<%BANNER%>

The Incorporation of Chia (Salvia hispanica Lamiaceae) into Baked Food Products

Permanent Link: http://ufdc.ufl.edu/UFE0042223/00001

Material Information

Title: The Incorporation of Chia (Salvia hispanica Lamiaceae) into Baked Food Products
Physical Description: 1 online resource (116 p.)
Language: english
Creator: Lewis, Devin
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: alpha, chia, fact, hedonic, linolenic, salvia
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: THE INCORPORATION OF CHIA (SALVIA HISPANICA LAMIACEAE) SEEDS INTO BAKED FOOD PRODUCTS Chia is an herbaceous summer annual which is botanically known as Salvia hispanica Lamiaceae. The seed of this annual is harvested in the fall and has been consumed in Southern Mexico and Central America for generations. Once a staple food of ancient Mesoamerican civilizations, chia has currently emerged as a high source of polyunsaturated fatty acids. The oil of the chia seed has an omega- 3 fatty acid content of up to 68% versus the 57% found in flax seed. The fatty acid contained in chia, ?-linolenic acid (ALA), is metabolized in the body to produce eicosapentanoic and docosahexanoic acids, respectively, and has been found to reduce the risks of coronary heart disease in humans. The objectives of this study were 1) To demonstrate similar consumer acceptance between control products and those incorporated with chia seeds, 2) To investigate which form of chia seed withstands processing and baking best, defined as retaining the highest amount of the chia seed?s omega-3 character. Consumer acceptance data was obtained through the use of hedonic testing at UF/IFAS FSHN taste panel. On separate occasions consumers were asked to rate and rank a cookie or muffin sample, one control and two treated samples (25% and 40% ground chia muffin; 15%ground/15% whole seed and 5% ground/10%whole chia cookie) . Panelist used a nine-point hedonic scale to rate the following attributes: overall acceptance, texture, flavor, and appearance. The hedonic scale was anchored with 1=?dislike extremely? and 9=?like extremely?. The nine-point Food Action rating scale (FACT) was used to determine the consumption attitudes of the panelist. The anchors for the FACT scale were 1=?I would eat this only if forced? and 9=?I would eat this at every opportunity?. Analysis of variance (ANOVA) was used to analyze panelist data, with means separation using Tukey?s HSD test. There was no significant (p < 0.05) difference between the control and the 25% chia muffin in overall acceptability. The same result was found for the cookie control and the 5% ground/10%whole chia cookie sample. Subsequent fatty acid analysis was performed on the muffin and cookie samples, but including the chia in whole, ground, and combination forms at the aforementioned chia levels. It was found that the highest content of ?-linolenic acid was obtained in both ground seed cookie and muffin samples. The muffin and cookie samples were also subjected to water activity, L*a*b* color, and textural analysis, to measure the chia product physical properties versus those of the control samples. Significant differences between the muffin treatments and the control muffin where found for the appearance attribute, L and b color-values, hardness, springiness, and cohesiveness textural attributes. The control cookie and the treatment cookies had significant differences for all attributes. It was found that the ground muffin and ground cookie treatments retained the most omega-3 character of the tested treatments. It is the position of this study that the chia seed is a viable ingredient in food products to be consumed by the public.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Devin Lewis.
Thesis: Thesis (M.S.)--University of Florida, 2010.
Local: Adviser: Goodrich, Renee M.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042223:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042223/00001

Material Information

Title: The Incorporation of Chia (Salvia hispanica Lamiaceae) into Baked Food Products
Physical Description: 1 online resource (116 p.)
Language: english
Creator: Lewis, Devin
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: alpha, chia, fact, hedonic, linolenic, salvia
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: THE INCORPORATION OF CHIA (SALVIA HISPANICA LAMIACEAE) SEEDS INTO BAKED FOOD PRODUCTS Chia is an herbaceous summer annual which is botanically known as Salvia hispanica Lamiaceae. The seed of this annual is harvested in the fall and has been consumed in Southern Mexico and Central America for generations. Once a staple food of ancient Mesoamerican civilizations, chia has currently emerged as a high source of polyunsaturated fatty acids. The oil of the chia seed has an omega- 3 fatty acid content of up to 68% versus the 57% found in flax seed. The fatty acid contained in chia, ?-linolenic acid (ALA), is metabolized in the body to produce eicosapentanoic and docosahexanoic acids, respectively, and has been found to reduce the risks of coronary heart disease in humans. The objectives of this study were 1) To demonstrate similar consumer acceptance between control products and those incorporated with chia seeds, 2) To investigate which form of chia seed withstands processing and baking best, defined as retaining the highest amount of the chia seed?s omega-3 character. Consumer acceptance data was obtained through the use of hedonic testing at UF/IFAS FSHN taste panel. On separate occasions consumers were asked to rate and rank a cookie or muffin sample, one control and two treated samples (25% and 40% ground chia muffin; 15%ground/15% whole seed and 5% ground/10%whole chia cookie) . Panelist used a nine-point hedonic scale to rate the following attributes: overall acceptance, texture, flavor, and appearance. The hedonic scale was anchored with 1=?dislike extremely? and 9=?like extremely?. The nine-point Food Action rating scale (FACT) was used to determine the consumption attitudes of the panelist. The anchors for the FACT scale were 1=?I would eat this only if forced? and 9=?I would eat this at every opportunity?. Analysis of variance (ANOVA) was used to analyze panelist data, with means separation using Tukey?s HSD test. There was no significant (p < 0.05) difference between the control and the 25% chia muffin in overall acceptability. The same result was found for the cookie control and the 5% ground/10%whole chia cookie sample. Subsequent fatty acid analysis was performed on the muffin and cookie samples, but including the chia in whole, ground, and combination forms at the aforementioned chia levels. It was found that the highest content of ?-linolenic acid was obtained in both ground seed cookie and muffin samples. The muffin and cookie samples were also subjected to water activity, L*a*b* color, and textural analysis, to measure the chia product physical properties versus those of the control samples. Significant differences between the muffin treatments and the control muffin where found for the appearance attribute, L and b color-values, hardness, springiness, and cohesiveness textural attributes. The control cookie and the treatment cookies had significant differences for all attributes. It was found that the ground muffin and ground cookie treatments retained the most omega-3 character of the tested treatments. It is the position of this study that the chia seed is a viable ingredient in food products to be consumed by the public.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Devin Lewis.
Thesis: Thesis (M.S.)--University of Florida, 2010.
Local: Adviser: Goodrich, Renee M.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042223:00001


This item has the following downloads:


Full Text





THE INCORPORATION OF CHIA (SALVIA HISPANICA LAMIACEAE) SEEDS INTO
BAKED FOOD PRODUCTS




















By

DEVIN C. LEWIS


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2010

































2010 Devin C. Lewis
































To my mother, who has been my emotional and spiritual support against all odds









ACKNOWLEDGMENTS


I would like to thank my advisor, Dr. Renee Goodrich-Schneider for granting me

the opportunity to pursue my master's degree and for all the invaluable guidance

throughout this experience. I am thoroughly appreciative for the expertise and

assistance given to be my committee members, Dr. Charles Sims and Dr. Al Wysocki.

I would also like to thank my lab mates, Lemane Delva, Yael Spektor, and

Brianna Mahoney for all their help and suggestions. A special thank you to the FSHN

taste panel staff for their assistance during the long and numerous taste panel sessions.

Most of all, I thank my family and close friends whose support and understanding gave

me the strength to succeed through accomplishing my goals.









TABLE OF CONTENTS

page

ACKNOW LEDGM ENTS .......... ............................ .. ......................................... 4

LIST O F TA BLES .......... ..... ..... ............................................................. ........ 7

LIST O F FIG URES........................................... ............... 10

A BST RA CT ............... ... ..... ......................................................... ...... 12

CHAPTER

1 INTRODUCTION .............. .......... .......... ......... 14

2 LITERATURE REVIEW .................. ................... ......... 19

2.1 Botanical and Physical Characteristics........................................ 19
2.2 Historical Perspective .......... ......... .......................... 21
2.3 Physiochemical Properties .................. ............................................... 22
2.4 Seed Physiology and Lipid Synthesis..................................................... 24
2.5 Fatty Acid Nomenclature ............... .......... ... .. ........................ 26
2.6 Alpha-Linolenic Acid Metabolism in Humans ..................................... 27
2.6.1 Fatty A cid C conversion ...................... ........................... ............... 27
2.6.2 Reactive protein interactions .............. ...... ...... ..... .... ........... 29
2.7 Consumer Demand for and Attitudes toward Functional Foods ................... 30
2.8 U.S. Oilseed Farming Industry Analysis ................. ......... .. ........... .... 38
2.9 Consumer Sensory Testing Overview ..................... ................. .. 43
2.9.1 H edonic Scale ........................... ......... .................... ............... 44
2 .9.2 F .A .C .T S cale ....................... .... ....................... .. ........... 46
2 .1 0 O bje ctiv e s ............................................................................................ 4 7

3 METHODS AND MATERIALS .................. ......................... 50

3.1 P product Form ulation ............. ................ .......................................... ...... 50
3.2 Taste Panel Product Assessment ........ ............. ........... .............. 53
3.3 Chia Product Physical Analysis ............... .......................... ............. 54
3.4 Chia Product Fatty Acid Analysis ......... .. ............. ............. .... ...... ..... 56

4 RESULTS AND DISCUSSION ........._.. ..... ........................... ......... 64

4.1 Consumer Taste Panel Testing ................... .............. .............. 64
4.1.1 Chia Seed Muffin ........ ............................... 64
4.1.2 C hia S eed C ookie................... .......................... ............................ 67
4.2 Product Physical Analysis ..... ............... ...................72
4.2.1 Chia Muffin Analysis .... ....... ......... ............... ... ............... 72
4 .2 .1.1 W after activity ............................... ...................... .. .. .... ... 73









4.2.1.2 Color analysis ....... ................................................. 74
4.2.1.3 Textural analysis......................... ....................... 76
4.2.2 Chia Cookie Analysis ............. .......... .... ................... 78
4.2.2.1 W after activity ..................................................... ....... ............... 79
4.2.2.2 Color analysis .................................... ........ ........... 80
4.2.2.3 Textural analysis......................... ....................... 82
4 .3 P product C hem ica l A na lysis.................................................... ... .. ............... 83
4.3.1 Chia Muffin Product ............ ............ ......... ............... 84
4 .3 .2 C h ia C oo kie P roduct................................. ................ ... ................. 85

5 CONCLUSIONS ................ ......... .......... ......... 105

APPENDIX

A MUFFIN AND COOKIE INGREDIENT MANUFACTURERS ............................... 107

B MOISTURE LOSS PERCENTAGE.............................. .............. 108

C FORMULATION COST ANALYSIS ........................................................... 109

LIST OF REFERENCES .................................... ........................... .. ....... 110

BIO G RA PH ICA L SKETC H ......................................................................... ...... 116




























6









LIST OF TABLES


Table page

3-2 Food Action Rating Scale .......... ......... ......... .. .. ............... 61

3-3 Evaluated chia muffin product formulations.............. ......... ................ 63

3-4 Evaluated chia cookie product formulations ....... ..... ................................... 63

4-1 Developed chia muffin and chia cookie products surface L*A*B* color-value
analysis and water activity analysis results at 19-21 C ............... ............... 87

4-2 Developed chia muffin product separation of means measuring the attribute
of water activity (Aw) (19-21 C) when comparing variation across sample
treatm ents ......................................... ............ 87

4-3 Developed chia muffin product separation of means measuring the attribute
of L* surface color-value when comparing product color variation across
sample treatments ...... ........ ......................... ...... ......... 87

4-4 Developed chia muffin product separation of means measuring the attribute
of surface a*color-value when comparing product color variation across
sam ple treatments ........... ........... .......................... ............... 87

4-5 Developed chia muffin product separation of means measuring the attribute
of b* color-value when comparing color variation across sample treatments..... 88

4-6 Developed chia cookie product separation of means measuring the attribute
of water activity (Aw) (19-21 C) when comparing variation across sample
treatm ents ......................................... ............ 88

4-7 Developed chia cookie product separation of means for the attribute of L*
color-value when comparing variation across sample treatments. ................... 88

4-8 Developed chia cookie product separation of means measuring the attribute
of a* color-value when comparing variation across sample treatments.............. 88

4-9 Developed chia cookie product separation of means measuring the attribute
of b* color-value when comparing variation across sample treatments ........... 89

4-10 Developed chia muffin product compression test analysis and chia cookie
product triple beam break test analysis results........................... ... ............... 89

4-11 Developed chia muffin product separation of means, measuring the physical
attribute of hardness when comparing variation across sample treatments....... 89









4-12 Developed chia muffin product means of separation measuring the attribute
of springiness when comparing variation across sample treatments ................ 89

4-13 Developed chia muffin product separation of means measuring the attribute
of cohesiveness when comparing variation across sample treatments ............. 90

4-14 Developed chia muffin product separation of means measuring the attribute
of gumminess when comparing variation across sample treatments............... 90

4-15 Developed chia muffin product separation of means measuring the attribute
of chewiness when comparing variation across sample treatments ................ 90

4-16 Developed chia cookie product separation of means measuring the cookie
"break" characteristic when comparing variation across sample treatments. ..... 90

4-17 Developed chia muffin product sensory evaluation panel preference test
results measuring six qualitative attributes. ............ .................................... 91

4-18 Developed chia muffin product separation of means measuring the attribute
of overall acceptability when comparing variation across sample treatments..... 91

4-19 Developed chia muffin product separation of means measuring the attribute
of appearance when comparing variation across sample treatments ................ 91

4-20 Developed chia muffin product means of separation measuring the attribute
of flavor when comparing variation across sample treatments....................... 92

4-21 Developed chia muffin product separation of means measuring the attribute
of texture when comparing variation across sample treatments ................... 92

4-22 Developed chia muffin product separation of means measuring collected
FACT scale data when comparing variation across sample treatments ............ 92

4-23 Developed chia muffin product separation of means measuring compiled
ranked totals when comparing variation across sample treatments. ................ 92

4-24 Developed chia cookie product sensory evaluation panel preference test
results measuring six qualitative attributes. ............ .................................... 93

4-25 Developed chia cookie product separation of means measuring the attribute
of overall acceptance when comparing variation across sample treatments..... 93

4-26 Developed chia cookie product separation of means measuring the attribute
of appearance when comparing variation across sample treatments ................ 93

4-27 Developed chia cookie product separation of means measuring the attribute
of flavor when comparing variation across sample treatments....................... 94









4-28 Developed chia cookie product separation of means measuring the attribute
of texture when comparing variation across sample treatments ............ ...... 94

4-29 Developed chia cookie product separation of means measuring FACT scale
collected data when comparing variation across sample treatments ............... 94

4-30 Developed chia cookie product separation of means measuring compiled
ranked totals when comparing variation across sample treatments. ................ 94

4-31 Developed chia muffin product measured fatty acid content results ................. 95

4-32 Developed chia cookie product measured fatty acid content results ............... 95

4-33 Developed chia muffin and chia cookie product measured omega-3 and
omega-6 fatty acid analysis statistical outcomes ......... ... ........ ............. 95

4-34 Developed chia muffin product separation of means measuring linoleic(18:2)
fatty acid content when comparing variation across sample treatments ............ 95

4-35 Developed chia muffin product separation of means measuring a-linolenic
(18:3) fatty acid content when comparing variation across sample treatments... 95

4-36 Developed chia cookie product separation of means measuring linoleic (18:2)
fatty acid content when comparing variation across sample treatments
means............................................ ........... 96

4-37 Developed chia cookie product separation of means measuring a-linolenic
(18:3) fatty acid content when comparing variation across sample treatments
m e a n s ................................................................................. 9 6

B-1 Calculated Chia m uffin product.......................................... ......................... 108

B-2 Calculated Chia cookie product.............................. .... ............... 108

C-1 Chia muffin .......................... .......... ............... 109

C-2 Chia cookie................................. ............... 109









LIST OF FIGURES


Figure page

2-1 Alpha linolenic acid structure ............................ ..................... 49

2-2 Docosahexanoic acid structure ............... ................. ................................... 49

2-3 Eicosapentanoic acid structure ............... .................................................. 49

3-1 Hedonic nine point scale ............................................................................. 61

3-2 Sample ballot used in chia product taste panel testing ................. .............. 62

4-1 Developed chia muffin product control sample compression test graphical
outcome e, trial 1, reps 1-3 ...................................... ............... .............. 96

4-2 Developed chia muffin product control sample compression test graphical
o utco m e tria l 1, re p 4 ........................................... ................ ........... 97

4-3 Developed ground seed chia muffin product compression test graphical
outcome e, trial 1, reps 1-3 ...................................... ............... .............. 97

4-4 Developed ground seed chia muffin product compression test graphical
o utco m e tria l 1, re p 4 ........................................... ................ ........... 98

4-5 Developed whole seed chia muffin product compression test graphical
outcome e, trial 1, reps 1-3 ........................................ ............... .............. 98

4-6 Developed whole seed chia muffin product compression test graphical
o utco m e tria l 1, re p 4 ........................................... ................ ........... 99

4-7 Developed combination chia seed muffin product compression test graphical
outcome e, trial 1, reps 1-3 ...................................... ............... .............. 99

4-8 Developed combination chia seed muffin product compression test graphical
outcome e, trial 1, rep 4 ................................................................ ................ 100

4-9 Developed chia seed cookie product control sample break test graphical
outcome e, trial 1, reps 1-3 .................................... ............... .............. 100

4-10 Developed chia seed cookie product control sample break test graphical
outcome, trial 1, rep 4 ............................................................. 101

4-11 Developed ground seed chia cookie product break test graphical outcome,
trial 1, reps 1-3 ................. .... ......... ................. .. ............. 101

4-12 Developed ground seed chia cookie product break test graphical outcome,
trial 1, rep 4.... ......................................................... 102









4-13 Developed whole seed cookie product break-test graphical outcome, trial 1,
reps 1-3 .................... .. .............. ......... ............................. 102

4-14 Developed whole seed chia cookie product break-test graphical outcome,
tria l 1 re p 4 .................. ........................... ....... .......... ...... 1 0 3

4-15 Developed combination chia seed cookie product break-test graphical
outcome e trial 1, reps 1-3 ............................................................................ 103

4-16 Developed combination chia seed cookie product break-test graphical
outcome e, trial 1, rep 4 ................................................................. ............... 104










Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

THE INCORPORATION OF CHIA (SALVIA HISPANICA LAMIACEAE) SEEDS INTO
BAKED FOOD PRODUCTS
By

Devin C. Lewis

August 2010

Chair: Renee Goodrich-Schneider
Major: Food Science and Human Nutrition

Chia is an herbaceous summer annual which is botanically known as Salvia

hispanica Lamiaceae. The seed of this annual is harvested in the fall and has been

consumed in Southern Mexico and Central America for generations. Once a staple

food of ancient Mesoamerican civilizations, chia has currently emerged as a high source

of polyunsaturated fatty acids. The oil of the chia seed has an omega- 3 fatty acid

content of up to 68% versus the 57% found in flax seed. The fatty acid contained in

chia, a-linolenic acid (ALA), is metabolized in the body to produce eicosapentanoic and

docosahexanoic acids, respectively, and has been found to reduce the risks of coronary

heart disease in humans.

The objectives of this study were 1) To demonstrate similar consumer acceptance

between control products and those incorporated with chia seeds, 2) To investigate

which form of chia seed withstands processing and baking best, defined as retaining the

highest amount of the chia seed's omega-3 character.

Consumer acceptance data was obtained through the use of hedonic testing at

UF/IFAS FSHN taste panel. On separate occasions consumers were asked to rate and

rank a cookie or muffin sample, one control and two treated samples (25% and 40%









ground chia muffin; 15%ground/15% whole seed and 5% ground/10%whole chia

cookie) Panelist used a nine-point hedonic scale to rate the following attributes:

overall acceptance, texture, flavor, and appearance. The hedonic scale was anchored

with 1="dislike extremely" and 9="like extremely". The nine-point Food Action rating

scale (FACT) was used to determine the consumption attitudes of the panelist. The

anchors for the FACT scale were 1 =" would eat this only if forced" and 9="1 would eat

this at every opportunity". Analysis of variance (ANOVA) was used to analyze panelist

data, with means separation using Tukey's HSD test. There was no significant (p< 0.05)

difference between the control and the 25% chia muffin in overall acceptability. The

same result was found for the cookie control and the 5% ground/10%whole chia cookie

sample. Subsequent fatty acid analysis was performed on the muffin and cookie

samples, but including the chia in whole, ground, and combination forms at the

aforementioned chia levels. It was found that the highest content of a-linolenic acid was

obtained in both ground seed cookie and muffin samples. The muffin and cookie

samples were also subjected to water activity, L*a*b* color, and textural analysis, to

measure the chia product physical properties versus those of the control samples.

Significant differences between the muffin treatments and the control muffin where

found for the appearance attribute, L and b color-values, hardness, springiness, and

cohesiveness textural attributes. The control cookie and the treatment cookies had

significant differences for all attributes. It was found that the ground muffin and ground

cookie treatments retained the most omega-3 character of the tested treatments. It is

the position of this study that the chia seed is a viable ingredient in food products to be

consumed by the public.










CHAPTER 1
INTRODUCTION


In today's society, consumers are far more informed with respect to the foods they

consume than past generations. This increase in knowledge proportionately increases

the demand for food products that are more nutritious, but remain tasty and satisfying

as well. When venturing to the health food aisle of the average grocer consumers are

bombarded by the large variety of foods that are labeled as "high in..." and "a good

source of... ". Although these products undergo some governmental scrutiny, it is still

left up to the consumer to identity those products that can ultimately provide the sought

after minerals and nutrients in the highest quantities. Consumers are more aware of

food and dietary issues and are monitoring and adjusting what they consume as they

have become more proactive and diligent in improving their overall health through their

daily diet (Constance 2008).

One market that many of these products strive to gain a share of is that of the

"health and wellness" market. According to the Natural Marketing Institute, retail sales

of health and wellness consumer packaged goods reached $102.8 billion in the U.S. in

2007, demonstrating a growth of 15% over the previous year (Teklenburg 2009). This

market tends to be stronger due to the rise in coronary heart disease and other

diseases witnessed in the United States. Coronary heart disease (CHD) is the leading

cause of heart attacks and stroke in this country. This disease is defined as the

narrowing of the small blood vessels that supply blood and oxygen to the heart. CHD is

commonly caused by a condition known as atherosclerosis. Atherosclerosis occurs

when fatty material and plaque buildup on the walls of the arteries. This accumulation









causes the arteries to narrow, which not only limits, but can completely halt the flow of

blood to the heart (Goldstein 2007). As of 2006, more than 17 million people over the

age of 20 suffer from this disease and it causes at least 500,000 deaths annually. It is

estimated that approximately every 25 seconds, an American will suffer a coronary

event and approximately every minute, someone will die of one (Lloyd-Jones and others

2009). Some of the risk factors involved with CHD are diabetes, high blood pressure,

elevated levels of low density lipoprotein concentrations and triacylglycerols in blood

plasma. To combat these undesirable conditions, health officials urge lifestyle changes

that emphasize a transformation in the amount and types of dietary fats consumed. It

has been well documented that to decrease these risk factors, a shift from the

consumption of saturated to polyunsaturated fatty acids is necessary (Watkins and

Bierenbaum 2001).

Through research, omega-3 fatty acid consumption has emerged as one of the

key components in the fight to lower the aforementioned risk factors. This fatty acid is

metabolized in the body to produce eicosapentanoic and docosahexanoic acids,

respectively. Typically, the source of omega-3 fatty acids has been relegated to high oil

fish, flax seed endosperm and its oil fraction (Finley and Shahidi 2001). High oil fish

produce substantial amounts of the fatty acid, but are also accompanied by an

unwanted and many times intolerable "fishy" aroma. Although high oil fish are the best

source of these fatty acids, the use of their oils in food products has been slow due to

the unstable nature of these oils. This instability produces off-flavors which are the

result of degraded products from hydroperoxide formed from the highly unsaturated

lipids in the fish bodies. The oxidation of the fatty acids can produce off flavors that









range from mild to offensive. This oxidative degradation is the product of the degree of

unsaturation that composes the chemical matrix of omega-3 fatty acids (Gebauer and

others 2006). Recently, there have been many attempts to increase the stability of the

aforementioned oils. It was found that oxidation can be limited by interesterification to

create borage oil, or partially hydrolyzing the oil and combining it with emulsifiers.

Although these methods produced positive results, it is not certain if they can be used in

food products and gain consumer acceptability after this type of manipulation (Gebauer

and others 2006).

Flax seed is not associated with the aforementioned aroma, but tends not to

contain the highest fatty acid content possible from a plant source. The highest content

of omega- 3 fatty acids from a plant source can be obtained from chia, also known as

Salvia hispanica Lamiaceae. The oil expressed from chia has been measured at 68%

w/w a-linolenic acid (ALA) as opposed to the 57% w/w found in flaxseed (Ayerza 1995).

a-linolenic acid is a polyunsaturated fatty acid that is metabolized through the sequential

activities of A-5 and A-6 desaturases. These desaturases elongate the original carbon

chain to create eicosapentanoic (EPA) and docosahexanoic (DHA) fatty acids. Both

fatty acids are essential, and recommended daily dosages are 500-1800 mg per day or

1300 mg of ALA. EPA and DHA are also beneficial to the function of cell membranes,

the central nervous and immune systems. ALA has also been found to aid in the

production of various eicosanoids such as leukotrienes and prostaglandins. These

eicosanoids can be derived from omega-3 and omega-6 fatty acids act as signaling

molecules. They exert complex control over the central nervous system, inflammation

and immunity in the human body (DeCaterina and Basta 2001). Due to higher levels of









the estrogen hormone, women have been found to metabolize these fatty acids more

efficiently than their male counterparts (Burdge and Wooten 2002). Another health

benefit that chia provides is derived from its total dietary fiber content. On a dry weight

basis, chia seeds contain 37-41g total dietary fiber per 100g (Reyes-Caudillo and others

2008). In accordance with the American Dietary Association guideline of a 3:1 ratio of

insoluble to soluble dietary fiber intake for adults; the seed of S. hispanica yields 32.8-

34.9 grams insoluble and 6.16-6.84 grams of soluble fiber per 100 grams. Some of the

reported effects of both types of dietary fiber are the reduction of cholesterolemia,

modification of glycemic and insulinaemic responses, and more efficient intestine

function (Reyes-Caudillo and others 2008). The main component of the insoluble

dietary fiber fraction is lignin, which makes up 39-41% of the total dietary fiber. Lignin is

a complex organic polymer, derived from wood and is an integral part of plant wall

secondary structure. It has been directly linked to the aforementioned

hypocholesterolemic effects through its tendency to absorb bile acids (Reyes-Caudillo

and others 2008). The fiber-rich fraction of chia seeds has also been observed to have

a high antioxidant activity (488.8 pmol Trolox equivalents (TE)/g), which is similar to

high antioxidant beverages such as tea (631 pmol TE/100ml) and freeze-dried coffee

(450 pmol TE/g). This high antioxidant activity is attributed to the presence of quercetin,

caffeic, and chlorogenic acids. Given the above, it has been purported that as little as

7grams of chia fiber rich fraction may contribute to the trapping of free radicals, meeting

total dietary antioxidant level (3549 pmol TE/day) in a Mediterranean diet (Sauro-Calixto

and Gohi 2006).









The focus of this study is to demonstrate similar consumer acceptance between

control products and those incorporated with chia seeds, and to investigate which form

of chia seed withstands processing and baking best, which is defined as retaining the

highest amount of the chia seed's omega-3 character. To attain these objectives,

consumer acceptance data was obtained through the use of hedonic testing at UF/IFAS

FSHN taste panel. On separate occasions consumers were asked to rate and rank a

cookie or muffin sample, one control and two treated samples (25% and 40% ground

chia muffin; 15%ground/15% whole seed and 5% ground/10%whole chia cookie).

Panelist used a nine-point hedonic scale to rate the following attributes: overall

acceptance, texture, flavor, and appearance. The hedonic scale was anchored with

1="dislike extremely" and 9="like extremely". The nine-point Food Action rating scale

(FACT) was used to determine the consumption attitudes of the panelist. Anchors for

the FACT scale were 1 =" would eat this only if forced" and 9="1 would eat this at every

opportunity". The muffin and cookie samples were also subjected to water activity,

L*a*b* color, and textural analysis, to compare the chia product physical properties

versus those of the control samples. It is hypothesized that the products incorporated

with chia seeds will find acceptance amongst consumers and the physical attributes of

said products will not be significantly different (p>0.05) than control samples.









CHAPTER 2
LITERATURE REVIEW


2.1 Botanical and Physical Characteristics

Chia is an herbaceous summer annual, botanically classified as Salvia hispanica

Lamiaceae. Common names for the seeds of this species are chiaa", chiaa sage", and

"Spanish sage". Closely related to mint, the Salvia genus consists of at least 900

species within the Lamiaceae family (Ayerza and Coates 2004). This summer annual

produces a scalene ellipsoidal seed, which has a thickness of 0.81-1.40 mm, average

width of 1.3-1.4 mm, and is approximately 2 mm in length. Its appearance is said to be

similar to that of rapeseed or quinoa (Vilche and others 2003). Seed color typically

varies between black and white, with a much smaller percent of white seeds due to the

encoding of a single recessive gene. The white seeds also tend to be larger than the

black seeds (Cahill and Provance 2002). On a dry basis, the average seed moisture is

6.6% to 7.2% for black and white seeds, respectively, which lends itself to the storage

stability of the seeds. The bulk density, a ratio of the mass of the seeds to its total

volume, was measured between 0.667 and 0.722 g cm-3. Bulk density is used to

determine the capacity of agricultural storage and transport systems. The true density,

which is used in the sourcing of agricultural separation equipment, was measured

between 0.931 and 1.075 g cm-3 (Ixtaina and others 2008). The resistance to airflow

during aeration and drying procedures is determined by the porosity of the mass of

seeds. The porosity of the seed is the fraction of the space within the grain not

occupied by the grain (Thompson and Issacs 1967). The porosity of chia seeds was

found to be 22.9- 35.9%, and continuous, leaving ample space for the flow of oxygen

during aeration and carbon dioxide while stored in silos. When tested for its frictional









properties, chia seeds were measured to have an average angle of repose of 17.10.4.

The seed shape and smooth outer surface were responsible for the lower value of

repose causing the seeds to slide onto one another. Values of static coefficient of

friction were 0.28 0.01 on galvanized sheets and 0.31 0.01 on mild steel sheets,

which are lower than values of cumin, sunflower and sesame seeds (Ixtaina and others

2008).

The seeds of this light-sensitive annual are generally harvested in the fall, during

its early vegetative stage which is immediately prior to the plants shooting period when

a-linolenic content is at its highest (Peiretti and Gai 2009). Although native to southern

Mexico and northern Guatemala, chia is grown throughout Central and South America.

The environmental conditions best suited for plant maturation and higher fatty acid

content are described as arid to semi-arid. These ecosystems typically exhibit cooler

temperatures (12-24C) in conjunction with elevations ranging from 398 to 500 meters.

Average rainfall in the arid to semi-arid regions was measured at 186-187 millimeters

during crop growing cycles (Ayerza and Coates 2004).

Although there have been changes in morphology due to human selection, both

domesticated and wild type varieties of S. hispanica remain morphologically distinct

from related taxa, especially in regards to anther, calyx and seed morphology. This

observation led to the classification of chia within the taxonomic section, Potiles, within

the Salvia subgenus Calosphace (Epling 1940). According to Haque and Ghoshal

(1980), indicators such as reproductive isolation, a unique chromosome number

(2n=12), the lowest of the genus, and an autogamous (self-fertilizing) breeding system,

it is suggested that the wild-type variety contributed to domesticated S. hispanica.









These contributions coupled with unintentional human-selected traits over time, produce

the domesticated chia variety harvested currently. The human-selected traits for

domesticated chia include: apical dominance, increased branching, increased seed

size, decreased pubescence, seed coat colors and patterns, and closed calyxes. The

trait that solidifies the domestication of the S. hispanica is the closed calyx (Cahill 2003).

Calyx closure completely eliminates seed dispersal, rendering limited survival of the

species outside of human cultivation.

2.2 Historical Perspective

During pre-Colombian times, chia served as a fixture in the culture of several

meso-American civilizations. After harvesting, the seeds had many uses within daily life

including but not limited to culinary and medicinal purposes. The Aztec capital,

Tenochtitlan was reported to receive between 5 and 15 tons of chia seed annually.

According to the 16th century Codex Mendoza and Matricula do Los Tributos, 21 of the

38 Aztec provincial states presented the seeds as tribute to Aztec gods (Berdan and

Anawalt 1996). Its importance as a staple food item only trailed that of beans and corn,

but was consumed more frequently than amaranth. Post-harvest chia seeds were

roasted and ground into flour known as chiapinolli. This flour was then incorporated into

tortillas, tamales, or eaten as gruel. The ground seeds were also used as an ingredient

in Chianatoles, an Aztec beverage. Today, ground chia beverages have been replaced

by those comprised of whole seeds, water, lemon, and sugar or fruit juices, and called

"Chia fresca" (Cahill 2003). These versatile seeds were not only consumed, but were

used for medicinal purposes also. Citations in the Badianus Manuscript refer to

infusions including whole chia seeds to aid in the uptake of medicines and treatment of

respiratory malaise (de la Cruz 1952). Chia was also used to treat eye obstructions and









infections by placing seeds directly under the eyelid. Kidney problems were alleviated

through the consumption of a handful of seeds combined with two liters of water.

Extracted oils served as a base in skin emollients, lacquer for primitive pottery

manufacture, craft and body paints (Cahill 2003). Mixchiaviticac, a Nahua (indigenous

people of Mexico) word, refers to the circles painted on the cheeks of Aztec deities

(Hauman 1991) According to Spanish manuscripts; chia was known as the "running

food" due to its consumption by Aztec messengers for endurance. Due to Spanish

colonization efforts and influence, the cultivation of chia seeds in meso-American

society declined substantially (Kreiter 2005).

2.3 Physiochemical Properties

Throughout history chia has been used in various capacities, but it is not until

recent years that its physiochemical properties have been investigated. Chia seeds

contain 250-390 g oil/kg of fresh matter. The fatty acids of the oil fraction extracted from

the seeds are polyunsaturated, their primary components are linoleic (C18:2n-6; 170-260

g/kg of total fatty acid) and a-linolenic acid (C18:3n-3; 500-570 g/kg of the fatty acid)

(Ayerza, 1995). Once introduced into an aqueous substance, chia seeds tend to exude

a polysaccharide mucilage that remains tightly bound to the seed. Removal of the

mucilage can be performed using 6N urea at pH 7.4 for 7 hours, which gives a 4.5% dry

weight yield of the polysaccharide. The polysaccharide has been identified as being

composed of D-xylosyl, D-glucosyl and 4-O-methyl-a-D-glucopyrranosyluronic acid in

ratios of 2:1:1. These components are arranged in a linear tetrasaccharide sequence

(Lin and others 1994). A low content of uronic acid is a sole indicator of no pectin being

associated with the polysaccharide mucilage (Reyes-Caudillo and others 2008).

Defatted chia that remains after oil extraction leaves behind a fiber fraction that is









33.9g/100g and 17g/100g of protein (Craig and Sons 2004). The total dietary fiber

content of the seeds is 36.9-39.9g/100g, while insoluble and of soluble fiber are 32.8-

34.9 grams 6.16-6.84 grams per 100 grams, respectively. The primary component of

the insoluble dietary fraction is lignin, occupying 39-41% of the total dietary fiber. Lignin

is thought to protect the unsaturated fats in the chia seed by building a strong and

resistant structure. This structure is supported by spaces in the cell walls filled with

lignin, cross-linking various plant polysaccharides. The presence of cellulose and

hemicellulose are verified by the neutral sugars found in the insoluble dietary fiber

fraction (Reyes-Caudillo and others 2008). The water-holding capacity of the defatted

fiber fraction was 15.41g/ g fiber. The water-holding capacity is the ability of a moist

material to retain water when subjected to an external centrifugal gravitational force or

compression. This capacity is composed of the linked water, hydrodynamic water and

physically trapped water, which contributes the greatest proportion to the capacity. It is

suggested that the high water-holding capacity is influenced by the polysaccharide

mucilages (Vazquez-Ovando and others 2009). In contrast, the oil-holding capacity of

chia seeds has a tendency to be low (2.02g /g sample). It is theorized that the particle

size of the fiber fraction is not small enough to hold higher amounts of oil, since smaller

particles typically have more contact surfaces. The capacity of water absorption is an

indication of a structure's ability to absorb water when immersed in water or in contact

with a constantly moist surface. The water absorption capacity of the chia fibrous

fraction is 11.73g water/g sample. The fibrous fraction of the chia seed also has

consistent emulsifying properties. The emulsifying capacity of a molecule demonstrates

its ability to facilitate solubilization or dispersion of two immiscible liquids. The emulsion









activity of the chia fibrous fraction was 53.26 ml/100ml. The stability of the emulsions

formed by the seed fiber fraction was 94.84ml/100ml which could possibly be attributed

to the chia seed protein fraction because most proteins are strong emulsifiers. These

emulsifying properties are also displayed in regard to the absorption of bile acids, and

increasing fecal excretion, which would limit small intestine uptake (Vazquez-Ovando

and others 2009). The total chia seed phenolic content was in the range of 0.880-

0.92110.008 mg/g chia seed extract in GAE (gallic acid equivalents). Flavonols were

found to be in the highest concentrations, followed by quercetin, kaempferol, with lower

amounts of chlorogenic and caffeic acids detected. The antioxidant activity of chia

seeds is comparable to Trolox at 220ppm GAE when measuring radical scavenging

activity. In the 3-carotene inhibition linolenic acid model (3-CLAMS), chia seed extracts

demonstrated an ability to stabilize reactive oxygen species responsible for the

oxidation of linolenic acid (Reyes-Caudillo and others 2008). Singlet oxygen is a

precursor of hydrogen peroxide and hydroxyl radical formation (Siddhuraju and others

2002). Extracts of chia seeds not only showed comparable activity to Trolox in the

inhibition of peroxidation, but displayed properties of singlet oxygen quenchers. The

seed extracts act as antioxidants via the hydrogen donating capacity of their phenolic

groups. The activities of these polyphenolics also demonstrate a metal-chelating

potential that plays a role in the protection against iron- and copper- induced free

radicals (Reyes-Caudillo and others 2008).

2.4 Seed Physiology and Lipid Synthesis

The seed of the S. hispanica is known for its high polyunsaturated fatty acid

content associated with its lipid fraction. The sole site of the de novo fatty acid









synthesis within the plant cells is the plastid, more specifically the elaioplast. Generally,

the accumulation of total lipids takes on a sigmodial pattern and can be divided into

three periods. During the initial period, structural lipids phospholipidss and glycolipids)

are present, but are only a small proportion of the total seed weight. Triacylglycerols are

collectively absent at this point, and the predominant fatty acids at this stage are

palmitic, oleic, linoleic, and a-linolenic. As the secondary period is initiated, more rapid

seed growth begins, stimulating triacylglycerol accumulation. Due to this accumulation,

a substantial increase in the seed's total dry weight begins and continues for at least a

two week period. As this occurs, triacylglycerols become the predominant lipid

component as phospholipids and glycolipids quantities decrease. At the end of this

stage, triacylglycerols are 90% of the total lipid fraction and overall fatty acid

composition. When the growing seed reaches the final period leading to full maturity

the dry weight per seed increases very little which is accompanied by a progressive

decrease in moisture content. Phosphotidylcholine is thought to play a major role in

polyunsaturated fatty acid formation. One theory suggests that as the development of

the seed cotyledon (part of seed embryo and major storage organ) occurs, phospholipid

acyltransferase catalyzes an acyl exchange between oleyol-CoA enzymes and fatty

acids at position two of the phosphotidylcholine. This reaction provides polyunsaturated

acyl CoAs that work with acyl CoAs derived from the plastid via the glycerol-3-

phosphate pathway which can via lipid synthesis be incorporated into triacylglycerols

(Styme and Glad 1981). These triacylglycerols accumulate in the seed's cell cytoplasm

as lipid droplets, and are spherically-shaped during the early and mid-stages of seed

development. Eventually, they become compressed and tightly packed in the oil-rich









cells of mature seeds. Neutral lipids are typically located within discrete organelles

known as lipid or oil bodies. These oil bodies typically have a mean diameter of 1 pm.

During seed development, the size of the oil body remains constant as triacylglyceride

accumulation is accompanied by an increase in the number of oil bodies in the cell

(Rest and Vaughn 1972). The major lipid organ within chia seeds is the cotyledon. The

cotyledon contains a small percentage of saturated fatty acids, but not enough for the

reserved oils to be considered fats. Unsaturated fatty acids generally favor position two

of the triacylglycerol and constitute more than 33% of the total fatty acid, the "spill over"

congregates at position one (Trealese and Doman 1984).

2.5 Fatty Acid Nomenclature

The derivatives of hydrocarbons, fatty acids, are carboxylic acids with hydrocarbon

chains. These hydrocarbon chains can range from 4 to 36 carbons long (C4- C36),

although the most common carbon chains are from 4 to 24 carbons in length. Some

fatty acid structures are unbranched and completely saturated. Saturation in fatty acid

nomenclature denotes the absence of double bonds within the structure. Unsaturation,

on the other hand, denotes at least one double bond in the hydrocarbon chain. Fatty

acids with only one double bond are termed monounsaturated while multiple double

bonds are categorized as polyunsaturated fatty acids. The most simplified

nomenclature for these carbon chains specifies the chain length and number of double

bonds, separated by a colon; for example 18-carbon stearic acid with no double bonds

would be noted as 18:0. When these structures contain double bonds there are two

options for noting these bonds. The first option is the delta (A) system. In this system,

the carbons are counted starting at the carboxyl (COOH) end of the chain. The

positions of the double bonds are noted by superscript numbers following delta, ex:









18:3(A 9,12) (Nelson and Cox 2005). The second option for noting double bonds within

fatty acid structures is the omega (w) system. In this system, the carbons are counted

from the methyl (CH3) end of the structure. This double bond position is noted by the

number after the w- symbol, ex: 20:4 w6 or 20:4 n-6 (Nawar 1996). The latter

convention is the more commonly used and is the notation used in the following

discussion.

2.6 Alpha-Linolenic Acid Metabolism in Humans

The primary fatty acid derived from the lipid fraction of chia seeds is a-linolenic

acid (ALNA), C18:3 n-3. The adequate intake, and intake associated with a low

prevalence of inadequacy, is set for ALNA is 1.6g/ day for men and 1.1 g/day for women

(Gebauer and others 2006).

2.6.1 Fatty Acid Conversion

Within the human body, ALNA is converted to eicosapentanoic acid (EPA, C20:5n-

3) by the sequential activities of A6 and A5 desaturases and carbon chain elongation.

Docosapentaenoic acid (DPA) C22:5 n-3 is formed via the addition of C2 to EPA which is

then converted to C24:5 n-3 and C24:6n-3 by A6 desaturation (Burdge and Wooten 2002).

Docosahexanoic acid (DHA) is synthesized from C24:6n-3 by peroxisomal 3-oxidation,

which shortens the carbon chain by C2. Although the A6 desaturase is the rate-limiting

step in the synthesis of DHA, its overall regulation is still unclear. This synthesis occurs

to a lesser extent than that of DPA and EPA (Sprecher 2000). ALNA is considered to

be an essential fatty acid since it is not produced by the body. Long chain

polyunsaturated fatty acids (LCPUFA) are important structural components of cell

membranes, and their interactions with phospholipids ensure cell functionality. The

primary conversion site is the liver followed by enterocytes (Burdge and others 2002).









After ALNA conversion, very low density lipoproteins transport newly synthesized EPA

and DPA away from the liver to other parts of the body (Burdge and Wooten 2002).

During metabolic conversions of ALNA, omega-3 and omega-6 fatty acids compete for

metabolic enzymes; this competition also exists during esterification into plasma

phospholipids and triglycerides (Mozaffarian 2005). It has been noted that through this

competitive nature, increases in dietary concentration of C18:2 n-6 caused a decrease in

the synthesis of n-3 LCPUFA and vice versa. Optimal conversion of ALNA to EPA/DPA

is expected when the diet is low in both n-6 fatty acids and n-3 LCPUFA, particularly

C18:2 n-6. Simultaneous increases in 18:3 n-3 and 18:2 n-6 generally decreasel8:3 n-3

conversion as well (Brenna 2002). From 13C labeled fatty acid studies, the highest levels

of ALNA are found in triacylglycerols (TAG) and the chylomicron-enriched fraction

(CRF) of the triacylglycerols due to ALNA exportation from enterocytes within

chylomicron TAG. Chylomicrons, being the primary transporters that they are, tend to

be reused in this cycle. It was also noted that 13C labeled ALNA persisted in the total

blood plasma TAG for up to 73 hours, suggesting incorporation into other lipoproteins

by the liver. At this point a-linolenic acid becomes a part of the nonesterfied fatty acid

pool from adipose tissue that provides a short term supply to other tissues (Burdge and

others 2002). These activities are also reserved for EPA and DPA since they too can

be identified in both total plasma and CRF TAG as desaturation and elongation occur

within enterocytes. This action is consistent with reports of desaturase activities in

microsomes prepared from human intestinal epithelium (Garg and others 1992). ALNA

is then released into the circulation via chylomicrons by lipoprotein lipase activity.

Plasma cholesterol esters act as a long term source of ALNA while in circulation,









whereas EPA, DPA, and DHA (in women) are primarily associated with

phosphatidylcholine (Burdge and Wooten 2002). In healthy men, it has been observed

that DPA concentrations increase before EPA concentrations within the total plasma

TAG; while there is an opposite effect demonstrated in the CRF TAG. This is possibly

due to LCPUFA associated with the CRF TAG being derived from enterocyte

metabolism, and total TAG. EPA and DPA are the net products of enterocyte and

hepatic (liver action) synthesis which may mask the precursor-product relationship

between the metabolites. The inhibition of DPA to DHA synthesis in healthy males is

attributed to a 3-oxidation step in DHA synthesis that acts as a point of metabolic

control. It is theorized that the down-regulation of this synthesis is attributed to the need

for DHA being met via diet alone negating the need for synthesis (Burdge and others

2002). In direct contrast, for healthy women, there was shown to be an increase in the

conversion of DPA from ALNA due to an up-regulation of the desaturation/elongation

pathway downstream of EPA synthesis by oestrogenn" (Silverstolpe and others 1981).

Variations in metabolic capacity for ALNA desaturation and elongation may be due in

part to different oestrogenn" exposure rather than diet alone. This results in increased

conversion of DPA to DHA, consistent with increased flux through peroxisomal 3-

oxidation steps, which is a proposed locus of pathway control (Sprecher 2000).

2.6.2 Reactive protein interactions

An elevation in C-reactive proteins (CRP) is strongly associated with clinical

definitions of atherothrombotic disease (Libby and others 2002). Atherothrombosis is

defined as atherosclerotic plaque disruption with superimposed thrombosis (blood within

blood vessels) formation (Viles-Gonzalez and others 2004). The C-reactive protein

exerts a direct pro-inflammatory effect on human endothelium. An increase in CRP









levels exhibits synergy with hypercholesterolemia to increase CVD risk in men and

women. N-3 polyunsaturated fatty acids suppress pro-inflammatory cytokine production

by peripheral blood cells and inhibits lymphocyte proliferation. These fatty acids lessen

inflammatory response that are important to the initiation of atherothrombosis. From

epidemiological and clinical studies, it has been found that a-linolenic acids, which are

derived from chia, demonstrate cardioprotective effects (de Lorgeril and others 1994).

Endothelial activation and enhanced expression of cell adhesion molecules are early

events leading to atherogenesis (plaque formation within the arteries) (Cybulsky and

Gimbrone 1991). DHA metabolism from a high ALNA diet results in a dose dependent

inhibition of vascular cell adhesion molecules, E-selectin, and intercellular cell adhesion

molecule-1, to a lesser extent, which have all been shown to have an inverse

association with CRP. Patients with higher CRP levels typically have a diminished

cholesterol lowering response (29%) when compared to patients with lower CRP.

Therefore, higher CRP levels without dietary and other interventions can impose overall

cardio-vascular risk (Zhao and others 2004).

2.7 Consumer Demand for and Attitudes toward Functional Foods

The purchasing public is the most important segment of the U.S. food system.

Their role as food-decision makers has a large impact in the success or failure of

today's marketed food products (Sloan 1994a). Whether marketed food products are

taken home for consumption or left on the grocer's shelf hinges on the food preferences

of the individual consumer. Food preferences play an important role in food selection

because they provide an indication of the amount of satisfaction an individual

anticipates from eating a food. Preferences are a result of physiological and

psychological development and social experiences, and are related to the degree of









liking of a food. Liked foods are those that are familiar, considered pleasant, and are

typically consumed, yielding food preference as a predecessor of consumption (Asp

1999). Attitudes are defined as mental states, learned predispositions, psychological

tendencies, or evaluated judgments about objects which guide behavior toward said

objects (Tudoran and others 2009). Intentions, on the other hand, represent "a willful

state of choice where one makes a self-implicated statement as to a future course of

action" (Baggozi 1983). According to Fishbein and Ajzen (1975), intention is the most

immediate determinant of behavior and implicitly, the most direct predictor of engaging

in a behavior. It is these three ideals, food preference, consumer attitude, and purchase

intention that must be engaged, to render a food product that is repeatedly purchased

by consumers. By appealing to consumer food preferences, gaining positive feedback/

attitudes toward the food product during the initial product trial, purchase intentions can

be directed toward repeat buying. This mind-set is the driving force behind functional

food marketing.

One of the largest trends that has gathered and sustained momentum in the U.S.

is the increasing awareness of the role of diet and proper nutrition to maintain health

and prevent disease. To this end, today's consumers are seeking foods that can

provide added health benefits, thereby creating a market for functional foods.

Functional foods are food items that have been purposefully designed to provide added

benefits beyond normal nutrition value in such a way that it improves health (Urala and

Lahteenmaki 2003) and reduces disease (Diplock and others 1999). Functional foods

are not pills, but remain food and are part of a normal food pattern (Diplock and others

1999). The idea of the health effects associated with functional foods is based on a









single product and its delivery of functional components (Urala and Lahteenmaki 2007).

To take full advantage of this market opportunity three conditions must be met: 1) There

must be a consumer need or problem that requires a solution. 2) There must be an

awareness that the consumer or people have a problem. 3) Consumers must be willing

to spend money to solve the problem or satisfy the need they have identified. All three

of these elements are in place in regard to functional foods. The basic selling

proposition for functional foods is that they promote health in a convenient way (Beck-

Larsen and others 2001). It has been shown that even during times of financial crisis;

consumers are willing to spend for functional benefits. To meet consumer need, pure

ingredient selling is no longer sufficient, today's food companies must sell "value-added

solutions" that are relevant to consumers (Bleiel 2010). According to Urala and

Lahteenmaki (2007), the use of functional food provides a new, convenient tool for

improving health, yielding a pleasure not only based on improved health, but improved

mood and performance with the possibility of preventing disease. Consumers perceive

the fortification as a bonus or stimulus, which provides hedonic expectations, and can

further justify their purchasing decision (Tudoran and others 2009). It has been found

that consumer acceptance of functional foods is not unconditional, the buying public is

not ready to compromise taste for health, the purported benefits do not allow any trade-

offs for flavor. Some of the unwanted flavor characteristics that arise in functional foods

are bitter, acid, astringent and salty off-flavors (Verbeke 2006). 60% of consumers

consider taste more important than nutritional information, at least in most foods (Harris

1997). Other key purchase intention factors include convenience, quality, and

price/value. Even with the health benefits associated with functional foods there are still









some consumers that are skeptical of these products. Previous research reported that

nonusers of functional foods cite a lack of consumer knowledge, low perceived

importance or interest in this food category, and price as a reason to not use the

products. Some consumers tend to have suspicions of the efficacy of the functional

food promise and its possible side effects. Trust in health-related information plays an

essential role in functional food choices, the most crucial factor affecting consumer

acceptance (Niva 2006).

Attitudes, lifestyle factors, demographic factors i.e. age, gender and education,

strongly affect the acceptability or intention to use functional a food. Females, 25+

years, were consistently found to have a higher perceived concern for health regarding

food over men in this same age range and younger, consequently responding more

aggressively to proven health benefits. This finding could be attributed to females

purchasing food for others within a family setting. The purchase intent for consumers

65+years was lower due to a reticence to try new products (Bower and others 2003).

To appeal to those demographic groups that have lower purchase intentions, functional

foods must approach a status of traditionally healthy foods, and fit into an individual's

normal eating pattern to be effective (Patch and other 2005). According to Bower and

others (2003), a higher purchase intent was demonstrated when product sensory

aspects were "liked" more. The increased "liking" of the food product also caused a

significant increase in the price consumers were willing to pay for said products despite

a higher price point due to health benefits.

In 2009, the top four health issues consumers were extremely concerned about

were: retention of mental shapness-65%, heart disease-62%, cancer-61%, and









maintaining ability to continue with normal activity while aging-59% (Health Focus

2009). As these health issues concern consumers, they in turn create a demand for

functional food products. Anticipating this demand, food companies clamor for a share

in this market. In 2009, nearly 46% of food shoppers said they were very concerned

about nutrition, which is up 5% over 2008. The functional food segment also outpaced

the overall food industry growth rate of 1.6%. This observation is due impart to the

functional food bread/grains category's 3%growth and a 2% growth of functional snack

foods (NBJ 2010). According to Mintel (2008), 6 in 10 adults bought a functional food or

beverage, which is up from 2007's 48%. Within the U.S., food and beverage sales in

this segment reached $37.4 billion in 2009, showing 2.7% growth over the prior year

(NBJ 2010). During this period, fiber, omega-3 fatty acids, vitamins, calcium, and

antioxidants were the top five ingredients consumers sought after (Mintel 2009a). By

years end, 2009, the dollar sales for food and beverages touting an omega-3 claim rose

42% (Nielsen 2010).

Omega-3 enriched product market analysis: In today's society it has become

commonplace to find multiple products touting health benefits via the addition of omega-

3 fatty acids lining the average grocer's shelves. While price and taste are the most

important drivers of food and beverage purchase decisions, 62% of American

consumers considered healthfulness to have great or some impact on their choices

according to the 2008 IFIC (International Food Information Council) Food and Health

Survey. This figure is down from 65% in 2007, but still shows despite a sluggish

economy, food choices are seen by consumers as important to underlying good health

(Tecklenburg 2009). As consumers become more aware of the health benefits derived









from fatty acids, many products that include an omega-3 ingredient will begin to find life

and longevity amongst shoppers. Omega-3 enhanced products entered the retail

market in 2003, but it was not until 2006 that such products reached mainstream U.S.

supermarkets. A large number of the first entries into this category came from small

marketers with limited distribution, i.e. health/natural food and specialty stores. Due to

these conditions, sales were not often tracked by in-store scanner data. Substantial

amounts of these products were also sold over the internet and through mail order, two

avenues also not tracked by monitoring services, therefore total sales of omega-

enriched products are estimates from various trade magazines, information from food

distributors and retail grocery chains. The global retail market for omega-3 enhanced

foods and beverages is estimated at $4.6 billion in 2007, a 33.5% increase over 2006

estimates of $3.4 billion (Gray 2009). Packaged Foods projects that the global retail

market of omega-3 enhanced foods will approach $8.2 billion by 2012. This projection

reflects a CAGR (Compound Annual Growth Rate) of 31.7% between 2003 and 2012.

Two factors support the growth of this market: innovations in formulation products using

DHA and EPA, which allows more omega-3 product enhancement, and a growing

consumer awareness of the benefits of consuming this fatty acid. The omega-3 market

is typically dominated by grain-based products, holding 74% of the market in 2007, but

with innovations in formulation techniques other categories are increasing their shares

in the market. The number of milk, nondairy and yogurt beverage product introductions

increased in 2006 from 76 SKU (stock keeping units) to 120 SKUs in 2007. Yogurt

introductions rose from 20 SKUs in 2006 to 68 in 2007. Snack bars, which have been a

mainstay since 2003, increased from 57 SKUs to 72 SKUs introduced in 2007. It is









projected that in the next two to three years that several percent shares will shift as

more beverage, dairy, and oil/butters/margarine/cooking sprays/ spreads continue to

gain market share (Gray 2009).

According to a 2008, IFIC Food and Health Survey, 72% of respondents had

heard of omega-3 fatty acids, which is up from 63% in 2006. The Health and Wellness

Trends Report, published by the Natural Marketing Institute, noted that many

consumers feel as if they are deficient in omega-3s more than any other nutrient

including fiber and calcium. Consumer familiarity with foods that can provide benefits

beyond basic nutrition, or functional foods is at an all- time high, with 92% of Americans

being able to name a food and its health benefits (Constance 2008). This familiarity is

demonstrated in the purchasing tendencies of consumers, as well. In 2008, due to the

dramatic rise in fuel prices, consumers tended to focus on a single shopping venue that

offered the greatest range of the products they needed. As for the purchasing of "better

for you" food products, such as those enriched with omega-3s, convenience and variety

were key factors. A large assortment of these products was initially found in

health/natural food stores, but the issue of convenience began to overshadow other

considerations. In 2007, according to Packaged Facts, traditional super markets

accounted for 49% of all omega-enhanced foods and beverages, followed by

health/natural food stores-31%, mass merchandiser and specialty stores-6% each, club

stores-5%, and convenience, dollar, drugstores and other- 3%.

Data Monitor's Product Scan noted in 2005, 648 SKUs with "high omega-3" or

"high omega" tags or claims introduced into global markets with 346 in the U.S. In

2007, this figure increased to 1,156 global SKUs and 520 in the U.S. With market









introductions such as these, omega-3s have a great potential to become a ubiquitous

component of all fortified foods as calcium and vitamin D are added to many foods. On

the contrary, the omega enriched sector has been plagued by product withdrawals and

misinformation on the part of consumers. It was found that some consumers lacked

trust in the efficacy of the enhanced food products, and the health claims were

perceived as hype rather than being scientifically established. Some products were also

believed to have low dose levels. Recommended dose levels of EPA/ DHA are 400mg-

1000mg per day for adults. Conservative per serving recommendations for most foods

and beverages incorporated with omega-3s is 65-250 mg on the lower end for products

consumed frequently by children and the higher end for adult consumption.

Formulations with less than 65 mg lead to negligible effects (Gray 2009).

Initially, the omega-3 enriched food and beverage industry was dominated by

small, entrepreneurial companies. In recent years, a number of large international

corporations have become involved in this sector. By the end of 2008, more than 500

companies worldwide marketed foods that listed "omega", "DHA", or "EPA" on the label.

Of the 474 products filed globally in 2007, only a few had more than one or two of these

products introduced. The leading manufacturers of nonfish omega products in 2008

were Rockin' Roll Gourmet (32 SKU), Unilever (30 SKU), Navitas Naturals (15 SKU),

Kellogg and Wegman's (both 14 SKU) (Gray 2009).

From a survey conducted by Prepared Foods in 2008, it was found that in 2000 it

took an average of just under ten months for a new product to reach the market, in 2008

this increased to eleven months. Line extensions introduction time frames increased

from five months in 2000 to over six months in 2008. These time frames increases,









according to manufacturers surveyed were attributed to the increasing use of health

claims and the emerging science of the ingredients (Gray 2009).

2.8 U.S. Oilseed Farming Industry Analysis

Currently, chia production within the United States is nonexistent. This section is

meant to provide an overview of the oilseed farming industry within the United States

and demonstrate how chia seed cultivation would fit into this agricultural sector. The

oilseed industry is comprised of establishments that engage in growing fibrous oilseed

producing plants or producing oilseed seeds. This industry is dominated primarily by

soybean production, followed by sunflower, canola, and flax seed oils. Mustard and

safflower oils are produced as well, but to a much lesser extent than the aforementioned

seed varieties. These oilseeds are field crops that are primarily used as a source of

vegetable oil, and also serve as an important source of livestock feed.

Flax, which shows a strong similarity to chia although containing a lower oil

percentage, is the fourth largest crop produced by the industry, with 6.7% of value and

8.5% of the harvested acreage in 2008. In 2005, following an increase in prices,

planted acreage doubled yielding a larger crop and lower prices. Demand has since

remained strong due to its wide array of uses. 2009 saw flax account for 10% of the

harvested acreage and a share of its production value increased 7.9% (Ibis world 2009).

Oilseed processors represent the majority of domestic sales at 48.5% and 51.2%

of sales volume. In 2006, of the total supply of 2,854 million pounds of canola seed

harvested, oil processors produced 8.9 million pounds of oil and 1,400 million pounds of

meal, which was used for livestock feed, food and industrial purposes. The second

largest market for oilseeds is exportation. In terms of volume, seed export accounted

for 32.7% of production in 2006. U.S. oilseed exportation has grown strongly over









current period, but remains a net importer of oilseeds. The value of oilseed exports

grew at yearly 13.1% over the five years to 2008; imports on the other hand, grew at a

yearly rate of 39.2% which produced a widening industry trade deficit from $1.2 million

in 2003 to $7.6 million in 2008. The United States generally exports processed

oilseeds, as larger markets for raw seed exists in Mexico and Canada (Ibis World 2009).

As an agricultural industry, oil seed farming has a low level of ownership

concentration when compared to other economic sectors. The average U.S. oilseed

farm covers 134 acres, with annual sales of $208,000 and a market share of 0.04% (Ibis

World 2009). Concentration levels have begun to rise due to the decrease in farm

number by 6.1 % per year in the five years to 2009. This decrease is attributed to

competition and cost pressures forcing smaller farms to close as the industry moves to

large-scale production. An increase in concentration in oilseed farming has been driven

by greater mechanization and the shift to more farm specialization. Higher

concentrations may lead to greater profitability as fixed costs fall relative to production.

Oilseed crops also tend to require less farm labor, so large farms are more profitable

than smaller farms. The state of North Dakota has dominated the oilseed industry over

the past five years with greater than 50% of the plantings and production nationally (Ibis

World 2009).

The oilseed farming industry has seen revenues peak, plunge, and peak again

before falling in 2009. The value of oilseed farming peaked in 2008, when oilseed

prices were at their strongest. As the demand for food crops for biofuel producers grew,

so did biofuel and feedstock prices. A similar, effect was seen as crude oil prices fell,

the value of using oilseeds in alternative fuel production followed suit. The demand for









oilseeds is a function of activity of the vegetable oil and fat manufactures. Oil

production is an important source of demand for oilseeds, changes in output levels of oil

and fat processors will alter the demand for raw inputs. Like many commodities, the

demand for oilseeds is sensitive to price changes. Large increases in the price of

oilseeds can constrain demand as fat and oil processors move to switch production to

alternate materials such as animal-based fats. Between 2002 and 2007, the average

price of oilseeds such as canola, flax, and sunflower seeds increased in response to

higher demand and smaller harvests. Seed meal is a major component in protein feed

used by commercial poultry and pork producers. Increases in livestock production

yields a greater demand for oil seed for animal feed is volatile and can fluctuate due to

supplies of competing protein feed sources and the availability of natural forage (Ibis

World 2009).

The factors forming the basis of competition among oilseed growers includes

production costs, quality, and range of products produced. Internal competition within

this industry in the U.S. is moderate since growers do not interact strategically.

Typically, no single grower is large enough to singly have a significant effect on the

market. Producers generally respond to price signals from the market to decide

production. Costs of production are a key competitive factor among growers due to the

homogeneity of oil seeds, prices per farmer are the same therefore more efficient

growth yields higher profits. Oilseeds are differentiated by quality standards, grades

and assessments are based on protein and oil contents, premium grades can demand

higher prices. Demand for particular seed varieties can also drive up prices received by

grocers. Recent food safety concerns have enabled producers to charge higher prices









even though quality is difficult to control due to a large number of exogenous factors.

Oilseeds are generally grown in combination with other crops, and a higher demand for

other crops yields more planting s and less oil seeds causing lower production and less

industry revenues. U.S. oilseed growers mainly compete against the soybean industry,

which dominates the oilseed sector. It is forecasted that external competition will rise

due to the demand for vegetable oils for edible and industrial purposes (Ibis World

2009).

Oilseed farming requires special equipment for planting and harvesting. The

introduction of labor-saving equipment has caused the level of capital to increase over

time. This sector of farming is classified as capital intensive and its estimated labor to

capital ratio is approximately 1:6:1. Oilseed producers spend $1.60 worth of labor for

every $1 worth of depreciation incurred on capital equipment. Land, a non-depreciable

asset, has no contribution to the industry's total depreciation costs. Due to the high-

mechanization of oilseed farming, trends have begun to shift toward replacing on-farm

services with off-farm services, both factors reducing labor cost requirements.

According to the 2007 Agricultural Census, plant and machinery on the average oilseed

farm is worth approximately $1,723,863, which is higher than other farming sectors at

$88,357 (2007 Census of Agriculture).

The oilseed farming industry is a minor part of the U.S. grain growing sector, which

tends to be overshadowed by the soybean farming industry. This agriculture sector is

expected to generate $716.6 million compared to the $27.5 billion of the soybean

industry. This disparity is due to many factors, but one of the primary factors is the

limitation of geographic regions suitable for oilseed production. Over the next five years,









demand for oilseeds is expected to benefit from increased demand for food oils without

trans fats. A greater understanding of the effects of trans fats has stimulated demand

for oilseed in U.S. and international markets yielding a strong increase in the price of oil

with lower trans fat content (Ibis World 2009)

As for profitability within this industry, it is estimated to have risen over the current

period. Growth in prices received by farmers has exceeded growth in price of farming

inputs; but wage growth has been constrained. There has also been some varied

profitability over the current period due to price variation, planting, yield and farm inputs.

Farm input costs have risen currently, but crop production costs have fallen from their

heights seen in mid 2008. The falling of sort commodity (commodities grown and not

mined) prices had an equal effect on farm inputs, leaving producers unwilling to pay

high fertilizer and farm chemical prices from the soft commodity boom. Although costs

have increased at a yearly rate of 4.5% over the five years to 2009, they have fallen

about 17% between 2008 and 2009. Strong increases in prices had a negative effect on

farm profitability, however in this industry; prices had risen at a greater rate than farm

input costs, yielding higher profit levels (Ibis World 2009).

After experiencing volatile conditions over the last five years, the oilseed farming

industry is forecasted to experience consistent growth over the upcoming five year

period, with a slight dip in 2013. This growth will be due to the demand for seed oils in

the retail and industrial markets (Ibis World 2009). Chia seed, with its high oil content

and various components for industrial usage could gain a share of this market. Chia

cultivation would provide an alternate crop to be rotated with other crop cultivation

during the year. The primary concern would be locating the proper growing area within









the U.S. and making chia cultivation profitable for producers. This profitability would be

created by the addition of chia meal to livestock feed and chia seed oil use in biofuel

and industrial formulations.

2.9 Consumer Sensory Testing Overview

"Sensory evaluation is a scientific discipline used to evoke, measure, analyze, and

interpret reactions to those characteristics of foods and materials as they are perceived

by the senses of sight, smell, taste, touch and hearing."-Anonymous, Sensory

Evaluation Division of the Institute of Food Technologist. Consumer testing is one of

the most important activities in product development. The primary purpose of consumer

affective tests is to assess the personal response by current and potential customers of

a product, product ideas, or specific product characteristics. Consumer evaluation

concerns itself with testing certain products using untrained people who are or will be

the ultimate users of the product (ASTM 1979). Two types of sensory evaluation exist,

objective and subjective evaluation. Objective sensory evaluations are used to garner a

sense of some sensory attributes. This type of evaluation includes: analytical sensory

tests, expert panel tests, and difference tests. On the other hand there is subjective

testing which also may also be called preference testing. During this evaluation style,

panelists are presented with a choice of samples and must state which sample is

preferred or most accepted. It must be made clear and understood that these two

evaluation methods are not interchangeable (Fuller 2005). Sensory tests are typically

conducted during product development for product development guidance, to screen

products, to identify those products that are disliked and those that match or exceed a

specific target product for acceptance. The overall goal of sensory evaluation in the









food industry is to enhance quality to improve appearance, flavor and texture as it is

perceived by consumers to guide/ influence their food choices (Resurreccion 1998).

Before sensory testing begins, there must be a clear understanding of what is

required from the test. Do the developers want to determine whether they have

produced at the best formulation of a product with respect to the criterion? Or do they

want to determine whether this particular product is as good as, or better than, the

competition product. Once these important questions are answered, then the sensory

evaluation process truly serves its purpose.

2.9.1 Hedonic scale

During taste panel sessions, panelist were asked to use two different nine-point

scales to rate product attributes, the first being the hedonic scale, followed by the FACT

scale. The hedonic scale, also known as a "degree of liking" scale, was developed at

the Food Research Division of the Quartermaster Food and Container Institute of the

Armed Forces, Chicago, Ill. The scale introduced the hedonic value concept, which

refers to the psychological range of dislike, positioned at the lower end of the scale and

like positioned at the upper end of the scale (Peryam and Girardot 1952). The hedonic

scale assumes that consumer preferences exist on a continuum and that preferences

can be categorized by responses based on likes and dislikes (Lawless and Heymann

1998). Initially, the scale was to be used to predict the food choices of soldiers. The

method was designed for people with no food testing experience, hence the

descriptions and scale points. When using the scale, panelists are encouraged to

report an immediate and naive response without any conscious effort to remember or

judge (Peryam and Girardot 1952). The focus of the hedonic scale development was: 1)

to detect small differences in the direct response to similar foods. 2) To detect gross









differences in the direct response to foods, even when time, subject, and test conditions

are allowed to vary. 3) In field questionnaire surveys, were to reveal differences in

group-reference attitudes toward foods (Peryam and Girardot 1952). In developing the

terms for the categories of the scale there was concern about which terms had

consensual meaning among the population. Great care was taken to remove terms that

could be thought of as ambiguous or having double meanings across the population

(Lawless and Heymann 1998). An example of such a term is "average". In an original

study by Jones and Thurstone (1955) test groups equated "average" with "like

moderately" whereas in today's society the same word may have a negative

connotation. The descriptive categories or terms were to have low variability in

meaning, no bimodality and little skew. Developers used Thurstone's model for

categorical judgment as a means of measuring the psychological scale values for words

(Lawless and Heymann 1998). The number values on the scale were developed using

51 words or phrases which formed the candidate list, and taken from a pilot study with

900 soldiers chosen to represent all enlisted personnel. Each phrase was presented on

a form with a rating scale from -4 to + 4 with a check-off format. The subjects were

instructed to read each phrase and assign an integer value from -4 to +4, with zero as

an option. This method presumed that integers were themselves an interval scale of

psychological magnitude. Following the Thurstonian model, which did not use raw

numbers, this scale was converted to standard deviations as units of measure by way of

converting the scale in to z-values. The final assessment of the scale was to have

eleven categories, but due to equipment restrictions nine categories resulted (Peryam

and Girardot 1952). "Government paper was only 8" wide and we found that typing









eleven categories horizontally was not possible. So we sacrificed a modicum of

precision for a real improvement in efficiency at the moment. Probably, at no great

loss"-David Peryam (Peryam 1989).

2.9.2 FACT scale

The second scale used to evaluate the chia seed product was the FACT scale.

This scale is a behaviorally-oriented approach to scaling food acceptability. The scale

is based on attributes and actions, combining statements about frequency of

consumption and motivationally related statements, producing an action-based index of

food acceptance. It was reasoned by scale developer, Howard Schutz of Hunt Foods

and Industries Inc., Fullerton, Ca, that behaviors might not always match-up with

acceptance as scale on the traditional nine-point hedonic scale (Lawless and Heymann

1998) The FACT scale was developed using twenty subjects to rank 18 statements that

reflected an effective action toward food, 1-the most positive food attitude to 18- the

most negative food attitude. The 18 statements were then scaled using Guilford's

composite standard technique (Schutz 1965). Selected from the eighteen statements

were nine statements which represented approximately equal scale intervals,

approximate equality was used because it lends more confidence to statistical analysis

of the ratings. The nine categories also provided a means to make direct comparisons

of the usefulness of the scale to the hedonic scale. This direct comparison was

demonstrated during the FACT scale development process by way of a study using 100

participants to generate means for 54 foods using both scales. The FACT scale's

means were consistently lower than hedonic means, but demonstrated less skew than

the hedonic scale. As for the relationship between mean preference rating and the

percent dislike falling below the midpoint of the scale, these values were the same









shape for both scales. From this study, it was also noted that there was a high degree

of correlation (r=+.97) between the two scales. The two scales were found not to be

interchangeable, but best used in complement as an overall measure of food

acceptance (Schutz 1965). The food action rating scale has been employed in

numerous research projects such as: "Efficacy of fruit purees as partial replacements in

a chocolate cake and cookie recipe." W. Landis and L. Altman, 1996. J. Amer. Diet.

Assoc. 96: a48 and "Effects of fat level and cooking methods on physical and sensory

characteristics of restructured beef steaks." M. Penfield and others, 1988. J. Food

Quality (11) 5: 349-356.



2.10 Objectives

This study has several objectives. The first is to heighten the public's awareness

of the health benefits of chia through incorporation into food products that hold a place

in the normal eating pattern. It has been well-documented that chia seeds contain the

highest known content of a-linolenic acid from a plant source. This fatty acid contributes

to the reduction of low density lipoprotein concentrations in humans (Finley and Shahidi

2001). Reductions such as these also contribute to a smaller risk of coronary heart

disease (CHD), one of the leading forms of mortality. Regular consumption has also

been shown to lower the c-reactive protein in type-2 diabetics, which is a risk factor for

CHD (Kreiter 2005). In the United States, when the word chia is mentioned, the potted

plant product the "Chia Pet" is typically visualized. This is a preconception that can be

beneficial and detrimental in the same instance. The benefit is that the potted plant

product will give consumers a frame of reference so the seed product is not completely

foreign. After this association is built, it could possibly give way to consumers actually









sampling any product containing chia seeds. The association may also create a

curiosity about the seed in general and engage the consumer to become more informed

about chia. On the other hand, the same reference may be constructed, but not by a

positive means. This could lead to the consumer not perceiving chia seriously and

disregarding the seed and products formulated from it as well.

The next objective of this study is to demonstrate that chia seeds can be utilized

in the production of food products that are accepted by consumers. These products

will contain chia seeds in various forms (whole, ground, and combination). High

consumer acceptance could possibly lead to the commercialization of these products.

Although the health benefits of chia are established, it is obvious that these benefits will

not be fully realized unless the seed is actually consumed. The United States Food and

Drug Administration recognizes chia as a foodstuff; ancient recordings as well as

research studies and human testimonials identify the seed as being palatable. After the

products are created and have been found to obtain attributes the consuming public

finds acceptable, the final objective can then be completed.

This objective includes comparing and testing the physical attributes of the

created food products to a control equivalent, measuring the quality of the products.

These bench-marking exercises will provide some sense of how the created products

compare quality-wise with similar products that are not incorporated with the seed.

During this phase of the study, fatty acid content of the chia seed products will also be

analyzed to ascertain which form of the chia seeds maintain the highest level of omega-

3 character after processing and baking.

























H H



Figure 2-1. Alpha linolenic acid structure


Figure 2-2. Docosahexanoic acid structure


H
I
o ,o


Figure 2-3.Eicosapentanoic acid structure











CHAPTER 3
METHODS AND MATERIALS

3.1 Product Formulation

The seed of the Salvia hispanica L. are known for their high omega-3 fatty acid oil

and various other health benefits. Although the benefits are many, without a proper

vehicle the healthful attributes of the seeds would be lost to consumers. The primary

vehicles used in this study to transfer the aforementioned healthful properties were chia

muffin and chia cookie product. The formulation for the control muffin product began

with a basic muffin recipe. The initial muffin control formulation was prepared numerous

times, adjusting ingredient quantities, or removing certain ingredients completely until a

control muffin with acceptable flavor, color, and texture was reached (Table 3-2). These

changes included the addition of a raisin puree for its humectant/ moisture retention

properties. The addition of brown sugar to enhance the sweetness profile and balance

that of the white sugar already present in the formulation. The raisin puree consisted of

golden raisin that were softened in heated water, then blended using a two-speed

Waring Commercial blender (Torrington, CT USA), model 7011S, in a 2.88:1 raisin to

water ratio. Initially, commercially-blended apple pie and cake spice products were used

during the control formulation process, but they were removed and replaced by their

separate components (cinnamon, allspice, and nutmeg) that were scaled individually for

better flavor control. Honey was used as a sweetener, but did not achieve the sought

after sweetness profile. The vegetable oil quantity was reduced by half to reduce the

"greasy" feel left by the control muffin. Melted unsalted butter was added to the

formulation to enhance mouth-feel, top browning and provide crisp edges around the









muffin products (Cross 2006). Although the ingredients and their quantities changed

during the control muffin formulation, the method of preparation remained consistent.

The producers of the ingredients used in the muffin formulation can be found in

Appendix A. The muffin method was used to mix the ingredient components of the

control muffin. This method is a classic technique used in quick bread preparation. All

the dry ingredients were sifted and combined, being careful to completely incorporate

both leavening agents, baking powder and baking soda. Then the wet ingredients were

combined including the sugars, white/brown, and melted butter. Finally, the wet

ingredients were added to the dry ingredients and mixed until all the components were

incorporated, without dry flour pockets, but still lumpy. The pans were muffin sprayed

lightly with canola pan coating and filled % full with the muffin batter using a level #24

scoop. While formulating the muffin control, each muffin batch was prepared to yield six

muffins and was baked in a White-Westinghouse 30" freestanding range (Augusta, GA

USA), model WWEF3002KW at 176.6C (3500F) for 18 minutes. The oven was

preheated at 190.5C (3750F), and then lowered to 176.6C upon introduction of the

muffins. The doneness of each batch was checked by piercing each muffin with a

toothpick, a dry toothpick indicating that the muffins were done. The muffins were left to

cool in the pans for one minute, and then the batch was flipped out and left to cool on

glazing racks. The muffin product was stored overnight in paper towel-lined,

Tupperware containers. When preparing chia muffins, the seeds were incorporated

into the control muffin via the substitution method. According to this method, the chia

seeds were substituted for the flour quantity that was removed, keeping all the other

ingredient quantities consistent. Before the seeds were combined with the other dry









ingredients, they were ground using a Kitchen Aid Blade Coffee Grinder (St. Joseph,

MI USA), model BCG 100WGH and finally sifted to remove unground hulls. The seeds

were added in various quantities in ground, whole or combination whole and ground

forms. It must be noted that the muffin product with whole seeds had no flour removed,

only seeds added. The tested muffin treatments were: 25% ground, 40% ground, 25%

whole seed, and 10% whole/15% ground muffins (Table 3-4).

The other chia seed product formulated during this study was a chia seed cookie.

The original recipe was taken from a sugar cookie recipe found online at allrecipes.com.

The only change made to the original formulation was in its procedure. Originally, the

formulation required no chilling period, but a chilling period of 40-45 minutes was added

to the cookie preparation method. The cookie control was prepared via the creaming

method. After all the dry ingredients were sifted and combined, softened butter was

folded into white sugar until smooth. At this point, the other wet ingredients were

added, and finally the dry ingredients were incorporated in three parts. After the

ingredients were completely combined, the dough was refrigerated for 40-45 minutes to

stiffen the dough. This step was added to help the control cookie hold a uniform shape

and reduce its spreading. Chia seeds were added to the control cookie formulation by

the substitution method also. The seeds were ground and sifted by the same procedure

used for the chia muffin product. When whole seeds were to be added to the

formulation, no flour was removed to compensate for the seed addition. After the

proper chilling time was achieved the cookie dough was portioned using a level #50

scoop. The chia cookies were baked on a nonstick sheet pans in a White-

Westinghouse 30" freestanding range (Augusta, GA USA), model WWEF3002KW at









190C (3750F) for ten minutes. Since the cookies were so small there was no need to

rotate this product for even baking. Initially, there was the issue of the bottoms of the

cookies over browning, but this was alleviated by baking the cookie product on

parchment paper. Chia seeds were added to the cookie product for testing in the

following ratios: 15% whole, 15% ground, and 5% ground/10% whole combination

(Table 3-5). Producers of the ingredients used in the chia cookie formulations can be

found in Appendix A.

3.2 Taste Panel Product Assessment

All cookie and muffin products were evaluated in at the University of

Florida/Institute of Food and Agricultural Sciences- Food Science and Human Nutrition

sensory panel facility. All treatments of both chia products were evaluated on separate

dates using different groups of untrained panelists. Each panelist was given three

randomly coded samples, along with water and crackers for palate cleansing. Initially,

panelists were asked demographic questions (Table 3-1), then presented with a ballot

(Figure 3-3), and were asked to assess the three products before them. Panelists were

asked to assess the samples from left to right using a nine-point hedonic scale (Figure

3-1) for each of the following attributes: overall acceptability, appearance, flavor, and

texture. The two final assessments panelist were asked to make were: to choose how

often they would consume each sample using the Food Action (FACT) rating scale

(Figure 3-2) and to rank all three samples from most to least preferred. A nine-point

hedonic scale, (Figure 3-1) anchored by 1="dislike extremely" and 9="like extremely"

was used to capture panelist opinions of the products. The panelists were also asked to

express their consumption attitudes using the nine-point FACT (Food Action) rating

scale. This scale was anchored by 1="1 would eat this only if forced" and 9="1 would eat









this at every opportunity". These tests were performed using 75 untrained panelists, 31

males and 44 females. The age distribution of the panelists was as follows: 67.7% ages

18-24, 22.6% ages 25-34, 3.2% ages 35-44, and 6.5% ages 45-54. The data obtained

from these sessions was tabulated by the Compusense program, (Compusense Five

3.6 Sensory Analysis Software for Windows, Compusense, Guelph, Canda) two-way

ANOVA, in concert with means separation via Turkey's HSD (Honestly Significant

Difference) test was used to ascertain if there were any significant (p<0.05) differences

between the tested attributes of the chia products from those of the control products.

The null hypothesis (Ho) for the panel testing sessions was "there are no significant

differences between the attributes of the chia muffin products and the attributes of the

control muffin", and the alternative hypothesis (HA) being "there are significant

differences between the attributes of the chia muffin product and the attributes of the

control muffin. "The confidence or alpha (a)-level used in the data analysis was 0.05 or

5%. The determination of significance was rendered via p-value calculations, whereas

the p-value must be less than the alpha level of 0.05. Previous panel testing evaluated

whole seed and combination whole/ground seed muffins. From these efforts, it was

found that consumers consistently preferred ground seed muffins, which led to the test

samples of the final taste panel sessions. The final taste panel assessed the consumer

preference of a control, 40% ground, and a 25% ground chia seed muffin. The

objective in the data collection was to compare the variation across the samples for

each attribute.

3.3 Chia Product Physical Analysis

The chia seed cookie and muffin products were subjected to water activity (Aw),

color, and textural analyses. The water activity of the chia seed products was measured









using the Aqua Lab water activity meter, Pullman, WA 99163 USA, model CX-2. The

meter was verified to be calibrated using a dilute NaCI standard at 20C. The water

activity of each cookie and muffin product sample variation (control, whole seed, ground

seed, and combination whole/ground seed) was measured by three replications over

five trials. Color analysis of the muffin and cookie products was performed using a

Minolta Chroma meter 200B, Ramsey, NJ 07446 USA, which measured the intensity of

reflectance at each wavelength. The colorimeter was calibrated using a white

calibration plate, and was set to record sample L*, a*, b* values, also known as the

Hunter Lab color scale. The L* values representing the lightness/darkness, 0=black and

100= white. The a* values denoted (-) green and (+) red and the b* values measured (-)

blue and (+) yellow of the sample products. Color analysis was performed on the

crowns of the muffing samples and the top surface of the chia cookie product. Each

cookie and muffin product sample variation's color was measured by three replications

over five trials. The texture of the chia muffin products were measured using an Instron

Universal Testing Machine, Norwood, MA 02062 USA, model 4411, at a cross head

speed of 10 mm/min with a static 1.2 kg load cell. A 10mm aliquot using a size #6 cork

borer was removed from each chia muffin sample. Each sample was subjected to

compression testing using a #13 plunger. The samples were compressed twice to 50%

of their original height, according to Grigelmo-Miguel and others, 1999. During this test,

the Instron measured the force needed to compress each sample. The attributes

measured during this test included sample hardness-first compression maximum force,

cohesiveness-ratio second cycle max. load/ first cycle max load, gumminess-hardness x

cohesiveness, and chewiness- springiness x gumminess. Muffin samples underwent









four compression replicates over four trial runs. The chia cookie samples were

subjected to a three-point break test (Gaines 1991). This test was performed using the

Instron noted above and a # 4 blade at a crosshead speed of 50mm/min. The bottom

beams that the chia cookies rested on were spaced 4 cm apart, the diameter of the

cookies averaged 5.4-6.5 cm. The three-point snap test was used to assess the

hardness or brittle nature of the chia cookie product by measuring the force needed to

snap each cookie. Textural analyses of the chia cookie were replicated four times over

four trial runs. The results of the Aw, color and textural analyses were subjected to

ANOVA tables to identify significant (p<0.05) differences between the cookie and muffin

products and their control products.

3.4 Chia Product Fatty Acid Analysis

To ascertain which chia seed variation within the cookie and muffin products

withheld the largest amounts of omega-3 fatty acid character; products were subjected

to fatty acid analysis. Each sample variation underwent this analysis a total of three

times to generate data for statistical analysis. The cookie and muffin products

underwent similar analyses, according to the following standard methods: American

Association of Cereal Chemistry- method 62-05, Association of Official Analytical

Chemists- methods 922.06, 996.06, and American Oil Chemists Society- method Ce

1h-05.

The initial steps in sample preparation are as follows: each muffin sample was

placed on parchment paper and cut into eight equal pieces, cookie samples were

broken into eight pieces ensuring sample crumbs were not lost. The divided samples

were left on the parchment paper to dry at room temperature (20-220C) until the

samples were approximately equal with the moisture of the ambient air, 18-22 hours.









After drying, the samples were ground finely enough to pass through a 20-mesh sieve.

The grinding was executed through the use of a food mill to create as little heat from

friction as possible.

Next, the samples were subjected to acid hydrolysis to extract the fat /lipid fraction.

This was done by placing 2 grams of sample in a 50 mL beaker, and adding 2 mL of

alcohol and stirred to moisten all the particles to prevent clumping. Then 10 mL HCL

were added to the beaker and mixed well. The beaker was then placed in a water bath

held at 70-800 C, and stirred at frequent intervals for 30- 40 minutes. Afterward 10 mL

of alcohol was added and the mixture was allowed to cool. The mixture was then

transferred to a Mojonnier fat-extraction apparatus. The holding beaker was rinsed into

the extraction tube with 25 mL ether, added in three portions. After closing the flask and

shaking for one minute, the mixture was left to stand until the upper liquid was clear.

The ether-fat solution was drawn off through a filter of cotton pledget in the funnel stem,

into a 125 mL flask containing porcelain beads. The liquid remaining in the tube was re-

extracted twice, each time using only 15 mL of ether. These extracts were filtered into

the original beaker. The extracted lipid was then dried in an oven at 100C for

approximately 90 minutes and weighed. The sample was then left to air dry to a

constant weight (30 minutes) and weighed once again. The results are reported as

percent fat by acid hydrolysis.

The next portion of the fatty acid analysis consists of preparing the fatty acid

methyl esters (FAMES) from the extracted lipid. The extracted lipid was dissolved in 2-

3 mL of chloroform and 2-3 mL of diethyl ether. Then the mixture was transferred to a 3

dram glass vial and evaporated to dryness in a 40C water bath. After properly cooled,









2.0 mL of 7% BF3 and 1.0 mL toluene were added to the vial. The vial was then sealed

with a screw cap top with a Teflon/silicone septum. The vial was then heated in the

oven for 45 minutes at 100C, during this heating the vial is gently shaken every 10

minutes. After 45 minutes had elapsed, the vial was cooled to room temperature and

5.0 ml of water, 1.0mL of hexane, and 1.0g Na2SO4are added. The vial was then

capped and shaken for one minute. After the layers had time to separate, the top layer

was transferred to another vial containing 1.0g Na2SO4. The removed top layer contains

the prepared fatty acid methyl esters. Once completely collected, 1 L (15-20 pg

equivalent) of sample was injected into the DSQ Trace GC Ultra (GC), Fisher Scientific,

Pittsburgh, PA 15275 USA. The GC operating conditions used during experiment trials

are as follows: injection port temperature-2500C, detector temperature- 250C, and

oven temperature- 180C. The carrier gas used was helium, with a column head

pressure of 286 kPa (41 psi), flow rate of 1.0 mL/min, 26 cm/s linear velocity, a split ratio

of 100:1. The individual FAMES were identified via the use of their retention times and

by comparison with mixed and individual FAME reference standards, all standards were

obtained from Nu Check Prep, Elysian, MN 56028, USA. The reference standards used

for FAME comparisons are as follows: triglyceride internal standard-C11:o triundecanoin,

FAME mixed standard solution-GLC-85, Nu Check Prep, MN 56028, USA; C11:o FAME

standard solution- C11:o undecanoic methyl ester, individual FAME standards-C4:o-

tetranoic methyl ester, C6:0- hexanoic methyl ester, C14:0- tetradecanoic methyl ester,

C14:1- 9-tetradecenoic methyl ester, C18:o- octadecanoic methyl ester, C18:1-9-

octadecenoic methyl ester, C18:2-9,12-octadecadienoic methyl ester, C18:3- 9,12,15-

octadecatrienoic methyl ester, C20:1-8-eicosenoic methyl ester, C20:2-11,14-









eicosadienoic methyl ester, C20:3-11,14,17- eicosatrienoic methyl ester and C22:0-

docosanoic methyl ester.









Table 3-1. Demographic Questions.

Question # 1.
Please indicate your gender.
O Male
0 Female

Question # 2.
Male:
Please indicate your age range.
o18-24
o 25-34
o 35-44
o 45-54
o 55-64
o Over 65

Question # 3.
Female:
Please indicate your age range.
018-24
o 25-34
0 35-44
0 45-54
0 55-64
o Over 65











dislike dislike dislike dislike
extremely very much moderately slightly


neither
like nor
dislike


1 2 | 3 | 4 | 5 | 6 7 | 8 | 9
Figure 3-1.Hedonic nine point scale




Table 3-2. Food Action Rating Scale


like like like very
slightly moderately much


like
extremely


I would eat this product only if I were forced to.

I would eat this product if there were no other foods choices.
I would hardly ever eat this product.

I don't like this product, but would eat it on occasion.

I would eat this if available, but wouldn't go out of my way.

I like this product and would eat it now and then.

I would frequently eat this product.

I would eat this product very often.

I would eat this product every opportunity I had.









Question #4-Sample <>


Please indicate how much you like or dislike the following attributes in

sample <>

Overall Acceptability


Dislike Dislike Dislike Dislike
extremely very much moderately slightly

I1 2 | 3 W 4 |


Neither
like nor
dislike


Like Like Like very
slightly moderately much


I 5 E 6 I 7 W 8


Appearance


Dislike Dislike
extremely very much

1 I1 2 ]


Dislike
moderately

3


Dislike
slightly

1 4


Neither
like nor Like
dislike slightly

I 5 I 6


Like Like very
moderately much

S7 1 8


Flavor


Dislike Dislike Dislike Dislike
extremely very much moderately slightly

I1 2 ] 3 W 4 |


Neither
like nor
dislike


Like Like Like very
slightly moderately much


I 5 I 6 E 7 8


Like
extremely

9 1


Texture


Neither
Dislike Dislike Dislike Dislike like nor Like Like Like very
extremely very much moderately slightly dislike slightly moderately much

I 1 2 ] 3 4 I 5 | 6 | 7 [ 8
Figure 3-2. Sample ballot used in chia product taste panel testing


Like
extremely

9 i


Like
extremely

9 i


Like
extremely

9 I










Table 3-3. Evaluated chia muffin product formulations
40% Ground seed 25% Ground seed 25% Whole seed 10%Whole/15% Ground
Ingredient Control (g) (g) (g) (g) seed (g)
All purpose flour 125 75 93.7 125 93.75
Granulated sugar 100 100 100 100 100
Brown sugar 25 25 25 25 25
Liquid eggs 40.6 40.6 40.6 40.6 40.6
UHT milk 125.1 125.1 125.1 125.1 125.1
Salt 1.7 1.7 1.7 1.7 1.7
Baking powder 2.3 2.3 2.3 2.3 2.3
Baking soda 1.15 1.15 1.15 1.15 1.15
Vegetable oil 7.3 7.3 7.3 7.3 7.3
Vanilla extract 2 2 2 2 2
Almond extract 1.5 1.5 1.5 1.5 1.5
Golden raisins 15.7 15.7 15.7 15.7 15.7
Nutmeg 1 1 1 1 1
Cinnamon 1.5 1.5 1.5 1.5 1.5
Allspice 1.5 1.5 1.5 1.5 1.5
Butter 46.7 46.7 46.7 46.7 46.7
Chia seeds 0 50 31.2 31.2 12.5/18.7

Table 3-4. Evaluated chia cookie product formulations
15% Ground 15% Whole seed
Ingredient Control (g) seed (g) (g) 10%Whole/5% Ground seed (g)
All purpose flour 70 59.5 70 59.5
Granulated sugar 60 60 60 60
Liquid eggs 14.1 14.1 14.1 14.1
Baking powder 0.5 0.5 0.5 0.5
Baking soda 1 1 1 1
Vanilla extract 1.5 1.5 1.5 1.5
Chia seeds 0 10.5 10.5 3.5/7









CHAPTER 4
RESULTS AND DISCUSSION

4.1 Consumer Taste Panel Testing

4.1.1 Chia seed muffin

After the formulation of the control muffin was completed, variations of this control

were tested for consumer acceptance at the University of Florida (UF)/ Institute of Food

and Agricultural Sciences (IFAS), Food Science and Human Nutrition (FSHN) Taste

Panel Facility, Gainesville, FL. The muffin products were assessed by panelist using

acceptability and preference testing. Three different muffin samples (control, 25%

ground, 40% ground), randomly number-coded, were presented to the panelist.

The first attribute to be assessed by panelists was overall acceptability. The

hedonic means for this attribute were between 6.16-6.41, correlating to likes slightly on

the hedonic scale, with standard deviations from 1.39-1.69 (Table 4-17). From

calculations, the p-value for this attribute was 0.42, indicating no significant difference

across the muffin samples for this attribute. To confirm this finding, a Tukey's HSD test

was run and similar results were noted at the 5% alpha level. The Tukey's value for this

attribute was 0.507, no differences between any of the calculated means (Table 4-18)

was larger than this value. These findings led us to accept the null hypothesis that there

were no differences between the chia muffin products and the control as it pertained to

this attribute. Some of the panelist did point out the bread-like nature of the muffin

samples in regard to the overall acceptability of the muffin samples: "Drier muffin that

appears dry. More like bread i.e. banana bread than a muffin." There was no question

in the demographic section that asked if the panelist actually liked muffins and how

often they consumed them. This line of questioning may have been helpful in









ascertaining consumer preferences of muffins in general. In regards to the

aforementioned panelist comments, it must be noted that muffins are a type of quick

bread, hence the bread-like attribute. The next attribute assessed was muffin

appearance. The hedonic means of this attribute were between 6.07 and 6.81, with

standard deviations between 1.1-1.5 (Table 4-17), which correlates to the phrase "like

slightly" from the hedonic scale. The p-value for the appearance attribute was 0.0002,

leading us to reject the null hypothesis that there are no differences between the chia

muffin products and the control muffin due to a significant difference among sample

treatments. To determine where this difference was, mean separation via the Tukey's

test was consulted. The Tukey's value was 0.42 at a 5% alpha-level. It was elucidated

that there was a significant difference between the control and 40% ground seed

muffins. The 25% ground seed muffin was not significantly different than either of the

other two treatments (Table 4-19). This result was expected due the higher ground chia

seed content of the 40% muffin which gave this sample a highly noticeable, darker

color, different than that of the control muffin. There were only two negative comments

from panelists about the color of the 40% muffin, citing the sample as having "the worst

look". As for the attribute of flavor, the hedonic scale means (Table 4-17) were between

6.08 and 6.47, with standard deviations between 1.60-1.83, which translates to "likes

slightly' on the hedonic scale. The p-value for the flavor attribute at an alpha-level of

0.05 was 0.26. A p-value such as this is an indication of no significant difference across

the sample treatments, leading us to fail to reject the null hypothesis that there were no

differences between the chia muffin products and the control muffin. This was further

verified by the Tukey's HSD test, rendering a value of 0.61 at a 5% significance level.









In regard to flavor, panelists gave mixed comments. A portion of the panelists

expressed satisfaction with the flavor, while others found the samples to be bland.

During initial control formulations, this is an issue that was dealt with on several

occasions. The control muffin was noted to have a good initial flavor that fell flat quickly.

It is possible that the addition of higher spice levels and more sugar could solve this

issue. There were a number of panelist that commented that the samples could have

been sweeter. These comments are understandable since it has been reported that the

American palette prefers sweeter products than other countries. Although all

components of the muffin samples were held constant except the level of chia seed

addition, increased levels of sugar seem to be a sensible addition to samples with

higher ground seed levels. The texture attribute results were very similar to those of the

flavor attribute. The hedonic scale means for this attribute were between 6.00 and 6.29

with standard deviations of 1.61-1.74 (Table 4-17). The mean values displayed a "likes

slightly" assessment across the various muffin treatment textures. At an alpha-level of

0.05, the p-value of this attribute was 0.47, which demonstrates no significant

differences between the sample treatments, leading us to fail to reject the null

hypothesis that there were differences between the chia muffin products and the control

muffin. Although unnecessary, the Tukey's value was 0.572 (Table 4-21), confirming

the p-value result. Both chia muffin samples received positive comments about the

moistness of the samples, but it seemed from their comments that panelist found the

40% muffin to be the moister of the samples, even though this is not displayed in the

data. The 25% muffin received comments pertaining to the dryness of the sample, and

this observation is supported by the % moisture loss data (Appendix B). From this data,









it was found that a higher content of ground seeds led to higher moisture retention. The

Food Action (FACT) rating scale was used to determine the consumption behavior of

panelists toward the three muffin treatments. The FACT scale means for the muffin

sample were 4.91-5.21 with standard deviations from 1.46 to 1.86 (Table 4-17). These

mean values coincide with consumption behaviors equivalent to those described as "I

would eat this if available, but would not go out of my way." The p-value for the panelist

consumption behavior was 0.36, indicating no significant differences among consumer

consumption attitudes across the muffin treatments, allowing us to fail to reject the null

hypothesis that there were no differences between the chia muffin products and the

control muffin. Taking the aforementioned panelist comments into consideration and

adjusting the samples accordingly may have a positive effect on consumption behavior

data. In the final assessment panelist were asked to rank the muffin samples from most

to least preferred. The data obtained from panelist subjected to the Friedman's Test, a

non-parametric randomized block analysis of variance. The null hypothesis for this test

was "there is no difference in rank data for the repeated measure (the actual ranking)."

The calculated Friedman's statistic for this evaluation was 1.62 with 2 degrees of

freedom. At an alpha level of 0.05, the p-value of 0.44 at the aforementioned alpha-

level indicating no significant difference in the rank data (Table 4-17). This finding

allows us to fail to reject the null hypothesis.

4.1.2 Chia seed cookie

Once the cookie formulation process was completed, and a viable control product

was produced, sensory testing commenced. Sensory panel testing was executed at the

UF/IFAS FSHN Taste Panel Facility, Gainesville, FL. The chia cookie product was

assessed through the use of acceptability and preference testing. The cookie samples









were evaluated by a total of 75 untrained volunteer panelists, 43 male and 32 females.

The age distribution of these panelists is as follows: 76.7% ages 18-24 and 23.3% ages

25-34.

Panelist received three randomly coded samples (control, 5%ground/10%whole

seed, and 15%ground/15%whole seed) with crackers and water for palate cleansing.

Once the samples were received, panelists were asked to assess each sample from left

to right using a nine-point hedonic scale (Figure 3-1) to rate the following attributes:

overall acceptability, texture, flavor and appearance. At the conclusion of the product

assessment, panelist were asked to give an overall ranking of the three samples and to

comment on their consumption behavior towards the three samples using the Food

Action (FACT) rating scale. From preliminary testing, the final sample field was

narrowed to contain a control, 15% ground/15% whole seed, and 5% ground/10%whole

seed cookie. The compiled data was analyzed using the Compusense program via

one-way Anova and the Tukey's HSD (Honestly Significant Difference) test for mean

separation. Cookie sample rank data was analyzed using the Friedman's test, a non-

parametric randomized block analysis of variance. The alpha-level for the taste panel

data analysis was 0.05 with significance demonstrated with p-values less than the

assigned alpha-level. The null hypothesis (Ho) of this test was"there are no significant

differences between the attributes of the chia cookie treatments versus the attributes of

the cookie control." The alternative hypothesis (HA) was as follows: "there are

significant differences among the attributes of the chia cookie treatments and the

attributes of the cookie control."









The first attribute of the chia cookie samples to be assessed was overall

acceptability. The hedonic means of the treatments for this attribute ranged from 5.45-

6.29 with standard deviations ranging from 1.57-1.73 (Table 4-24). On the hedonic

scale the mean values translate to the descriptors: "neither like nor dislike" and "like

slightly". The p-value, at an alpha-level of 0.05, for this attribute was calculated to be

0.0001, indicating a significant difference among the treatments. This allowed us to the

reject the null hypothesis of there are no differences among the attributes of the chia

cookie treatments vs. the control cookie. Mean separation via Tukey's HSD test

rendered a HSD value of 0.483. It was found that at a 0.05 a-level there was a

significant difference between the 15% ground/ 15% whole chia seed cookie and the

other two treatments, control and 5%ground/10% whole seed cookie samples (Table 4-

25). The control and 5% ground/10% whole seed cookie had no significant differences,

but according to these results we must reject the null hypothesis in regards to this

attribute. From panelist comments, it was noted that several panelist did not like the

quantity of or the addition of whole chia seeds in the cookies. One panelist commented,

"Seemed like a healthy cookie which turned me off immediately, cookies shouldn't be

healthy." The next evaluated attribute was cookie appearance. The hedonic means fell

within 5.67-7.00 with standard deviations between 1.44 and 1.83 (Table 4-24). Means

of this nature correlate to the descriptors "like slightly" and "like moderately" on the

hedonic scale. The p-value for this attribute, at the aforementioned a-level was

calculated to be 0, demonstrating a significant difference across the treatments.

According to the Tukey's test (Table 4-26); the control sample was significantly different

from the other two treatments leading us to reject the null hypothesis that there are no









differences between the attributes of the chia cookie treatments vs. the control cookie.

This result was expected due to the presence of whole seeds in the chia cookie

samples and the slightly darker color from the ground seed component. The third

attributed assessed by panelists was flavor. The means for the three treatments ranged

from 5.41 to 5.96 with standard deviations between1.62 and 2.02 (Table 4-24).

According to the hedonic scale the mean values correlate to "neither like nor dislike"

and "like slightly". The calculated p-value for this attribute at a 5% alpha-level was 0.03

(Table 4-24). After employing the Tukey's HSD test for means separation, it was found

that there were no significant differences between any of the cookie treatments. From

the calculated data we must fail to reject the null hypothesis that there are no

differences between the attributes of the chia cookie treatments vs. the control cookie

as it relates to this attribute. Panelist often commented on an unfamiliar flavor, possibly

that of the chia seeds. Since the 15%/15% cookie contained the higher chia content, it

consequently was set apart from the other samples. Some panelist went as far as

calling the unfamiliar flavor "unpleasant". The next evaluated attribute was cookie

texture. The treatment hedonic means for this attribute ranged from 5.52 to 6.92, with

standard deviations from 1.50 to 1.81 (Table 4-24). Attribute hedonic means of this

nature coincide with the hedonic phrases: "like slightly" and "like moderately",

respectively. The p-value for this attribute at an a-level of 0.05, was 0.0 indicating a

significant difference between treatments. This difference was pin-pointed using the

Tukey's HSD test. The Tukey's value was 0.495, and it was found that the control and

5% ground/10% whole seed cookies were both significantly different than the 15%

ground/15% whole seed cookie, but not from one another (Table 4-28). This once again









leads us to reject the null hypothesis that there are no differences between the

attributes of the chia cookie vs. the control cookie. This separation was expected, one

panelist did comment on the 15%/15% cookie having a "weird grainy" texture. The food

action rating scale was used to assess the consumption behavior of the panelists as it

related to each cookie treatment. The FACT scale means for the treatments ranged

from 4.35-4.96, with standard deviations from 1.56-1.83 (Table 4-24). These means

coincide with the following the FACT scale phrases: "I don't like this product, but would

eat on occasion" and "I would eat this if available, but would not go out of my way",

respectively. The p-value for this rating was 0.008 at an a-level of 0.05, indicating a

significant difference across the treatments (Table 4-24). The 15% ground/ 15% whole

seed cookie was found to be significantly different when compare to the control and

5%ground/10% whole seed cookie, hence rejecting the null hypothesis of there being

no differences between the attributes of the chia cookie treatments vs. the control

cookie (Table 4-29). Finally, panelists were asked to rank the three cookie samples

from most to least preferred. The rank totals ranged from 122-180 (Table 4-24). The

data generated from this test was analyzed via the Friedman test, which rendered a

calculated value of 22.5 versus a critical value of 5.99, with a p-value of 0. A

comparison of the two values yields significant differences across the treatments. A p-

value such as this at an a-level of 0.05 leads us to reject the null hypothesis due to the

differences across the treatments. The 15%whole/15% ground cookie product was less

preferred the other two sample treatments when rank totals were compared. No

differences were found among the control and 5%whole/10%ground chia cookie

product. From the data tabulation, it was found that the higher chia content in the









15%/15% cookie, even though not much higher than the 5%/10% cookie, had attributes

that were consistently significantly different than the control and 5%/10% cookie

samples during panelist evaluations.

4.2 Product Physical Analysis

4.2.1 Chia Muffin Analysis

Consumer preference data generated from taste panel testing was analyzed to

determine which chia muffin sample consumers found most acceptable. From this data,

it was determined that the 25% ground chia muffin was the most accepted by

consumers. With this knowledge, other chia muffin products were produced with 25%

chia being added to these products in different forms, i.e. ground, whole and

combination (whole and ground). Each form of chia seed was added to the control

formula via the substitution method except for the whole seed muffin. For this muffin,

whole seeds were added without removing any of the flour component. Each variety of

chia muffin was subjected to a battery of tests: water activity (Aw), L*a*b* color analysis

and textural analysis. During the chia muffin product's physical analysis, the null

hypothesis (Ho) that was tested is as follows: "there are no significant (p>0.05)

differences between the tested physical attributes of the chia product samples versus

the control muffin. The alternate hypothesis (HA), being the opposite of the null was

"there are significant (p<0.05) differences between the tested physical attributes of the

chia product samples versus the control muffin. To fail to reject (accept) or reject the

null hypothesis, the p-value, calculated from one-way ANOVA tables was used at the

significance level of 0.05 to measure data consistency. The p-value is defined as "the

probability of obtaining a value of the test statistic that is as likely or more likely to reject

Ho as the actual observed value of the test statistic. This probability is computed









assuming that the null hypothesis is true (Ott and Longnecker 2004). If significant

differences were obtained, the data was then subjected to means separation via the

Duncan's Multiple Range test. The coefficient of determination (r2) was also consulted

to ascertain the magnitude of correlation strength of the tested variable and the seed

form of the muffin product.

4.2.1.1 Water activity

Perishability is a major concern for all packaged products. One determining factor

of product perishability is the association of water with non-aqueous constituents of said

products. The strength of this association dictates the shelf-life of a product due to its

support of activities that lead to product degradation e.g. microorganism growth,

oxidation, and Maillard browning. To measure this association, a ratio of vapor

pressure in the water of the sample/ the vapor pressure of pure water (p/po) is employed

which is also known as water activity (Aw) (Fennema 1996).

During the water activity testing of the muffin products, an aliquot of each muffin

was removed and placed in a testing receptacle. Water activity data was measured by

the Aqua Lab Cx-200 at room temperature (20C). A total of five trials were run with

three replications within each trial for each muffin type. The data generated was

collected and enter into the SAS program. Using this program, a two-way ANOVA was

performed comparing variation across the treatments. During this comparison there

were 60 observations in total. The mean Aw reading of the muffin samples was 0.87.

When the p-value of 0.97 was calculated at the 0.05 alpha level, no significant

differences were found amongst the water activity levels of the muffin samples. To

further confirm these results, a means separation test was run using the muffin water

activity (Table 4-2). Once again, there were no significant differences found between









the muffin treatments, leading us to accept the null hypothesis that there are no

differences between the physical attributes of the chia product samples vs. the control

muffin. The coefficient of determination (r2) displayed a lack of correlation (0.004)

between the Awvalues and the various muffin sample treatments. The results of this test

were attributed to the samples being very similar, it was hypothesized that the whole

seed products may render some variation in Aw due to the polysaccharide mucilage that

is excreted from chia seeds once immersed in a polar solvent. During these trials this

was not the case.

4.2.1.2 Color analysis

The surface color of each muffin crown was measured using the Minolta Chroma

meter 200B. The color of each muffin treatment was analyzed over five trials with three

replications during each trial. During the test runs, light/darkness (L*), red/blue (a*), and

yellow/green (b*) light reflectance were measured. Once these attributes were

measured, the compiled data was analyzed using the SAS program. A one-way

ANOVA was employed to compare variation across the treatments for each color

variable analyzed: L*, a*, and b*. The mean L* value, a measure of lightness

(100=white) and darkness (0=black), of the chia muffin products was 42.93. When

accounting for the calculated p-value (<0.001), a significant difference was found

between the chia muffin products for the attribute of light/darkness of color. This finding

was also confirmed using a separation of the means (Table 4-3). At an a-level of 0.05,

the control and 25% whole seed muffins were not significantly different from each other.

The 25% ground seed muffin and the combination (ground and whole seed) muffin were

not significantly different from each other either. On the other hand, both

aforementioned pairs of muffins were significantly different from each other in terms of









light/darkness values, leading us to reject the null hypothesis that there are no

differences between the physical attributes of the chia product samples vs. the control

muffin. The coefficient of determination (r2), demonstrated a rather small (0.48) yet

positive correlation between the L* color value and the muffin sample treatments. This

result was expected due the similar nature of the paired muffin samples. The pairs

have similar light/ dark measurements because their differences are quite minute. The

mean a* color value of the muffin samples was 5.49, yielding more red reflectance than

green. The calculated p-value of this color variable was 0.0008, demonstrating

significant differences between the sample treatments at the 0.05 alpha level.

Separation of the sample a* color value means via the means separation test (Table 4-

4), also at an a-level of 0.05, was performed. This test concluded that all of the sample

treatments were significantly different than the control muffin, rejecting the null

hypothesis. The coefficient of determination (r2) for the a* value was 0.257, although a

positive correlation, the magnitude between the a* color value and muffin treatments is

very low. The final color value, b*, is a measure of yellow and blue light wave

reflectance of the chia muffin samples. The mean b* color value of the muffin samples

was 16.79, resulting in a higher reflection of yellow light than blue light. The calculated

p-value for this color variable at a 0.05 alpha level was less than 0.0001, demonstrating

a significant difference between muffin sample b* color values. According to Duncan's

Multiple Range test (Table 4-5), the control and 25% whole muffin samples were not

significantly different from each other in regard to their b* color value. The combination

and 25% ground muffin samples were not significantly different from one another either,

but both samples were significantly different than the two aforementioned samples









leading to the rejection of the null hypothesis. Visually, the control and whole seed

muffins looked very similar. Each of these samples were golden brown, with slightly

darker edges, the only different between the two samples visually was the appearance

of the black and white chia seeds throughout the whole seed muffin. The combination

and ground seed muffins were equally darker brown than the aforementioned samples,

there only contrast was the whole seeds of the combination muffin.

4.2.1.3 Textural analysis

The texture of each chia muffin product was tested and compared to that of the

control muffin. All muffins samples had a 10 mm aliquot removed from the whole muffin

using a #6 cork borer. Texture measurements were performed over four trials with four

replicates using the Instron Universal Testing Machine, model 4411. During each trial,

the 10 mm muffin aliquot was compressed twice to 50% of their original height using a

#13 plunger. The textural attributes assessed were hardness, springiness,

cohesiveness, gumminess, and chewiness. The null hypothesis for this test was "there

were no significant (p>0.05) differences between the tested attributes of the chia muffin

products versus the control muffin. After testing each muffin sample, the data was

compiled and analyzed with SAS using a two-way ANOVA, comparing variation across

the treatments for 64 observations. The first attribute to be analyzed was muffin

hardness, which can also be defined as first cycle (first compression) maximum force

measured in newtons (N). The mean hardness of all the muffin samples was 1.01 N.

The calculated p-value for this attribute was 0.0079, indicating a significant difference

amongst the samples. After applying a means separation test to the data (Table 4-11),

it was concluded that the combination (10% whole/ 15%gound seed) muffin was the

only sample that was statistically different from the control muffin. It is thought that this









result is due to the removal of a portion of the flour component in the combination muffin

and replacing it with ground chia, thereby weakening the muffin matrix. It can be

argued that the same action takes place in the ground seed muffin, but the whole seeds

tend to create inconsistent leavened cavities that weaken the muffin matrix. Although

only one sample was different from the control, it was enough to reject the null

hypothesis. The r2 value for this attribute was 0.17, a positive but weak correlation

between muffin hardness and the various forms of seed in the muffin products. The

next tested attribute was muffin springiness or recovery (S) in millimeters. The overall

mean muffin springiness was 5.42S, and at an alpha level of 0.05, the p-value was

calculated to be 0.02. A p-value as such demonstrates a significant difference in the

springiness across the muffin samples. This difference was found to be between the

25% whole muffin and the ground and combination muffins, according to a separation of

the means (table 4-12). The control muffin was found not to be significantly different

than any of the individual muffin samples. A result such as this leads us to fail to reject

the null hypothesis that there are no differences between the physical attributes of the

chia product samples vs. the control muffin due to the means separation which confirms

the null. The r2 value for muffin springiness was 0.14, positive, but weak correlation

between muffin springiness and the different seed treatments. Cohesiveness is defined

as a ratio of the second cycle energy maximum load (J) / first cycle energy maximum

load (J). Each individual compression is defined as a cycle. When comparing the

cohesiveness of the chia muffin products and the control muffin, the mean value of 2.53

J. At an alpha level of 0.05, the p-value for this attribute was 0.84, yielding no

significant difference across the muffin treatments. This finding leads us to fail to reject









the null hypothesis. The cohesiveness r2 value was 0.01, indicating a lack of correlation

between muffin cohesiveness and the seed treatments. Muffin gumminess is defined

as first cycle maximum force x cohesiveness and is measured in newtons (N). During

testing, this attribute was found to have an overall mean of 2.36 N, with a p-value of

0.17 at an a-level of 0.05. A p-value of this magnitude indicates that there are no

significant differences across the muffin treatments for the gumminess attribute. The r2

value for this attribute was 0.07, displaying a lack of correlation between gumminess

and the various seed treatments. The final analyzed attribute, chewiness is related to

how a consumer perceives a product in terms of labored mastication. Chewiness is

defined as springiness xgumminess (N x mm). This attribute had an overall mean of

12.84 Nxmm, and a p-value of 0.03 at a 0.05 alpha level. This p-value is rather close to

0.05, but still not close enough to result in an insignificant difference amongst

treatments. According to the means separation test (Table 4-15), the 25% whole muffin

and the combination muffin were significantly different from one another leading, but

neither sample was significantly different than the control, according to the means

separation, leading us to fail to reject the null hypothesis. The r2 value for this attribute

was 0.12, resulting in a weak correlation between chewiness and the different seed

treatments.

4.2.2 Chia Cookie Analysis

From consumer preference data, it was determined that the chia cookie product

with a 15% seed addition was most preferred by consumers. In response to this finding,

three chia cookie products were produced with 15% seeds added in whole, ground, and

combination 5% whole and 10% ground forms. The seed additions were rendered via

the substitution method, substituting the added seeds for an equivalent amount of flour.









The whole cookie product was the only treatment with no flour removed; the 15% seed

addition was combined directly to the control formulation. After the various cookie

samples were produced each variation was subjected to water activity (Aw), L*a*b* color

analysis, and a triple-beam snap test. The data collected from these tests were

compiled and analyzed with SAS. The p-value generated from ANOVA tables was used

to determine significance (p<0.05) differences among samples. During the chia cookie

product's physical analysis, the null hypothesis (Ho) that was tested is as follows: "there

are no significant (p>0.05) differences between the tested physical attributes of the chia

product samples versus the tested physical attributes of the of the control cookie. The

alternate hypothesis (HA), being the opposite of the null was "there are significant

(p<0.05) differences between the tested physical attributes of the chia product samples

versus the tested physical attributes of the control cookie. To fail to reject (accept) or

reject the null hypothesis, the p-value, calculated from two-way NOVA tables was used

at the significance level of 0.05 to measure data consistency. If significant differences

were obtained, the data was then subjected to means separation via the Duncan's

Multiple Range test. The coefficient of determination (r2) was also consulted to

ascertain the magnitude of correlation strength of the tested variable and the seed form

of the cookie product.

4.2.2.1 Water activity

The four cookie samples (control, whole, ground, and combination seed) were

subjected to water activity (Aw) analysis. An aliquot of each cookie was used for this

analysis which was conducted at room temperature (20C). Water activity data was

obtained through the use of the Aqua Lab Cx-200. The Aw of each cookie sample was

measured over five trials with three replications during each trial. The mean Aw of the









cookie samples was 0.47. Upon separating the means (Table 4-6); it was found that the

control cookie and the whole seed cookie were not significantly different. The results

lead us to reject the null hypothesis that there are no differences between the physical

attributes of the chia product samples vs. the control cookie The ground and

combination cookie products were not significantly different from each other, but these

two variations were different than the control and whole seed cookies (Table 4-6). The

water activity results were as such because of the similar nature of the cookie samples.

The whole and control cookies only differed due to the addition of the whole seeds.

While the combination and ground seed cookies shared the ground seed component

leading the samples water activities to mimic on another. The r2 value of this test was

0.05 (Table 4-1) indicating a positive, but weak correlation between product Aw and the

various forms of seed used in the cookie products.

4.2.2.2 Color analysis

Each chia cookie sample was subjected to L*a*b* color analysis. This analysis

was carried out using the Minolta Chroma meter 200B. To obtain product color data,

the colorimeter was calibrated with a white calibration plate, and the intensity of

reflectance of each wavelength of the cookie surface was measured. The

measurements were rendered using the Hunter Lab color scale producing L*a*b*

values. The L* value is an indicator of product lightness (100=white) and darkness

(black=0), the mean L* value of the chia seed cookie products was 58.29. The p-value

of this test at the 0.05 a-level was <0.0001. This finding indicates that there is some

significance between the L* value, light/darkness of the chia cookie samples. To

confirm this significance, a means separation test was employed. From the Duncan's

mean separating technique, at a 5% a-level, the control and 15% whole cookie products









were found not to be significantly different from each other (Table 4-7). This was also

the case between the 15% ground and combination (whole and ground) chia cookie

products. A significant difference was found between the pairs of aforementioned

cookie treatments leading us to reject the null hypothesis. These results followed the

same pattern as those of the water activity testing, due to the similar nature of the

components of the samples. The r2 value for this tested variable was 0.33, which is an

indication of a positive, but weak correlation between the L* value and he different seed

treatments. The next variable in the color analysis to be considered is the a* color

value. The mean a* value for the chia cookie products was 12.05 (Table 4-1),

demonstrating more red light reflectance than green. At an a-level of 0.05, the p-value

for a* was 0.38, no significant differences between samples (Table 4-8). This result was

confirmed by the means separation test as well, leading us to accept the null hypothesis

that there are no differences between the physical attributes of the chia product

samples vs. the control cookie. The r2 value for the a* color value was 0.05, which

shows a lack of correlation between the a* color variable and the different chia

treatments used in the cookie products. The final color variable b* was measured to

have a mean of 25.73 (Table 4-1), demonstrating a higher level of yellow reflected light

than blue. At the established alpha level of 0.05, the p-value of the b* color variable

was 0.0033 yielding a significant difference. After means separation, it was elucidated

that all of the treatments were different than the control with respect to this color

variable (Table 4-9), leading us to reject our null hypothesis that there are no

differences between the physical attributes of the chia product samples vs. the control

cookie. The r2 value for this color value was 0.21, although positive it indicates a weak









correlation between the b* color value and the various treatments of the seed in the

cookie product. Visually, the control and whole seed chia cookie product were very

similar. Both samples had a pale yellow hue, and browned edges. The whole seed

cookie was speckled with black and white chia seeds. The combination and ground

chia seed cookie products were visually darker than the aforementioned samples due to

the ground seed component. The ground outer seed coats caused the darker color of

both samples, with their only contrast being the unground seeds of the whole seed chia

cookie.

4.2.2.3 Textural analysis

The final test the chia cookie product was subjected to was the three-point break

test. The three-point break test was carried out over four trials with four replications,

using the Instron Universal Testing Machine, model 4411. During these trials, the

cookies averaged 5.4-6.5 cm. A third beam, a #6 blade was attached to a static 1.2kg

load cell, with a crosshead speed of 50 mm/min. The Instron measured the maximum

compressive load (N) needed to snap each cookie sample. The collected data was

analyzed using the SAS program; a two-way ANOVA was performed to compare

variation across the treatments. To determine significant differences, a calculated p-

value was employed. The null hypothesis (Ho) for this test was "there were no

significant (p>0.05) difference in the amount of force needed to snap each chia cookie

sample vs. the force needed to snap the control cookie". The alternative hypothesis (HA)

was "there was significant (p<0.05) difference in the amount of force needed to snap

each chia cookie sample vs. the force needed to snap the control cookie". During this

comparison, there were a total of 64 observations. The mean maximum compressive

load of the three point break test was 19.23N. The p-value of this test was calculated to









be 0.003 at the 0.05 alpha level. This p-value indicates a significant difference across

the sample treatments for this test. To determine this difference, Duncan's Multiple

Range test was employed. At the 0.05 alpha level, the 15% ground chia cookie product

was statistically different than all other cookie sample treatments (Table 4-16), leading

us to reject the null hypothesis. This result was surprising due to the moisture loss

percentage results (Appendix B). From this data, the control cookie demonstrated the

highest moisture loss, with the ground seed and combination cookie samples close

behind it. Given these results, it was expected that more force would have been

needed to snap the drier cookie samples, but this was not the observation. The r2 value,

coefficient of determination, for this test was 0.26. This value makes a weak, but

positive correlation between the three-point break test and the various seed treatments

used in the chia cookie formulations.

4.3 Product Chemical Analysis

Following the analysis of the textural characteristics of the chia products, analysis

of the product fatty acid components commenced. The fatty acids that were the focus

of this analysis were alpha-linolenic acid (18:3), followed by linoleic acid (18:2).

Although alpha-linolenic acid is the primary omega-3 fatty acid, it is important to note

the quantity of its omega-6 counterpart because they both compete for the same carbon

elongation pathways. During this analysis, all the fatty acids extracted from the lipid

layer of each chia product was measured paying particular attention to the 18:2 and

18:3 fatty acids. After this was completed, the variation across the muffin and cookie

treatments was assessed to establish if a significant (p<0.05) difference existed

between the treatments. This information was obtained through the use ANOVA. The

null hypothesis (Ho) of this analysis was "there are significant differences across the









chia muffin and cookie treatments in regard to omega-3 and omega-6 fatty acid

content." The alternative hypothesis (HA) was "there are no significant differences

between the chia muffin and cookie treatments in regard to the omega-3 and omega-6

fatty acid content". The alpha-level of the statistical analysis was 0.05. When

necessary means separation was employed using the Duncan's Multiple Range test.

4.3.1 Chia Muffin Product

It must be noted that all the following quantities discussed in this and the

accompanying section are reported in grams per 100 grams of product. The control

muffin was found to contain an average of 12.42% total fat over 3 trials (Table 4-36).

When separated into its fatty acid components there was a greater content of linoleic

acid versus a-linolenic acid, although there was no chia seed addition to the control

muffin. This finding is attributed to the canola oil and butter in the muffin formulation.

The combination (ground/whole) chia seed muffin was measured to have a total fat

percentage of 13.97, with 18:2 and 18:3 fatty acid levels of 1.09 and 0.95, respectively.

The largest total fat percentage was obtained from the ground seed muffin, 14.25%.

The greater content of total fat was in direct proportion with a higher measured level of

alpha-linolenic acid, the highest amongst all muffin treatments, 1.47g. The lowest 18:3

was measured in the whole seed muffin which also contained the lowest level of total fat

amongst samples with chia seeds at 12.85%.These findings were rather surprising

because it was originally hypothesized that the whole seed muffin would contain higher

levels of the 18:3 fatty acid. The whole seed product was perceived in this fashion

because the chia seeds were mechanically uncompromised whereas the ground seeds

were. Initially, it was thought that the grinding of the seed would cause some loss of the

omega-3 character of the seeds. From this chemical analysis, this was not the case.









Generally, when chia seeds come in contact with a polar liquid it releases a

polysaccharide mucilage. It is possible that the ground seeds create a stronger web via

this mucilage formation and ground seed coat, repelling any omega-3 release. On the

other hand, the whole seed mucilage created of equal strength, holds the individual

seeds in a segregated suspension from one another. These and other theories must be

researched further for validation. During the statistical analysis of the muffin product's

fatty acid content, it was found that there was no significant difference across the

treatments for linoleic acid content. This observation is due to a p-value of 0.14, the

coefficient of determination (r) was 0.47, demonstrating a mild correlation between

muffin seed variation and linoleic acid content (Table 4-31). The means separation also

confirmed no significant difference amongst the treatments (Table 4-32). As for a-

linolenic acid, the statistical analysis displayed a significant difference across the seed

treatments. The means amongst the treatments was 0.74g, with a p-value of <0.0001at

an a-level of 0.05. This leads us to fail to reject the null hypothesis. The r2 value for this

test also displayed a strong positive correlation between the seed treatments and the

18:3 fatty acid. According to the Duncan's Multiple Range test, the control and whole

seed muffins were not significantly different from one another, but the ground and

combination seed muffins were different from each other and all other treatments. This

finding was expected due to the measured fatty acid content from earlier testing.

4.3.2 Chia Cookie Product

The chia cookie product was subjected to similar chemical and statistical analysis

as the chia muffin product. As aforementioned fatty acid values are expressed as

"grams per 100 grams" of product. The control cookie was found to have a total fat

content of 21.48%, with a minimal amount 18:2 and 18:3 fatty acids. The detected









quantities were attributed to the butter within the cookie formulation, although not a

dominant fatty acid in butter 18:2 and 18:3 is present in small amounts. The

combination (ground/whole seed) cookie was measured to have a total fat content of

12.21%. After fatty acid extraction this cookie treatment displayed the second lowest

quantity of 18:2 and 18:3 acids (Table 4-37). The lowest 18:2 and 18:3 content was

measured in the whole seed cookie at 0.83 and 0.23, respectively. This cookie

treatment also had 21.34% total fat content. The measurement of such a small fatty acid

content directly mimics that of the chia muffin product. The linoleic acid content across

the chia cookie treatments had a mean of 0.93g, and a p-value of 0.02 at a 0.05

confidence level (Table 4-31). A p-value of this nature is an indication of a significant

difference across the treatments leading us to fail to reject the null hypothesis. The

means of this fatty acid were separated via Duncan's Multiple Range test. This means

separation found the control and whole seed cookies were not significantly different

from one another, but the combination cookie was not significantly different than any

cookie sample (Table 4-34). The alpha-linolenic fatty acid content of the chia cookies

was found to have a mean of 0.63g across all treatments. The p-value for this fatty acid

was <0.0001 at a 0.05 confidence level, demonstrating a significant difference across

the seed treatments. From this observation, we are inclined to fail to reject the null

hypothesis. The r2 value of this test was 0.96, demonstrating a strong correlation

between the seed treatments and the 18:3 fatty acid. According to the multiple range

test, the only two treatments that are not significantly different from one another are the

control and whole seed cookie (Table 4-35). This result was expected due to the lower

fatty acid content of these seed treatments.









Table 4-1. Developed chia muffin and chia cookie products surface L*A*B* color-value
analysis and water activity analysis results at 19-21 C.
Sample Variable r2 Mean p-value
Muffin Aw 0 0.87 0.97
L* 0.48 42.93 <0.0001
a* 0.25 5.49 0.0008
b* 0.43 16.79 <0.0001
Cookie Aw 0.15 0.47 0.025
L* 0.33 58.29 <0.0001
a* 0.05 12.05 0.38
b* 0.21 25.73 0.003

Table 4-2. Developed chia muffin product separation of means measuring the attribute
of water activity (Aw) (19-21 C) when comparing variation across sample
treatments.
Treatment N Mean Duncan Group*
control 15 0.87 A
25% ground 15 0.87 A
combination 15 0.87 A
25% whole 15 0.87 A
*means with same letter are not significantly different.

Table 4-3. Developed chia muffin product separation of means measuring the attribute
of L* surface color-value when comparing product color variation across
sample treatments.
Treatment N Mean Duncan Group*
control 15 45.77 A
25% ground 15 39.63 B
combination 15 39.51 B
25% whole 15 46.8 A
*means with same letter are not significantly different.

Table 4-4. Developed chia muffin product separation of means measuring the attribute
of surface a*color-value when comparing product color variation across
sample treatments.
Treatment N Mean Duncan Group*
Control 15 6.54 A
25% ground 15 4.88 B
Combination 15 5.44 B
25% whole 15 5.10 B
*means with same letter are not significantly different.









Table 4-5. Developed chia muffin product separation of means measuring the attribute
of b* color-value when comparing color variation across sample treatments.
Treatment N Mean Duncan Group*
control 15 19.20 A
25% ground 15 14.31 B
combination 15 15.46 B
25% whole 15 18.20 A
*means with same letter are not significantly different.

Table 4-6. Developed chia cookie product separation of means measuring the attribute
of water activity (Aw) (19-21 C) when comparing variation across sample
treatments.
Treatment N Mean Duncan Group*
control 15 0.50 A
15% ground 15 0.45 B
combination 15 0.45 B
15% whole 15 0.50 A
*means with same letter are not significantly different.

Table 4-7. Developed chia cookie product separation of means for the attribute of L*
color-value when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
control 15 61.18 A
15% ground 15 54.54 B
combination 15 55.89 B
15% whole 15 61.55 A
*means with same letter are not significantly different.

Table 4-8. Developed chia cookie product separation of means measuring the attribute
of a* color-value when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
control 15 2.27 A
15% ground 15 42.3 A
combination 15 2.54 A
15% whole 15 1.11 A
*means with same letter are not significantly different.









Table 4-9. Developed chia cookie product separation of means measuring the attribute
of b* color-value when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
Control 15 28.28 A
15% ground 15 23.99 B
Combination 15 24.84 B
15% whole 15 25.8 B
*means with same letter are not significantly different.

Table 4-10. Developed chia muffin product compression test analysis and chia cookie
product triple beam break test analysis results.
Sample Variable r2 Mean p-value
Muffin Hardness 0.17 1 0.007
Springiness 0.14 5.42 0.02
Cohesiveness 0.01 2.53 0.84
Gumminess 0.07 2.36 0.17
Chewiness 0.10 12.84 0.04
Cookie break-test 0.26 19.23 0.0003

Table 4-11. Developed chia muffin product separation of means, measuring the physical
attribute of hardness when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
control 16 1.06 A
25% ground 16 1.05 A
combination 16 0.61 B
25% whole 16 1.30 A

*means with same letter are not significantly different.

Table 4-12.Developed chia muffin product means of separation measuring the attribute
of springiness when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
control 16 5.69 AB
25% ground 16 4.80 B
combination 16 4.77 B
25% whole 16 6.43 A
*means with same letter are not significantly different.









Table 4-13.Developed chia muffin product separation of means measuring the attribute
of cohesiveness when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
control 16 2.41 A
25% ground 16 2.54 A
combination 16 2.23 A
25% whole 16 2.93 A
*means with same letter are not significantly different.

Table 4-14. Developed chia muffin product separation of means measuring the attribute
of gumminess when comparing variation across sample treatments.


Treatment N Mean Duncan Group*
control 16 2.27 AB
25% ground 16 2.34 AB
combination 16 1.32 B
25% whole 16 3.51 A
*means with same letter are not significantly different.

Table 4-15. Developed chia muffin product separation of means measuring the attribute
of chewiness when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
control 16 13.3 AB
25% ground 16 11.97 AB
combination 16 6.45 B
25% whole 16 19.64 A
*means with same letter are not significantly different.

Table 4-16. Developed chia cookie product separation of means measuring the cookie
"break" characteristic when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
control 16 16.33 B
15% ground 16 27.36 A
combination 16 19.68 B
15% whole 16 13.55 B
*means with same letter are not significantly different.


e












Table 4-17. Developed chia muffin product sensory evaluation panel preference test
results measuring six qualitative attributes.
Attribute p-value Treatment Mean* Standard dev.
Control 6.41 1.51
25% ground 6.39 1.39
Overall accept. 0.42 40% ground 6.16 1.69
Control 6.81 1.19
25% ground 6.45 1.45
Appearance 0 40% ground 6.07 1.50
Control 6.47 1.69
25% ground 6.41 1.60
Flavor 0.26 40% ground 6.08 1.83
Control 6.00 1.74
25% ground 6.13 1.61
Texture 0.47 40% ground 6.29 1.70
Control 5.21 1.68
25% ground 5.12 1.46
FACT 0.36 40% ground 4.91 1.68
Control 145
25% ground 146
Rank 0.44 40% ground 159
*rank "means" are panelists rank totals; lower value equals most preferred

Table 4-18.Developed chia muffin product separation of means measuring the attribute
of overall acceptability when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 6.41 A
2-25% ground 75 6.39 A
3-40% ground 75 6.16 A
*means with same letter are not significantly different.

Table 4-19. Developed chia muffin product separation of means measuring the attribute
of appearance when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 6.81 A
2-25% ground 75 6.45 AB
3-40% ground 75 6.07 B 1
*means with same letter are not significantly different.









Table 4-20. Developed chia muffin product means of separation measuring the attribute
of flavor when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 6.47 A
2-25% ground 75 6.41 A
3-40% ground 75 6.08 A
*means with same letter are not significantly different.

Table 4-21. Developed chia muffin product separation of means measuring the attribute
of texture when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 6.00 A
2-25% ground 75 6.13 A
3-40% ground 75 6.29 A
*means with same letter are not significantly different.

Table 4-22. Developed chia muffin product separation of means measuring collected
FACT scale data when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 5.21 A
2-25% ground 75 5.12 A
3-40% ground 75 4.91 A
*means with same letter are not significantly different.

Table 4-23. Developed chia muffin product separation of means measuring compiled
ranked totals when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 145 A
2-25% ground 75 146 A
3-40% ground 75 159 A
*means with same letter are not significantly different.











Table 4-24. Developed chia cookie product sensory evaluation panel preference test
results measuring six qualitative attributes.
Attribute p-value Treatment Mean* Standard dev.
control 6.29 1.73
5% ground/10%whole 6.21 1.57
Overall accept. 0 15% ground/15%whole 5.45 1.74
control 7.00 1.44
5% ground/10%whole 5.67 1.56
Appearance 0 15% ground/15%whole 6.00 1.83
control 5.96 2.02
5% ground/10%whole 5.41 1.62
Flavor 0.03 15% ground/15%whole 5.96 1.85
control 6.92 1.50
5% ground/10%whole 5.52 1.71
Texture 0 15% ground/15%whole 6.45 1.81
control 4.96 1.83
5% ground/10%whole 4.92 1.56
FACT 0 15% ground/15%whole 4.35 1.83
control 122
5% ground/10%whole 148
Rank 0 15% ground/15%whole 180
*rank "means" are panelists rank totals; lower value equals most preferred

Table 4-25. Developed chia cookie product separation of means measuring the attribute
of overall acceptance when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 6.29 A 3
2-5% ground/10%whole 75 6.21 A 3
3-15% ground/15%whole 75 5.45 B
*means with same letter are not significantly different.

Table 4-26. Developed chia cookie product separation of means measuring the attribute
of appearance when comparing variation across sample treatments.


Tukey
Treatment N Mean Group* Sig. diff. than sample

1-Control 75 7.00 A 23
2-5% ground/10%whole 75 5.67 B
3-15% ground/15%whole 75 6.00 B
*means with same letter are not significantly different.









Table 4-27. Developed chia cookie product separation of means measuring the attribute
of flavor when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 5.96 A
2-5% ground/10%whole 75 5.41 A
3-15% ground/15%whole 75 5.96 A
*means with same letter are not significantly different.

Table 4-28. Developed chia cookie product separation of means measuring the attribute
of texture when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 6.92 A 3
2-5% ground/10%whole 75 5.52 A 3
3-15% ground/15%whole 75 6.45 B
*means with same letter are not significantly different.

Table 4-29. Developed chia cookie product separation of means measuring FACT scale
collected data when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 4.96 A 3
2-5% ground/10%whole 75 4.92 A 3
3-15% ground/15%whole 75 4.35 B
*means with same letter are not significantly different.


Table 4-30. Developed chia cookie product separation of means measuring compiled
ranked totals when comparing variation across sample treatments.
Tukey
Treatment N Mean Group* Sig. diff. than sample
1-Control 75 122 B
2-5% ground/10%whole 75 148 B
3-15% ground/15%whole 75 180 A 1 2
*means with same letter are not significantly different.









Table 4-31. Developed chia
Treatment
control
25%ground
25% whole
10%whole/15% ground


*expressed as grams/100 grams


muffin product measured fatty acid content results.
18:2* 18:3* 20:1* total fat%
0.83 0.18 0.08 12.42
1.22 1.47 0.02 14.25
0.92 0.32 0.06 12.85
1.09 0.95 0.02 13.97


Table 4-32. Developed chia cookie product measured fatty acid content results.
Treatment 18:2* 18:3* 20:1* total fat%
control 0.79 0.11 0.04 21.48
15%ground 1.14 1.35 0.08 22.56
15% whole 0.83 0.23 0.05 21.34
10%whole/5% ground 0.96 0.83 0.05 21.12

Table 4-33. Developed chia muffin and chia cookie product measured omega-3 and
omega-6 fatty acid analysis statistical outcomes.
Sample Fatty acid R2 Mean p-value
Muffin 18:2 0.47 1.01 0.14
18:3 0.94 0.74 <0.0001
Cookie 18:2 0.67 0.93 0.02
18:3 0.95 0.63 <0.0001

Table 4-34. Developed chia muffin product separation of means measuring linoleic
(18:2) fatty acid content when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
Control 3 0.83 A
25% ground 3 1.21 A
combination 3 1.09 A
25% whole 3 0.91 A
*means with same letter are not significantly different.

Table 4-35. Developed chia muffin product separation of means measuring a-linolenic
(18:3) fatty acid content when comparing variation across sample treatments.
Treatment N Mean Duncan Group*
Control 3 0.22 C
25% ground 3 1.46 A
combination 3 0.94 B
25% whole 3 0.32 C
*means with same letter are not significantly different.









Table 4-36. Developed chia cookie product separation of means measuring linoleic
(18:2) fatty acid content when comparing variation across sample treatments
means.
Treatment N Mean Duncan Group*
Control 3 0.78 B
15% ground 3 1.14 A
combination 3 0.95 AB
15% whole 3 0.82 B
*means with same letter are not significantly different.
Table 4-37. Developed chia cookie product separation of means measuring a-linolenic
(18:3) fatty acid content when comparing variation across sample treatments
means.
Treatment N Mean Duncan Group*
Control 3 0.11 C
15% ground 3 1.34 A
combination 3 0.83 B
15% whole 3 0.23 C
*means with same letter are not significantly different.


Figure 4-1. Developed chia muffin product control sample compression test graphical
outcome, trial 1, reps 1-3.


0
U Replicate #
Ir I / -- I



-1
0 1 3 4 5 6 7
Time (sec)











































Figure 4-2. Developed chia muffin product control sample compression test graphical
outcome, trial 1, rep 4.


2.0




3 / 3
AReplicate #



0.0 "



-1 .0 i: i i i i i i i i i i i i i
0 1 2 3 4 5 6 7 8 9 10
Time (sec)




Figure 4-3. Developed ground seed chia muffin product compression test graphical
outcome, trial 1, reps 1-3.





97


-, 3


0

,> Replicate #
o 4

E
C



-1. --------- i------- i----------------
0 1 2 3 4 5 6 7
Time (sec)










































Figure 4-4. Developed ground seed chia muffin product compression test graphical
outcome, trial 1, rep 4.


Replicate #
--- 1
------- 2
3


0 1


3 4 5 6 7
Time (sec)


Figure 4-5. Developed whole seed chia muffin product compression test graphical
outcome, trial 1, reps 1-3


03






0 2 3 4 5 7 9 10
Time (Replicat









0 1 2 3 4 5 6 7 8 9 0
Time (sec)










































Figure 4-6. Developed whole seed chia muffin product compression test graphical

outcome, trial 1, rep 4.


Figure 4-7. Developed combination chia seed muffin product compression test graphical

outcome, trial 1, reps 1-3.


z3

0


m 4
o /
E





0
0 1 2 3 4 5 6 7 8 9 10
Time (sec)


z
"-" 1.6

1.4
a 1.2

1.0
a. 0.8 :
E
o 0.6
UI
0.4

0.2 T'
0.0
0 1 2 3 4 5 6 7 8 9 10
Time (sec)


_________ 1









































Figure 4-8. Developed combination chia seed muffin product compression test graphical
outcome, trial 1, rep 4.


^ ^ ^ y J ..-- --. -.- ^ -- ^ -- --, -







--------------I-----------------------i------ --- ---I-------------I--------------------i------




Compressi..e extension (mm)




Figure 4-9. Developed chia seed cookie product control sample break test graphical
outcome, trial 1, reps 1-3.





100


2.0


-' 1.6
1.8
1.6
Z
- 1.4
S1.2

1.0 Replicate #
-0 /1
L0.
E 0.6
U0
0.4

0.2
0.0Tl A
0 1 2 3 4 5 6 7
Time (sec)

































3 pr 5
Compressive extension rmnn)


Figure 4-10. Developed chia seed cookie product control sample break test graphical
outcome, trial 1, rep 4.


Figure 4-11. Developed ground
trial 1, reps 1-3.


seed chia cookie product break test graphical outcome,


101


Replicate #


3 4 5
Compressive extension (mn-)


6 7






















I,t 4 I


0 2 3 4 5 6 7




Figure 4-12. Developed ground seed chia cookie product break test graphical outcome,
trial 1, rep 4.


Compressive extension (mni)



Figure 4-13. Developed whole seed cookie product break-test graphical outcome, trial
1, reps 1-3.


102


n~pllcatl
3

























IR p ---a t I


Figure 4-14. Developed whole seed chia cookie product break-test graphical outcome,
trial 1, rep 4.


Figure 4-15. Developed combination chia seed cookie product break-test graphical
outcome trial 1, reps 1-3.


103


Compressive extension (mrn)


Replicate


6 7


3 4 5
Conpre55ive extension Emm3
































3 4 5
Co.npressive extension mmr)


Figure 4-16. Developed combination chia seed cookie product break-test graphical
outcome, trial 1, rep 4.


104









CHAPTER 5
CONCLUSIONS

The link between diet and health is a connection that remains at the forefront of

the minds of consumers across the nation. American consumers understand that

including the right foods in ones diet can prevent the onset of many diseases such as

cancer and heart disease. As the baby-boomer generation comes of age, a more

knowledgeable and discerning consumer now searches for healthier alternatives. This

search has created a great demand for functional foods of all varieties. The consuming

public demands products with added health benefits, but not at the cost of flavor and

taste. Prior research has even shown that consumers do not mind spending more for a

product with proven health benefits. The key to obtaining a share of the functional food

market is by providing a tasty product that can be easily amalgamated into ones

traditional everyday diet. Obtaining this attribute is key because consumers will only eat

foods they feel comfortable with and this element translates into repeat buying.

From this research, we introduced two new vehicles that can be easily introduced

in the consumer diet and fill the need of lacking omega-3s. These products call upon

the omega-3 content of the Salvia hispanica Lamiaceae, also known as chia or Spanish

sage. Once a staple in ancient Meso-American society, chia has been found to harbor

an oil with the highest content of a-linolenic acid, found in a plant source. Through

carbon chain elongation, a-linolenic acid can be transformed into eicosapentanoic and

docosahexanoic fatty acids, respectively. The chosen vehicles for consumer dietary

incorporation were muffin and cookie food products. From taste panel testing, minor

formulation adjustments are still needed to optimize flavor and texture, but the final

product outlook is promising. Physical attribute testing of the two chia products


105









confirmed that the enhanced baked goods had minimal significant differences from

control muffin and cookie samples. Chemical analysis of both the cookie and muffin

products confirms that the essential fatty acid character of the chia seeds is retained

during baking. The highest quantity of omega-3 retention was measured in the both

ground seed cookie and muffin products. A formulation cost analysis demonstrated that

production of the enhanced baked products is only a few cents more than the non-

enhanced control products (Appendix C).

Further study of the developed products is still needed to address some persisting

issues. The first of these issues is shelf-stability, although product water activity was

assessed, packaged shelf-stability was not. a-linolenic acid being confined in the lipid

layer of the chia seed can oxidize as all other lipids do which can lead to off aromas and

flavors within the products. Proper packaging and storage temperatures to retard

oxidation are other aspects that have yet to be explored. A final unresolved issue was

the mechanism that causes whole chia seeds to have lower omega-3 levels than

ground chia seeds after baking. While most likely due to experimental sample

preparation, analytical methods may need to be adjusted due to the extremely hard

nature of the chia seed. Resolving some of these questions would be the next step in

the commercialization of a chia product with consumer health benefits.


106









APPENDIX A

MUFFIN AND COOKIE INGREDIENT MANUFACTURERS

allspice- McCormick Ground Allspice- Hunt Valley, MD

almond extract- McCormick Pure Almond Extract- Hunt Valley, MD

baking powder- Rumford Baking Double Acting Baking Powder- Terre Haute, IN

baking soda- Arm & Hammer Pure Baking Soda- Princeton, NJ

brown sugar-Domino Light Brown Sugar- Yonkers, NY

butter- Wholesome Farm Unsalted Sweet Cream Butter- Houston, TX

chia- Greens Chia- Vero Beach, FL

cinnamon- McCormick Ground Cinnamon- Hunt Valley, MD

eggs- Publix Eggstirs- Lakeland, FL

flour- Pillsbury's Best All Purpose Flour- Orville, OH

milk- Parmalat UHT Whole Milk- Wallington, NJ

nutmeg- McCormick Ground Nutmeg- Hunt Valley, MD

raisins- Sunmaid Golden Raisins- Kingsburg, CA

salt- Publix Salt- Lakeland, FL

vanilla extract- Publix Pure Vanilla Extract- Lakeland, FL

vegetable oil- Crisco Pure Canola Oil-Orville, OH

white sugar- Publix Extra Fine Granulated Sugar- Lakeland, FL


107










APPENDIX B
MOISTURE LOSS PERCENTAGE


Table B-1. Calculated Chia muffin product
standard
Sample trial 1 trial 2 trial 3 trial 4 trial 5 mean dev.
Control 8.73 9.30 7.60 7.96 8.57 8.43 0.67
25% Whole 7.53 8.73 7.16 7.56 8.21 7.84 0.62
25% Ground 8.25 8.40 9.17 9.01 9.80 8.93 0.63
10% Whole/15%
Ground 8.34 9.28 10.44 10.85 10.04 9.79 1.00

Table B-2. Calculated Chia cookie product
standard
Sample trial 1 trial 2 trial 3 trial 4 trial 5 mean dev.
Control 10.29 9.56 9.01 8.08 9.14 9.22 0.81
15% Whole 9.04 7.22 7.61 7.01 8.32 7.84 0.84
15% Ground 8.80 8.62 9.12 9.69 8.79 9.00 0.42
5% Whole/10%Ground 8.41 10.15 8.34 8.90 9.08 8.98 0.73


108









APPENDIX C
FORMULATION COST ANALYSIS


Table C-1. Chia muffin
Sample cost per muffin cost per 6 muffin batch
Control 0.34 2.01
25% Whole 0.56 3.39
25% Ground 0.56 3.36
10% Whole/15% Ground 0.56 3.36

Table C-2. Chia cookie
Sample cost per cookie cost per 9 cookie batch
control 0.07 0.61
15% Whole 0.12 1.08
15% Ground 0.12 1.07
5% Whole/10%Ground 0.12 1.07


109









LIST OF REFERENCES


2007 Census of Agriculture United States Summary and State data vol.1 part 51

American Association of Cereal Chemists. 1983. Approved methods of the American
Association of Cereal Chemists. St. Paul, Minn., USA: AACC. method 62-05.

American Oil Chemists' Society. 1989. In Official methods and recommended practices
of the American Oil Chemists' Society. Champaign, Ill: The Society. method Ce 1h.

Anonymous. 2009. Oilseed Farming in the U.S.: 11112: Ibis World 36.

AOAC International. 2000. Latimer GW, Horwitz W. (eds.) Official Methods of AOAC
International. Arlington, Va., USA: AOAC International. p 20-23.

Asp EH. 1999. Factors affecting food decisions made by individual consumers. Food
Policy 24: 287-294.

ASTM, Committee E-18. 1979. ASTM Manual on Consumer Sensory Evaluation, ASTM
Special Technical Publication 682, E.E. Schaefer, ed American Society for Testing
and Materials, Philidelphia, Pa. p 28-30.

Ayeza R. 1995. Oil content and fatty acid composition of chia (Salvia hispanica L.) from
five northwestern locations in Argentina. J. Amer. Oil Chemists' Society 72: 1079-
1081.

Ayera R, Coates W. 2004. Composition of Chia (Salvia hispanica) grown in six tropical
and subtropical ecosystems of South America. Trop Sci 44:131-135.

Bagozzi, RP. 1983. A holistic methodology for modeling consumer response to
innovation. Operations Research 31: 128-176.

Bech-Larsen T, Grunert KG, Poulsen JB. 2001April. The acceptance of functional foods
in Denmark, Finland, and the United States. A study of consumers' conjoint
evaluations of the qualities of functional food and perceptions of general health
factors and cultural values: The Aarhus School of Business 25. Available from:
Aarhus University, Denmark.

Berdan F, Anawalt PR. (eds.) 1996. Codex Mendoza. Berkeley: University of California
Press Ltd. (Originally composed ca. 1541-1542).

Bleiel J. 2010. Functional foods from the perspective of the consumer: How to make it a
Success? Intl Dairy J. 20: 303-306.

Bower JA, Saadat MA, Whitten C. 2003.Effect of liking, information and consumer
characteristics on purchase intention and willingness to pay more for a fat spread
with a proven health benefit. Food Quality and Pref. 14: 65-74.


110









Brenna, JT. 2002. Efficiency of conversion of a-linolenic acid to long chain n-3 fatty
acids in men. Current Opinion in Clinical Nutrition and Metabolic Care. 5:127-132.

Burdge GC, Wooten SA. 2002. Conversion of a-linolenic acid to eicosapentanoic,
docosapentanoic, and docosahexanoic acid in young men. Brit J Nutr 88: 411-
420.

Burdge GC, Jones AE, Wooten SA. 2002. Eicosapentanoic and docosapentaenoic
acids are the principal products of a-linolenic acid metabolism n young men. Brit J
Nutr 88: 355-363.

Cahill JP. 2003. Ethnobotany of chia, Salvia hispanica L. (Lamiaceae). Economic
Botany 57 (4) 604-618.

Cahill JP, Provance MC. 2002. Genetics of qualitative traits in domesticated chia (Salvia
hispanica L.). J. Hered. 93:52-55.

Constance J.2008. Food Ingredient Trends Addressing Specific Diseases and other
health conditions: Packaged Facts. Rockville Md. 112.

Craig R, Sons M. 2004. Application for approval of whole chia (Salvia hispanica L.) seed
and ground whole chia as novel food ingredients. Advisory committee for novel
foods and processes. Ireland: Company David Armstrong. p 1-29.

Cross N. 2006. Muffins. In: Hui Y. Handbook of Food Science, Technology, and
Engineering Vol.1. Boca Raton, Fl: CRC Press. p16-2 -16-8.

de Caterina R, Cybulsky MA, Clinton SK, Gimbrone Jr. MA, Libby P. 1995. Omega-3
fatty acids and endothelial leukocyte adhesion molecules. Prostaglandins Leukot
Essent Fatty Acids 52: 191-195.

de Caterina R, Basta G. 2001. n-3 Fatty acids and the inflammatory response-biological
background. European Heart J Supplements 3, Suppl D: D42-D49.

de la Cruz, Martinus. 1940. Badianus manuscript. Translated by John Badianus. (Codex
Barberini Latin 241). Baltimore: The John Hopkins Press. (Originally composed
1552).

de Lorgeril M, Renaud S, Mamelle N, Salen P, Martin JL, Monjaud I, Guidollet J,
Touboul P, Delaye J. 1994. Mediterranean alpha-linolenic acid rich diet in
secondary prevention of coronary heart disease. Lancet 334: 1454-1459.

Diplock AT, Aggett PJ, Ashwell M, Bornet F, Fern EB, Roberfroid MB. 1999. Scientific
concepts of functional foods in Europe: Consensus document. Brit J Nutr 81:1, sl-
27.

Epling CC. 1940. A Revision of Salvia, Subgenus Calosphace. Verlag des
Repertoriums. Berlin: Fabeckstr p 49.


111









Finley J, Shahidi F. 2001. The chemistry, processing, and health benefits of highly
unsaturated fatty acids: An overview. In: Finley J, Shahidi F. Omega-3 fatty acids-
Chemistry, Nutrition, and Health effects. Washington, DC: Oxford University Press.
p 2-9.

Fishbein M, Ajzen I. 1975.Belief, attitude, intention and behavior: an introduction to
theory and research. Reading, MA: Addison-Wesley

Fuller GW. 2005. New Food Product Development. Boca Raton: CRC Press.p188-189.

Food Marketing Institute. 2009. U.S. food shopper trends.

Gaines CS. 1991. Instrumental measurement of hardness of cookies and crackers.
Cereal Foods World. 36: 989,991-994, 996.

Garg ML, Keelan M, Thomson ABR, Clandin MT. 1992. Desaturation of linolenic acid in
the small bowel is increased by short -term fasting and by dietary content of
linolenic acid. Biochemica Biophysica Acta 1126:17-25.

Goldstein LB. 2007 Prevention and management of stroke. In: Libby P, Bonow RO,
Mann DL, Zipes DP. Braunwald's Heart Disease: A Textbook of Cardiovascular
Medicine. 8th ed. Philadelphia, Pa: Saunders Elsevier. p 2.

Gray L. 2009. Omega Fatty Acids: Trends in the Worldwide Food and Beverage
Markets, 2nd ed: Packaged Facts. Rockville Md. 186.

Grigelmo-Miguel N, Boladeras EC, Martin-Belloso O. 1999. Development of high-fruit-
dietary-fibre muffins. Eur Food Res Technol. 210: 123-128.

Gebaur SK, Psota TL, Harris WS, Kris-Etherton PM. 2006. N-3 fatty acid dietary
recommendation and food sources to achieve essentiality and cardiovascular
benefits. Am J Nutr 83:1526-35s.

Guilford JP. 1954. Psychometric method. New York, NY: McGraw-Hill. p 120-129.

Haque MS, Ghoshal KK. 1980. Karyotypes and chromosome morphology in the genus
Salvia L. Cytologia 45: 627-640.

Harris JM. 1997. The impact of food product characteristics on consumer purchasing
behavior: The case of frankfurters. J Food Distribution Research 28:92-97.

Haumann F. 1991. U.S. efforts focus on a variety of new crops. International News on
Fats, Oils and Related Materials 2: 685-690.

HealthFocus. 2009. International trend study. Healthfocus Intl., St. Petersburg, Fl.
www.healthfocus.com.


112









Ixtaina VY, Nolasco SM, Tomas MC. 2008. Physical properties of chia (Salvia Hispanica
L.) seeds. Industrial Crops and products 28: 286-293.

Jones LV, Thurstone LL. 1955. The psychophysics of semantics: an experimental
investigation. J Applied Psych 39: 31-36.

Kreiter T. 2005. Seed of wellness: Return of a supergrain. Saturday Evening Post
40:106-107

Lawless HT, Heymann H. 1998. Sensory evaluation of food; principles and practices.
New York, NY. Chapman and Hall. p 450-455.

Libby P, Ridker PM, Maseri A. 2002. Inflammation and atherosclerosis. Circulation
105:1135-1143.

Lin KY, Daniel JR, Whistler RL. 1994. Structure of chia seed polysaccharide exudates.
Carbohydrate Polymers 23:13-18.

Lloyd-Jones D, Adams RJ, Brown TM, Carnethon M, Dai S, DeSimone G, Ferguson TB,
Ford E, Furie K, Gillespie C, Go A, Greenlund K, Haase N, Hailpern S, Ho PM,
Howard V, Kissela B, Kittner S, Lackland D, Lisabeth L, Marelli A, McDermott MM,
Meigs J, Mozaffarian D, Mussolino M, Nichol G, Roger V, Rosamond W, Sacco R,
Sorlie P, Stafford R,Thom T, Wasserthiel-Smoller S, Wong ND, Wylie-Rosett.
2010. Heart Disease and Stroke Statistics-2010 Update A Report From the
American Heart Association. Circulation. Dallas, TX: American Heart Association.
DOI: 10.1161/CIRCULATIONAHA.109.192667.170.

Mintel. 2009a. Functional foods-U.S. Aug.

Mozaffarian, D. 2005. Does alpha-linolenic acid intake reduce the risk of coronary heart
disease? A review of the evidence. Alternative Therapies (3)11: 24-30.

NawarW.1996. Lipids. In: Fennema O. Food Chemistry. 3rd ed. Boca Raton, Fl: CRC
Press. p 227-228.

Nielsen. 2010. U.S. healthy eating trends part 1: commitment trumps the economic
pinch. Press release, Jan 26. Nielsen Co., Schaumburg, Ill. www.nielsen.com.

Niva M. 2006. Can we predict who adopts health-promoting foods? Users of functional
foods in Finland. Scandinavian Journal of Food and Nutrition, 50(1):13-24.

Nelson DL, Cox MM. 2005. Lehninger Principles of Biochemistry 4th ed. New York, NY.
WH Freeman and Company. p.343-344.

Nutrition Business Journal. 2010. U.S. functional sales slow, but category outpaces
overall food sector in '09. Nutr Bus J. Feb 1. www.nutritionbusinessjournal.com.


113









O' Connor EL, White KM. 2010. Willingness to trial functional foods and vitamin
supplements: The role of attitudes, subjective norms, and dread of risks. Food
Quality and Pref. 21: 75-81.

Ott RL, Longnecker MT. 2004. A first course in statistical methods. Belmont, Ca.
Thomson Brooks/Cole. p 200-201.

Packaged Facts. 2008. Nutritional supplements in the U.S. Nov. Packaged Facts, New
York, N.Y. www.packagedfacts.com.

Patch CS, Williams PG, Tapsell LC. 2005. Attitudes and intentions toward purchasing
novel foods enriched with omega-3 fatty acids. J Nutr Educ Behav.37:235-241.

Peiretti PG, Gai F. 2009. Fatty acid and nutritive quality of chia (Salvia hispanica L.)
seeds and plant during growth. Animal Feed Sci and Tech 148:267-275.

Peryam DR. 1989. Reflections. In sensory evaluation. In celebration of our beginnings.
Philadelphia, Pa. American Society for Testing and Materials. p 31-30.

Peryam DR, Girardot NF. 1952. Advanced taste-test method. Food Eng 24(July):58-
61,194.

Resurreccion AV. 1998. Consumer sensory testing for product development.
Gaithersburg, Ga. Aspen Publishers pl-6.

Reyes-Caudillo, Tecante A, Valdivia-Lopez MA. 2008. Dietary fiber content and
antioxidant activity of phenolic compounds present in Mexican chia (Salvia
hispanica L.) seeds. Food Chem 107: 656-663.

Robinett S.2008.Easy sugar cookies. Seattle,WA: All recipes.Available from
allrecipes.com/Recipe/Easy-Sugar-Cookies/Detail.aspx. Accessed March 4, 2009.

Saura-Calixto F, Gohi I. 2006. Antioxidant capacity of the Spanish Mediterranean diet.
Food Chem 94: 442-447.

Schutz HG. 1965. A food action rating scale for measuring food acceptance. J Food Sci
30: 365-74.

Siddhuraju P, Mohan PS, Becker K. 2002. Studies on the antioxidant activity of Indian
Laburnum ICassia fistula L.): A preliminary assessment of crude extracts from
stem bark, leaves, flowers and fruit pulp. Food Chem. 79: 61-67.

Silfverstolpe G, Johnson P, Samsice G, Svanborg A, Gustafson. 1981. Effects induced
by two different estrogens on serum individual phospholipids and serum lecithin
fatty acid composition. Hormone & Metabolic Research 13: 141-145.

Sloan AE. 1998. Food industry forecast: consumer trends to 2020 and beyond.
FoodTechnology 52(1): 37-44.


114









Sprecher H. 2000. Metabolism of highly unsaturated n-3 and n-6 fatty acids. Biochimica
et Biophysica Acta. 1486:219-231.

Tecklenburg E. 2009 Food Flavors and Ingredients Outlook 2009: Packaged Facts
Rockville Md.120.

Thompson RA, Isaacs GW. 1967. Porosity determination of grains and seeds with air
comparison pycnometer. Trans. ASAE 10: 693-696.

Tudoran A, Olsen SO, Dopico DC. 2009. The effect of health benefit information on
consumers' health value, attitudes and intentions. Appetite 52: 568-579.

Urala N, Lahteenmaki L. 2003. Strength of health-related claims and their perceived
advantage. Int J Food Science and Technology 38: 218-826.

Urala N, Lahteenmaki L. 2007. Consumers' changing attitudes toward functional foods
Food Quality and Pref. 18: 1-12.

Watkins TR, Bierenbaum ML. 2001. Dietary polyunsaturated fat in cardiovascular
disease: Boon or bane? In: Finley J, Shahidi F. Omega-3 fatty acids- Chemistry,
Nutrition, and Health effects. Washington, DC: Oxford University Press. p 66-76.

Vazquez-Ovando A, Rosado-Rubio G, Chel-Guerrero L, Betancur-Ancona D. 2009.
Physicochemical properties of a fibrous fraction from chia (Salvia hispanica L.).
LWT-Food Science and Technology 42: 168-177.

Verbeke W. 2006. Functional foods: Consumer willingness to compromise on taste for
health? Food Quality and Pref. 17: 126-131.

Vilche C, Gely M, Santalla E. 2003. Physical properties of quinoa seeds. Biosyst. Eng.
86: 59-65.

Viles-Gonzalez, Fuster V, Badminon JJ.2004. Atherothrombosis: a widespread disease
with unpredictable and life-threatening consequences. European Heart Journal.
(14) 25:1197-1207.


115









BIOGRAPHICAL SKETCH

Devin Lewis was born and raised in Houston, Texas. At a very young age he was

intrigued by the sights and sounds of his mother and grandmother toiling in the kitchen.

This led him to pursue a Bachelor of Science in culinary arts from Johnson & Wales

University, Miami. FL. After graduating Summa Cum Laude in May 2002, Devin began

honing his culinary skills while working with several notable chefs. While working in the

kitchen, Devin became interested in chemical aspects of food and product development.

Devin further explored his new found interest by attending a Research Chef's

Association Convention. Shortly thereafter, Devin decided to pursue his master's

degree at the University of Florida, majoring in food science. Graduating in August

2010, Devin hopes to pursue a career in food chemistry and food product research and

development.


116





PAGE 1

1 THE INCORPORATION OF CHIA ( SALVIA HISPANICA LAMIAC E AE) SEED S INTO BAKED FOOD PRODUCTS By DEVIN C. LEWIS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FO R THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Devin C. Lewis

PAGE 3

3 To my mother, who has been my emotional and spi ritual support against all odds

PAGE 4

4 ACKNOWLE DGMENTS I would like to thank my advisor, Dr. Renee GoodrichSchn e ider for granting me the opportunity to pursue my masters degree and for all the invaluable guidance throughout this experience. I am thoroughly appreciative for the expertise and assistance given to be my committee members, Dr. Charles Sims and Dr. Al W ysocki. I would also like to thank my lab mates, Lemne Delva, Yael Spektor, and Brianna Mahoney for all their help and suggestions. A special thank you to the FSHN taste panel staff for their assistance during the long and numerous taste panel sessions. Most of all, I thank my family and close friends whose support and understanding gave me the strength to succeed through accomplishing my goals.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 7 LIST OF FIGURES ........................................................................................................ 10 ABSTRACT ................................................................................................................... 12 CHAPTER 1 INTRODUCTION .................................................................................................... 14 2 LITERATURE REVIEW .......................................................................................... 19 2.1 Botanical and Physical Characteristics ........................................................... 19 2.2 Historical Perspective ..................................................................................... 21 2.3 Physiochemical Properties ............................................................................. 22 2.4 Seed Physiology and Lipid Synthesis ............................................................. 24 2.5 Fatty Acid Nomenclature ................................................................................ 26 2.6 AlphaLinolenic Acid Metabolism in Humans ................................................. 27 2.6.1 Fatty Acid Conversion ......................................................................... 27 2.6.2 Reactive protein i nteractions ............................................................... 29 2.7 Consumer Demand for and Attitudes toward Functional Foods ..................... 30 2.8 U.S. Oilseed Farming Industry Analysis ......................................................... 38 2.9 Consumer Sensory Testing Overview ............................................................ 43 2.9.1 Hedonic S cale ..................................................................................... 44 2.9.2 F A C T S cale ..................................................................................... 46 2.10 Objectives ...................................................................................................... 47 3 METHODS AND MATERIALS ................................................................................ 50 3.1 Product Formulation ....................................................................................... 50 3.2 Taste Panel Product Assessment .................................................................. 53 3.3 Chia Product Physical Analysis ...................................................................... 54 3.4 Chia Product Fatty Acid Analysis ................................................................... 56 4 RESULTS AND DISCUSSION ............................................................................... 64 4.1 Consumer Taste Panel Testing ...................................................................... 64 4.1.1 Chia S eed M uffin ................................................................................. 64 4.1.2 Chia S ee d C ookie ................................................................................ 67 4.2 Product Physical Analysis .............................................................................. 72 4. 2.1 Chia Muffin Analysis ............................................................................ 72 4.2.1.1 Water activity ........................................................................... 73

PAGE 6

6 4. 2.1.2 Color analysis .......................................................................... 74 4.2.1.3 Textural analysis ...................................................................... 76 4.2.2 Chia Cookie Analysis ............................................................................ 78 4.2.2.1 Water activity ........................................................................... 79 4.2.2.2 Color analysis .......................................................................... 80 4.2.2.3 Textural analysis ...................................................................... 82 4.3 Product Chemical Analysis ............................................................................. 83 4.3.1 Chia Muffin Product ............................................................................. 84 4.3.2 Chia Cookie Product ............................................................................ 85 5 CONCLUSIONS ................................................................................................... 105 APPENDIX A MUFFIN AND COOKIE INGREDIENT MANUFACTUERS ................................... 107 B MOISTURE LOSS PERCENT AGE ....................................................................... 108 C FORMULATION COST ANALYSIS ...................................................................... 109 LIST OF REFERENCES ............................................................................................. 110 BIOGRAPHIC AL SKETCH .......................................................................................... 116

PAGE 7

7 LIST OF TABLES Table page 3 2 Food Action Rating Scale ................................................................................... 61 3 3 Evaluated c hia muffin product formulations ........................................................ 63 3 4 Evaluated chia cookie product formulations ....................................................... 63 4 1 Developed chia muffin and chia cookie product s surface L*A*B* color value analysis and water activity analysis results at 1921 C. .................................... 87 4 2 Developed chia muffin product separation of means measuring the attribute of water activity (Aw) (1921C) when comparing variation across sample treatments. .......................................................................................................... 87 4 3 Developed chia muffin product separation of means measuring the attribute of L* surface color value when comparing product color variation across sample treatments. ............................................................................................. 87 4 4 Developed chia muffin product separation of means measuring the attribute of surface a*color value when comparing product color variation across sample tr eatments. ............................................................................................. 87 4 5 Developed chia muffin product separation of means measuring the attribute of b* color value when comparing color variation across sample treatments. .... 88 4 6 Developed chia cookie product separation of means measuring the attribute of water activity (Aw) (1921C) when comparing variation across sample treatments. .......................................................................................................... 88 4 7 Developed chia cookie product separation of means for the attribute of L* color value when comparing variation across sample treatments. ..................... 88 4 8 Developed chia cookie product separation of means measuring the attribute of a* color value when comparing variation across sample treatments. ............. 88 4 9 Developed chia cookie product separation of means measuring the attribute of b* color value when comparing variation across sample treatments. ........... 89 4 10 Developed chia muffin product compression test analysis and chia cookie product triple beam break test analysis results. .................................................. 89 4 11 Developed chia muffin product separation of means, measuring the physical attribute of hardness when comparing variation across sample treatments. ...... 89

PAGE 8

8 4 12 Developed chia muffin product means of separation measuring the attribute of springiness when comparing variation across sample treatments. ................ 89 4 13 Developed chia muffin pr oduct separation of means measuring the attribute of cohesiveness when comparing variation across sample treatments. ............. 90 4 14 Developed chia muffin product separation of means measuring the attribute of gumminess when comparing variation across sample treatments. ................ 90 4 15 Developed chia muffin product separation of means measuring the attribute of chewiness when comparing variation acros s sample treatments. ................. 90 4 16 Developed chia cookie product separation of means measuring the cookie break characteristic when comparing variation across sample treatments. ..... 90 4 17 Developed chia muffin product sensory evaluation panel preference test results measuring six qualitative attributes. ........................................................ 91 4 18 Developed chia muffin product separation of means measuring the attribute of overall acceptability when comparing variation across sample treatments. .... 91 4 19 Developed chia muffin product separation of means measuring the attribute of appearance when comparing variation across sample treatments. ................ 91 4 20 Developed chia muffin product means of separation measuring the attribute of flavor when comparing variati on across sample treatments. ......................... 92 4 21 Developed chia muffin product separation of means measuring the attribute of texture when comparing variation across sample treatments. ....................... 92 4 22 Developed chia muffin product separation of means measuring collected FACT scale data when comparing variation across sample treatments. ............ 92 4 23 De veloped chia muffin product separation of means measuring compiled ranked totals when comparing variation across sample treatments. .................. 92 4 24 Developed chia cookie product sensory evaluation panel preference test results measuring six qualitative attributes. ........................................................ 93 4 25 Developed chia cookie product separation of means measuring the attribute of overall acceptance when comparing variation acr oss sample treatments. .... 93 4 26 Developed chia cookie product separation of means measuring the attribute of appearance when comparing variation across sample treatments. ................ 93 4 27 Developed chia cookie product separation of means measuring the attribute of flavor when comparing variation across sample treatments. ......................... 94

PAGE 9

9 4 28 Developed chia cookie product separation of means measuring the attribute of texture when comparing variation across sample treatments. ........................ 94 4 29 Developed chia cookie product separation of means measuring FACT scale collected data when comparing variation across sample treatments. ................. 94 4 30 Developed chia cookie product separation of means measuring compiled ranked totals when comparing variation across sample treatments. .................. 94 4 31 Developed chia muffin product measured fatty acid content results. .................. 95 4 32 Developed chia cookie product measured fatty acid content results. ................. 95 4 33 Developed chia muffin and chia cookie product measured omega3 and omega6 fatty acid analysis statistical outcomes. ............................................... 95 4 34 Developed chia muffin product separation of means measuring linoleic(18:2) fatty acid content when comparing variation across sample treatments. ............ 95 4 35 D linolenic (18:3) fatty acid content when comparing variation across sample treatments. .. 95 4 36 Developed chia cookie product separatio n of means measuring linoleic (18:2) fatty acid content when comparing variation across sample treatments means. ................................................................................................................ 96 4 37 linolenic (1 8:3) fatty acid content when comparing variation across sample treatments means. ................................................................................................................ 96 B 1 Calculated Chia muffin product ......................................................................... 108 B 2 Calcula ted Chia cookie product ........................................................................ 108 C 1 Chia muffin ....................................................................................................... 109 C 2 Chia cookie ....................................................................................................... 109

PAGE 10

10 LIST OF FIGURES Figure p age 2 1 Alpha linolenic acid structure ............................................................................. 49 2 2 Docosahexanoic acid structure .......................................................................... 49 2 3 Eicosapentanoic acid structure .......................................................................... 49 3 1 Hedonic nine point scale .................................................................................... 61 3 2 Sample ballot used in chi a product taste panel testing ....................................... 62 4 1 Developed chia muffin product control sample compression test graphical outcome, trial 1, reps 1 3. ................................................................................... 96 4 2 Developed chia muffin product control sample compression test graphical outcome, trial 1, rep 4. ........................................................................................ 97 4 3 Developed ground seed chia muffin product compression test graphical outcome, tri al 1, reps 1 3. ................................................................................... 97 4 4 Developed ground seed chia muffin product compression test graphical outcome, trial 1, rep 4. ........................................................................................ 98 4 5 Developed whole seed chia muffin product compression test graphical outcome, trial 1, reps 1 3 .................................................................................... 98 4 6 Developed whole seed chia muffin product compression test graphical outcome, trial 1, rep 4. ........................................................................................ 99 4 7 Developed combination chia seed muffin product compression test graphical outcome, trial 1, reps 1 3. ................................................................................... 99 4 8 Developed combination chia s eed muffin product compression test graphical outcome, trial 1, rep 4. ...................................................................................... 100 4 9 Developed chia seed cookie product control sample break test graphical outcome, trial 1, reps 1 3. ................................................................................. 100 4 10 Developed chia seed cookie product control sample break test graphical outcome, trial 1, rep 4. ...................................................................................... 101 4 11 Developed ground seed chia cookie product break test graphical outcome, trial 1, reps 13. ................................................................................................. 101 4 12 Developed ground seed chia cookie product break test graphical outcome, trial 1, rep 4. ...................................................................................................... 102

PAGE 11

11 4 13 Developed whole seed cookie product break test graphical outcome, trial 1, reps 13. .......................................................................................................... 102 4 14 Developed whole seed chia cookie product break test graphical outcome, t rial 1, rep 4. ...................................................................................................... 103 4 15 Developed combination chia seed cookie product break test graphical outcome trial 1, reps 13. .................................................................................. 103 4 16 Develo ped combination chia seed cookie product break test graphical outcome, trial 1, rep 4. ...................................................................................... 104

PAGE 12

12 Abst ract of Thesis Presented to the Graduate School of the University of Fl orida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE INCORPORATION OF CHIA ( SALVIA HISPANICA LAMIACEAE) SEED S INTO BAKED FOOD PRODUCTS By Devin C. Lewis August 2010 Chair: Ren e Goodrich Schneider Major : Food Science and Human Nutrition Chia is an herbaceous summer annual which is botanically known as Salvia hispanica Lamiaceae. The seed of this annual is harvested in the fall and has been consumed in Southern Mexico and Central America for generations. Once a staple food of ancient Mesoamerican civilizations, chia has currently emerged as a high source of polyunsaturated fatty acids. The oil of the chia seed has an omega3 fatty acid content of up to 68% versus the 57% found in flax seed. The fatty a cid contained in linolenic acid (ALA) is metabolized in the body to produce eicosapentanoic and docosahexanoic acids, respectively, and has been found to reduce the risks of coronary heart disease in hu mans. The objectives of this study were 1) To demonstrate s imilar consumer acceptance between control products and those incorporated with chia seeds, 2) To investigate which form of chia seed withstands processing and baking best, defined as retaining the highest amount of the chia seeds omega3 character. Consu mer acceptance data was obtained through the use of hedonic testing at UF/IFAS FSHN taste panel. On separate occasions consumers were asked to rate and rank a cookie or muffin sample, one control and two treated samples (25% and 40%

PAGE 13

13 ground chia muffin; 15% ground/15% whole seed and 5% ground/10%whole chia cookie) Panelist used a ninepoint hedonic scale to rate the following attributes: overall acceptance, texture, flavor, and appearance. The hedonic scale was anchored with 1=dislike extremely and 9=l ike extremely The ninepoint Food Action rating scale (FACT) was used to determine the consumption attitudes of the panelist. The anchors for the FACT scale were 1=I would eat this only if forced and 9=I would eat this at every opportunity Analys is of variance (ANOVA) was used to analyze panelist data, with means separation using Tukeys HSD test. There was no significant (p< 0.05) difference between the control and the 25% chia muffin in overall acceptability. The same result was found for the c ookie control and the 5% ground/10%whole chia cookie sample. Subsequent fatty acid analysis was performed on the muffin and cookie samples, but including the chia in whole, ground, and combination forms at the aforementioned chia levels. It was found thalinolenic acid was obtained in both ground seed cookie and muffin samples. The muffin and cookie samples were also subjected to water activity, L*a*b* color, and textural analysis, to measure the chia product physical properties versu s those of the control samples. Significant differences between the muffin treatments and the control muffin where found for the appearance attribute, L and b color values, hardness, springiness, and cohesiveness textural attributes. The control cook ie and the treatment cookies had significant differences for all attributes. It was found that the ground muffin and ground cookie treatments retained the most omega3 character of the tested treatments. It is the position of this study that the chia see d is a viable ingredient in food products to be consumed by the public.

PAGE 14

14 CHAPTER 1 INTRODUCTION In todays society, con sumers are far more informed with respect to the foods they consume than past generations. This increase in knowledge proportionately increases the demand for food products that are more nutritious, but remain tasty and satisfying as well. When venturing to the health food aisle of the average grocer consumers are bombarded by the large variety of foods that are labeled as high in and a good source of Although these products undergo some governmental scrutiny, it is still left up to the consumer to identity those products that can ultimately provide the sought after minerals and nutrients in the highest quantities. Consumers ar e more aware of food and dietary issues and are monitoring and adjusting what they consume as they have become more proactive and diligent in improving their overall health through their daily diet ( Constance 2008). One market that many of these products s trive to gain a shar e of is that of the health and wellness market. According to the Natural Marketing Institute, retail sales of health and wellness consumer packaged goods reached $102.8 billion in the U.S. in 2007, demonstrating a growth of 15% over the previous year (T eklenburg 2009). This market tends to be stronger due to the rise in coronary heart disease and other diseases witnessed in the United States. Coronary heart disease (CHD) is the leading cause of heart attacks and stroke in this countr y. This disease is defined as the narrowing of the small blood vessels that supply blood and oxygen to the heart. CHD is commonly caused by a condition known as atherosclerosis. Atherosclerosis occurs when fatty material and plaque build up on the walls of the arteries. This accumulation

PAGE 15

15 causes the arteries to narrow, which not only limits, but can completely halt the flow of blood to the heart (Goldstein 2007). As of 2006, more than 17 million people over the age of 20 suffer from this disease and it c auses at least 500,000 deaths annually It is estimated that approximately every 25 seconds, an American will suffer a coronary event and approximately every minute, someone will die of one ( Lloyd Jones and others 2009) Some of the risk factors involved with CHD are diabetes, high blood pressure, elevated levels of low density lipoprotein concentrations and triacylglycerols in blood pl asma. To combat these undesirable conditions health officials urge lifestyle changes that emphasize a transformation in t he amount and types of dietary fats consumed. It h as been well documented that to decrease these risk factors, a shift from the consumption of saturated to polyunsaturated fatty acids is necessary (Watkins and Bierenbaum 2001). Throug h research, omega3 fatty acid consumption has emerged as one of the key components in the fight to lower the aforementioned risk factors. This fatty acid is metabolized in the body to produce eicosapentanoic and docosahexanoic acids respectively Typically, the source of ome ga 3 fatty acids has been relegated to high oil fish, flax seed endosperm and its oil fraction ( Finley and Shahidi 2001). H igh oil fish produce substantial amounts of the fatty acid, but are also accompanied by an unwanted and many times intolerable fishy aroma. Although high oil fish are the best source of these fatty acids, the use of their oils in food products has been slow due to the unstable nature of these oils. This instability produces off flavors which are the result of degraded products from hydroperoxide formed from the highly unsaturated lipids in the fish bodies. The oxidation of the fatty acids can produce off flavors that

PAGE 16

16 range from mild to offensive. This oxidative degradation is the product of the degree of unsaturation that composes the chemical matrix of omega3 fatty acids (Gebauer and others 2006) Recently, there have been many attempts to increase the stability of the aforementioned oils. It was found that oxidation can be limited by interesterification to create borage oil, or partially hydrolyzing the oil and combining it with emulsifiers. Although these methods produced positive results, it is not certain if they can be used in food products and gain consumer acceptability after this type of manipulation (Gebauer and others 2006). Flax seed is not associated with the aforementioned aroma, but tends not to contain the highest fatty acid content possible from a plant source. The highest content of omega 3 fatty acids from a plant source can be obtained from chia, also known as Salvia hispani ca Lamiaceae. The oil expressed from chia has been measured at 68% w/w linolenic acid (ALA) as opposed to the 57% w/w found in flaxseed (Ayerza 1995). linolenic acid is a polyunsaturated fatty acid that is metabolized through the sequential 6 desaturases. These desaturases elongate the original c arbon chain to create eicosapentanoic (EPA) and docosahexanoic (DHA) fatty acids. Both fatty acids are essential, and recommended daily dosages are 5001800 mg per day or 1300 mg of ALA. EPA and DHA are also beneficial to the function of cell membranes, t he central nervous and immune systems. ALA has also been found to aid in the production of various eicosanoids such as leukotrienes and prostaglandins. These eicosanoids can be derived from omega3 and omega6 fatty acids act as signaling molecules. They exert complex control over the central nervous system, inflammation and immunity in the human body (DeCaterina and Basta 2001). Due to higher levels of

PAGE 17

17 the estrogen hormone, women have been found to metabolize these fatty acids more efficiently than thei r male counterparts ( Burdge and Wooten 2002) Another health benefit that chia provides is derived from its total dietary fiber content. On a dry weight basis chia seeds contain 3741g total dietary fiber per 100g (Reyes Caudillo and others 2008). In accordance with the American Dietary Association guideline of a 3:1 ratio of insoluble to soluble dietary fiber intake for adults; the seed of S hispanica yields 32.834.9 grams insoluble and 6.166.84 grams of soluble fiber per 100 grams. Some of the rep orted effects of both types of dietary fiber are the reduction of cholesterolemia, modification of glycemic and insulinaemic responses, and more efficient intestine function (Reyes Caudillo and others 2008) The main component of the insoluble dietary fib er fraction is lignin, which makes up 3941% of the total dietary fiber. Lignin is a complex organic polymer, derived from wood and is an integral part of plant wall secondary structure. It has been directly linked to the aforementioned hypocholesterolemic effects through its tendency to absorb bile acids (Reyes Caudillo and others 2008). The fiber rich fraction of chia seeds has also been observed to have a high antioxidant activity (488.8 mol Trolox equivalents (TE)/g), which is similar to high antioxidant beverages such as tea (631 mol TE/100ml) and freezedried coffee (450 mol TE/g). This high antioxidant activity is attribut ed to the presence of querc etin caffeic, and chlorogenic acids. Given the above, it has been purported that as little as 7g rams of chia fiber rich fraction may contribute to the trapping of free radicals, meeting total dietary antioxidant level (3549 mol TE/day) in a Mediterranean diet (Sauro Calixto and Go i 2006).

PAGE 18

18 The focus of this study is to demonstrate similar consumer acceptance between control products and those incorporated with chia seeds, and to investigate which form of chia seed withstands processing and baking best, which is defined as retaining the highest amount of the chia seeds omega3 character. To attain t hese objectives, consumer acceptance data was obtained through the use of hedonic testing at UF/IFAS FSHN taste panel. On separate occasions consumers were asked to rate and rank a cookie or muffin sample, one control and two treated samples (25% and 40% g round chia muffin; 15%ground/15% whole seed and 5% ground/10%whole chia cookie) Panelist used a ninepoint hedonic scale to rate the following attributes: overall acceptance, texture, flavor, and appearance. The hedonic scale was anchored with 1= dislike extremely and 9=like extremely The nine point Food Action rating scale (FACT) was used to determine the consumption attitudes of the panelist. A nchors for the FACT scale were 1=I would eat this only if forced and 9=I would eat this at every opp ortunity. The muffin and cookie samples were also subjected to water activity, L*a*b* color, and textural analysis, to compare the chia product physical properties versus those of the control samples. It is hypothesized that the products incorporated wi th chia seeds will find acceptance amongst consumers and the physical attributes of said products will not be significantly different (p>0.05) than control samples.

PAGE 19

19 CHAPTER 2 LITERATURE REVIEW 2.1 Botanical and Physical Characteristics Chia is an herb ac eous summer annual, botanically classified as Salvia hispani ca Lamiaceae. Common names for the seeds of this species are chia, chia sage, and Spanish sage. C losely related to mint t he Salvia genus consists of at least 900 species with in the Lamiaceae family ( Ayerza and Coates 2004). This summer annual produces a scalene ellipsoidal seed, which has a thickness of 0.811.40 mm, average width of 1.31.4 mm, and is approximately 2 mm in length. Its appearance is said to be similar to that of rapeseed or quinoa ( Vilche and others 2003). Seed color typically varies between black and white, with a much smaller percent of white seeds due to the encoding of a single recessive gene. The white seeds also tend to be larger than the black seeds (Cahill and Prov ance 2002). On a dry basis, the average seed moisture is 6.6% to 7.2% for black and white seeds, respectively, which lends itself to the storage stability of the seeds. The bulk density, a ratio of the mass of the seeds to its total volume, was measured between 0.667 and 0.722 g cm3. Bulk density is used to determine the capacity of agricultural storage and transport systems. The true density, which is used in the sourcing of agricultural separation equipment, was measured between 0.931 and 1.075 g cm3 (Ixtaina and others 2008). The resistance to airflow during aeration and drying procedures is determined by the porosity of the mass of seeds. The porosity of the seed is the fraction of the space within the grain not occupied by the grain (Thompson and Issacs 1967). The porosity of chia seeds was found to be 22.935.9%, and continuous, leaving ample space for the flow of oxygen during aeration and carbon dioxide while stored in silos. When tested for its frictional

PAGE 20

20 properties, chia seeds were measure d to have an average angle of repose of 17.10.4. The seed shape and smooth outer surface were responsible for the lower value of repose causing the seeds to slide onto one another. Values of static coefficient of friction were 0.28 0.01 on galvanized sheets and 0.31 0.01 on mild steel sheets, which are lower than values of cumin, sunflower and sesame seeds (Ixtaina and others 2008). The seeds of this light sensitive annual are generally harvested in the fall, during its early vegetative stage which i s immediately prior to the plants shooting period when linolenic content is at its highest (Peiretti and Gai 2009). Although native to southern Mexico and northern Guatemala, chia is grown throughout Central and South America. The environmental conditions best suited for plant maturation and higher fatty acid content are described as arid to semi arid. These ecosystems typically exhibit cooler temperatures (1224C) in conjunction with elevations ranging from 398 to 500 meters. Average rainfall in the arid to semi arid regions was measured at 186187 millimeters during crop growing cycles ( Ayerza and Coates 2004). Although there have been changes in morphology due to human selection, both domesticated and wild type varieties of S. hispanica remain morphologically distinct from related taxa, especially in regards to anther, calyx and seed morphology. This observation led to the classification of chia within the taxonomic section, Potiles, within the Salvia subgenus Calosphace ( Epling 1940). According t o Haque and Ghoshal (1980) indicators such as reproductive isolation, a unique chromosome number (2n=12), the lowest of the genus, and an autogamous (self fertilizing) breeding system, it is suggested that the wildtype variety contributed to domesticated S. hispanica.

PAGE 21

21 These contributions coupled with unintentional humanselected traits over time, produce the domesticated chia variety harvested currently. The humanselected traits for domesticated chia include: apical dominance, increased branching, inc reased seed size, decreased pubescence, seed coat colors and patterns, and closed calyxes. The trait that solidifies the domestication of the S. hispanica is the closed calyx (Cahill 2003). Calyx closure completely eliminates seed dispersal, rendering li mited survival of the species outside of human cultivation. 2.2 Historical Perspective During preColombian times, chia served as a fixture in the culture of several meso American civilizations. After harvesting, the seeds had many uses within daily life i ncluding but not limited to culinary and medicinal purposes. The Aztec capital, Tenochtitlan was reported to receive between 5 and 15 tons of chia seed annually. According to the 16th century Codex Mendoza and Matricula do Los Tributos, 21 of the 38 Aztec provincial states presented the seeds as tribute to Aztec gods (Berdan and Anawalt 1996). Its importance as a staple food item only trailed that of beans and corn, but was consumed more frequently than amaranth. Post harvest chia seeds were roasted and g round into flour known as chiapinolli. This flour was then incorporated into tortillas, tamales, or eaten as gruel. The ground seeds were also used as an ingredient in Chianatoles, an Aztec beverage. Today, ground chia beverages have been replaced by those comprised of whole seeds, water, lemon, and sugar or fruit juices, and called Chia fresca (Cahill 2003). These versatile seeds were not only consumed, but were used for medicinal purposes also. Citations in the Badianus Manuscript refer to infusi ons including whole chia seeds to aid in the uptake of medicines and treatment of respiratory malaise (de la Cruz 1952). Chia was also used to treat eye obstructions and

PAGE 22

22 infections by placing seeds directly under the eyelid. Kidney problems were alleviated through the consumption of a handful of seeds combined with two liters of water. Extracted oils served as a base in skin emollients, lacquer for primitive pottery manufacture, craft and body paints ( Cahill 2003). Mixchiaviticac, a Nahua (indigenous peo ple of Mexico) word, refers to the circles painted on the cheeks of Aztec deities (Hauman 1991) According to Spanish manuscripts; chia was known as the running food due to its consumption by Aztec messengers for endurance. Due to Spanish colonization e fforts and influence, the cultivation of chia seeds in mesoAmerican society declined substantially (Kreiter 2005). 2.3 Physiochemical Properties Throughout history chia has been used in various capacities, but it is not until recent years that its physioc hemical properties have been investigated. Chia seeds contain 250390 g oil/kg of fresh matter. The fatty acids of the oil fraction extracted from the seeds are polyunsaturated, their primary components are linoleic (C18:2n6; 170260 g/kg of total fatty linolenic acid ( C18:3n3; 500570 g/kg of the fatty acid) (Ayerza, 1995). Once introduced into an aqueous substance, chia seeds tend to exude a polysaccharide mucilage that remains tightly bound to the seed. Removal of the mucilage can be per formed using 6N urea at pH 7.4 for 7 hours, which gives a 4.5% dry weight yield of the polysac charide The polysaccharide has been identified as being composed of D xylosyl, D glucosyl and 4O methyl D glucopyrranosyluronic acid in ratios of 2:1:1. These components are arrang ed in a linear tetrasaccharide sequence ( Lin and others 1994) A low content of uronic acid is a sole indicator of no pectin being associated with the polysaccharide mucilage (Reyes Caudillo and others 2008). Defatted chia that remains after oil extraction leaves behind a fiber fraction that is

PAGE 23

23 33.9g/100g and 17g/100g of protein (Craig and Sons 2004). The total dietary fiber content of the seeds is 36.939.9g/100g, while insoluble and of soluble fiber are 32.834.9 grams 6.166.84 grams per 100 grams, respectively. The primary component of the insoluble dietary fraction is lignin, occupying 3941% of the total dietary fiber. Lignin is thought to protect the unsaturated fats in the chia seed by building a strong and resistant structure. This structure is s upported by spaces in the cell walls filled with lignin, cross linking various plant polysaccharides. The presence of cellulose and hemicellulose are verified by the neutral sugars found in the insoluble dietary fiber fraction (Reyes Caudillo and others 2008). The water holding capacity of the defatted fiber fraction was 15.41g/ g fiber. The water holding capacity is the ability of a moist material to retain water when subjected to an external centrifugal gravitational force or compression. This capacity is composed of the linked water, hydrodynamic water and physically trapped water, which contributes the greatest proportion to the capacity. It is suggested that the high water holding capacity is influenced by the polysaccharide mucilages (Vazquez Ovando and others 2009). In contrast, the oil holding capacity of chia seeds has a tendency to be low (2.02g /g sample). It is theorized that the particle size of the fiber fraction is not small enough to hold higher amounts of oil, since smaller particles typically have more contact surfaces. The capacity of water absorption is an indication of a structures ability to absorb water when immersed in water or in contact with a constantly moist surface. The water absorption capacity of the chia fibrous fraction is 11.73g water/g sample. The fibrous fraction of the chia seed also has consistent emulsifying properties. The emulsifying capacity of a molecule demonstrates its ability to facilitate solubilization or dispersion of two immiscible liquids. The emulsi on

PAGE 24

24 activity of the chia fibrous fraction was 53.26 ml/100ml. The stability of the emulsions formed by the seed fiber fraction was 94.84ml/100ml which could possibly be attributed to the chia seed protein fraction because most proteins are strong emulsifier s. These emulsifying properties are also displayed in regard to the absorption of bil e acids, and increasing fecal excretion, which would limit small intestine uptake (Vazquez Ovando and others 2009). The total chia seed phenolic content was in the range of 0.880 0.92110.008 mg/g chia seed extract in GAE (gallic acid equivalents). Flavonols were found to be in the highest concentrations, followed by quercetin, kaempferol, with lower amounts of chlorogenic and caffeic acids detected. The antioxidant acti vity of chia seeds is comparable to Trolox at 220ppm GAE when measuring radical scavenging CLAMS), chia seed extracts demonstrated an ability to stabilize reactive oxygen species responsible for the oxidation of linolenic acid (Reyes Caudillo and others 2008). Singlet oxygen is a precursor of hydrogen peroxide and hydroxyl radical formation (Siddhuraju and others 2002). Extracts of chia seeds not only showed comparable activity to Trolox in the inhibition of peroxidation, but displayed properties of singlet oxygen quenchers. The seed extracts act as antioxidants via the hydrogen donating capacity of their phenolic groups. The activities of these polyphenolics also demonstrate a metal chelating potential that plays a role in the protection against ironand copper induced free radicals (Reyes Caudillo and others 2008). 2.4 Seed Physiology and Lipid Synthesis The seed of the S. hispanica is known for its high polyunsaturated fatty acid con tent associated with its lipid fraction. The sole site of the de novo fatty acid

PAGE 25

25 synthesis within the plant cells is the plastid, more specifically the elaioplast. Generally, the accumulation of total lipids takes on a sigmodial pattern and can be divided into three periods. During the initial period, structural lipids (phospholipi ds and glycolipids) are present but are only a small proportion of the total seed weight Triacylglycerols are collectively absent at this point, and the predominant fatty aci ds at this stage are linolenic. As the secondary period is initiated, more rapid seed growth begins, stimulating triacylglycerol accumulation. Due to this accumulation, a substantial increase in the seeds total dry weight begins and continues for at least a two week period. As this occurs, triacylglycerols become the predominant lipid component as phospholipids and glycolipids quantities decrease. At the end of this stage, triacylglycerols are 90% of the total lipid fraction and overall fatty acid composition. When the growing seed reaches the final period leading to full maturity the dry weight per seed increases very little which is accompanied by a progressive decrease in moisture content. Phosphotidylcholine is tho ught to play a major role in polyunsaturated fatty acid formation. One theory suggests that as the development of the seed cotyledon (part of seed embryo and major storage organ) occurs, phospholipid acyltransferase catalyzes an acyl exchange between oley ol CoA enzymes and fatt y acids at position two of the phosphotidylcholine. This reaction provides polyunsaturated acyl CoAs that work with acyl CoAs derived from the plastid via the glycerol 3 phosphate pathway which can via lipid synthesis be incorporated into triacylglycerols (Styme and Glad 1981). These triacylglycerols accumulate in the seeds cell cytoplasm as lipid droplets, and are spherically shaped during the early and midstages of seed development. Eventually, they become compressed and tightl y packed in the oil rich

PAGE 26

26 cells of ma ture seeds. Neutral lipids are typi cally located within discrete organelles known as lipid or oil bodies. These oil bodies typically have a mean diameter of 1m. During seed development, the size of the oil body remai ns constant as triacylglyceride accumulation is accompanied by an increase in the number of oil bodies in the cell (Rest and Vaughn 1972). The major lipid organ within chia s eeds is the cot y l e don. The cot y l e don contains a small percentage of saturated fat ty acids, but not enough for the reserved oils to be considered fats. Unsaturated fatty acids generally favor position two of the triacylglycerol and constitute more than 33% of the total fatty acid, the spill over congregates at position one (Tre a le se and Doman 1984). 2.5 Fatty Acid Nomenclature The derivatives of hydrocarbons, fatty acids, are carboxylic acids with hydrocarbon chains. These hydrocarbon chains can range from 4 to 36 carbons long (C4C36), although the most common carbon chains are fr om 4 to 24 carbons in length. Some fatty acid structures are unbranched and completely saturated. Saturation in fatty acid nomenclature denotes the absence of double bonds within the structure. Unsaturation, on the other hand, denotes at least one double bond in the hydrocarbon chain. Fatty acids with only one double bond are termed monounsaturated while multiple double bonds are categorized as polyunsaturated fatty acids. The most simplified nomenclature for these carbon chains specifies the chain leng th and number of double bonds, separated by a colon; for example 18carbon stearic acid with no double bonds would be noted as 18:0. When these structures contain double bonds there are two options fo the carbons are counted starting at the carboxyl (COOH) end of the chain. The positions of the double bonds are noted by superscript numbers following delta, ex:

PAGE 27

27 9, 1 2) (Nelson and Cox 2005). The second option for noting double bonds within from the methyl (CH3) end of the structure. This double bond position is noted by the number symbol, ex: 20:4 6 (Nawar 1996). The latter convention is the more commonly used and is the notation used in the following discussion. 2.6 Alpha Linolenic Acid Metabolism in Humans The primary fatty acid derived from the lipid f linolenic acid (ALNA), C18:3 n 3. The adequate intake, and intake associated with a low prevalence of inadequacy, is set for ALNA is 1.6g/ day for men and 1.1g/day for women (Gebauer and others 2006). 2.6.1 Fatty Acid Conversio n Within the human body, ALNA is converted to eicosapentanoic acid (EPA, C20:5 n3) Docosapentaenoic acid (DPA) C22:5 n3 is formed via the addition of C2 to EPA which is t hen converted to C24:5 n3 and C24:6n3 Docosahexanoic acid (DHA) is synthesized from C24:6n3 oxidation, which shortens the carbon chain by C2 A lthough the limiting step in the synthesis of DHA, its overall regulation is still unclear. This synthesis occurs to a lesser extent than that of DPA and EPA (Sprecher 2000). ALNA is considered to be an essential fatty acid since it is not produced by the body. Long chain p olyunsaturated fatty acids (LCPUFA) are important structural components of cell membranes, and their interactions with phospholipids ensure cell functionality. The primary conversion site is the liver followed by enterocytes (Burdge and others 2002).

PAGE 28

28 Afte r ALNA conversion, very low density lipoproteins transport newly synthesized EPA and DPA away from the liver to other parts of the body (Burdge and Wooten 2002). During metabolic conversions of ALNA, omega3 and omega6 fatty acids compete for metabolic enzymes; this competition also exists during esterification into plasma phospholipids and triglycerides (Mozaffarian 2005). It has been noted that through this competitive nature, increases in dietary concentration of C18:2 n6 caused a decrease in the synthesis of n 3 LCPUFA and vice versa. Optimal conversion of ALNA to EPA/DPA is expected when the diet is low in both n6 fatty acids and n3 LCPUFA, particularly C18:2 n6. Simultaneous increases in 18:3 n3 and 18:2 n6 generally decrease18:3 n3 convers ion as well (Brenna 2002). From 13C labeled fatty acid studies, the highest levels of ALNA are found in triacylglycerols (TAG) and the chylomicronenriched fraction (CRF) of the triacylglycerols due to ALNA exportation from enterocytes within chylomicron T AG. Chylomicrons, being the primary transporters that they are, tend to be reused in this cycle. It was also noted that 13C labeled ALNA persisted in the total blood plasma TAG for up to 73 hours, suggesting incorporation into other lipoproteins by the l linolenic acid becomes a part of the nonesterfied fatty acid pool from adipose tissue that provides a short term supply to other tissues (Burdge and others 2002). These activities are also reserved for EPA and DPA since they too can be identified in both total plasma and CRF TAG as desaturation and elongation occur within enterocytes. This action is consistent with reports of desaturase activities in microsomes prepared from human intestinal epithelium (Garg and others 1992). ALNA is then released into the circulation via chylomicrons by lipoprotein lipase activity. Plasma cholesteryl esters act as a long term source of ALNA while in circulation,

PAGE 29

29 whereas EPA, DPA, and DHA (in women) are primarily associated with phosphatidylcholine ( Burdge and Wooten 2002). In healthy men, it has been observed that DPA concentrations increase before EPA concentrations within the total plasma TAG; while there is an opposite effect demonstrated in the CRF TAG. This is possibly due to LCPUFA associated with the CRF TAG being derived from enterocyte metabolism, and total TAG. EPA and DPA are the net products of enterocyte and hepatic (liver action) synthesis which may mask the precursor product relationship between the metabolites. The inhibition of DPA to DHA synthesis in healthy males is oxidation step in DHA synthesis that acts as a point of metabolic control. It is theorized that the downregulation of this synthesis is attributed to the need for DHA being met via diet alone negati ng the need for synthesis (Burdge and others 2002). In direct contrast, for healthy women, there was shown to be an increase in the conversion of DPA from ALNA due to an upregulation of the desaturation/elongation pathway downstream of EPA synthesis by oestrogen (Silverstolpe and others 1981). Variations in metabolic capacity for ALNA desaturation and elongation may be due in part to different oestrogen exposure rather than diet alone. This results in increased conversion of DPA to DHA, consistent wi oxidation steps, which is a proposed locus of pathway control (Sprecher 2000). 2.6.2 Reactive protein interactions An elevation in C reactive proteins (CRP) is strongly associated with clinical definitions of atherot hrombotic disease (Libby and others 2002). Atherothrombosis is defined as atherosclerotic plaque disruption with superimposed thrombosis (blood within blood vessels) formation (Viles Gonzalez and others 2004). The C reactive protein exerts a direct proin flammatory effect on human endothelium. An increase in CRP

PAGE 30

30 levels exhibits synergy with hypercholesterolemia to increase CVD risk in men and women. N 3 polyunsaturated fatty acids suppress proinflammatory cytokine production by peripheral blood cells and inhibits lymphocyte proliferation. These fatty acids lessen inflammatory response that are important to the initiation of atherothrombosis. From linolenic acids, which are derived from chia, demonstrate cardioprotec t ive effects (de Lorgeril and others 1994). Endothelial activation and enhanced expression of cell adhesion molecules are early events leading to atherogenesis (plaque formation within the arteries) (Cybulsky and Gimbrone 1991). DHA metabolism from a high ALNA diet results in a dose dependent inhibition of vascular cell adhesion molecules, E selectin, and intercellular cell adhesion molecule1, to a lesser extent, which have all been shown to have an inverse association with CRP. Patients with higher CRP levels typically have a diminished cholesterol lowering response (29%) when compared to patients with lower CRP. Therefore, higher CRP levels without dietary and other interventions can impose overall cardiovascular risk (Zhao and others 2004). 2.7 Consumer Demand for and Attitudes toward Functional Foods The purchasing public is the most important segment of the U.S. food system. Their role as fooddecision makers has a large impact in the success or failure of todays marketed food products (Sloan 1994a). Whether marketed food products are taken home for consumption or left on the grocers shelf hinges on the food preferences of the individual consumer. Food preferences play an important role in food selection because they provide an indication of the amount of satisfaction an individual anticipates from eating a food. Preferences are a result of physiological and psychological development and social experiences, and are related to the degree of

PAGE 31

31 liking of a food. Liked foods are those that are familiar, considered pleasant, and are typically consumed, yielding food preference as a predecessor of consumption (Asp 1999). Attitudes are defined as mental states, learned predispositions, psychological tendencies, or evaluated j udgments about objects which guide behavior toward said objects (Tudoran and others 2009). Intentions, on the other hand, represent a willful state of choice where one makes a self implicated statement as to a future course of action (Baggozi 1983). Ac cording to Fishbein and Ajzen (1975), intention is the most immediate determinant of behavior and implicitly, the most direct predictor of engaging in a behavior. It is these three ideals, food preference, consumer attitude, and purchase intention that must be engaged, to render a food product that is repeatedly purchased by consumers. By appealing to consumer food preferences, gaining positive feedback/ attitudes toward the food product during the initial product trial, purchase intentions can be directed toward repeat buying. This mindset is the driving force behind functional food marketing. One of the largest trends that has gathered and sustained momentum in the U.S. is the increasing awareness of the role of diet and proper nutrition to maintain health and prevent disease. To this end, todays consumers are seeking foods that can provide added health benefits, thereby creating a market for functional foods. Functional foods are food items that have been purposefully designed to provide added benefi ts beyond normal nutrition value in such a way that it improves health (Urala and Lhteenmki 2003) and reduces disease (Diplock and others 1999). Functional foods are not pills, but remain food and are part of a normal food pattern (Diplock and others 1999). The idea of the health effects associated with functional foods is based on a

PAGE 32

32 single product and its delivery of functional components (Urala and Lhteenmki 2007). To take full advantage of this market opportunity three conditions must be met: 1) T here must be a consumer need or problem that requires a solution. 2) There must be an awareness that the consumer or people have a problem. 3) Consumers must be willing to spend money to solve the problem or satisfy the need they have identified. All thre e of these elements are in place in regard to functional foods. The basic selling proposition for functional foods is that they promote health in a convenient way (Beck Larsen and others 2001). It has been shown that even during times of financial cris is; consumers are willing to spend for functional benefits. To meet consumer need, pure ingredient selling is no longer sufficient, todays food companies must sell valueadded solutions that are relevant to consumers (Bleiel 2010). According to Urala and Lhteenmki (2007), the use of functional food provides a new, convenient tool for improving health, yielding a pleasure not only based on improved health, but improved mood and performance with the possibility of preventing disease. Consumers percei ve the fortification as a bonus or stimulus, which provides hedonic expectations, and can further justify their purchasing decision (Tudoran and others 2009). It has been found that consumer acceptance of functional foods is not unconditional, the buying public is not ready to compromise taste for health, the purported benefits do not allow any tradeoffs for flavor. Some of the unwanted flavor characteristics that arise in functional foods are bitter, acid, astringent and salty off flavors (Verbeke 2006) 60% of consumers consider taste more important than nutritional information, at least in most foods (Harris 1997). Other key purchase intention factors include convenience, quality, and price/value. Even with the health benefits associated with functi onal foods there are still

PAGE 33

33 some consumers that are skeptical of these products. Previous research reported that nonusers of functional foods cite a lack of consumer knowledge, low perceived importance or interest in this food category, and price as a reas on to not use the products. Some consumers tend to have suspicions of the efficacy of the functional food promise and its possible side effects. Trust in healthrelated information plays an essential role in functional food choices, the most crucial fact or affecting consumer acceptance (Niva 2006). Attitudes, lifestyle factors, demographic factors i.e. age, gender and education, strongly affect the acceptability or intention to use functional a food. Females, 25+ years, were consistently found to have a higher perceived concern for health regarding food over men in this same age range and younger, consequently responding more aggressively to proven health benefits. This finding could be attributed to females purchasing food for others within a family set ting. The purchase intent for consumers 65+years was lower due to a reticence to try new products (Bower and others 2003). To appeal to those demographic groups that have lower purchase intentions, functional foods must approach a status of traditionall y healthy foods, and fit into an individuals normal eating pattern to be effective (Patch and other 2005). According to Bower and others (2003), a higher purchase intent was demonstrated when product sensory aspects were liked more. The increased lik ing of the food product also caused a significant increase in the price consumers were willing to pay for said products despite a higher price point due to health benefits. In 2009, the top four health issues consumers were extremely concerned about wer e: retention of mental shapness 65%, heart disease62%, cancer 61%, and

PAGE 34

34 maintaining ability to continue with normal activity while aging 59% (Health Focus 2009). As these health issues concern consumers, they in turn create a demand for functional food pr oducts. Anticipating this demand, food companies clamor for a share in this market. In 2009, nearly 46% of food shoppers said they were very concerned about nutrition, which is up 5% over 2008. The functional food segment also outpaced the overall food i ndustry growth rate of 1.6%. This observation is due impart to the functional food bread/grains categorys 3%growth and a 2% growth of functional snack foods (NBJ 2010). According to Mintel (2008), 6 in 10 adults bought a functional food or beverage, whic h is up from 2007s 48%. Within the U.S., food and beverage sales in this segment reached $37.4 billion in 2009, showing 2.7% growth over the prior year (NBJ 2010). During this period, fiber, omega3 fatty acids, vitamins, calcium, and antioxidants were the top five ingredients consumers sought after (Mintel 2009a). By years end, 2009, the dollar sales for food and beverages touting an omega3 claim rose 42% (Nielsen 2010). O mega 3 e nriched p roduct m arket a nalysis: I n todays society it has become commonplace to find multiple products touting health benefits via the addition of omega3 fatty acids lining the average grocers shelves. While price and taste are the most important drivers of food and beverage purchase decisions, 62% of American consume rs considered healthfulness to have great or some impact on their choices according to the 2008 IFIC (International Food Information Council) Food and Health Survey. This figure is down from 65% in 2007, but still shows despite a sluggish economy, food choices are seen by consumers as important to underlying good health (Tecklenburg 2009). As consumers become more aware of the health benefits derived

PAGE 35

35 from fatty acids, many products that include an omega3 ingredient will begin to find life and longevity amongst shoppers. Omega3 enhanced products entered the retail market in 2003, but it was not until 2006 that such products reached mainstream U.S. supermarkets. A large number of the first entries into this category came from small marketers with limited distribution, i.e. health/natural food and specialty stores. Due to these conditions, sales were not often tracked by instore scanner data. Substantial amounts of these products were also sold over the internet and through mail order, two avenues also not tracked by monitoring services, therefore total sales of omegaenriched products are estimates from various trade magazines, information from food distributors and retail grocery chains. The global retail market for omega3 enhanced foods and beverages is estimated at $4.6 billion in 2007, a 33.5% increase over 2006 estimates of $3.4 billion (Gray 2009). Packaged Foods projects that the global retail market of omega3 enhanced foods will approach $8.2 billion by 2012. This projection reflects a CAGR (Compound Annual Growth Rate) of 31.7% between 2003 and 2012. Two factors support the growth of this market: innovations in formulation products using DHA and EPA, which allows more omega3 product enhancement, and a growing consumer awareness of the ben efits of consuming this fatty acid. The omega3 market is typically dominated by grainbased products, holding 74% of the market in 2007, but with innovations in formulation techniques other categories are increasing their shares in the market. The number of milk, nondairy and yogurt beverage product introductions increased in 2006 from 76 SKU (stock keeping units) to 120 SKUs in 2007. Yogurt introductions rose from 20 SKUs in 2006 to 68 in 2007. Snack bars, which have been a mainstay since 2003, increased from 57 SKUs to 72 SKUs introduced in 2007. It is

PAGE 36

36 projected that in the next two to three years that several percent shares will shift as more beverage, dairy, and oil/butters/margarine/cooking sprays/ spreads continue to gain market share (Gray 2009) According to a 2008, IFIC Food and Health Survey, 72% of respondents had heard of omega3 fatty acids, which is up from 63% in 2006. The Health and Wellness Trends Report, published by the Natural Marketing Institute, noted that many consumers feel as if they are deficient in omega3s more than any other nutrient including fiber and calcium. Consumer familiarity with foods that can provide benefits beyond basic nutrition, or functional foods is at an all time high, with 92% of Americans being able to name a food and its health benefits (Constance 2008). This familiarity is demonstrated in the purchasing tendencies of consumers, as well. In 2008, due to the dramatic rise in fuel prices, consumers tended to focus on a single shopping venue that offered the greatest range of the products they needed. As for the purchasing of better for you food products, such as those enriched with omega3s, convenience and variety were key factors. A large assortment of these products was initially found in health/ natural food stores, but the issue of convenience began to overshadow other considerations. In 2007, according to Packaged Facts, traditional super markets accounted for 49% of all omegaenhanced foods and beverages, followed by health/natural food stores 31%, mass merchandiser and specialty stores 6% each, club stores 5%, and convenience, dollar, drugstores and other 3%. Data Monitors Product Scan noted in 2005, 648 SKUs with high omega3 or high omega tags or claims introduced into global markets with 346 in the U.S. In 2007, this figure increased to 1,156 global SKUs and 520 in the U.S. With market

PAGE 37

37 introductions such as these, omega3s have a great potential to become a ubiquitous component of all fortified foods as calcium and vitamin D are added to many foods. On the contrary, the omega enriched sector has been plagued by product withdraw a ls and misinformation on the part of consumers. It was found that some consumers lacked trust in the efficacy of the enhanced food products, and the health c laims were perceived as hype rather than being scientifically established. Some products were also believed to have low dose levels. Recommended dose levels of EPA/ DHA are 400mg 1000mg per day for adults. Conservative per serving recommendations for mos t foods and beverages incorporated with omega3s is 65250 mg on the lower end for products consumed frequently by children and the higher end for adult consumption. Formulations with less than 65 mg lead to negligible effects (Gray 2009). Initially, the omega 3 enriched food and beverage industry was dominated by small, entrepreneurial companies. In recent years, a number of large international corporations have become involved in this sector. By the end of 2008, more than 500 companies worldwide marketed foods that listed omega, DHA, or EPA on the label. Of the 474 products filed globally in 2007, only a few had more than one or two of these products introduced. The leading manufacturers of nonfish omega products in 2008 were Rockin Roll Gourmet (32 SKU), Unilever (30 SKU), Navitas Naturals (15 SKU), Kellogg and Wegmans (both 14 SKU) (Gray 2009). From a survey conducted by Prepared Foods in 2008, it was found that in 2000 it took an average of just under ten months for a new product to reach the market, in 2008 this increased to eleven months. Line extensions introduction time frames increased from five months in 2000 to over six months in 2008. These time frames increases,

PAGE 38

38 according to manufacturers surveyed were attributed to the increasing use of health claims and the emerging science of the ingredients (Gray 2009). 2.8 U.S. Oilseed Farming Industry Analysis Currently, chia production within the United States is nonexistent. This section is meant to provide an overview of the oilseed farming industry within the United States and demonstrate how chia seed cultivation would fit into this agricultural sector. The oilseed industry is comprised of establishments that engage in growing fibrous oilseed producing plants or producing oilseed seeds. This industry is dominated primarily by soybean production, followed by sunflower, canola, and flax seed oils. Mustard and safflower oils are produced as well, but to a much lesser extent than the aforementioned seed varieties. These oilseeds are field cr ops that are primarily used as a source of vegetable oil, and also serve as an important source of livestock feed. Flax, which shows a strong similarity to chia although containing a lower oil percentage, is the fourth largest crop produced by the indust ry, with 6.7% of value and 8.5% of the harvested acreage in 2008. In 2005, following an increase in prices, planted acreage doubled yielding a larger crop and lower prices. Demand has since remained strong due to its wide array of uses. 2009 saw flax ac count for 10% of the harvested acreage and a share of its production value increased 7.9% (Ibis world 2009). Oilseed processors represent the majority of domestic sales at 48.5% and 51.2% of sales volume. In 2006, of the total supply of 2,854 million pounds of canola seed harvested, oil processors produced 8.9 million pounds of oil and 1,400 million pounds of meal, which was used for livestock feed, food and industrial purposes. The second largest market for oilseeds is exportation. In terms of volume, seed export accounted for 32.7% of production in 2006. U.S. oilseed exportation has grown strongly over

PAGE 39

39 current period, but remains a net importer of oilseeds. The value of oilseed exports grew at yearly 13.1% over the five years to 2008; imports on the other hand, grew at a yearly rate of 39.2% which produced a widening industry trade deficit from $1.2 million in 2003 to $7.6 million in 2008. The United States generally exports processed oilseeds, as larger markets for raw seed exists in Mexico and Canada (Ibis W orld 2009). As an agricultural industry, oil seed farming has a low level of ownership concentration when compared to other economic sectors. The average U.S. oilseed farm covers 134 acres, with annual sales of $208,000 and a market share of 0.04% (Ibis World 2009). Concentration levels have begun to rise due to the decrease in farm number by 6.1% per year in the five years to 2009. This decrease is attributed to competition and cost pressures forcing smaller farms to close as the industry m oves to large scale production. An increase in concentration in oilseed farming has been driven by greater mechanization and the shift to more farm specialization. Higher concentrations may lead to greater profitability as fixed costs fall relative to pro duction. Oilseed crops also tend to require less farm labor, so large farms are more profitable than smaller farms. The state of North Dakota has dominated the oilseed industry over the past five years with greater than 50% of the plantings and product ion nationally (Ibis World 2009 ) The oilseed farming industry has seen revenues peak, plunge, and peak again before falling in 2009. The value of oilseed farming peaked in 2008, when oilseed prices were at their strongest. As the demand for food crops f or biofuel producers grew, so did biofuel and feedstock prices. A similar, effect was seen as crude oil prices fell, the value of using oilseeds in alternative fuel production followed suit. The demand for

PAGE 40

40 oilseeds is a function of activity of the veget able oil and fat manufactures. Oil production is an important source of demand for oilseeds, changes in output levels of oil and fat processors will alter the demand for raw inputs. Like many commodities, the demand for oilseeds is sensitive to price cha nges. Large increases in the price of oilseeds can constrain demand as fat and oil processors move to switch production to alternate materials such as animal based fats. Between 2002 and 2007, the average price of oilseeds such as canola, flax, and sunfl ower seeds increased in response to higher demand and smaller harvests. Seed meal is a major component in protein feed used by commercial poultry and pork producers. Increases in livestock production yields a greater demand for oil seed for animal feed is volatile and can fluctuate due to supplies of competing protein feed sources and the availability of natural forage (Ibis World 2009). The factors forming the basis of competition among oilseed growers includes production costs, quality, and range of pr oducts produced. Internal competition within this industry in the U.S. is moderate since growers do not interact strategically. Typically, no single grower is large enough to singly have a significant effect on the market. Producers generally respond to price signals from the market to decide production. Costs of production are a key competitive factor among growers due to the homogeneity of oil seeds, prices per farmer are the same therefore more efficient growth yields higher profits. Oilseeds are di fferentiated by quality standards, grades and assessments are based on protein and oil contents, premium grades can demand higher prices. Demand for particular seed varieties can also drive up prices received by grocers. Recent food safety concerns have enabled producers to charge higher prices

PAGE 41

41 even though quality is difficult to control due to a large number of exogenous factors. Oilseeds are generally grown in combination with other crops, and a higher demand for other crops yields more planting s and less oil seeds causing lower production and less industry revenues. U.S. oilseed growers mainly compete against the soybean industry, which dominates the oilseed sector. It is forecasted that external competition will rise due to the demand for vegetable oils for edible and industrial purposes (Ibis World 2009). Oilseed farming requires special equipment for planting and harvesting. The introduction of labor saving equipment has caused the level of capital to increase over time. This sector of farming is classified as capital intensive and its estimated labor to capital ratio is approximately 1:6:1. Oilseed producers spend $1.60 worth of labor for every $1 worth of depreciation incurred on capital equipment. Land, a nondepreciable asset, has no contribution to the industrys total depreciation costs. Due to the highmechanization of oilseed farming, trends have begun to shift toward replacing onfarm services with off farm services, both factors reducing labor cost requirements. According to the 2007 A gricultural Census, plant and machinery on the average oilseed farm is worth approximately $1,723,863, which is higher than other farming sectors at $88,357 (2007 Census of Agriculture). The oilseed farming industry is a minor part of the U.S. grain growing sector, which tends to be overshadowed by the soybean farming industry. This agriculture sector is expected to generate $716.6 million compared to the $27.5 billion of the soybean industry. This disparity is due to many factors, but one of the primary f actors is the limitation of geographic regions suitable for oilseed production. Over the next five years,

PAGE 42

42 demand for oilseeds is expected to benefit from increased demand for food oils without trans fats. A greater understanding of the effects of trans fats has stimulated demand for oilseed in U.S. and international markets yielding a strong increase in the price of oil with lower trans fat content (Ibis World 2009) As for profitability within this industry, it is estimated to have risen over the current period. Growth in prices received by farmers has exceeded growth in price of farming inputs; but wage growth has been constrained. There has also been some varied profitability over the current period due to price variation, planting, yield and farm inputs. Farm input costs have risen currently, but crop production costs have fallen from their heights seen in mid 2008. The falling of sort commodity (commodities grown and not mined) prices had an equal effect on farm inputs, leaving producers unwilling to pay high fertilizer and farm chemical prices from the soft commodity boom. Although costs have increased at a yearly rate of 4.5% over the five years to 2009, they have fallen about 17% between 2008 and 2009. Strong increases in prices had a negative effect on farm profitability, however in this industry; prices had risen at a greater rate than farm input costs, yielding higher profit levels (Ibis World 2009) After experiencing volatile conditions over the last five years, the oilseed farming industry i s forecasted to experience consistent growth over the upcoming five year period, with a slight dip in 2013. This growth will be due to the demand for seed oils in the retail and industrial markets (Ibis World 2009). Chia seed, with its high oil content a nd various components for industrial usage could gain a share of this market. Chia cultivation would provide an alternate crop to be rotated with other crop cultivation during the year. The primary concern would be locating the proper growing area within

PAGE 43

43 t he U.S. and making chia cultivation profitable for producers. This profitability would be created by the addition of chia meal to livestock feed and chia seed oil use in biofuel and industrial formulations. 2.9 Consumer Sensory Testing Overview Sensory evaluation is a scientific discipline used to evoke, measure, analyze, and interpret reactions to those characteristics of foods and materials as they are perceived by the senses of sight, smell, taste, touch and hearing. Anonymous, Sensory Evaluation Divi sion of the Institute of Food Technologist. Consumer testing is one of the most important activities in product development. The primary purpose of consumer affective tests is to assess the personal response by current and potential customers of a produc t, product ideas, or specific product characteristics. Consumer evaluation concerns itself with testing certain products using untrained people who are or will be the ultimate users of the product (ASTM 1979). Two types of sensory evaluation exist, object ive and subjective evaluation. Objective sensory evaluations are used to garner a sense of some sensory attributes. This type of evaluation includes: analytical sensory tests, expert panel tests, and difference tests. On the other hand there is subjectiv e testing which also may also be called preference testing. During this evaluation style, panelists are presented with a choice of samples and must state which sample is preferred or most accepted. It must be made clear and understood that these two evaluation methods are not interchangeable (Fuller 2005). Sensory tests are typically conducted during product development for product development guidance, to screen products, to identify those products that are disliked and those that match or exceed a speci fic target product for acceptance. The overall goal of sensory evaluation in the

PAGE 44

44 food industry is to enhance quality to improve appearance, flavor and texture as it is perceived by consumers to guide/ influence their food choices (Resurreccion 1998). Befo re sensory testing begins, there must be a clear understanding of what is required from the test. Do the developers want to determine whether they have produced at the best formulation of a product with respect to the criterion? Or do they want to determi ne whether this particular product is as good as, or better than, the competition product. Once these important questions are answered, then the sensory evaluation process truly serves its purpose. 2.9.1 Hedonic scale During taste panel sessions, panelist were asked to use two different ninepoint scales to rate product attributes, the first being the hedonic scale, followed by the FACT scale. The hedonic scale, also known as a degree of liking scale, was developed at the Food Research Division of the Qu artermaster Food and Container Institute of the Armed Forces, Chicago, Ill. The scale introduced the hedonic value concept, which refers to the psychological range of dislike, positioned at the lower end of the scale and like positioned at the upper end o f the scale (Peryam and Girardot 1952). The hedonic scale assumes that consumer preferences exist on a continuum and that preferences can be categorized by responses based on likes and dislikes (Lawless and Heymann 1998). Initially, the scale was to be us ed to predict the food choices of soldiers. The method was designed for people with no food testing experience, hence the descriptions and scale points. When using the scale, panelists are encouraged to report an immediate and nave response without any conscious effort to remember or judge (Peryam and Girardot 1952). The focus of the hedonic scale development was: 1) to detect small differences in the direct response to similar foods. 2) To detect gross

PAGE 45

45 differences in the direct response to foods, even w hen time, subject, and test conditions are allowed to vary. 3) In field questionnaire surveys, were to reveal differences in groupreference attitudes toward foods (Peryam and Girardot 1952). In developing the terms for the categories of the scale there w as concern about which terms had consensual meaning among the population. Great care was taken to remove terms that could be thought of as ambiguous or having double meanings across the population (Lawless and Heymann 1998). An example of such a term is average. In an original study by Jones and Thurstone (1955) test groups equated average with like moderately whereas in todays society the same word may have a negative connotation. The descriptive categories or terms were to have low variability in meaning, no bimodality and little skew. Developers used Thurstones model for categorical judgment as a means of measuring the psychological scale values for words (Lawless and Heymann 1998). The number values on the scale were developed using 51 words or phrases which formed the candidate list, and taken from a pilot study with 900 soldiers chosen to represent all enlisted personnel. Each phrase was presented on a form with a rating scale from 4 to + 4 with a check off format. The subjects were instr ucted to read each phrase and assign an integer value from 4 to +4, with zero as an option. This method presumed that integers were themselves an interval scale of psychological magnitude. Following the Thurstonian model, which did not use raw numbers, this scale was converted to standard deviations as units of measure by way of converting the scale in to z values. The final assessment of the scale was to have eleven categories, but due to equipment restrictions nine categories resulted (Peryam and Girardot 1952). Government paper was only 8 wide and we found that typing

PAGE 46

46 eleven categories horizontally was not possible. So we sacrificed a modicum of precision for a real improvement in efficiency at the moment. Probably, at no great loss David Peryam (Peryam 1989). 2.9.2 FACT scale The second scale used to evaluate the chia seed product was the FACT scale. This scale is a behaviorally oriented approach to scaling food acceptability. The scale is based on attributes and actions, combining statements about frequency of consumption and motivationally related statements, producing an actionbased index of food acceptance. It was reasoned by scale developer, Howard Schutz of Hunt Foods and Industries Inc., Fullerton, Ca, that behaviors might not always matchup with acceptance as scale on the traditional ninepoint hedonic scale (Lawless and Heymann 1998) The FACT scale was developed using twenty subjects to rank 18 statements that reflected an effective action toward food, 1the most positive food attitude t o 18the most negative food attitude. The 18 statements were then scaled using Guilfords composite standard technique (Schutz 1965). Selected from the eighteen statements were nine statements which represented approximately equal scale intervals, approx imate equality was used because it lends more confidence to statistical analysis of the ratings. The nine categories also provided a means to make direct comparisons of the usefulness of the scale to the hedonic scale. This direct comparison was demonstr ated during the FACT scale development process by way of a study using 100 participants to generate means for 54 foods using both scales. The FACT scales means were consistently lower than hedonic means, but demonstrated less skew than the hedonic scale. As for the relationship between mean preference rating and the percent dislike falling below the midpoint of the scale, these values were the same

PAGE 47

47 shape for both scales. From this study, it was also noted that there was a high degree of correlation (r=+. 97) between the two scales. The two scales were found not to be interchangeable, but best used in complement as an overall measure of food acceptance (Schutz 1965). The food action rating scale has been employed in numerous research projects such as: Eff icacy of fruit purees as partial replacements in a chocolate cake and cookie recipe. W. Landis and L. Altman, 1996. J. Amer. Diet. Assoc. 96: a48 and Effects of fat level and cooking methods on physical and sensory characteristics of restructured beef st eaks. M. Penfield and others, 1988. J. Food Quality (11) 5: 349356. 2.10 Objectives This study has several objectives. The first is to heighten the publics awareness of the health benefits of chia through incorporation into food products that hold a place in the normal eating pattern. It has been well documented that chia seeds contain the linolenic acid from a plant source. This fatty acid contributes to the reduction of low density lipoprotein concentrations in humans (Finley and Shahidi 2001) Reductions such as these als o contribute to a smaller risk of coronary heart disease (CHD), one of the leading forms of mortality. Regular consumption has also been shown to lower the c reactive protein in type2 diabetics, which is a risk factor for CHD (Kreiter 2005). In the United States, when the word chia is mentioned, the potted plant product the Chia Pet is typically visualized. This is a preconception that can be beneficial and detrimental in the same instance. The benefit is that the potted plant product will give consum ers a frame of reference so the seed product is not completely foreign. After this association is built, it could possibly give way to consumers actually

PAGE 48

48 sampling any product containing chia seeds. The association may also create a curiosity about the seed in general and engage the consumer to become more informed about chia. On the other hand, the same reference may be constructed, but not by a positive means. This could lead to the consumer not perceiving chia seriously and disregarding the seed and pr oducts formulated from it as well. The next objective of this study is to demonstrate that chia seeds can be utilized in the production of food products that are accepted by consumers. These products will contain chia seeds in various forms (whole, ground, and combination). High consumer acceptance could possibly lead to the commercialization of these products. Although the health benefits of chia are established, it is obvious that these benefits will not be fully realized unless the seed is actually consumed. The United States Food and Drug Administration recognizes chia as a foodstuff; ancient recordings as well as research studies and human testimonials identify the seed as being palatable. After the products are created and have been found to obtain attributes the consuming public finds acceptable, the final objective can then be completed. This objective includes comparing and testing the physical attributes of the created food products to a control equivalent, measuring the quality of the products. These benchmarking exercises will provide some sense of how the created products compare quality wise with similar products that are not incorporated with the seed. During this phase of the study, fatty acid content of the chia seed products wil l also be analyzed to ascertain which form of the chia seeds maintain the highest level of omega3 character after processing and baking.

PAGE 49

49 Figure 21. Alpha linolenic acid structure Figure 2 2. Docosahexanoic acid structure Figure 23. Eicosapentanoic acid structure

PAGE 50

50 CHAPTER 3 METHODS AND MATERIALS 3.1 Product Formulation The seed of the Salvia hispanica L. are known for their high omega3 fatty acid oil and various other health benefits. Although the benefits are many, without a proper vehicle the healthful attributes of the seeds would be lost to consumers. The primary vehicles used in this study to transfer the aforementioned healthful properties were chia muffin and chia cookie product. The formulation for the control muff in product began with a basic muffin recipe. The initial muffin control formulation was prepared numerous times, adjusting ingredient quantities, or removing certain ingredients completely until a control muffin with acceptable flavor, color, and texture was reached (Table 32). These changes included the addition of a raisin puree for its humectant/ moisture retention properties. The addition of brown sugar to enhance the sweetness profile and balance that of the white sugar already present in the formul ation. The raisin puree consisted of golden raisin that were softened in heated water, then blended using a twospeed Waring Commercial blender ( Torrington, CT USA ) model 7011S, in a 2.88:1 raisin to water ratio. Initially, commercially blended appl e pie and cake spice products were used during the control formulation process, but they were removed and replaced by their separate components (cinnamon, allspice, and nutmeg) that were scaled individually for better flavor control. Honey was used as a s weetener, but did not achieve the sought after sweetness profile. The vegetable oil quantity was reduced by half to reduce the greasy feel left by the control muffin. Melted unsalted butter was added to the formulation to enhanc e mouth feel, top browni ng and provide crisp edges around the

PAGE 51

51 muffin products ( Cross 2006). Although the ingredients and their quantities changed during the control muffin formulation, the method of preparation remained consistent. The producers of the ingredients used in the m uffin formulation can be found in Appendix A. The muffin method was used to mix the ingredient components of the control muffin. This method is a classic technique used in quick bread preparation. All the dry ingredients were sifted and combined, being c areful to completely incorporate both leavening agents, baking powder and baking soda. Then the wet ingredients we re combined including the sugars, white/brown, and melted butter. Finally, the wet ingredients were added to the dry ingredients and mixed until all the components were incorporated, without dry flour pockets, but still lumpy. The pans were muffin sprayed lightly with canola pan coating and filled full with the muffin batter using a level #24 scoop. While formulating the muffin control, eac h muffin batch was prepared to yield six muffins and was baked in a WhiteWestinghouse 30 freest anding range (Augusta GA USA ) model WWEF3002KW at 176.6C (350F) for 18 minutes. The oven was preheated at 190.5C (375F), and then lowered to 176. 6C upon introduction of the muffins. The doneness of each batch was checked by piercing each muffin with a toothpick, a dry toothpick indicating that the muffins were done. The muffins were left to cool in the pans for one minute, and then the batch was f lipped out and left to cool on glazing racks. The muffin product was stored overnight in paper towel lined, Tupperware containers. When preparing chia muffins, the seeds were incorporated into the control muffin via the substitution method. According to this method, the chia seeds were substituted for the flour quantity that was removed, keeping all the other ingredient quantities consistent. Before the seeds were combined with the other dry

PAGE 52

52 ingredients, they were ground using a Kitchen Aid Blade Coffee Grinder ( St. Joseph, MI USA ) model BCG 100WGH and finally sifted to remove unground hulls. The seeds were added in various quantities in ground, whole or combination whole and ground forms. It must be noted that the muffin product with whole seeds had no flour removed, only seeds added. The tested muffin treatments were: 25% ground, 40% ground, 25% whole seed, and 10% whole/15% ground muffins (Table 34). The other chia seed product formulated during this study was a chia seed cookie. The original recipe wa s taken from a sugar cookie recipe found online at allrecipes.com. The only change made to the original formulation was in its procedure. Originally, the formulation required no chilling period, but a chilling period of 4045 minutes was added to the co okie preparation method The cookie control was prepared via the creaming method. After all the dry ingredients were sifted and combined, softened butter was folded into white sugar until smooth. At this point, the other wet ingredients were added, and finally the dry ingredients were incorporated in three parts. After the ingredients were completely combined, the dough was refrigerated for 4045 minutes to stiffen the dough. This step was added to help the control cookie hold a uniform shape and reduc e its spreading. Chia seeds were added to the control cookie formulation by the substitution method also. The seeds were ground and sifted by the same procedure used for the chia muffin product. When whole seeds were to be added to the formulation, no f lour was removed to compensate for the seed addition. After the proper chilling time was achieved the cookie dough was portioned using a level #50 scoop. The chia cookies were baked on a nonstick sheet pans in a WhiteWestinghouse 30 freestanding range ( Augusta, G A USA ) model WWEF3002KW at

PAGE 53

53 190C (375F) for ten minutes. Since the cookies were so small there was no need to rotate this product for even baking. Initially, there was the issue of the bottoms of the cookies over browning, but this was alleviated by baking the cookie product on parchment paper. Chia seeds were added to the cookie product for testing in the following ratios : 15% whole, 15% ground, and 5% ground/10% whole combination (Table 35). Producers of the ingredients used in the chia cookie formulations can be found in Appendix A. 3.2 Taste Panel Product Assessment All cookie and muffin products were evaluated in at the University of Florida/Institute of Food and Agricultural Sciences Food S cience and Human Nutrition sensory panel facility. All treatments of both chia products were evaluated on separate dates using different groups of untrained panelists. Each panelist was given three randomly coded samples, along with water and crackers for palate cleansing. Initially, panelist s were asked demographic questions (Table 31 ), then presented with a ballot (Figure 3 3), and were asked to assess the three products before them. Panelists were asked to assess the samples from left to right using a ninepoint hedonic scale (Figure 3 1) for each of the following attributes: overall acceptability, appearance, flavor, and texture. The two final assessments panelist were asked to make were: to choose how often they would consume each sample using the Food Action (FACT) rating scale (Figure 3 2) and to rank all three samples from most to least preferred. A ninepoint hedonic scale, ( Figure 3 1) anchored by 1=dislike extremely and 9=like extremely was used to capture panelist opinions of the products. The panelists were also asked to exp ress their consumption attitudes using the ninepoint FACT (Food Action) rating scale. This scale was anchored by 1=I would eat this only if forced and 9=I would eat

PAGE 54

54 this at every opportunity. These tests were performed using 75 untrained panelists, 31 males and 44 females. The age distribution of the panelists was as follows: 67.7% ages 1824, 22.6% ages 2534, 3.2% ages 3544, and 6.5% ages 4554. The data obtained from these sessions was tabulated by the Compusense program (Compusense Five 3.6 S ensory Analysis Software for Windows, Compusense, Guelph, Canda) twoway ANOVA, in concert with means separation via Turkeys HSD (Honestly Significant Difference) test was used to ascertain if there were any significant (p<0.05) differences between the t ested attributes of the chia products from those of the control products. The null hypothesis (H0) for the panel testing sessions was there are no significant differences between the attributes of the chia muffin products and the attributes of the control muffin, and the alternative hypothesis (HA) being there are significant differences between the attributes of the chia muffin product and the attributes of the level used in the data analysis was 0.05 or 5%. The determination of significance was rendered via pvalue calculations, whereas the pvalue must be less than the alpha level of 0.05. Previous panel testing evaluated whole seed and combination whole/ground seed muffins. From these efforts, it was found that consumers consistently preferred ground seed muffins, which led to the test samples of the final taste panel sessions. T he final taste panel assessed the consumer preference of a control, 40% ground, and a 25% ground chia seed muffin. The objective in the data collection was to compare the variation across the samples for each attribute. 3.3 Chia Product Physical Analysi s The chia seed cookie and muffin products were subjected to water activity (AW), color, and textural analyses. The water activity of the chia seed products was measured

PAGE 55

55 using the Aqua Lab water activity meter, Pullman, WA 99163 USA, model CX 2. The meter was verified to be calibrated using a dilute NaCl standard at 20C. The water activity of each cookie and muffin product sample variation (control, whole seed, ground seed, and combination whole/ground seed) was measured by three replications over five tr ials. Color analysis of the muffin and cookie products was performed using a Minolta Chroma meter 200B, Ramsey, NJ 07446 USA, which measured the intensity of reflectance at each wavelength. The colorimeter was calibrated using a white calibration plate, and was set to record sample L*, a*, b* values, also known as the Hunter Lab color scale. The L* values representing the lightness/darkness, 0=black and 100= white. The a* values denoted ( ) green and (+) red and the b* values measured ( ) blue and (+) yel low of the sample products. Color analysis was performed on the crowns of the muffing samples and the top surface of the chia cookie product. Each cookie and muffin product sample variations color was measured by three replications over five trials. The texture of the chia muffin products were measured using an Instron Universal Testing Machine, Norwood, MA 02062 USA, model 4411, at a cross head speed of 10 mm/min with a static 1.2 kg load cell. A 10mm aliquot using a size #6 cork borer was removed from each chia muffin sample. Each sample was subjected to compression testing using a #13 plunger. The samples were compressed twice to 50% of their original height, according to GrigelmoMiguel and others, 1999. During this test, the Instron measured the force needed to compress each sample. The attributes measured during this test included sample hardness first compression maximum force, cohesiveness ratio second cycle max. load/ first cycle max load, gumminess hardness x cohesiveness, and chewiness sp ringiness x gumminess. Muffin samples underwent

PAGE 56

56 four compression replicates over four trial runs. The chia cookie samples were subjected to a threepoint break test (Gaines 1991). This test was performed using the Instron noted above and a # 4 blade at a crosshead speed of 50mm/min. The bottom beams that the chia cookies rested on were spaced 4 cm apart, the diameter of the cookies averaged 5.46.5 cm. The threepoint snap test was used to assess the hardness or brittle nature of the chia cookie product by measuring the force needed to snap each cookie. Textural analyses of the chia cookie were replicated four times over four trial runs. The results of the AW, color and textural analyses were subjected to ANOVA tables to identify significant (p<0.05) differences between the cookie and muffin products and their control products. 3.4 Chia Product Fatty Acid Analysis To ascertain which chia seed variation within the cookie and muffin products withheld the largest amounts of omega3 fatty acid character; products were subjected to fatty acid analysis. Each sample variation underwent this analysis a total of three times to generate data for statistical analysis. The cookie and muffin products underwent similar analyses, according to the following standard m ethods: American Association of Cereal Chemistry method 6205, Association of Official Analytical Chemists methods 922.06, 996.06, and American Oil Chemists Society method Ce 1h05. The initial steps in sample preparation are as follows: each muffin sam ple was placed on parchment paper and cut into eight equal pieces, cookie samples were broken into eight pieces ensuring sample crumbs were not lost. The divided samples were left on the parchment paper to dry at room temperature (2022C) until the sampl es were approximately equal with the moisture of the ambient air, 1822 hours.

PAGE 57

57 After drying, the samples were ground finely enough to pass through a 20mesh sieve. The grinding was executed through the use of a food mill to create as little heat from fri ction as possible. Next, the samples were subjected to acid hydrolysis to extract the fat /lipid fraction. This was done by placing 2 grams of sample in a 50 mL beaker, and adding 2 mL of alcohol and stirred to moisten all the particles to prevent clumping. Then 10 mL HCL were added to the beaker and mixed well. The beaker was then placed in a water bath held at 7080 C, and stirred at frequent intervals for 3040 minutes. Afterward 10 mL of alcohol was added and the mixture was allowed to cool. The mixture was then transferred to a Mojonnier fat extraction apparatus. The holding beaker was rinsed into the extraction tube with 25 mL ether, added in three portions. After closing the flask and shaking for one minute, the mixture was left to stand unti l the upper liquid was clear. The ether fat solution was drawn off through a filter of cotton pledget in the funnel stem, into a 125 mL flask containing porcelain beads. The liquid remaining in the tube was reextracted twice, each time using only 15 mL of ether. These extracts were filtered into the original beaker. The extracted lipid was then dried in an oven at 100C for approximately 90 minutes and weighed. The sample was then left to air dry to a constant weight (30 minutes) and weighed once agai n. The results are reported as percent fat by acid hydrolysis. The next portion of the fatty acid analysis consists of preparing the fatty acid methyl esters (FAMES) from the extracted lipid. The extracted lipid was dissolved in 23 mL of chloroform an d 23 mL of diethyl ether. Then the mixture was transferred to a 3 dram glass vial and evaporated to dryness in a 40C water bath. After properly cooled,

PAGE 58

58 2.0 mL of 7% BF3 and 1.0 mL toluene were added to the vial. The vial was then sealed with a screw ca p top with a Teflon/silicone septum. The vial was then heated in the oven for 45 minutes at 100C, during this heating the vial is gently shaken every 10 minutes. After 45 minutes had elapsed, the vial was cooled to room temperature and 5.0 ml of water, 1.0mL of hexane, and 1.0g Na2SO4 are added. The vial was then capped and shaken for one minute. After the layers had time to separate, the top layer was transferred to another vial containing 1.0g Na2SO4 The removed top layer contains the prepared fatty equivalent) of sample was injected into the DSQ Trace GC Ultra (GC), Fisher Scientific, Pittsburgh, PA 15275 USA. The GC operating conditions used during experiment trials are as follows: inject ion port temperature250C, detector temperature250C, and oven temperature180C. The carrier gas used was helium, with a column head pressure of 286 kPa (41psi), flow rate of 1.0 mL/min, 26 cm/s linear velocity, a split ratio of 100:1. The individual FAMES were identified via the use of their retention times and by comparison with mixed and individual FAME reference standards, all standards were obtained from Nu Check Prep, Elysian, MN 56028, USA. The reference standards used for FAME comparisons are as follows: triglyceride internal standardC11:0 triundecanoin, FAME mixed standard solutionGLC 85, Nu Check Prep, MN 56028, USA; C11:0 FAME standard solutionC11:0 undecanoic methyl ester, individual FAME standards C4:0tetranoic methyl ester, C6:0h exanoic methyl ester, C14:0tetradecanoic methyl ester, C14:19 tetradecenoic methyl ester, C18:0octadecanoic methyl ester, C18:19 octadecenoic methyl ester, C18:29,12 octadecadienoic methyl ester, C18:39,12,15octadecatrienoic methyl ester, C20:18 eicosenoic methyl ester, C20:211,14-

PAGE 59

59 eicosadienoic methyl ester, C20:311,14,17eicosatrienoic methyl ester and C22:0docosanoic methyl ester.

PAGE 60

60 Table 31. Demographic Questions. Question # 1. Please indicate your gender. 0 Male 0 Female Question # 2. Male: Please indicate your age range. 0 18 24 0 25 34 0 35 44 0 45 54 0 55 64 0 Over 65 Question # 3. Female: Please indicate your age range. 0 18 24 0 25 34 0 35 44 0 45 54 0 55 64 0 Over 65

PAGE 61

61 dislike extremely dislike ve ry much dislike moderately dislike slightly neither like nor dislike like slightly like moderately like very much like extremely 1 2 3 4 5 6 7 8 9 Figure 31.Hedonic nine point scale Table 32. Food Action Rating Scale 1 I woul d eat this product only if I were forced to. 2 I would eat this product if there were no other foods choices. 3 I would hardly ever e at this product. 4 I dont like this product, but would eat it on occasion. 5 I would eat this if availa b le, but wouldn 't go out of my way. 6 I like this product and would eat it now and then. 7 I would frequently eat this product. 8 I would eat this product very often. 9 I would eat this product every opportunity I had.

PAGE 62

62 Question #4Sample <> Please indicate how much you like or dislike the following attributes in sample < > Overall Acceptability Dislike extremely Dislike very much Dislike moderately Dislike slightly Neither like nor dislike Like slightly Like moderately Like very much Like e xtremely 1 2 3 4 5 6 7 8 9 Appearance Dislike extremely Dislike very much Dislike moderately Dislike slightly Neither like nor dislike Like slightly Like moderately Like very much Like extremely 1 2 3 4 5 6 7 8 9 Flavor Dislike extremely Dislike very much Dislike moderately Dislike slightly Neither like nor dislike Like slightly Like moderately Like very much Like extremely 1 2 3 4 5 6 7 8 9 Texture Dislike extremely Dislike very much Dislike moderately Dislike slightly Neither like nor dislike Like slightly Like moderately Like very much Like extremely 1 2 3 4 5 6 7 8 9 Figure 32. Sample ballot used in chia product taste panel testing

PAGE 63

63 Table 3 3 Evaluated chia muffin product formulations Ingredient Control (g) 40% Ground seed (g) 25% Ground seed (g) 25% Whole seed (g) 10%Whole/15% Ground seed (g) All purpose flour 125 75 93.7 125 93.75 Granulated sugar 100 100 100 100 100 Brown sugar 25 2 5 25 25 25 Liquid eggs 40.6 40.6 40.6 40.6 40.6 UHT milk 125.1 125.1 125.1 125.1 125.1 Salt 1.7 1.7 1.7 1.7 1.7 Baking powder 2.3 2.3 2.3 2.3 2.3 Baking soda 1.15 1.15 1.15 1.15 1.15 Vegetable oil 7.3 7.3 7.3 7.3 7.3 Vanilla extract 2 2 2 2 2 Almon d extract 1.5 1.5 1.5 1.5 1.5 Golden raisins 15.7 15.7 15.7 15.7 15.7 Nutmeg 1 1 1 1 1 Cinnamon 1.5 1.5 1.5 1.5 1.5 Allspice 1.5 1.5 1.5 1.5 1.5 Butter 46.7 46.7 46.7 46.7 46.7 Chia seeds 0 50 31.2 31.2 12.5/18.7 Table 34 Evaluated chia cookie pr oduct formulations Ingredient Control (g) 15% Ground seed (g) 15% Whole seed (g) 10%Whole/5% Ground seed (g) All purpose flour 70 59.5 70 59.5 Granulated sugar 60 60 60 60 Liquid eggs 14.1 14.1 14.1 14.1 Baking powder 0.5 0.5 0.5 0.5 Baking soda 1 1 1 1 Vanilla extract 1.5 1.5 1.5 1.5 Chia seeds 0 10.5 10.5 3.5/7

PAGE 64

64 CHAPTER 4 RESULTS AND DISCUSSION 4.1 Consumer Taste Panel Testing 4.1.1 Chia seed muffin After the formulation of the control muffin was completed, variations of this control were tested for consumer acceptance at the University of Florida (UF)/ Institute of Food and Agricultural Sciences (IFAS), Food Science and Human Nutrition (FSHN) Taste Panel Facility, Gainesville, FL. The muffin products were assessed by panelist using acceptability and preference testing. Three different muffin samples (control, 25% ground, 40% ground), randomly number coded, were presented to the panelist. The first attribute to be assessed by panelists was overall acceptability. The hedonic means for this attribute were between 6.166.41, correlating to likes slightly on the hedonic scale, with standard deviations from 1.391.69 (Table 417). From calculations, the pvalue for this attribute was 0.42, indicating no significant difference across the muffin samples for this attribute. To confirm this finding, a Tukeys HSD test was run and similar results were noted at the 5% alpha level. The Tukeys value for this attribute was 0.507, no differences between any of the calculated means (Table 418) was larger than this value. These findings led us to accept the null hypothesis that there were no differences between the chia muffin products and the control as it pertained to this attribute. Some of the panelist did point out the breadlike nature of the muffin sampl es in regard to the overall acceptability of the muffin samples: Drier muffin that appears dry. More like bread i.e. banana bread than a muffin. There was no question in the demographic section that asked if the panelist actually liked muffins and how often they consumed them. This line of questioning may have been helpful in

PAGE 65

65 ascertaining consumer preferences of muffins in general. In regards to the aforementioned panelist comments, it must be noted that muffins are a type of quick bread, hence the bread like attribute. The next attribute assessed was muffin appearance. The hedonic means of this attribute were between 6.07 and 6.81, with standard deviations between 1.11.5 (Table 4 17), which correlates to the phrase like slightly from the hedonic scale. The pvalue for the appearance attribute was 0.0002, leading us to reject the null hypothesis that there are no differences between the chia muffin products and the control muffin due to a significant difference among sample treatments. To determi ne where this difference was, mean separation via the Tukeys test was consulted. The Tukeys value was 0.42 at a 5% alphalevel. It was elucidated that there was a significant difference between the control and 40% ground seed muffins. The 25% ground s eed muffin was not significantly different than either of the other two treatments (Table 419). This result was expected due the higher ground chia seed content of the 40% muffin which gave this sample a highly noticeable, darker color, different than tha t of the control muffin. There were only two negative comments from panelists about the color of the 40% muffin, citing the sample as having the worst look. As for the attribute of flavor, the hedonic scale means ( Table 417) were between 6.08 and 6.47, with standard deviations between 1.601.83, which translates to likes slightly on the hedonic scale. The pvalue for the flavor attribute at an alphalevel of 0.05 was 0.26. A pvalue such as this is an indication of no significant difference across th e sample treatments, leading us to fail to reject the null hypothesis that there were no differences between the chia muffin products and the control muffin. This was further verified by the Tukeys HSD test, rendering a value of 0.61 at a 5% significance level.

PAGE 66

66 In regard to flavor, panelists gave mixed comments. A portion of the panelists expressed satisfaction with the flavor, while others found the samples to be bland. During initial control formulations, this is an issue that was dealt with on sever al occasions. The control muffin was noted to have a good initial flavor that fell flat quickly. It is possible that the addition of higher spice levels and more sugar could solve this issue. There were a number of panelist that commented that the samples could have been sweeter. These comments are understandable since it has been reported that the American palette prefers sweeter products than other countries. Although all components of the muffin samples were held constant except the level of chia seed addition, increased levels of sugar seem to be a sensible addition to samples with higher ground seed levels. The texture attribute results were very similar to those of the flavor attribute. The hedonic scale means for this attribute were between 6.00 a nd 6.29 with standard deviations of 1.611.74 ( Table 4 17). The mean values displayed a likes slightly assessment across the various muffin treatment textures. At an alphalevel of 0.05, the p value of this attribute was 0.47, which demonstrates no sig nificant differences between the sample treatments, leading us to fail to reject the null hypothesis that there were differences between the chia muffin products and the control muffin. Although unnecessary, the Tukeys value was 0.572 (Table 421), conf irming the pvalue result. Both chia muffin samples received positive comments about the moistness of the samples, but it seemed from their comments that panelist found the 40% muffin to be the moister of the samples, even though this is not displayed in the data. The 25% muffin received comments pertaining to the dryness of the sample, and this observation is supported by the % moisture loss data (Appendix B). From this data,

PAGE 67

67 it was found that a higher content of ground seeds led to higher moisture retention. The Food Action (FACT) rating scale was used to determine the consumption behavior of panelists toward the three muffin treatments. The FACT scale means for the muffin sample were 4.915.21 with standard deviations from 1.46 to 1.86 (Table 417). T hese mean values coincide with consumption behaviors equivalent to those described as I would eat this if available, but would not go out of my way. The pvalue for the panelist consumption behavior was 0.36, indicating no significant differences among c onsumer consumption attitudes across the muffin treatments, allowing us to fail to reject the null hypothesis that there were no differences between the chia muffin products and the control muffin. Taking the aforementioned panelist comments into considera tion and adjusting the samples accordingly may have a positive effect on consumption behavior data. In the final assessment panelist were asked to rank the muffin samples from most to least preferred. The data obtained from panelist subjected to the Friedmans Test, a nonparametric randomized block analysis of variance. The null hypothesis for this test was there is no difference in rank data for the repeated measure (the actual ranking). The calculated Friedmans statistic for this evaluation was 1.6 2 with 2 degrees of freedom. At an alpha level of 0.05, the pvalue of 0.44 at the aforementioned alphalevel indicating no significant difference in the rank data (Table 417). This finding allows us to fail to reject the null hypothesis. 4.1.2 Chia s eed cookie Once the cookie formulation process was completed, and a viable control product was produced, sensory testing commenced. Sensory panel testing was executed at the UF/IFAS FSHN Taste Panel Facility, Gainesville, FL. The chia cookie product was assessed through the use of acceptability and preference testing. The cookie samples

PAGE 68

68 were evaluated by a total of 75 untrained volunteer panelists, 43 male and 32 females. The age distribution of these panelists is as follows: 76.7% ages 1824 and 23.3% ages 2534. Panelist received three randomly coded samples (control, 5%ground/10%whole seed, and 15%ground/15%whole seed) with crackers and water for palate cleansing. Once the samples were received, panelists were asked to assess each sample from left to right using a ninepoint hedonic scale (F igure 3 1) to rate the following attributes: overall acceptability, texture, flavor and appearance. At the conclusion of the product assessment, panelist were asked to give an overall ranking of the three samples and to comment on their consumption behavior towards the three samples using the Food Action (FACT) rating scale. From preliminary testing, the final sample field was narrowed to contain a control, 15% ground/15% whole seed, and 5% ground/10%whole seed cookie. The compiled data was analyzed using the Compusense program via oneway Anova and the Tukeys HSD (Honestly Significant Difference) test for mean separation. Cookie sample rank data was analyzed using the Friedmans test, a nonparametric random ized block analysis of variance. The alphalevel for the taste panel data analysis was 0.05 with significance demonstrated with pvalues less than the assigned alphalevel. The null hypothesis (H0) of this test was there are no significant differences b etween the attributes of the chia cookie treatments versus the attributes of the cookie control. The alternative hypothesis (HA) was as follows: there are significant differences among the attributes of the chia cookie treatments and the attributes of t he cookie control.

PAGE 69

69 The first attribute of the chia cookie samples to be assessed was overall acceptability. The hedonic means of the treatments for this attribute ranged from 5.456.29 with standard deviations ranging from 1.571.73 (Table 424). On the hedonic scale the mean values translate to the descriptors: neither like nor dislike and like slightly. The p value, at an alphalevel of 0.05, for this attribute was calculated to be 0.0001, indicating a significant difference among the treatments This allowed us to the reject the null hypothesis of there are no differences among the attributes of the chia cookie treatments vs. the control cookie. Mean separation via Tukeys HSD test level there was a significant difference between the 15% ground/ 15% whole chia seed cookie and the other two treatments, control and 5%ground/10% whole seed cookie samples ( Table 425) The control and 5% ground/ 10% whole seed cookie had no significant differences, but according to these results we must reject the null hypothesis in regards to this attribute. From panelist comments, it was noted that several panelist did not like the quantity of or the addition of whole chia seeds in the cookies. One panelist commented, Seemed like a healthy cookie which turned me off immediately, cookies shouldnt be healthy. The next evaluated attribute was cookie appearance. The hedonic means fell within 5.677.00 with standard deviations between 1.44 and 1.83 ( Tab le 4 24). Means of this nature correlate to the descriptors like slightly and like moderately on the hedonic scale. The plevel was calculated to be 0, demonstrating a significant difference across th e treatments. According to the Tukeys test (Table 426 ) ; the control sample was significantly different from the other two treatments leading us to reject the null hypothesis that there are no

PAGE 70

70 differences between the attributes of the chia cookie treatm ents vs. the control cookie. This result was expected due to the presence of whole seeds in the chia cookie samples and the slightly darker color from the ground seed component. The third attributed assessed by panelists was flavor. The means for the thr ee treatments ranged from 5.41 to 5.96 with standard deviations between1.62 and 2.02 (Table 424). According to the hedonic scale the mean values correlate to neither like nor dislike and like slightly. The calculated pvalue for this attribute at a 5% alphalevel was 0.03 (Table 424 ) After employing the Tukeys HSD test for means separation, it was found that there were no significant differences between any of the cookie treatments. From the calculated data we must fail to reject the null hypothesis that there are no differences between the attributes of the chia cookie treatments vs. the control cookie as it relates to this attribute. Panelist often commented on an unfamiliar flavor, possibly that of the chia seeds. Since the 15%/15% cookie cont ained the higher chia content, it consequently was set apart from the other samples. Some panelist went as far as calling the unfamiliar flavor unpleasant. The next evaluated attribute was cookie texture. The treatment hedonic means for this attribute ranged from 5.52 to 6.92, with standard deviations from 1.50 to 1.81 (Table 424). Attribute hedonic means of this nature coincide with the hedonic phrases: like slightly and like moderately, respectively. The plevel of 0.05, was 0.0 indicating a significant difference between treatments. This difference was pinpointed using the Tukeys HSD test. The Tukeys value was 0.495, and it was found that the control and 5% ground/10% whole seed cookies were both significantly different than the 15% ground/15% whole seed cookie, but not from one another (Table 428). This once again

PAGE 71

71 leads us to reject the null hypothesis that there are no differences between the attributes of the chia cookie vs. the control cookie. This separation was expected, one panelist did comment on the 15%/15% cookie having a weird grainy texture. The food action rating scale was used to assess the consumption behavior of the panelists as it related to each cookie treatment. The FACT scale means for the treatments ranged from 4.35 4.96, with standard deviations from 1.561.83 (Table 424). These means coincide with the following the FACT scale phrases: I dont like this product, but would eat on occasion and I would eat this if available, but would not go out of my way, respectively. The plevel of 0.05, indicating a significant difference across the treatments (Table 424). The 15% ground/ 15% whole seed cookie was found to be significantly different when compare to the control and 5%ground/10% whole seed cookie, hence rejecting the null hypothesis of there being no differences between the attributes of the chia cookie treatments vs. the control cookie (Table 429). Finally, panelists were asked to rank the three cookie samples from most to least preferred. The rank totals ranged from 122180 ( Table 424). The data generated from this test was analyzed via th e Friedman test, which rendered a calculated value of 22.5 versus a critical value of 5.99, with a pvalue of 0. A comparison of the two values yields significant differences across the treatments. A plevel of 0.05 leads us to reject the null hypothesis due to the differences across the treatments. The 15%whole/15% ground cookie product was less preferred the other two sample treatments when rank totals were compared. No differences were found among the control and 5%whole/10%g round chia cookie product. From the data tabulation, it was found that the higher chia content in the

PAGE 72

72 15%/15% cookie, even though not much higher than the 5%/10% cookie, had attributes that were consistently significantly different than the control and 5%/ 10% cookie samples during panelist evaluations. 4.2 Product Physical Analysis 4.2.1 Chia Muffin Analysis Consumer preference data generated from taste panel testing was analyzed to determine which chia muffin sample consumers found most acceptable. From this data, it was determined that the 25% ground chia muffin was the most accepted by consumers. With this knowledge, other chia muffin products were produced with 25% chia being added to these products in different forms, i.e. ground, whole and combination (whole and ground). Each form of chia seed was added to the control formula via the substitution method except for the whole seed muffin. For this muffin, whole seeds were added without removing any of the flour component. Each variety of chia muffin w as subjected to a battery of tests: water activity (Aw), L*a*b* color analysis and textural analysis. During the chia muffin products physical analysis, the null hypothesis (H0) that was tested is as follows: there are no significant (p>0.05) differences between the tested physical attributes of the chia product samples versus the control muffin. The alternate hypothesis (HA), being the opposite of the null was there are significant (p<0.05) differences between the tested physical attributes of the chi a product samples versus the control muffin. To fail to reject (accept) or reject the null hypothesis, the pvalue, calculated from oneway ANOVA tables was used at the significance level of 0.05 to measure data consistency. The pvalue is defined as the probability of obtaining a value of the test statistic that is as likely or more likely to reject H0 as the actual observed value of the test statistic This probability is computed

PAGE 73

73 assuming that the null hypothesis is true (Ott and Longnecker 2004). If s ignificant differences were obtained, the data was then subjected to means separation via the Duncans Multiple Range test. The coefficient of determination (r2) was also consulted to ascertain the magnitude of correlation strength of the tested variable and the seed form of the muffin product. 4.2.1.1 Water a ctivity Perishability is a major concern for all packaged products. One determining factor of product perishability is the association of water with nonaqueous constituents of said products. The str ength of this association dictates the shelf life of a product due to its support of activities that lead to product degradation e.g. microorganism growth, oxidation, and Maillard browning. To measure this association, a ratio of vapor pressure in the wat er of the sample/ the vapor pressure of pure water (p/po) is employed which is also known as water activity (Aw) (Fennema 1996). During the water activity testing of the muffin products, an aliquot of each muffin was removed and placed in a testing recepta cle. Water activity data was measured by the Aqua Lab Cx 200 at room temperature (20C). A total of five trials were run with three replications within each trial for each muffin type. The data generated was collected and enter into the SAS program. Us ing this program, a twoway ANOVA was performed comparing variation across the treatments. During this comparison there were 60 observations in total. The mean Aw reading of the muffin samples was 0.87. When the p value of 0.97 was calculated at the 0.05 alpha level, no significant differences were found amongst the water activity levels of the muffin samples. To further confirm these results, a means separation test was run using the muffin water activity (Table 42). Once again, there were no signifi cant differences found between

PAGE 74

74 the muffin treatments, leading us to accept the null hypothesis that there are no differences between the physical attributes of the chia product samples vs. the control muffin. The coefficient of determination (r2) displayed a lack of correlation (0.004) between the Aw values and the various muffin sample treatments. The results of this test were attributed to the samples being very similar, it was hypothesized that the whole seed products may render some variation in Aw due to the polysaccharide mucilage that is excreted from chia seeds once immersed in a polar solvent. During these trials this was not the case. 4.2.1.2 Color analysis The surface color of each muffin crown was measured using the Minolta Chroma meter 200B. The color of each muffin treatment was analyzed over five trials with three replications during each trial. During the test runs, light/darkness (L*), red/blue (a*), and yellow/green (b*) light reflectance were measured. Once these attributes were measured, the compiled data was analyzed using the SAS program. A oneway ANOVA was employed to compare variation across the treatments for each color variable analyzed: L*, a*, and b*. The mean L* value, a measure of lightness (100=white) and darkness (0=black), of the chia muffin products was 42.93. When accounting for the calculated pvalue (<0.001), a significant difference was found between the chia muffin products for the attribute of light/darkness of color. This finding was also confirmed using a separati on of the means (Table 4level of 0.05, the control and 25% whole seed muffins were not significantly different from each other. The 25% ground seed muffin and the combination (ground and whole seed) muffin were not significantly different fr om each other either. On the other hand, both aforementioned pairs of muffins were significantly different from each other in terms of

PAGE 75

75 light/darkness values, leading us to reject the null hypothesis that there are no differences between the physical attri butes of the chia product samples vs. the control muffin. The coefficient of determination (r2), demonstrated a rather small (0.48) yet positive correlation between the L* color value and the muffin sample treatments. This result was expected due the simi lar nature of the paired muffin samples. The pairs have similar light/ dark measurements because their differences are quite minute. The mean a* color value of the muffin samples was 5.49, yielding more red reflectance than green. The calculated pvalue of this color variable was 0.0008, demonstrating significant differences between the sample treatments at the 0.05 alpha level. Separation of the sample a* color value means via the means separation test (Table 4level of 0.05, was performed. This test concluded that all of the sample treatments were significantly different than the control muffin, rejecting the null hypothesis. The coefficient of determination (r2) for the a* value was 0.257, although a positive correlation, the magnitude between the a* color value and muffin treatments is very low. The final color value, b*, is a measur e of yellow and blue light wave reflectance of the chia muffin samples. The mean b* color value of the muffin samples was 16.79, resulting in a higher reflection of yellow light than blue light. The calculated p value for this color variable at a 0.05 alpha level was less than 0.0001, demonstrating a significant difference between muffin sample b* color values. According to Duncans Multiple Range test (Table 45), the control and 25% whole muffin samples were not significantly different from each other i n regard to their b* color value. The combination and 25% ground muffin samples were not significantly different from one another either, but both samples were significantly different than the two aforementioned samples

PAGE 76

76 leading to the rejection of the nul l hypothesis. Visually, the control and whole seed muffins looked very similar. Each of these samples were golden brown, with slightly darker edges, the only different between the two samples visually was the appearance of the black and white chia seeds t hroughout the whole seed muffin. The combination and ground seed muffins were equally darker brown than the aforementioned samples, there only contrast was the whole seeds of the combination muffin. 4.2.1.3 Textural analysis The texture of each chia muffi n product was tested and compared to that of the control muffin. All muffins samples had a 10 mm aliquot removed from the whole muffin using a #6 cork borer. Texture measurements were performed over four trials with four replicates using the Instron Univ ersal Testing Machine, model 4411. During each trial, the 10 mm muffin aliquot was compressed twice to 50% of their original height using a #13 plunger. The textural attributes assessed were hardness, springiness, cohesiveness, gumminess, and chewiness. The null hypothesis for this test was there were no significant (p>0.05) differences between the tested attributes of the chia muffin products versus the control muffin. After testing each muffin sample, the data was compiled and analyzed with SAS using a two way ANOVA, comparing variation across the treatments for 64 observations. The first attribute to be analyzed was muffin hardness, which can also be defined as first cycle (first compression) maximum force measured in newtons (N). The mean hardness o f all the muffin samples was 1.01 N. The calculated pvalue for this attribute was 0.0079, indicating a significant difference amongst the samples. After applying a means separation test to the data (Table 411), it was concluded that the combination (10% whole/ 15%gound seed) muffin was the only sample that was statistically different from the control muffin. It is thought that this

PAGE 77

77 result is due to the removal of a portion of the flour component in the combination muffin and replacing it with ground chi a, thereby weakening the muffin matrix. It can be argued that the same action takes place in the ground seed muffin, but the whole seeds tend to create inconsistent leavened cavities that weaken the muffin matrix. Although only one sample was different f rom the control, it was enough to reject the null hypothesis. The r2 value for this attribute was 0.17, a positive but weak correlation between muffin hardness and the various forms of seed in the muffin products. The next tested attribute was muffin spr inginess or recovery (S) in millimeters. The overall mean muffin springiness was 5.42S, and at an alpha level of 0.05, the pvalue was calculated to be 0.02. A pvalue as such demonstrates a significant difference in the springiness across the muffin sam ples. This difference was found to be between the 25% whole muffin and the ground and combination muffins, according to a separation of the means ( table 412). The control muffin was found not to be significantly different than any of the individual muff in samples. A result such as this leads us to fail to reject the null hypothesis that there are no differences between the physical attributes of the chia product samples vs. the control muffin due to the means separation which confirms the null. The r2 value for muffin springiness was 0.14, positive, but weak correlation between muffin springiness and the different seed treatments. Cohesiveness is defined as a ratio of the second cycle energy maximum load (J) / first cycle energy maximum load (J). Eac h individual compression is defined as a cycle. When comparing the cohesiveness of the chia muffin products and the control muffin, the mean value of 2.53 J. At an alpha level of 0.05, the pvalue for this attribute was 0.84, yielding no significant difference across the muffin treatments. This finding leads us to fail to reject

PAGE 78

78 the null hypothesis. The cohesiveness r2 value was 0.01, indicating a lack of correlation between muffin cohesiveness and the seed treatments. Muffin gumminess is defined as fir st cycle maximum force x cohesiveness and is measured in newtons (N). During testing, this attribute was found to have an overall mean of 2.36 N, with a pvalue of level of 0.05. A pvalue of this magnitude indicates that there are no significant differences across the muffin treatments for the gumminess attribute. The r2 value for this attribute was 0.07, displaying a lack of correlation between gumminess and the various seed treatments. The final analyzed attribute, chewiness is related to how a consumer perceives a product in terms of labored mastication. Chewiness is defined as springiness gumminess (N mm). This attribute had an overall mean of 12.84 Nmm, and a pvalue of 0.03 at a 0.05 alpha level. This p value is r ather close to 0.05, but still not close enough to result in an insignificant difference amongst treatments. According to the means separation test (Table 415), the 25% whole muffin and the combination muffin were significantly different from one another leading, but n either sample was significantly different than the control, according to the means separation, leading us to fail to reject the null hypothesis. The r2 value for this attribute was 0.12, resulting in a weak correlation between chewiness and the different seed treatments. 4.2.2 Chia Cookie Analysis From consumer preference data, it was determined that the chia cookie product with a 15% seed addition was most preferred by consumers. In response to this finding, three chia cookie products were produced with 15% seeds added in whole, ground, and combination 5% whole and 10% ground forms. The seed additions were rendered via the substitution method, substituting the added seeds for an equivalent amount of flour.

PAGE 79

79 The whole cookie product was the only treatment with no flour removed; the 15% seed addition was combined directly to the control formulation. After the various cookie samples were produced each variation was subjected to water activity (Aw), L*a*b* color analysis, and a triplebeam snap test. The dat a collected from these tests were compiled and analyzed with SAS. The pvalue generated from ANOVA tables was used to determine significance (p<0.05) differences among samples. During the chia cookie products physical analysis, the null hypothesis (H0) t hat was tested is as follows: there are no significant (p>0.05) differences between the tested physical attributes of the chia product samples versus the tested physical attributes of the of the control cookie. The alternate hypothesis (HA), being the opposite of the null was there are significant (p<0.05) differences between the tested physical attributes of the chia product samples versus the tested physical attributes of the control cookie. To fail to reject (accept) or reject the null hypothesis, the pvalue, calculated from twoway NOVA tables was used at the significance level of 0.05 to measure data consistency. If significant differences were obtained, the data was then subjected to means separation via the Duncans Multiple Range test. The coef ficient of determination (r2) was also consulted to ascertain the magnitude of correlation strength of the tested variable and the seed form of the cookie product. 4.2.2.1 Water activity The four cookie samples (control, whole, ground, and combination seed) were subjected to water activity (Aw) analysis. An aliquot of each cookie was used for this analysis which was conducted at room temperature (20C). Water activity data was obtained through the use of the Aqua Lab Cx 200. The Aw of each cookie sample w as measured over five trials with three replications during each trial. The mean Aw of the

PAGE 80

80 cookie samples was 0.47. Upon separating the means (Table 46); it was found that the control cookie and the whole seed cookie were not significantly different. The results lead us to reject the null hypothesis that there are no differences between the physical attributes of the chia product samples vs. the control cookie The ground and combination cookie products were not significantly different from each other, but these two variations were different than the control and whole seed cookies (Table 46 ). The water activity results were as such because of the similar nature of the cookie samples. The whole and control cookies only differed due to the addition of th e whole seeds. While the combination and ground seed cookies shared the ground seed component leading the samples water activities to mimic on another. The r2 value of this test was 0.05 ( Table 41) indicating a positive, but weak correlation between product Aw and the various forms of seed used in the cookie products. 4.2.2.2 Color analysis Each chia cookie sample was subjected to L*a*b* color analysis. This analysis was carried out using the Minolta Chroma meter 200B. To obtain product color data, the colorimeter was calibrated with a white calibration plate, and the intensity of reflectance of each wavelength of the cookie surface was measured. The measurements were rendered using the Hunter Lab color scale producing L*a*b* values. The L* value is an indicator of product lightness (100=white) and darkness (black=0), the mean L* value of the chia seed cookie products was 58.29. The pvalue level was <0.0001. This finding indicates that there is some significance between the L* value, light/darkness of the chia cookie samples. To confirm this significance, a means separation test was employed. From the Duncans level, the control and 15% whole cookie products

PAGE 81

81 were found not to be signifi cantly different from each other (Table 47). This was also the case between the 15% ground and combination (whole and ground) chia cookie products. A significant difference was found between the pairs of aforementioned cookie treatments leading us to reject the null hypothesis. These results followed the same pattern as those of the water activity testing, due to the similar nature of the components of the samples. The r2 value for this tested variable was 0.33, which is an indication of a positive, but weak correlation between the L* value and he different seed treatments. The next variable in the color analysis to be considered is the a* color value. The mean a* value for the chia cookie products was 12.05 (Table 41), demonstrating more red light re level of 0.05, the pvalue for a* was 0.38, no significant differences between samples (Table 48). This result was confirmed by the means separation test as well, leading us to accept the null hypothesis that there are no differences between the physical attributes of the chia product samples vs. the control cookie. The r2 value for the a* color value was 0.05, which shows a lack of correlation between the a* color variable and the different chia treatments used in the cook ie products. The final color variable b* was measured to have a mean of 25.73 (Table 41), demonstrating a higher level of yellow reflected light than blue. At the established alpha level of 0.05, the pvalue of the b* color variable was 0.0033 yielding a significant difference. After means separation, it was elucidated that all of the treatments were different than the control with respect to this color variable (Table 49), leading us to reject our null hypothesis that there are no differences between the physical attributes of the chia product samples vs. the control cookie. The r2 value for this color value was 0.21, although positive it indicates a weak

PAGE 82

82 correlation between the b* color value and the various treatments of the seed in the cookie product. Visually, the control and whole seed chia cookie product were very similar. Both samples had a pale yellow hue, and browned edges. The whole seed cookie was speckled with black and white chia seeds. The combination and ground chia seed cookie product s were visually darker than the aforementioned samples due to the ground seed component. The ground outer seed coats caused the darker color of both samples, with their only contrast being the unground seeds of the whole seed chia cookie. 4.2.2.3 Textural analysis The final test the chia cookie product was subjected to was the threepoint break test. The threepoint break test was carried out over four trials with four replications, using the Instron Universal Testing Machine, model 4411. During these tr ials, the cookies averaged 5.46.5 cm. A third beam, a #6 blade was attached to a static 1.2kg load cell, with a crosshead speed of 50 mm/min. The Instron measured the maximum compressive load (N) needed to snap each cookie sample. The collected data w as analyzed using the SAS program; a twoway ANOVA was performed to compare variation across the treatments. To determine significant differences, a calculated pvalue was employed. The null hypothesis (H0) for this test was there were no significant (p >0.05) difference in the amount of force needed to snap each chia cookie sample vs. the force needed to snap the control cookie. The alternative hypothesis (HA) was there was significant (p<0.05) difference in the amount of force needed to snap each chia cookie sample vs. the force needed to snap the control cookie. During this comparison, there were a total of 64 observations. The mean maximum compressive load of the three point break test was 19.23N. The pvalue of this test was calculated to

PAGE 83

83 be 0.00 3 at the 0.05 alpha level. This pvalue indicates a significant difference across the sample treatments for this test. To determine this difference, Duncans Multiple Range test was employed. At the 0.05 alpha level, the 15% ground chia cookie product w as statistically different than all other cookie sample treatments (Table 416), leading us to reject the null hypothesis. This result was surprising due to the moisture loss percentage results (Appendix B). From this data, the control cookie demonstrated the highest moisture loss, with the ground seed and combination cookie samples close behind it. Given these results, it was expected that more force would have been needed to snap the drier cookie samples, but this was not the observation. The r2 value, coefficient of determination, for this test was 0.26. This value makes a weak, but positive correlation between the threepoint break test and the various seed treatments used in the chia cookie formulations. 4.3 Product Chemical Analysis Following the analysis of the textural characteristics of the chia products, analysis of the product fatty acid components commenced. The fatty acids that were the focus of this analysis were alphalinolenic acid (18:3), followed by linoleic acid (18:2). Although alphalinolenic acid is the primary omega3 fatty acid, it is important to note the quantity of its omega 6 counterpart because they both compete for the same carbon elongation pathways. During this analysis, all the fatty acids extracted from the lipid layer of each chia product was measured paying particular attention to the 18:2 and 18:3 fatty acids. After this was completed, the variation across the muffin and cookie treatments was assessed to establish if a significant (p<0.05) difference existed between t he treatments. This information was obtained through the use ANOVA. The null hypothesis (H0) of this analysis was there are significant differences across the

PAGE 84

84 chia muffin and cookie treatments in regard to omega3 and omega6 fatty acid content. The al ternative hypothesis (HA) was there are no significant differences between the chia muffin and cookie treatments in regard to the omega3 and omega6 fatty acid content. The alphalevel of the statistical analysis was 0.05. When necessary means separat ion was employed using the Duncans Multiple Range test. 4.3.1 Chia Muffin Product It must be noted that all the following quantities discussed in this and the accompanying section are reported in grams per 100 grams of product. The control muffin was fou nd to contain an average of 12.42% total fat over 3 trials (Table 436). When separated into its fatty acid components there was a greater content of linoleic linolenic acid, although there was no chia seed addition to the control muffin. T his finding is attributed to the canola oil and butter in the muffin formulation. The combination (ground/whole) chia seed muffin was measured to have a total fat percentage of 13.97, with 18:2 and 18:3 fatty acid levels of 1.09 and 0.95, respectively. T he largest total fat percentage was obtained from the ground seed muffin, 14.25%. The greater content of total fat was in direct proportion with a higher measured level of alphalinolenic acid, the highest amongst all muffin treatments, 1.47g. The lowest 18:3 was measured in the whole seed muffin which also contained the lowest level of total fat amongst samples with chia seeds at 12.85%.These findings were rather surprising because it was originally hypothesized that the whole seed muffin would contain higher levels of the 18:3 fatty acid. The whole seed product was perceived in this fashion because the chia seeds were mechanically uncompromised whereas the ground seeds were. Initially, it was thought that the grinding of the seed would cause some loss of the omega3 character of the seeds. From this chemical analysis, this was not the case.

PAGE 85

85 Generally, when chia seeds come in contact with a polar liquid it releases a polysaccharide mucilage. It is possible that the ground seeds create a stronger web vi a this mucilage formation and ground seed coat, repelling any omega3 release. On the other hand, the whole seed mucilage created of equal strength, holds the individual seeds in a segregated suspension from one another. These and other theories must be researched further for validation. During the statistical analysis of the muffin products fatty acid content, it was found that there was no significant difference across the treatments for linoleic acid content. This observation is due to a pvalue of 0 .14, the coefficient of determination ( r2 ) was 0.47, demonstrating a mild correlation between muffin seed variation and linoleic acid content (Table 431). The means separation also confirmed no significant difference amongst the treatments (Table 432). linolenic acid, the statistical analysis displayed a significant difference across the seed treatments. The means amongst the treatments was 0.74g, with a p value of <0.0001at level of 0.05. This leads us to fail to reject the null hypothesis. The r2 value for this test also displayed a strong positive correlation between the seed treatments and the 18:3 fatty acid. According to the Duncans Multiple Range test, the control and whole seed muffins were not significantly different from one another, but the ground and combination seed muffins were different from each other and all other treatments. This finding was expected due to the measured fatty acid content from earlier testing. 4.3.2 Chia Cookie Product The chia cookie product was subjected to similar chemical and statistical analysis as the chia muffin product. As aforementioned fatty acid values are expressed as grams per 100 grams of product. The control cookie was found to have a total fat content of 21.48%, with a minimal amount 18:2 and 18:3 fatty acids. The detected

PAGE 86

86 quantities were attributed to the butter within the cookie formulation, although not a dominant fatty acid in butter 18:2 and 18:3 is present in small amounts. The combination (ground/whole seed) cookie was meas ured to have a total fat content of 12.21%. After fatty acid extraction this cookie treatment displayed the second lowest quantity of 18:2 and 18:3 acids (Table 437). The lowest 18:2 and 18:3 content was measured in the whole seed cookie at 0.83 and 0.23 respectively. This cookie treatment also had 21.34% total fat content. The measurement of such a small fatty acid content directly mimics that of the chia muffin product. The linoleic acid content across the chia cookie treatments had a mean of 0.93g, and a pvalue of 0.02 at a 0.05 confidence level (Table 431). A pvalue of this nature is an indication of a significant difference across the treatments leading us to fail to reject the null hypothesis. The means of this fatty acid were separated via D uncans Multiple Range test. This means separation found the control and whole seed cookies were not significantly different from one another, but the combination cookie was not significantly different than any cookie sample (Table 434). The alphalinole nic fatty acid content of the chia cookies was found to have a mean of 0.63g across all treatments. The pvalue for this fatty acid was <0.0001 at a 0.05 confidence level, demonstrating a significant difference across the seed treatments. From this obser vation, we are inclined to fail to reject the null hypothesis. The r2 value of this test was 0.96, demonstrating a strong correlation between the seed treatments and the 18:3 fatty acid. According to the multiple range test, the only two treatments that are not significantly different from one another are the control and whole seed cookie (Table 435). This result was expected due to the lower fatty acid content of these seed treatments.

PAGE 87

87 Table 41. Developed chia muffin and chia cookie products surface L*A*B* color value analysis and water activity analysis results at 1921 C. Sample Variable r 2 Mean p value Muffin Aw 0 0.87 0.97 L* 0.48 42.93 <0.0001 a* 0.25 5.49 0.0008 b* 0.43 16.79 <0.0001 Cookie Aw 0.15 0.47 0.025 L* 0.33 58.29 <0 .0001 a* 0.05 12.05 0.38 b* 0.21 25.73 0.003 Table 42. Developed chia muffin product separation of means measuring the attribute of water activity (Aw) (1921C) when comparing variation across sample treatments. Treatment N Mean Duncan Group* c ontrol 15 0.87 A 25% ground 15 0.87 A combination 15 0.87 A 25% whole 15 0.87 A *means with same letter are not significantly different. Table 43. Developed chia muffin product separation of means measuring the attribute of L* surface color value whe n comparing product color variation across sample treatments. Treatment N Mean Duncan Group* control 15 45.77 A 25% ground 15 39.63 B combination 15 39.51 B 25% whole 15 46.8 A *means with same letter are not significantly different. Table 44. Devel oped chia muffin product separation of means measuring the attribute of surface a*color value when comparing product color variation across sample treatments. Treatment N Mean Duncan Group* Control 15 6.54 A 25% ground 15 4.88 B Combination 15 5.44 B 2 5% whole 15 5.10 B *means with same letter are not significantly different.

PAGE 88

88 Table 4 5. Developed c hia muffin product separation of means measuring the attribute of b* color value when comparing color variation across sample treatments Treatment N Mean D uncan Group* control 15 19.20 A 25% ground 15 14.31 B combination 15 15.46 B 25% whole 15 18.20 A *means with same letter are not significantly different. Table 4 6. Developed c hia cookie product separation of means measuring the attribute of water activity (Aw) (1921C) when comparing variation across sample treatments Treatment N Mean Duncan Group* control 15 0.50 A 15% ground 15 0.45 B combination 15 0.45 B 15% whole 15 0.50 A *means with same letter are not significantly different. Table 4 7. Developed c hia cookie product separation of means for the attribute of L color value when comparing variation across sample treatments. Treatment N Mean Duncan Group* control 15 61.18 A 15% ground 15 54.54 B combination 15 55.89 B 15% whole 15 6 1.55 A *means with same letter are not significantly different. Table 4 8. Developed c hia cookie product separation of means measuring the attribute of a* color value when comparing variation across sample treatments. Treatment N Mean Duncan Group* cont rol 15 2.27 A 15% ground 15 42.3 A combination 15 2.54 A 15% whole 15 1.11 A *means with same letter are not significantly different.

PAGE 89

89 Table 49. Developed chia cookie product separation of means measuring the attribute of b* color value when c omparing variation across sample treatments. Treatment N Mean Duncan Group* Control 15 28.28 A 15% ground 15 23.99 B Combination 15 24.84 B 15% whole 15 25.8 B *means with same letter are not significantly different. Table 4 10. Developed chia muffin product compression test analysis and chia cookie product triple beam break test analysis results. Sample Variable r 2 Mean p value Muffin Hardness 0.17 1 0.007 Springiness 0.14 5.42 0.02 Cohesiveness 0.01 2.53 0.84 Gumminess 0.07 2.36 0.17 Ch ewiness 0.1 0 12.84 0.04 Cookie break test 0.26 19.23 0.0003 Table 411. Developed c hia muffin product separation of means, measuring the physical attribute of hardness when comparing variation across sample treatments. Treatment N Mean Duncan Group* co ntrol 16 1.06 A 25% ground 16 1.05 A combination 16 0.61 B 25% whole 16 1.30 A *means with same letter are not significantly different. Table 4 12. Developed c hia muffin product means of separation measuring the attribute of springiness when comparing variation across sample treatments. Treatment N Mean Duncan Group* control 16 5.69 AB 25% ground 16 4.80 B combination 16 4.77 B 25% whole 16 6.43 A *means with same letter are not significantly different.

PAGE 90

90 Table 4 13. Developed chia muffin product separation of means measuring the attribute of cohesiveness when comparing variation across sample treatments. Treatment N Mean Duncan Group* control 16 2.41 A 25% ground 16 2.54 A combination 16 2.23 A 25% whole 16 2.93 A *means with same le tter are not significantly different. Table 4 14. Developed c hia muffin product separation of means measuring the attribute of gumminess when comparing variation across sample treatments. Treatment N Mean Duncan Group* control 16 2.27 AB 25% ground 16 2.34 AB combination 16 1.32 B 25% whole 16 3.51 A *means with same letter are not significantly different. Table 4 15. Developed c hia muffin product separation of means measuring the attribute of chewiness when comparing variation across sample treatm ents. Treatment N Mean Duncan Group* control 16 13.3 AB 25% ground 16 11.97 AB combination 16 6.45 B 25% whole 16 19.64 A *means with same letter are not significantly different. Table 416. Developed c hia cookie product separation of means measuring the cookie break characteristic when comparing variation across sample treatments. Treatment N Mean Duncan Group* control 16 16.33 B 15% ground 16 27.36 A combination 16 19.68 B 15% whole 16 13.55 B *means with same letter are not significantly dif ferent.

PAGE 91

91 Table 4 17. Developed c hia muffin product sensory evaluation panel preference test results measuring six qualitative attributes. Attribute p value Treatment Mean* Standard dev. Overall accept. 0.42 Control 6.41 1.51 25% ground 6.39 1.39 40% ground 6.16 1.69 Appearance 0 Control 6.81 1.19 25% ground 6.45 1.45 40% ground 6.07 1.50 Flavor 0.26 Control 6.47 1.69 25% ground 6.41 1.60 40% ground 6.08 1.83 Texture 0.47 Control 6.00 1.74 25% ground 6.13 1.61 40% ground 6. 29 1.70 FACT 0.36 Control 5.21 1.68 25% ground 5.12 1.46 40% ground 4.91 1.68 Rank 0.44 Control 145 25% ground 146 40% ground 159 *rank means are panelists rank totals; lower value equals most preferred Table 4 18. Developed c hia mu ffin product separation of means measuring the attribute of overall acceptability when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 6.41 A 2 25% ground 75 6.39 A 3 40% ground 75 6.16 A *means with same letter are not significantly different. Table 4 19. Developed c hia muffin product separation of means measuring the attribute of appearance when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. th an sample 1 Control 75 6.81 A 2 25% ground 75 6.45 AB 3 40% ground 75 6.07 B 1 *means with same letter are not significantly different.

PAGE 92

92 Table 4 20. Developed c hia muffin product means of separation measuring the attribute of flavor when compar ing variation across sample treatments Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 6.47 A 2 25% ground 75 6.41 A 3 40% ground 75 6.08 A *means with same letter are not significantly different. Table 4 21. Developed c hia muf fin product separation of means measuring the attribute of texture when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 6.00 A 2 25% ground 75 6.13 A 3 40% ground 75 6.29 A *means wit h same letter are not significantly different. Table 4 22. Developed chia muffin product separation of means measuring collected FACT scale data when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Co ntrol 75 5.21 A 2 25% ground 75 5.12 A 3 40% ground 75 4.91 A *means with same letter are not significantly different. Table 4 23. Developed chia muffin product separation of means measuring compiled ranked totals when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 145 A 2 25% ground 75 146 A 3 40% ground 75 159 A *means with same letter are not significantly different.

PAGE 93

93 Table 424. Developed chia cookie product sens ory evaluation panel preference test results measuring six qualitative attributes. Attribute p value Treatment Mean* Standard dev. Overall accept. 0 control 6.29 1.73 5% ground/10%whole 6.21 1.57 15% ground/15%whole 5.45 1.74 Appeara nce 0 control 7.00 1.44 5% ground/10%whole 5.67 1.56 15% ground/15%whole 6.00 1.83 Flavor 0.03 control 5.96 2.02 5% ground/10%whole 5.41 1.62 15% ground/15%whole 5.96 1.85 Texture 0 control 6.92 1.50 5% ground/10%whole 5.52 1.71 15% gro und/15%whole 6.45 1.81 FACT 0 control 4.96 1.83 5% ground/10%whole 4.92 1.56 15% ground/15%whole 4.35 1.83 Rank 0 control 122 5% ground/10%whole 148 15% ground/15%whole 180 *rank means are panelists rank totals; lower value equals most preferred Table 4 25. Developed chia cookie product separation of means measuring the attribute of overall acceptance when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 6.29 A 3 2 5% g round/10%whole 75 6.21 A 3 3 15% ground/15%whole 75 5.45 B *means with same letter are not significantly different. Table 4 26. Developed chia cookie product separation of means measuring the attribute of appearance when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 7.00 A 2 3 2 5% ground/10%whole 75 5.67 B 3 15% ground/15%whole 75 6.00 B *means with same letter are not significantly different.

PAGE 94

94 Table 4 27. Developed chia cookie product separation of means measuring the attribute of flavor when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 5.96 A 2 5% ground/10%whole 75 5.41 A 3 15% ground/15%whole 75 5.96 A *means with same letter are not significantly different. Table 4 28. Developed chia cookie product separation of means measuring the attribute of texture when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 6.92 A 3 2 5% ground/10%whole 75 5.52 A 3 3 15% ground/15%whole 75 6.45 B *means with same letter are not significantly different. Table 4 29. Developed chia cookie product separation of means measuring FACT scale collected data when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 4.96 A 3 2 5% ground/10%whole 75 4.92 A 3 3 15% ground/15%whole 75 4.35 B *means with same letter are not significantly different. Table 430. Developed chia cookie product separation of means measuring compiled rank ed total s when comparing variation across sample treatments. Treatment N Mean Tukey Group* Sig. diff. than sample 1 Control 75 122 B 2 5% ground/10%whole 75 148 B 3 15% ground/15%whole 75 180 A 1 2 *means with same letter are not significantly different.

PAGE 95

95 Table 431. Developed chia muffin product measured fatty acid content results. Treatment 18:2* 18:3* 20:1* total fat% control 0.83 0.18 0.08 12.42 25%groun d 1.22 1.47 0.02 14.25 25% whole 0.92 0.32 0.06 12.85 10%whole/15% ground 1.09 0.95 0.02 13.97 *expressed as grams/ 100 grams Table 432. Developed chia cookie product measured fatty acid content results. Treatment 18:2* 18:3* 20:1* total fat% control 0.79 0.11 0.04 21.48 15%ground 1.14 1.35 0.08 22.56 15% whole 0.83 0.23 0.05 21.34 10%whole/5% ground 0.96 0.83 0.05 21.12 Table 433. Developed c hia muffin and chia cookie product measured omega3 and omega6 fatty a cid analysis statistical outcome s. Sample Fatty acid R2 Mean p value Muffin 18:2 0.47 1.01 0.14 18:3 0.94 0.74 <0.0001 Cookie 18:2 0.67 0.93 0.02 18:3 0.95 0.63 <0.0001 Table 434. Developed c hia muffin product separation of means measuring linoleic (18:2) fatty acid content w hen comparing variation across sample treatments. Treatment N Mean Duncan Group* Control 3 0.83 A 25% ground 3 1.21 A combination 3 1.09 A 25% whole 3 0.91 A *means with same letter are not significantly different. Table 435. Developed chia muffin p roduct se paration of means measuring linolenic (18:3) fatty acid content when comparing variation across sample treatments. Treatment N Mean Duncan Group* Control 3 0.22 C 25% ground 3 1.46 A combination 3 0.94 B 25% whole 3 0.32 C *means with same letter are not significantly different.

PAGE 96

96 Table 436. Developed chia cookie product se paration of means measuring linoleic (18:2) fatty acid content when comparing variation across sample treatments means Treatment N Mean Duncan Group* Control 3 0.78 B 1 5% ground 3 1.14 A combination 3 0.95 AB 15% whole 3 0.82 B *means with same letter are not significantly different. Table 4 37. Developed chia cookie product separation of means measuring linolenic (18:3) fatty acid content when comparing variation across sample treatments means Treatment N Mean Duncan Group* Control 3 0.11 C 15% ground 3 1.34 A combination 3 0.83 B 15% whole 3 0.23 C *means with same letter are not significantly different. Figure 4 1. Developed chia m uffin product control sample compression test graphical outcome trial 1, reps 1 3.

PAGE 97

97 Figure 42. Developed chia muffin product control sample compression test graphical outcome, trial 1, rep 4. Figure 4 3. Developed g round seed chia muffin product compression test graphical outcome trial 1, reps 1 3.

PAGE 98

98 Figure 4 4. Developed g round seed chia muffin product compression test graphical outcome trial 1, rep 4. Figure 45. Developed w hole seed chia muffin pr oduct compression test graphical outcome trial 1, reps 1 3

PAGE 99

99 Figure 4 6. Developed whole seed chia muffin product compression test graphical outcome trial 1, rep 4. Figure 4 7. Developed c ombination chia seed muffin product compression test graphical outcome trial 1, reps 1 3.

PAGE 100

100 Figure 4 8. Developed c ombination chia seed muffin product compression test graphical outcome, trial 1, rep 4. Figure 4 9. Developed chia seed cookie product control sample break test graphical outcome trial 1, reps 1 3.

PAGE 101

101 Figure 410. Developed chia seed cookie product control sample break test graphical outcome trial 1, rep 4 Figure 4 11. Developed g round seed chia cookie product break test graphical outcome, trial 1, reps 13.

PAGE 102

102 Figure 412. Developed g round s eed chia cookie product break test graphical outcome, trial 1, rep 4. Figure 4 13. Developed w hole seed cookie product break test graphical outcome, trial 1, reps 1 3.

PAGE 103

103 Figure 4 14. Developed w hole seed chia cookie product break test graphical outc ome trial 1, rep 4 Figure 4 15. Developed c ombination chia seed cookie product break test graphical outcome trial 1, reps 13.

PAGE 104

104 Figure 416. Developed c ombination chia seed cookie product break test graphical outcome trial 1, rep 4.

PAGE 105

105 CHAPTER 5 CONCLUSIONS The link between diet and health is a connection that remains at the forefront of the minds of consumers across the nation. American consumers understand that including the right foods in ones diet can prevent the onset of many diseases such as cancer and heart disease. As the baby boomer generation comes of age, a more knowledgeable and discerning consumer now searches for healthier alternatives. This search has created a great demand for functional foods of all varieties. The consuming public demands products with added health benefits, but not at the cost of flavor and taste. Prior research has even shown that consumers do not mind spending more for a product with proven health benefits. The key to obtaining a share of the functional food m arket is by providing a tasty product that can be easily amalgamated into ones traditional everyday diet. Obtaining this attribute is key because consumers will only eat foods they feel comfortable with and this element translates into repeat buying. From this research, we introduced two new vehicles that can be easily introduced in the consumer diet and fill the need of lacking omega3s. These products call upon the omega3 content of the Salvia hispanica Lam i ac e ae, also known as chia or Spanish sage. On ce a staple in ancient MesoAmerican society, chia has been found to harbor linolenic acid, found in a plant source. Through linolenic acid can be transformed into eicosapentanoic and docosahexanoic fatty acids, respectively. The chosen vehicles for consumer dietary incorporation were muffin and cookie food products. From taste panel testing, minor formulation adjustments are still needed to optimize flavor and texture, but the final product outlook is promising. Physical attribute testing of the two chia products

PAGE 106

106 confirmed that the enhanced baked goods had minimal significant differences from control muffin and cookie samples. Chemical analysis of both the cookie and muffin products confirm s that the essential fatty acid character of the chia seeds is retained during baking. The highest quantity of omega 3 retention was measured in the both ground seed cookie and muffin products. A formulation cost analysis demonstrated that production of t he enhanced baked products is only a few cents more than the nonenh anced control products (Appendix C). Further study of the developed products is still needed to address some persisting issues. The first of these issues is shelf stability, although product water activity was assessed, packaged shelf linolenic acid being confined in the lipid layer of the chia seed can oxidize as all other lipids do which can lead to off aromas and flavors within the products Proper packaging and storage temperatures to retard oxidation are other aspects that have yet to be explored. A final unresolved issue was the mechanism that causes whole chia seeds to have lower omega3 levels than ground chia seeds after baking. W hile most likely due to experimental sample preparation, analytical methods may need to be adjusted due to the extremely hard nature of the chia seed R esolving some of these questions would be the next step in the commercialization of a chia product with consumer health benefits.

PAGE 107

107 APPENDIX A MUFFIN AND COOKIE INGREDIENT MANUFACTUERS allspice McCormick Ground AllspiceHunt Valley, MD almond extract McCormick Pure Almond Extract Hunt Valley, MD baking powder Rumford Baking Double Acting Baking Powder Terre Haute, IN baking sodaArm & Hammer Pure Baking SodaPrinceton, NJ brown sugar Domino Light Brown Sugar Yonkers, NY butter Wholesome Farm Unsalted Sweet Cream Butter Houston, TX chia Greens ChiaVero Beach, FL cinnamonMcCormick Ground CinnamonHunt Valley, MD eggs Publix Eggstirs Lakeland, FL flour Pillsburys Best All Purpose Flour Orville, OH milk Pa r malat UHT Whole Milk Wallington, NJ nutmeg McCormick Ground Nutmeg Hunt Valley, MD raisins Sunmaid Golden Raisins Kingsburg, CA salt Publix Salt Lakeland, FL vanilla extract Publix Pure Vanilla Extract Lakeland, FL vegetable oil Crisco Pur e Canola Oil Orville, OH white sugar Publix Extra Fine Granulated Sugar Lakeland, FL

PAGE 108

108 APPENDIX B MOISTURE LOSS PERCENTAGE Table B 1. Ca lculated Chia muffin product Sample trial 1 trial 2 trial 3 trial 4 trial 5 mean standard dev. Control 8.73 9.30 7.60 7.96 8.57 8.43 0.67 25% Whole 7.53 8.73 7.16 7.56 8.21 7.84 0.62 25% Ground 8.25 8.40 9.17 9.01 9.80 8.93 0.63 10% Whole/ 15% Groun d 8.34 9.28 10.44 10.85 10.04 9.79 1.00 Table B 2. Calculated Chia cookie product Sample trial 1 trial 2 trial 3 trial 4 trial 5 mean standard dev. Control 10.29 9.56 9.01 8.08 9.14 9.22 0.81 15% Whole 9.04 7.22 7.61 7.01 8.32 7.84 0.84 15% Ground 8.80 8.62 9.12 9.69 8.79 9.00 0.42 5% Whole/10%Ground 8.41 10.15 8.34 8.90 9.08 8.98 0.73

PAGE 109

109 APPENDIX C FORMULATION COST ANALYSIS Table C 1. Chia muffin Sample cost per muffin cost per 6 muffin batch Control 0.34 2.01 25% Whole 0.56 3.39 25% Ground 0 .56 3.36 10% Whole/15% Ground 0.56 3.36 Table C 2. Chia cookie Sample cost per cookie cost per 9 cookie batch control 0.07 0.61 15% Whole 0.12 1.08 15% Ground 0.12 1.07 5% Whole/10%Ground 0.12 1.07

PAGE 110

110 LIST OF REFERENCES 2007 Census of Agriculture U nited States Summary and State data vol.1 part 51 American Association of Cereal Chemists. 1983. Approved methods of the American Association of Cereal Chemists. St. Paul, Minn., USA: AACC. method 62 05. Americ an Oil Chemists' Society. 1989. In Official m ethods and recommended practices of the American Oil Chemists' Society. Champaign, Ill: The Society. method Ce 1h Anonymous. 2009. Oilseed Farming in the U.S. : 11112: Ibis World 36. AOAC International. 2000. Latimer GW, Horwitz W. (eds.) Official Methods of AOAC International. Arlington, Va., USA: AOAC International. p 2023. Asp EH. 1999. Factors affecting food decisions made by individual consumers. Food Policy 24: 287294. ASTM, Committee E 18. 1979. ASTM Manual on Consumer Sensory Evaluation, ASTM Spec ial Technical Publication 682, E.E Schaefer, ed American Society for Testing and Materials, Philidelphia, Pa. p 2830. Ayeza R. 1995. Oil content and fatty acid composition of chia (Salvia hispanica L.) from five northwestern locations in Argentina. J. Amer. Oil Chemists Society 72: 10791081. Ayera R, Coates W. 2004. Composition of Chia (Salvia hispanica) grown in six tropical and subtropical ecosystems of South America. Trop Sci 44:131135. Bagozzi, RP.1983 A holistic methodology for model ing consumer r esponse to innovation. Operations Research 31: 128 176. Bech Larsen T, Grunert KG, Poulsen JB. 2001April. The acceptance of functional foods in Denmark, Finland, and the United States. A study of consumers conjoint evaluations of the qualities of functional food and perceptions of general health factors and cultural values: The Aarhus School of Business 25. Available from: Aarhus University, Denmark. Berdan F, Anawalt PR. (eds.) 1996. Codex Mendoza. Berkeley : University of California Press Ltd. (Originally composed ca. 1541 1542) Bleiel J. 2010. Functional foods from the perspective of the consumer: How to make it a Success? Intl Dairy J. 20: 303306. Bower JA, Saadat MA, Whitten C. 2003.Effect of liking, information and consumer characteristics on purchase intention and willingness to pay more for a fat spread with a proven health benefit. Food Quality and Pref. 14: 6574.

PAGE 111

111 linolenic acid to long chain n3 fatty acids in men. Current Opinion in Clinical Nutrition and Metabolic Care. 5:127132. linolenic acid to eicosapentanoic, docosapentanoic, and docosahexanoic acid in young men. Brit J Nutr 88: 411420. Burdge GC, Jones AE, Wooten SA. 2002. Eicosapentanoic and docosapentaenoic linolenic acid metabolism n young men. Brit J Nutr 88: 355363. Cahill JP. 2003. Ethnobotany o f chia, Salvia hispanica L. ( Lam i aceae). Economic Botany 57 (4) 604618. Cahill JP, Provance MC. 2002. Genetics of qualitative traits in domesticated chia (Salvia hispanica L.). J. Hered. 93:5255. Constance J.2008. Food Ingredient Trends Addressing Specific Diseases and other health conditions: Packaged Facts. Rockville Md. 112. Craig R, Sons M. 2004. Application for approval of whole chia (Salvia hispanica L.) seed and ground whole chia as novel food ingredients. Advisory committe e for novel foods and processes. Ireland: Company David Armstrong. p 129. Cross N. 2006. Muffins. In: Hui Y. Han dbook of Food Science, Technology, and Engineering Vol.1. Boca Raton, Fl: CRC Press. p162 168. de Caterina R, Cybulsky MA, Clinton SK, Gimbr one Jr. MA, Libby P. 1995. Omega3 fatty acids and endothelial leukocyte adhesion molecules. Pros t aglandins Leukot Essent Fatty Acids 52: 191195. de Caterina R, Basta G. 2001. n3 Fatty acids and the inflammatory responsebiological background. European Heart J Supplements 3, Suppl D : D42 D49. de la Cruz, Marti nus. 1940. Badianus manuscript. Translated by John Badianus. (Codex Barberini Latin 241) Baltimore : The John Hopkins Press (Originally composed 1552). de Lorgeril M, Renaud S, Mamelle N, Salen P, Mart in JL, Monjaud I, Guidollet J, Touboul P, Delaye J. 1994. Mediterranean alphalinolenic acid rich diet in secondary prevention of coronary heart disease. Lancet 334: 14541459. Diplock AT, Aggett PJ, Ashwell M, Bornet F, Fern EB, Roberfroid MB. 1999. Sci entific concepts of functional foods in Europe: Consensus document. Brit J Nutr 81:1 s1 27. Epling CC. 1940. A Revision of Salvia, Subgenus Calosphace. Verlag des Repertoriums Berlin : Fabeckstr p 49.

PAGE 112

112 Finley J, Shahidi F. 2001. The chemistry, processing, and health benefits of highly unsaturated fatty acids: An overview. In: Finley J, Shahidi F Omega 3 fatty acids Chemistry, Nutrition, and Health effects. Washington, DC: Oxford University Press. p 29. Fishbein M, Ajzen I. 1975. Belief, attitude, intenti on and behavior: an introduction to theory and research. Reading, MA: AddisonWesley. Fuller GW. 2005. New Food Product Development. Boca Raton: CRC Press.p188189. Food Marketing Institute. 2009. U.S. food shopper trends. Gaines CS. 1991. Instrumental measurement of hardness of cookies and crackers. Cereal Foods World. 36: 989,991994, 996. Garg ML, Keelan M, Thomson ABR, Clandin MT. 1992. Desaturation of linolenic acid in the small bowel is increased by short term fasting and by dietary content of lino lenic acid. Biochemica Biophysica Acta 1126:1725. Goldstein LB. 2007 Prevention and management of stroke. In: Libby P, Bonow RO, Mann DL, Zipes DP. Braunwald's Heart Disease: A Textbook of Cardiovascular Medicine. 8th ed. Philadelphia, Pa: Saunders Elsevi er. p 2. Gray L. 2009. Omega Fatty Acids : Trends in the Worl d wide Food and Beverage Markets, 2nd ed: Packaged Facts. Rockville Md. 186. Grigelmo Miguel N, Boladeras EC, MartinBelloso O. 1999. Development of highfruit dietary fibre muffins. Eur Food Res Tec hnol. 210: 123128. Gebaur SK, Psota TL, Harris WS, Kris Etherton PM. 2006. N 3 fatty acid dietary recommendation and food sources to achieve essentiality and cardiovascular benefits. Am J Nutr 83:152635s. Guilford JP. 1954. Psychometric method. New York NY: McGraw Hill. p 120 129. Haque MS, Ghoshal KK. 1980. Karyotypes and chromosome morphology in the genus Salvia L. Cytologia 45: 627 640. Harris JM. 1997. The impact of food product characteristics on consumer purchasing behavior: The case of frankfurters. J Food Distribution Research 28:9297. Haumann F 1991. U.S. efforts focus on a variety of new crops. International News on Fats, Oils and Related Materials 2: 685 690. HealthFocus. 2009. International trend study. Healthfocus Intl., St. Petersburg, F l. www.healthfocus.com.

PAGE 113

113 Ixtaina VY, Nolasco SM, Tomas MC. 2008. Physical properties of chia (Salvia Hispanica L.) seeds. Industrial Crops and products 28: 286293. Jones LV, Thurstone LL. 1955. The psychophysics of semantics: an experimental investigation. J Applied Psych 39: 3136. Kreiter T. 2005. Seed of wellness: Return of a supergrain. Saturday Evening Post 40:106107 Lawless HT, Heymann H. 1998. Sensory evaluation of f ood; principles and practices. New York NY Chapman and Hall. p 450455. Libby P, R idker PM, Maseri A. 2002. Inflammation and atherosclerosis. Circulation 105:11351143. Lin KY, Daniel JR, Whistler RL. 1994. Structure of chia seed polysaccharide exudates. Carbohydrate Polymers 23:1318. LloydJones D, Adams RJ, Brown TM, Carnethon M, Dai S, DeSimone G, Ferguson TB, Ford E, Furie K, Gillespie C, Go A, Greenlund K, Haase N, Hailpern S, Ho PM, Howard V, Kissela B, Kittner S, Lackland D, Lisabeth L, Marelli A, McDermott MM, Meigs J, Mozaffarian D, Mussolino M, Nichol G, Roger V, Rosamond W, Sacco R, Sorlie P, Stafford R,Thom T, Wasserthiel Smoller S, Wong ND, Wylie Rosett. 2010. Heart Disease and Stroke Statistics 2010 Update A Report From the American Heart Association. Circulation. Dallas, TX: American Heart Association. DOI:10.1161/CIRCULA TIONAHA.109.192667.170. Mintel. 2009a. Functional foods U.S. Aug. Mozaffarian, D. 2005. Does alphalinolenic acid intake reduce the risk of coronary heart disease? A review of the evidence. Alternative Therapies ( 3)11: 24 30. Nawar W.1996. Lipids. In: Fennema O. Food Chemistry. 3rd ed. Boca Raton, Fl: CRC Press. p 227228. Nielsen. 2010. U.S. healthy eating trends part 1: commitment trumps the economic pinch. Press release, Jan 26. Nielsen Co., Schaumburg, Ill. www.nielsen.com. Niva M. 2006. Can we predi ct who adopts healthpromoting foods? Users of functional foods in Finland. Scandinavian Journal of Food and Nutrition, 50(1): 13 24. Nelson DL, Cox MM. 2005. Lehninger Principles of Biochemistry 4th ed. New York, NY. WH Freeman and Company. p.343344. Nutr ition Business Journal. 2010. U.S. functional sales slow, but category outpaces overall food sector in 09. Nutr Bus J. Feb 1. www.nutritionbusinessjournal.com.

PAGE 114

114 O Connor EL, White KM. 2010. Willingness to trial functional foods and vitamin supplements: The role of attitudes, subjective norms, and dread of risks. Food Quality and Pref. 21: 7581. Ott RL, Longnecker MT. 2004. A first course in statistical methods. Belmont, Ca. Thomson Brooks/Cole. p 200201. Packaged Facts. 2008. Nutritional supplements in th e U.S. Nov. Packaged Facts, New York, N.Y. www.packagedfacts.com. Patch CS, Williams PG, Tapsell LC. 2005 Attitudes and intentions toward purchasing novel foods enriched wi th omega 3 fatty acids. J Nutr Educ Behav.37:235241 Peiretti PG, Gai F. 2009. Fat ty acid and nutritive quality of chia (Salvia hispanica L.) seeds and plant during growth. Animal Feed Sci and Tech 148:267275. Peryam DR. 1989. Reflections. In sensory evaluation. In celebration of our beginnings. Philadelphia, Pa. American Society for T esting and Materials. p 3130. Peryam DR, Girardot NF. 1952. Advanced tas te test method. Food Eng 24(July):5861,194. Resurreccion AV. 1998. Consumer sensory testing for product development. Gaithersburg, Ga. Aspen Publishers p16. Reyes Caudillo, Tecante A, Valdivia Lopez MA. 2008. Dietary fiber content and antioxidant activity of phenolic compounds present in Mexican chia (Salvia hispanica L.) seeds. Food Chem 107: 656663. Robinett S.2008.Easy sugar cookies. Seattle,WA: All recipes.Available from allr ecipes.com/Recipe/Easy Sugar Cookies/Detail.aspx Accessed March 4, 2009. SauraCalixto F, Go i I 2006. Antioxidant capacity of the Spanish Mediterranean diet. Food Chem 94: 442 447. Schutz HG. 1965. A food action rating scale for measuring food acceptance. J Food Sci 30: 36574. Siddhuraju P, Mohan PS, Becker K. 2002. Studies on the antioxidant activity of Indian Laburnum ICassia fistula L.): A preliminary assessment of crude extracts from stem bark, leaves, flowers and fruit pulp. Food Chem. 79: 6167. Silfverstolpe G, Johnson P, Samsice G, Svanborg A, Gustafson. 1981. Effects induced by two different estrogens on serum individual phospholipids and serum lecithin fatty acid composition. Hormone & Metabolic Research 13: 141145. Sloan AE. 1998. Food indus try forecast: consumer trends to 2020 and beyond. Food Technology 52(1) : 37 44.

PAGE 115

115 Sprecher H. 2000. Metabolism of highly unsaturated n3 and n6 fatty acids. Biochimica et Biophysica Acta. 1486:219231. Tecklenburg E. 2009 Food Flavor s and Ingredients Outlook 2009: Packaged Facts Rockville Md.120. Thompson RA, Isaacs GW. 1967. Porosity determination of grains and seeds with air comparison pycnometer. Trans. ASAE 10: 693 696. Tudoran A, Olsen SO, Dopico DC. 2009. The effect of health benefit information on cons umers health value, attitudes and intentions. Appetite 52: 568579. Urala N, Lhteenmki L. 2003 Strength of health related claims and their perceived advantage. Int J Food Science and Technology 38: 218 826. Urala N, Lhteenmki L. 2007. Consumers changing attitudes toward functional foods Food Quality and Pref. 18: 112. Watkins TR, Bierenbaum ML. 2001. Dietary polyunsaturated fat in cardiovascular disease: Boon or bane? In: Finley J, Shahidi F Omega 3 fatty acids Chemistry, Nutrition, and Health eff ects. Washington, DC: Oxford University Press. p 6676. Vazquez Ovando A, RosadoRubio G, Chel Guerrero L, Betancur Ancona D. 2009. Physicochemical properties of a fibrous fraction from chia (Salvia hispanica L.). LWT Food Science and Technology 42: 1681 77. Verbeke W. 2006. Functional foods: Consumer willingness to compromise on taste for health? Food Quality and Pref. 17: 126131. Vilche C, Gely M, Santalla E. 2003. Physical properties of quinoa seeds. Biosyst. Eng. 86: 59 65. Viles Gonzalez, Fuster V, B adminon JJ.2004. Atherothrombosis: a widespread disease with unpredictable and lifethreatening consequences. European Heart Journal. (14) 25:11971207.

PAGE 116

116 BIOGRAPHICAL SKETCH Devin Lewis was born and raised in Houston, Texas. At a very young age he was intr igued by the sights and sounds of his mother and grandmother toiling in the kitchen. This led him to pursue a B achelor of S cience in culinary arts from Johnson & Wales University, Miami. FL. After graduating Summa Cum Laude in May 2002, Devin began honing his culinary skills while working with several notable chefs. While working in the kitchen, Devin became interested in chemical aspects of food and product development. Devin further explored his new found interest by attending a Research Chefs Association Convention. Shortly thereafter, Devin decided to pursue his m asters degree at the University of Florida, majoring in food science. Graduating in August 2010, Devin hopes to purs ue a career in food chemistry and food product research and development.