse health effects of trans fatty acids has dealt
with CVD, trans fatty acids have been implicated as increasing the risk of several other diseases
including diabetes and Alzheimer’s disease. There has been less scientific consensus concerning
the role of trans-fats in these diseases than its apparent effects on CVD. A study by Salmeron
and coworkers showed that when energy from trans fatty acids was increased by 2%, the relative
risk for diabetes was increased by 1.39 (Salmeron and others 2001). In an epidemiological study
Hu and coworkers reported that trans fatty acids adversely affected glucose metabolism and
insulin resistance (Hu and others 2001). Investigations by van Dam et al. and Meyer et al.
however, showed no association between trans fatty acids and diabetes (van Dam and others
2002; Meyer and others 2001).
7
A study of 815 people over the age of 65 implicated trans fatty acids as having a relative
risk of 2.4 for the onset of Alzheimer’s disease (Morris and others 2003). A recent study of mice
fed diets high in trans-fats showed changes in the fatty acid profile of the brain shifting from
docosahexaenoic acid (DHA) to docosapentaenoic acid (DPA), but did not show any of the
major brain neuropathological hallmarks of Alzheimer’s disease (Phivilay and others 2009).
Recent Government Regulations
Based upon growing scientific evidence relating trans-fats to an increased incidence of
CHD and a citizen petition’s in 1994, the FDA, proposed in 1999 that the quantity of trans-fats
in products be included on food labels, specifically, in the section dealing with the amount and
percent daily value for saturated fats, a footnote was included indicating the amount of trans
fatty acids per serving of the product (Services, 2003). Following the FDA’s proposal, the
Institute of Medicine issued a report indicating a direct association between trans fatty acids and
an increased incidence of CHD. They failed, however, to establish a daily reference value,
simply stating that trans-fat consumption should be as low as possible (Food and nutrition board,
2002). In 2003, the FDA issued its final ruling requiring the posting of trans-fat on the
nutritional labels of all products. This law went into effect on January 1st 2006. The trans-fat
column would be listed as a sub grouping of total fat and would not contain a %DV. Any
product that contained less then 0.5g of trans-fat per serving could be listed as 0 g trans-fat, or
alternatively would not be required to list trans-fat on the nutritional label (Services, 2003).
Many believe that 0.5 g is too high of a value to be listed as 0 g because a person consuming
multiple servings of products containing 0.5g trans-fats could easily consume over the maximum
recommended intake of 1% of energy or 2 g of trans fatty acids, as recommended by the
8
American Heart Association, without knowing it (American Heart Association, 2008). The
FDA, unlike several other countries, did not decide to limit the amount of trans-fats that a
product could contain. In 2003, Denmark banned any product that contained over 2% trans-fat.
This regulation included commercial oils practically eliminating partially hydrogenated oils from
the food supply. Even fast food has low a level of trans-fats in Denmark. For example the
cooking oil used to fry McDonalds French fries has, on average, 24% trans-fats in the United
States while it has less than 1% in Denmark (Stender and others 2006). In 2008, Sweden
adopted the same laws as Denmark concerning trans-fat levels.
Several US cities have enacted laws limiting the quantity of trans-fats that can be used by
restaurants. As and example, in 2006, New York City banned the use of trans-fats in oils
employed for deep frying (Tucker and Markt, 2006). These regulations came with a public
education campaign to inform the consumers of the health effects attributed to trans-fats and
how to choose foods that are low in them. Other cities have enacted similar trans-fat regulations
including Philadelphia in 2007 (City of Philadelphia, 2007a) and San Francisco in 2008,
although San Francisco’s regulation is voluntary. As of 2010, the state of California will be the
first state to ban the use of trans-fats for deep frying, except for donuts, food businesses making
donuts will have until 2011 before the restriction of using partially hydrogenated oils will apply
(State of California, 2008).
Trans-Fat Alternatives
With these regulations two things have become apparent: the need for fats and oils free of
trans-fat, and an effective and efficient way to determine the amount of trans-fat contained in a
product. One concern while reformulating products is that while replacing trans-fats, the level of
9
unsaturated fats should be maximized while providing the same sensory attributes and
maintaining a similar cost. There are several alternatives at hand to replace partially
hydrogenated oils. These include naturally stable oils and fats, interesterified oils, fractionated
oils, and trait-enhanced oils from new oilseed varieties. Palm, corn, and cottonseed oils are
relatively stable and do not require partial hydrogenation for most commercial applications.
Interesterification also produces relatively stable oils. Depending on what oil stocks are
used and the degree of interesterification, these oils can be utilized in many commercial
applications. This technology is highly developed in Europe for the production of margarines
and shortenings (Kodali and List, 2005). Industrially, interesterification can be accomplished by
batch or continuous methods. During a batch process, the raw lipid is sparged with nitrogen and
heated to 120-150oC under vacuum to remove any residual moisture. Trace amounts of water
will deactivate the catalyst. The mixture is then cooled to 70-100oC followed by the addition of
the catalyst, usually in the form of methylated or ethylated sodium or potassium at a
concentration of <0.4%. The reaction proceeds for 30-60 min, is then neutralized with CO2, and
then terminated by the addition of water. The water and catalyst are centrifuged out and the
catalyst is recovered.
During interesterification, fatty acids are exchanged within triacylglycerols and between
triacylglycerols resulting in a completely random distribution of fatty acids within the
triacylglycerol molecules and a random distribution of triacylglycerols within the lipid. If three
completely pure stocks in equal proportions of tristearin, triolein, and trilinolein, are
interesterified, the resulting composition of triacylglycerols will be as follows:
SSS, 3.7% SOO, 11.1%
OOO, 3.7% SLL, 11.1%
10
LLL, 3.7% OOL, 11.1%
SSO, 11.1% OLL, 11.1%
SSL, 11.1% SOL, 22.2%
where S = stearic acid, L = linoleic acid, and O = oleic acid. These percentages can change if the
reaction temperature is below the melting point of the highest melting point triacylglycerol,
usually a trisaturate. This kind of interesterification, called directed interesterification, will result
in a product enriched in the highest melting point molecules. Directed interesterification is
useful to increase the solid fat content without affecting the content of unsaturated fats. During
the reaction when trisaturates form, they will crystallize out of the liquid phase. They can then
be removed and the reaction will proceed until all of the trisaturates have propagated out of the
mixture. The final product from equal proportions of tristearin, triolein, and trilinolein, will have
the following composition:
SSS, 33.3%
OOO, 8.3%
OOL, 24.99%
OLL, 24.99%
LLL, 8.3%
To increase the rate of settling, gentle agitation is employed. Interesterification can change the
physical properties such as solid fat content and melting point without affecting the content of
fatty acids. There is also no cis/trans isomerization, thereby making it a useful technique to
replace partial hydrogenation. It has also been used to produce zero-trans fat margarines and
shortenings.
11
Fractionation is the process in which lower melting triacylglycerols, oleins, are separated
from higher melting triacylglycerols, stearines. It is commonly used on palm oil, palm kernel
oil, coconut oil, and animal fats. Fractionation of palm oil is accomplished at 20oC resulting in
70% palm olein, and 30% palm stearine. Palm stearine is used in the formulations of margarines
and shortenings. The palm olein may undergo a second fractionation at a lower temperature to
produce super palm olein used in frying applications, and palm mid-fraction, used in cocoa butter
substitutes. To increase the rate of settling, gentle agitation is employed. If too vigorous
agitation is used, it will cause the crystals to break, hindering the settling rate. To decrease
viscosity, fractionation is sometimes preformed in a solvent such as hexane.
The use of certain varieties of oil, such as high-oleic corn oil and low-linolenic canola oil
can replace partially hydrogenated oils in deep frying applications. These specialty oils,
however, tend to be higher in price than the partially hydrogenated oils previously, used because
many of them are not produced in commercially-viable quantities for large scale operations.
Increasing production of these specialty crops would cut down on cost and increase supply;
however, to develop oil at an industrially useful scale can take up to 6 or 7 years for a crop to be
planted until it is available for use by the consumer.
Oil seeds have been produced with modified fatty acid compositions by traditional seed
breeding procedures and by modern genetic methods to manufacture oils that meet specific
consumer needs. Monsanto is a leader in the latter approach. Mutants from oil plants have been
identified that produce oil with varying fatty acid compositions. Canola oil, which is widely
consumed as cooking oil, is in fact the result of a low eurcic rapeseed oil cultivar (Tarrago-Trani
and others 2006). High-oleic sunflower oils were produced in Russia in 1976 by a selective
sunflower breeding program (Fick, 1983). These oils are more stable in deep-frying applications
12
and have an extended shelf life compared to traditional sunflower oil. The National Sunflower
Association released a mid-oleic oil NuSun in 1998 (NuSun, 2006).
Genetic engineering offers a more direct way to alter the fatty acid compositions of oil
seed crops (Flickinger and Huth, 2004; Hammond and Fehr, 1984). Genetic engineering of
soybeans has produced soybean oil with a lower content of α-linolenic acid (C18:3 n-3), which is
prone to rancidity, and higher amounts of oleic acid thereby making it more suitable for deep-
frying (Hammond and Fehr, 1984). Canola has been engineered to contain low amounts of α-
linolenic acid and higher amounts of oleic acid; soybean oil has been made with increased levels
of saturated fatty acids, such as stearic or palmitic acid for use in margarines and shortenings.
Soybean oil has also been produced with decreased amounts of saturated fatty acids to improve
its nutritional value (Hammond and Fehr, 1984).
Fatty acid supplementation of laying hen feeds
Laying hen diets are often supplemented with fat or oil to increase the weight of the eggs
produced (Sell and others 1987). This is especially important when sorghum or barley is
substituted for corn, which contains higher amounts of fat (Coon and others 1988). Adding fat to
laying hens’ diets can substantially increase the cost of the feed, so economically feed-grade fat
and oils come from many different places, including spent grease from restaurants, rendered
animal fats, and refuse from oil refining plants. All of these sources have the possibility of
containing high levels of trans fatty acids.
Lipids deposited into the egg yolk are formed in the liver with fatty acids acquired
through the diet or synthesized de novo. Supplying hens with dietary fat from the diet reduces
hepatic fatty acid synthesis, increasing yolk formation and egg weight (March and MacMillan,
13
1990). The increased use of dietary lipids during yolk formation leads to a fatty acid composition
similar to that of the diet. Fat absorbed from the intestine is transported to the liver where some
desaturation and elongation of saturated fatty acids takes place. Linoleic acid is the only fatty
acid essential to hens, with a minimum requirement of 1% of dietary feed. Higher levels of
linoleic acid can produce heavier eggs (Subcommittee on Poultry Nutrition, 1994).
Recent research has focused on changing the n-3:n-6 ratio to better reflect the needs of
the human diet. Laying hens that were feed diets rich in omega-3 fatty acids, such as flaxseed,
showed a marked increase of omega-3 fatty acids in the yolk fat. (Leeson and Caston, 1990)
showed that a diet containing 30% flaxseed increased the linolenic acid composition from 0.38%
to 11.50% of total fat. Small increases in long chain n-3 fatty acids also occurred. Studies have
also shown that supplementation with fish oil will increase ALA and long chain n-3 fatty acids
but impart negative sensory attributes to the eggs.
Trans Fatty Acid Analysis: Gas Chromatography (GC)
The other concern that was triggered by the enactment of the FDA’s labeling mandate is
how to accurately measure the trans-fat content in a product. There are several methods that can
be used including GC-FID, Fourier transform infrared (FTIR) spectroscopy, Raman
spectroscopy, NMR spectroscopy, and reversed phase HPLC (Akoh and Min, 2002). The
American Oil Chemists’ Society (AOCS) has several official methods for the determination of
trans-fat, and these have been updated since the initial meeting by the FDA in 1999. Specifically,
AOCS method Ce 1f-96 (AOCS Official method Ce 1f-96, 2002) for the determination of trans
fatty acids for hydrogenated and refined oils by capillary GLC is used to identify and quantify
individual fatty acid methyl esters (FAMEs). Prior to GC analysis, fatty acid moieties of
14
triacylglycerols must be released from the glycerol backbone and the resultant the non-volatile
fatty acids must then be derivatized to FAMEs. As and illustration, methyl oleate has a boiling
point of only 219oC compared to 360oC for oleic acid, allowing FAMEs to volatilize at
temperatures reasonable for GC (MSDS, 2007b; MSDS, 2005). The approved method for
derivitizing fatty acids into FAMEs is AOCS method Ce 2-66 (AOCS Official Method Ce 2-66,
1997). This method calls for the use of boron triflouride (BF3). BF3, as a derivatizing agent, has
some major disadvantages according to Cristie; the use of BF3, which has a relatively short shelf
life even when refrigerated, can cause artefacts to form during derivitization and can lead to a
significant loss of polyunsaturated fatty acids (Christie, 1989). Its major advantage, however, is
that substantial amounts of fatty acids can be derivatized at a given time (~1g) and it only takes
15 min for complete derivitization. Cristie recommends the use of 6% (v/v) sulfuric acid in
methanol at 65oC for 12 h (Christie, 1989), because it employs a low temperature and does not
generate artefacts. The time can be shortened if constant stirring is used. Other options for
derivitizing agents include BCl3 and 1% (v/v) HCl in water.
The challenging part of trans fatty acid analysis is resolving the cis and trans isomers
from each other. There are two columns reviewed for use in trans fatty acid analysis, the SP-
2560 and Cp-88 (Ratnayake and others 2006). Both columns have a very polar bis-cyanopropyl
stationary phase and are 100-m in length, have a 0.25-mm-inner diameter, and a film thickness of
0.20-μm. Even with optimized operating conditions, considerable overlap still exists especially
when there is a high concentration of oleic acid in the lipid. However, cis isomers are eluted
after the oleic acid peak while the trans isomers are eluted before (Buchanam, 2008). This
makes determination of total trans-fat applicable even if all the trans peaks are not resolved from
one another. Peaks are identified based on retention times of known standards.
15
One way to isolate trans fatty acids, for GC analysis, from their cis counterparts is by
silver-ion TLC. By impregnating silica with AgNO3, trans FAMEs can be separated from cis
FAMEs. The silver ions form complexes with the double bonds in the FAMEs. Cis isomers
form more stable complexes with Ag+ because they are less sterically hindered than trans
FAMEs, therefore, trans FAMEs will migrate further than cis FAMEs. The separation of
FAMEs follows the order of polyunsaturated FAMEs (cc, ct, tt, for C18:2 in order of decreasing
Rf values), monounsaturated cis-isomers, monounsaturated trans-isomers, and then saturated
FAMEs. The TLC plates can be developed either in a mixture of heptane and diethyl ether or
toluene at -25oC. Silver-ion LC and silver-ion HPLC have also been used to isolate trans
FAMEs.
Trans Fatty Acid Analysis: Fourier Transform Infrared Spectroscopy
FTIR spectroscopy is widely used to determine the amount of trans-fat in foodstuffs.
There are several AOCS official methods available for the determination of trans fatty acids for
hydrogenated and refined oils. AOCS methods Cd 14-95 (AOCS Official method Cd 14-95,
2000) and Cd 14d-99 (AOCS Official method Ce 14d-99, 1999) are the official methods for
determining trans fatty acid content using FTIR spectroscopy. The most practical method, Cd
14d-99, uses a horizontal attenuated total reflectance (ATR) cell as this allows for direct analysis
of either triacylglycerols or FAMEs without having to weigh or quantitatively dilute the samples
in volatile solvents. This method is able to accurately detect trans fatty acids down to 1.0% by
weight (AOCS Official method Ce 14d-99, 1999). Instead of transmission or partial reflections
horizontal ATR cells utilize the internal reflection of a substance, in this case FAMEs, to
measure how much of the IR waves are internally reflected. The depth of penetration of the IR
16
light is extremely small, so accurate weighing is not needed, only that the cell is completely
covered with no air bubbles (Mossoba and others 2007b). Many ATR cells can be temperature
controlled, because the lipid needs to be in the liquid state. To ensure that triacylglycerols are in
the liquid state the cell should be heated to around 65oC, but for FAMEs, the temperature should
be maintained at ~40oC. The temperature must not only keep the sample in a liquid state it must
also be kept relatively constant, as the reflective property of a liquid is temperature dependent.
The main problem with FTIR methodologies is that the detection limit is around 1% trans-fat per
gram total fat, so small quantities are difficult to detect (AOCS Official method Ce 14d-99,
1999; Mossoba and others 2007a; Sherazi and others 2009). This stems from the fact that the
trans-fat peak at 966 cm-1 lies on a sloping baseline. Some recent advances in methods have
improved the accuracy at low levels by eliminating the baseline problem. For instance, one
study showed that by using reference oils identical to the sample oil but devoid of trans-fat, the
detection limit was lowered to 0.1% (Mossoba and others 2007a). By using a reference oil
identical to the sample but devoid of trans-fat for the background, it produces a straight baseline
because it subtracts out the equivalent slope that is seen in the sample itself. However,
determining and finding reference oils to use is difficult when analyzing samples that were not
industrially processed oils. Another way of increasing the accuracy at low levels is by using the
second derivative absorbance spectroscopy (Mossoba and others 2007a). By using the 2nd
derivative, the slope of the baseline becomes straight. The 2nd derivative is also useful for
identifying peaks that are shoulders of other peaks. For regulatory practices, an FTIR reading of
3.57% trans-fat would translate into 0.5-g trans-fat per serving (14g) of oil. Consequently
readings below this value will result in a trans-fat content of 0-g on the nutrition label, making
the low limit of the detection a non-issue for regulatory procedures.
17
Chapter 2
MATERIALS, METHODS, RESULTS, AND DISCUSSION
Chemicals
Soybean oil and vegetable shortening for feed trials were purchased from Kroger’s
(Athens, GA). Tallow was obtained from Welch, Holme & Clark Co., Inc. (Newark, NJ). ACS-
grade anhydrous methanol, hexanes, chloroform, carbon disulfide, silver nitrate, and sulfuric acid
were purchased from Fisher Scientific Ltd. (Suwanee, GA), whereas anhydrous sodium sulfate,
acetonitrile, and sodium hydroxide were acquired from VWR International Inc. (Suwanee, GA).
Hydroquinone, 14% (v/v) boron trifluoride in methanol, TLC plates (20 x 20 cm, 250 μm, on
polyester), diethyl ether, 2',7'-dichlorofluorescein, hydrochloric acid, and a Supelco 37
component FAME mixture were purchased from Sigma Chemical Company (St. Louis, MO).
Absolute (200 proof) enthanol was obtained from AAPER Alcohol and Chemical Co.
(Shelbyville, KY). Triheptadecanoin (>99%), methyl heptadecanoate (>99%), methyl oleate
(>99%), and methyl elaidate (>99%) were acquired from Nu-Chek Prep Inc. (Elysian, MN).
Industrial-grade nitrogen, scientific air, UHP helium, and UHP hydrogen were obtained from
Airgas National Welders (Toccoa, GA).
Feed Trials
Fifty adult Single Comb White Leghorn (SCWL) laying hens were maintained on a diet
free of animal by-products for a period of 30 days. After that time, hens were divided into five
18
groups of ten hens each and fed diets containing different types of fats to ascertain the effect of
varied trans-fat levels and various types of lipids on the trans-fat levels in eggs. The five diets
were as follows: I - Standard corn/soy laying ration with 0% added fat/oil; II - Standard corn/soy
laying ration with 4% (w/w) added soybean oil; III - Standard corn/soy laying ration with 4%
(w/w) added tallow; IV - Standard corn/soy laying ration with 4% (w/w) added shortening; and
V - Standard corn/soy laying ration with 4% (w/w) added tallow shortening blend. During the
feeding period, both feed consumption and egg production were monitored, as was any gain or
loss in body weight. Each hen was individually leg banded so as to ensure proper identification.
Eggs from each of the five groups of hens were collected on day 7, 14, and 21. Five egg yolks,
from each of the five groups, were pooled and homogenized separately before analyzing for
trans-fats.
Acid Hydrolysis
Lipids that are covalently and ionically bonded to proteins and carbohydrates are released
by digestion with hydrochloric acid. Feed samples were ground, while still frozen to avoid
deterioration of the samples. Approximately 32 g of ground feed was transferred to a 250-mL
Erlenmeyer flask to which 60 mL of ethanol and 60 mL of 8 N HCl were added. The
Erlenmeyer flask was covered with a watch glass placed in an 80oC water bath for 40 min with
occasional swirling . After cooling to room temperature, the lipids were extracted as follows.
Lipid Extraction
Total lipids were extracted by the Bligh & Dyer method (Bligh and Dyer, 1959) with
slight modifications. Briefly, 25 g of separated egg yolks were weighed into a 250-mL
19
Erlenmeyer flask, followed by 35 mL of deionized water to adjust the moisture content to 80%.
The contents were then blended with a Polytron® PT 3100 homogenizer for 1 min at a speed of
15,000 rpm (Kinematica, Inc. Bohemia, NY). Then, 50 mL of methanol and 25 mL of
chloroform were added creating a monophasic system. To create a diphasic system 25 mL of
chloroform were added to the egg yolk solution, followed by 25 mL of deionized water. The
lower organic phase was comprised of nearly 100% chloroform and extracted lipids, while the
upper phase was comprised of almost entirely methanol-water. Finally, an additional 35.5 mL of
chloroform were added to aid in separation. The egg yolk mixture was homogenized between
each addition for complete lipid extraction. A 2-min quiescent period facilitated separation
before filtering under slight vacuum through a Büchner funnel lined with Whatman #1 filter
paper. The filter paper and solids were returned to the Erlenmeyer flask. To this, 37.5-mL
chloroform and ~10 mg of hydroquinone (as an antioxidant) were added. Contents were blended
for 1 min and again filtered as described above. The supernatant containing any residual non-
polar lipids was transferred to a 250-mL separatory funnel and the phases were allowed to
separate overnight. The bottom chloroform layer was collected through anhydrous sodium
sulfate to remove any remaining moisture. The lipid extract was then transferred to a 250-mL
round bottom flask and concentrated with a Büchi Rotovapor R-210 using a V-700 vacuum
pump connected to a V-850 vacuum controller (Büchi Corporation, New Castle, DE) at 35oC and
25 kPa. Collected lipids were transferred to a 20-mL amber vial and the chloroform was
evaporated off with an N-EVAP 111 nitrogen evaporator (Organomation Associates, Inc. Berlin,
MA). Lipids were extracted in triplicates and were stored under a nitrogen headspace at -80oC to
retard lipid oxidation.
20
Fatty Acid Methylation
Fatty acid profiles of egg lipids were determined by GC-FID after converting
triacylglycerols to their respective FAME (Christie, 1989). Approximately 80 mg of extracted
lipids were weighed into a 5-mL Reacti-vialTM (Pierce, Rockford, IL). Then, 100 µL of 20%
(w/v) triheptadecanoin in hexanes were added to each Reacti-vialTM as an internal standard. The
triacylglycerols were then hydrolyzed and derivatized with 2-mL transmethylation reagent
consisting of 6% (v/v) concentrated H2SO4 in anhydrous methanol with ~5 mg of hydroquinone.
Each Reacti-vialTM was incubated at 65°C overnight with stirring on a Pierce Reacti-Therm III
incubating block (Pierce, Rockford, IL, USA). The following day samples were removed from
the incubator and allowed to cool to room temperature. Then, 1 mL of deionized water was
added to each Reacti-vialTM and the contents were vortexed for ~30 s. FAMEs were extracted 3x
with 1.5-mL aliquots of hexanes and subsequently washed 2x with 1.5 mL of deionized water.
The extracted FAMEs were collected in a clean test tube and then dried under nitrogen gas using
the N-EVAP nitrogen evaporator. The FAMEs were dissolved in 3 mL of carbon disulfide, a
100-μL aliquot was transferred to a 2-mL amber GC vial and diluted with 900-μL carbon
disulfide; the vial was then capped and crimped for GC analysis.
Gas Chromatographic Analysis
FAMEs from diet rations and egg samples were analyzed for trans fatty acids using an
Agilent Technologies 6890N Network GC System with capillary split/splitless inlet with
electronic pneumatics control (EPC) and a flame ionization detector (FID) for packed and
capillary columns (Agilent Technologies, Wilmington, DE). A 100-m, highly polar,
biscyanopropyl column with an inner diameter of 0.25 mm and a film thickness of 2 μm (SP-
21
2560, Supelco, Bellefonte, PA) was employed. The initial oven temperature was held at 140oC
for 5 min then ramped at 5oC/min to a final temperature of 240oC, which was maintained for 15
min resulting in a total runtime of 45 min. The injector temperature was set at 250oC. An
Agilent 7683 auto-sampler tray module equipped with a 7683B auto-injector module and a 10-
μL syringe was employed for injecting 1-μL samples through a pre-pierced 11-mm inlet septum
(P/N: 5183-4761-100) into the GC inlet containing a split liner packed with glass wool (P/N:
5183-4647). The GC was operated in split mode at a split ratio of 50:1 with helium as the carrier
and gas flow at 59.2 mL/min. The column pressure was 276.62 kPa. The detector was heated to
250oC and the FID flame was generated from hydrogen at a flow of 40 mL/min, air at 450
mL/min, and a nitrogen makeup gas flow of 23.9 mL/min. A Supelco 37 component FAME
mixture was used to identify and quantify individual FAMEs by retention time mapping. A
relative response factor was calculated for each FAME using methyl heptadecanoate as an
internal standard. Response factors are the concentration of an analyte divided by the response
to that analyte.
Ri = Wsi / Psi
where, Ri = response factor for FAME i; Wsi = mg of FAME i injected into the inlet; and Psi =
peak area corresponding to FAME i.
A relative response factor is defined as the response factor of the internal standard divided by the
response factor of the analyte.
RRfi = Rfis/Rfi
where, RRfi = relative response factor for FAME i; Rfis = response factor for the internal
standard; and Rfi = response factor for FAME i. Each FAME will have a different response to
the FID depending on chain length, saturation, and cis/trans configuration. For positively-
22
identified compounds, employment of relative response factors improves the accuracy of the
data. For unidentified compounds the relative response factor for the closest positively identified
fatty acid was used.
Silver-Ion Thin Layer Chromatography (Ag+-TLC)
Silver Ion TLC provides a means to separate cis/trans isomers, and to increase the
resolution of trans fatty acids by GC-FID. Trans-fat isomers of C18:1 may be coeluted or
appear as shoulders off the much larger oleic acid peak. By eliminating the cis components from
the chromatogram all trans peaks are able to be visualized. A 20 x 20 cm 250-µm silica G TLC
plate was submerged into 20% (w/v) AgNO3 in CH3CN for 20 min. Before analysis, the silica
plate was activated at 120oC for 1 h. One hundred microliters of FAMEs were spotted as a long
contiguous band across the origin on the silica plate and developed in hexanes:diethyl ether
(80:10, v/v). The plates were sprayed with 0.1% (w/v) 2',7'-dichlorofluorescein in ethanol and
viewed under UV light to visualize the bands. The band corresponding to the eluted trans
FAMEs was scraped off the plate and the FAMEs were extracted 2x with 25-mL aliquots of
diethyl ether. The ether was washed 2x with water to remove any residual silica, and then dried
using the N-EVAP system to remove the organic solvent. Subsequently, 25 µL of methyl
heptadecanoate, as an internal standard and 0.5-mL carbon disulfide were added to the isolated
trans fraction, followed by GC-FID analysis as already described.
Linearity: Gas Chromatography
The linearity of the peak area response versus concentration for the standard
triheptadecanoin was studied from 75-600 μg/mL. The calibration curve was generated with
23
eight different concentrations (i.e., 75, 150, 225, 300, 375, 450, 525, 600 μg/mL) using the least-
squares method.
Precision and Accuracy: Gas Chromatography
Assay precision was determined by repeatability (intra-day) and intermediate precision
(inter-day). Intra-day was evaluated by assaying samples at the same concentration and during
the same day. The inter-day was studied by comparing the assay on three different days
(Demirkaya and Kadioglu, 2007). The accuracy of this analytic method was evaluated by
checking three different concentrations of triheptadecanoin.
Limit of Detection (LOD) and Limit of Quantification (LOQ): Gas Chromatography
LOD and LOQ were determined by an empirical method that consisted of analyzing a
series of standard solutions containing decreasing amounts of triheptadecanoin. The LOD and
LOQ were defined as having a signal-to-noise (S/N) ratio of 3 to 1 and 10 to 1, respectively
(Bononi and Tateo, 2008).
Recovery: Gas Chromatography
The analytical recovery of triheptadecanoin was assessed by direct comparison of
concentrations of triheptadecanoin to methyl heptadecanoate. Three replicates at four different
concentration levels (i.e., 75, 150, 300, and 600 μg/mL) was employed.
24
Analysis by Fourier Transform Infrared Spectroscopy
Extracted lipids from diet and egg samples were transmethylated according to AOCS
official method Ce 2-66. This method was used for FTIR analysis because it was completed in a
shorter amount of time, thereby, allowing the FTIR method to maintain its rapidity in analyzing
trans-fats. Moreover, a greater quantity of sample can be transmethylated by this method, which
is important, because a larger quantity of sample is needed for FTIR analysis compared with GC
analysis. One gram of egg lipids was transferred to a 250-mL flat bottom flask. Then, 10 mL of
0.5 N NaOH in anhydrous methanol and several glass boiling beads were added to the flask,
which was then connected to a condenser. The mixture was refluxed for 10 min, forming free
fatty acids from triacylglycerols. A subsequent 12-mL addition of 14% (v/v) boron trifluoride in
methanol, as the transmethylating reagent, was made through the top of the condenser. After 2
min, 5 mL of hexanes were introduced and the mixture was refluxed for an additional minute.
The contents were transferred to a 250-mL separatory funnel where FAMEs were extracted 2x
with 50-mL aliquots of hexanes. The organic layer was passed through anhydrous sodium
sulfate to remove any moisture and collected in a 100-mL round bottom flask. The hexanes were
evaporated off using the Büchi Rotovapor system, and solvent residue was removed under a
stream of nitrogen with the N-EVAP.
A 300-μL aliquot of FAMEs was pipetted onto a Thermo Smart Ark Horizontal
Attenuated Total Reflectance (ATR) cell installed in a Nicolet 6700 FTIR spectrophotometer
(Thermo Fisher Scientific, Rockford, IL). The ATR cell was maintained at a temperature of
45oC to ensure samples were in a liquid state. The ATR cell was cleaned with a dustless wipe
prior to each sampling. Thirty two scans over a wavenumber range of 4000 to 400 cm-1 were
taken at a resolution of 4 cm-1. A new background was collected prior to each sample, and the
25
results were expressed as absorbances. A series of standards were made by accurately weighing
proportions of methyl elaidate and methyl oleate ranging from 0.5 to 10% methyl elaidate by
weight. Using these standards, a calibration curve was generated using the TQ Analyst software
program. Partial least squares regression analysis of second derivative spectra in the
wavenumber range of 978 cm-1 to 955 cm-1 was employed to regress the standards. The Norris
derivative filter was used to smooth the spectra, reducing random noise, thus decreasing the
random error in the calibration model.
Linearity: Fourier Transform Infrared Spectroscopy
The linearity of the peak area versus %trans-fat per g total fat for mixtures of methyl
elaidate and methyl oleate ranging from 0.5 to 10% methyl elaidate by weight was determined.
The calibration curve was generated with eight percentage proportions (0.5, 1, 2, 3, 4, 5, 6, 7, 8,
9, and 10%) using the least-squares method.
Precision, Accuracy, and Limit of Detection (LOD): Fourier Transform Infrared
Spectroscopy
Assay precision was determined by repeatability (intra-day) and intermediate precision
(inter-day). Intra-day was evaluated by assaying samples, at the same concentration and during
the same day. The inter-day was studied by comparing the assay on three different days
(Demirkaya and Kadioglu, 2007). The accuracy of this analytic method was evaluated by
assessing two different concentrations of methyl elaidate in methyl oleate (5 and 8%).
26
Statistical Methods
Eggs were collected at 7, 14, and 21 days after initiation of the experimental diets. The
eggs for a diet were pooled on each day and analyzed in triplicate (n=9). The experimental unit
was the hen, and the measurement unit was the pool of five egg yolks. A one-way ANOVA for
egg composition and diet was performed using Minitab Statistical Software. Tukey’s multiple
range test, with a family error rate of five, was employed for the comparison of fatty acid profiles
in the diets to determine if significant (p < 0.05) differences existed.
Results and Discussion
The weights of whole eggs (g) including the shell, the whole liquid egg, the egg white,
and the egg yolk were recorded every week for each diet. Over the three-week period, the
average weights of the shells, egg whites, and yolks were recorded and found not to be
statistically (p > 0.05) different amongst the test diets. The average whole egg weight did not
vary over the three-week collection period, but was statistically (p=0.001) different between hens
fed different diets. Egg weights from hens fed diets I, II, III, and IV were not statistically (p >
0.05) different. This, however, was not the case for diet V; egg weights from hens fed diet V
showed the lowest average egg weight of 59.05 g. A summary of egg weights is provided in
Table 1.
Gas Chromatography
Figures 2 through 11 depict GC-FID chromatograms of egg yolk lipid extracts (i.e., after
derivatization to FAMEs) from hens fed diets I-V as well as chromatograms for each of the five
trial diets. The chromatograms were completely resolved (i.e., with excellent resolution between
27
peaks, Rs > 2.0), and in some cases fully resolving the cis/trans isomers. The baseline was very
stable even with a temperature ramp of 140 to 240oC. The peaks indicate a Gaussian distribution
thereby indicating that the column was not overloaded with the sample.
A Supelco 37 component FAME mixture was used to identify and quantify individual
FAMEs by retention time mapping. Each of the 37 components in the standard was identified
and their retention times recorded. Samples’ FAMEs retention times were compared to those of
the Supelco 37 component FAME mixture. Table 2 lists the retention times, response factors
and relative response factors for each of the 37 compounds in the standard. Unidentified peaks
were tentatively identified based by retention times when compared to positively-identified
FAMEs. Peak 3, as shown in Figure 5, was not included in the Supelco 37 component FAME
mixture but eluted between palmitic acid, C16:0 (peak 2), and palmitoleic acid, C16:1 n-7 (peak
4), it is suspected that this peak corresponds to an isomer of palmitoleic acid, C16:1 n-9.
Silver-ion TLC was used to separate trans-monounsaturated FAMEs from cis-
monounsaturated FAMEs. The Rf value for cis-monounsaturated FAMEs was 0.28, whereas the
Rf value for trans-monounsaturated FAMEs was 0.34, and for saturated FAMEs it was 0.43.
The GC-FID chromatogram of the trans isomers contained in shortening compared with its full
fatty acid profile and is depicted in Figure 12. This figure indicates that a very small portion
(i.e., < 3% of all trans fatty acids present in the sample) of trans FAMEs coeluted with the oleic
acid peak. This chromatogram is consistent for all lipids containing trans fatty acids from this
study. Figure 13 identifies the five trans fatty acid peaks present at detectable levels from GC-
FID chromatograms analyzed by Juanéda and coworkers (Juanéda and others 2007). The first
peak, 6t-10t C18:1, eluted as a shoulder off the 11t C18:1 peak. This was followed by the
elution of the isomers 12t C18:1, 13t-14t C18:1, and finally 15t C18:1, which coeluted with the
28
oleic acid peak in the full chromatogram. The major trans fatty acid in the sample was found to
be 11t C18:1 (vaccenic acid). Therefore, all fatty acids eluting directly before oleic acid are
trans fatty acids, while those eluting afterwards are cis isomers of C18:1. For calculations of
total trans-fats present, the peak areas of 6t-14t C18:1 were used.
The trans-fat contents of diet I (i.e., standard corn/soy-laying ration with 0% added
fat/oil) and diet II (i.e., standard corn/soy-laying ration + 4% (w/w) soybean oil) were below the
detection limit for the GC-FID system. It, however, cannot be assumed that the diets were
devoid of all trans-fats. Not unexpectedly, the trans-fat contents of egg samples collected from
the hens fed either diet I or II for 21 days were also below the limit of detection of the GC
system. Stating this another way: hens fed diets devoid of detectable levels of endogenous
trans-fats, produced eggs that did not contain detectable levels of trans-fats. Unlike diets I and
II, diets III, IV, and V contained some level of fats known to come from the partial
hydrogenation process of edible oils. The trans-fat levels in diets III, IV, and V (i.e., standard
corn/soy-laying ration + 4% tallow, shortening, and tallow-shortening blend, respectively) were
found to contain 2.78 ± 0.08, 3.25 ± 0.03, and 2.85 ± 0.05 g trans-fat/100-g dietary fat,
respectively, whereas, the egg yolk from hens fed these diets showed only 3.33 ± 0.32, 4.16 ±
0.03, and 3.52 ± 0.05 g trans-fat/100-g dietary fat, respectively. Therefore, approximately one
quarter of the trans-fats available in each ration was deposited in the egg yolks of hens
consuming this feed. The percentages correspond to under 0.1% trans-fats (g trans-fat/g whole
egg) in a serving of eggs. It was observed that eggs contained similar proportions of both elaidic
acid (9t C18:1) and vaccenic acid (11t C18:1) in the yolks despite a much higher concentration
of vaccenic acid in the diet. This may be a metabolic consequence, as desaturation of vaccenic
acid is seen in humans.
29
Over 98% by weight of the lipids in the recovered lipid extracts comprised six fatty acids,
namely myristic acid (C14:0), palmitic acid (C16:0), palmitoleic acid (C16:1 n-7), stearic acid
(C18:0), oleic acid (C18:1 n-9), linoleic acid (C18:2 n-6), γ-linolenic acid (C18:3 n-6), and
arachidic acid (C22:0). Of these, oleic acid is the dominant fatty acid contributing to greater than
40% (g/total g lipid). The weight percentages of the aforementioned fatty acids for each diet are
shown in Table 3, while saturated, monounsaturated, polyunsaturated, and trans-fatty acids are
summarized in Table 4. The myristic acid content was the highest, at 0.6% (w/w), in hens fed
diets III, IV, and V; diets III, IV, and V each contained approximately 2% myristic acid. A
summary of the fatty acid profile of the feeds is reported in Table 5. Although diets I and II
possessed no detectable levels of myristic acid, in egg yolks, a myristic acid content of 0.43%
was detected. Palmitic acid, on the other hand, remained relatively constant at around 26.2% and
there was no correlation determined between how much palmitic acid was present in the feed.
Eggs from hens fed diets I, IV, and V contained the highest levels of palmitic acid, 26.9, 26.4,
and 26.5% respectively, and were statistically (p < 0.01) different from eggs of hens fed diets II
and III, which contained 25.3 and 25.9% palmitic acid, respectively. Stearic acid showed the
opposite relationship towards diets compared to palmitic acid; that is, eggs from hens fed diets I,
IV, and V contained the least amount of stearic acid (i.e., 9.2, 9.0, and 9.0%), while eggs from
hens fed diets II and III contained the highest percentage of stearic acid (i.e., 10.4 and 9.7%).
Arachidic acid varied from 1.8% in eggs from hens fed diet V, to 2% in eggs from hens fed diets
II, III, and IV, and to 2.3% in eggs from hens fed diet I. None of the feeds contained a detectable
level of arachidic acid. Even though there were variations in the levels of the four primary
saturated fats – namely myristic, palmitic, stearic, and arachidic acid – a relatively constant
amount of saturated fatty acids, 38.1 ± 0.46%, of total fat resulted.
30
Egg yolks contained low levels of palmitoleic acid ranging from 1.8% in eggs from hens
fed diet II, to 2.5% in eggs from hens fed diet III, to 2.9% in eggs from hens fed diets IV and V,
and finally to 3.2% in eggs from hens fed diet I. Oleic acid constitutes the major portion of
lipids in egg yolks. As the oleic acid content in the feed increased (i.e., 20.5, 21.1, 27.6, 29.2, and
29.6% for diets II, I, IV, V, and III, respectively), the oleic acid levels in the eggs also
systematically increased (i.e., 40.0, 43.0, 45.9, 46.9, and 47.6% for diets II, I, IV, V, and III,
respectively). Monounsaturated fatty acids showed a similar pattern as that of oleic acid likely
because oleic acid constitutes the bulk of the monounsaturated fatty acids.
Linoleic acid showed the most variation amongst the diets and eggs produced. Diets I
and II contained 54% linoleic acid and the hens gave eggs with a linoleic acid content ranging
from 14.5 to 18.8%. Eggs from hens fed diets III, IV, and V, which contained only 25% linoleic
acid, were found to have 10.2, 11.3, and 10.7% linoleic acid contents, respectively. Furthermore,
all eggs contained very small quantities of γ-linolenic acids ranging from 0.2 to 1.0%. Note,
however, that no α-linolenic acid (C18:3 n-3) was detected in any of the diets or eggs of this
study. The quantities of polyunsaturated and monounsaturated fats in the eggs varied depending
on how much of each was contained in the hens feed.
The GC method was evaluated for the recovery of FAMEs from triacylglycerols during
analysis, linearity, precision and accuracy, as well as LOD and LOQ. The recovery of FAMEs
was analyzed by comparing the peak areas of triheptadecanoin to that of methyl heptadecanoate
at four different concentrations (i.e., 75, 150, 300, and 600 μg/mL). It was determined that
triheptadecanoin had a recovery of only 81.33 ± 1.5% compared to methyl heptadecanoate which
had a 100% recovery; a summary of this data is presented in Table 6. It can be assumed that if
the triheptadecanoin had only an 81% recovery, then the triacylglycerols from the lipid extracts
31
should also have a similar recovery. Based on this reasoning, triheptadecanoin was deemed to be
the better choice of an internal standard for GC analysis and was so used when applicable. The
linearity of triheptadecanoin was assessed by generating a calibration curve using eight different
concentrations of the standard (i.e., 75, 150, 225, 300, 375, 450, 525, 600 μg/mL) and the least-
squares method. The correlation coefficient, r, was found to be 0.999 (n = 24), the slope was
0.4033, and the intercept was -13.441. The method was analyzed for intra-day precision
[%relative standard deviation (%RSD)], and accuracy [percent relative error]. The %RSD was
0.86, 5.54, and 1.11%, whereas the percent relative error was -5.09, -3.32, and 5.40%, for
triheptadecanoin concentrations of 160, 320, and 520 μg/mL, respectively. The inter-day
precision was determined to be 8.64, 6.15, and 4.92%, while the inter-day accuracy was -11.03, -
4.71, -0.41%, for triheptadecanoin concentrations of 160, 320, and 520 μg/mL, respectively.
Table 7 summarizes the precision and accuracy results for GC analyses. The LOQ, defined as
the lowest concentration of measurable value of a standard solution of triheptadecanoin, was
determined to be 6 μg/mL while the LOD was approximately 3 μg/mL.
Fourier Transform Infrared (FTIR) Spectroscopy
Methods have been developed to determine trans-fat percent using neat oils mainly from
edible oil processing industries (Sherazi and others 2009). These methods have been examined
for use in identifying the trans-fat content in diet and egg samples. Initially, extracted egg lipids
were employed as such, which eliminated the need for their derivatization to FAMEs. The
extracted lipids, however, showed an interference band at the wavenumber = 966 cm-1 (Figure
14), which interferes with the quantification of the trans-fats. Therefore, the extracted lipids
were transmethylated to FAMEs in an effort to either reduce or eliminate the interference band.
32
Transmethylated samples showed no interference at the characteristic trans absorption band of
966 cm-1.
Figures 15 and 16 depict the trans band at the 966 cm-1 of prepared samples as the second
derivative of absorbance versus wavenumber and absorbance versus wavenumber, respectively.
The trans band can be visualized better by taking the second derivative of absorbance
mathematically. This action removes the sloping baseline in the FTIR spectra; thus, making area
measurements more accurate at lower trans-fat levels and at the same time improves the
detection limit of the assay (Mossoba and others 2007a).
Figure 17 shows the results of the TQ Analyst software for the partial least squares (PLS)
regression with and without taking the second derivative of absorbance readings. It can be seen
that the PLS curve, using the second derivative spectrum, affords slightly less variance below 2%
trans-fat levels. This provides greater accuracy in the detection of trans-fat of samples that
possess trans-fat levels below 2%.
The trans-fat content of each diet, as well as pure tallow and shortening, was determined
by FTIR spectroscopy. Diets III, IV, and V had trans-fat concentrations of 3.33 ± 0.32, 4.16 ±
0.03, and 3.52 ± 0.05 g trans-fat/100-g dietary fat, respectively, while pure tallow and shortening
were measured as containing 5.44 ± 0.19 and 7.79 ± 0.56 g trans-fat/100-g dietary fat,
respectively. The intra-day precision for the FTIR method was 0.21 and 0.14%, and the intra-
day accuracy was -3.90 and 4.08% for concentrations of 5.11 and 7.88% methyl elaidate in
methyl oleate, respectively. Inter-day precision was determined to be 0.31% at both
concentrations whereas inter-day accuracy was -2.68 and -2.26% for concentrations of 5.11 and
7.88% methyl elaidate in methyl oleate, respectively. Precision and accuracy results are shown
in Table 8. The linearity was assessed using 11 different concentrations of methyl elaidate (i.e.,
33
0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10%). The correlation coefficient, r, was found to be 0.995 (n =
22), the slope was 0.1016, and the intercept was 0.0104. Figure 15 depicts the absorbance of
trans band at the 966 cm-1 wavenumber of prepared standards of methyl elaidate in methyl
oleate. It can also be noted that below 1% trans-fat, the baseline has a significant amount of
noise associated with it compared to the trans peak signal, thus, making accurate measurements
of the peak area quite difficult. The LOD and LOQ for this FTIR spectroscopic method were
determined to be 1 and 2.4%, respectively.
The FTIR results were compared to those obtained by GC-FID and an attempt was made
to establish a correlation between these two analytical techniques. The FTIR data for trans-fat
content in egg yolk lipid extracts was below the detection limit of the method. FTIR methods
have been cited as having limits of detection of 1% trans-fat per gram total fat. A correlation
coefficient of 0.988 was determined when comparing GC-FID and FTIR data of the hen diets
and pure fat samples, as depicted in Figure 18. No correlation could be made using egg yolk
data because the % trans-fat could not be quantified by the FTIR method.
Conclusions
Based on these findings, approximately 25% of the trans-fat available in the ration was
deposited into the egg yolks giving trans-fat levels of about 0.1 g/100 g whole egg. Using fats
and oils with high levels of trans-fat, such as grease from restaurants, rendering of animal
carcasses, and by-products from vegetable oil refining, to supplement laying hen feed will not
result in appreciable levels of trans-fat in the hens’ eggs. However, public perception of this
practice (i.e., supplementing rations with trans-fat) is not favorable. The type of fat
supplemented into the feed should be chosen knowing that the diet can significantly effect the
34
lipid composition of the egg mainly in its proportions of mono-to-polyunsaturated fatty acids. In
this study, saturated fatty acid levels remained relatively constant.
GC-FID analysis is a more sensitive technique at detecting low concentrations of trans-
fats or fatty acids in general compared with FTIR spectroscopy. It also provides information
about the types of fatty acids contained in the egg yolk, whereas FTIR analysis gives only the
percentage of trans-fat within the lipid extract. The FTIR method provided more precise and
accurate measurements of trans-fat at elevated concentrations compared to GC-FID. If extremely
low levels of trans-fat are not a concern, as is the case for nutrition labeling purposes (NB, if the
trans-fat content is less than 0.5 g per serving, then it is indicated as 0 g on the label), then the
FTIR approach offers a rapid and cost-effective means of determining trans-fat contents in
foodstuffs.
35
References
Akoh CC & Min DB. 2002. Food Lipids: Chemistry, Nutrition, and Biotechnology. New York:
Marcel Dekker, Inc.
Anonymous. 2005. Oleic acid material safety data sheet.
<http://msds.chem.ox.ac.uk/ME/methyl_oleate.html>.
Anonymous. 2006. Sunflower Oil. 2/11/2009
<http://www.sunflowernsa.com/oil/default.asp?contentID=43>.
City of Philadelphia. 2007a. Foods Containing Artificial Trans Fats. (070511).
Anonymous. 2007b. Safety data for methyl oleate. 5/16/08
<http://msds.chem.ox.ac.uk/ME/methyl_oleate.html>.
California State Assembly. 2008. Food facilities: trans-fats. (97).
AOCS Official method Cd 14-95. 2000. AOCS Official method Cd 14-95, Isolated trans isomers
infrared spectrometric method Champaign IL: AOCS Press.
AOCS Official method Ce 1f-96. 2002. AOCS Official method Ce 1f-96, Determination of cis-
and trans fatty acids in hydrogenated and refined oils and fats by capillary GLC.
Champaign IL: AOCS Press.
AOCS Official Method Ce 2-66. 1997. AOCS Official method Ce 2-66, Preparation of methyl
esters of long-chain fatty acids. Champaign IL: AOCS Press.
AOCS Official method Ce 14d-99. 1999. AOCS Official method Ce 14d-99, Rapid
determination of isolated trans geometric isomers in fats and oils by attenuated total
reflection infrared spectroscopy. Champaign IL: AOCS Press.
36
Baer DJ, Judd JT, Clevidence BA & Tracy RP. 2004. Dietary fatty acids affect plasma markers
of inflammation in healthy men fed controlled diets: a randomized crossover study. Am.
J. Clin. Nutr. 79:969-973.
Bligh EG & Dyer WJ. 1959. A rapid method of total lipid extraction and purification. Can. J.
Biochem. Physiol. 37:911-917.
Bononi M & Tateo F. 2008. Liquid chromatography/tandem mass spectrometry analysis of
chloramphenicol in propolis extracts available on the italian market. J.Food Comp. Anal.
21:84-89.
Buchanam MD. 2008. Fast GC analysis of detailed cis/trans fatty acid methyl esters (FAMEs) on
the 75m SP-2560 capillary column. The Reporter. Sigma-Aldrich.
Christie WW. 1989. Gas Chromatography and Lipids. AYR, Scotland: The Oily Press.
Coon CN, Obi I & Hamre ML. 1988. Use of barley in laying hen diets. Poult. Sci. 67:1306-1313.
Demirkaya F & Kadioglu Y. 2007. Simple GC-FID method development and validation for
determination of α-tocopherol (vitamin E) in human plasma. J. Biochem. Biophys. Meth.
70:363-368.
Fick G. 1983. Genetics and breeding of sunflower. J. Am. Oil Chem. Soc. 60:1252-1253.
Flickinger B & Huth P. 2004. Dietary fats and oils: Technologies for improving cardiovascular
health. Curr. Atheroscler. Rep. 6:468-476.
Food and nutrition board iomotna. 2002. Letter report on dietary reference intakes for trans fatty
acids. Institute of Medicine.
Gunstone FD. 2007. The Lipid Handbook. New York: Chapman and Hall.
Hammond E & Fehr W. 1984. Improving the fatty acid composition of soybean oil. J. Am. Oil
Chem. Soc. 61:1713-1716.
37
Hu FB, Stampfer MJ, Manson JE, Rimm E, Colditz GA, Rosner BA, Hennekens CH & Willett
WC. 1997. Dietary fat intake and the risk of coronary heart disease in women. N. Engl. J.
Med. 337:1491-1499.
Hu FB, van Dam RM & Liu S. 2001. Diet and risk of Type II diabetes: the role of types of fat
and carbohydrate. Diabetologia 44:805-817.
Huth PJ. 2007. Do ruminant trans fatty acids impact coronary heart disease? Lipid Technol.
19:59-62.
Jakobsen MU, Overvad K, Dyerberg J & Heitmann BL. 2008. Intake of ruminant trans fatty
acids and risk of coronary heart disease. Int. J. of Epidemiol. 37:173-182.
Jiang J, Bjoerck L, Fonden R & Emanuelson M. 1996. Occurrence of conjugated cis-9, trans-11-
octadecadienoic acid in bovine milk: Effects of feed and dietary regimen. J. Dairy Sci.
79:438-445.
Juanéda P, Ledoux M & Sébédio J-L. 2007. Analytical methods for determination of trans fatty
acid content in food. Eur. J. Lipid Sci. Technol. 109:901-917.
Katan MB, Zock PL & Mensink RP. 1995. Trans fatty acids and their effects on lipoproteins in
humans. Annu. Rev. Nutr. 15:473-493.
Kodali DR & List GR. 2005. Trans Fats Alternatives. Champaign, Illinois: AOCS Press.
Kramer J, Blackadar C & Zhou J. 2002. Evaluation of two GC columns (60-m SUPELCOWAX
10 and 100-m CP sil 88) for analysis of milkfat with emphasis on CLA, 18:1, 18:2 and
18:3 isomers, and short- and long-chain FA. Lipids 37:823-835.
Kung H-C, Hoyert DL, Jiaquan X & L. MS. 2008. Deaths: Final Data for 2005. Division of Vital
Statistics. p. 1-121.
38
Kyoko Hayakawa, Yu-Yen Linko & Pekka Linko. 2000. The role of trans fatty acids in human
nutrition. Eur. J. Lipid Sci. Technol. 102:419-425.
Leeson S & Caston LJ. 1990. Dietary flax and egg composition. Poult. Sci. 69:1617-1620.
Lopez-Garcia E, Schulze MB, Meigs JB, Manson JE, Rifai N, Stampfer MJ, Willett WC & Hu
FB. 2005. Consumption of trans fatty acids is related to plasma biomarkers of
inflammation and endothelial dysfunction. J. Nutr. 135:562-566.
March BE & MacMillan C. 1990. Linoleic acid as a mediator of egg size. Poult. Sci. 69:634-639.
Meyer KA, Kushi LH, Jacobs DR, Jr. & Folsom AR. 2001. Dietary fat and incidence of type 2
diabetes in older Iowa women. Diabetes Care 24:1528-1535.
Micha R & Mozaffarian D. 2008. Trans fatty acids: Effects on cardiometabolic health and
implications for policy. Prost. Leuk. Essent. Fatty Acids 79:147-152.
Morris MC, Evans DA, Bienias JL, Tangney CC, Bennett DA, Aggarwal N, Schneider J &
Wilson RS. 2003. Dietary fats and the risk of incident alzheimer disease. Arch. Neurol.
60:194-200.
Mossoba MM, Kramer JKG, Milosevic V, Milosevic M & Azizian H. 2007a. Interference of
Saturated Fats in the Determination of Low Levels of trans Fats (below 0.5%) by
Infrared Spectroscopy. J. Am. Oil Chem. Soc. 84:339-342.
Mossoba MM, Milosevic V, Milosevic M, Kramer JKG & Azizian H. 2007b. Determination of
total trans fats and oils by infrared spectroscopy for regulatory compliance. Anal.
Bioanal. Chem. 389:87-92.
Mozaffarian D. 2006. Trans fatty acids - Effects on systemic inflammation and endothelial
function. Atheroscler. Suppl. 7:29-32.
39
Mozaffarian D, Katan MB, Ascherio A, Stampfer MJ & Willett WC. 2006. Trans fatty acids and
cardiovascular disease. N. Engl. J. Med. 354:1601-1615.
Mozaffarian D, Pischon T, Hankinson SE, Rifai N, Joshipura K, Willett WC & Rimm EB. 2004.
Dietary intake of trans fatty acids and systemic inflammation in women. Am. J. Clin.
Nutr. 79:606-612.
Phivilay A, Julien C, Tremblay C, Berthiaume L, Julien P, Giguère Y & Calon F. 2009. High
dietary consumption of trans fatty acids decreases brain docosahexaenoic acid but does
not alter amyloid-beta and tau pathologies in the 3xTg-AD model of Alzheimer's disease.
Neuroscience 159:296-307.
Ratnayake NWM. 2004. Overview of methods for determination of trans fatty acids by gas
chromatography, silver-ion thin-layer chromatography, and gas chromatography/mass
spectrometry. J. AOAC Int. 87:523-539.
Ratnayake NWM, Hansen SL & Kennedy MP. 2006. Evaluation of the CP-Sil 88 and SP-2560
GC Columns Used in the Recently Approved AOCS Official Method Ce 1h-05:
Determination of cis-, trans-,Saturated, Monounsaturated, and Polyunsaturated Fatty
Acids in Vegatable or Non-ruminant Animal Oils and Fats by Capillary GLC Method. J.
Am. Oil Chem. Soc. 83:475-488.
Salmeron J, Hu FB, Manson JE, Stampfer MJ, Colditz GA, Rimm EB & Willett WC. 2001.
Dietary fat intake and risk of type 2 diabetes in women. Am. J. Clin. Nutr. 73:1019-1026.
Sell JL, Scheideler SE & Rahn BE. 1987. Influence of different phosphorus phase-feeding
programs and dietary calcium level on performance and body phosphorus of laying hens.
Poult. Sci. 66:1524-1529.
40
Services HaH. 2003. Food labeling; trans fatty acids in nutrition labeling; consumer research to
consider nutrient content and health claims.
Sherazi STH, Kandhro A, Mahesar SA, Bhanger MI, Talpur MY & Arain S. 2009. Application
of transmission FT-IR spectroscopy for the trans fat determination in the industrially
processed edible oils. Food Chem. 114:323-327.
Stender S, Dyerberg J & Arne A. 2006. High levels of industrially produced trans fat in popular
fast foods. N. Engl. J. Med. 354:1650-1652.
Subcommittee on Poultry Nutrition NRC. 1994. Nutrient Requirements of Poultry. Washington,
D.C.: National Academy Press.
Tarrago-Trani MT, Phillips KM, Lemar LE & Holden JM. 2006. New and existing oils and fats
used in products with reduced trans-fatty acid J. Am. Diet. Assoc. 106:867-880.
Tucker A & Markt S. 2006. Board of health votes to phase out artificial trans fat from New York
City’s restaurants. 3/12/09 <http://www.nyc.gov/html/doh/html/pr2006/pr114-06.shtml>.
Vaclavik VC, Elizabeth W. 2007. Essentials of Food Science. New York: Springer Publishing.
van Dam RM, Willett WC, Rimm EB, Stampfer MJ & Hu FB. 2002. Dietary fat and meat intake
in relation to risk of type 2 diabetes in men. Diabetes Care 25:417-424.
41
List of Tables
Table 1: Egg weights (g) of hens fed diets of varying levels of trans-fat for 21 days
Day
Diet*† 7 14 21
Ia 60.738 ± 3.12 62.30 ± 3.55 62.06 ± 4.42
IIa 61.327 ± 5.40 61.89 ± 4.56 63.30 ± 5.14
IIIa 59.123 ± 4.83 59.07 ± 4.23 61.50 ± 4.94
IVa 63.009 ± 3.89 62.27 ± 3.11 63.16 ± 4.12
Vb 57.945 ± 2.81 60.09 ± 3.80 59.11 ± 3.12 †Mean ± standard deviation (n = 12)
* Diet I – standard corn/soy-laying ration with 0% added fat/oil; Diet II – Standard corn/soy laying ration with 4% added soybean oil;
Diet III – Standard corn/soy laying ration with 4% added tallow; Diet IV - Standard corn/soy laying ration with 4% added shortening;
Diet V - Standard corn/soy laying ration with 4% added tallow/shortening blend
.a,b,c - Diets with different superscripts were significantly different (p<0.05) as determined by Tukey’s test
42
Table 2: Response factors of FAMEs as determined by GC-FID using Supelco 37 component FAME mixture
Peak # RT* Compound Name Response Factor† Relative Response Factor‡
1 10.46 Methyl Butyrate 0.0042 0.53 2 11.06 Methyl Hexanoate 0.0030 0.73 3 12.12 Methyl Octanoate 0.0027 0.81 4 13.89 Methyl Decanoate 0.0025 0.88 5 15.08 Methyl Undecanoate 0.0024 0.92 6 16.45 Methyl Dodecanoate 0.0024 0.94 7 17.97 Methyl Tridecanoate 0.0023 0.98 8 19.59 Methyl Myristate 0.0022 0.99 9 21.04 Methyl Myristoleate 0.0023 0.97 10 21.24 Methyl Pentadecanoate 0.0021 1.04 11 22.71 Methyl cis-10-pentadecenoate 0.0023 0.97 12 22.91 Methyl Palmitate 0.0022 1.03 13 24.12 Methyl Palmitoleate 0.0022 1.00 14 24.54 Methyl Heptadecanoate 0.0022 1.00 15 25.72 Methyl cis-10-Heptadecenoate 0.0022 1.02 16 26.11 Methyl Stearate 0.0021 1.05 17 26.90 Methyl Elaidate 0.0021 1.05 18 27.13 Methyl Oleate 0.0021 1.04 19 27.92 Methyl Linoleaidate 0.0022 1.01 20 28.62 Methyl Linoleate 0.0022 1.03 21 29.09 Methyl Arachidate 0.0021 1.06
43
22 29.72 Methyl y-Linolenate 0.0022 1.00 23 30.03 cis-11-Eicosenoic Acid Methyl Ester 0.0021 1.08 24 30.30 Methyl Linolenate 0.0022 1.01 25 30.49 Methyl Heneicosanoate 0.0020 1.10 26 31.47 cis-11,14-Eicosadienoic Acid Methyl Ester 0.0022 1.03 27 31.92 Methyl Behenate 0.0021 1.08 28 32.58 cis-8,11,14-Eicosatrienoic Acid Methyl Ester 0.0023 0.98 29 32.87 Methyl Eurcate 0.0021 1.04 30 33.16 cis-11,14,17-Eicosatrienoic Acid Methyl Ester 0.0025 0.90 31 33.32 Methyl Arachidonate 0.0020 1.09 32 33.46 Methyl Tricosanoate 0.0022 1.00 33 34.36 cis-13,16-Docosadienoic Acid Methyl Ester 0.0021 1.05 34 34.80 Methyl Tetracosanoate 0.0021 1.08 35 35.34 cis-5,8,11,14,17--Eicosapentaenoic Acid Methyl Ester 0.0023 0.97 36 35.84 cis-15-Tetracosenoate 0.0020 1.09 37 39.89 cis-4,7,10,13,16,19-Docosahexanoic Acid Methyl Ester 0.0024 0.91
*RT – Retention time
†Response factor: Ri = Wsi / Psi where, Ri = response factor for FAME i; Wsi = mg of FAME i injected into the inlet; and Psi = peak
area corresponding to FAME i.
‡ Relative response factor: RRfi = Rfis/Rfi where, RRfi = relative response factor for FAME i; Rfis = response factor for the internal
standard; and Rfi = response factor for FAME i.
44
Table 3: Fatty acid profile of egg yolk lipids from hens consuming diets with varying levels of trans-fat
Fatty acid Diet* % Fatty acid† Fatty acid Diet* % Fatty acid†
Ia 0.48 ± 0.01 Ia,b 9.16 ± 0.48 IIb 0.49 ± 0.04 IIc 10.70 ± 0.19 IIIc 0.59 ± 0.06 IIIa 9.72 ± 0.66 IVc 0.62 ± 0.02 IVb 8.96 ± 0.34
Myristic Acid
(C14:0)
Vc 0.63 ± 0.03
Stearic Acid (C18:0)
Vb 8.99 ± 0.57
Ia 26.83 ± 0.48 Ia 2.30 ± 0.08 IIb 25.29 ± 0.24 IIb,c 2.07 ± 0.09
IIIb,c 25.85 ± 0.63 IIIb 1.97 ± 0.15 IVc 26.33 ± 0.39 IVb 1.97 ± 0.06
Palmitic Acid
(C16:0)
Va,c 26.40 ± 0.52
Arachidic Acid
(C20:0)
Vc 1.82 ± 0.10
Ia 3.16 ± 0.24 Ia 14.51 ± 0.27 IIb 1.85 ± 0.02 IIb 17.94 ± 0.33 IIIc 2.49 ± 0.33 IIIc 10.20 ± 0.57 IVd 2.93 ± 0.14 IVd 11.28 ± 0.38
Palmitoleic Acid
(C16:1 n-7)
Vd 2.86 ± 0.10
Linoleic Acid (C18:2 n-6)
Vc,d 10.72 ± 0.42
Ia 42.90 ± 0.62 Ia 0.44 ± 0.03 IIb 40.04 ± 0.20 IIb 0.92 ± 0.05 IIIc 47.44 ± 1.22 IIIc 0.27 ± 0.03 IVd 45.74 ± 0.67 IVa 0.39 ± 0.12
Oleic Acid (C18:1 n-9)
Vc,d 46.73 ± 0.74
γ-Linolenic Acid
(C18:3 n-6)
Vc 0.30 ± 0.02
45
†Mean (g fatty acid / 100 g total lipid) ± standard deviation (n = 12)
* Diet I – standard corn/soy-laying ration with 0% added fat/oil; Diet II – Standard corn/soy
laying ration with 4% added soybean oil; Diet III – Standard corn/soy laying ration with 4%
added tallow; Diet IV - Standard corn/soy laying ration with 4% added shortening; Diet V -
Standard corn/soy laying ration with 4% added tallow/shortening blend
.a,b,c - Diets with different superscripts were significantly different (p<0.05) as determined by
Tukey’s test
46
Table 4: Percentage of fatty acids in egg yolk lipids from hens consuming diets with varying levels of trans-fat
Fatty acid Diet* % Fatty acid† Fatty acid Diet* % Fatty acid†
Ia 0.00 ± 0.00 Ia 46.07 ± 0.42
IIa 0.00 ± 0.00 IIb 41.63 ± 0.54
IIIb 0.68 ± 0.05 IIIc 49.94 ± 0.97
IVb 0.81 ± 0.05 IVd 48.66 ± 0.63
Trans Fatty Acids
Vb 0.76 ± 0.04
Monounsaturated Fatty Acids
Vc 49.65 ± 0.73
Ia 38.90 ± 0.49 Ia 14.95 ± 0.28
IIb 38.54 ± 0.32 IIb 19.72 ± 0.75
IIIa,b 37.98 ± 0.21 IIIc 10.47 ± 0.60
Ivb 38.23 ± 0.39 IVd 11.67 ± 0.41
Saturated Fatty Acids
Vb 38.26 ± 0.73
Polyunsaturated Fatty Acids
Vc,d 11.01 ± 0.41
47
†Mean (g fatty acid / 100 g total lipid) ± standard deviation (n = 12)
* Diet I – standard corn/soy-laying ration with 0% added fat/oil; Diet II – Standard corn/soy laying ration with 4% added soybean oil;
Diet III – Standard corn/soy laying ration with 4% added tallow; Diet IV - Standard corn/soy laying ration with 4% added shortening;
Diet V - Standard corn/soy laying ration with 4% added tallow/shortening blend
.a,b,c - Diets with different superscripts were significantly different (p<0.05) as determined by Tukey’s test
48
Table 5: Fatty acid composition of feeds with and without added trans-fats
Feeds*
Fatty Acid Diet I Diet II Diet III Diet IV Diet V
Myristic acid (C14:0) 0.00 0.00 2.04 2.11 2.18
Palmitic acid (C16:0) 15.85 12.92 21.86 22.00 22.20
Palmitoleic acid (C16:1 n-7) 0.09 0.00 1.81 1.70 1.84
Stearic acid (C18:0) 2.78 3.79 11.57 12.39 12.22
Oleic acid (C18:1 n-9) 21.09 20.53 29.56 27.55 29.15
Linoleic acid (C18:2 n-6) 54.24 53.44 25.06 25.57 24.68
Arachidic acid (C20:0) 0.00 0.00 0.00 0.00 0.00
Trans fatty acid isomers (4t-16t C18:1)
0.00 0.00 2.78 3.25 2.85
* Diet I – standard corn/soy-laying ration with 0% added fat/oil; Diet II – Standard corn/soy laying ration with 4% added soybean oil; Diet III – Standard corn/soy laying ration with 4% added tallow; Diet IV - Standard corn/soy laying ration with 4% added shortening; Diet V - Standard corn/soy laying ration with 4% added tallow/shortening blend
49
50
Table 6: Recovery of triheptadecanoin and methyl heptadecanoate during GC-FID analysis
Lipid Type Amount Injected (ng) Mass Recovered (ng) Percent Recovered Peak area
FAME* 600 600.07 100.01 370.70 FAME 300 298.82 99.61 183.40 FAME 150 151.59 101.06 91.87 FAME 75 74.55 99.40 43.97 TAG† 600 478.69 79.78 295.23 TAG 300 245.63 81.88 150.33 TAG 150 122.91 81.94 74.03 TAG 75 61.31 81.74 35.73
*FAME – Fatty Acid Methyl Ester
†TAG - Triacylglycerol
51
Table 7: Precision and accuracy of trans-fat analysis using GC-FID
Concentration added* Intra-day (μg/mL) Observed Precision† Accuracy‡
160 151.85 ± 1.31 0.8640 -5.0939 320 309.47 ± 4.25 5.5438 -3.2919 520 548.08 ± 1.51 1.1098 5.4002 Concentration added * Inter-day (μg/mL) Observed Precision† Accuracy‡ 160 142.34 ± 12.31 8.6445 -11.0345 320 304.92 ± 18.76 6.1513 -4.7124 520 517.89 ± 25.46 4.9158 -0.4066
*Concentration of triheptadecanoin in CS2
†Precision as percent relative standard deviation (%RSD)
%RSD = 100 * standard deviation / mean
‡Accuracy as percent relative error
Percent relative error = 100 * (observed – added) / added
52
Table 8: Precision and accuracy of trans-fat analysis using FTIR-ATR
Intra-day
%Methyl Elaidate* Observed Precision† Accuracy‡
5.11 4.91 ± 0.01 0.2139 -3.8990
7.88 7.56 ± 0.01 0.1377 -4.0812
Inter-day
%Methyl Elaidate* Observed Precision† Accuracy‡
5.11 4.97 ± 0.02 0.3084 -2.6791
7.88 7.70 ± 0.02 0.3088 -2.2632 *Methyl elaidate in methyl oleate
†Precision as percent relative standard deviation (%RSD)
%RSD = 100 * standard deviation / mean
‡Accuracy as percent relative error
Percent relative error = 100 * (observed – added) / added
53
List of Figures
Figure 1: Partial hydrogenation reaction mechanism – Figure 1 shows the cis/trans
isomerization, double bond migration, and saturation reactions during partial hydrogenation of
oils.
54
55
Figure 2: GC-FID Chromatogram of egg yolk lipid from diet I
1.Myristic Acid, 2. Palmitic Acid, 3. cis-9-Hexanoic acid, 4. Palmitoleic Acid 5. Heptadecanoic
Acid (IS), 6. Stearic Acid, 7. Oleic Acid, 8. cis-11-Octadecenoic Acid, 9. Linoleic Acid, 10.
Arachidic Acid
56
57
Figure 3: GC-FID Chromatogram of egg yolk lipid from diet II
1.Myristic Acid, 2. Palmitic Acid, 3. cis-9-Hexanoic acid, 4. Palmitoleic Acid 5. Heptadecanoic
Acid (IS), 6. Stearic Acid, 7. Oleic Acid, 8. cis-11-Octadecenoic Acid, 9. Linoleic Acid, 10.
Linolenic Acid, 11. Arachidic Acid
58
59
Figure 4: GC-FID Chromatogram of egg yolk lipid from diet III
1.Myristic Acid, 2. Palmitic Acid, 3. cis-9-Hexanoic acid, 4. Palmitoleic Acid 5. Heptadecanoic
Acid (IS), 6. Stearic Acid, 7. Trans Fatty Acids 8. Oleic Acid, 9. cis-11-Octadecenoic Acid, 10.
Linoleic Acid, 11. Arachidic Acid
60
61
Figure 5: GC-FID Chromatogram of egg yolk lipid from diet IV
1.Myristic Acid, 2. Palmitic Acid, 3. cis-9-Hexanoic acid, 4. Palmitoleic Acid 5. Heptadecanoic
Acid (IS), 6. Stearic Acid, 7. Trans Fatty Acids 8. Oleic Acid, 9. cis-11-Octadecenoic Acid, 10.
Linoleic Acid, 11. Arachidic Acid
62
63
Figure 6: GC-FID Chromatogram of egg yolk lipid from diet V
1.Myristic Acid, 2. Palmitic Acid, 3. cis-9-Hexanoic acid, 4. Palmitoleic Acid 5. Heptadecanoic
Acid (IS), 6. Stearic Acid, 7. Trans Fatty Acids 8. Oleic Acid, 9. cis-11-Octadecenoic Acid, 10.
Linoleic Acid, 11. Arachidic Acid
64
65
Figure 7: GC-FID Chromatogram of diet I
1.Palmitic Acid, 2. Heptadecanoic Acid (IS), 3. Stearic Acid, 4. Linoleic Acid, 5 Oleic Acid, 6.
Linolenic Acid.
66
67
Figure 8: GC-FID Chromatogram of diet II
1.Palmitic Acid, 2. Heptadecanoic Acid (IS), 3. Stearic Acid, 4. Linoleic Acid, 5 Oleic Acid, 6.
Linolenic Acid
68
69
Figure 9: GC-FID Chromatogram of diet III
1. Myristic Acid, 2. Palmitic Acid, 3. Palmitoleic Acid 4. Heptadecanoic Acid (IS), 5. Stearic
Acid, 6. Trans Fatty Acids 7. Oleic Acid, 8 Linoleic Acid, 9. Linolenic Acid
70
71
Figure 10: GC-FID Chromatogram of diet IV
1. Myristic Acid, 2. Palmitic Acid, 3. Palmitoleic Acid 4. Heptadecanoic Acid (IS), 5. Stearic
Acid, 6. Trans Fatty Acids 7. Oleic Acid, 8 Linoleic Acid, 9. Linolenic Acid
72
73
Figure 11: GC-FID Chromatogram of diet V
1. Myristic Acid, 2. Palmitic Acid, 3. Palmitoleic Acid 4. Heptadecanoic Acid (IS), 5. Stearic
Acid, 6. Trans Fatty Acids 7. Oleic Acid, 8 Linoleic Acid, 9. Linolenic Acid
74
75
Figure 12: Overlay of silver ion-TLC chromatogram and original chromatogram for shortening.
Figure 2 shows (1) the resulting chromatogram of silver ion-TLC isolation of trans fatty acids
for shortening and the full fatty acid profile in the same region.
76
77
Figure 13: Chromatogram of silver ion-TLC separation of trans monounsaturated fraction of
shortening
1. 6t-10t C18:1, 2. 11t C18:1, 3. 12t C18:1, 4. 13t-14t C18:1, 5. 15t C18:1
78
79
Figure 14: Interference band of egg lipids during FTIR analysis - Egg lipids were
transmethylated to remove the interference band at 967 cm-1 seen in (2) extracted egg lipids
compared to the spectra of (1) prepared FAMEs of extracted egg lipids.
80
81
Figure 15: Second derivative spectra of absorbance of trans band at 967 cm-1 - Prepared
standards of 0.5 to 10% methyl elaidate in methyl oleate. (1) 10; (2) 9; (3) 8; (4) 7; (5) 6; (6) 5;
(7) 4; (8) 3; (9) 2; (10) 1; (11) 0.5% methyl elaidate.
82
A
bsor
banc
e (2
nd d
eriv
ativ
e)
Wavenumbers (cm-1)
83
Figure 16: Spectra of absorbance of trans band at 967 cm-1 - Prepared standards of 0.5% to
10% methyl elaidate in methyl oleate. (1) 10%; (2) 9%; (3) 8%; (4) 7%; (5) 6%; (6) 5%; (7) 4%;
(8) 3%; (9) 2%; (10) 1%; (11) 0.5% methyl elaidate.
84
85
Figure 17: TQ Analyst calibrations of prepared standards of 0.5% to 10% (w/w) methyl elaidate
in methyl oleate - Comparison of TQ Analyst calibration curve of trans standards using the
second derivative of absorbance versus using absorbance.
86
87
Figure 18: Correlation of GC versus FTIR data – Fig. 17 shows a strong correlation between the
results obtained by GC-FID compared with FTIR
88
y = 1.2735x - 1.8486R2 = 0.9759
0
1
2
3
4
5
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9
g trans -fat / 100 g total fat by FT-IR
g tra
ns-fa
t / 1
00 g
tota
l fat
by
GC
-FID
89