EVALUATION OF AN AUTOMATED HYDROLYSIS AND EXTRACTION METHOD FOR

QUANTIFICATION OF FAT IN CEREAL FOODS

by

JACLYN ELLIS ROBINSON

(Under the Direction of Sandra E. Kays)

ABSTRACT

The utility of an automated acid hydrolysis and extraction (AHE) system was evaluated

for extraction of fat for the quantification of total, saturated, polyunsaturated, monounsaturated,

and trans fat in cereal products. Oil extracted by the AHE system was assessed: gravimetrically

for total fat; by capillary GC analysis for total fat, lipid classes, and trans fat; and by ATR-FTIR

spectroscopy for quantification of trans fat. Results were compared with parallel determinations

using the standard AOAC Method 996.01. For gravimetric and gas chromatographic evaluations

the AHE system results were equivalent to the AOAC 996.01 results (α = 0.01), indicating that

the AHE system can be used to measure total fat, lipid classes, and trans fat with sufficient

accuracy for nutrition labeling. The AHE and AOAC results were not equivalent for all samples

assessed by ATR-FTIR spectroscopy, thus for this application the AHE system may be better

suited for rough screening.

INDEX WORDS: total fat; trans fat; nutrition labeling; Soxtec®; cereal products; ATR-

FTIR; AOAC 996.01

EVALUATION OF AN AUTOMATED HYDROLYSIS AND EXTRACTION METHOD FOR

QUANTIFICATION OF FAT IN CEREAL FOODS

by

JACLYN ELLIS ROBINSON

B.S., Florida A & M University, 2004

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

2006

© 2006

Jaclyn Ellis Robinson

All Rights Reserved

EVALUATION OF AN AUTOMATED HYDROLYSIS AND EXTRACTION METHOD FOR

QUANTIFICATION OF FAT IN CEREAL FOODS

by

JACLYN ELLIS ROBINSON

Major Professor: Sandra Kays

Committee: Rakesh Singh Louise Wicker

Electronic Version Approved: Maureen Grasso Dean of the Graduate School The University of Georgia December 2006

DEDICATION

To all who are a part of my extensive support network…WE did it!

iv

ACKNOWLEDGEMENTS

So many people have contributed to the completion of this document, it would be

remiss of me not to acknowledge them. I have been extremely lucky to have Dr. Sandra

Kays as my major professor. Dr. Kays has been all things and more I needed from a

mentor, and working with her has been an honor. Her guidance, advice, kindness,

encouragement, and training have been invaluable throughout this process. I, too, deeply

value the encouragement, technical instruction, and patience (in answering all of the a

million and one questions I asked) of Yookyung Kim and Michelle Huk. It’s been an

absolute pleasure working together. I would like to thank Dr. Rakesh Singh, and

everyone in his lab for always making me feel welcome, especially Dr. Nepal Singh for

his instruction in using ATR-FTIR. I am also grateful for Dr. Dave Himmelsbach; his

input was extremely valuable in the interpretation of FTIR results. Dr. Louise Wicker

has truly been an inspiration to me. My entire graduate school experience would not

have been the same if it were not for her guidance, assistance, and advice.

My husband, parents, brother, family, and special friends have always believed in

me and supported me in whatever I’ve tried to do in life. They delight in my successes

and anguish in any slight misfortune that may come my way. They are my laughter in

misery, direction in confusion, and angels in desperation. Their unconditional love is

priceless and the greatest gift one could ever have.

Finally, the apostle Paul wrote in Phillipians 4:13, “I can do all things through

Christ who strenghthens me.” This quote encompasses the story of my past, present, and

future; I know my strength, and peace come solely from my Lord and Savior Jesus

Christ. He is my motivation; all I ever am and all I will ever be; and to Him I owe my all.

v

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS.............................................................................................................v

LIST OF TABLES....................................................................................................................... viii

LIST OF FIGURES ....................................................................................................................... ix

CHAPTER

1 INTRODUCTION AND LITERATURE REVIEW .....................................................1

1. Introduction ...........................................................................................................2

2. Lipids and processed cereal foods.........................................................................3

3. Nutrition labeling regulation of fat content in cereal products ...........................13

4. Quantitation of lipid content of cereal products ..................................................18

5. Objective .............................................................................................................32

6. References ...........................................................................................................34

2 EVALUATION OF AN AUTOMATED HYDROLYSIS AND EXTRACTION

METHOD FOR QUANTIFICATION OF FAT IN CEREAL FOODS..................45

1. Abstract ...............................................................................................................46

2. Introduction .........................................................................................................47

3. Materials and methods.........................................................................................49

4. Results .................................................................................................................56

5. Discussion ...........................................................................................................57

6. References ...........................................................................................................61

3 SUMMARY AND CONCLUSIONS ..........................................................................70

vi

APPENDICES ...............................................................................................................................74

A EXPERIMENTAL CEREAL FOOD PRODUCTS.....................................................75

B MACRONUTRIENT COMPOSITION OF EXPERIMENTAL CEREAL

PRODUCTS BASED ON NUTRITION LABEL DECLARATIONS...................76

C LIST OF INGREDIENTS FOR EXPERIMENTAL CEREAL PRODUCTS BASED

ON NUTRITION LABEL DECLARATIONS.......................................................77

D FLOW CHART FOR AOAC 996.01 DETERMINATION OF TOTAL FAT, LIPID

CLASSES, AND TRANS FAT................................................................................79

E FLOW CHART FOR THE AHE DETERMINATION OF TOTAL FAT, LIPID

CLASSES, AND TRANS FAT................................................................................80

F EQUATIONS FOR TOTAL FAT, LIPID CLASSES, AND TRANS FAT

DETERMINATION................................................................................................81

G THEORETICAL CONVERSION FACTORS FOR THE DETERMINATION OF

EACH FATTY ACID EXPRESSED AS TRIGLYCERIDES AND/OR FATTY

ACID .......................................................................................................................83

H STABILITY OF FAT IN GROUND PRODUCTS STORED AT -20 °C FOR 7

DAYS DETERMINED BY MODIFIED AOAC METHOD 996.01......................84

I STABILITY OF FAT IN FAME STORED AT -20 °C FOR 7 DAYS DETERMINED

BY MODIFIED AOAC METHOD 996.01 ............................................................85

J STABILITY OF COMPONENTS OF EXTRACTED OIL STORED AT -20 °C FOR

4 DAYS DETERMINED BY MODIFIED AOAC METHOD 996.01...................86

vii

LIST OF TABLES

Page

Table 2.1: Cereal products used in the study .................................................................................63

Table 2.2: Measurement of total fat (%) in cereal products by the standard GC method and by an

automated gravimetric method......................................................................................64

Table 2.3: Determination of total, saturated, polyunsaturated, and monounsaturated fat (%)

extracted from cereal products by the standard method and an automated method......65

Table 2.4: Determination of trans fat (%) in oil extracted from cereal products by the standard

method and an automated method.................................................................................67

Table 2.5: ATR-FTIR determination of trans fat (%) in oil extracted from cereal products by the

standard method and an automated method ..................................................................68

viii

LIST OF FIGURES

Page

Figure 1.1: Stucture of stearic acid ................................................................................................41

Figure 1.2: Cis and trans geometric isomers of C18:1 fatty acid ..................................................42

Figure 1.3: Soxhlet apparatus ........................................................................................................43

Figure 1.4: The AHE system .........................................................................................................44

Figure 2.1: ATR-FTIR spectra of oil extracted from a medium fat cereal product by the standard

method (A) and an automated method (B)....................................................................69

ix

CHAPTER 1

INTRODUCTION AND LITERATURE REVIEW

1

1. Introduction

The Nutrition Labeling and Education Act (NLEA) regulates nutrition labeling, ensuring

that declarations are presented accurately and uniformly (Summers, 2003). For the NLEA to be

effective, accurate and repeatable methods are required for the determination of nutrients in

foods. AOAC 996.01 is a universally accepted method for the determination of total, saturated,

polyunsaturated, monounsaturated, and in a modified form, trans fat in cereal products with

sufficient accuracy to satisfy NLEA regulations (Ratnayake, 2004, Ngehngwainbi et al., 1997).

It involves the hydrolyzation of ground cereal foods, extraction of lipids using diethyl and

petroleum ether solvents, evaporation of solvents, saponification and methylation of fatty acids,

and capillary GC quantification of fatty acid methyl esters (FAME) (AOAC, 2000c). Although

AOAC 996.01 is accurate and reliable, the analysis is labor-intensive and time-consuming, and

involves work with caustic and flammable chemicals, many of which require specific disposal.

An automated hydrolysis and extraction (AHE) system that is available commercially,

offers an alternative to manual hydrolysis and extraction methods, such as AOAC 996.01. The

method involves a combination of automated acid hydrolysis and rinsing of the sample in a

closed system followed by reflux boiling and automated Soxhlet extraction of the lipid, also in a

closed system (de Castro et al., 1998). The percentage of total lipid is obtained gravimetrically.

In addition, the extracted fat can be recovered and total fat and lipid classes measured by

capillary GC analysis as in AOAC method 996.01. Because the AHE system is an automated

and closed system: the operator has less contact and exposure to solvents and fumes; the

operator’s time and attention may be directed towards other activities during extraction; and the

results are less likely to be affected by operator error (Helaleh et al., 2005a). Six samples can be

analyzed simultaneously with one unit. Less solvent overall is consumed using the AHE system;

2

in fact, 80% of the solvent can be recovered and reused (Anonymous, 2006). The design of the

AHE hydrolyzation unit provides for the rinsing of non-lipid aqueous moieties from the

hydrolyzed sample, removing elements that might, otherwise, cause overestimation of

gravimetric total lipid. In theory, this should provide for the accurate determination of total lipid

without use of a gas chromatographic step. Recovery of the lipid, and subsequent saponification

and methylation, allow for determination of total, saturated, polyunsaturated, monounsaturated,

and trans fat by capillary GC analysis. The extracted lipid can also be recovered and analyzed

by ATR-FTIR spectroscopy for the determination of trans fatty acids.

The accuracy of the AHE system for extraction of lipids for the analysis of total fat and

lipid components compared to the extraction of lipids by AOAC Method 996.01 has not been

reported. Thus, its potential for analysis of lipids for nutrition labeling and monitoring is

unknown.

The objective of this study is to evaluate the AHE system for the determination of total

fat gravimetrically and for the extraction of lipid for the capillary GC determination of total,

saturated, polyunsaturated, monounsaturated and trans fat. A diverse range of cereal products

with a wide range of total fat and trans fat were used for the study, and the results were evaluated

against those using AOAC Method 996.01 as the standard. In addition, the AHE system was

evaluated for the extraction of lipid for Fourier Transform Infrared (FT-IR) determination of

trans fat.

2. Lipids and processed cereal products

2.1. History of cereals

Thousands of years ago grains were the main energy and protein source for humans. In

fact, the word “cereal” is derived from the name for the Greek goddess of the earth, Demeter

3

(Roman Ceres) (Dendy, 2001). To the ancients, food sprouting from the ground appeared to be a

magical and quite mysterious phenomenon; the ancient Greeks believed Demeter revealed to

humanity the secrets of sowing and plowing. The development of farming and animal

domestication were two extremely important events in the history of mankind, enabling people to

abandon nomadic lifestyles and form stable, fixed communities, freeing up more time for the

pursuit of other knowledge. The first farming communities were located between the Tigris and

Euphrates rivers, a region near the modern day country of Iraq. Since this time, cereals have

remained a major crop worldwide (Dendy, 2001).

2.2. The types and morphology of the cereal grain

Cereals are grains or edible seeds of plants belonging to the monocot, angiosperm

Gramineae family or “grass” family. Cereals come from two types of plants: grassy plants and

canes. Wheat (Triticum aestivum L.), rye (Secale cereale L.), oats (Avena sativa L.), rice (Oryza

sativa L.), and barley (Hordeum vulgare L.) are of the grassy type; and maize (Zea mays L.

subsp. mays), sorghum [Sorghum bicolor (L.) Moench], and millets (Pennisetum spp.) are of the

cane type. Cereals are grown all over the world in a variety of climates; sorghum and millets

grow in semiarid climates; deep water rice grows under water; rice and millets grow in very hot

climates; and rye and oats can grow in cold climates. Pseudocereals are grains that are

harvested, prepared, and consumed similarly to cereals, but originate from plants other than

grasses. Amaranth (Amaranthus hypochondriacus L.), buckwheat (Fagopyrum esculentum

Moench), and quinoa (Chenopodium quinoa Wild.) are examples of pseudocereals (Dendy,

2001).

Cereal grains, or caryopses, consist of a seed covered by a fruit coat or pericarp. The

pericarp is enclosed by an outer layer called the testa or seed coat. The germ (or embryo) and the

4

endosperm (or embryo food source) lie in the interior of the grain, beneath the pericarp and testa.

The caryopses may be covered by a husk or glume (Dendy, 2001).

2.3. Nutritional value of cereals

For the majority of the earth’s population, cereals and cereal-based foods are a major

source of energy and nutrients. Cereals supply approximately forty different necessary nutrients

(Ranhotra, 1991). Whole grain cereal products are often associated with low calorie and low fat

contents and are recognized as a rich source of complex carbohydrate, dietary fiber, and certain

vitamins. Soluble fibers found in cereal grains such as oats and barley are widely considered to

be functional, reducing low density lipoprotein (LDL) cholesterol, and thereby risk of coronary

heart disease (Lane, 2001; Hasler, 1998).

Processing practices commonly applied to many cereals often alter the grain’s original

nutritive value. Milling wheat removes the outer layer and germ of the grain, collectively known

as the bran. While milling removes highly oxidizable lipids and increases digestibility, the

nutrients and fiber of the bran are also removed. In addition, cooking practices, time, pH,

moisture, light, oxygen, enzymes, oxidants and additives all may alter the original nutrient

content of the grain (Ranhotra, 1991).

2.4. Consumption of cereals

In almost every region of the world, cereals are a staple food. Cereals may be eaten

whole, in porridges, in leavened and unleavened breads, processed into snack and breakfast

products, used as sources for starch and glucose, or fermented into beverages, such as beer and

spirits (Dendy, 2001).

Americans consumed, on average, 200 pounds of flour and cereal products in 2000

compared to 147 pounds/year and 135 pounds/year in the early eighties and seventies,

5

respectively (Putnam, 2002). Because of the high consumption of cereals and cereal products,

cereals have often been used as enrichment vehicles, the latest nutrient enrichment being folate

in 1998 (Lane, 2001). With a decline in popularity of low carbohydrate diets, consumption of

cereal and bakery products is expected to increase (USDA, 2006).

2.5. Lipid content of cereals

Lipids are a minor constituent of cereals. The ether extractable lipids, i.e., free lipids or

crude fat, range from 1.5-2% for whole grain barley, rice, rye, triticale, and wheat and from 3-7%

by weight for whole grain oats, millets, corn, and sorghum. The germ is the part of the grain

highest in lipids, even though it constitutes only a small part of the entire kernel (Chung, 1991).

Cereal lipids can be broken into two classes: polar lipids—such as glycolipids and

phospholipids—and nonpolar lipids—such as steryl esters, long chain free fatty acids, and

mono-, di-, and tri- glycerides. All cereals contain more nonpolar lipids than polar. Most fats in

cereals are found as esters of glycerol or its derivatives; however, very small portions are esters

of sterols, carotenoids, and tocopherols. All cereal grains are high in unsaturated fatty acid

content, with linoleic (18:2) acid being the major unsaturated fatty acid and palmitic (16:0) being

the predominant saturated fatty acid (Chung, 1991). The major fatty acids in cereals are 16:0,

16:1, 18:0, 18:1, 18:2, and 18:3 (Becker, 2000).

2.6. Commercially available cereal products

Cereal grains are processed into an extremely diverse range of products around the world.

Examples of processed cereal products available in the United States are breakfast cereals,

breads, cookies, cakes, pastries, donuts, pastas, popcorn, rice snacks, granola, pretzels, and

baking mixes. They consist of whole grains, processed grains, or flours, and can include

numerous ingredients such as eggs, sugar, fat, and dairy products.

6

A number of processing methods are employed in the manufacture of cereal foods. Many

contain one or more cereal flours obtained by grinding of dried, milled—and in the case of

barley—malted grains. Numerous cereal products, such as donuts, corn chips and rice-based

snacks, are fried; while others, such as, cookies, cakes, crackers, breads, pastries, and pretzels are

baked. Some breakfast cereals and many snack products are extruded, with or without

expansion. Pastas are generally dried before packaging, and boiled directly prior to

consumption. The more highly processed a cereal product is, the lower it tends to be in

nutritional value, often being high in fat, sugar and calories (Ranhotra, 1991). Some breakfast

cereal products, although highly processed, are fortified with vitamins and minerals, as indicated

in the Nutrition Facts labels of the products.

The wide variety of additives used in cereal products greatly contributes to the diversity

of the food group. Additives can include fruit, fruit juices, and essential oils; dehydrated

vegetables like onions, peppers, and potatoes; molasses, honey, high fructose corn syrup, and

noncaloric sweeteners; nuts and nut butters; annatto and beet juice (as coloring agents); cheeses;

chocolate; spices, such as paprika, cinnamon, cardomon, fennel, nutmeg, and tumeric;

antioxidants and preservatives (to help extend the shelf life of the product); and gums and

thickening agents.

2.7. Lipids in processed cereal products

The vast majority of fats present in processed cereal products are added fats. Although

the total fat content of cereal grains is very low, the percentage total fat for processed cereal

products varies widely, ranging from 0-40%. Solid fats added to cereal foods are called

shortenings, and liquid fats are called oils (Becker, 2000). Shortenings may be a mixture of both

animal and vegetable fats and oils (Dendy, 2001; Matz, 1991). Lipids in cereal products can be

7

naturally occurring from plant or animal origins or can be hydrogenated or partially

hydrogenated oils.

2.7.1. Animal and vegetable lipids

Some of the major shortenings and oils used as ingredients in cereal foods are soybean

oil, peanut oil, canola oil, safflower oil, cottonseed oil, corn oil, sunflower oil, palm oil, palm

kernel oil, coconut oil, butter, margarine, and lard or beef tallow (Matz, 1991). Soybean,

safflower, sunflower, cottonseed, corn, and peanut oils are all high in unsaturated fatty acids

(>75%) and low in saturated fat. In addition canola, peanut, and soybean oil are rich in

monounsaturated fat (>35%). Coconut oil, butter, palm oil and lard or beef tallow have high

saturated fatty acid content (40-85%) (Wardlaw et al., 2007). Soybean oil (unhydrogenated) and

cottonseed oil are high in polyunsaturated fatty acids, about 53% of which are linoleic acid

(18:2). Sunflower oil and canola oil are both, also, high in unsaturated fatty acids, with linoleic

(18:2) and oleic (18:1) being the predominant fatty acids, respectively. About 44% of palm oil is

saturated palmitic acid (16:0), while coconut oil and palm kernel oil are high in both lauric (12:0)

and myristic (14:0) saturated fatty acids (Gunstone, 2002).

2.7.2. Hydrogenated lipids

During the seventies, eighties, and nineties the food industry in the United States replaced

fats high in saturated fatty acids—such as butter, lard, and palm oil—in many products with

more hydrogenated vegetable oils, which were lower in saturated fatty acids and cholesterol

(Canning, 2005). Apart from yielding products with a seemingly healthier appeal, hydrogenated

oils had a longer shelf-life, higher melting points, and better flavor stability than butter, palm oil,

and lard (Canning, 2005; Craig-Schmidt, 2000).

8

Oils are fully or partially hydrogenated by bubbling hydrogen through the oil at elevated

temperature in the presence of a catalyst, such as nickel, and in the absence of oxygen. Complete

hydrogenation results in saturated fatty acids (Craig-Schmidt, 2000). Partial hydrogenation may

shift double bonds found in unsaturated fatty acids, forming positional isomers, or may transform

naturally occurring cis-configurated double bonds to more thermodynamically stable trans-

double bonds, forming geometric isomers (Matz, 1991). Trans isomers have higher melting

points and are less reactive and, therefore, more stable to oxidative stress, than cis isomers

(Matz, 1991; Craig-Schmidt, 2000). Thus, hydrogenation raises a fat’s melting point yielding

fats that are semisolid at room temperature, a highly desirable trait for blending and creaming for

many cookies, crackers, snacks, and other baked goods (Dendy, 2001). Soybean and cottonseed

oils are the most common vegetable oils chosen as raw ingredients for the manufacture of

hydrogenated shortenings and are the most commonly occurring partially hydrogenated oils in

cereal products (Matz, 1991).

The partial hydrogenation of fats results in isomeric fatty acids. Overall the most

commonly consumed trans fatty acid in human diets is 18:1(t) (Craig-Schmidt, 2000). The

occurance of trans fats in cereal products containing partially hydrogenated oil can vary from 0-

15 % of the product and 0.2-43 % of the lipids in the product (Kim et al., unpublished data). The

overwhelming majority, about 90 to 95%, of all isomeric fatty acids consumed in the United

States originates from the commercial hydrogenation of oils. Biohydrogenation, from the actions

of microorganisms in the gut of ruminant animals, also results in a small percentage of naturally

occuring isomeric fatty acids (Matz 1991). For example, 4-8% of fatty acids in milk are trans-

fatty acids resulting from biohydrogenation (Gunstone, 2002).

9

With mandatory labeling of trans fat in affect as of January 1, 2006 (CFR, 2006), many

food companies have attempted to reformulate products formally containing trans fatty acids to

contain reduced amounts or no trans fatty acids. The efforts to reduce trans fatty acids in

products have been focused primarily in four directions. The first is the modification of

hydrogenation to reduce the amount of trans geometric isomers produced. The second makes

use of biotechnology to modify plant oils to have fatty-acid compositions more stable to

oxidative stress. The third is blending liquid oils with high solid oil fractions from palm, palm

kernel, and coconut oils to generate shortenings and margarines. Finally, the fourth approach

involves the interesterification of fatty acids, a process that rearranges fatty acids bonded to the

glycerol backbone to change the melting and crystallization properties of the fats. The

interesterified oil is then blended with a fully hydrogenated oil. Products reduced in trans fatty

acids using the latter method may still have partially hydrogenated oil on the food label but will

contain no trans fatty acids (Hunter, 2005; Tarrago-Trani, 2006).

2.8. Functions of added fat in cereal products

Added fats in cereal products greatly influence the taste, and texture of the finished

product. The main functions of fat in processed cereal products are to provide a short, rich,

tender texture, impart structure, and facilitate proper flavor release (Faridi, 1991; Bakal, 1991).

In cakes, fats influence the texture of the finished product by assisting the dispersion of small

pockets of air throughout the aqueous batter during mixing; the air pockets later act as centers for

expansion during baking, a phenomenon that strongly contributes to the texture of the baked

cake. In puff doughs, firm fats, such as lard, tallow, hydrogenated and partially hydrogenated

vegetable oils, butter, and/or margarine, are thinly deposited between dough, creating layers that

separate and expand during baking, making for a pastry with a crispy, flaky texture (Faridi, 1991;

10

Gunstone, 2002). Fats lubricate gluten in yeast-raised dough, lubricate slicing blades during

slicing of breads, and act as emulsifiers, thereby, stabilizing, and softening the texture of foods

(Bakal, 1991). In addition to being major ingredients, partially hydrogenated and fractionated

soybean and cottonseed oils are used as coating sprays for crackers, doughnuts, ice cream cones,

wafers and other cereal products, providing a glossy finish, a moisture barrier for extended shelf-

life, lubrication and anticaking properties. (Gunstone, 2002; Faridi, 1991).

The functionality of fats in foods is greatly dependent on the physical properties of the

fat. The physical properties are determined by the surrounding temperature, the length of the

longest carbon chain, and the degree of unsaturation (Matz, 1991; Faridi, 1991). The fats used in

many formulations must be semisolid at room temperature (20 to 25°C), in order to cream well

during mixing with other ingredients and maintain structural integrity; yet, must have little solid

fat above body temperature (37°C), in order to avoid leaving an oily or waxy residue in the

mouth and provide proper flavor release (Faridi, 1991).

Frying oils tend to have different desirable qualities from fats and oils used solely as

ingredients. Frying oils are unique in that they are heat-transferring processing agents, as well as

contributors to flavor, texture, and structure. High quality frying oils will have a high smoke

point and be resistant to hydrolysis and oxidation at high temperatures (Matz, 1991). This is

essential, because the heating of fats may cause an acceleration in degradation of fatty acids,

which in turn may negatively alter taste, functionality, and nutrition (Becker, 2000).

2.9. Consumption of fatty acids and health

The body needs, but is not able to produce, oleic acid (C18:1, n-6) and linoleic acid

(C18:2, n-3). They are known as essential fatty acids and must be acquired from the diet. It is

11

recommended that they make up approximately 3% of a days total caloric intake; most diets in

the United States exceed this requirement.

A lifestyle including regular exercise and a diet low in total fat intake is a healthy way to

control weight and prevent obesity, a factor that is a risk for many diseases, namely

cardiovascular disease, hypertension, diabetes mellitus, gall bladder disease, and certain types of

cancer (Hausman, D., 2002; Jequier, 2002). To help achieve this goal, the American Heart

Association (Krauss et al., 2000) recommends that fat intake be limited to 20-30% of total

energy intake with no more than 7-10% of that energy coming from saturated fat and trans fat

combined.

In more recent years, the quality of fat consumed rather than the quantity has been

implicated as more relevant to the risk of developing chronic diseases, such as coronary heart

disease (CHD) (Ascherio, 2002). For quite some time the public has been aware that increased

consumption of saturated fatty acids tends to increase plasma low density lipoprotein cholesterol

(LDL), a phenomenon that has been associated with increased risk for heart disease. However,

increased consumption of saturated fatty acids also raises beneficial high density lipoprotein

(HDL) cholesterol, although not enough to counteract the increase in LDL cholesterol (Bruckner,

2000; Canning, 2005). Myristic (14:0), lauric (12:0), and palmitic (16:0) are the saturated fatty

acids associated with the most hypercholesterolemic effects, with myristic being perhaps the

most nocuous. Stearic acid (18:0, Fig. 1.1.) is considered to be neutral in the way it affects

plasma cholesterol levels (Mensink, 2002). Americans consume approximately 12-14%

saturated fatty acids as a percentage of total caloric intake (Hunter, 2005).

Consumption of cis-polyunsaturated fatty acids lowers all classes of lipoprotein, LDL,

HDL, and VLDL (very low density lipoprotein) (Mensink, 2002; Sacks, 2002). Unsaturated

12

trans fatty acids, in contrast to cis-unsaturated fatty acids, have been shown to increase LDL and

decrease HDL, an effect worse than that of saturated fatty acids (Hunter, 2005; Sacks, 2002;

Bruckner, 2000; Hunter, 2000; Mensink, 2002). Trans fatty acids have a linear structure (Fig.

1.2.) more closely resembling that of saturated fatty acids (Hunter, 2005; Craig-Schmidt, 2000).

It has been estimated that Americans consume 2-3% trans fatty acids as a percentage of total

caloric intake (Hunter, 2005).

3. Nutrition labeling regulation of fat content in cereal products

Although it is pertinent to educate the public on negative health risks associated with high

intakes of food components, such as saturated and trans fat, and on the benefits to health of

nutrients such as monounsaturated fatty acids, it is equally important that consumers accurately

relate nutrient content with specific foods. An obvious way to inform consumers about the

nutrient content of foodstuffs is through posting, or labeling, of nutrition content on the

packaging of food products. Nutrition labeling of foods was legislated in 1990 in the form of the

Nutrition Labeling and Education Act (NLEA) (Summers, 2003).

3.1. History of the Nutrition Labeling and Education Act

Prior to the NLEA many companies were labeling their products with nutrition details.

Regulations were insufficient, however, to provide clear, uniform information for the diverse

number of food products on the market. Consumers, aware of growing evidence supporting a

link between diet and health, were confused and wanted better regulated labels (Hutt, 1995).

There was much controversy surrounding whether the Food and Drug Administration

(FDA) had legal authority to regulate nutrition labeling under the Food, Drug, and Cosmetics

Act. Congress gave the FDA authority over nutrition labeling with the passing of the NLEA in

October of 1990 (Summers, 2003). President George H. Bush signed the act into law the

13

following November, and it was made effective in May of 1994. The hope was that the NLEA

would clear consumer confusion, allowing healthier dietary choices; encourage food

manufacturers to provide increased nutritional value with their products; provide labeled

macronutrient information in the context of the ideal diet; and ensure consistent terminology and

format uniformity (Scarbrough, 1995).

Only products regulated by the FDA are covered by the NLEA. Although most foods fall

under FDA jurisdiction, there are a few products that do not. Meat and poultry products are

regulated by the United States Department of Agriculture, Food Safety and Inspection Service

(USDA, FSIS), and alcoholic beverages are regulated by the Bureau of Alcohol, Tobacco, and

Firearms (BATF). Furthermore, all advertising, including food advertising, is regulated by the

Federal Trade Commission (FTC), thus, any statements made to advertise or promote a product

are regulated by the FTC. The FDA has a working relationship with all of these agencies

(Scarbrough, 1995). Although meat and poultry products do not fall under NLEA legislation, the

USDA, FSIS adheres to NLEA guidelines.

3.2. The Nutrition Facts panel and labeling of lipid content

The Nutrition Facts panel on processed and/or packaged foods is a result of the NLEA. It

performs many functions. For the FDA, it serves as a regulatory tool, and a source of declared

values to compare with analytically determined values. For the manufacturer, the panel serves as

a vehicle to promote their product. For the consumer, the panel serves as a valuable resource,

divulging information about a product’s nutritional value, and facilitating wise food choices by

comparison with similar products on the market. The first five components listed on the label in

order are calories, calories from fat, total fat, saturated fat, and trans fat. Polyunsaturated and

monounsaturated fat are optional and listed below trans fat (Scarbrough, 1995).

14

Total, saturated, polyunsaturated and monounsaturated fat values may be expressed as

zero on the nutrition label if they are present in quantities less than 0.5 grams. Fat content of less

than 5 grams must be expressed to the nearest 0.5 gram increment. Fat equal to or greater than 5

grams may be expressed to the nearest 1gram increment. Total fat declarations must always be

present on the Nutrition Facts panel, whereas, saturated fat may be deleted from a product’s

Nutrition Facts panel when total fat content of the product is less than 0.5 g, no claim is made

about fat or cholesterol, and calories from saturated fat are, also, not declared. Whenever

saturated fat is deleted the statement “Not a significant source of saturated fat” must follow the

nutrient list. Polyunsaturated and monounsaturated fat declaration becomes mandatory when

one of them is listed or a claim about fatty acids or cholesterol is made, unless the food is fat free

(Code Federal Regulations, 2006).

3.3. Trans-fat labeling

In 1998 the Center for Science in the Public Interest (CSPI), a consumer advocacy

organization, filed a petition for mandatory trans fat labeling. The FDA responded by issuing a

proposed rule in November of 1999, soliciting comments from the food industry, trade

associations, consumers, consumer advocacy organizations, academia, health care professionals,

and national and international governments about the scientific, economic, policy and legal basis

of the rule. In the final rule, the FDA cited research from scientific articles, expert panels, and

studies from the Institute of Medicine, National Academy of Science, 2001 Report of the

National Cholesterol Education Program, and the 2000 Dietary Guidelines. Research indicated

that diets high in trans fats increased serum low density lipoprotein cholesterol, a major risk

factor for coronary heart disease (CHD). The final rule was issued in July 2003 and made

effective on January 1, 2006 (Federal Register, 2003).

15

The final rule required mandatory declaration of trans fat on the Nutrition Facts panel on

a separate line immediately following saturated fat. No % Daily Value would be listed, because

scientific evidence at that time for determination of a Daily Value for trans fat was deemed

inadequate. The official definition provided for trans fat was “unsaturated fatty acids that

contain one or more isolated (i.e., nonconjugated) double bonds in a trans configuration (Federal

Register, 2003).” This is a definition based solely on structural chemistry, and not on the health

effects of trans fat. Foods with less than 0.5 g trans fat are listed as zero trans fat; no declaration

is necessary if total fat is less than 0.5 g and no health claims are made about fat, fatty acids, or

cholesterol (Federal Register, 2003).

For the implementation of trans fat labeling, the FDA estimated a one time cost to the

food industry of approximately $40 to $250 million due to: the quantification of trans fat in

products; reformatting of the Nutrition Facts panel to include trans fat; and reformulation efforts

to reduce the amount of trans fat in products. The FDA estimates that 3 years after the effective

date, trans fat labeling will prevent 600-1200 heart attacks and 250-500 deaths associated with

CHD each year. Nine hundred million to $1.8 billion a year may be saved in medical costs, loss

of productivity, and pain and suffering (Federal Register, 2003).

3.4. Daily Values

One of the goals of the NLEA is to convey a product’s nutritional information from the

perspective of a fraction of a complete day’s diet. This concept gave rise to the Daily Value

chart. Daily Values were a new dietary reference term to appear on the food label. The % Daily

Values are based on recommendations for consumption of macro- and micro- nutrients

associated with health. Ideally, the sum of % Daily Values for any given nutrient from all foods

consumed in a day should equal 100%. This would signify that the individual’s dietary needs for

16

that particular nutrient were met (Wardlaw et al., 2007). Daily Values are based on Daily

Reference Values and Reference Daily Intakes. Daily Reference Values (DRVs) are nutrient

intake standards for protein, carbohydrate, and some dietary components lacking Recommended

Dietary Allowances, such as fat, saturated fat, cholesterol, fiber, sodium, and potassium.

Reference Daily Intakes (RDIs) are nutrient intake standards based on the old Recommended

Dietary Allowance (RDAs) for essential vitamins and minerals, and selected groups, such as

proteins. The Daily Value percentage declared on the Nutrition Facts panel is determined using

the DRVs and RDIs and based on a 2,000 calorie diet. The Daily Value for total fat is 65g, and

the Daily Value for saturated fat is 20g (Summers, 2003). The FDA has not set a Daily Value

for trans fat, citing insufficient information available to establish a recommended amount for

daily trans fat consumption (Federal Register, 2003).

3.5. Compliance with the FDA food labeling regulations for lipids

The FDA mandates no specific methods for determination of a manufacturer’s labeled

values, but only requires that those values are in agreement with FDA official definitions for

total, saturated, and trans fat. Accuracy of labeling depends on the nutrient class. Nutrients such

as calories, sugars, total fat, saturated fat, cholesterol, and sodium are deemed misbranded if the

nutrient content is greater that 20% in excess of the value for that nutrient declared on the label

(CFR, 2006).

For monitoring purposes, the FDA analyzes specific products from a composite made of

12 independent packages of the product, thereby minimizing package to package variation. An

Association of Official Analytical Chemists (AOAC) or other FDA approved method is used to

compare the analyzed value with the labeled value. If a product is found misbranded, the FDA

may take several types of actions to correct the problem, all of which may incur negative

17

publicity to the manufacturer (Heimbach, 1995). The action taken depends on the severity of the

violation and may vary in harshness from a warning letter to seizure of product and prosecution

(Summers, 2003).

3.6. Uniformity of terminology for lipids

The NLEA provides a set of universal definitions for previously undefined or ill-defined

terms. Total fat is defined as “total lipid fatty acids and expressed as triglycerides”; saturated fat

as the “sum of all fatty acids containing no double bonds”; polyunsaturated fat as the sum of all

“cis,cis-methylene-interrupted fatty acids”; and monounsaturated fat as the sum of all “cis-

monousaturated fatty acids (CFR, 2006).” The official definition provided for trans fat was the

sum of “unsaturated fatty acids that contain one or more isolated (i.e., nonconjugated) double

bonds in a trans configuration” (Federal Register, 2003).

4. Quantitation of lipid content of cereal products

4.1. NLEA implications for fat analysis

The regulations set by the NLEA require that any method for the analysis of total fat for

nutrition labeling of foods to effectively extract fat, measure individual fatty acids, and convert

each fatty acid to its equivalent triglyceride weight. Trans, saturated, monounsatured, and

polyunsaturated fatty acids need not be converted to triglyceride equivalents, but are simply

individually measured, summed, and declared by their weight (Ratnayake, 2004).

At the onset of the NLEA, there was a need for method development, validation, and

standardization for many nutrient components. For the macronutrient fat, in particular, no

official AOAC methods existed that satisfied NLEA regulations (Shapiro, 1995). The AOAC

International (originally the Association of Official Analytical Chemists) is an organization that

evaluates, validates, and publishes official standard methods for analysis, and methods are

18

evaluated using interlaboratory collaborative studies. If a method performs satisfactorily in its

full collaborative study, the AOAC method committee bestows first action approval status to the

method, and it is published in the Official Methods of the AOAC (OMA) book. The

collaborative study is published in the Journal of AOAC International (JAOAC). Upon review

of collaborative studies, taking into account the statistical results and the safety of the methods,

among other parameters, the Official Methods Board of review votes whether or not to endow

final action approval (AOAC, 2000g). Due to the nature of the extensive approval process,

AOAC official methods are often not the fastest, most sensitive, simple, or up-to-date procedures

(Shapiro, 1995).

With the implementation of the NLEA, the AOAC assembled a task force in 1991 to

assess the applicability of existing AOAC methods in relation to the new regulations. The goals

of this task force were:

“To identify currently available methods that satisfy labeling requirements, to identify

methods that needed revision or modifications, to identify nutrient/matrix combinations

that need method development and validation, and to identify reference materials for the

validation and quality assurance processes” (Shapiro, 1995).

This task force evaluated 1080 nutrient/matrix combinations and found that 88% of methods

were adequate for nutrition labeling purposes. Fat was one of the nutrient components,

nonetheless, posing the most problems. To assist with method development, the AOAC task

force suggested that the FDA set official definitions for each nutrient component. The FDA

complied and set the official definitions for total, saturated, monounsaturated, and

19

polyunsaturated fats that are in place at present (Shapiro, 1995). The official definition for fat

changed the previous classical definition for fat from “any of various substances that are soluble

in organic solvents” to “the sum of total fatty acids expressed as triglycerides” (Ashraf-

Khorassani et al., 2002).

4.2. Traditional gravimetric methods for fat analysis

Traditional gravimetric methods for fat analysis can be applied to a variety of matrixes,

e.g., cereal products, raw cereal grains, meats, animal feed, nuts, butter, and spices (Shapiro,

1995). Traditional methods may involve hydrolysis of the sample, organic solvent extraction of

soluble materials from the matrix, evaporation of the solvent, and weighing of the extracted lipid

(Ratnayake, 2004). Some traditional gravimetric methods for the determination of fat in cereal

products are AOAC 992.06 for fat in flour and AOAC 923.05 for lipids in flour (AOAC, 2000g).

4.2.1. Soxhlet fat extraction—A traditional fat extraction technique

For over 100 years, Soxhlet extraction methods have been a standard reference against

which other techniques of extraction were evaluated. The Soxhlet is a cross between a

continuous and discontinuous system. It is continuous in that the solvent is recycled through the

system, and it is discontinous in that fat is measured a single sample at a time (de Castro et al.,

1998). The sample, which is usually dried and may be hydrolyzed, is measured into an

extraction thimble, and the weight of the thimble and sample is recorded. The thimble is placed

in the extraction chamber of a Soxhlet apparatus (Fig. 1.3.) and gradually filled with solvent

condensed from a heated boiling flask. When the solvent begins to overflow, a siphon drains the

solute/solvent solution and directs it back into the boiling flask. As the solvent evaporates from

the flask, the vapor rises into a condenser and re-condenses to participate in the cycle again. In

this way, the sample is repeatedly in contact with fresh solvent. This process is repeated for 12-

20

72 hours (Shapiro, 1995). After extraction is complete, the weight of the thimble, containing the

defatted sample, is measured; the loss in weight of the sample is determined and used to

indirectly calculate sample fat content as “any of various substances that are soluble in organic

solvents”. The low cost of equipment makes multiple analyses feasible through purchasing of

additional units of the apparatus. Extractions performed using Soxhlet tend to be more effective

than shaking or stirring techniques (de Castro et al., 1998).

Several disadvantages are associated with Soxhlet methods. Fats vary in structure and

polarity, therefore, solvent selection is a vital factor in any fat extraction (Shapiro, 1995). No

solvent or combination of solvents will extract selectively for all lipids (Eller et al., 2001).

Organic solvents commonly used range from nonpolar solvents, such as hexane, petroleum ether,

and diethyl ether; to solvents intermediate in polarity, such as acetone; to more polar solvents

such as chloroform, methanol and ethanol. Different combinations of these solvents may yield

different fat extraction results (King et al., 1996). The duration of extraction may range from

several hours to a few days, and no agitation of the sample may be used to speed up the process

(de Castro et al., 1998). Because the process is long and involves heat, thermal decomposition of

analytes may occur. In addition to time limitations, large volumes of solvents, which are often

hazardous and require expensive disposal, are required.

4.2.2. Soxtec® extraction: an automated fat extraction technique

Several companies manufacture automated instruments, such as the Soxtec® Auto Fat

Extraction System (Foss North America, Eden Prairie, MI, USA), that perform multiple

accelerated Soxhlet-style extractions. Soxtec® instruments use smaller amounts of sample and

solvent, and produce similar results to traditional Soxhlet methods. The first prototypes were

designed in the 1970s and commercialized in 1982 (de Castro et al., 1998; Sporring et al., 2005).

21

The principle of the methodology involves a combination of reflux boiling and traditional

Soxhlet extraction (de Castro et al., 1998). The sample is usually dried and may or may not be

hydrolyzed, depending on the methodology employed. The sample is weighed into thimbles and

inserted into the extraction unit. Aluminum cups are, subsequently, inserted into the extraction

unit, directly under each sample-containing extraction thimble. The solvent is added to the

instrument in a closed system. The extraction thimbles are lowered into boiling solvent. After

boiling, the sample is rinsed by refluxing of the solvent, and the solute/solvent solution is

collected in the aluminum collection cups. The solvent is evaporated away from the extracted fat

and recovered. The fat remaining in the cup is weighed and the percentage fat calculated (de

Castro et al., 1998; Anonymous, 2006).

Soxtec® instruments have a large number of applications (de Castro et al., 1998) and can

be used for any application a Soxhlet apparatus is used for. Extractions with Soxtec® units are

common in the food, feed, and environmental industrial segments, with particular interests for fat

and pesticide analysis (Anonymous, 2006).

The Soxtec® system offers many advantages over the conventional Soxhlet procedure.

The simpler design achieves similar extraction yield and precision with a drastic shortening of

extraction time, in some cases 90% shorter duration (de Castro et al., 1998; Giner et al., 1996;

Helaleh et al., 2005b; Helaleh et al., 2005a). The instrument allows for the analysis of six

samples simultaneously with one unit. The solvents used in Soxhlet extractions tend to be

hazardous and flammable and require expensive disposal. Because it is an automated, closed

system, the operator is not exposed to solvents or vapors released during the procedure (Helaleh

et al., 2005a). Also, less solvent overall is consumed using the Soxtec® rather than the Soxhlet

system. In fact, 80% of the solvent can be recovered and reused (Anonymous, 2006).

22

4.3. Limitations of traditional fat methods for nutrition labeling purposes

Traditional fat analyses have been proven inadequate for nutritional labeling purposes for

a number of reasons. Fat extractions are largely matrix dependent, and many sample matrix-

related variables, such as particle size and moisture content, affect the extraction process. The

presence of food processing aids may encapsulate the fat or bind strongly to the fat, making it

inaccessible to the solvent resulting in underestimation of fat content. High levels of water-

soluble carbohydrates, e.g., certain sugars, glycol, and lactic acid, in baked goods and other

processed cereal products may interfere with extraction, causing some of the fatty acids present

to be retained in the sample matrix, also resulting in underestimation of fat content. To ensure

that all fatty acids are recovered during extractions, some scientists have attempted longer, more

vigorous extractions; however, overly vigorous extractions may obscure results, increasing

chances that non-fat material will be released into the organic solvent phase (Shapiro, 1995).

To facilitate the release of fat into the solvent, hydrolytic pretreatment procedures can be

employed. The ideal purpose of hydrolysis is to cleave bound fatty acid moieties from non-fat

structures (King et al., 1996). In reality, hydrolysis is often not specific to fat substances and can

cause more non-fat substances to be leached from the sample matrix during solvent extraction

(Shapiro, 1995) resulting in overestimation of fat content. Non-fatty acid substances that might

be extracted into organic solvents include glycerol, low molecular weight carbohydrates,

polymerization products, amino acids, and urea salts (Ratnayake, 2004).

Many studies have illustrated that with or without hydrolytic pretreatment, accurate fat

quantification is unlikely using traditional gravimetric methods, especially for low fat products

(Rader et al., 1995; Eller et al., 2001; Ashraf-Khorassani et al., 2002). Simple ether extraction

gravimetric methods tend to underestimate fat, and gravimetric methods that combine hydrolysis

23

and ether extraction tend to overestimate fat (Zou et al., 1999; Ranhotra et al., 1996; Rader et al.,

1995). The difference in weight of extract may be small, but for low fat samples small

differences in weight may correspond to large differences in accuracy of percentage fat detected

in the sample (Rader et al., 1995; Ali et al., 1997). Also with any gravimetric measurement, the

percentage of saturated, unsaturated, and trans fatty acids is not obtained.

4.4. Measurement of total fat and saturated, unsaturated, and monounsaturated fatty acids by

hydrolytic extraction gas chromatographic methods

4.4.1. General principles and methodology

AOAC Method 996.01 is a hydrolytic extraction gas chromatographic method that can be

used to quantify total, saturated, unsaturated, and monounsaturated fat in cereal-based products

with sufficient accuracy to satisfy current nutrition labeling requirements (Ratnayake, 2004,

Ngehngwainbi et al., 1997). The method determines total fat as the sum of fatty acids expressed

as triglycerides and saturated, unsaturated, and monounsaturated fats as the sum of individual

fatty acids. The method was developed to yield results that were consistent with the new official

definitions for fat established by the FDA with the onset of the NLEA (Ngehngwainbi et al.,

1997; House et al., 1994). The method is more accurate than the hydrolytic gravimetric methods

in that bound lipids are released by hydrolysis but quantitation of the extracted lipid by capillary

gas chromatography (GC) is specific for fatty acids, therefore, total lipid is not overestimated due

to the release of non-lipid ether soluble materials. The method was submitted as a possible

AOAC method in November of 1995 and approved for first action for cereal products in 1996

(Ngehngwainbi et al., 1997; OMA, 2000). Although the method was intended for cereal

products only, Ali et al. (1997) demonstrated that it could be applied to a wide range of food

products, including meats, candies, and oils (Ratnayake, 2004).

24

AOAC Method 996.01 involves grinding of the food sample followed by digestion with

hot 8N HCl to facilitate the release of bound lipids. The hydrolyzed sample is extracted using

equal proportions of two solvents: diethyl ether and petroleum ether. The solvents are

evaporated away, and the extracted fatty acids are saponified and methylated to fatty acid methyl

esters (FAME). Fatty acid content is determined by capillary GC with flame ionization detection

(FID). The method stipulates use of a 30 m x 0.25 mm id x 0.2 µm column with nonbonded 90%

cyanopropyl, and 10% phenyl siloxane. Comparison of the test sample with a mixed standard of

FAMEs, containing from C:8 to C:22 fatty acids and including C:13, C:15, and C:17, is used to

identify fatty acid peaks from each test sample. An internal standard, glyceryl tridecanoate

(C:13) is added at the beginning of the procedure and is used in the quantification of each fatty

acid. For calculation of total fat, each fatty acid is converted to its triglyceride equivalent weight

and summed to give total fat expressed as triglyceride equivalents. Saturated, unsaturated, and

monounsaturated fatty acids are expressed as the direct sum of their individual fatty acids.

Initially each laboratory will optimize the specific GC conditions with mixed standard FAME, an

internal standard, practice samples, and a standard reference material, before test sample analysis

(Ngehngwainbi et al., 1997). The committee for AOAC method 996.01 determined that the

method was suitable for expressing fat content in cereal products for nutrition labeling purposes

for values ranging from 0.5% to 13% total fat (Ngehngwainbi et al., 1997). For samples with

>13% total fat the sample size is reduced.

A similar method to AOAC 996.01 is AOAC 996.06. AOAC 996.06 differs from AOAC

996.01 in that: triundecanoin (C:11) is used as the internal standard rather than tridecanoin

(C:13) (however, C:13 may be more advantageous as it is less volatile) (Ratnayake, 2004);

pyrogallic acid is added to minimize oxidation; hydrolysis is alkaline or acid hydrolysis; the

25

carrier gas is hydrogen rather than helium; and the column is longer (60m). Method 996.06 can

be used for a wide variety of foods including, meats, cheeses, other dairy products (alkaline

hydrolysis) or cereal products (acid hydrolysis) (AOAC, 2000c; AOAC, 2000d).

4.4.2. Limitations

Extreme caution is required of the operator when conducting analyses such as AOAC

996.01. Several reagents used are caustic (8N HCl), and/or flammable (petroleum ether, diethyl

ether). Boron trifluoride, used during the methylation of fatty acids into FAME, may be fatal if

inhaled. The method uses large quantities of solvents (petroleum ether and diethyl ether) and

requires specific disposal of the chemical waste, i.e., petroleum ether, diethyl ether, boron

trifluoride, heptane, and methanolic sodium hydroxide. For protection, workers are required to

wear gloves and safety glasses and work under a properly vented hood throughout most of the

procedure (Ngehngwainbi et al., 1997). AOAC method 996.01 also is very time consuming and

labor intensive, taking at least 8 hours of almost constant attention on the part of the operator.

An additional drawback of the method is the relatively short length of the column, which would

prevent good separation of cis- from trans- isomers of unsaturated fatty acids (Ratnayake, 2004).

4.5. Measurement of trans fatty acids

Measurement of trans fatty acids in processed cereal products begins with the extraction

of oil from the product by a technique similar to that in AOAC methods 996.01 and 996.06.

Trans fatty acids can then be quantified in the extracted oil by several techniques, which include

capillary GC of FAME, thin-layer chromatography, silver-ion liquid chromatography, Fourier

transform infrared (FTIR) spectroscopy and mass spectrometry (MS). The technique

selected depends on the complexity of the sample, the amount of trans fat in the sample, the

purpose of the analysis and the accuracy required (Ratnayake, 2004).

26

4.5.1. Capillary gas chromatography

Gas chromatographic methods are the most commonly used methods to quantify lipids

and component fatty acids because of the accuracy and specificity of the technique. A GC

method that could determine total fat and lipid components, including trans fat, simultaneously

would be advantageous (Mossoba et al., 2003). Existing capillary GC methods, such as AOAC

996.01, have been modified to achieve this objective using capillary columns with highly polar

stationary phases. The separation of fatty acids is based on carbon chain length, the degree of

unsaturation, and the geometry and position of the double bonds within the fatty acid. The

shortest chain length fatty acids are the first ones to reach the detector. For fatty acids of the

same chain length, more saturated fatty acids precede progressively increasing unsaturated fatty

acids, and trans positional isomers elute before cis isomers (Mossoba et al., 2003). Optimum

GC operating conditions need to be established for individual applications. Optimization of GC

conditions and experienced, justified discretion in peak identification make for a successful trans

fat analysis. GC methods usually have a detection limit of as low as 0.5 % trans fatty acid as a

percentage of total fat (Mossoba et al., 2003).

The major modification of AOAC methods 996.01 and 996.06 for trans fatty acid

analysis is in the length of the column required. While 996.01 recommends a 30m x 0.25 mm id

x 0.2 μm film, nonbonded 90% cyanopropyl, 10% phenyl siloxane capillary column, for total fat

analysis (AOAC, 2000c) and 996.06 recommends a 60m fused silica capillary column, a 100m x

0.25 mm id x 0.2μm 100% cyanopropylsilicone stationary phase column is required for trans

analysis (AOAC, 2000d). Two suitable 100m columns currently available are SP 2560 or CP Sil

88. Both columns give much better separation of 18:1 fatty acid cis and trans isomers, than the

27

30m and 60m capillary columns (Mossoba et al., 2003; Ratnayake, 2004). In addition to longer

column length, GC conditions are optimized to give the best separation of trans and cis isomers.

AOCS Ce1h-05 is the Association of Oil Chemists’ Society’s (AOCS) official GC

method for qualitative and quantitative analysis of trans fatty acid isomers. It was developed for

use in the determination of saturated, and cis and trans monounsaturated and polyunsaturated

fatty acids in hydrogenated and refined vegetable oils and fats, but may be used for oil extracted

from a variety of food products. The method is not applicable for analysis of dairy, ruminant or

marine long chain polyunsaturated fats and oils, or any other products supplemented with or

containing conjugated linoleic acid (CLA). It is very similar to the GC procedures for AOAC

996.01 and AOAC 996.06. It calls for use of 100-120m fused silica capillary column, with a

highly polar stationary phase (Mossoba et al., 2003; AOCS, 2006).

4.5.2. Fourier Transform Infrared (FTIR) spectroscopy

Infrared spectroscopy methods may overcome any difficulties in separation of cis and

trans fatty acid isomers associated with the capillary GC methods. Infrared methods can rapidly

measure trans fat if present as more than 5% of total fat and are generally performed on oils.

The C-H out of plane deformation, found in double bonds with the trans configuration,

selectively absorbs IR radiation at 966 cm-1. This unique quality is exploited in IR trans fat

quantification methodologies (Mossoba et al., 2003).

An FTIR instrument has three parts: an IR radiation source; an interferometer, which

allows the detection of wavelengths in the mid IR region (4000-600cm-1); and a detector. During

analysis, the sample selectively absorbs IR energy, and changes in energy reaching the detector

over time are used to make an interferogram. The mathematical Fourier transformation is used

28

to convert the raw spectrum from energy versus time, to energy versus frequency, therefore

yielding a single beam spectrum.

The sample single beam spectrum is a profile of the energy reaching the detector, and is a

product of the emittance of the IR source and the absorption of all IR absorbing materials in the

radiation path. Some IR absorbing materials that may be in the path of the radiation include the

test sample, water vapor, and CO2. A reference raw spectrum, containing no test sample, is

obtained and converted into the corresponding single beam reference spectrum. The single beam

reference spectrum is ratioed to the single beam sample spectrum to obtain a transmission or

absorbance spectrum; the energy from this spectrum is assumed to be due only to the sample

being tested. The absorbance spectrum is compared with a calibration curve, and the

concentration is estimated. Free fatty acids must be methylated before examination, because the

O-H out-of-plane deformation in the COOH functional group tends to interfere with the 966 cm-1

trans absorption band. The individual types of trans fatty acids present cannot be obtained using

FTIR methodology, but currently only the sum of total (isolated) trans fatty acids are required

for nutrition labeling of food products (Mossoba et al., 2003). Two modes are commonly used

for FTIR trans fat quantification: transmission FTIR and attenuated total reflectance (ATR).

4.5.2.1. Transmission FTIR

Transmission FTIR is the traditional method for trans fat analysis. Transmission FTIR

uses liquid cells made of sodium chloride salt crystals separated by a thin Teflon spacer. Only

transmitted radiation, the energy that passes through the sample, reaches the detector (Mossoba

et al., 2003). A calibration plot is generated using various concentrations of solutions, usually

carbon disulfide solutions of methyl elaidate (9t-18:1) (Mossoba et. al, 2003; Ratnayake, 2004).

29

Absorbance follows Beer’s law and is dependent upon the molar absorption coefficient of

analyte, the pathlength of IR light, and the concentration of analyte (Mossoba et al., 2003).

AOCS Cd14-95 and AOAC 965.34 are two standard methods for quantification of trans

fat using transmission FTIR. In both methods, FAME are weighed and dissolved in CS2 at a

concentration of 0.2g FAME/10 mL CS2. FAME are analyzed using a 1 mm fixed pathlength in

transmission cells. Calibration standards are made using known proportions of methyl elaidate

(9t-18:1) and methyl oleate (9c-18:1). Two linear regression calibrations are made, one for

samples containing less than or equal to 10% trans fat and one for samples with greater than

10% trans fat. These methods assume that the major component is methyl elaidate (Mossoba et

al., 2003; AOAC, 2000e).

4.5.2.2. Attenuated Total Reflectance (ATR)-FTIR

ATR-FTIR is easier and more convenient compared to transmission FTIR (Mossoba et

al., 2003). A fat sample is placed on a crystal. Under conditions of total internal reflection, IR

light waves are completely reflected inside the crystal. As light begins to bounce inside the

crystal, evanescent waves directed away from the crystal surface proliferate and travel through

the melted fat sample. The intensity of the evanescent waves generated attenuates quickly as the

waves travel over distance and are absorbed by trans fat. The single beam spectrum obtained is

a measure of the attenuation of total internally reflected light by the trans fat in a test sample. A

reference or background interferogram/single beam spectrum, is obtained using a trans-free

sample. The background spectrum of the trans-free sample is ratioed to the single beam

spectrum of the trans fat test sample, removing any O-H interference with the 966 cm-1 peak and,

therefore, eliminating the need for methylation preparation of fatty acids. Test samples are

compared with a calibration curve prepared using standards with known amounts of trans fat.

30

AOCS Cd 14d-99 and AOAC 2000.10 are two standard methods for trans fat analysis using

ATR-FTIR. Because the weighing of FAME and dilution with the volatile solvent CS2, is not

necessary, these methods are simpler and faster than the official transmission FTIR methods

(Mossoba et al., 2003; AOAC, 2000a).

4.6. Measurement of total lipid/fat and saturated, polyunsaturated, monounsaturated, and trans

fatty acids by an automated hydrolysis and extraction (AHE) system

4.6.1. General principle and methodology

The Soxtec® Auto Fat Extraction System was recently extended with the addition of a

closed semi-automated acid hydrolysis step to precede solvent extraction. Initially the sample is

weighed into glass capsules fitted with a disposable filter. Six sample capsules are placed in a

holder, which is lowered into a hydrolysis unit (Soxcap™ Hydrolysis Unit, Foss North America,

Eden Prairie, MI, USA) containing hot acid. The unit then allows hydrolysis and subsequent

rinsing of the samples in a closed system. The rinsing or washing removes aqueous, non-lipid,

hydrolyzed materials that could otherwise be extracted into the solvent with resultant

overestimation of fat content. After hydrolysis and rinsing, the glass capsule containing the

remaining sample is fitted with a cellulose thimble on the top, inverted, dried and placed in a

Soxtec® Auto Fat Extraction unit where lipids are extracted as described previously for the

Soxtec® system. The total lipid extracted can be measured gravimetrically. Alternatively, lipid

can be recovered, saponified, methylated and analyzed by capillary GC. The accuracy of this

automated system for extraction of lipids for the analysis of total fat and lipid components

compared to the extraction of lipids by AOAC Method 996.01 has not been reported. Thus, its

potential for analysis of lipids for nutrition labeling and monitoring is unknown.

The Soxcap™ 2047 Hydrolysis unit (Figure 1.4.) and the Soxtec® 2050 Auto Fat

31

Extraction System unit (Figure 1.4.) can be purchased from Foss North America (Eden Prairie,

MN) as a package for $28,500 (direct telephone quote, November 2006). Initial equipment costs

of the AHE system are quite high, however, these may be off-set by reduced labor costs.

4.6.2. Advantages of the AHE system compared to AOAC Method 996.01 for extraction of

lipid

A major advantage of the AHE system compared to AOAC Method 996.01 for extraction

of total lipid is that the acid hydrolysis step is semi-automated and the extraction step is fully

automated allowing the operator’s time and attention to be focused elsewhere. The hydrolysis

and rinsing is performed without any sample transfer and with minimum manual handling of the

samples due to the use of batch handling tools. Automation of the processes also reduces

operator error. Both the hydrolysis and extraction are performed in closed systems, reducing

operator exposure to hot acid, solvent, and fumes; in addition the hydrolysis unit has a fume

removal system so the procedure does not need to be performed in a fume hood as does the

AOAC procedure. Both hydrolysis and extraction units will process six samples at a time.

Finally, the amount of solvent required for analysis of each sample is less than used in the

AOAC method and, as previously stated for the Soxtec® unit, up to 80% of the solvent can be

recovered and reused (Foss Tecator, 2006).

5. Objective

The objective of this study is to evaluate the AHE system for the determination of total

fat gravimetrically and for the extraction of lipid for the capillary GC determination of total,

saturated, polyunsaturated, monounsaturated and trans fat. A diverse range of cereal products

with a wide range of total fat and trans fat will be used for the study, and the results will be

evaluated against those using AOAC Method 996.01 as the standard. In addition, the AHE

32

system will be evaluated for the extraction of lipid for Fourier transform infrared (FT-IR)

determination of trans fat.

33

6. References

Aguilera, J., & Gloria, H. (1997). Determination of oil in fried potato products by differential

scanning calorimetry. Journal of Agricultural and Food Chemistry, 45(3), 781-785.

Ali, L., Angyal, G., Weaver, C., & Rader, J. (1997). Comparison of capillary column gas

chromatographic and AOAC gravimetric procedures for total fat and distribution of fatty

acids in foods. Journal of Food Chemistry, 58(1-2), 149-160.

Anonymous. (2006). Soxtec™ systems, http://www.foss.dk/solutions/productsdirect/soxtec

systems.aspx, November 2006.

AOAC (2000a). Official methods of analysis. (2000.10) Determination of total isolated trans

unsaturated fatty acids in fats and oils. (17th ed.) USA: AOAC International.

AOAC (2000b). Official methods of analysis. (922.06) Fat in flour. (17th ed.). USA: AOAC

International.

AOAC (2000c). Official methods of analysis. (996.01) Fat (total, saturated, unsaturated, and

monounsaturated) in cereal products. (17th ed.). USA: AOAC International.

AOAC (2000d). Official methods of analysis. (996.06) Fat (total, saturated, and unsaturated) in

foods. (17th ed.). USA: AOAC International.

AOAC (2000e). Official methods of analysis. (965.34) Isolated trans isomers in margarines and

shortenings. (17th ed.). USA: AOAC International.

AOAC (2000f). Official methods of analysis. (923.05) Lipids in flour. (17th ed.). USA: AOAC

International.

AOAC (2000g). Official methods validation program. In W. Horwitz, Official methods of

analysis (17th ed.). (pp. xxiii-xxvii). USA: AOAC International.

34

AOCS (2006). Official Methods and Recommended Practices of the American Oil Chemists’

Society. (Ce 1h-05) Determination of cis-, trans-, saturated, monounsaturated, and

polyunsaturated fatty acids in vegetable or non-ruminant animal oils and fats by capillary

GLC. (5th ed.). Champaign, USA: AOCS Press.

Ascherio, A. (2002). Epidemiologic studies on dietary fats and coronary heart disease. American

Journal of Medicine, 113(9B), 9S-12S.

Ashraf-Khorassani, M., Ude, M., Doane-Weideman, T., Tomczak, J., & Taylor, L. (2002).

Comparison of gravimetry and hydrolysis/derivatization/gas chromatography-mass

spectrometry for quantitative analysis of fat from standard reference infant formula

powder using supercritical fluid extraction. Journal of Agricultural and Food Chemistry,

50, 1822-1826.

Bakal, A. (1991). Fat reduction in baked goods. In C. Haberstroh, & C. Morris, Fat and

cholesterol reduced foods: technologies and strategies, (pp. 185-192). Houston, TX:

Gulf.

Becker, R. (2000). Fatty acids in food cereal grains and grain products. In C. Chow, Fatty acids

in foods and their health implications, (pp. 271-284). New York: Marcell Dekker.

Bruckner, G. (2000). Fatty acids and cardiovascular diseases. In C. Chow, Fatty acids in foods

and their health implications, (pp. 843-864). New York: Marcell Dekker.

Canning, K. (2005). Trans trouble. Snack Food and Wholesale Bakery, 94, 34-37.

Chung, O. (1991). Cereal lipids. In K. Lorenz, & K. Kulp, Handbook of cereal science and

technology, (pp. 497-554). New York: Marcel Dekker.

Craig-Schmidt, M., & Holzer, B. (2000). Fatty acid isomers in foods. In C. Chow, Fatty acids in

foods and their health implications, (pp. 307-356). New York: Marcell Dekker.

35

Code of Federal Regulation; (2006). Title 21, Volume 2, Part 101.9. Washington, DC, USA: US

Government Printing Office.

de Castro, M., & Garcia-Ayuso, L. (1998). Soxhlet extraction of solid materials: an outdated

technique with a promising innovative future. Analytica Chemica Acta, 369, 1-10.

Dendy, D., & Dobraszczyk, B. (2001). Cereals and cereal products: chemistry and technology.

Gaithersburg, Maryland: Aspen.

Eller, F., & King, J. (2001). Supercritical fluid extraction of fat from ground beef: effects of

water on gravimetric and GC-FAME fat determinations. Journal of Agricultural and

Food Chemistry, 49, 4609-4614.

Faridi, H. (1991). Soft wheat products. In K. Lorenz, & K. Kulp, Handbook of cereal science

and technology, (pp. 683-740). New York: Marcel Dekker.

Federal Register. (2003). Food labeling: trans fatty acids in nutrition labeling, nutrient content

claims, and health claims. Vol. 68, 41433-41506.

Formanek, R., Ed. FDA Office of Public Affairs. (2003). Revealing trans fats. FDA Consumer

Magazine, 37, http://www.fda.gov/fdac/features/2003/503_fats.html.

Gunstone, F. (2002). Food applications of lipids. In C. Akoh and D. Min, Eds, Food lipids:

chemistry, nutrition, and biotechnology, (2nd ed.) (pp. 663-694). New York: Marcell

Dekker.

Hasler, C. (1998). Functional foods: their role in disease prevention and health promotion.

Journal of Food Technology, 52(11), 63-70.

Hausman, D., Higbee, D., & Grossman, B. (2002). Dietary fats and obesity. In C. Akoh and D.

Min, Eds., Food lipids: chemistry, nutrition, and biotechnology (2nd ed.) (pp. 663-694).

New York: Marcell Dekker.

36

Heimback, J., & Egan, S. (1995). Compliance with nutrient content declaration requirements. In

R. Shapiro, Nutrition labeling handbook, (pp. 509-550). New York: N. Dekker.

Helaleh, M., Al-Omair, A., Ahmed, N., & Gevao, B. (2005a). Quantitative determination of

organochlorine pesticides in sewage sludges using Soxtec, Soxhlet and pressurized liquid

extractions and ion trap mass-mass spectrometric detection. Analytical and Bioanalytical

Chemistry, 382, 1127-1134.

Helaleh, M., Al-Omair, A., Nisar, A., & Gevao, B. (2005b). Validation of various extraction

techniques for the quantitative analysis of polycyclic aromatic hydrocarbons in sewage

sludges using gas chromatography-ion trap mass spectrometry. Journal of

Chromatography A, 1083, 153-160.

Hunter, J. (2005). Dietary levels of trans-fatty acids: Basis for health concerns and industry

efforts to limit use. Nutrition Research, 25, 499-513.

Hunter, J. (2000). Safety and health effects of isomeric fatty acids. In C. Chow, Fatty acids in

foods and their health implications, (pp. 667-686). New York: Marcell Dekker.

Hutt, P. (1995). A brief history of FDA regulations relating to the nutrient content of food. In R.

Shapiro, Nutrition labeling handbook, (pp. 1-28). New York: N. Dekker.

Jéquier, E., & Bray, G. (2002). Low-fat diets are preferred. American Journal of Medicine,

113(9B), 41S-46S.

King, J., Eller, F., Snyder, J., Johnson, J., McKeigh, F., & Stites, C. (1996). Extraction of fat

from ground beef for nutrient analysis using analytical supercritical fluid extraction.

Journal of Agricultural and Food Chemistry, 44, 2700-2704.

Krauss, R., Eckel, R., Howard, B., Appel, L., Daniels, S., Deckelbaum, R., Erdman, J., Kris-

Etherton, P., Goldberg, I., Kotchen, T., Lichtenstein, A., Mitch, W., Mullis, R., Robinson,

37

K., Wylie-Rosett, J., St. Jeor, S., Suttie, J., Tribble, D., & Bazzarre, T. (2000). AHA

Dietary Guidelines: Revision 2000: A statement for healthcare professionals from the

Nutrition Committee of the American Heart Association. Circulation, 102, 2284-2299.

Lane, R. (2001). Cereals and cereal products. Journal of AOAC International, 84(1), 230-232.

Matz, S. (1991). The chemistry and technology of cereals as food and feed, 2nd ed. New York:

Van Nostrand Reinhold.

Mensink, R., & Plat, J. (2002). Dietary fats and coronary heart disease. In C. Akoh and D. Min,

Food lipids: chemistry, nutrition, and biotechnology, (2nd ed.) (pp. 603-636). New York:

Marcell Dekker.

Mossoba, M., Kramer, J., Delmonte, P., Yurawecz, M., & Rader, J. (2003). Official methods for

the determination of trans fat. Champaign, Ill: AOCS Press.

Ngeh-Ngwainbi, J., Lin, J., & Chandler, A. (1997). Determination of total, saturated,

unsaturated, and monounsaturated fats in cereal products by acid-hydrolysis and capillary

gas-chromatography—collaborative study. Journal of AOAC International, 80(2), 359-

372.

Putnam, J., Allshouse, J., & Kantor, L. (2002). US per capita food supply trends: more calories,

refined carbohydrates, and fats. Food Review, 25(3), 2-15.

Rader, J., Angyal, G., O’Dell, R., Weaver, C., Sheppard, A., & Bueno, M. (1995).

Determination of total fat and saturated fat in foods by packed column gas-liquid

chromatography after acid hydrolysis. Food Chemistry, 54(4), 419-427.

Ranhotra, G., Gelroth, J., & Vetter, J. (1996). Determination of fat in bakery products by three

different techniques. Cereal Foods World, 41(7), 620-622.

38

Ranhotra, G. (1991). Nutritional quality of cereals and cereal-based foods. In K. Lorenz, & K.

Kulp, Handbook of cereal science and technology, (pp. 845-862). New York: Marcel

Dekker.

Ratnayake, W. (2004). Overview of methods for the determination of trans fatty acids by gas

chromatography, silver-ion thin-layer chromatography, silver-ion liquid chromatography,

and gas chromatography/mass spectrometry. Journal AOAC International, 87(2), 523-

539.

Sacks, F., & Katan, M. (2002). Randomized clinical trials on the effects of dietary fat and

carbohydrate on plasma lipoproteins and cardiovascular disease. American Journal of

Medicine, 113(9B), 13S-24S.

Scarbrough, F. (1995). Perspectives on nutrition labeling education act. In R. Shapiro, Nutrition

labeling handbook, (pp. 29-52). New York: N. Dekker.

Shapiro, R. (1995). Analytical methods and databases for nutrition labeling. In R. Shapiro,

Nutrition labeling handbook, (pp. 551-586). New York, N. Dekker.

Sporring, S., Bowadt, S., Svensmark, B., & Bjorklund, E. (2005). Comprehensive comparison of

classic Soxhlet extraction with Soxtec extraction, ultrasonication extraction, supercritical

fluid extraction, microwave assisted extraction and accelerated solvent extraction for the

determination of polychlorinated biphenyls in soil. Journal of Chromatography A, 1090,

1-9.

Summers, J. (2003). Food labeling compliance review, (3rd ed.). Ames: Blackwell Publishing/

Iowa State Press.

39

Tarrago-Trani, M., Phillips, K., Lemar, L., & Holden, J. (2006). New and existing oils and fats

used in products with reduced trans-fatty acid content. Journal of the American Dietetic

Association. 106. 867-880.

USDA. (2006). Briefing room, economic research service, Food CPI, prices, and expenditures:

outlook for food prices in 2006 Page, http://www.ers.usda.gov/briefing/ CPI

foodandexpenditures/outlook.htm.

Wardlaw, G, & Hampl, J. (2007). Perspectives in Nutrition, (7th ed.). New York: McGraw-Hill.

Zou, W., Lusk, C., Messer, D., & Lane, R. (1999). Fat contents of cereal foods: comparison of

classical with recently developed extraction techniques. Journal of AOAC International,

82(1), 141-150.

40

Fig. 1. 1. Structure of stearic acid

41

Fig. 1.2. Cis and trans geometric isomers of C18:1 fatty acid

trans-9-octadecenoic (elaidic) acid cis-9-octadecenoic (oleic) acid

42

Fig. 1.3 Soxhlet apparatus

43

Fig. 1.4. The AHE system

2050 Auto Fat Extraction unit Soxcap™ 2047 Hydrolysis unit

44

CHAPTER 2

EVALUATION OF AN AUTOMATED HYDROLYSIS AND EXTRACTION METHOD FOR

QUANTIFICATION OF FAT IN CEREAL FOODS1

_____________________________ Robinson, J.E., Singh, R., and Kays, S.E. To be submitted to Food Chemistry

45

1. Abstract

The utility of an automated acid hydrolysis and extraction (AHE) system was evaluated

for extraction of fat for the quantification of total, saturated, polyunsaturated, monounsaturated,

and trans fat in cereal products. Oil extracted by the AHE system was assessed: gravimetrically

for total fat; by capillary GC analysis for total fat, lipid classes, and trans fat; and by ATR-FTIR

spectroscopy for quantification of trans fat. Results were compared with parallel determinations

using the standard AOAC Method 996.01. For gravimetric and gas chromatographic evaluations

the AHE system results were equivalent to the AOAC 996.01 results (α = 0.01), indicating that

the AHE system of lipid extraction can be used to measure total fat, lipid classes, and trans fat

with sufficient accuracy for nutrition labeling. The AHE and AOAC results were not equivalent

for all samples assessed by ATR-FTIR spectroscopy, thus for this application the AHE system

may be better suited for rough screening.

46

2. Introduction

The quantification of fat in foods and its declaration on the nutrition facts label are

mandated as part of the Nutrition Labeling and Education Act (NLEA) (Scarbrough, 1995).

Under the NLEA, statement of the amount of total, saturated, and trans fat in processed foods is

mandatory, and statement of the polyunsaturated and monounsaturated content of foods is

voluntary (Code Federal Regulations, 2006; Federal Register, 2003).

To obtain accurate information on the fat content of various foods and for monitoring of

this information, accurate and repeatable methods are required for the analysis of total fat and

lipid classes in foods. AOAC 996.01 is a universally accepted method for the determination of

total, saturated, unsaturated, and monounsaturated fat in cereal-based products and has sufficient

accuracy and repeatability to satisfy current nutrition labeling requirements (Ratnayake, 2004,

Ngehngwainbi et al., 1997). A modification of AOAC 996.01 may be used for analysis of trans

fatty acids in cereal products (Mossoba et al., 2003). AOAC 996.01 and its modification are

identical up to gas chromatographic analysis and involve hydrolysis of the ground sample,

extraction of fat into diethyl and petroleum ether solvents, evaporation of the solvents,

methylation of the extracted fat, and quantification of fatty acids by gas chromatography (GC)

(AOAC, 2000a). The modification for trans fat requires a longer GC column and GC conditions

that optimize separation of trans and cis isomers. AOAC 996.01 is more accurate than

traditional Soxhlet gravimetric methods for crude fat, in that lipid extraction is more complete

and quantification of the extracted lipid by capillary GC is specific for fatty acids.

Although the method is accurate and repeatable, AOAC 996.01 and similar methods are

laborious procedures, requiring careful attentiveness throughout the duration of analysis. They

are also time-consuming, taking 8 hours to perform with additional time for capillary GC and its

47

interpretation. The protocol consumes large volumes of diethyl ether and petroleum ether,

solvents which are hazardous, flammable and require specific disposal.

An automated hydrolysis and extraction (AHE) system that is available commercially,

offers an alternative to the manual hydrolysis and extraction required for AOAC 996.01. The

method involves a combination of automated acid hydrolysis and rinsing of the sample in a

closed system followed by reflux boiling and automated Soxhlet extraction of the lipid, also in a

closed system (de Castro et al., 1998). The percentage of total lipid is obtained gravimetrically.

In addition, the extracted fat can be recovered and total fat and lipid classes measured by

capillary GC as in AOAC method 996.01. Because the AHE system is automated and closed:

the operator has less contact with and exposure to solvents and fumes; the operator’s time and

attention may be directed towards other activities during extraction; and the results are less likely

to be affected by operator error (Helaleh, et al., 2005). Six samples can be analyzed

simultaneously with one unit. Less solvent overall is consumed using the AHE system, in fact,

80% of the solvent can be recovered and reused (Anonymous, 2006). The design of the AHE

hydrolyzation unit provides for the rinsing of non-lipid aqueous moieties from the hydrolyzed

sample, removing elements that might, otherwise, cause overestimation of gravimetric total lipid.

In theory, this should provide for the accurate determination of total fat without use of a gas

chromatographic step. Recovery of the lipid, and subsequent saponification and methylation,

allow for determination of total, saturated, polyunsaturated, monounsaturated, and trans fat by

capillary GC analysis. The accuracy of the AHE system for extraction of lipids for the analysis

of total fat and lipid components compared to the extraction of lipids by AOAC Method 996.01

has not been reported. Thus, its potential for analysis of lipids for nutrition labeling and

monitoring is unknown.

48

The objective of this study was to evaluate the AHE system for the determination of total

fat gravimetrically and for the extraction of fat for the capillary GC determination of total,

saturated, polyunsaturated, monounsaturated and trans fat. A diverse range of cereal products

with added fat was used for the study, and the results were evaluated against those using AOAC

Method 996.01 as the standard. In addition, the AHE system was evaluated for the extraction of

lipid for Fourier transform infrared (FT-IR) determination of trans fat.

3. Materials and methods

3.1. Samples and sample preparation

Twelve cereal products with a wide range of grains were purchased from local

commercial grocery retailers (Table 2.1.). Based on the Nutrition Facts panel information for

each product, total fat content ranged from 4-40% and trans fat from 0-15%. Products also had a

wide range in sugar (0-50%), fiber (0-6%), and protein (2-10%) content. To ensure that a variety

of cereal products were represented, products were selected from four categories: snacks,

cookies and crackers, baking mixes, and breakfast products. A high fat (total fat ≥ 25%),

medium fat (25% < total fat ≥ 13%), and low fat (>13% total fat) cereal product was selected for

each category, except for the breakfast product category, which contained one medium fat

product, and two low fat products. Overall, the products had a wide variety of ingredients

including fruits, nuts, flavors, spices, sweeteners, fats, flavor enhancers, gums, emulsifiers,

leavening agents, and preservatives. Frying, baking, extruding, milling, and malting processes

were all represented by products included in the study.

Products were ground using a household coffee grinder made by Kitchen Aid (St.

Josephs, MI, USA) to reduce the particle size and obtain a homogeneous sample. The ground

products were transferred to polyethylene bags, stored at -20°C, and analyzed for total, saturated,

49

polyunsaturated, monounsaturated, and trans fat within three to four days. It was established in

duplicate samples that total fat, lipid classes, and trans fat were stable over 7 days at -20°C

(Appendix H).

3.2. Reagents and standards

Chloroform and methanol were HPLC grade and heptane was capillary GC grade. All

other reagents were ACS grade. The Standard Reference Material (SRM) 1846, a spray-dried

infant formula, was purchased from the U.S. Department of Commerce National Institute of

Standards and Technology (NIST) (Gaithersburg, MD, USA). SRM 1846 contains 27.1 ± 0.59%

total fat calculated on an as delivered basis. All AOAC 996.01 and modified AOAC 996.01

analyses were verified using SRM 1846 in parallel with samples. The internal standard for GC

analysis was glyceryl tridecanoate, purchased from Sigma-Aldrich (St. Louis, MO, USA), and

was prepared to a concentration of 20mg/mL in chloroform.

KEL-FIM-FAME-5 mix was obtained from Matreya (St. Pleasant Gap, PA, USA) and is

a 19 component fatty acid methyl esters (FAME) mixed standard containing methyl esters of the

following fatty acids (mg/ml in parentheses): C8:0 (0.3), C10:0 (0.5), C12:0 (1.0), C13:0 (0.5),

C14:0 (0.5), C14:1 (0.3), C15:0 (0.3), C16:0 (2.0), C16:1 (1.0), C17:0 (0.5), C18:0 (1.0), C18:1tr

(0.4), C18:1c (3.0), C18:2 (2.0), C20:0 (0.3), C18:3 (1.0), C20:1 (0.3), C 22:0 (0.3), C22:1 (0.3).

Supelco 37 Component FAME mix was purchased from Supelco (Bellafonte, PA, USA)

and is a mixed standard with methyl esters of the following fatty acids (mg/ml in parentheses):

C4:0 (0.4), C6:0 (0.4), C8:0 (0.4), C10:0 (0.4), C11:0 (0.2), C12:0 (0.4), C13:0 (0.2), C14:0

(0.4), C14:1 (0.2), C15:0 (0.2), C15:1 (0.2), C16:0 (0.6), C16:1 (0.2), C17:0 (0.2), C17:1 (0.2),

C18:0 (0.4), C18:1n9t (0.2), C18:1n9c (0.4), C18:2n6t (0.2), C20:0 (0.4), C18:3n6 (0.2), C20:1

(0.2), C18:3n3 (0.2), C21:0 (0.2), C20:2 (0.2), C22:0 (0.4), C20:3n6 (0.2), C22:1n9 (0.2), C

50

20:3n3 (0.2), C20:4n6 (0.2), C23:0 (0.2), C22:2 (0.2), C24:0 (0.4), C20:5n3 (0.2), C24:1 (0.2),

C22:6n3 (0.2).

3.3. AOAC Method 996.01 for total, saturated, polyunsaturated, and monounsaturated fat

Total, saturated, polyunsaturated, and monounsaturated fat in cereal products and the

standard reference material (SRM 1846) were determined by AOAC Method 996.01 (AOAC,

2000a). Two cereal products were analyzed per day. Briefly, ground products were weighed in

triplicate into Mojonnier tubes using 2g sample for products with 13% fat or less (based on

nutrition label values). Sample size was reduced for samples with >13% fat. One mL of internal

standard (20mg/mL glyceryl tridecanoate) was added to each sample, and samples were

hydrolyzed with 8N HCl at 80 ± 2°C. Lipids were extracted from the hydrolyzed sample matrix

with diethyl ether and petroleum ether. The organic ether phase, containing the extracted fat,

was decanted into flasks and the solvents evaporated. The ether extract was saponified and

methylated. The FAME obtained were suspended in n-heptane and immediately analyzed for

total fat and fatty acids using an Agilent Technologies 6890N Gas Chromatograph, operating

with flame ionization detector and Restek 2330 30 m x 250 μm i.d. x 0.2 μm film thickness

column (10% cyanopropylphenyl-90% biscyanopropyl polysiloxane) purchased from Restek

(Bellefonte, PA, USA). Helium was the carrier gas, with a gas flow velocity of 24 cm/s.

Injection temperature was 250 °C, and the detector temperature was 275 °C. Hydrogen and air

flows were set to 34 mL/min and 300 mL/min., respectively, at a split ratio of 50:1. Oven

temperature programming consisted of an initial temperature of 120 °C held for 4 minutes,

followed by an increase in temperature of 5 °C/min. until 230 °C, with a hold time of 5 minutes.

FAME were measured against the C:13 internal standard; the KEL-FIM-FAME-5 mix was run in

parallel with the samples and used in the identification and quantification of individual fatty

51

acids (Appendix F & G). Each fatty acid was converted to its triglyceride equivalent weight and

triglycerides summed to obtain total fat. The sum of individual fatty acids was used directly to

obtain saturated, polyunsaturated, and monounsaturated fat. Total fat and lipid classes were

reported on a dry weight basis. Dry weight was determined on individual samples independently

at 105°C in a forced air oven (AOAC Method 935.29).

3.4. Modified AOAC method 996.01 for trans fat

For analysis of trans fat content, cereal samples were hydrolyzed and the oils extracted,

saponified, and methylated as described for AOAC method 996.01 (AOAC, 2000a). As for the

unmodified AOAC method, two samples were analyzed simultaneously in triplicate. The FAME

in heptane were stored in GC vials at -20 °C and analyzed within 24 hours. It was established in

duplicate samples that total fat and lipid components as FAME were stable at -20 °C for 7 days

(Appendix I). The FAME were analyzed for trans fatty acid content using an Agilent

Technologies 6890N Gas Chromatograph, operating with flame ionization detector and Supelco

2560 fused silica 100 m x 250 μm i.d. x 0.2 μm film thickness column purchased from Supelco

(Bellafonte, PA, USA). Helium was the carrier gas, with a gas flow velocity of 18 cm/s.

Injection temperature was 200°C, and the detector temperature was 250 °C. Hydrogen and air

flows were set to 40 mL/min and 450 mL/min., respectively, at a split ratio of 50:1. Oven

temperature programming consisted of an initial temperature of 120 °C held for 5 minutes, an

increase in temperature of 3 °C/min. until 240 °C, and a hold time of 20 minutes at 240 °C.

FAME were measured against C:13 internal standard; Supelco 37 component FAME mix was

used in the identification and quantification of individual fatty acids. The sum of all fatty acids

containing trans isomers was used directly to obtain trans fat, which was reported on a dry

weight basis. The modified AOAC 996.01 method was used to measure total, saturated,

52

polyunsaturated, monounsaturated, and trans fat for investigation of the stability of ground

products and FAME at -20 °C.

3.5. Automated hydrolysis extraction (AHE) method for total, saturated, polyunsaturated, and

monounsaturated fat

Ground samples were weighed into glass Soxcap capsules (Foss North America, Eden

Prairie, MI, USA) and fitted with corresponding disposable polyester filters (Foss North

America, Eden Prairie, MI, USA). Two samples were analyzed in triplicate each day. The

quantity of sample used was as described for AOAC Method 996.01. For samples to be analyzed

by capillary GC, 1 mL tridecanoate (C13:0) internal standard (20 mg/mL) was added. Samples

were then hydrolyzed with the Soxcap™ 2047 Hydrolysis Unit (Foss North America, Eden

Prairie, MI, USA). The glass capsules were loaded into a tray that holds six capsules, the tray

lowered into the hydrolysis unit, and the samples allowed to hydrolyze for one hour in boiling

4N HCl. After hydrolysis, the samples were rinsed with water sufficiently to increase the pH of

the rinse water to that of the tap water. The tops of the capsules were fitted with a cellulose

thimble (22 mm × 28 mm i.d.; Foss North America, Eden Prairie, MI, USA), inverted, and

transferred to freeze-drying jars and freeze-dried to constant weight (20 hours) using a Virtis

Freezemobile 24EL (Gardiner, NY, USA). The capsules with cellulose thimbles intact and

containing the dried, hydrolyzed samples, were fitted with metal adaptors, and loaded into a

Soxtec® 2050 Auto Fat Extraction System (Foss North America, Eden Prairie, MI, USA) for

solvent extraction. Briefly, previously weighed aluminum cups were placed beneath each

extraction thimble in the extraction unit, and petroleum ether, as solvent, was added to each of

six extraction chambers. Extraction programming consisted of the sample being in contact with

boiling petroleum ether for 40 minutes, a sample rinsing stage of 40 minutes, a recovery stage of

53

10 minutes, and an evaporation/drying stage of 5 minutes. After extraction, to determine total fat

gravimetrically (AHE-G method), excess petroleum ether was evaporated from the aluminum

cups in a vacuum oven, the cups weighed, and % total fat calculated.

To determine total, saturated, polyunsaturated, and monounsaturated fat (AHE-GC

method), the lipid, from the aluminum cups was transferred to a boiling flask by rinsing with

petroleum ether, the solvent was evaporated, and saponification, methylation, and capillary GC

analysis performed, as for AOAC 996.01. Trans fat was determined, using the capillary GC

conditions described for the modified AOAC 996.01 method.

3.6. Attenuated total reflection (ATR)-FTIR analysis of trans fat—modified AOAC 2000.10

3.6.1. Trans calibration standards

Standards for preparation of a calibration curve were obtained from a Reference Sample

Set purchased from the American Oil Chemists’ Society (Urbana, IL, USA). These included

margarine oil, canola oil, and vegetable shortening having 12.01%, 27.31%, and 47.15% trans

fat as a percentage of total fat, respectively. A soy oil containing 0% trans fat was purchased

from Nexsoy (Springfield, IL, USA). All trans percentages were verified using the modified

AOAC method 996.01.

3.6.2. Preparation of test samples

Oils were extracted from a high fat, medium fat and low fat sample by the modified

AOAC Method 996.01and by the AHE method. All extracted oils were transferred from

extraction vessels to sample vials using disposable pipettes. Sample vials were flushed with N2

gas, sealed, and stored at -20 °C. FTIR analysis was performed within 3 days using the modified

AOAC Method 2000.10 (AOAC, 2000b). It was established in duplicate samples that changes in

trans fatty acid % did not occur during 4 days storage at -20 °C (Appendix J). Prior to

54

quantitative analysis, fats were removed from the freezer and allowed to thaw to ambient

temperature. All samples were placed in a water bath (65 ± 2 °C) briefly before analysis to melt

any solid fats.

3.7.3. Infrared determination of trans fat

All infrared analyses were performed with a Thermo Nicolet 6700 Fourier Transform

Infrared Spectrometer (FTIR) (Madison, WI, USA) equipped with deuterated triglycine sulfate

(DTGS) detector. FTIR operating parameters used were as follows: 4000-350 cm-1 wavelength

range with 32 scans collected and averaged; resolution of 4 cm-1, and Happ-Genzel apodization

functions. Oils were deposited using disposable pipettes onto the ZnSe Smart ARK ATR cell

until the oil completely covered the surface of the cell. The single beam sample absorbance

spectrum was collected and saved. A background was taken before each sample single beam

spectrum was collected. The crystal was cleaned between runs by rinsing with hexane and

wiping with soft lint-free tissue paper. This procedure was followed for the trans free oil,

calibration standards, and test samples.

3.7.4. Calculations

The trans free oil absorbance spectrum was subtracted from each calibration standard and

test sample spectrum to obtain horizontal baselines for area integrations. The wavenumbers from

1000 cm-1 to 900 cm-1 were selected from each standard and test sample spectrum. The area

under the 966 cm-1 peak between the region of 990 and 945 cm-1 was electronically integrated

using OMNIC 7.0 software. The best-fit first-order regression equation for the area vs. % trans

fat plot for the trans calibration standards was determined. The % trans fat for each unknown

test sample was interpolated by substituting the area for each test sample in the following

equation:

55

Trans fat as % of total fat = area – intercept slope

Results were expressed to the nearest 0.1%.

3.8. Statistical analysis

Values for total, saturated, polyunsaturated, monounsaturated, and trans fat were

expressed as the mean ± standard deviation of 3 replicates for each parameter. Differences

between the two methods were determined by Student’s t-test (α = 0.01) within individual

products.

4. Results

The values for total fat determined by the AOAC-GC method and the AHE-G method

were equivalent for all products (Table 2.2; p>0.01). It was observed that values for total,

saturated, polyunsaturated, and monounsaturated fat measured by the AOAC-GC and AHE-GC

methods were also equivalent in 11 of 12 products for each parameter (Table 2.3; p>0.01). The

statistical differences observed between the methods were 0.58% for total fat determination (for

one product out of the 12), 0.17% for saturated fat (one product), 0.12% and 0.16% for

polyunsaturated fat (two products) and 0.16% for monounsaturated fat (one product). The

standard error of the laboratory (SEL) or pooled standard deviation of the repeatability (ASTM,

1995) of the AOAC method for total fat is 0.33%; for saturated fat is 0.25%; for polyunsaturated

fat is 0.14%; and for monounsaturated fat is 0.33%. Thus, the differences are small and within

the accuracy usually encountered for AOAC Method 996.01 in this laboratory. They occur in a

different product for each of the four parameters and are, thus, not significant in the overall

evaluation of the method. For the GC determination of trans fat (Table 2.4), no significant

difference or trend was observed between methods for five of the seven samples. For the

56

remaining samples the differences were 0.05 and 0.44 %, and these differences are also within or

close to the accuracy expected of the modified AOAC method (SEL 0.19%).

AOAC 996.01 extracted oils and AHE extracted oils produced statistically equivalent

trans fat results, as determined by ATR-FTIR spectroscopy for two of the three samples.

However, the differences in values between methods were 0.9% (NS), 4.3% (NS), and 8.6%

(p<0.01) trans fat (Table 2.5). For all three of the cereal products tested, AHE mean values for

trans fat in extracted oil were less than AOAC 996.01 values. The largest difference was

observed in the medium fat cereal product. For oil extracted from the high fat product the FTIR

spectra were very similar for the two extraction methods (data not shown). For the medium fat

sample the FTIR spectrum for the oil extracted with the AOAC method had an extra peak and

larger peak area than the spectrum of the oil extracted with the AHE method (Figures 2.1 and

2.2). The spectra of the oil extracted from the low fat sample showed the same trends as those

from the medium fat sample but to a lesser extent (data not shown).

5. Discussion

An automated AHE lipid extraction system has been demonstrated to allow

determination of fat in cereal products gravimetrically with comparable accuracy to AOAC

Method 996.01. A variety of cereal products having a wide range of fat, sugar, fiber, and protein

content have been tested. The AHE-G method is a distinct advantage over the AOAC 996.01

method for determination of total fat due to its increased safety for the operator, the closed

automatic system and the ability to reuse solvents.

Traditionally total lipid gravimetric measurements have included some non-lipid moieties

that were freed during hydrolyzation and extracted along with the lipid, causing the

overestimation of total fat in foods. On the other hand, traditional crude fat measurements,

57

lacking hydrolysis, have incompletely extracted bound lipids, causing the underestimation of

total fat (Rader et al., 1995; Zou et al., 1999; Ranhotra et al., 1996; Ali et al., 1997). The AHE-

G method is an advantage over these methods in that it allows for the rinsing of water soluble,

non-lipid, moieties from the sample after acid hydrolysis. In theory, the removal of these non-

lipid materials from the sample should allow for a gravimetric measurement of total lipid to be

obtained that is an accurate estimation of total fat. In the cereal products tested, AHE-

gravimetric total fat results were statistically equivalent to AOAC 996.01 total fat results. The

completeness and selectiveness of the lipid extraction process in the AHE method appears to

allow for the accurate estimation of total fat through gravimetric means.

The AHE system is also adequate for the extraction of lipid for the GC analysis of a wide

variety of cereal products for the determination of total, saturated, polyunsaturated and

monounsaturated fat. Although the GC part of the analysis is identical, the AHE system of lipid

extraction has the same advantages of operator safety, with a closed system, lower labor input

and ability to regenerate the solvents. There was a trend in nine out of twelve samples towards

lower values for total fat using the AHE-GC method, but not for the AHE-G method. The trend

may be due to a minute loss of lipid in the extra transfer step for triacylglycerides and fatty acids

with the AHE-GC method, i.e. from aluminum extraction cups to boiling flasks just prior to

saponification and methylation of fatty acids. The modified AOAC 996.01 and AHE-GC trans

fat results were very similar with no trends observed in direction of results, supporting the

conclusion that the AHE method of lipid extraction performs comparably to the standard AOAC

method.

The adequacy of the AHE system has been evaluated for the extraction of oils from

cereal products in preparation for ATR-FTIR analysis of trans fat. Although the AOAC 996.01

58

extracted oils and AHE extracted oils produced statistically equivalent ATR-FTIR trans fat

results in two of the three products, the differences between methods in the medium fat product

and low fat product were 8.6% and 4.3%, respectively, a variation too large for accurate nutrition

labeling of trans fat. The AHE and AOAC 996.01 FTIR trans fat values have a substantial

amount of variation within each oil extraction method and between oil extraction methods, i.e.,

high standard deviations.

During FT-IR collection, oils were warmed directly before scanning. The samples were

then scanned on an ATR device at room temperature. Solidification of some oils could have

occurred and possibly affected the results causing greater variation. Therefore, it may be useful

in the future to collect the data using a heated ATR-cell, insuring that the oil is completely

melted during data collection.

AHE mean determinations were less than AOAC 996.01 values for all cereal products

tested. For two samples (the medium and low fat samples) the FTIR spectra differed

considerably in shape and area between the two methods. The difference between the methods

was not observed in the high fat sample suggesting that the two extraction methods may treat

trans isomers in individual samples differently. Although no difference in total trans occured

between the samples as measured by GC analysis, there may be a difference in distribution of

isomers, which would affect the IR spectra. Thus further study is required to determine if the

AHE system causes changes in the trans isomers. The accuracy of ATR-FTIR trans fat

prediction following the AHE extraction technique may be suitable for rapid screening of

products for trans content rather than for nutrition labeling.

Future work could include the careful study of gas chromatographic traces of AHE and

AOAC 996.01 extracted oils from a variety of products, searching for any differences between

59

the trans fatty acid profiles. Mass spectrometry and GC-GC techniques may also be used in the

further identification of trans isomers or interfering substances.

60

6. References

Anonymous. (2006). Soxtec™ systems, http://www.foss.dk/solutions/productsdirect/soxtec

systems.aspx, November 2006.

AOAC (2000a). Official methods of analysis. (996.01) Fat (total, saturated, unsaturated, and

monounsaturated) in cereal products. (17th ed.). USA: AOAC International.

AOAC (2000b). Official methods of analysis. (2000.10) Determination of total isolated trans

unsaturated fatty acids in fats and oils (17th ed.). USA: AOAC International.

ASTM (1995). Standard practices for infrared multivariate quantitative analysis. American

Society for Testing and Materials, E1655-94. Conshohocken, PA.

Code of Federal Regulations. (2006). Title 21, Volume 2, Part 101.9. Washington, DC, USA:

US Government Printing Office.

de Castro, M., & Garcia-Ayuso, L. (1998). Soxhlet extraction of solid materials: an outdated

technique with a promising innovative future. Analytica Chemica Acta, 369, 1-10.

Federal Register. (2003). Food labeling: trans fatty acids in nutrition labeling, nutrient content

claims, and health claims. Vol. 68, 41433-41506.

Helaleh, M., Al-Omair, A., Ahmed, N., & Gevao, B. (2005). Quantitative determination of

organochlorine pesticides in sewage sludges using Soxtec, Soxhlet and pressurized liquid

extractions and ion trap mass-mass spectrometric detection. Analytical and Bioanalytical

Chemistry, 382, 1127-1134.

Mossoba, M., Kramer, J., Delmonte, P., Yurawecz, M., & Rader, J. (2003). Official methods for

the determination of trans fat. Champaign, Ill: AOCS Press.

61

Ngehngwainbi, J., Lin, J., & Chandler, A. (1997). Determination of total, saturated, unsaturated,

and monounsaturated fats in cereal products by acid-hydrolysis and capillary gas-

chromatography—collaborative study. Journal of AOAC International, 80, 359-372.

Ratnayake, W. (2004). Overview of methods for the determination of trans fatty acids by gas

chromatography, silver-ion thin-layer chromatography, silver-ion liquid chromatography,

and gas chromatography/mass spectrometry. Journal AOAC International, 87, 523-539.

Scarbrough, F. (1995). Perspectives on nutrition labeling education act. In R. Shapiro,

Nutrition labeling handbook, (pp. 29-52). New York: N. Dekker.

62

Table 2.1. Cereal products used in the study Percentage1

Product Group Product Grains2

Total fat

Carbo-hydrate Sugars Protein

Dietary fiber

Corn chips* Corn 37.9 51.7 0.0 6.9 3.4

Snack mix* wheat, barley, rye 20.0 66.7 3.3 10.0 3.3

Snack products

Pretzels wheat, barley 3.6 82.1 10.7 7.1 3.6

Crackers with peanut butter*

wheat, barley 25.6 59.0 10.3 10.3 2.6

Oatmeal cookies with raisins*

wheat, oats 21.4 64.3 28.6 7.1 3.6

Cookies and crackers

Chocolate wafer snacks Wheat 8.7 87.0 39.1 4.3 4.3 Pie crust mix* Wheat 35.0 65.0 0.0 5.0 0.0 All-purpose baking mix* Wheat 15.0 65.0 2.5 7.5 0.0

Baking mixes

White cake mix Wheat 8.1 81.4 48.8 2.3 2.3

Granola oats, wheat 12.5 72.9 25.0 10.4 6.3

Toaster pastries*

wheat, corn 9.6 71.2 30.8 3.8 1.9

Breakfast products

Corn crunch corn, oats 5.5 85.2 44.4 3.7 3.7 1Percentage is based on nutrition label declarations. 2Grains are listed in order of predominance in the products. *Products used for trans fat analysis.

63

Table 2.2. Measurement of total fat (%) in cereal products by the standard GC method and by an automated gravimetric method1

Product group Product AOAC-GC2 AHE-G2

Corn chips 29.59 ± 0.21 30.28 ± 0.19 Snack mix 20.48 ± 0.16 20.32 ± 0.15

Snack products

Pretzels 4.82 ± 0.05 4.62 ± 0.08 Crackers with peanut butter 24.75 ± 0.26 25.00 ± 0.37 Oatmeal cookies with raisins 22.12 ± 0.94 21.80 ± 0.70

Cookies and crackers

Chocolate wafer snacks 10.24 ± 0.11 9.90 ± 0.22 Pie crust mix 45.35 ± 0.89 42.89 ± 0.55 All-purpose baking mix 16.36 ± 0.20 16.70 ± 0.17

Baking mixes

White cake mix 8.56 ± 0.14 8.52 ± 0.14 Granola 13.26 ± 0.14 12.64 ± 0.26 Toaster pastries 10.49 ± 0.13 10.53 ± 0.12

Breakfast products

Corn crunch 6.20 ± 0.02 6.02 ± 0.06 1Values are means ± SD, for triplicate analyses performed on the same day. No significant difference between methods, students paired t-test, p>0.01.

2AOAC-GC, oil extracted from products by AOAC 996.01 prior to GC analysis; AHE-G, oil extracted by automated hydrolysis and extraction system prior to gravimetric analysis.

64

Table 2.3.

Determination of total, saturated, polyunsaturated, and monounsaturated fat (%) extracted from cereal products by the standard method and an automated method1

Component Product group Product AOAC-GC2 AHE-GC2

Total fat Corn chips 29.59 ± 0.21 29.81 ± 0.47 Snack mix 20.48 ± 0.16 20.21 ± 0.26

Snack products

Pretzels 4.82 ± 0.05 4.66 ± 0.11

Crackers with peanut butter 24.75 ± 0.26 24.15 ± 0.13 Oatmeal cookies with raisins 22.12 ± 0.94 21.36 ± 0.05

Cookies and crackers Chocolate wafer snacks 10.24 ± 0.11 a 9.66 ± 0.02 b

Pie crust mix 45.35 ± 0.89 45.45 ± 0.57 All-purpose baking mix 16.36 ± 0.20 15.85 ± 0.26

Baking mixes White cake mix 8.56 ± 0.14 8.42 ± 0.09

Granola 13.26 ± 0.14 13.03 ± 0.06 Toaster pastries 10.49 ± 0.13 10.60 ± 0.15

Breakfast products

Corn crunch 6.20 ± 0.02 5.91 ± 0.08 Saturated fat Corn chips 4.78 ± 0.11 5.02 ± 0.14 Snack mix 3.74 ± 0.02 3.84 ± 0.07

Snack products

Pretzels 0.86 ± 0.03 0.85 ± 0.12 Crackers with peanut butter 4.46 ± 0.05 a 4.29 ± 0.03 b

Oatmeal cookies with raisins 4.64 ± 0.20 4.49 ± 0.02

Cookies and crackers Chocolate wafer snacks 1.25 ± 0.08 1.12 ± 0.002

Pie crust mix 11.03 ± 0.33 11.32 ± 0.10 All-purpose baking mix 4.03 ± 0.05 3.84 ± 0.10

Baking mixes White cake mix 3.36 ± 0.08 3.23 ± 0.10

Granola 7.30 ± 0.03 7.12 ± 0.04 Toaster pastries 2.34 ± 0.02 2.31 ± 0.03

Breakfast products

Corn crunch 3.93 ± 0.03 3.84 ± 0.06 Corn chips 3.09 ± 0.01 3.06 ± 0.04 Snack mix 4.93 ± 0.03 4.83 ± 0.08

Polyunsaturated fat

Snack products

Pretzels 2.63 ± 0.05 2.51 ± 0.07 Crackers with peanut butter 5.05 ± 0.05 4.91 ± 0.01 Oatmeal cookies with raisins 1.91 ± 0.07 1.83 ± 0.01

Cookies and crackers Chocolate wafer snacks 3.00 ± 0.02 a 2.84 ± 0.01 b

Pie crust mix 5.27 ± 0.13 5.44 ± 0.08 All-purpose baking mix 2.56 ± 0.27 2.31 ± 0.02

Baking mixes White cake mix 3.22 ± 0.04 3.22 ± 0.02

65

Granola 2.07 ± 0.03 2.05 ± 0.01 Toaster pastries 1.11 ± 0.02 1.09 ± 0.02

Breakfast products

Corn crunch 1.07 ± 0.01 a 0.95 ± 0.01 b

Corn chips 20.46 ± 0.14 20.43 ± 0.31 Snack products Snack mix 10.91 ± 0.11 10.65 ± 0.10

Monounsaturated fat

Pretzels 1.11 ± 0.02 1.09 ± 0.03 Crackers with peanut butter 14.15 ± 0.15 13.89 ± 0.10 Oatmeal cookies with raisins 14.59 ± 0.63 14.10 ± 0.04

Cookies and crackers Chocolate wafer snacks 5.54 ± 0.07 5.27 ± 0.01

Pie crust mix 25.97 ± 0.73 26.70 ± 0.38 All-purpose baking mix 9.22 ± 0.12 9.01 ± 0.13

Baking mixes White cake mix 1.59 ± 0.02 1.59 ± 0.02

Granola 3.20 ± 0.08 3.19 ± 0.02 Toaster pastries 6.58 ± 0.12 a 6.74 ± 0.10 b

Breakfast products

Corn crunch 0.87 ± 0.01 0.81 ± 0.004

1Values are means ± SD, for triplicate analyses performed on the same day; Different superscript letters in a row indicate a statistically significant difference (p<0.01), students paired t-test. 2AOAC-GC, oil extracted from products by AOAC 996.01 prior to GC analysis; AHE-GC, oil extracted by automated hydrolysis and extraction system prior to GC analysis.

66

Table 2.4. Determination of trans fat (%) in oil extracted from cereal products by the standard method and an automated method1

Lipid extraction method Product group Product AOAC-

modified GC2AHE-

modified GC2 Snack products Corn chips 12.56 ± 0.08 12.56 ± 0.22 Snack mix 5.15 ± 0.08 5.00 ± 0.17

Crackers with peanut butter 2.87 ± 0.04 a 2.82 ± 0.04 bCookies and crackers Oatmeal cookies with raisins 8.18 ± 0.35 7.97 ± 0.11 Baking mixes Pie crust mix 15.47 ± 0.33 15.58 ± 0.14 All-purpose baking mix 4.41 ± 0.09 a 4.85 ± 0.06 b

Breakfast products Toaster pastries 3.12 ± 0.06 3.18 ± 0.07 1Values are means ± SD, for triplicate analyses performed on the same day; Different superscript letters in a row indicate a statistically significant difference (p<0.01), students paired t-test. 2AOAC-modifed GC, oil extracted from products by AOAC 996.01 prior to GC analysis optimized for trans fat; AHE-GC, oil extracted by automated hydrolysis and extraction system prior to GC analysis optimized for trans fat.

67

Table 2.5 ATR-FTIR determination of trans fat (%) in oil extracted from cereal products by the standard method and an automated method1

Sample Lipid extraction method AOAC2 AHE2

High fat sample 36.3 ± 0.7 35.4 ± 1.6

Medium fat sample 37.8 ± 3.6 a 29.2 ± 1.6 b

Low fat sample 26.4 ± 1.6 22.1 ± 2.8

1Trans fat values are expressed as a percentage of total fat and are means ± SD for n = 3 replications; Different superscript letters within a row indicate a statistically significant difference (p<0.01). 2AOAC, extraction of oil by AOAC 996.01 method; AHE, extraction of oil by automated hydrolysis and extraction system.

68

966 cm-1 966 cm-1

Fig. 2.1. ATR-FTIR spectra of oil extracted from a medium fat cereal product by the standard method (A) and an automated method (B).

1000 970 940

0.5

1000 970 940

0.5

A

990 cm-1

Absorbance

Frequency (cm-1)

Absorbance

Frequency (cm-1)

990 cm-1945 cm-1

B

945 cm-1

69

CHAPTER 3

SUMMARY AND CONCLUSIONS

70

The utility of a semi-automated method for the quantification of total, saturated,

polyunsaturated, monounsaturated, and trans fat in cereal products was investigated. AOAC

996.01 is a widely accepted method that can be used to quantify fat in cereal products for

nutrition labeling purposes. Although the method is accurate and repeatable, AOAC 996.01 and

similar methods are extremely laborious, and time-consuming procedures that require the use of

hazardous solvents and chemicals.

An automated hydrolysis and extraction (AHE) system commercially available, the

Soxtec® Soxcap™ 2047 and Auto Fat Extraction system, offers an alternative to manual

hydrolysis and extraction methods, and has many advantages. Some of the advantages gained by

use of the system are: the operator has less exposure to fumes and solvents; the operator’s time

and attention may be directed towards other activites during extraction; 80% of the solvent can

be recovered and reused; and the results are less likely to be affected by operator error (Helaleh

et al., 2005a). The design of the AHE hydrolysis unit allows for the rinsing of non-lipid water-

soluble moieties from the hydrolyzed sample, removing non-lipid components that might,

otherwise, cause overestimation of gravimetric total fat. In theory, this should provide for the

accurate determination of total fat without use of a gas chromatographic step. However,

recovery of the lipid, and subsequent saponification and methylation, would allow for

determination of total, saturated, polyunsaturated, monounsaturated, and trans fat by capillary

GC analysis.

Oils from twelve diverse cereal products were extracted using the AHE system and

AOAC Method 996.01. The extracted oils from both methods were subjected to GC analysis for

the quantification of total, saturated, polyunsaturated, monounsaturated, and trans fat and the

results compared. In addition the oil extracted by the AHE system was also weighed for

71

gravimetric determination of total fat. Oil extracted from a high fat, medium fat, and low fat

sample by both the AHE system and AOAC 996.01 methods was subjected to ATR-FTIR

analysis for the spectroscopic quantification of trans fat and the results for the two fat extraction

techniques compared.

It was found that the AHE-GC results for total, saturated, polyunsaturated, and

monounsaturated fat were similar to those obtained from AOAC Method 996.01. Likewise, the

AHE-GC trans fat results were comparable to the trans fat results obtained using the modified

AOAC Method 996.01. The AHE-gravimetric total fat measurements were very close to the

total fat results obtained from the AOAC method and the AHE-GC method. For two of the three

samples tested using ATR-FTIR spectroscopy, the trans fat results for oils extracted using the

AHE system were equivalent to those obtained from oils extracted using the AOAC method. In

the third sample tested using ATR-FTIR, the AOAC-extracted oil trans fat results were

significantly larger than the AHE-extracted oil results. Also for two of the three samples the

FTIR spectra differed considerably in shape and area between the two methods. The results

suggest that the two extraction methods may treat trans isomers in samples differently.

The AHE system to determine total fat, lipid classes, and trans fat in cereal products

appears to be of sufficient accuracy for nutrition labeling purposes and has potential for use in

food analysis laboratories, in industry and in regulatory settings. Its convenience, accuracy,

reliability, and safety make it an attractive alternative to manual extraction methods, such as

AOAC 996.01. The AHE-GC method yields results very similar to standard method total,

saturated, polyunsaturated, monounsaturated, and trans fat results. The AHE system can be used

to gravimetrically determine total fat values that are similar to total fat values obtained using

AOAC 996.01, eliminating the time-consuming and potentially hazardous saponification and

72

methylation of fatty acids and the need for GC analysis. Further investigation is needed to

determine if the AHE system is suited for the extraction of oils from cereal samples to determine

trans fat content by ATR-FTIR spectroscopy.

73

APPENDICES

74

Appendix A

EXPERIMENTAL CEREAL FOOD PRODUCTS

Snack Products

• Cornitos Original Corn Chips: The Kroger Co., Cincinnati, OH

• Gardettos Original Recipe Snack Mix: General Mills, Inc., Minneapolis, MN

• Rold Gold Honey Wheat Braided Twists Pretzels: Frito-Lay, Inc., Plano, TX

Cookies and Crackers

• Austin Peanut Butter Crackers: Kellogg Co., Battle Creek, MI

• Country Style Oatmeal Cookies Baked with Raisins: Kellogg Co., Battle Creek, MI

• Oreo Thin Crisps Baked Chocolate Wafer Snacks: Kraft Foods North America, East

Hanover, NJ

Baking Mixes

• Betty Crocker Pie Crust Mix: General Mills, Inc., Minneapolis, MN

• Bisquick Original All-Purpose Baking Mix: General Mills, Inc., Minneapolis, MN

• Pillsbury Moist Supreme Premium Cake Mix: The J.M. Smucker Co., Orrville, OH

Breakfast Products

• Quaker Granola Oats & Honey: The Quaker Oats Co., Chicago, IL

• Toaster Treats: The Kroger Co., Cincinnati, OH

• Cap’n Crunch Sweetened Corn and Oat Cereal: The Quaker Oats Co., Chicago, IL

75

Appendix B

MACRONUTRIENT COMPOSITION OF EXPERIMENTAL CEREAL PRODUCTS BASED ON NUTRITION LABEL DECLARATIONS

Co sition mpo

P kcal/g ercentage

Product Group Product

Cal

orie

s

Tot

al F

at

Sat

urat

ed F

at

Tra

ns F

at

Tota

l C

arbo

hydr

ate

Die

tary

Fib

er

Suga

rs

Prot

ein

Snack products Corn Chips 5.5 37.9 5.2 15.5 51.7 3.4 0.0 6.9

Snack Mix 5.0 20.0 3.3 6.7 66.7 3.3 3.3 10.0

Pretzels 3.9 3.6 0.0 0.0 82.1 3.6 10.7 7.1 Cookies and crackers

Crackers with Peanut Butter 5.1 25.5 3.8 5.1 59.0 2.6 10.3 10.3

Oatmeal Cookies with Raisins 4.6 21.4 5.4 8.9 64.3 3.6 28.6 7.1

Chocolate wafer snacks 4.4 8.7 0.0 0.0 87.0 4.3 39.1 4.3

Baking Mixes Pie Crust Mix 5.5 35.0 10.0 12.5 65.0 0.0 0.0 5.0

All-Purpose Baking Mix 4.0 15.0 3.8 3.8 65.0 0.0 2.5 7.5

White Cake Mix 4.0 8.1 3.5 0.0 81.4 2.3 48.8 2.3

Breakfast products Granola 4.4 12.5 8.3 0.0 72.9 6.3 25.0 10.4 Toaster

Pastries 3.9 9.6 1.9 1.9 71.2 1.9 30.8 3.8

Corn Crunch 4.1 5.5 3.7 0.0 85.2 3.7 44.4 3.7

76

Appendix C

LIST OF INGREDIENTS FOR EXPERIMENTAL CEREAL PRODUCTS BASED ON NUTRITION LABEL DECLARATIONS

Product Group Product Ingredients

Corn Chips Corn, Partially Hydrogentated Corn Oil, Salt Snack Mix Enriched flour bleached, partially hydrogenated soybean oil, rye

flour. Contains 2% or less: salt, sesame seed, monosodium glutamate, yeast, sugar, corn syrup, malt, worcestershire sauce, maltodextrin, malted corn syrup, baking soda, onion, garlic, wheat gluten, corn starch, color added, sodium-stearoyl lactylate, mono- and diglycerides, caraway, disodium guanylate, disodium inosinate, nonfat milk, soy flour, BHT

Snack products

Pretzels Enriched flour bleached, whole wheat flour, honey, sugar, corn oil, salt, ammonium bicarbonate, sodium bicarbonate, and malted barley flour

Cookies and crackers

Crackers with Peanut Butter

Enriched flour, peanut butter, partially hydrogenated soybean and/or cottonseed oil, sugar, high fructose corn syrup, dextrose, salt. Contains 2% or less: malted barley flour, leavening, soy lecithin, cornstarch, sodium sulfite, cheddar cheese, yellow #6, red pepper, buttermilk, whey, disodium phosphate

Oatmeal Cookies with Raisins

Enriched flour, sugar, partially hydrogenated soybean oil, oats, raisins. Contains 2% or less: high fructose corn syrup, molasses, salt, baking soda, cinnamon, natural flavors, soy lecithin, nonfat dry milk, eggs

Chocolate wafer snacks

Enriched flour, sugar, high fructose corn syrup, high oleic canola oil, cocoa, cornstarch, leavening, artificial color, salt, powdered sugar, artificial flavor

Baking mixes

Pie Crust Mix Enriched flour bleached, partially hydrogenated soybean and/or cottonseed oil, modified corn starch, salt, corn syrup solids, sodium caseinate, color added

All-Purpose Baking Mix

Enriched flour bleached, partially hydrogenated soybean and/or cottonseed oil, leavening, dextrose, salt

White Cake Mix

Sugar, enriched bleached flour, partially hydrogenated soybean oil, wheat starch, baking powder. Contains 2% or less: propylene glycol monoesters, dextrose, corn starch, salt, cellulose, mono- and diglycerides, xanthan gum, artificial flavor, cellulose gum, polysorbate 60, nonfat milk, soy lecithin, TBHQ, citric acid

Breakfast products

Granola Whole grain rolled oats, whole grain rolled wheat, brown sugar, coconut oil, nonfat dry milk, almonds, whey, whey protein concentrate, honey, natural flavors, sunflower oil

77

Toaster Pastries

Enriched wheat flour, blueberry filling (corn syrup, high fructose corn syrup, fruit from concentrate, glycerine, modified corn starch, precooked corn meal, natural and artificial flavor, citric acid, salt, red 40, blue 1), sugar, partially hydrogenated soybean and/or cottonseed oils, corn syrup, whey leavening, salt, titanium dioxide, turmeric, potassium sorbate, gelatin, modified soy protein, vitamin A palmitate, reduced iron, niacinamide, color added, pyridoxine hydrochloride, riboflavin, thiamin mononitrate, folic acid, red 40, blue 1

Corn Crunch Corn flour, sugar, oat flour, brown sugar, coconut oil, salt, niacinamide, yellow 5, reduced iron, zinc oxide, yellow 6, thiamin mononitrate, BHT, pyridoxine hydrochloride, riboflavin, folic acid

78

Appendix D

FLOW CHART FOR AOAC 996.01 DETERMINATION OF TOTAL FAT, LIPID CLASSES, AND TRANS FAT

Ground sample

Acid hydrolysis HCl, 80 ± 2 °C

Extraction of fatty acids diethyl ether and petroleum ether

Separation of organic solvent phase and aqueous phases and evaporation of solvents

Methylation of fatty acids NaOH, BF3

Suspension of fatty acid methyl esters (FAME) into heptane solvent

GC quantification of FAME total, saturated, polyunsaturated, and monounsaturated

fat AOAC-GC

GC quantification of FAME trans fat

AOAC modified GC

79

Appendix E

FLOW CHART FOR THE AHE DETERMINATION OF TOTAL FAT, LIPID CLASSES AND TRANS FAT

Ground sample

Automated acid hydrolysis HCl, boiling

Freeze drying

Automated extraction and evaporationpetroleum ether

Methylation of fatty acids NaOH, BF3

Suspension of FAME into heptane solvent

GC quantification of FAME total, saturated, polyunsaturated, and monounsaturated fat

GC quantification of FAME trans fat

Gravimetric measurement of total fat

80

Appendix F

EQUATIONS FOR TOTAL FAT, LIPID CLASSES, AND TRANS FAT DETERMINATION

• Individual fatty acid response factor (Ri):

Ri= Psi × WsC13:0

PsC13:0 × Wsis Psi = peak area of a fatty acid methyl ester (FAME), i, in injected standard mixture WsC13:0 = mg of C13:0 FAME in FAME standard mixture PsC13:0 = peak area of C13:0 FAME in injected FAME standard mixture Wsis = mg of a FAME, i, in standard FAME mixture

• Amount of each fatty acid (g), expressed as methyl esters, in test sample (FME):

FME= Pti × WtC13:0 × 1.006 × 1000

PtC13:0 Ri

Pti = peak area of a fatty acid methyl ester (FAME), i, in injected test sample PtC13:0 = peak area of internal standard C13:0, in injected test sample WtC13:0 = mg of internal standard C13:0, in test sample

• Amount of each fatty acid (g), expressed as triglycerides, in test sample (FTG):

FTG = FME × fTG fTG = theoretical conversion factor constant (Appendix G)

• Amount of each fatty acid (g), expressed as fatty acids, in test sample (FFA):

FFA = FME × fFA fFA = theoretical conversion factor constant (Appendix G)

• Percentage total fat, expressed as triglycerides, in test sample (Total Fat (%)):

Total Fat = Σ FTG × 100 W

W = weight (g) of test sample

• Percentage saturated fat, expressed as fatty acids, in test sample (Saturated Fat (%)):

Saturated Fat = Σ saturated (FFA) × 100

81

W • Percentage polyunsaturated fat, expressed as fatty acids, in test sample (Polyunsaturated

Fat (%)):

Polyunsaturated Fat = Σ polyunsaturated (FFA) × 100 W

• Percentage monounsaturated fat, expressed as fatty acids, in test sample

(Monounsaturated Fat (%)):

Monounsaturated Fat = Σ monounsaturated (FFA) × 100 W

• Percentage trans fat, expressed as fatty acids, in test sample (Trans Fat (%)):

Trans Fat = Σ trans (FFA) × 100 W

• Percentage total lipid (TL (%)): TL = Wpost − Wpre − WC13:0 × 100

W

Wpost = weight (g) of aluminum cup after extraction Wpre = weight (g) of aluminum cup before extraction WC13:0 = weight (g) of internal standard W = initial weight (g) of test sample

• Percentage dry matter (DM (%)):

DM = WSD − WPD × 100 WS WSD = weight (g) of test sample and aluminum weigh pan after drying WPD= weight (g) of aluminum weigh pan dried WS = initial weight (g) of test sample Reference:

AOAC (2000). Official methods of analysis. (996.01) Fat (total, saturated, unsaturated, and

monounsaturated) in cereal products. (17th ed.). USA: AOAC International.

82

Appendix G

THEORETICAL CONVERSION FACTORS FOR THE DETERMINATION OF EACH

FATTY ACID EXPRESSED AS TRIGLYCERIDES AND/OR FATTY ACID

Fatty Acid fTG fFA8:0 0.9915 0.9114 10:0 0.9928 0.9247 12:0 0.9937 0.9346 14:0 0.9945 0.9421 15:0 0.9948 0.9453 16:0 0.9950 0.9481

16:1 cis 0.9950 0.9477 17:0 0.9953 0.9507 18:0 0.9955 0.9530

18:1 cis or trans 0.9955 0.9527 18:2 0.9954 0.9524 18:3 0.9954 0.9520 20:0 0.9959 0.9570 20:1 0.9959 0.9568 22:0 0.9962 0.9604 22:1 0.9962 0.9602

Reference:

AOAC (2000). Official methods of analysis. (996.01) Fat (total, saturated, unsaturated, and

monounsaturated) in cereal products. (17th ed.). USA: AOAC International.

83

Appendix H

STABILITY OF FAT IN GROUND PRODUCTS STORED AT -20 °C FOR 7 DAYS DETERMINED BY MODIFIED AOAC METHOD 996.011

Day 1 Day 2 Day 4 Day 7 Total fat, % High fat

sample 27.78 ± 0.32 28.43 ± 0.28 28.71 ± 0.10 27.90 ± 0.23

Medium fat sample 16.32 ± 0.09 16.17 ± 0.01 16.29 ± 0.04 16.17 ± 0.15

Saturated fat, % High fat sample 4.48 ± 0.05 4.76 ± 0.20 4.64 ± 0.02 4.51 ± 0.04

Medium fat sample 4.09 ± 0.03 4.05 ± 0.00 4.08 ± 0.004 4.03 ± 0.04

Polyun- saturated fat, %

High fat sample 3.07 ± 0.02 3.10 ± 0.01 3.15 ± 0.04 3.02 ± 0.05

Medium fat sample 2.41 ± 0.01 2.40 ± 0.00 2.40 ± 0.01 2.40 ± 0.02

Monoun-saturated fat, %

High fat sample 19.12 ± 0.23 19.44 ± 0.05 19.78 ± 0.04 19.26 ± 0.13

Medium fat sample 9.20 ± 0.05 9.11 ± 0.01 9.18 ± 0.03 9.12 ± 0.08

Trans fat, % High fat sample 11.76 ± 0.15 12.11 ± 0.09 12.09 ± 0.10 11.90 ± 0.09

Medium fat sample 4.52 ± 0.01 4.49 ± 0.01 4.54 ± 0.07 4.45 ± 0.02

1Values are means ± SD, for duplicate analyses.

84

Appendix I

STABILITY OF FAT IN FAME STORED AT -20 °C FOR 7 DAYS DETERMINED BY MODIFIED AOAC METHOD 996.011

Day 1 Day 2 Day 4 Day 7 Total fat, % High fat

sample 27.78 ± 0.32 27.82 ± 0.34 27.76 ± 0.36 27.68 ± 0.35

Medium fat sample 16.32 ± 0.09 16.32 ± 0.09 16.32 ± 0.09 16.29 ± 0.17

Saturated fat, %

High fat sample 4.48 ± 0.05 4.49 ± 0.06 4.49 ± 0.06 4.48 ± 0.06

Medium fat sample 4.09 ± 0.03 4.09 ± 0.03 4.08 ± 0.03 4.07 ± 0.03

Polyun- saturated fat, %

High fat sample 3.07 ± 0.03 3.08 ± 0.03 3.04 ± 0.04 3.02 ± 0.05

Medium fat sample 2.28 ± 0.19 2.41 ± 0.01 2.40 ± 0.01 2.40 ± 0.01

Monoun-saturated fat, %

High fat sample 19.12 ± 0.32 19.14 ± 0.34 19.13 ± 0.36 19.08 ± 0.35

Medium fat sample 9.19 ± 0.07 9.19 ± 0.06 9.20 ± 0.06 9.22 ± 0.12

Trans fat, % High fat sample 11.76 ± 0.15 11.77 ± 0.16 11.78 ± 0.15 11.73 ± 0.15

Medium fat sample 4.52 ± 0.01 4.46 ± 0.02 4.52 ± 0.01 4.51 ± 0.07

1Values are means ± SD, for duplicate analyses.

85

Appendix J

STABILITY OF COMPONENTS OF EXTRACTED OIL STORED AT -20 °C FOR 4 DAYS DETERMINED BY MODIFIED AOAC METHOD 996.011

Day 1 Day 2 Day 4 Total fat, % High fat sample 27.78 ± 0.32 28.17 ± 0.28 28.53 ± 0.23 Medium fat sample 16.32 ± 0.09 16.17 ± 0.05 16.29 ± 0.15 Saturated fat, % High fat sample 4.48 ± 0.05 4.56 ± 0.04 4.62 ± 0.03 Medium fat sample 4.09 ± 0.03 4.04 ± 0.01 4.08 ± 0.04 Polyunsaturated fat, % High fat sample 3.07 ± 0.02 3.08 ± 0.04 3.12 ± 0.04 Medium fat sample 2.41 ± 0.01 2.39 ± 0.01 2.39 ± 0.02 Monounsaturated fat, % High fat sample 19.12 ± 0.23 19.41 ± 0.19 19.66 ± 0.15 Medium fat sample 9.20 ± 0.05 9.12 ± 0.03 9.19 ± 0.08 Trans fat, % High Fat Sample 11.76 ± 0.15 11.90 ± 0.11 12.11 ± 0.06

Medium Fat Sample 4.52 ± 0.01 4.47 ± 0.05 4.47 ± 0.04

1Values are means ± SD, for duplicate analyses.

86