ABSTRACT HATHORN, CHELLANI S. Development of a Peanut

ABSTRACT

HATHORN, CHELLANI S. Development of a Peanut-Sweetpotato Ready to Use Therapeutic Food and Evaluation of Peanut Skins in Peanut Products. (Under the direction of Dr. Timothy H. Sanders).

Severe acute malnutrition (SAM) affects 20 million children under the age of five.

Vitamin A deficiency (VAD) is the leading cause of childhood blindness in the world and an estimated 250,000 to 500,000 children become blind every year. Ready to use therapeutic foods (RUTF) are soft enriched foods containing mostly peanuts, milk powder, and vitamin/minerals. RUTFs were developed to eliminate SAM and incorporating β-carotene rich sweetpotatoes in RUTFs may be a viable way of addressing VAD. The main objective of this work was to develop a peanut-sweetpotato RUTF and evaluate nutritional composition, descriptive sensory analysis and consumer acceptance of formulations containing peanuts/sweetpotato flakes (SPF)/milk powder, and other standard ingredients. A secondary objective was to evaluate the potential of peanut skins to improve nutritional properties and enhance the shelf life of peanut products.

Eleven formulations of peanut paste fortified with SPF and milk powder were evaluated for nutritional composition and sensory characteristics. A lexicon incorporating roasted peanut and sweetpotato flavor descriptors was developed. Increased β-carotene and oxidative stability index (ORAC) were observed in the test formulations. In the samples evaluated, β-carotene ranged from 230 to 3870 µg/100g, while ORAC ranged from 4030 to

5025 µMol/100g. The addition of milk powder improved protein quality, resulting in increased concentration of some specific amino acids. From the eleven formulations, two were selected for the evaluation of nutritional composition, flavor using descriptive sensory analysis and consumer acceptance. A commercial RUTF (MANA), and one peanut only formulation were evaluated. The three formulations contained 49/15/20, 28/20/30 and

56/0/30 percentages of peanuts/SPF/milk powder, respectively, and other standard ingredients. β-carotene in 49/15/20 and 28/20/30 was 2040 µg/100g and 3070 µg/100g, respectively. MANA and the 56/0/30 formulation contained 350 µg/100g and 420 µg/100g, respectively. The combination of milk powder with other ingredients used in this study resulted in a more complete amino acid profile than peanuts alone. Formulations containing

SPF were characterized by roast peanutty, sweet aromatic, baked sweetpotato/dried apricot/floral aromatics. The formulation containing 15% SPF was one of the two most overall liked by consumers and overall flavor liking was highest in this formulation.

To potentially add antioxidants to RUTF’s, peanut skins were incorporated into peanut paste and peanut butter. Peanut skins blended up to 20.0% w/w resulted in increased

(P<0.05) total phenolic content of peanut paste from 12.9 to 31.9 mg/GAEg and 14.1 to 28.1 mg GAE/g for peanut butter. Oxygen radical absorbance capacity (ORAC) of peanut paste and peanut butter increased as the concentration of peanut skins increased, and ranged from

4041 to 20063 and 5702 to 20376 µMol Trolox/100g. The addition of low levels of skins did not result in flavor changes but more than 5% peanut skins resulted in a decrease in roast peanut intensity and an increase in woody/hulls/skins, bitter and astringency intensities. Subsequent storage of peanut paste and peanut butter containing skins resulted in a decrease in shelf life inversely related to the concentration of peanut skins added. A study was conducted to identify the cause of reduced shelf life. Oxidative stability index (OSI) of peanut skin oil (1.1-4.1 hrs) was significantly lower (P<0.05) than oil from peanuts (9.8 –

15.7 hrs). Analysis of peanut skins indicated low levels of -tocopherol and high levels of

Cu and Fe. OSI, low -tocopherol, and high copper and iron concentrations in peanut skins suggest a cause for reduced shelf-life of peanut products containing peanut skins.

by Chellani S. Hathorn

A dissertation submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Food Science

Raleigh, North Carolina

2013

APPROVED BY:

______Dr. Timothy H. Sanders, Dr. Lisa O. Dean Committee Chair

______Dr. MaryAnne Drake Dr. Van Den Truong DEDICATION

To my mom and dad

ii BIOGRAPHY

Chellani Hathorn was born and raised in Lexington, MS. Upon graduating from high school,

Chellani went on attend Alcorn State University in Lorman, MS earning a Bachelor of

Science in Agriculture. Chellani received a Master of Science in Food Science from

Tuskegee University in Tuskegee, AL. Later, Chellani started the PhD program in Food

Science at North Carolina State University in Raleigh, NC.

iii ACKNOWLEDGMENTS

First and foremost, I would like to thank my Lord and Savior, Jesus Christ, for this amazing opportunity. I have grown to become a much stronger individual on this jouney, but

I know there is much more to come.

I would like to acknowledge my advisor, Dr. Timothy H. Sanders for giving me the opportunity to work in your lab. Every time I visited your office, you always took time to talk to me and see what I wanted or needed. Thank you for encouraging me both academically and personally. I truly appreciate everything you have done for me. Also I would like to acknowledge my committee members, Dr. Lisa Dean, Dr. MaryAnne Drake and Dr. Den Truong. Dr. Dean, thank you for taking time to answer my questions and helping me solve problems. Dr. Drake, thank you for your suggestions and help with the sensory portion of my work. Dr. Truong, thank you for encouraging me throughout my time in school.

I am forever grateful to the USDA/ARS Market Quality and Handling Research Unit

(MQHRU). Keith, you have such a giving spirit! I truly appreciate all of your help. If I started to list all the ways you have helped, the list would go on. Jim, my fellow

Mississippian, thank you for helping me in the lab whenever I needed it. Sabrina, you have always been extremely kind, helpful and you were always ready to answer my questions.

Michael, thank you for assisting me with roasting peanuts. Thanks to Dr. Jack Davis, Kristen

Price, Dr. Brittany White and Jeff Whaley for their help and assistance. To all of my office

iv mates in room 221 (both past and present), thank you for all the laughs and fun times that we had. I would like to thank Judy Cooper for her kind words of encourgaement. Dr. Leon

Boyd, thanks for encouraging me to apply to NCSU and I appreciate you stopping by the office to check and see how things were progressing even after you had retired.

I would like to acknowledge my parents, Louis Blake and Claranett Hathorn for their prayers, support, sacrifices and always believing in me. Thanks for always being there! To

Shintri and Blake, thanks for being the best sister and brother ever! We always have such a great time laughing and talking whenever we are together, as well as when we are on the phone. Adrian, thank you for always being there, pushing me, praying with me and constantly speaking the word of God. You have truly supported me through this process.

You are awesome! Danielle, Erica, Dee, and Jessica, thank you for encouraging me and always showing the love of God. You all are such wonderful and supportive friends. I would like to thank the Carter family for their continuous support and love.

I’ve been blessed to have many wonderful people in my life. I thank each of you for always loving me and for being there whenever I needed you. I love each of you from the bottom of my heart!

TABLE OF CONTENTS

LIST OF TABLES ……………………………………………………………………….....xi

LIST OF FIGURES ……………………………………………………………………….xiv

CHAPTER 1. LITERATURE REVIEW…………………………………………………...1

Peanuts………………………………………………………………………………………..2

Introduction ………………………………………………………………………...... 2

Peanut maturity ……………………………………………………………………..5

Peanut skins …………………………………………………………………………6

Peanut flavor ………………………………………………………………...... 13

Storage stability and peanut flavor ………………………………………………..18

Bitter taste …………………………………………………………………………20

Transduction of bitter taste ………………………………………………………..21

Bitter - phenolic compounds and tannins …………………………………………22

Astringency ………………………………………………………………………..23

Astringent compounds …………………………………………………………….24

Texture …………………………………………………………………………….25

Texture of peanut products ………………………………………………………..28

Sweetpotato………………………………………………………………………………….30

Introduction ………………………………………………………………………………30

Composition ……………………………………………………………………….30

Harvest, curing and processing ……………………………………………………33

Sweetpotato Flakes ………………………………………………………………..34

Descriptive sensory and consumer acceptance ……………………………………37

Flavor constituents ………………………………………………………………...41

Milk powder ………………………………………………………………………………..44

Introduction ………………………………………………………………………...... 44

Composition ……………………………………………………………………….45

Oxidation …………………………………………………………………………..47

Flavor ………………………………………………………………………...... 47

Flavor as affected by storage ……………………………………………………...48

Ready to Use Therapeutic Foods/Ready to Use Supplementary Foods ...……………....49

Introduction …………………………………………………………………...... 50

RUTF/RUSF ………………………………………………………………………54

Plumpy’nut ………………………………………………………………………..56

MANA …………………………………………………………………………….56

Clinical evidence of using RUTF …………………………………………………57

References ……….………………………………………………………………………….58

CHAPTER 2. FLAVOR AND ANTIOXIDANT CAPACITY OF PEANUT PASTE AND PEANUT BUTTER SUPPLEMENTED WITH PEANUT SKINS ……………...74

vii

Abstract ……………………………………………………………………………………76

Introduction …………………………………………………………………………...... 78

Materials and Methods …………………………………………………………………….81

Color ……………………………………………………………………………….81

Sample Extraction …………………………………………………………………81

Total Phenolics …………………………………………………………………….82

Hydrophilic – ORAC ………………………………………………………………83

Lipophilic – ORAC …………………………………………………………...... 83

Descriptive Sensory Analysis ……………………………………………………...84

Statistical Analysis …………………………………………………………………85

Results and Discussion ………………………………………………………………...... 86

Conclusions …………………………………………………………………...... 91

References …….…………………………………………………………………...... 92

CHAPTER 3. EVALUATION OF PEANUT SKINS: POTENTIAL TO ENHANCE SHELF-LIFE AND OIL CHARACTERIZATION …………………………………….103

Abstract ……………………………………………………………………………………105

Introduction ………………………………………………………………………………..107

Materials and Methods …………………………………………………………………….109

Oxygen Radical Absorbance Capacity (ORAC) ………………………………….109

Oxidative Stability Index ………………………………………………………….110

viii

Peroxide Value ………………………………………………………………...... 110

Fatty Acid Profile ……………………………………………………………...... 111

Tocopherol Analysis ……………………………………………………………...111

Heavy Metal Analysis by ICP ……………………………………………………..112

Statistical Analysis ………………………………………………………………..112

Results and Discussion …………………………………………………………………....113

Conclusion ………………………………………………………………………………....119

References …….……………………………………………………………………………120

CHAPTER 4. SENSORY AND NUTRITIONAL OPTIMIZATION OF A PEANUT SWEETPOTATO READY TO USE THERAPEUTIC TYPE FOOD (RUTF) ...……135

Abstract …………………………………………………………………………………...137

Introduction ……………………………………………………………………………….139

Materials and Methods …………………………………………………………………....141

Color, Moisture, Fat ……………………………………………………………....141

β-carotene analysis ………………………………………………………………..141

ORAC Sample Extraction ………………………………………………………...142

Hydrophilic – ORAC ……………………………………………………………...143

Amino Acids ……………………………………………………………...... 143

Development of a Descriptive Sensory Peanut-Sweetpotato Lexicon ……………144

Descriptive Sensory Analysis ……………………………………………………..144

Statistical Analysis ………………………………………………………………..145

Results and Discussion …………………………………………………………………...146

Conclusion ………………………………………………………………………………..150

References .……………………………………………………………………………….151

CHAPTER 5. NUTRITIONAL COMPOSITION, DESCRIPTIVE SENSORY ANALYSIS, AND CONSUMER ACCEPTANCE OF AN OPTIMIZED THERAPEUTIC READY-TO-USE PEANUT-SWEETPOTATO BASED FOOD………………………………………………………………………………………162

Abstract …………………………………………………………………………………...164

Introduction …………………………………………………………………………...... 166

Materials and Methods …………………………………………………………………....170

Color, Moisture, Fat …………………………………………………………...... 170

Viscosity …………………………………………………………………………..171

β-Carotene Analysis ……………………………………………………………....171

Amino Acid Content ………………………………………………………...... 172

Descriptive Sensory Analysis (DSA) ……………………………………………..172

Consumer Acceptance Testing …………………………………………………....173

Statistical Analysis ………………………………………………………………..174

Results and Discussion ………………………………………………………………...... 175

Conclusion ………………………………………………………………………………...182

References .………………………………………………………………………………..183

LIST OF TABLES

CHAPTER 1.

Table 1. Peanut types and primary growing region ………………………………………...2

Table 2. Tocopherol contents in normal, mid and high-oleic cultivar ……………………...4

Table 3. Fatty acid content of different peanut genotypes ………………………………….4

Table 4. Lexicon of peanut flavors ………………………………………………………..15

Table 5. Peanut aroma model system ……………………………………………………..18

Table 6. Terms used in the texture profile analysis ...……….…………………………….26

Table 7. Proximate composition of sweetpotato …..……………………………………...31

Table 8. Concentration of amino acids in sweetpotato flour …..………………………….32

Table 9. Nutritional facts label for instant sweetpotato flakes ....……………………...... 36

Table 10. β-carotene content of sweetpotato flakes after storage under various packaging conditions ……………………………………………………..……..37

Table 11. Descriptive sensory lexicon for sweetpotatoes ………...……………………….38

Table 12. Sensory attributes of sweetpotato cultivars ……………………………………..40

Table 13. Volatile compounds identified in cooked sweetpotatoes ……………………….43

Table 14. Amino acid composition of milk powder ………………………………….…....46

Table 15. Prevalences of protein energy malnutrition among children under 5 years of age in developing countries, 1995 …………………………………...52

CHAPTER 2.

Table 1. Intensity of attributes of peanut paste reference containing 2% tannic acid …………………………………………………………………….100

Table 2. L and a value of peanut paste containing peanut skins ……………………...... 100

Table 3. Total phenolics (TP) and oxygen radical absorbance a b b capacity (ORAC) of peanut paste containing peanut skins ………………………101 c Table 4. Oxygen radical absorbance capacity (ORAC) of peanut skins …………………...101

Table 5. Descriptive sensory analysis of peanut paste containing peanut skins …………...102

Table 6. Descriptive sensory analysis of peanut butter containing peanut skins ………...... 102

CHAPTER 3.

Table 1. Peanut skins and corresponding peanut designation …………………………...... 124

Table 2. Fatty acid composition of oil from peanut skins and peanuts ………………….....125

Table 3. OSI (110 C) of oil from peanuts ………………………………………………....126

Table 4. Total tocopherol content of oil from peanut skins and peanuts …………………..127

Table 5. Copper and iron content in peanut skins and peanuts ………………………….....128

CHAPTER 4.

Table 1. Peanut-sweetpotato paste formulations …………………………………………..155

Table 2. Color, moisture and fat content of peanut-sweetpotato formulations ……………156

Table 3. β-carotene and ORAC of peanut-sweetpotato formulations ……………………..157

xii

Table 4. Total amino acid content (g/100g) of peanut-sweetpotato formulations ………...159

Table 5. Descriptors and definition of peanut-sweetpotato lexicon …………………….....161

Table 6. Descriptive sensory analysis of peanut-sweetpotato formulations …………….....161

CHAPTER 5.

Table 1. Percentages of ingredients used in optimized peanut-sweetpotato formulations ………………………………………………………………………188

Table 2. Color, moisture, fat, viscosity and β-carotene of commercial and formulated RUTFs ...…………………………………………………………………...... 189

Table 3. Amino acid content (g/100) in RUTF ………………………………………….....190

Table 4. Descriptors and definition of peanut-sweetpotato lexicon …………………….....191

Table 5. Descriptive sensory analysis of MANA and RUTF formulations ………………..192

Table 6. Demographic information and consumer characteristics for RUTF consumer studies ……………………………………………………………….....193

Table 7. Consumer consumption interest for RUTF …………………………………….....194

Table 8. Liking attribute means from consumer acceptance testing of RUTF ………….....195

Table 9. Consumer just about right (JAR) scores for RUTF ……………………….……...196

Table 10. Overall liking means for cluster segmentation ……………………………….....197

xiii

LIST OF FIGURES

CHAPTER 1.

Figure 1. Steps of autoxidation: initiation, propagation, and termination …………………...9

Figure 2. General structure of flavonoid ………………………………………………...….11

Figure 3. Comparative aroma profile of roasted peanut paste (black and yellow maturity class) and optimized aroma model …………...………………………...17

Figure 4. Proposed mechanism for PRP-polyphenol binding and subsequent protein aggregation and complex formation …………………………………..…25

Figure 5. Generalized instrumental texture profile curve obtained with the General Foods Texturometer …………………………………………………….27

Figure 6. Structure of β-carotene ..………………………………………………………….33

Figure 7. Causes of death among child under 5 years of age, 2000-2003, worldwide ………………………………………………………………………..51

Figure 8. Direct and indirect causes of malnutrition …..………..…………………………53

Figure 9. F-75 and F-100 therapeutic milk in dehydrated form …………………………...55

CHAPTER 3.

Figure 1. Hydrophilic ORAC of peanut paste containing peanut skins stored at 30° C ..…131

Figure 2. Hydrophilic ORAC of peanut butter containing peanut skins stored at 30° C...... 132

Figure 3. Oxidative stability index (110°C) of peanut paste containing peanut skins stored at 30°C ……………………………………………………………………133

Figure 4. Oxidative stability index (110 °C) of peanut butter containing peanut skins stored at 30°C ……………………………………………………………………134

xiv

Figure 5. Peroxide value (meq/kg) of peanut paste containing peanut skins stored at 30°C …………………………………………………………………………..135

Figure 6. Peroxide value (meq/kg) of peanut butter containing peanut skins stored at 30°C …………………………………………………………………………..136

CHAPTER 5.

Figure 1. Partial least squares correlation biplot of clusters for RUTF ……………………198

APPENDICES

Appendix 1……………………………………………………………………………….....200

xvi

CHAPTER 1

LITERATURE REVIEW

PEANUT

Introduction

Peanuts [Arachis hypogea L.] are an important agricultural commodity that originated in South America. They are grown in many different regions including Asia, Africa, North and South America, and are commonly referred to as groundnuts or groundpeas in many parts of the world. The majority of peanuts grown in the United States (USA) are produced in the southern region. Georgia is the leading peanut producing state, followed by Texas,

Alabama, North Carolina, Florida, Virginia and Oklahoma. Peanuts are categorized into four market types and they are further separated by growing region and seed size (Table 1). The shell, skin (testa or seedcoat), and seed together makes up the peanut (Woodroof 1983).

Table 1. Peanut types and primary growing region

Types sp. ssp. Growing region Size runner Arachis hypogaea L. hypogaea Georgia, Alabama, Florida Medium virginia A. hypogaea L. hypogaea Virginia, North Carolina Large spanish A. hypogaea L. fastigiata Texas, Oklahoma Small valencia A. hypogaea L. Fastigiata New Mexico Intermediate

In the U.S., about 50% of the peanuts are processed into peanut butter and result in more than $850 million dollars in annual sales (American Peanut Council 2008). Peanut

butter, by standard of identity, contains a minimum of 90% peanuts (CFR 2008a). Products which do not meet the standard of identity for peanut butter are labeled as peanut spread or imitation peanut butter (McWatters and Young 1978). Due to advanced processing and packaging, peanut butter and similar products can be stored up to 5 years under optimum storage conditions (Shewfelt and Young 1977). Peanuts are also processed for snack nuts and confectionary products. About 15% of all peanuts grown in the U.S. are crushed for oil

(American Peanut Council 2008). Lipids in peanuts account for 40-50% of the dry weight and 80% of the fatty acids in peanuts are unsaturated (Hoffpauir 1953; Eheart et al. 1955;

Özcan and Seven 2003). The fatty acid composition of peanuts is about 48% oleic and 31% linoleic acid (Cobb and Johnson 1973) except for high oleic varieties which may contain up to 80% oleic acid (Norden et al. 1987). Genotype as well as production environment affect fatty acid composition (Table 3). Peanuts contain , β, , and δ tocopherols, a class of compounds with vitamin E activity (Table 2). The protein content of peanuts ranges between

21-36% (Hoffpauir 1953). Peanuts contain all essential amino acids, but lack adequate amounts of lysine, methionine and threonine (Pominski et al. 1991).

Table 2. Tocopherol contents in normal, mid and high-oleic cultivar (mg per 100g) Shin et al. (2009) Type -T β-T -T δ-T Total

Normal-oleic 10.9±1.4 b 0.33±0.14 a 10.4±2.4 b 0.85±0.40 a 22.4±3.3 a

Mid-oleic 11.7±1.0 a 0.35±0.07 a 11.2±0.8 ab 0.68±0.10 ab 2.39±1.4 a

High-oleic 9.8±1.3 c 0.26±0.07 b 11.7±1.4 a 0.66±0.08 b 22.4±1.7 a

Data represent the mean±SD of each sample assayed in triplicate. Means with different letter indicate significant differences.

Table 3. Fatty acid content of different peanut genotypes (Andersen et al. 1998; Isleib et al. 2006) Genotype 16:0 18:0 18:1 18:2 20:0 20:1 22:0 24:0 Florunner 10.40 2.33 52.80 27.10 1.35 1.28 3.10 1.68 SunOleic 95R 6.55 2.63 79.00 4.70 1.33 1.58 2.60 1.63 GA 2844 9.42 2.94 54.20 26.40 1.42 1.08 2.92 1.62 TX 896100 10.30 2.98 55.40 24.50 1.55 1.05 2.78 1.50 UF 91108 9.10 2.98 60.40 19.30 1.62 1.18 3.90 1.52 NC 12C 10.10 3.58 56.70 24.20 1.50 0.75 2.23 1.08 F 1334 5.63 2.18 81.20 3.46 1.20 1.94 2.77 1.86 N00090ol 5.53 4.05 80.46 2.48 1.79 1.35 2.76 1.37 NC 7 8.97 3.47 58.09 22.85 1.57 1.02 2.67 1.27 16:0 – Palmitic acid; 18:0 –Stearic acid; 18:1 – Oleic acid; 18:2 – Linoleic acid; 20:0 – Arachidic acid; 20:1 –Eicosenoic acid; 22:0 –Behenic acid; 24:0 –Lignoceric acid Peanut Maturity

The pod maturity profile commonly referred to as the hull-scrape method (Williams and Drexler 1981) is the preferred method for predicting and determining optimum peanut harvest date. This method is based on progressive changes in mesocarp color. A small portion of the exocarp of the pod is scraped off exposing the colors of the mesocarp, which

progressively range from white (immature) to black (mature) and have been divided into maturity classes corresponding to the colors: white, yellow, orange A, orange B, brown and black (Williams and Drexler 1981; Sanders et al. 1982).

Environmental conditions such as weather, disease, soil type, and crop rotation may affect the maturation of peanuts and thus the optimum harvest time (Sanders and Bett 1995).

Peanut maturity affects grade, yield, size, composition, quality and flavor (Sanders et al.

1989a; Sanders et al. 1989b; McNeil and Sanders 1998). The peanut plant has an indeterminate flowering pattern that results in a wide range of maturities on a plant at the time of harvest. The harvest of this range of maturity classes with inherent variability in seed sizes results in a maturity distribution within each shelled seed size lot which affects quality and flavor of peanut and peanut products (McNeil and Sanders 1998). Young et al. (1973) found that amino acid variations observed in peanuts were related to maturity and specifically, arginine and proline were present in higher concentrations in immature peanuts.

Young and Mason (1972) developed the arginine maturity index for peanut harvest date prediction which received some attention before development of the pod maturity index

(Williams and Drexler 1981) was developed. Arginine has been associated with off-flavors during roasting in peanuts (Young and Mason 1972; Young et al. 1973) and that corresponds with studies by Sanders et al. (1989a) and Sanders et al. (1989b) on the potential for low roasted flavor and higher incidence of off-flavors in immature peanuts. Alanine, isoleucine, phenylalanine and threonine decreased in peanuts as they matured, suggesting that these amino acids may be useful as an indicator of peanut maturity (Basha et al. 1980).

Pattee and Purcell (1967) reported that β-carotene and lutein were highest in immature peanuts; therefore, the oil from immature peanuts is normally lighter in color and highly correlated with peanut maturity (Holley and Young 1963). Pattee et al. (1969) suggested that a decrease in peanut oil color is due to a decrease in carotenoids as peanuts mature and this result in less yellow color in the oil. Some research suggest that oil color differences among peanuts may be a result of curing (Holley and Young 1963).

Peanut Skins

The skin (seed coat or testae) from peanuts is the thin paper-like layer that surrounds the seed. The epidermis or sclerenchyma, the middle parenchyma and the inner parenchyma make up the three unicellular layers of peanut skins (Rao and Murty 1994). Large quantities of peanut skins are produced from blanching and roasting, and are considered as waste products. Removal of the skin is done in preparation for the production of products such as peanut butter. Peanut skins are often dumped as waste but some are utilized as feed additives because they are inexpensive and are a source of fat and protein (Hill 2002). Use of peanut skins for animal feed is limited by the high tannin content (Sanders 1979), because tannins react with protein and form protein-tannin complexes. Tannins inhibit digestion by binding with protein and interfering with the absorption of nutrients. Approximately 60,000 tons of peanut skins are accumulated annually in the U.S. as a result of peanut processing.

Peanut skins contain various compounds with antioxidant activity. Antioxidants are a class of compounds that may inhibit chain peroxidation reactions. There has been increasing

interest in natural and synthetic antioxidants possibly being added to food products or extracts used for nutraceutical supplements. Antioxidants when added to lipid based foods may aid in reducing the formation of various off-flavors and other compounds that result from the oxidation of lipids. Butylated hydroxyanisole (BHA) and butylated hydroxytoluene

(BHT) are synthetic antioxidants which have been used widely as natural preservatives in foods (Velioglu et al. 1998). Currently, natural antioxidants are of interest to researchers interested in protecting and maintaining food quality. Houlihan et al. (1985) described the criteria of natural antioxidants for use in food. Antioxidants:

(a) should develop no flavor or colors when added at the prescribed concentration in the

product

(b) should have the desired solubility in the fat substrate to assure effectiveness

(d) should have carry-through properties

(e) must be cost effective

According to Boskou and Elmada (1999) antioxidants should:

(a) compete effectively with the substrate for reactive intermediates

(b) be readily repaired by the biological system. If the antioxidant is destroyed irreparably at every encounter with the oxidizing species then the system will rapidly reach a state where it is no longer protected.

(c) be accessible to the reactive intermediate in the microenvironment, i.e., the antioxidant has to be located in the same environment as the oxidizing radical species.

(d) be relatively unreactive to the substrate. The products should not be toxic to the system and must not take part in the reaction.

(e) be catalytic in its quenching mechanism.

Lipids are components of many foods and peanuts are susceptible to oxidation, since the majority of the lipids are unsaturated. Oxidation is catalyzed by light, heat, enzymes and metals leading to autoxidation, photooxidation, thermal or enzymatic oxidation (Shahidi and

Zhong 2010). Autoxidation is the reaction that occurs in the presence of atmospheric oxygen through a chain reaction of free radicals forming peroxides and hydroperoxides.

Autoxidation is generally responsible for oxidative off-flavors in oils. Autoxidation occurs in three steps: 1) initiation – formation of free radicals; 2) propagation –free-radical chain reaction; and 3) termination –formation of nonradical products (Figure 1). During the initiation step, hydrogen is removed from a fatty acid at the carbon atom next to the double bond producing a free radical. The free radical is formed and will combine with available oxygen to form a peroxyl-free radical which can in turn remove hydrogen from another unsaturated molecule forming a free radical, thus beginning the propagation step.

Figure 1. Steps of autoxidation: initiation, propagation, and termination (Shahidi 2000)

Propagation is the chain reaction step, which occurs until there is no available oxygen. Lipid hydroperoxides are produced during propagation and secondary oxidation

products such as aldehydes, ketones, alcohols, hydrocarbons, volatile organic acids and epoxy compounds are formed from the hydroperoxides (Shahidi and Zhong 2010). Lipid oxidation generally results in secondary products are associated with undesirable odors and flavors (Shahidi and Zhong 2010). Termination, the last stage of oxidation, occurs when free radicals start to react with one another, which completes the cycle.

Phenolic compounds are effective antioxidants and possess biological properties such as anti-carcinogenic, anti-mutagenic, anti-inflammatory and anti-microbial properties (Huang and Ferraro 1992, Puupponen-Pimiä et al. 2001, Rimando et al. 2002, Castilla et al. 2006,

Wang et al. 2006). Phenolic compounds are divided into two categories: flavonoids and nonflavonoids. Flavonoids have a 15 carbon backbone with two benzene rings (A and B) joined by a three carbon chain (Figure 2). Flavonoids make up the largest class of polyphenolic compounds and can be divided into several subclasses including flavanols, flavanones, flavones, anthocyanidins and flavonols. The subclasses of flavonoids are based on the connection of the B ring to the C ring, including the oxidation state and functional groups of the C ring (Beecher 2003). The effectiveness of antioxidants depends largely upon their structure. The antioxidant activity and potency of phenolic compounds are affected by the location and number of hydroxyl groups (Porto et al. 2003). The class of phenols known as nonflavonoid includes stilibenes (resveratrol) and phenolic acids (benzoic acid and caffeic acid).

Figure 2. General structure of flavonoid

Tannins are present in the seed of peanuts, but higher concentrations are present in the skin (Sanders 1977). Defatted peanut skins contain approximately 150 mg of total polyphenols per gram (Nepote et al. 2002). Proanthocyanidins, also known as condensed tannins, are polymers of flavonoids such as catechins. These are powerful antioxidant agents and may be more effective than antioxidants, such as vitamins C and E (Porto et al. 2003;

Beecher 2003). Proanthocyanidins have been identified in mature red peanuts by Karchesy and Hemingway (1986). Lou et al. (1999) reported six proanthocyanidins from the water- soluble fraction of peanut skins. Lou et al. (2001) also isolated eight new flavonoids and two alkaloids from peanut skins. Yu et al. (2005) reported three classes of compounds in peanut skins. These compounds include flavonoids (epigallocatechin and epicatechin), phenolic acids (chlorogenic and caffeic acids), and stilbenes (resveratrol). Sanders et al. (2000) reported that the skins from runner and virginia type peanuts contain approximately 0.65

µg/g resveratrol.

Nepote et al. (2005) used a solvent extraction method to obtain antioxidant compounds from peanut skins. The researchers found that 70% ethanol in water was the best solvent ratio for extracting the antioxidant compounds from peanut skins. In another study,

80% ethanol was used to extract phenolics from the skin of peanuts (Yu et al. 2006) and compounds with antioxidant activity included epigallocatechin, epicatechin, chlorogenic acids, caffeic acids and stilbenes. Seung et al. (2006) reported that the use of methanol to extract phenolic compounds from peanuts resulted in higher total phenolics than other solvents used.

The oxygen radical absorbance capacity (ORAC) assay is a method to determine the inhibition of peroxyl radical induced oxidation in foods and biological materials (Prior et al.

2003; Prior et al. 2005; Karadag et al. 2009). One of the advantages of the ORAC assay is that it can be used to determine both hydrophilic and lipophilic activity (Wu et al. 2004).

ORAC specifically measures peroxyl radical quenching of the fluorescence of fluorescein.

The antioxidant activity is quantified using the area under the fluorescence decay curve. The standard curve is generated from Trolox, a water-soluble analogue of vitamin E and ORAC is reported as micromoles of Trolox equivalents (TE) per gram or 100 g of sample. Ballard et al. (2009) reported that antioxidant activity using ORAC was 214,900 µMol TE /100g for peanut skins. Davis et al. (2010) reported that ORAC ranged from 154,440 to 216,030 µMol

TE / 100 g for peanut skins that were subjected to increased roasting.

Peanut skin extracts have been used in model food systems to provide antioxidant activity. Nepote et al. (2004) determined the shelf-life of honey roasted peanuts (HRP) with

and without peanut skin extracts. At 42 days of storage, HRP without peanut skin extract had higher peroxide values than HRP with peanut skin. Natural antioxidant extracts from peanut skins provided protection against lipid oxidation during storage, while HRP-butylated hydroxytolene (BHT) provided more. Sobolev and Cole (1999) demonstrated potential uses of peanut skins that included extracted oil, alcoholic and non-alcoholic beverages. O’Keefe and Wang (2006) examined phenolic compounds extracted from peanut skins as natural antioxidants in beef products and concluded that 200-400 ppm was the optimal concentration of extract to significantly reduce lipid oxidation and extend storage stability.

Peanut Flavor

Descriptive sensory analysis (DSA) is a comprehensive sensory tool used to generate quantitative and qualitative data. DSA utilizes a panel that is selected through screening and extensive training to develop panel abilities to identify and quantify attributes of foods (Drake 2007). Some DSA methods include Spectrum™, Flavor Profile®,

Quantitative Descriptive Analysis (QDA) and Texture Profile Analysis (TPA) (Murray et al.

2001). These methods all encompass the development of a language which is used to describe the samples in detail. Oupadissakoon and Young (1984) first developed a simple lexicon for peanuts. Later, Johnsen et al. (1988) developed a complete lexicon that described the flavor attributes of roasted peanuts (Table 4). Sanders et al. (1989b) later added the term fruity fermented to the lexicon to describe flavors associated with high temperature cured peanuts.

Amino acids and sugars have been considered as flavor precursors in roasted peanuts

(Newell et al. 1967; Mason et al. 1969). Newell et al. (1967) observed the changes in amino acid content during roasting of mature and immature peanuts, and suggested that aspartic acid, asparagine, glutamine, glutamic acid, phenylaline and histidine were precursors to roasted peanut flavor. Amino acids threonine, tyrosine and lysine are not major flavor precursor (Newell et al. 1967). Chung et al. (1994) used peptide modeling to distinguish between mature and immature peanut proteins. The authors reported that the proteins in mature and immature peanuts are structurally different. Oupadissakoon and Young (1984) used maximum R2 (MAXR) stepwise mathematical modeling using amino acid and sugar data as dependent variables to predict critical laboratory evaluation roast (CLER) score and roast peanut flavor intensity. A 10-variable model using raw peanut data had the best fit

(R2=0.762).

Fructose, glucose, sucrose and inositol were reported as precursors of flavor components in roasted peanuts (Newell et al. 1967; Oupadissakoon and Young 1984).

Hydrolysis of sucrose by invertase activity resulted in an increase in fructose and glucose in roasted peanuts (Newell et al. 1967; Mason et al. 1969); suggesting that fructose and glucose take part in the browning reactions occurring during roasting.

Table 4. Lexicon of peanut flavors developed by Johnsen et al. (1988)

Descriptor Definition Aromatics Roast Peanutty The aromatic associated with medium-roast peanuts (about 3-4 USDA color chips) and having fragrant character such as methylpyrazine Raw Beany The aromatic associated with light-roast peanuts (about 1- 2 on USDA color chips) having legume-like character (specify beans or pea if possible) Dark Roasted Peanut The aromatic associated with dark roast peanuts (4+ on USDA color chips) and having very brown or toasted character Sweet Aromatic The aromatics associated with sweet material such as caramel, vanilla, molasses, fruit (specify type) Woody/Hulls/Skins The aromatics associated with base peanut character (absence of fragrant top notes) and related to dry wood, peanut hulls and skins Cardboardy The aromatic associated with somewhat oxidized fats and oils and reminiscent of cardboard Painty The aromatic associated linseed oil, or oil based paint Burnt The aromatic associated with very dark roast, burnt starches, and carbohydrates (burnt toast or expresso coffee) Green The aromatic associated with uncooked vegetable/grasstwigs, cis-3-hexanal Earthy The aromatic associated with wet dirt and mulch Grainy The aromatic associated with raw grain (bran, starch, corn, sorghum) Fishy The aromatic associated with trimethylamine, cod liver oil or old fish Chemical/Plastic The aromatic associated with plastic and burnt plastics Skunky/Mercaptan The aromatic associated with sulfur compounds, such as mercaptan, which exhibit skunk-like character Tastes Sweet The taste on the tongue associated with sugars Sour The taste on the tongue associated with acids Salty The taste on the tongue associated with sodium ions Bitter The taste on the tongue associated with bitter agents such as caffeine or quinine Chemical Astringent The chemical feeling factor on the tongue, described as Feelings puckering/dry and associated with tannins or alum Metallic The chemical feeling factor on the tongue described as flat, metallic and associated with iron and copper

Peanut flavor results from a complex mixture of volatile compounds. Literature has long ascribed peanut flavor to pyrazines; however, recent studies by Chetschik et al. (2010) and Neta (2010) show that pyrazines are not a component of peanut flavor although they do increase during roasting. Chetschik et al. (2010) quantified odorants in raw and pan roasted peanut meal. Stable isotope dilution assay (SIDA), high-resolution gas chromatography- mass spectrometry (HRGC-MS), two-dimensional gas chromatography-mass spectrometry

(2D-HRGC-MS), odor activity values (OAVs) and aroma recombination experiments were used to characterize peanut flavor. Peanut flavor models without pyrazines and with pyrazines (2,5-dimethylpyrazine, methylpyrazine, trimethylpyrazine, 2-ethyl-6- methylpyrazine, 2,6-dimethylpyrazine, 2-ethyl-5-methylpyrazine, 2,3-dimethylpyrazine, and

2-ethyl-3-methylpyrazine) evaluated by a sensory panel concluded that there were no differences between the models. These results indicate that pyrazines do not contribute to the overall roasted peanut aroma. Neta (2010) identified methional, 1-octen-3-one, nonanal, hexanal, octanal, 2-acetyl-1-pyrroline, carbon disulfide, and phenylacetaldehyde as major contributors of roasted peanut aroma. Similarly, Chetschik et al. (2010) reported hexanal, 2- acetyl-1-pyrroline and phenylacetaldehyde as key aroma compounds in roasted peanuts.

Neta (2010) used a series of flavor and flavor compound identification studies to identify major contributors to peanut flavor. Aroma extract dilution analysis (AEDA), headspace solid-phase microextraction (HS-SPME), and gas chromatography-olfactometry

(GC-O) were methods used in determining roasted peanut flavor volatiles. Neta (2010)

compared an optimized peanut aroma model system to normal roasted peanut and found a similarity rating of 8 on a 10 point scale (Figure 3 and Table 5).

Figure 3. Comparative aroma profile of roasted peanut paste (black and yellow maturity class) and optimized peanut aroma model (Neta 2010)

Table 5. Optimized peanut aroma model system (Neta 2010)

Compounds Concentration (ppb) 2,3-pentanedione 200 2-acetyl-1-pyrroline 3,000 2-methyoxy-4-vinylphenol 0 Phenylacetaldehyde 25,000 2-methylbutanal 800 carbon disulfide 100 dimethyl trisulfide 100 Hexanal 1,200 Maltol 24,500 Nonanal 1,000 Octanal 1,000 2,5-dimethylpyrazine 6,000 2-ethyl-3-methylpyrazine 100 2-ethyl-3,5-dimethylpyrazine 1,000 acetic acid 0 1-octen-3-one 100 pentanoic acid 0 Methional 100 2,3-diethyl-5-methylpyrazine 100

Storage Stability and Peanut Flavor

Off-flavors and flavor-fade in peanuts have been associated with storage of roasted peanuts (Bett et al. 1994; Braddock et al. 1995; Warner et al. 1996; Lee 2001). Flavor fade is known as a loss of flavor attributes associated with fresh-roasted peanuts, usually occurring during storage (Warner et al. 1996; Abegaz et al. 2004).

Reed et al. (2002) studied the effect of storage and water activity on flavor fade in high oleic and normal oleic peanuts. Water activity during storage of peanuts did have an effect on flavor attributes. The loss of roasted peanutty flavor was lowest in high oleic

peanuts with aw 0.60, followed by high oleic peanuts with aw 0.19, followed by normal oleic peanuts with aw 0.60 and normal oleic peanuts with aw 0.19. Normal oleic peanuts were higher in painty intensity than high oleic peanuts. Increased moisture systems have been studied in peanut butter products (Felland and Koehler 1997) and results indicated that peanut butter with increased moisture had decreased roasted peanut flavor. Abegaz et al.

(2004) reported the role of moisture on flavor changes in peanut butter.

Baker et al. (2002) measured peroxide values, moisture content and sensory attributes during 14 weeks of storage in peanuts stored with different aw. Peroxide values (PV) increased over time for all treatments, while the highest PVs were observed in peanuts held at aw 0.67. Lipid oxidation is influenced by water activity, and controlling this factor can reduce the extent of oxidation occurring in peanuts and peanut products. Roasted peanut flavor decreased with time and the decrease was greater at higher aw.

High oleic acid peanuts have improved oil stability due to lower concentration of polyunsaturated fatty acids. According to Braddock et al. (1995), high oleic peanuts stored at

25°C for 74 days had lower painty and cardboardy intensities than normal oleic peanuts.

After 74 days of storage, hexanal, a secondary oxidation product was twice as high in normal oleic peanuts as in high oleic peanuts. Pattee et al. (1999) studied low temperature (-23°C) effects on peanut paste over long term storage and concluded lipid oxidation can occur even at low temperature.

Williams et al. (2006) evaluated flavor fade in roasted peanuts during short term storage. Roast peanutty intensity decreased significantly after 21 days. In addition, there

were decreases in sweet taste and significant increases in bitter, and cardboardy intensity and hexanal. After 14 and 21 days, peroxide values were 5.0 and 6.0 meq/kg, respectively.

BITTER TASTE

There are five basic tastes: salty, sour, sweet, bitter and umami. Bitter taste is stimulated by quinine, caffeine and hops and is elicited at the back of the tongue (Robichaud and Noble 1990; Meilgaard et al. 1999). Phenolic compounds such as catechins, gallic acid, and tannins have been described as having a bitter taste. Other bitter-tasting compounds include amino acids, peptides, sulfimides, ureas, thioureas (6-n-propyl-thiouracil), phenylthiocarbamide (PTC), esters, lactones, and terpenoids. However, the mechanism of bitter perception is complex and poorly understood. The chemical compounds that are perceived as bitter are diverse and many of the chemical structures are not similar

(Drewnowski et al. 2001).

Bitterness, like sweetness, is sensed by G protein coupled receptors (GPCRs) coupled to the G protein gustducin (Kinnamon and Cumming 1992; Margolskee 1993) in the apical membranes of taste receptor cells (TRCs) (Margolskee 2002). G proteins are heterotrimeric proteins composed of α, β, and γ subunits (Spielman 1998). Bitter, like other basic tastes is mediated by taste receptor proteins residing on the surfaces of taste receptor cells (TRCs) within the taste buds of the tongue (Breslin 2001; Drewnowski 2001). Data suggest that

T2Rs or TRBs receptors, described as the second family of taste receptors have only been found in TRCs positive for the expression of gustducin (Margolskee 2002).

Stimuli for a single taste may come from several different types of chemicals. In the case of bitterness, for example, caffeine, morphine, and potassium chloride are all bitter. The first step in taste recognition takes place in the taste pore, where molecules that are perceived enter the taste pore and interact with receptor molecules and channels within the microvillar membrane of the TRCs (McLaughlin and Margolskee 1994). The neurons make contact with taste cells at the synapse, a specialized region between the receiving end of the neuron, and the sending end of the taste cell. Information is then passed from the TRC to the neuron via chemical transmitters called neuro-transmitters secreted by the taste cell into the synapse.

When the neurons detect these transmitters they react to them with a nerve impulse that is transmitted to the brain (McLaughlin and Margolskee 1994). This process of receiving sensory information that is translated into a useful signal to the nervous system is called sensory transduction (McLaughlin and Margolskee 1994).

Transduction of Bitter Taste

Data suggest that different mechanisms are utilized in the transduction of different taste stimuli (Kinnamon and Margolskee 1996). Some findings suggest that G-proteins (a family of proteins involved in transmitting chemicals) may stimulate enzyme-activated inositol triphosphate (IP3), which leads to the release of calcium from cell reserves. Another mechanism proposed for the perception of bitter and sweet compounds involves G-protein, α- gustducin, that activates enzyme phosphodiesterase to decrease intracellular AMP (cAMP)

(Spielman 1998). G protein α subunit, α-gusducin is associated with the gustatory system and with the cell-surface receptors (McLaughlin et al. 1992).

Both ion-channel events and receptor-mediated transduction mechanisms are involved in bitter taste chemoreception in species other than man (Kinnamon and Cummings

1992). Ionic taste stimuli such as salts, acids and some bitter compounds interact directly with apically located ion channels to depolarize taste cells. Amino acids, sweet stimuli and other bitter compounds bind to specific membrane receptors usually coupled to G-proteins and secondary messenger systems. Many people find bitter taste to be unpleasant.

Evolutionary biologists suggest that a dislike for bitter foods may have evolved because it enabled people to avoid toxic chemicals. For this, it may be appropriate to suggest that bitter receptors are encoded by a large family of genes that have evolved to provide recognition of a wide and diverse range of chemicals (Maehashi and Huang 2009).

Bitter – Phenolic Compounds and Tannins

Phenolic compounds are found widely in nature and possess antioxidant activity due to their redox properties (Kãhkõnen et al. 1999; Pereira et al. 2009). These hydrogen- donating antioxidants can react with reactive oxygen and reactive nitrogen species in a termination reaction, which breaks the cycle of generation of new radicals. Bate-Smith and

Swain (1962) defined tannins as having molecular weights between 500 and 3000 Da.

However low-molecular-weight phenolic compounds tend to be bitter, while higher- molecular-weight phenolic compounds are more likely to be astringent (Noble 1994).

Astringency

Astringency is an oral sensation resulting in the binding and precipitation of salivary proteins by polyphenols (Gawel 1998). Astringent compounds precipitate proteins, and are perceived as a dry, rough sensation or shrinking, drawing or puckering of the epithelium

(ASTM 1995). Lee and Lawless (1991) generated terms that describe astringent characteristics as ‘drying’, ‘puckering’, and ‘rough’. The mechanism behind astringency is not fully understood. However, astringency is produced by a variety of stimuli, including tannins and other polyphenols (Gawel 1998), tannin substituents such as catechins

(Thorngate and Noble 1995), acids (Rubico and McDaniel 1992; Hartwig and McDaniel

1995; Lawless et al. 1996) and aluminum salts (Lee and Lawless 1991).

Guinard et al. (1986) investigated the time-intensity of astringency in wines containing moderate amounts of phenols (200 to 800 mg/L). Observations from this study indicated that repeated exposure resulted in an increase in maximum intensity. Lee and

Lawless (1991) evaluated the time-intensity of varying concentrations of tannic acid and tartaric acid, and reported a significant decrease in sensation over time. While some acids are considered to be astringent, the low pH of these acids may be the cause of astringency

(Lawless et al. 1996). Beecher et al. (2008) found that whey protein beverages with a lower pH had a higher perceived astringency than those with a higher pH. The authors proposed that astringency was associated with positively charged whey proteins and negatively charged saliva proteins.

Astringent Compounds

Saliva is a water-like substance consisting of 98% water, some enzymes, electrolytes, mucus and antibacterial compounds. The salivary glands, where saliva is produced, keep the mouth and other parts of the digestive system moist. One of the functions of saliva is to break down food in the mouth. There are hypotheses which allude to the interaction between polyphenols and salivary proteins present in the mouth. To date, precipitation of proteins may be the most widely and accepted theory. Mucoproteins and proline-rich protein (PRP) help make up salivary proteins (Gawel 1998). The binding and precipitation of proline-rich proteins is described in three stages (Figure 4). In the first stage, the binding of multiple multidentate polyphenols to several sites on the protein causes the previously randomly coiled protein to coil around the polyphenol, making the protein more compact. In the second stage, the polyphenol fractions of the protein-phenol complexes cross-link forming polyphenol bridges and create protein dimmers. In the last stage, the dimers aggregate to form large complexes and precipitate. While polyphenols have been described as bi-dentate linkers and reactive proteins, where the proteins and polyphenols combine to form soluble complexes and these can grow to colloidal size at which point they scatter light and become larger leading to sediment formation. Phenolic compounds are capable of reducing the lubricating capacity of saliva. The reduction/loss in lubrication is often described as oral friction. When this happens, it leads to the widely described drying affect that is often associated with astringency. The flow rate of saliva has been linked to the perception of

astringency. It was demonstrated that subjects with high salivary flow had a less intense perception of astringency than low-flow subjects (Fischer and Noble 1994).

Figure 4. Proposed mechanism for PRP-polyphenol binding and subsequent protein aggregation and complex formation (Jöbstl et al. 2004)

Texture

The texture of food is important in consumer acceptance. Texture of foods has been studied extensively using instrumental and sensory methods (Szczesniak et al. 1963; Civille and Szczesniak 1973; Civille and Liska 1975). Textural characteristics have been defined and classified as the texture complex of a food in terms of mechanical properties, geometrical properties, fat and moisture content (Brandt et al. 1963). Dr. Alina Surmacka Szczesniak has been recognized for her many contributions to food texture and rheology. Since the early foundation laid by Dr. Szczesniak, researchers have developed sound sensory methods for measuring and evaluating food texture. A texture profile method based on the flavor profile method was developed by the Arthur D. Little Corporation (Brandt et al. 1963). Following that, Civille and Szczesniak (1973) and Civille and Liska (1975) built on the basic principles of texture profiling described by Brandt et al. (1963) who developed a comprehensive

guideline for training a texture profile panel. Bourne (1978) developed the texture profile analysis, which has been widely used to measure the texture of many foods (Table 6 and

Figure 5). Texture Profile Analysis (TPA) is a two-cycle uniaxial compression test for characterization of texture features including hardness, cohesiveness, brittleness, gumminess, adhesiveness, chewiness and elasticity (Table 6 and Figure 5).

Table 6. Selected terms used in texture profile analysis (Breene 1975;Bourne 2002)

Terms Definitions

Fracturability Force at the first significant break during compression Hardness The peak force of the first compression Cohesiveness The area of the second compression divided by the compression if the first compression Springiness The height that the food recovers during the time that elapses between the end of the first bite and the start of the second bite Gumminess The product of hardness x cohesiveness Chewiness The product of gumminess x springiness (which equals hardness x cohesiveness x springiness)

Figure 5. Instrumental texture profile curve (Breene 1975)

Rheology is the science of the deformation and flow of materials under well-defined conditions (Steffe 1992). It is the study of the manner in which materials respond to applied stress or strain. Stress is defined as force per unit (σ = force/area) (Steffe 1992). Stress can be expressed as pound per square inch (psi) or kilograms per square centimeters (kg/cm2).

Normal stress (σn) is stress acting perpendicular to a surface. Shear stress (σs or T ) is stress acting parallel to a surface. Principal stresses (σ1, σ2, σ3) are stresses acting on planes with shear stress (by convention, σ1 > σ2 >σ3). Strain is a measure of the deformation, also known as the change in shape. There are two ways to measure strain: longitudinal strain (shortening or lengthening) and strain stress (Steffe 1992). Fluid rheology can be understood as the material science of fluids.

Texture of Peanut Products

Spreadablility of foods that are elastoplastic or viscoplastic have been studied (Sun and Gunasekaran 2009). Bingham model and Casson model yield stress of stabilized and unstabilized peanut butters were determined (Citerne et al. 2001). The stabilized peanut butter had a yield stress of 374 Pa (Bingham) and 363 Pa (Casson), while the unstabilized was 27 Pa (Bingham) and 22 Pa (Casson). The larger yield stresses of the stabilized suspension could be attributed to the stabilizing agent used. Peanut oil is liquid at room temperature thus separates out of unstabilized peanut butter (peanut paste). Stabilizers, used to prevent oil separation, are hydrogenated oils such as cottonseed, rapeseed, soybean and palm oil. In peanut butter, stabilizers form crystal matrices entrapping and preventing the oil from rising to the surface of the product (Woodroof 1983; Totlani and Chinnan 2007).

Muego et al. (1990) used textural profile analysis to measure the textural characteristics of peanut butter and peanut paste. Results indicated that peanut paste was harder than commercial and laboratory produced peanut butter. The firmness and adhesiveness of three Haitian peanut spreads and 1 peanut butter made in the U.S. was determined by Hinds et al. (2002). The four products had similar firmness. The U.S. peanut butter was significantly more adhesive than the Haitian peanut spreads. This could be due to the addition of stabilizer in the U.S. peanut butter, while the three Haitian peanut spreads did not contain stabilizer. Totlani and Chinnan (2007) examined textural changes in peanut butter with 0, 0.5, 1.0, 1.5 and 2.0% stabilizer and stored at 35 C for three months and reported that the firmness of peanut butter increased in relation to the amount of stabilizer as

well the length of storage. They attributed the increase in firmness to the formation and strengthening of the crystal network, which was formed by the stabilizer.

SWEETPOTATO

Introduction

Sweetpotato (Ipomoea batatas) is a dicotyledonus plant belonging to the morning glory family. In the United States (U.S.), sweetpotatoes are commercially grown in New

Jersey, Texas, Iowa, Illinois, Indiana, Kentucky, Mississippi and North Carolina. North

Carolina leads the U.S. in the production of the sweetpotato (North Carolina Sweetpotato

Commission 2010). Sweetpotato ranks as the seventh most important crop in the world

(FAO 2004a; Loebenstein 2009) and third most important starchy food crop (Padda and

Picha 2008). Sweetpotatoes are grown in many developing countries, while over 80% of the crop is grown in developed countries (Woolfe 1992). The sweetpotato grows well in warm climates and is thought to be a tropical plant, which naturally thrives best in the South

Atlantic and Gulf Coast States (Grubb and Guilford 2008).

Composition

Sweetpotatoes are a rich source of fiber, protein, carotenoids, polyphenols, vitamins

(C, folate, and B6) and minerals (calcium, iron, and potassium) (Table 7) (Woolfe 1992;

Bovell-Benjamin 2007). Carbohydrates account for over 90% of sweetpotato dry matter, with starch varying between 57-90% depending on the variety. Hagenimana et al. (1994) reported the three major amylolytic enzymes in sweetpotatoes as -amylase, β-amylase and starch phosphorylase. β-amylase is the most abundant of the three.

Table 7. Proximate Composition of Sweetpotato (Wolfe 1992)

Component (%)

Starch 70

Total Sugars 10

Total Fiber 10

Total Protein 5

Lipid 1

Ash 3

Vitamins, organic acids and others 4

In developing countries, sweetpotatoes are an important source of protein where meat (i.e. animal protein) consumption is typically low. However, plant proteins have a relatively low biological value, a measure of the proportion of absorbed protein from a food. Walter et al.

(1983) reported the essential (indispensable) amino acid content of oven-dried sweet potato flour made from “Jewel” and “Centennial” sweetpotatoes (Table 8).

Table 8. Concentration of amino acids in sweetpotato flour (Walter et al. 1983)

Essential Amino Acids Jewel (g/100g) Centennial (g/100g)

Threonine 5.32 5.57

Valine 6.67 7.55

Methionine 0.97 1.19

Half-cystine 1.22 1.37

Isoleucine 3.94 4.38

Tyrosine 5.85 6.51

Phenylalaline 5.94 6.33

Lysine 3.82 4.47

Sweetpotatoes are one of the major food sources of carotenoids, and orange varieties of sweetpotatoes are richest in β-carotene. The basic structure of β-carotene is made up of eight isoprene units, which are cyclized at each end (Figure 6). These isoprene units are joined together to make up a conjugated chain which is common to all carotenoids. β- carotene, a precursor of vitamin-A, is absorbed in the intestines and stored in the liver where two molecules of vitamin A are formed from one molecule of β-carotene. β-carotene exists as several different isomers including all-trans-β-carotene and cis-β-carotene, with all-trans-

β-carotene being the most common and stable form (Hornung et al. 2005; Cerezo 2012).

The concentration of β-carotene in sweetpotatoes has been reported as high as 190

µg/g (Simonne et al. 1993). Van Jaarsveld et al. (2006) found that degradation of β-carotene

occurred in boiled sweetpotatoes. Bengtsson et al. (2008) reported degradation of β-carotene using three processing methods (boiling, steaming and deep-frying). Loss of β-carotene is due to degradation or isomerization when exposed to varying processes or environmental conditions (Bengtsson et al. 2008). Becoff et al. (2009) reported 16-34% loss of all-trans-β- carotene in flour made from dried sweetpotato chips.

Figure 6. Structure of β-carotene

Harvesting, Curing and Processing

Sweetpotatoes are generally harvested 100 to 140 days after planting. After harvesting, sweet potatoes are cured in a storage facility at 85ºF with a relative humidity ranging from 85 to 90% (Kemble 2004). Curing promotes the healing of wounds and reducing losses due to shrinkage and disease and enhances the organoleptic properties (such as texture and flavor) (Edmunds et al. 2008). Improper curing is associated with shorted shelf life, increased sprouting during storage and weight loss (Edmunds et al. 2008).

Processing sweetpotatoes into products such as flour and flakes extends the shelf-life and create alternative uses for sweetpotatoes. Therefore, spray drying and drum drying are necessary to make products such as sweetpotato flour and sweetpotato flakes (SPF). Spray

drying is a widely used industrial process involving the conversion of a liquid into a dried powder (Walstra et al. 1999; Verdurmen and de Jong 2003). Spray drying involves feeding a concentrated liquid product through an atomizer to form small droplets which are dried in the drying chamber and deposited into a collection vessel. The atomizer controls the rate at which the material is fed into the drying chamber (Patel et al. 2009).

Sweetpotato Flakes (SPF)

Drum drying is a processing technology which dries food using a large rotary metal cylinder that is steam heated internally (Walstra et al. 1999). A puree is applied to the hot drum and forms a thin film, and the dried film is scraped from the drum by a steel blade, collected and ground. Film formation is composed of two layers, where the first layer adheres to the drums and is referred to as the burn-on layer (Wadsworth et al. 1967). As the moisture content of the puree is reduced, the puree sticks to the surface forming a thin layer.

The second layer called the friction layer, differs in thickness and adheres by surface tension

(Wadsworth et al. 1967). This process is used for dried sweetpotato products, such as SPF.

SPF contain <5% moisture and can be rehydrated with hot water using a 1:1 by volume ratio. A shelf-life of two years is expected when SPF are stored in a cool dry place.

SPF contains 70% and 25% of the daily requirements of vitamins A and C, respectively

(Table 9). Emenhiser et al. (1999) reported that the β-carotene content of SPF was 39.8 µg/g.

However, SPFs have not been as widely accepted among consumers as white potato flakes.

Several researchers have suggested that consumer acceptability is limited due to a strong

hay-like odor that develops after 6 months of storage at 23ºC (Alexandridia and Lopez 1979).

Other reports have indicated the development of hay-like off-flavor after 29 days (Buttery et al. 1961; Walter et al. 1972).

Emenhiser et al. (1999) studied the β-carotene content in SPF packaged under different storage conditions (Table 10). The initial β-carotene content was 39.8 µg/g. SPF stored in polypylene film and air headspace retained 33% of β-carotene after 210 days.

While, SPF stored in nylon laminate and air headspace retained 52% of β-carotene after 210 days of storage.

Table 9. Nutritional facts label for sweetpotato flakes (BFC 2008)

Table 10. β-carotene content of sweetpotato flakes after storage under various packaging conditions (Emenhiser et al. 1999)

Packaging Conditions Days Polypropylene Film, Nylon Film, Air Nylon Film, Nylon Film, Stored Air Headspace Headspace Under Vacuum Oxygen Absorber β-carotene Content µg/g (% β-carotene retained)

0 39.8 39.8 39.8 39.8 30 34.2 (86.0) 35.1 (88.4) ------60 27.1 (68.2) 32.3 (81.3) ---- 45.0 (>100.0) 120 20.9 (52.5) 27.8 (69.9) 33.3 (83.7) 45.0 (>100.0) 210 13.2 (33.1) 20.8 (52.4) 25.6 (64.5) 39.4 (99.4)

Descriptive sensory and consumer acceptance

Leighton et al. (2010) developed a sensory lexicon which included terms to describe texture, aroma, flavor and aftertaste of orange flesh sweetpotato (OFSP) (Table 11).

Leighton et al. (2010) reported that OFSP differed from white flesh sweetpotato (WFSP) in that WFSP had a damp soil, slightly undercooked potato, earthy aroma and a less intense sweetpotato aroma. OFSP were reported to be sweeter, with flavor characteristics of yellow vegetables.

Table 11. Descriptive sensory lexicon for sweetpotatoes (Leighton et al. 2010)

Attribute Description Aroma Earthy Aromatic notes associated with damp soil, wet foliage or slightly undercooked potatoes Sweetpotato Aromatic associated with cooked sweetpotato, typical of WFSP. Burn An aromatic associated with vegetables that were burnt while cooking. Texture – initial impression: squeeze sweetpotato lightly between fingers, holding it on the skin side Moistness Hold sample between forefingers and evaluate the amount of wetness/juiciness released by the sample, which is visible when squeezing sample lightly. Firm Degree to which the sample retains its shape after lightly squeezing it. First bite Denseness The solidness/compactness of the sample. Moistness The amount of moistness/wetness of the sample in the mouth. Mastication Fiber Using a fork, gently break piece off sweet potato to observe fibers and then evaluate the amount of stringy fibers perceived in the mouth. Adhesive (Stickness/pasty) The amount to which the sample sticks to any of the mouth surfaces such as teeth, gums or palate and is perceived as pasty. Grainy Use the tongue to press the sweetpotato on to the palate. Degree to which surface is uneven, amount of graininess or roughness of particles on chewing. Flavor Vegetable sweet Taste characteristics of sweet vegetable varieties, such as sweet corn, sweetpotato, butternut or sweet carrots. Sweetpotato Flavor notes associated with the taste of cooked WFSP. Yellow vegetables (Butternut, carrots, pumpkin) Taste associated with yellow starchy vegetables such as butternut, pumpkin, carrots, and, to a lesser degree, squash. Aftertaste Sweet An aftertaste that leaves a sweetness on the tongue and in the mouth that is pleasant.

WFSP – white-fleshed sweetpotato

Leksrisompong et al. (2012) published a flavor and texture lexicon for sweetpotatoes

(Table 12). The researchers also evaluated consumer acceptance of 12 sweetpotato cultivars.

The sensory lexicon included terms on appearance, aroma, flavor, and texture of sweetpotatoes. Orange flesh sweepotatoes were less firm and dense than yellow and purple flesh sweepotato. Purple flesh sweetpotatoes were perceived as more firm, dense, less moist and higher in chalkiness. The majority of orange and purple flesh sweetpotates had higher overall aroma and brown sugar aroma compared to yellow flesh sweetpotatoes. Principal component analysis revealed that orange flesh sweetpotatoes were characterized mainly by canned carrot aroma and flavor, and dried apricot/floral aroma and flavor. Determination of consumer acceptability of sweetpotatoes consisted of consumers evaluating the samples with and without blind folds (Leksrisompong et al. 2012). Results suggest that consumers scored differently between treatments when they were allowed to see the samples. The overall liking of sweetpotatoes was driven by flavor liking followed by texture liking.

Table 12. Sensory attributes of sweetpotato cultivars (Leksrisompong et al. 2012)

Term Definition Visual Color homogeneity Degree of evenness of color Moisture Degree of surface moisture Fibrousness Amount of stringy fibers present Texture in mouth Firmness Amount of force necessary to compress the sample fully between the tongue and the palate Denseness Degree to which the sample is solid; compactness of the cross section Moistness Degree to which the sample is moist Smoothness Smoothness of chewed mass Cohesiveness Degree to which sample holds together after chewing Fibrousness Amount of stringy fibers received Residual fiber Amount of stringy fibers perceived after swallowing Chalkiness Degree to which the mouth feels chalky, like raw potato, very fine particles, often perceived on the roof of the mouth Aromatics Overall The overall orthonasal aroma impact Brown sugar Aromatic associated with brown sugar Potato Aromatic associated with white baked potato Earthy/canned carrot Earthy aromatic associated with canned carrot Dried apricot/floral Floral aromatics associated with dried apricot Vanilla Aromatics associated with vanilla and vanillin Flavor in mouth Brown sugar In-mouth aromatic associated with brown sugar Earthy/canned carrot In-mouth earthy aromatic associated with canned carrot Dried apricot/floral In-mouth floral aromatics associated with dried apricot White baked potato In-mouth aromatic associated with white baked potato Vanilla In-mouth aromatic associated with vanilla and vanillin Sour taste Basic taste stimulated by acid Sweet taste Basic taste stimulated by sugar Bitter taste Basic taste associated with caffeine Umami Basic taste associated with monosodium glutamate Astringent Sensation of drying, drawing and/or puckering of any of the mouth surfaces

Tomlins et al. (2007) used descriptive sensory analysis and evaluated consumer acceptability of OFSP to pale-fleshed sweetpotato (PFSP). Orange color, creamy color, pumpkin flavor, watery texture, starchiness, hardness/texture, coarse texture, yellow color, fibrous texture and sweet taste were the descriptive terms generated and used to describe differences between sweetpotato cultivars Polista and Sinia B (PFSP), and Karote DSM and

Resisto (OFSP). Principle component analysis (PCA) illustrated relationships among descriptive terms and sweetpotato cultivars. Resisto was characterized by watery texture, pumpkin flavor and orange color. Sinia B was characterized by creamy color, coarse texture, yellow color and sweet taste. Karote DSM was characterized by orange color. Polista was characterized by starchiness, hard texture and fibrous texture. Consumer acceptability testing revealed that OFSP were preferred over PFSP.

Dansby and Bovell-Benjamin (2003a) used descriptive sensory analysis to describe the appearance, flavor and texture of an extruded sweetpotato cereal. Two different formulations were used, one containing 100% sweetpotato flour (SPF) and the second containing 75% SPF/ 25% whole wheat bran (WWB). A 100% WWB cereal and a commercial cereal were compared. Color intensity was highest in the 75% SPF / 25% WWB and 100% WWB formulation. Sweet taste was most apparent in the 100% WWB followed by the 100% SPF formulation. The degree of liking among sweetpotato cereals (100% SPF and 75% SPF / 25% WWB) were not different and were liked slightly among the test population of school age children (Dansby and Bovell-Benjamin 2003b).

Flavor Constituents

The identification of compounds associated with cooked sweetpotato flavor was reported by Horvat et al. (1991) (Table 13). Maltol (3-hydroxy-2-methyl -4-pyrone) was identified as a principal aroma component in baked sweetpotatoes (Sun et al. 1995). Wang and Kays (2000) qualitatively and quantitatively reported the differences in volatile

constituents among baked, boiled and microwaved Jewel sweetpotato and identified the compounds that were related to specific cooked product aromas. A wide variety of baked sweetpotato volatiles, including hydrocarbons, acids, alcohols, aldehydes, esters, furans, ketones, and nitrogen containing compounds have been identified (Kays 1992). Volatiles released from cooked sweetpotatoes are derived from two primary sources: (1) the increased volatility of compounds already present; and (2) chemical reactions occurring during cooking that result in de novo synthesis of products (Kays 1992). Cooking methods influence heat penetration and internal temperature of sweetpotatoes. Each of these factors influences the synthesis of new compounds (Wang and Kays 2000). Out of 36 compounds identified from the sweetpotato, 54.3 and 6.4% compounds were from boiled and microwaved sweetpotatoes, respectively (Wang and Kays 2000). Baked sweetpotatoes contained all of the odor active compounds and had higher concentrations than other cooking methods.

Kays et al. (2005) used sucrose equivalents (SE) to measure relative sweetness among sweetpotato varieties. The majority of sweetpotatoes were high in sucrose which ranged between 29 to 37 g/100g (dry mass). The carbohydrate content of six varieties of sweetpotatoes was determined during different stages of growth and development (LaBonte et al. 2000). Sucrose was the major sugar in all six cultivars at all stages of development.

Beauregard, Jewel and Travis varieties had increases in total sugars from 7 to 19 weeks after transplanting, although they increased at different rates.

Table 13. Volatile compounds identified in cooked sweetpotatoes (Horvat et al. 1991)

Compounda Jewel (area %) Tainung 57 (area %) 99 (area %)

Toluene tr tr Tr Pyridine + xylene tr tr Tr Furfural 1.1 1.3 1.0 2-acettylfuran tr tr tr Benzaldehyde 2.7 1.0 1.5 5-methyl-2-furfual 10.1 2.1 5.9 Limonene tr tr tr Cineole tr tr -- Phenylacetaldehyde 2.6 tr 2.2 Linalool + nonanal 4.3 1.2 tr -terpineol 4.4 tr tr β-cyclocitral tr tr tr Bornyl acetate 10.5 1.1 4.3 -copaene 5.8 2.3 13.8 Caryophyllene 1.7 tr 2.7 Sesquiterpene tr tr tr hydrocarbon I Sesquiterpene tr tr tr hydrocarbon II (E)-β-farnesene 2.5 tr 1.9 -cadinene 1.3 -- 1.0 β-ionone 4.6 1.8 7.4 Sesquiterpene 1.2 tr -- hydrocarbon III Sesquiterpene tr 1.3 tr hydrocarbon IV Palmitic acid 2.5 1.1 1.5 aIdentified by GC/MS and GC retention times Tr = < 1% based on GC peak areas

MILK POWDER

Introduction

Milk and dairy products are major components of the human diet in Western countries. Bovine milk is consumed more than any other type. Fluid milk, cheese, butter, whole milk powder, skimmed milk powder, yogurt, fermented milk products, casein and infant formula make up total world milk utilization (Fox 2003).

The United States (US) is the largest dairy processor in the world, producing 800,000 mt of dry milk powder (DMP) each year (USDEC 2011). DMP is easily instantized, convenient, and economical source of dairy solids. DMP is classified as either non-fat dry milk powder or whole milk powder (WMP). Using DMP instead of fluid milk reduces transport costs, storage space and eliminates refrigerated shipping and warehousing of the product (Verdurmen and Jong 2003). In addition, milk powders exist in roller-dried and spray-dried form, the latter being the most common.

Composition

Whole milk contains, about 87% water and skim milk contains about 91% water.

Milk powders are highly nutritious containing fat, protein, minerals, vitamins, lactose and water. WMP has a composition of 2.0-4.5% moisture, 26.0-29% fat, 24.5-27.0% protein,

36.0-38.5% lactose and 5.5-6.5% ash (DMI 2005; Kim et al. 2009; USDEC 2011).

According to the code of federal regulations (CFR), WMP should have a fat content between

26-40%, while the moisture should be no more than 5% based on milk solids (CFR 2008c).

Non-fat milk powder (NFMP) is obtained by the removal of water from pasteurized skim milk (CFR, 2008c). The final product should be no more than 1.5% milk fat and the moisture content should be less than 5% (CFR 2008c). NFMP is composed of 3.0-4.0% moisture, 0.6-1.25% fat, 34.0-37.0% protein, 49.5-52.0% lactose, and 8.2-8.6% ash (USDEC

2011). Carbohydrates are a major component in milk powder, and accounts for 54% of the total solids, non-fat portion of milk (USDEC 2011). Whey proteins and caseins are two types of proteins in milk. Casein makes up 80% of bovine milk proteins. Milk proteins contain all 9 essential amino acids required by humans and the amino acid composition is presented in Table 14. DMP can also be fortified with vitamins and minerals.

Final quality is an important consideration in the preservation of milk powders

(Verdurmen and Jong 2003). Processing of milk powder involves the removal of water at the lowest possible cost under stringent hygiene conditions while retaining all the desirable natural properties of the milk – color, flavor, solubility and nutritional value. During milk powder production, water is removed by boiling the milk under reduced pressure at low temperature in a process known as evaporation (USDEC 2011). The resulting concentrated milk is then dried to remove further moisture.

Table 14. Amino Acid Composition of Milk Powder (USDEC 2011)

Amino Acids Whole Milk Powder (g/100g) Skim Milk Powder (g/100g)

Tryptophan 0.37 0.51

Threonine 1.19 1.63

Isoleucine 1.59 2.19

Leucine 2.58 3.54

Lysine 2.09 2.87

Methionine 0.66 0.91

Cystine 0.24 0.33

Phenylalanine 1.27 1.75

Tyrosine 1.27 1.75

Valine 1.76 2.42

Arginine 0.95 1.31

Histidine 0.71 0.98

Alanine 0.91 1.25

Aspartic acid 2.00 2.74

Glutamic acid 5.51 7.57

Glycine 0.56 0.77

Proline 2.55 3.50

Serine 1.43 1.97

Hydroxyproline

Oxidation

Liang (2000) used fluorescence and oxidation parameters to investigate the changes in (WMP) stored at 37, 50, 60 and 70ºC. Whole milk powder extracted with chloroform- methanol and assayed by transmission spectrofluorometry suggested that lipids are critical components contributing to the formation of intrinsic fluorescence. Fresh WMP had an excitation peak with a maximum wavelength of around 270 nm and oxidized samples had two peaks with wavelength maximum at 270 and 350 nm. The emission spectra for the oxidized sample showed an intensive and broad peak at 440 nm. The fluorescence intensity at 350 nm excitation and 440 nm emission increased significantly during oxidation.

However, the spectra of skim milk powder did not change in wavelength during storage, suggesting that skim milk powder had little to no oxidation during storage. The author also suggests that fluorescence intensity is a better index of milk powder oxidation than peroxide value.

Flavor

Drake et al. (2002) developed a descriptive flavor lexicon for milk powder, including descriptor definitions, references and instructions to prepare references (Appendix 1). The lexicon has 22 terms including cook/sulfurous, caramelized/butterscotch, brothy/potato-like and milkfat/lactone. Kobayashi et al. (2008) used gas chromatography-mass spectrometry

(GC-MS), gas chromatorgraphy-olfactometry (GC-O), gas chromatography atomic emission detector (GC-AED), aroma extract dilution analysis (AEDA), descriptive sensory analysis

and aroma models to identify compounds characteristic of high-heat skim milk powder. Key odorants in high heat skim milk powder have been described as animal, brothy, metallic/mushroom-like and vitamin-like (Kobayashi et al., 2008). Karagül-Yücee et al.

(2002) reported the volatile flavor compounds of stored nonfat milk using gas- chromatography-olfactometry and descriptive sensory. Aldehydes, ketones, and free fatty acids were responsible for generation of flavors in NFMP. (E,E)-2,4-decadienal (fried/fatty), o-aminoacetophenone (corn tortilla), acetic acid and hexanoic acids (sour/vinegar), butanoic acid (cheesy), pentanoic acid (sweaty), 2,5-dimethyl-4-hydroxy 3(2H)-furanone and 2- methyl-3-hydroxy-4H-pyran-4-one (burnt sugar), 3-(methylthio)propanal (boiled potato), octanoic, decanoic and dodecanoic acids (waxy), p-cresol (cowy/barny), 3-methylindole

(fecal), dimethyl trisulfide (cabbage) and phenylacetic acid (rose-like) and 1-octen-3-one

(mushroom) exhibited high aroma impact using AEDA (Karagül-Yücee et al. 2002).

Flavor as affected by storage

Caudle et al. (2005) evaluated the consumer acceptability of 4 different products containing SMP. Reconstituted SMP was used as an ingredient to make vanilla ice cream, strawberry yogurt, white chocolate bars and hot cocoa. Consumer acceptability tests indicated that products containing SMP were not liked as much as products containing fluid skim milk. In a study reported by Carunchia et al. (2007), skim milk powder (SMP) and whole milk powder (WMP) were stored for 36 months at 21ºC. Sweet aromatic in SMP,

cooked caramelized and milk fat/lactone in WMP decreased in flavor intensity during storage.

Lloyd et al. (2009) studied WMP stored under nitrogen and air at 2 and 23º C.

Cooked milk, sweet aromatic and milkfat flavors were present initially as they are typically associated with fresh WMP. After 12 months, WMP stored under air at 23 º C was lower in milkfat and sweet aromatic flavors than all other treatments. WMP stored under nitrogen and at 2º C had no development of the descriptor “painty”, while painty developed in samples stored under air between 3 and 6 months. Painty was not detected in samples stored under nitrogen. Lloyd et al. (2009) also reported the relationship between sensory perception and volatile compounds during the storage study of WMP stored under nitrogen and air. Grassy flavor correlated with 1-octen-3-ol, hexanal, heptanal and nonanal. Painty correlated with 2- methylbutanal, 1-octen-3-ol, 3-octen-2-one, hexanal, heptanal, octanal and nonanal.

READY-TO-USE THERAPEUTIC FOODS (RUTF)

Introduction

Malnutrition, a major public health problem, is a primary contributor to child mortality and total global disease burden (WHO 1999; Müller and Krawinkel 2005; WHO

2007). Individuals are considered to be malnourished or suffer from under nutrition if their diet does not provide adequate calories and protein for maintenance and growth, or they cannot fully utilize the food they eat due to illness (Müller and Krawinkel 2005). Several different nutrition disorders may develop depending on which nutrients are lacking.

In 2000, 189 nations made an agreement to relieve people from extreme poverty and multiple deprivations, and these goals became known as the Millennium Development Goals

(MDG). MDG are eight international development goals agreed to by world leaders to achieve by the year 2015. Those goals are: (1) eradicating extreme poverty and hunger, (2) achieving universal primary education, (3) promoting gender equality and empowering women, (4) reducing child mortality rates, (5) improving material health, (6) combating

HIV/AIDS, malaria, and other diseases, (7) ensuring environmental sustainability and (8) developing a global partnership for development. These goals have not been fully achieved and efforts are still being made to combat these issues.

From 2000-2002 more than 850 million people in developing countries have been reported to be undernourished (FAO 2004b). Nearly 20 million children under the age of five suffer from severe acute malnutrition (SAM) and 1 million die each year (WHO 2007).

SAM is defined by a very low weight for height, by severe wasting or by the presence of nutritional oedema. Protein-energy malnutrition in children is defined by measurements that fall below the normal weight for age (underweight), height for age (stunting), and weight for height (wasting) (Pinstrup-Anderson et al.1993; WHO 2007). A child between the ages 6-59 months with an arm circumference of less than 110 mm is also a sign of SAM (WHO 2007).

Children under the age of 5 are particularly at risk from malnutrition because of demanding dietary requirements (Figure 7). In addition, children living in rural areas are almost twice as likely to be underweight than children in urban households. In particular, children in Africa and Asia have higher incidence of stunting, being underweight and wasting (Table 16).

Figure 7. Causes of death among child under 5 years of age, 2000- 2003, worldwide (Müller and Krawinkel 2005)

Table 15. Prevalences of protein-energy malnutrition among children under 5 years of age in developing countries, 1995 (Müller and Krawinkel 2005)

Several demographic trends in developing countries may hinder efforts to reduce SAM among children. For instance, impoverished adults tend to have limited access to food, insufficient resources to take care of children, and substandard living can all lead to malnutrition (Figure 8). These specific populations can also have larger families and the additional population also exerts greater pressure on shrinking arable agricultural land and contributes to ecological degradation in the rain-fed agricultural, semiarid, and nomadic pasture areas in such countries.

Figure 8. Direct and indirect causes of malnutrition (Müller and Krawinkel 2005)

RUTF

The burden of child mortality due to SAM primarily remains absent from the international health agenda and there are few national policies aimed at addressing the issue comprehensively. Community based approaches are one way of combating issues such as

SAM. The standard treatment for SAM was originally with therapeutic milk (Formula-75 and Formula-100), which was developed in the 1980s by Nutriset, a French private company specializing in therapeutic foods, which had to be administered in a hospital setting (Guimón and Guimón 2012). Formula-75 (F-75) and Formula-100 (F-100) was designed as phase 1 and phase 2 treatments. F-75, phase 1, is given as a stabilization treatment for SAM. A malnourished child admitted to a hospital will likely have infections or diseases of some sort, and also the organs are incapable of metabolizing normal quantities of proteins, fat or sodium. Therefore, it is necessary to start treatment consisting of a low protein, fat and sodium, but rich in carbohydrates. F-100 is given during Phase II of treatment and is known as the recovery phase. The dry skim milk powder formulation (F-75 and F-100) contains vegetable fat, whey, maltodextrin, sugar and a vitamin and mineral mix. Some disadvantages of F-75 and F-100 include having to be reconstituted with drinking water, requires heat, clean utensils, precise measurement, the finished product only retains its properties for a few hours and can not be left at room temperature (Guimón and Guimón 2012).

Figure 9. F-75 and F-100 therapeutic milk in dehydrated form

Later in the 1990s, focus was geared towards developing a better version of the F-100 in the form of a ready-to-use-therapeutic food (RUTF). RUTF are soft enriched foods that do not require additional water, can be consumed easily by children from the age of six months and does not require specialize storage (WHO 2007). In the mid 1990s, French scientist

André Briend teamed up with Michel Lescanne, the director of Nutriset to develop RUTFs.

Several forms of RUTF were tried by the scientists such as enriched pancakes, doughnuts and biscuits. They also tried RUTF in the form of a chocolate bar with similar composition to F100; however, the bar melted easily and had an unacceptable taste when vitamins and proteins were added (Guimón and Guimón 2012).

Plumpy’nut

Plumpy’nut® is peanut-base RUTF used to treat SAM and is manufactured by

Nutriset. Nutriset has a patent (US 6,346,284) for a complete food or nutritional supplement

(Briend and Lescanne 2002), which would cover products such as Plumpy’nut®.

Plumpy’nut® has a two year shelf life and requires no water, preparation, or refrigeration, and does not support the growth of bacteria because of low-moisture and high fat content

(Linneman et al. 2007). Unlike therapeutic milk, Plumpy'nut® can be administered at home and without medical supervision. The ingredients in Plumpy'nut® include peanut-paste, sucrose, vegetable fat and skimmed milk powder, enriched with vitamins and minerals. The fat suspension of Plumpy’nut® makes the addition of vitamins and minerals very stable

(Guimón and Guimón 2012).

MANA

Mother-administered-nutritive- aid (MANA), similar to Plumpy’nut®, is a peanut- base RUTF developed by MANA Nutrition (Matthews, NC). MANA Nutrition seeks to develop and provide solutions to address severe cases of malnutrition. The fortified peanut paste is formulated to provide a child’s basic daily nutritional need by consuming three servings per day. Just like Plumpy’nut, MANA does not require preparation before consumption and does not require medical supervision. Other examples of RUTF include

Medika Mumba (fortified peanut paste), Unimix (fortified maize flour and soybeans) and

Wawa Mum (fortified chickpeas).

Clinical evidence of using RUTF

In Malawi, a food insecure country, a large feeding trial study using locally produced

RUTF was conducted with severely and moderately malnourished children (Linneman et al.

2007). The RUTF consisted of 25% peanut butter, 28% sucrose, 30% full-cream, 15% vegetable oil, and 1.4% imported vitamin and mineral supplement. The study included 2,131 and 806 severely malnourished and moderately malnourished children, respectively. Of the

original 2,131 and 806 participants, 89 and 85% recovered with weight gain. In another

study where standard therapy (F-100 and hospitalization) and home-based therapy (RUTF) was used, the outcome of the study revealed that home-based therapy with RUTF resulted in higher rates of recovery (Ciliberto et al. 2005). Patel et al. (2005) evaluated the effectiveness of peanut-based and corn-soy based supplementary foods. Children receiving peanut-based

supplement had greater weight gain than those receiving corn-soy based supplement.

REFERENCES

Abegaz EG, Kerr WL, Koehler PE. 2004. The role of moisture in flavor changes of model peanut confections during storage. Lebensm Wiss Technol 37:215-25.

Alexandridia N, Lopez A. 1979. Lipid changes during processing and storage of sweetpotato flakes. J Food Sci 44:1186-90.

American Peanut Council. 2008. Peanuts: A Brief History. Available at: http://www.peanutsusa.org.uk/Europe/index.cfm?fuseaction=home.page&pid=72.

Andersen PC, Hill K, Gorbet DW, Brodbeck BV. 1998. Fatty acid and amino acid profiles of selected peanut cultivars and breeding lines. J Food Comp Anal 11:100-11.

ASTM. 1995. Standard definition of terms relating to sensory evaluation of materials and products. In Annual Book of ASTM Standards. Vol. 15.07 American Society for Testing and Materials, Philadelphia, Pennsylvania.

Ballard, T. S., Mallikarjunan, P., Zhou, K. and O’Keefe, S.F. 2009. Optimizing the Extraction of Phenolic Antioxidants from Peanuts Skins Using Response Surface Methodology. J. Agric. Food Chem. 57: 3064-3072.

Baker, G.L, Cornell, J.A., Gorbet, D.W., O’Keefe, S.F., Sims, C.A., and Talcott, S.T. 2002. Storage water activity affects oxidation and sensory properties of high oleic peanuts. J. Food Sci. 67:1600-1603.

Basha SMM, Cherry JP, Young CT. 1980. Free and total amino acid composition of maturing seed from six peanut (Arachis hypogeal L) cultivars. Peanut Sci 7:32–7.

Bate-Smith EC, Swain T. 1962. Flavonoid compounds. In Comparative Biochemistry. Mason, H.S., Florkin, A.M., Eds.: Academic Press: New York. pp. 755-809.

Bechoff A, Dufour D, Dhuique-Mayer C, Marouzé C, Reynes M, Westby A. 2009. Effect of hot air, solar and sun drying treatments on provitamins A retention in orange flesh sweetpotato. J Food Eng 92:164-71.

Beecher GR. 2003. Overview of dietary flavonoids: nomenclature, occurrence and intake. Processings of the third international scientific symposium on tea and human health: role of flavonoids in the diet. J Nutr 3248-54S.

Beecher JW, Drake MA, Luck PJ, Foegeding EA. 2008. Factors regulating astringency of whey protein beverages. J Dairy Sci 91:2553-60.

Bengtsson A, Namutebi A, Alminger ML, Svanberg U. 2008. Effects of various traditional processing methods on the all-trans-β-carotene content of orange-fleshed sweetpotato. J Food Comp Anal 21:134-43.

Bett KL, Vercellotti JR, Lovegren NV, Sanders TH, Stinsch RT, Rasmussen GK. 1994. A comparison of flavor and compositional quality of peanuts from several origins. Food Chem 51: 21-7.

BFC. Bruce Foods Corporation. 2008. Product specification and nutritional facts. Instant sweetpotato flakes.

Biolatto A, Grigioni G, Irurueta M, Sancho AM, Taverna M, Pensel N. 2006. Seasonal variation in the odour characteristics of whole milk powder. Food Chem 103:960-67.

Boskou D, Elmadfa I. 1999. Frying of food: oxidation, nutrient, and non-nutrient antioxidants, biologically active compounds and high temperatures. CRC press: Boca Raton, FL. pp. 1-277.

Bourne MC. 1978. Texture profile analysis. Food Technol 32:62-6.

Bourne MC. 2002. Food Texture and Viscosity: Concept and Measurement. 2nd Ed. Acdemic Press. pp. 1-400.

Bovell-Benjamin AC. 2007. Sweetpotato: A review of its past, present and future role in human nutrition. Advances in Food Nutr Res 52: 1-59.

Braddock JC, Sims CA, O’Keefe SF. 1995. Flavor and Oxidative Stability of Roasted High Oleic Acid Peanuts J Food Sci 60:489-93.

Brandt MA, Skinner EZ, Coleman JA. 1963. Texture profile method. J Food Sci 28:404-9.

Breene WM.1975. Application of texture profile analysis to instrumental food texture evaluation. J Texture Stud 6:53-82.

Breslin PAS. 2001. Human gestation and flavour. J Flav Fragr 16:439-56.

Briend A, Lescanne M. 2002. Food or nutritional supplement, preparation method and uses. Patent No. US 6,346, 284. pp. 1-10.

Buttery RG, Hendel CE, Boggs MM. 1961. Autooxidation of potato granules. J Agri Food Chem. 9:245-8.

Carunchia Whetstine ME, Drake MA. 2007. The flavor and flavor stability of skim and whole milk powders. In: Cadwallader KR, Drake MA, McGorrin R, editors. Flavor of Dairy Products. Washington, DC.: ACS Publishing. pp. 217-51.

Castilla P, Echarri R, Dávalos A, Cerrato F, Ortega H, Teruel JL, Lucas MF, Gómez- Coronado D, Ortuño J, Lasunción MA. 2006. Concentrated red grape exerts antioxidant, hypolipidemic, and anti iinflammatory effects in both hemodialysis patients and healthy subjects. Am J Clin Nutr 84:252-62.

Caudle AD, Yoon Y, Drake M. 2005. Influence of flavor variability in skim milk powder on consumer acceptability of ingredient applications. J Food Sci 70:S427-31.

Cerezo J, Zúñiga J, Bastida A, Requena A. 2012. Antioxidant properties of β-carotene isomers and their role in photosytems: insights from Ab initio simulations. J Phys Chem A 116:3498-506.

CFR. Code of Federal Regulations. 2008a. Title 21, Part 164.150. http://www.gpoaccess.gov/CFR/INDEX.HTML. Accessed Dec 15, 2008.

CFR. Code of Federal Regulations. 2008b. Title 21, Part 131.147. http://gpoaccess.gov/CFR/INDEX.HTML. Accessed Dec. 30, 2008.

CFR. Code of Federal Regulations. 2008c. Title 21, Part 131.125. http://gpoaccess.gov/CFR/INDEX.HTML. Accessed Dec. 30, 2008.

Chetschik I, Granvogl M, Schieberle P. 2008. Comparison of the key aroma compounds in organically grown, raw west-african peanuts (Arachis hypogaea) and in ground, pan-roasted meal produced thereof. J Agric Food Chem. 56:10237-43.

Chetschik I, Granvogl M, Schieberle P. 2010. Quantitation of key peanut aroma compounds in raw peanuts and pan-roasted peanut meal. Aroma reconstruction and comparison with commercial peanut products. J Agric Food Chem. 58:11018-26.

Chung S-Y, Ullah AHJ, Sanders TH. 1994. Peptide mapping of peanut proteins: identification of peptides as potential indicators of peanut maturity. J Agric Food Chem 42:623-28.

Ciliberto MA, Sandiage H, Ndeskha MJ, Ashorn P, Briend A, Ciliberto HM, Manary MJ. 2005. Comparison of home-based therapy with ready-to-use- therapeutic food with standard therapy in the treatment of malnourished Malawian children: a controlled, clinical effectiveness trial. Am J Clin Nutr 81:864-70.

Citerne GP, Carreau PJ, Moan M. 2001. Rheological properties of peanut butter. Rheol Acta 40:86-96.

Civille GV, Liska IH. 1975. Modifications and applications to foods of the General Foods sensory texture profile terchnique. J Texture Stud 6:19-31.

Civille GV, Szczesniak AS. 1973. Guidelines to training a texture profile panel. J Texture Stud 4:204-23.

Cobb WY, Johnson BR. 1973. Physicochemical properties of peanuts. In Peanuts: Culture and Uses. Wilson, C.T. Amer Peanut Res Edu Assoc. Stillwater, OK. pp.209-264.

Dansby MA, Bovell-Benjamin AC. 2003a. Sensory characterization of a ready-to-eat sweetpotato breakfast cereal by descriptive analysis. J Food Sci 68:706-9.

Dansby MA, Bovell-Benjamin AC. 2003b. Physical properties and sixth graders’ acceptance of an extruded ready-to-eat sweetpotato breakfast cereal. J Food Sci 68:2607-12.

Davis JP, Dean LL, Price KM, Sanders TH. 2010. Roast effects on the hydrophilic and lipophilic capacities of peanut flours, blanched peanut seed and peanut skins. Food Chem 119:539-47.

DMI. Dairy Management Inc. 2005. Whole Milk Powder/Dry Whole Milk Ingredients. pp. 1-2.

Drake MA, Karagul-Yuceer Y, Cadwallader KR, Civille GV, Tong PS. 2002. Determination of the sensory attributes of dried milk powders and dairy ingredients. J. Sensory Stud 18:199-216.

Drake MA. 2007 Sensory analysis of dairy foods. J Dairy Sci 90:4925-37.

Drewnowski A. 2001. The science and complexity of bitter taste. Nutr. Rev. 59:163-9.

Drewnowski A, Gomez-Carneros C. 2000. Bitter taste, phytonutrients, and the consumer: a review. Am J Clin Nutr 72:1424-35.

Edmunds B, Boyette M, Clark C, Ferrin D, Smith T, Holmes, G. 2008. Postharvest Handling of Sweetpotatoes. North Carolina Cooperative Extension Service. p. 1-56.

Eheart JF, Young RW, Allison AH. 1955. Variety, type, year and location effects on the chemical composition of peanuts. J Food Sci 20:497-505.

Emenhiser C, Watkins RH, Simunovic N, Solomons N, Bulux J, Barrows J, Schwartz S. 1999. Packaging preservation of β-carotene in sweetpotato flakes using flexible film and an oxygen absorber. J Food Qual 22:63-73.

FAO. 2004a.Food and Agriculture Organization of the United Nations. Agricultural data. Available at: http://faostat.fao.org

FAO. 2004b. Food and Agriculture Organization of the United Nations. Undernouishment around the world. In: The state of food insecurity in the world 2004. Rome: The Organization.

Felland SL, Koehler PE. 1997. Sensory, chemical and physical changes in increased water activity peanut butter products. J Food Qual 20:145-56.

Fox PF. 2003. The major constituents of milk. In Dairy Processing: Improving quality. Gerrit Smit ed. CRC Press, Boca Raton, FL. pp. 5-41.

Fischer U, Noble AC. 1994. The effect of Ethanol, Catechin Concentrations, and pH on sourness and bitterness of wine. Am J Enol Vitic 45:6-10.

Gawel R. 1998. Red wine astringency a review. Aust J Grape Wine Res 4:74-95.

Grubb EH, Guilford WS. 2008. The Potato: A Compilation of Information from Every Available Source (1912). Kessinger Publishing, Whitefish, MT. pp.1-632.

Guimón J, Guimón P. 2012. How ready-to-use therapeutic food shapes a new technological regime to treat children malnutrition. Technol Forecast Soc. 79:1319-27.

Guinard J-X, Pangborn RM, Lewis MJ. 1986. The time-course of astringency in wine upon repeated ingestion. Am J Enol Vitic 37:184-9.

Hagenimana V, Vézina L-P, Simard RE. 1994. Sweetpotato -and β-Amylases: Characterizatioin and Kinetic Studies with endogenous inhibitors. J Food Sci 59:373-7.

Hale OM, McCormick WC. 1981. Value of Peanut Skins (testa) as a feed ingredient for growing-finishing swine. J Anim Sci 53:1006.

Hartwig P, McDaniel MR. 1995. Flavor characteristicsof lactic, malic, citric, and acetic acids at various pH levels. J Food Sci 60: 358-84.

Hill GM. 2002. Peanut by-products fed to cattle. Vet Clin Food Anim 18:295-315.

Hinds MJ, Jolly C M, Nelson RG, Donis Y, Prophete E. 2002. Consumer acceptability and physicochemical properties of Haitian peanut-butter-type products (Mambas) compared with U.S. peanut butter. Peanut Sci 29:102-9.

Hoffpauir CL. 1953. Peanut composition – relation to processing and utilization. J Agric Food Chem 1:668-71.

Holley KT, Young C. 1963. The relation of peanut maturity and pigments. Proceedings of Peanut Improvement Working Group. Stillwater, OK, pp. 52-9.

Hornung T, Skenderovic M, Motzkus M. 2005. Observation of all-trans-β-carotene wave packet motion on the electronic ground and excited dark state using degenerate four-wave mixing (DFWM) and pump –DFWM. Chem Phys Lett 402:283-8.

Horvat RJ, Arrendale RJ, Dull GG, Chapman GW, Kays SJ.1991. Volatile constituents and sugars of three diverse cultivars of sweetpotato. J Food Sci 56: 714-15.

Houlihan CM, Ho C-T, Chang SS. 1985. Natural antioxidants. In Flavor Chemistry of fats and oils. Min, D. B. and Smouse, T.H. (eds) American Oil Chemists’ Society, Champaign, IL. pp. 117.

Huang MT, Ferraro T. 1992. Phenolic compounds in food and cancer prevention: an overview. In Phenolic compounds in food and their effects on health II: Antioxidants and cencer prevention. Ho C-T, Lee CY, Huang MT Eds. American Chemical Society Series. New York, NY. Vol. 507. pp. 8-34.

Isleib TG, Pattee HE, Sanders TH, Hendrix KW, Dean LO. 2006. Compositional and sensory comparisons between normal- and high-oleic peanuts. J Agric Food Chem 54:1759-63.

Jensen RG. 1995. Handbook of milk composition. Academic Press. San Diego, CA.

Jöbstl E, O’Connell J, Fairclough JP, Williamson MP. 2004. Molecular model for astringency produced by polyphenol/protein interactions. Biomacromol 52:942-9.

Johnsen PB, Civille GV, Vercellotti JR, Sanders TH, Dus CA. 1988. Development of a lexicon for the descripton of peanut flavor. J Sensory Stud 3:9-17.

Kãhkõnen MP, Hopia AI, Vuorela HJ, Rauha JP, Pihlaja K, Kujala TS, Heinonen M. 1999. Antioxidants of plant extracts containing phenolic compounds. J Ag Food Chem 47:3954-62.

Karchesy JJ, Hemingway RW 1986. Condensed Tannins (4β → 8;2β → O→7)-Linked Procyanidins in Arachis Hypogea L. J Agric Food Chem 34: 966-70.

Karadag A, Ozcelik B,Saner S. 2009. Review of methods to determine antioxidant capacities. Food Anal Methods. 2:41-60.

Karagül-Yüceer Y, Cadwallader KR, Drake M. 2002. Volatile flavor components of stored nonfat dry milk. J Agric Food Chem. 50:305-12.

Kays SJ. 1992. The chemical composition of the sweetpotato. In: Sweetpotato Technology for the 21st Century. Hill, W.A., Bonsi, C.K., Loretan, P.A. Eds. Pp. 201-262.

Kays SJ, Wang Y, McLaurin WJ. 2005. Chemical and geographical assessment of the sweetness of the cultivated sweetpotato clones of the world. J Amer Soc Hort Sci 130:541- 97.

Kemble J. 2004. Harvesting and curing sweetpotatoes. Alabama Cooperative Extension System. pp. 1-2.

Kim EH-J, Chen XD, Pearce D. 2009. Surface composition of industrial spray-dried milk powders. 1. Development of surface composition during manufacture. J Food Eng 94:163-8.

Kinnamon SC, Cummings TAA. 1992. Chemosensory transduction mechanisms in taste. Rev Physiol 54: 715-31.

Kinnamon SC, Margolskee RF. 1996. Mechanisms of taste transduction. Curr Opin Neurobiol 6:506-13.

Kitts DD. 1994. Bioactive substances in food: Identification and potential uses. Can J Physiol Pharmacol 72:423-34.

Kobayashi N, Mizota Y, Kumazawa K, Nishimura O. 2008. Character impact odorants of high-heat skim milk powder isolated by simultaneous distillation-extraction. J Agric Food Chem 56:2469-76.

Kris-Etherton PM, Hecker KD, Bonanome A, Coval SM, Binkoski AE, Hilpert KF, Griel AE, Etherton TD. 2002. Bioactive Compounds in Foods: Their role in the prevention of cardiovascular disease and cancer. Am J Med 113:71S-88S.

LaBonte D, Picha DH, Johnson HA. 2000. Carbohydrate-related Changes in sweetpotato storage roots during development. J Amer Soc Hort Sci 125:200-4.

Lawless HT, Horne J, Giasi P. 1996. Astringency of organic acids is related to pH. Chem Senses 21:397-403.

Lee MH, Ho CT, Chang SS. 1981. Thiazoles, oxazoles and oxazolines identified in the volatile flavor of roasted peanuts. J Agric Food Chem 29:684-6.

Lee CB, Lawless HT. 1991. Time-course of astringent materials. Chem Senses 16:225-38.

Lee SY. 2001. Modeling shelf-life of whey-protein coated peanuts analyzed by static- headspace gas-chromatography. Institute of Food Technologist Meeting. IFT Session 73D- 30, New Orleans, LA.

Leighton CS, Schönfeldt HC, Kruger R. 2010. Quantitative descriptive sensory analysis of five different cultivars of sweet potato to determine sensory and textural profiles. J Sen Stud 25: 2-18.

Leksrisompong PP, Whitson ME, Truong VD, Drake MA. 2012. Sensory attributes and consumer acceptance of sweet potato cultivars with varying flesh colors. J Sen Stud. 27:56- 69

Linneman Z, Matilsky D, Hdekha M, Manary MJ, Maleta K, Manary MJ. 2007. A large scale operational study of home-based therapy with ready-to-use therapeutic food in childhood malnutritionin Malawi. Matern Child Nutr 3:206-15.

Liang J-H. 2000. Kinetics of fluorescence formation in whole milk powders during oxidation. Food Chem 71:459-63.

Lloyd MA, Hess SJ, Drake MA. 2009. Effect of nitrogen flushing and storage temperature on flavor and shelf-life of whole milk powder. J Dairy Sci 92:2409-22.

Loebenstein G, Thottappilly G, Fuentes S, Cohen J. 2009. Virus and phytoplasma diseases. In: The sweetpotato. Loebenstein, G. and Thottappilly, G. Springer, New York, NY. pp. 105- 34.

Lou H, Yamazaki Y, Sasaki T, Uchida M, Tanaka H, Oka S. 1999. A-type proanthocyanidins from peanut skins. Phytochem 54: 297-308.

Lou H, Yuan H, Yamazaki Y, Sasaki T, Oka S. 2001. Alkaloids and flavonoids from peanut skins. Planta Med. 67:345.

Maehashi K, Huang L. 2009. Bitter peptides and bitter taste receptors. Cell Mol Life Sci 66:1661-71.

Margolskee RF. 1993. The biochemistry and molecular biology of taste transduction. Curr Opin Neurobiol 3: 526-31.

Margolskee RF. 2002. Molecular mechanisms of bitter and sweet taste transduction. J Biol Chem 277:1-4.

Mason ME, Johnson B. 1966. Flavor components of roasted peanuts. Some low molecular weight pyrazines and a pyrrole. J Agr Food Chem 14:454-60.

Mason ME, Newell JA, Johnson BR, Koehler PE, Waller GR. 1969. Nonvolatile flavor components of peanuts. J Agric Food Chem 17:728-32.

McLaughlin SK, McKinnon PJ, Margolskee RF. 1992. Gustducin is a taste-cell specific G protein closely related to translation. Nature 357:563-9.

McLaughlin SK, Margolskee R. 1994. The sense of taste. Ameri Scient. 82:538.

McNeil KL, Sanders TH. 1998. Maturity effects on sensory and storage quality of roasted Virginia-type peanuts. J Food Sci 63:366-69.

McWatters KH, Young CT 1978. Quality and compositional characteristics of stabilized, unstabilized, and imitation peanut butter. J Food Sci 43:370-84

Meilgaard M, Civille GV, Carr BT. 1999. Spectrum terminology for descriptive analysis. 3rd Ed. CRC Press: Boca Raton, FL. pp. 177-94.

Muego KF, Resurreccion AVA, Hung Y-C.1990. Characterization of the textural properties of spreadable peanut based products. J Texture Stud. 21:61-73.

Müller O, Krawinkel M. 2005. Malnutrition and health in developing countries. CMAJ 173:279-86.

Murray JM, Delahunty CM, Baxter IA. 2001. Descriptive sensory analysis: past, present and future. Food Res Int 34:461-71.

Nelson KA, Labuza TP. 1992. Relationship between water and lipid oxidation rates: water activity and glass transition theory. In lipid oxidation in food. St. Angelo, A.J. Ed. ACS Symposium Series. Washington, DC. American Chemical Society. 93-103.

Nepote V, Grosso NR, Guzman CA. 2002. Extraction of antioxidant components from peanut skins. Gras Aceit 54: 391–5.

Nepote V, Mestrallet MG, Grosso NR. 2004. Natural antioxidant effect from peanut skins in honey-roasted peanuts. J Food Sci 69:S295-S300.

Nepote, V., Grosso, N.R., and C.A. Guzmán. 2005. Optimization of extraction of phenolic antioxidants from peanut skins. J Sci Food Agric 85:33-38.

Neta ERDC. 2010 Application of sensory and instrumental analyses for the characterization of flavor in roasted peanuts and farmstead cheddar cheese. Dissertation. North Carolina State University. p.1-265.

Newell JA, Mason ME, Matlock RS. 1967. Precursors of typical and atypical roasted peanut flavor. J Agric Food Chem 15:767-72.

Noble AC. 1994. Bitterness in wine. Physiol Behav. 6:1251-5.

Norden AJ, Gorbet DW, Knauft DA, Young CT. 1987. Variability in oil quality among peanut genotypes in the Florida breeding program. Peanut Sci 14:7-11.

North Carolina Sweetpotato Commission. 2010. NC sweetpotato fun facts and industry stats. Available at: http://www.ncsweetpotatoes.com/sweet-potato-facts/nc-sweet-potato-fun-facts- and-industry-stats.html. Accessed on September 14, 2011.

O’Keefe SF, Wang H. 2006. Effects of peanut skin extract on quality and storage stability of beef products. Meat Sci 73:278-86.

Oupadissakoon C, Young CT. 1984. Changes in free amino acids and sugars of peanuts during oil roasting. Peanut Sci. 11:6-9.

Özcan M, Seven S. 2003. Physical and chemical analysis and fatty acid composition of peanut, peanut oil and peanut butter from ÇOM and NC-7 cultivars. Gras Aceit 54:12-8.

Padda MS, Picha DH. 2008. Phenolic acid composition and antioxidant capacity of different heat-processed forms of sweetpotato cv. ‘Beauregard’. Inter J Food Sci Technol 43:1404–9.

Pereira DM, Valentão P, Pereira JA, Andrade PB. 2009. Phenolics: From chemistry to biology. Molecules 14:2202-11.

Patel MP, Sandige HL, Ndekha MJ, Briend A, Ashorn P, Manary MJ. 2005. Supplemental feeding with ready-to-use therapeutic food in Malawian children at of malnutrition. J Health Popul Nutr.23:351-7.

Patel RP, Patel MP, Suthar AM. 2009. Spray drying technology: an overview. Indian J Sci Technol 2:44-7.

Pattee HE, Purcell AE. 1967. Carotenoid pigments of peanut oil. J Amer Oil Chem Soc 44:328-30.

Patte HE, Purcell AE, Johns EB. 1969. Changes in carotenoid and oil content during maturation of peanut seeds. J Amer Oil Chem Soc 16:629-31.

Patte HE, Giesbrecht FG, Isleib TG. 1999. Sensory attribute variation in low-temperature- stored roasted peanut paste. J Agric Food Chem 47:2415-20.

Pinstrup-Anderson P, Burger S, Habicht JP, Peterson K. 1993. Protein-energy malnutrition. In: Jamison DT, Mosley WII, Measham AR, Bobnadilla JL. editors. Disease control priorities in developing countries. 2nd ed. Oxford University Press. Oxford (UK). pp.391-420.

Pominski L, Hamlet CA. 1991. Peanut butter quality as affected by inclusion of non-fat dry milk components. Peanut Sci 18:16-8.

Porto PA, Lde S, Laranjinha JAN, de Freitas, VAP. 2003. Antioxidant protection of low density lipoprotein by procyanidins: structure/activity relationships. Biochem Pharama 66:947-54.

Prior RL, Hoang H, Gu L, Bacchiocca M, Howard L, Hampsch-Woodhill M, Huang D, Ou B, Jacob R. 2003. Assays for Hydrophilic and Lipophilic antioxidant capacity (oxygen radical absorbance capacity (ORAC)) of plasma and other biological and food samples. J Ag Food Chem 51:3273-9.

Puupponen-Pimiä R, Nohynek L, Meier C, Kähkönen M, Heinonen M, Hopia A, Oksman- Caldentey K-M. 2001. Antimicrobial properties of phenolic compounds from berries. J Applied Micro 90:494-507.

Rao VR, Murty UR. 1994. Botany-morphology and anatomy. In The Groundnut Crop. Chapman and Hall. London, UK. pp.43-95.

Reed KA, Sims CA, Gorbet DW, O’Keefe SF. 2002. Storage water activity affects flavor fade in high and normal oleic peanuts. Food Res Int 35:769-74.

Rimando AM, Cuendet M, Desmarchelier C, Mehta RG, Pezzuto JM, Duke, SO. 2002. Cancer chemopreventive and antioxidant activities of pterostilbene, a naturally occurring analogue of resveratrol. J Agric Food Chem 50:3453-7.

Robichaud JL, Noble AC. 1990. Astringency and bitterness of selected phenolics in wine. J Sci Food Agric 53: 343-53.

Romeu-Nadal M, Chávez-Servín JL, Castellot AI, Rivero M, López-Sabater MC. 2007. Oxidation stability of the lipid fraction in milk powder formulas. Food Chem 100: 756-63.

Rubico SM, McDaniel MR. 1992. Sensory evaluation of acids by free-choice profiling. Chem Senses 17:273–89.

Sanders TH. 1977. Changes in tannin-like compounds of peanut fruit parts during maturation. Peanut Sci 4:51-3.

Sanders TH. 1979. Effect of variety, location and year on tannin content of peanut seed coats. Peanut Sci 6:62.

Sanders TH, Schubert AM, Pattee HE. 1982. Maturity methodology and postharvest physiology. In: Peanut Science and Technology; Patte HE, Young CT, Eds. Am. Research and Educ. Society. Yoakum, TX. pp. 624-54.

Sanders TH, Vercellotti JR, Crippen KL, Civille GV. 1989a. Effect of maturity on roast rolor and descriptive flavor of peanuts. J Food Sci 54:475-7.

Sanders TH, Vercellotti JR, Blankenship PD, Crippen KL, GV. Civille. 1989b. Interaction of maturity and curing temperature on descriptive flavor of peanuts. J Food Sci 54:1066-9.

Sanders TH, Bett KL. 1995. Effect of harvest date on maturity distribution and flavor of florunner peanuts. Peanut Sci 22:124-29.

Sanders TH, McMichael RW, Hendrix KW. 2000. Occurrence of resveratrol in edible peanuts. J Agric Food Chem 48:1243-46.

Schirack AV, Drake MA, Sanders TH, Sandeep KP. 2006. Characterization of Aroma-Active Compounds in Microwave Blanched Peanuts. J Food Sci 71:C513-20.

Schwambach SL, Peterson DG. 2006. Reduction of stale flavor development in low-heat skim milk powder via epicatechin addition. J Agric Food Chem 54:502-8.

Shahidi F. 2000. Lipids in flavor formation. In Flavor Chemistry. Risch SJ, Ho CT (eds). Industrial and Academic Research. ACS Symposium Series 756. Washington, DC: American Chemical Society. pp 24-43

Shahidi F, Zhong Y. 2010. Lipid oxidation and improving the oxidative stability. Chem Soc Rev 39:4067-79.

Shin E-C, Huang Y-Z, Pegg RB, Phillips RD, Eitenmiller RR. 2009. Commercial runner peanut cultivars in the United States: Tocopherol Composition. J Agric Food Chem 57:10289-95.

Seung CL, Seok MJ, So YK, Hae RP, Nam KC, Ahn DU. 2006. Effect of far-infrared radiation and heat treatment on the antioxidant activity of water extracts from peanut hulls. Food Chem 94: 489–93.

Shewfelt AL, Young CT. 1977. Storage stability of peanut-based foods: A review. J Food Sci 42:1148-52.

Shut CK, Waller GR. 1971. Volatile components of roasted peanuts: Comparative analyses of the basic fraction. J Food Sci 36:579-83.

Simonne AH, Kays SJ, Koehler PE, Eitenmiller RR. 1993. Assessment of β-carotene in sweetpotato breeding lines in relation to dietary requirements. Food Comp Anal 6:336-45.

Sobolev VS, Cole RJ. 1999. Note on utilization of peanut seed testa. J Sci Food Agric 84:105-11.

Spadaro JJ. 1979. Uses of defatted and partially defatted peanut flours. J Am Oil Chemist Soc 56:474-5.

Spielman AL. 1998. Gistducin and its role in taste. J Dent Res 77:539-44.

St. Angelo AJ, Kuck JC, Ory RL. 1975. Role of lipoxygenase and lipid oxidation in quality of oilseeds. J Agric Food Chem 27:229-34.

Stalker HT, Simpson CE. 1995. Germplasm resources in Arachis. In Advances in Peanut Science. Pattee HE and Stalker HT (eds). Amer Peanut Res Educ Soc. Stillwater, OK, pp 14-53.

Steffe JF. 1992. Introduction to Rheology. In Rheological Methods in Food Process Engineering. 2nd ed. Freeman Press. East Lansing, MI. pp. 1-48.

Sun A, Gunasekaran S. 2009. Measuring rheological characteristics and spreadability of soft foods using a modified squeeze-flow apparatus. J Texture Stud 40:275-87.

Sun JB, Severson RF, Schlotzhauer WS, Kays SJ. 1995. Identifying critical volatile in the flavor of baked “Jewel” sweetpotatoes [Ipomoea batatas (L.) Lam]. J Amer Soc Hort Sci 120:468-74.

Szczesniak AS, Brandt MA, Friedman H. 1963. Development of standard rating scales for mechanical parameters of texture and correlation between the objective and the sensory methods of texture evaluation. J Food Sci 28:397.

Thorngate JH, Noble AC. 1995. Sensory evaluation of bitterness and astringency of 3R(-) – epichatechin and 3S(+)-catechin. J Sci Food Agric 67:532-5.

Tomlins K, Ndunguru G, Stambul K, Joshua N, Ngendello T, Rwiza E, Amour R, Ramadhani B, Kapande A, Westby A. 2007. Sensory evaluation and consumer acceptability of pale-fleshed and orange-fleshed sweetpotato by school children and mothers with preschool children. J Sci Food Agric 87:2436-46.

Totlani VM, Chinnan MS. 2007. Effect of stabilizer levels and storage conditions on texture and viscosity of peanut butter. Peanut Sci 34:1-9.

Tsai T-T, Chiou RY-Y. 1989. Characterization of peanut proteins during roasting as affected by initial moisture content. J Agric Food Chem 37:1377-81.

USDEC. United States Dairy Export Council. 2011. Reference Manual for US Milk Powders: 2005 Revised Edition. http://www.usdec.org/home.cfm?navItem Number=82205. Accessed on September 14, 2011.

Van Jaarsveld PJ, Marais DW, Harme E, Nestel P, Rodriguez-Amaya DB. 2006. Retention of β-carotene in boiled, mashed orange-fleshed sweet potato. J Food Comp Anal 19:321-9.

Velioglu YS, Mazza G, Gao L, Oomah BD.1998. Antioxidant Activity and Total Phenolics in Selected Fruits, Vegetables, and Grain Products. J Agric Food Chem 46:4113-7.

Verdurmen REM, de Jong P. 2003. Optimising product quality and process control for powdered diary products. In Diary Processing: Improving quality. Gerrit. Smit (ed) CRC Press, Inc. Boca Rotan, FL pp. 333-65.

Wadsworth JI, Ziegler GM, Gallo AS, Spadaro JJ. 1967. Instant sweetpotato flakes- Factors affecting film thickness on double-drum drier. Trans ASAE. 157-63.

Walstra P, Geurts TJ, Noomen A, Jellema A, van Boekel MAJS. 1999. Concentration process. In: Diary Technology. Marcel Dekker, Inc. New York, NY. pp.265-265-319.

Walter WM Jr, Catignani GL, Yow LL, Porter DH. 1983. Protein nutritional value of sweetpotato flour. J Agric Food Chem 31:947-9.

Walter WM, Purcell AE. 1974. Lipid autoxidation in precooked dehydrated sweetpotato flakes stored in air. J Agr Food Chem 2:298-302.

Walter WM, Purcell AE, Hansen AP. 1972. Autoxidation of dehydrated sweetpotato flakes. The effect of solvent extraction on flake stability. J Agr Food Chem 20:1060-3.

Wang L, Tu Y-C, Lian T-W, Huang JT, Yen JH, Wu MJ. 2006. Distinctive antioxidant and anti-inflammatory effects of flavonols. J Agric Food Chem 54:9798-804.

Wang Y, Kays SJ. 2000. Contribution of volatile compounds to the characteristic aroma of baked “Jewel” sweetpotatoes [Ipomoea batata (L) Lam.} J Amer Soc Hort Sci 13:89-94.

Warner KJH, Dimick PS, Ziegler GR, Mumma RO, Hollender R. 1996. ‘Flavor-fade’ and off-flavors in ground roasted peanuts as related to selected pyrazines and aldehydes. J Food Sci 61:469-72.

Williams EJ, Drexler JS. 1981. A non-destructive method for determining peanut pod maturity. Peanut Sci 8:134-41.

Williams JP, Duncan SE, William RC, Mallikarjunan K, Eigel WN, O’Keefe SF. 2006. Flavor fade in peanuts during short-term storage. J Food Sci 71:S265-9.

Woodroof JG. 1983. Historical Background. In Peanuts: production, processing, products. Woodroof JG (ed) 3rd Ed. Avi Publishing Company, Inc. Roslyn, NY. pp. 1-18.

Woolfe J. 1992. Sweepotato: an untapped food resource. Cambridge Univ. Press. Cambridge, UK. pp.41-117.

World Health Organization. 1999. Management of severe malnutrition: A manual for physcians and other senior health workers. WHO: Geneva. pp.1-67.

World Health Organization/World Food Programme/United Nations System Standing Committee on Nutrition/The United Nations Children’s Fund. 2007. Community-based management of severe malnutrition. pp. 1-8.

Wu X, Gu L, Holden J, Haytowitz DB, Gebhardt SE, Beecher G, Prior RL. 2004. Development of a database for total antioxidant capacity in foods: a preliminary study. J Food Comp and Anal 17:407:22.

Young CT. 1973. Automated colorimetric measurement of free arginine in peanuts as a means to evaluate maturity and flavor. J Agr Food Chem 21:556-8.

Young, C.T., and Mason, M.E. 1972. Free arginine content of peanuts (Arachis hypogea L.) as a measure of seed maturity. J. Food Sci. 37: 722-725.

Young CT, Matlock RS, Mason ME, Waller GR. 1973. Effect of harvest date and maturity upon free amino acid levels in three varieties of peanuts. J Amer Oil Chem Soc 51:269-73.

Yu J, Ahmedna M, Goktepe I. 2005. Effects of processing methods and extraction solvents on concentration and antioxidant activity of peanut skin phenolics Food Chem 90:199.

Yu J, Ahmedna M, Goktepe I, Dai J. 2006. Peanut skin procyanidins: composition and antioxidant activities as affected by processing. J Food Comp Anal 19:364-71.

Zhang T, Oates CG. 1999. Relationship between α-amylase degradation and physico-chemical properties of sweet potato starches. Food Chem 65:157-63.

CHAPTER 2

Flavor and Antioxidant Capacity of Peanut Paste and Peanut Butter Supplemented with Peanut Skins

Chellani S. Hathorn1 and Timothy H. Sanders2,3

1Department of Food, Bioprocessing, Nutrition Sciences

North Carolina State University

Raleigh, NC 27695

2USDA-ARS-MQHRU

Department of Food, Bioprocessing, Nutrition Sciences

North Carolina State University

Raleigh, NC 27695

3Corresponding Author: Timothy H. Sanders

236 Schuab Hall, Campus Box #7624

Raleigh, NC 27695

Telephone: (919) 515-6312

Fax: (919) 513-8023

E-mail: [email protected]

Short title: Flavor and antioxidant of peanut products….

Journal section: Sensory and Food Quality

The content of this chapter has been published in Journal of Food Science

Abstract

Peanut skins (PS) are a good source of phenolic compounds. This study evaluated antioxidant properties and flavor of peanut paste and peanut butter enhanced with peanut skins. PS were added to peanut paste and peanut butter in concentrations of 0.0, 0.5, 1.0, 5.0,

10.0, 15.0 and 20.0 % (w/w). PS, peanut paste, and peanut butter used in the study had initial total phenolics contents of 158, 12.9, and 14.1 mg GAE/g, respectively. Hydrophilic oxygen radical absorbance capacity (H-ORAC) of peanut skins was 189,453 µMol Trolox/100g and addition of 5% PS increased H-ORAC of peanut paste and peanut butter by 52-63%.

Descriptive sensory analysis indicated that the addition of 1 % PS did not change intentsity of descriptors in the sensory profile of either peanut paste or peanut butter. Addition of 5%

PS resulted in significant differences in woody, hulls, skins; bitter; and astringent descriptors and 10% PS addition resulted in significant differences in most attributes toward more negative flavor.

Keywords: peanut, peanut butter, peanut skins, antioxidants, descriptive sensory analysis

Practical Application: Peanut skins are a low-value residue material from peanut processing which contain naturally occurring phenolic compounds. The use of this material to improve antioxidant capacity and shelf life of foods can add value to the material and improve the nutritional value of foods. The improved nutritional qualities and unchanged flavor profile occurring with low levels of peanuts skins in peanut paste and peanut butter suggest potential application of this technology in various food industries.

Introduction

Peanuts are an important crop in many parts of the world. Recent data suggest that production of peanuts in the United States (US) is about 2 million tons (USDA 2011).

Peanut skins (testae or seed coat), comprising about 3.0% (w/w) of a peanut seed, are low- value, residue materials resulting from peanut blanching and roasting. Removal of the skin is normally done in preparation for the production of products such as peanut butter.

Approximately 60,000 tons of peanut skins are accumulated annually in the U.S. as a result of peanut processing. Peanut skin use is generally limited to animal feeds (Ha and others

2007; Nepote and others 2004). The potential exists for value added use of this material to improve antioxidant capacity and shelf-life of lipid-containing foods.

The concentration of peanut skin tannins from six varieties of peanuts ranged from

289 to 468 mg/g (Sanders 1979). Karchesy and Hemingway (1986) reported 17% (w/w) procyanidins in peanuts skins. Lou and others (1999) identified six A-type procyanidins from the water-soluble fraction of peanut skin extracts. Procyanidins and other phenolic compounds may provide protection against oxidative stress, which has been implicated in atherosclerosis, diabetes mellitus, chronic inflammation, and some types of cancers in humans (Karadag and others 2009). Procyanidins were reported to have an antihyperglycemic effect in rats with induced diabetes (Pinent and others 2004; El-Alfy and others 2005). Plasma cholesterol levels were reduced in rats fed a diet containing procyanidins (Osakabe and Yamagishi 2009; Shimizu-Ibuka and others 2009). Further,

Frankel (1998) reported that phenolic compounds may reduce lipid oxidation in lipid-

containing foods. O’Keefe and Wang (2006) evaluated the effect of extracts from peanut skins on the storage stability of ground beef (250 g), and found that 200-400 ppm was the optimal concentration of extract to reduce lipid oxidation. Nepote and others (2004) observed that the addition of peanut skin extracts to honey roasted peanuts provided some protection against lipid oxidation.

The oxygen radical absorbance capacity assay (ORAC) is used to determine the inhibition of peroxyl radical induced oxidation in food and biological materials (Karadag and others 2009). ORAC specifically measures peroxyl radical quenching of fluorescence of fluorescein. Ballard and others (2009) reported the ORAC of peanut skins to be as high as

214,900 µMol Trolox/100g. Davis and others (2010) reported that ORAC of peanut skins increased with increased degree of roast.

Descriptive sensory analysis (DSA) is a powerful and comprehensive tool used in sensory science to generate quantitative and qualitative data. DSA can be used to evaluate quality control parameters, test the effects of ingredients, aid in evaluating processing methods, and can be correlated with other sensory data (Meilgaard and others 1999; McNeil and others 2002; Drake 2007). The peanut lexicon developed by Johnsen and others (1988), with an addition by Sanders and others (1989), describes both desirable and undesirable flavors of peanuts. This lexicon has been used as a common communication tool among researchers and various segments of the peanut industry. The peanut lexicon has been used extensively to relate flavor to maturity and curing (Sanders and others 1989), evaluate roast peanut flavor (Bett and others 1994), off-flavor development during storage (Pattee and

others 1999), and processing effects (Schirack and others 2006). Because of the antioxidant activity of phenolic compounds, the goal of this research was to evaluate the nutritional antioxidant properties and flavor of peanut paste and peanut butter enhanced with peanut skins.

Materials and Methods

Roasted peanut skins, peanut paste and peanut butter were obtained from Jimbo’s

Jumbos Inc. (Edenton, NC). Peanut paste was ground peanuts only while peanut butter contained added salt, sugar, and stabilizer which were less than 5% (w/w) of the peanut butter. Skins were milled using a laboratory grade Wiley Mill (Paul N. Gardner Company,

Inc., Pompano Beach, FL) fitted with a 0.5 mm sieve. Peanut skins were blended into peanut paste and peanut butter in concentrations of 0%, 0.5%, 1.0%, 5.0%, 10.0%, 15.0% and 20.0% w/w. All samples were prepared in triplicate and all subsequent analyses were performed in triplicate.

Color

Hunter L (0=black, 100=white) and a (+value=red, -value=green) value of all samples were determined using a HunterLab Colorimeter (HunterLab DP-9000™ Reston,

VA, USA).

Sample Extraction

Samples for lipophilic and hydrophilic ORAC analyses were extracted using a

Dionex (Sunnyvale, CA) ASE® 200 Accelerated Solvent Extractor (Prior and others 2003;

Davis and others 2010). Approximately 1.0 g of sample was weighed analytically and mixed with 25 g of clean sand. Samples and sand were transferred to a 22 ml extraction cell and extracted with 1:1 hexane:dichloromethane for lipophilic analysis. Lipophilic extracts

(peanut skins) were dried using nitrogen and adjusted to a final volume of 10 mL using acetone. Samples and sand were then extracted with 70:29.5:0.5 acetone:water:acetic acid

(AWA) and brought to 50 mL final volume with additional AWA in preparation for hydrophilic analysis.

Total Phenolics

Total phenolics were determined using the Folin-Ciocalteu method (Waterhouse,

2002) with modifications. Gallic acid (Sigma-Aldrich Co., St. Louis, MO) standards were prepared with 0.0, 50.0, 100.0, 150.0, 250.0, 500.0 mg/L. Approximately 0.1 mL of standard solution and hydrophilic extract samples were pipetted into test tubes, to which 7.9 mL of deionized water and 0.5 mL Folin reagent (Sigma-Aldrich Co., St. Louis, MO) were added to each of the standard and sample solutions. After 1 minute, 1.5 mL of sodium carbonate solution was added followed by vortexing. The sodium carbonate solution was prepared with

100g anhydrous sodium carbonate in 400 mL water, which was allowed to sit for 24 hours, filtered and brought to a final volume of 1 liter. Standards and samples remained at room temperature for 2h followed by absorbance measurement at 765 nm using a SAFIRE2 microplate reader equipped with version 6.1 Magellan reader software (Tecan US; Raleigh,

NC). Total phenolics were calculated as milligrams of gallic acid equivalents per gram (mg

GAE/g).

Hydrophilic - ORAC

Hydrophilic ORAC was determined using the procedure described by Prior and others (2003) and Davis and others (2010). All solutions and samples were prepared using pH 7.4 phosphate buffer. Solutions of 3.12, 6.25, 12.5, 25 and 50M of Trolox (Aldrich;

Milwaukee, WI) were used as the control standards. Approximately 130µL of standards and hydrophilic extracts were added to a Costar polystyrene flat-bottom black 96 microwell plate

(Corning; Acton, Massachusetts). Sixty micro liters of a 70 nM fluorescein (FL) solution was then added to the wells and then the plate was incubated in the SAFIRE2 for 15 min at

37 ºC. Next 60 L of a 153 mM 2,2’-azobis (2-amindino-propane) dihydrochloride (AAPH)

(Wako; Richmond, VA) solution was added. Fluorescence excitation and emission wavelengths were programmed to 485 and 535 nm. H-ORAC was calculated using a regression equation between the concentration of Trolox and the net area under the curve. H-

ORAC was reported as Trolox equivalents (µMol Trolox /100g).

Lipophilic - ORAC

The L-ORAC procedure was carried out as described by Prior and others (2003) and

Davis and others (2010). A 7% randomly methylated beta cyclodextrin (RMCD)

(Trappsol®; CTD, Inc.; High Springs, FL, USA) solution was prepared in 50% acetone: 50% water (7% RMCD). Solutions of standards prepared in 7% RMCD ranged from 200 to 1.56

M of Trolox (Aldrich; Milwaukee, WI). Approximately 21.5 nM of FL and 77 mM of

AAPH were prepared with phosphate buffer. Twenty-five micro liters of standards and

lipophilic extracts were added to the 96-microwell plate. One hundred twenty micro liters of

21.5 nM of fluorescein solution (prepared in 75 mM phosphate buffer), was added to samples using a multi-channel pipette and incubated in the SAFIRE2 for 15 min at 37°C. 80µL of a

70 mM AAPH solution was added using a multi-channel pipette. L-ORAC was calculated using a regression equation between the concentration of Trolox and the net area under the curve. L-ORAC was reported as Trolox equivalents (µMol Trolox /100g).

Descriptive Sensory Analysis

Evaluation of peanut paste and peanut butter samples were conducted by an experienced descriptive sensory panel (n=6, females; n=6, males; >500h experience) established using the Spectrum™ universal 15-point intensity scale. Panelists used the peanut lexicon described by Johnsen and others (1988) and Sanders and others (1989). To mask color differences of the various concentrations of peanut skins, all samples were presented in 2oz soufflé cups with lids under red lamps. Samples were equilibrated to and served at room temperature (22C).

A carboxymethyl cellulose (CMC) rinse (5.5 g/L, TIC Gums, Belcamp, MD) protocol described by Beecher and others (2008) was used to minimize astringency carryover effects.

Panelists were trained an additional 7 hours on bitter, astringency and familiarization with the

CMC protocol. After tasting the sample and expectorating, panelists were instructed to rinse with CMC, take a sip of water, a bite of cracker, then another sip of water followed by a two- minute timed waiting period. A peanut paste reference (Table 1), containing 2% w/w tannic

acid, was used as a warm-up sample prior to evaluating the test samples. Tannic acid, a plant polyphenol was added for increased bitter and astringent intensity in the reference peanut paste. The order of sample presentation was randomized for 4 replications and all samples were coded with random three digit codes.

Statistical Analysis

Analysis of variance (ANOVA) was generated using PROC GLM and comparison of means were made using Duncan’s post-hoc test (SAS version 9.1, Cary, NC). Significance was established at P<0.05.

Results and Discussion

The L value for peanut paste and peanut butter containing 0-20% (w/w) peanut skins ranged from 49.5 to 29.2 and 48.5 to 25.3, respectively (Table 2 and 3). The increase in darkness of both peanut products was directly related to the quantity of skins added. The a value, (green to magenta), increased significantly (P<0.05) from 7.1 to 8.7 and 8.5 to 9.6 for peanut paste and peanut butter, respectively, as the concentration of peanut skins increased

(Table 3). Stansbury and others (1950) reported dark red extracts from peanut skins.

Chukwumah and others (2009) evaluated the relationship between peanut skin color and polyphenolic compounds. The redness of the peanut skin extracts correlated well with the concentration of polyphenolic compounds present in the skins. Compounds, such as procyanidins, found primarily in woody or herbaceous plants, are colorless but may convert under atmospheric conditions or under light to red-brown pigments (Schwartz 1996). As such, differences in a value were observed in a collected sample of raw skins (a = 8.8) and roasted (a = 10.2) peanut skins, demonstrating an increase in red pigments in roasted skins.

The total phenolic content of peanut skins used in the present study was 158 mg

GAE/g. Peanut skin total phenolics have been reported to range from 36 to 280 mg GAE/g

(Francisco and Resurreccion 2008). Ballard and others (2009) reported that the total phenolic content of an ethanolic extract of peanut skins was 118 mg GAE/g. The phenolic content of roasted peanut skin extracts using water, methanol and ethanol solvents was 79.0,

96.7 and 125.0 mg GAE/g , respectively (Yu and others 2005). Yu and others (2006) published total phenolics results of peanut skins removed using direct peel, water blanching,

and roasted. Peanuts skins removed by direct peeling (130.8 mg GAE/g) and after roasting

(124.3 mg GAE/g) had higher antioxidant activity than skins removed by water blanching

(15.1 mg GAE/g). Water blanching of the skin causes phenolic compounds to leach out, resulting in a loss of skin color (Yu and others 2005). Based on data from our study, more than 95% of the phenolic compounds in peanut skins are hydrophilic and supports the concept of possible leaching of these compounds when skins are removed with water blanching. The total phenolic content of peanut paste increased (P<0.05) from 12.9 to 31.9 mg GAE/g across the range of PS added (Table 4) and from 14.1 to 28.1 mg GAE/g in peanut butter samples when peanut skins were added (Table 5). Currently, literature suggests that Americans consume about 1g of phenolic compounds daily; however, there is no recommended daily dietary intake of phenolic compounds (Scalbert and Williamson 2000;

Williamson and Holst 2008).

The lipophilic ORAC of roasted peanut skins was 5,617±223 µMol Trolox/100g, while the hydrophilic ORAC was 189,453±6963 µMol Trolox/100g (Table 4). These values are similar to data published by Ballard and others (2009) and Davis and others (2010).

ORAC reported for roasted almond skins ranged from 80,300 – 108,000 µMol Trolox/100g

(Garrido and others 2008). Davis and others (2010) found that roasting for longer times increased antioxidant activity of peanuts, peanut flour and skins. Increase in ORAC with roasting may be related to an increase in Maillard browning products which, in addition to phenolic compounds, have antioxidant activity.

The USDA (2007) report of the ORAC of approximately 277 foods included raw peanuts as 3,166 µMol Trolox/100g. The report does not state whether or not the skin was intact; however, based on published data, it is likely that the skin was removed. Davis and others (2010) reported slightly higher ORAC for raw blanched peanuts (3,750 µMol

Trolox/100g). ORAC for almonds was reported as 4,454 µMol Trolox/100g (USDA 2007).

The antioxidant activity of nuts is reduced when the skin is removed (Schmitzer and others

2011). Nuts such as walnuts and pecans, typically eaten with the skin intact, have an ORAC of 13,541 and 17,940, µMol Trolox/100g, respectively (USDA 2007). The ORAC of peanut paste and peanut butter increased as the concentration of peanut skins increased, and ranged from 4,041 to 20,063 and 5,702 to 20,376 µMol Trolox/100g. According to the USDA database (2007), peanut butter has an antioxidant capacity of 3,432 µMol Trolox/100g. The differences in antioxidant capacity for peanut butter in the database and the value in this study (5,702 µMol Trolox/100g) may be related to peanut variety, process parameters, and/or composition of the two products.

The addition of peanut skins at greater than 5% resulted in a decrease in roast peanutty intensity and an increase in woody/hulls/skins, bitter, and astringency intensities in peanut paste and peanut butter (Tables 5 and 6). The panel did not detect differences

(P>0.05) in roast peanutty intensity among samples through 5.0% skins in peanut paste but differences were detected at 10% added skins (Table 5). Similar results were found in peanut butter (Table 6). Peanut skin is approximately 3% of the weight of unblanched peanuts and

the results from this study provide evidence that slightly more than the normal peanut skin weight may be added to peanut paste without discernable difference in flavor.

Sweet aromatic is described as aromatics associated with sweet material such as caramel, vanilla, molasses and fruit (Johnsen and others 1988). For peanut paste, sweet aromatic did not become significantly (P<0.05) lower until 10% skins were added. The same was observed for peanut butter. Woody/hulls/skins is associated with base peanut character

(absence of fragrant top notes) and related to dry wood, peanut hulls and skins (Johnsen and others 1988). Since peanut skins are more closely associated with woody/hulls/skins, the addition of this material to peanut paste and peanut butter should and did increase base peanut and woody notes. Johnson (2007) used DSA to describe the flavor and aroma profile of almond skins as toasted and toasted:nutty, respectively. Panelists described the skins as having “mild flavor and aroma” but descriptors such as woody and earthy were also reported.

The authors suggested that almond skins could be added to foods with little change to flavor but did not report studies to demonstrate that fact.

Astringency intensity was higher in peanut paste than in peanut butter (Tables 5 and

6). Viscosity of a material has an effect on perceived astringency. Smith and others (1996) demonstrated that astringency intensity decreased as viscosity was increased in aqueous solutions of grape seed tannins using carboxymethyl cellulose (CMC). Similarly, Peleg and

Noble (1999) reported that perceived astringency decreased with increasing viscosity of cranberry juice using CMC. Also, pectin has been reported to reduce perceived astringency in catechin based solutions (Hayashi and others 2005). Buck (2010) measured the yield

stress, the minimum shear stress required to initiate flow, of peanut paste and commercial peanut butter. Peanut paste ranged from 0.58 – 2.02 kPa, while peanut butter had a yield stress of 10.61 kPa. Because of the addition of stabilizer to prevent oil separation, peanut butter is typically more viscous than peanut paste. The lower perceived astringency in peanut butter in the current study compared to peanut paste may be in part a result of increased viscosity. Sucrose has been previously demonstrated to reduce perceived astringency, likely due to concentration and viscosity of the test solution (Lyman and Green

1990; Breslin and others 1993).

Procyanidins have been identified as one of the key compounds in foods responsible for bitterness (Lopez and others 2007). The molecular structure of phenolic compounds can affect bitter and astringency perception. Low-molecular weight (< 500) phenolic compounds tend to be bitter, while higher-molecular (>500) compounds tend to be astringent (Lea and

Arnold 1978; Robichaud and Noble 1990; Noble 1994; Peleg and others 1999). Low molecular weight compounds with antioxidant activity found in peanut skins, such as caffeic acid (MW=180), chlorogenic acid (MW=354), ellagic acid (MW=302) and procyanidin monomers (MW=289) may be perceived as bitter (Peleg and others 1999; Yu and others

2005). While A-type procyanidin dimers (MW=575), B-type procyanidin dimers

(MW=577), A-type procyanidin trimers (MW=863), B-type procyanidin trimers (MW=865),

A-type procyanidin tetramers (MW=1149) and B-type procyanidin tetramers (MW=1151) found in peanut skins tend to be more astringent (Peleg and others 1999; Yu and others

2005). The mechanism of astringency perception may result from a decrease in saliva

lubrication caused by binding and precipitation of proteins (Charlton and others 2002; Jöbstl and others 2004). Increases in bitter and astringent may be described as unpleasant sensory attributes, which can negatively correlate with consumer liking (Young and others 2005).

Conclusion

The addition of peanut skins to peanut paste and peanut butter significantly increased the total phenolics and ORAC. Addition of 5% (w/w) of peanut skins to peanut paste and peanut butter resulted in overall reduced flavor. This study indicated a potential limited application for peanut skins in peanut paste and peanut butter and perhaps in other products to improve nutritional quality.

References

Ballard TS, Mallikarjuman P, Zhou K, O’Keefe SF. 2009. Optimizing the extraction of phenolic antioxidants from peanut skins using response surface methodology. J Agric Food Chem 57:3064-3072.

Beecher JW, Drake MA, Luck PJ, Foegeding EA. 2008. Factors regulating astringency of whey protein beverages. J Dairy Sci 91:2553-2560.

Bett KL, Vercellotti JR, Lovegran NV, Sanders TH, Hinsch RT, Rasmussen GK. 1994. A comparison of the flavor and compositional quality of peanuts from several origins. Food Chem 51:21-7.

Breslin PAS, Gilmore MM, Beauchamp GK, Green BG. 1993. Psychophysical evidence that oral astringency is a tactile sensation. Chem Senses 18: 405-417.

Buck VI. 2010. Rheological properties of peanut paste and characterization of fat bloom formation in peanut-chocolate confectionary. [DPhil dissertation]. Blacksburg, VA: Virginia Polytechnic Institute and State Univ. 137 p. Available from: Virginia Polytechnic Institute and State Univ. Libraries, Clemson, SC.

Charlton AJ, Baxter NJ, Khan ML, Moir AJ, Haslam E, Davies AP, Williamson MP. 2002. Polyphenol/peptide binding and precipitation. J Agric Food Chem 50:1593-1601.

Chukwumah Y, Walker LT, Verghese M. 2009. Peanut skin color: A biomarker for total polyphenolic content and antioxidative capacities of peanut cultivars. Int J Mol Sci 10: 4941-4952.

Davis JP, Dean LL, Price KM, Sanders TH. 2010. Roast effects on the hydrophilic and lipophilic capacities of peanut flours, blanched peanut seed and peanut skins. Food Chem. 119:539-547.

Drake MA. 2007. Invited review: Sensory analysis of dairy foods. J Dairy Sci 90:4925-4937.

El-Alfy AT, Ahmed AAE, Fatani AJ. 2005. Protective effect of red grape seeds proanthyocyanidins against induction of diabetes by alloxan in rats. Pharama Res 52: 264- 270.

Francisco MLDL, Resurreccion AVA. 2008. Functional components in peanuts. Crit Rev Food Sci Nutr 48:715-746.

Frankel EN. 1998. Lipid Oxidation. Scotland: The oily press ltd. 303p.

Garrido I, Monagas M, Gómez-Cordovés C, Bartolome, B. 2008. Polyphenols and antioxidant properties of almond skins: influence of industrial processing. J Food Sci 73: C106-C115.

Ha HV, Pokorny J, Sakurai H. 2007. Peanut skin antioxidants. J Food Lipids 14:298-314.

Hayashi N, Ujihara T, Kohata, K. 2005. Reduction of catechin astringency by the complexation of gallate-type catechins with pectin. Biosci Biotechnol Biochem 69:1306- 1310.

Jöbstl E, O’Connell J, Faircloough JP, Williamson MP. 2004. Molecular model for astringency produced by polyphenol/protein interactions. Biomacromolecules 52:942-949.

Johnsen PB, Civille GV, Vercellotti JR, Sanders TH, Dus, CA. 1988. Development of a lexicon for the description of peanut flavor. J Sensory Stud 3: 9-17.

Johnson HM. 2007. Almond skins as a natural antioxidant. [MSc thesis]. Clemson, SC: Clemson Univ. 86 p. Available from: Clemson Libraries, Clemson, SC.

Karadag A, Ozcelik B, Saner S. 2009. Review of methods to determine antioxidant capacities. Food Anal Methods 2:41-60.

Karchesy JJ, Hemingway RW. 1986. Condensed tannins (4β → 8;2β → O→ 7)-linked procyanidins in arachis hypogea L. J Agric Food Chem 34: 966-970.

Lea AGH, Arnold GM. 1978. The phenolics of ciders: Bitterness and astringency. J Sci Food Agric 29:478-483.

Lopez R, Mateo-Vivaracho L, Ferreira V. 2007. Optimization and validation of a taste dilution analysis to characterize wine taste. J Food Sci 72:S345-S351.

Lou H, Yamazaki Y, Sasaki T, Uchida M, Tanaka H, Oka S. 1999. A-type proanthocyanidins from peanut skins. Phytochem 54:297-308.

Lyman BJ, Green BG. 1990. Oral astringency: effects of repeated exposure and interactions with sweeteners. Chem Senses 15:151-164.

Mcneil KL, Sanders TH, Civille GV. 2002. Descriptive analysis of commercially available creamy style peanut butters. J Sensory Stud 17:391-414.

Meilgaard M, Civille GV, Carr BT. 1999. Sensory Evaluation Techniques. 3rd ed. New York: CRC Press. pp. 387.

Nepote V, Grosso NR, Guzman CA. 2004. Radical scavenging activity of extracts of argentine peanut skins (arachis hypogagea) in relation to its trans-resveratrol content. J Argentine Chem Soc 92:41-49.

Noble AC. 1994. Bitterness in wine. Physiol Behav 56:1251-1255.

O’Keefe SF, Wang H. 2006. Effects of peanut skin extract on quality and storage stability of beef products. Meat Sci 73:278-286.

Osakabe N, Yamagishi M. 2009. Procyanidins in theobroma cacao reduce plasma cholesterol levels in high cholesterol-fed rats. J Clin Biochem Nutr 45:131-136.

Pattee HE, Giesbrecht FG, Isleib TG. 1999. Sensory attribute variation in low-temperature stored roasted peanut paste. J Agric Food Chem 47:2415-2420.

Peleg H, Noble AC. 1999. Effect of viscosity, temperature and pH on astringency in cranberry juice. Food Qual Pref 10:343-347.

Peleg H, Gacon K, Schlich P, Noble AC. 1999. Bitterness and astringency of flavan-3-ol monomers, dimmers and trimers. J Sci Food Agric 79:1123-1128.

Pinent M, Blay M, Bladé MC, Salvadó MJ, Arola L, Ardévol, A. 2004. Grape seed-derived procyanidins have an antihyperglycermic effect in streptozocin-induced diabetes rats and insulinomimetic activity in insulin-sensitive cell lines. Endocrinol 145:4985-4990.

Prior RL, Hoang H, Gu LW, Wu XL, Bacchiocca M, Howard L, Hampschwoodill M, Huang DJ, Ou BX, Jacob R. 2003. Assays for hydrophilic and lipophilic antioxidant capacity (oxygen radical absorbance capacity (ORAC(FL)) of plasma and other biological and food samples. J Agric Food Chem 51:3273-3279.

Robichaud JL, Noble AC. 1990. Astringency and bitterness of selected Phenolics in wine. J Sci Food Agric 53:343-353.

Sanders TH. 1979. Effect of variety, location and year on tannin content of peanut seed coats. Peanut Sci 6:62-64.

Sanders T H, Vercellotti JR, Crippen KL, Civille GV. 1989. Effect of maturity on roast color and descriptive flavor of peanuts. J Food Sci 54:475-477.

Scalbert A, Williamson G. 2000. Dietary intake and bioavailability of polyphenols. J Nut 130: 2073S-2085S.

Schirack AV, Drake MA, Sanders TH, Sandeep KP. 2006. Impact of microwave blanching on the flavor of roasted peanuts. J Sensory Stud 21: 428-440.

Schmitzer V, Slatnar A, Veberic R, Stampar F, Solar A. 2011. Roasting affects phenolic composition and antioxidative activity of hazelnuts (Corylus avellana L.). J Food Sci 76:S14- 19.

Schwartz SJ, Von Elbe JH, Giusti MM. 1996. Colorants. In: Damodaran S, Parkin KL, Fennema OR, editors. Fennema’s Food Chemistry. 4th ed. Boca Raton, Fl.: CRC Press. p. 571-638

Shimizu-Ibuka A, Udagawa H, Kobayashi-Hattori K, Mura K, Tokue C, Takita T, Arai, S. 2009. Hypocholesterolemic effect of peanut skins and its fractions: a case record of rats fed on a high-cholesterol diet. Biosci Biotecnol Biochem 73: 205-208.

Smith AK, June H, Noble AC. 1996. Effects of viscosity on the bitterness and astringency of grape seed tannin. Food Qual Pref 7:161-166.

Stansbury MF, Field ET, Guthrue JD. 1950. The tannin and related pigments in the red skins (testa) of peanut kernels. Amer Oil Chem Soc 8:317-321.

United States Department of Agriculture (USDA). 2011. World agricultural production. Peanut area, yield and production. Table 13.

United States Department of Agriculture (USDA). 2007. Oxygen radical absorbance capacity (ORAC) of selected foods. p.1-34.

Waterhouse AL. 2002. Determination of total phenolics. In Current Protocols in Food Anal Chem I1.1-I1.1.8.

Williamson G, Holst B. 2008. Dietary reference intake (DRI) value for dietary polyphenols: are we heading in the right direction? British J Nutr 3: S55-S58.

Young ND, Sanders TH, Drake MA, Osborne J, Civille G. 2005. Descriptive analysis and US consumer acceptability of peanuts from different origins. Food Qual Pref 16:37-43.

Yu J, Ahmedna M, Goktepe I. 2005. Effects of processing methods and extraction solvents on concentration and antioxidant activity of peanut skin phenolics. Food Chem 90:199-206.

Yu J, Ahmedna M, Goktepe I, Dai J. 2006. Peanut skin procyanidins: composition and antioxidant activities as affected by processing. J Food Comp Anal 19: 364-371.

Acknowledgments

The authors thank the peanut sensory panel members and Ms. Kristin Price for meaningful contributions to this work.

Table 1. Intensity of attributes of peanut paste reference containing 2% tannic acid

Attribute Intensity Roast Peanutty 4.5 Sweet Aromatic 3.0 Dark Roast 3.0 Raw Beany 2.0 Woody/Hulls/Skins 3.0 Sweet Taste 2.5 Bitter 5.0 Astringency 4.0

Table 2. Hunter L and a value of peanut paste (PP) and peanut butter (PB) containing peanut skins (PS)

PP PB % PS L a L a 0 49.5 a 7.1 c 48.5 a 8.5 d 0.5 48.1 b 7.2 c 46.8 b 8.5 d 1.0 47.9 c 7.2 c 46.5 b 8.5 d 5.0 41.1 d 7.9 b 39.0 c 8.9 c 10.0 35.8 e 8.6 a 31.9 d 9.4 b 15.0 32.9 f 8.7 a 27.9 e 9.4 b 20.0 29.2 g 8.7 a 25.3 f 9.6 a aMeans in the same column with different letters are significantly different (P<0.05)

100

Table 3. Total phenolics (TP) and oxygen radical absorbance capacity (ORAC) of peanut a b b paste (PP) and peanut butter (PB) containing peanut skins (PS) c PP PB TP H-ORAC TP H-ORAC % PS (GAE/g) (µMol Trolox /100g) (GAE/g) (µMol Trolox /100g) 0 12.9 f 4,041 f 14.1 f 5,702 f 0.5 13.7 ef 4,625 f 14.7 f 6,059 ef 1.0 14.4 e 5,846 e 15.1 e 6,547 e 5.0 20.8 d 7,737 d 21.5 d 8,954 d 10.0 25.5 c 13,004 c 24.0 c 15,653 c 15.0 27.4 b 17,145 b 25.5 b 18,071 b 20.0 31.9 a 20,063 a 28.1 a 20,376 a aMeans in the same column with different letters are significantly different (P<0.05)

Table 4. Oxygen radical absorbance capacity (ORAC) of peanut skins

Samples µMol Trolox /100g± SD

Lipophilic 5,617±223

Hydrophilic 189,453±6963

Total 195,070

101

Table 5. Descriptive sensory analysis of peanut paste containing peanut skins

Conc. of PS Roast Sweet Dark Raw Woody/ Sweet Bitter Astringency Peanutty Aromatic Roast Beany Hulls/Skins Taste 0 4.4 a 3.0 a 3.0 a 2.0 a 3.2 a 2.7 ab 2.8 a 2.7 a 0.5 4.6 a 3.2 a 3.0 a 2.0 a 3.3 ab 2.9 a 2.9 ab 3.0 ab 1.0 4.3 a 3.0 a 3.0 a 2.0 a 3.4 ab 2.6 ab 2.9 ab 2.7 a 5.0 4.3 a 2.9 ab 2.9 a 2.1 a 3.6 b 2.5 b 3.3 b 3.4 b 10.0 3.5 b 2.6 b 2.9 a 2.2 a 4.6 c 2.2 c 4.3 c 4.4 c 15.0 2.7 c 2.2 c 2.8 a 2.1 a 4.9 c 1.9 d 4.6 c 4.3 c 20.0 1.9 d 1.6 d 2.9 a 2.4 b 5.7 d 1.5 e 5.3 d 5.3 d aMeans in the same column with different letters are significantly different (P<0.05)

Table 6. Descriptive sensory analysis of peanut butter containing peanut skins

Conc. of PS Roast Sweet Dark Raw Woody/ Sweet Bitter Astringency Peanutty Aromatic Roast Beany Hulls/Skins Taste 0 4.0 ab 2.9 ab 2.9a 1.9 b 3.0 ab 2.8 a 2.3 a 1.6 a 0.5 4.0 ab 3.0 ab 2.9 a 2.0 ab 2.9 a 3.0 a 2.5 ab 1.8 a 1.0 4.2 a 3.1 a 2.9 a 2.1 ab 3.3 ab 3.1 a 2.6 ab 1.9 a 5.0 3.9 ab 3.0 ab 2.9 a 2.2 a 3.6 b 2.8 a 3.1 b 2.6 b 10.0 3.5 b 2.7 bc 2.8 a 2.2 ab 4.3 c 2.7 a 4.1 c 3.5 c 15.0 2.8 c 2.5 c 2.9 a 2.2 a 4.3 c 2.6 a 4.4 c 3.7 c 20.0 2.2 d 2.0 d 2.9 a 2.1 ab 5.3 d 1.9 b 5.2 d 4.3 a aMeans in the same column with different letters are significantly different (P<0.05)

102

CHAPTER 3

Evaluation of peanut skins: potential to enhance shelf-life and oil characterization

103

Evaluation of peanut skins: potential to enhance shelf-life and oil characterization

Chellani S. Hathorn1, Lisa O. Dean2, Timothy H. Sanders2,3

1Department of Food, Bioprocessing, Nutrition Sciences North Carolina State University Raleigh, NC 27695

2USDA-ARS-MQHRU Department of Food, Bioprocessing, Nutrition Sciences North Carolina State University Raleigh, NC 27695

3Corresponding Author: Timothy H. Sanders 236 Schuab Hall, Campus Box #7624 Raleigh, NC 27695 Telephone: (919) 515-9108 Fax: (919) 513-8023 E-mail: [email protected]

The content of this chapter will be submitted for publication in the Journal of Food

Science

104

Abstract

Peanut skins contain high levels of phenolic antioxidant compounds. The initial objective of this study was to investigate the potential of peanut skins to enhance shelf life of peanut products. Milled peanut skins were added to peanut paste and peanut butter in concentrations of 0.0, 3.0, 10.0% (w/w) and then stored at 30°C for 24 wks. Oxygen radical absorbance capacity (ORAC) decreased similarly in all samples over time. Oxidative stability index (OSI) decreased and peroxide value (PV) increased to a greater degree in the samples with highest concentrations of peanut skins over 24 wk indicating a reduction in shelf life related to skin content. The resulting second objective of the work was to identify factors of skin that contributed to reduced peanut product shelf life. Fatty acid composition of peanuts and corresponding peanut skins were similar but OSI of skin oil was very low

(1.1-4.1 hrs) compared to peanut oil (9.8-15.7 hrs). Alpha tocopherol was very low to nondetectable in skin oil compared to peanut oil. High levels of Cu (39.6 to 56.0 ppm) and

Fe (107.7 to 206.1 ppm) were present in peanut skins, while significantly low levels were found in peanuts. Negligible levels of -tocopherol, low OSI, and high Cu and Fe levels in skins are all related to reduced shelf life of products and appeared to be responsible for reduced shorter shelf life when skins were added to peanut paste and peanut butter.

Keywords: peanut, peanut skins, peanut skin oil, tocopherols, heavy metals

105

Practical Application: The use of peanut skins in food applications demonstrate improved nutritional characteristics, and provides emerging knowledge regarding peanut skins to the peanut industry.

106

Introduction

Approximately 60,000 tons of peanut skins (seed coat or testae) are produced as a by- product of peanut processing and they are generally considered as waste products with limited use in animal feeds. Skin removal is done prior to further processing of peanuts to peanut butter and confectionery products. Peanut skins are an excellent source of phenolic compounds exhibiting high antioxidant activity. Proanthocyanidins, powerful antioxidants, have been identified in peanuts skins (Karchesy and Hemingway 1986; Lou and others

1999). Lou and others (2001) also isolated eight new flavonoids and two alkaloids from peanut skins. Yu and others (2005) reported three classes of compounds in peanut skins.

These compounds include flavonoids (epigallocatechin, epicatechin, catechin gallate and epicatechin gallate), phenolic acids (chlorogenic, caffeic acids, coumaric acid and ferulic acid), and stilbenes (resveratrol). Peanut skins contain 0.60-0.65 µg/g resveratrol (Sobolev and Cole 1999; Sanders and others 2000). Hathorn and Sanders (2012) reported that more than 95% of total ORAC was quantified in the hydrophilic ORAC in peanut skins.

Compounds found in peanut skins such as epigallocatechin and epicatechin, resveratrol, and proanthocyanidins are water-soluble antioxidants (Schwartz and others 1996), while tocopherols are lipid soluble antioxidants (Palozza and Krinsky 1992; Huang and others

1996).

Phenolic compounds (tannins) bind to and precipitate proteins and various other organic compounds including alkaloids. One of the main limitations for peanut skin use is

107

the high tannin content (289-462 mg tannin/g) (Sanders 1979). In animals, tannins inhibit digestion by binding with protein and interfering with absorption of nutrients (Hill 2002;

Hale and McCormick 1981; McBrayer and others 1983). Tannins are also reported to be responsible for the binding and precipitation of salivary proteins, causing astringency (Gawel

1998).

Peanuts contain approximately 48-50% oil. Peanut skins make up 3% w/w of the seed. Peanut skins contain as low as 4.9% (w/w) oil (Nepote and others 2002), while

Sobolev and Cole (2003) reported much higher depending on peanut type. The oil removed from peanut skins has a slight yellowish to light brown color. There is a lack of information on oil from peanut skins in the literature. Sobolev and Cole (2003) reported on the fatty acid content of oil from peanut skins. The initial objective of this study was to investigate the potential of peanut skins to enhance shelf life of peanut products. After finding a decrease in shelf life related to skin addition, the resulting second objective of the work was to identify compositional characteristics of skins that contributed to that decrease.

108

Materials and Methods

Peanut and peanut skins used in this study were collected from commercial facilities.

Peanut skins and peanut designations are given in Table 1. Skins were milled using a Wiley

Mill (Paul N. Gardner Company, Inc., Pompano Beach, FL) fitted with a 0.5 mm sieve. The skins were blended into peanut paste (PP) and peanut butter (PB) at 0%, 3% and 10% w/w.

Samples were stored at 30º C in a temperature controlled chamber (Isotemp® Incubator,

Model 304R, Fisher Scientific, Pittsburg, PA) and samples for analyses were collected at 0,

2, 4, 8, 16, and 24 weeks.

For oil removal from skins and peanuts, ground sample (100g) and hexane (400mL) were placed in beakers and stirred for 30 min. The solvent was decanted and then additional hexane was added to the skins and stirred for an additional 15 min. The total mixture was filtered through Whatman # 4 filter paper to remove solids. Hexane was removed from the extracted oil in a rotary evaporator (Büchi Corporation, New Castle, DE). The residual oil was placed in conical screw top tubes which were refrigerated until analyzed.

Oxygen Radical Absorbance Capacity (ORAC)

Milwaukee, WI) were used as the control standards. Approximately 130µL of standards and hydrophilic extracts were added to a Costar polystyrene flat-bottom black 96 microwell plate

109

(Corning; Acton, Massachusetts). Sixty µL of a 70 nM fluorescein (FL) solution was then added to the wells and then the plate was incubated in the SAFIRE2 for 15 min at 37 ºC.

Next 60 µL of a 153 mM 2,2’-azobis (2-amindino-propane) dihydrochloride (AAPH) (Wako;

Richmond, VA) solution was added. Fluorescence excitation and emission wavelengths were programmed to 485 and 535 nm. H-ORAC was calculated using a regression equation between the concentration of Trolox and the net area under the curve. H-ORAC was reported as Trolox equivalents (µMol Trolox /100g).

Oxidative Stability Index

Oxidative stability index (OSI) was determined according to AOCS Method12b-92

(AOCS, 2004). Five grams of oil was weighed into disposable glass tubes and placed into a heating block at 110ºC. Compressed air was introduced to the oils and volatiles generated by the oil were trapped in 50 ml of dionized water containing an electrode monitoring the conductivity. OSI was defined as the number of hours after placement in the heating block until the rate of increase in conductivity became exponential.

Peroxide Value

Peroxide value was determined in accordance to AOCS Cd 8-53 (AOCS, 2004).

Approximately, 5 grams of oil was titrated with sodium thiosulfate and a starch indicator.

PV is expressed as milliequivalents per kilogram of oil (meq/kg).

110

Fatty Acid Profile

Fatty acids were determined as fatty acid methyl esters according to AOCS Ce 2-66

(AOCS, 2004). Fatty acid methyl esters were analyzed using a Perkin Elmer Clarus 500 gas chromatograph (Shelton, CT) with an autosampler and a flame ionization detector (FID) fitted with a capillary column of 70% cyanopropyl polysilphenylene-siloxane as the stationary phase (30m length 0.25 mm i.d., 0.25 µm film thickness) (BPX-070, SGE

Analytical Science, Austin, TX, USA). The temperature gradient was programmed to increase from 60ºC with a 2 min hold time at 4ºC per min to 180ºC and then increased at

10ºC to a final temperature of 235º C. The total run time was 27.7 min. The injection was split at 150 mL/min. Helium was used as the carrier gas with a flow rate of 20 psi. The results were reported as percent of the total fatty acids based on peak areas (AOCS Ce 1f-96)

(AOCS, 2004). A standard mixture of fatty acid methyl esters (GLC-21A, NuChek Prep,

Elysian, MN) was run with each sample set to confirm retention times and identification of fatty acid methyl esters.

Tocopherol Analysis

Tocopherols were determined according to the method of Hashim and others (1993), by diluting oil samples with hexane. Samples were analyzed using a Luna 5-micron Silica column, 250 mm length, 4.60 mm i.d. Phenomenex (Torrance, CA) with a mobile phase of

1% isopropanol in hexane at a flow rate of 1.4 ml/min. The injection volume was 20 µl. A

Waters 2487 Dual Wavelength Absorbance Detector (Waters Corp., Milford, MA) was used

111

and set to a wavelength of 294 nm. The tocopherols were identified by comparison with standards purchased from Sigma (SigmaChemical, St. Louis, MO). Standards of , β, and

-tocopherol were diluted with hexane and their concentration was determined by the absorbance maximum of the solutions using UV spectroscopy according to Beer’s Law.

Concentrations of tocopherols in the samples were calculated using peak areas and the calculated concentrations of the standard solutions.

Heavy Metal Analysis

Cu and Fe were determined according to AOAC 985.01 using an Inductively Coupled

Plasma (ICP) Spectroscopic method (AOAC 1990). Samples were ashed in crucibles for 2 h at 500ºC and then cooled. Water was added to the dry ash using 10 drops of deionized water, and then 3-4 mL of 1:1 HNO3 was added carefully. HNO3 was evaporated by heating and then the sample was ashed an additional 1 h at 500ºC. Ash was dissolved in 10 mL HCl and brought to a final volume of 50 mL with deionized water. IPC emission, peak intensities at specific wavelengths were compared to a standard curve.

Statistical Analysis

Analysis of variance (ANOVA) was generated using PROC GLM and comparison of means were made using Duncan’s post-hoc test (SAS version 9.1, Cary, NC). Significance was established at P<0.05.

112

Results and Discussion

Antioxidant activity increased sequentially with the addition of peanut skins. ORAC for 0PP, 3PP and 10PP was 6374, 9571 and 12842 µMol Trolox/100g, respectively (Figure

1). After 24 wks of storage, there was a significant (P<0.05) decrease in antioxidant capacity for all peanut paste (PP) samples, ranging from 4253-10058 µMol Trolox/100g. For peanut butter samples, initial ORAC was 5970, 9024, 14591 µMol Trolox/100g for 0PB, 3PB and

10PB, respectively (Figure 2). ORAC significantly decreased (P<0.05) during storage, and ranged from 3252 to 9567 µMol Trolox/100g for all three peanut butter (PB) samples.

Hathorn and Sanders (2012) reported similar findings of PB containing peanut skins having higher ORAC than PP samples. At the end of the storage study, the ORAC of all PP and PB samples were lower and PB was the lower of the two. Both PP and PB were commercial samples with PB being slightly older. This difference in age may have contributed to the lower ORAC observed after 6 months (24 wks) of storage. The addition of

PS greatly increased the ORAC initially; and although the rate of decrease appeared to be similar for all samples, the samples with PS had and retained higher levels of ORAC throughout storage.

OSI is a relative measure of resistance to oxidation of a product. Initial (0 wk) mean

OSI values were 10.27, 9.21 and 9.54 (hrs) for 0PP, 3PP and 10PP samples, and 8.50, 8.02 and 7.59 (hrs) for 0PB, 3PB and 10PB samples, respectively (Figure 3 and 4). After 24 wks

113

of storage, the OSI for PP samples were 8.75, 7.18 and 6.68, respectively and for PB samples the OSI had significantly (P<0.05) decreased to 5.22, 5.02 and 4.89, respectively (Figure 3 and 4). These data indicate that increased addition of PS resulted in a sequential decrease in oxidative stability which suggests decreased shelf life. In a study where almond skins were added to soybean oil at levels ranging from 2 to 10%, oil stability did not increase (Johnson

2007). The authors suggested oil in the skins from the almonds may have been oxidized, causing no improvement in stability with the addition of almond skins. Johnson (2007) did not discuss the limited extraction of water soluble phenolic compounds in an oil matrix product. A high proportion of phenolic compounds in peanut skins (and almond skins) are water-soluble, and were not extracted into the oil matrix. Thus, they did not participate in prevention of oxidation of the oil matrix in the products. The data indicate that added skins reduced oxidative stability of the oil from the products which suggest that compounds in the skins contributed to the reduction of oil oxidative stability. Skin compounds or components contributing to reduction of oil stability would be oil soluble.

The mean PV for wk 0 was 10.74, 12.00, 11.27 mEq/kg for 0PP, 3PP, 10PP (Fig 5).

At 8 weeks of storage all samples were similar in PV; however, after 16 wks increases in PV were directly related to the concentrations of PS. This relationship was even more distinct at

24 wks where PV had increased significantly (P<0.05) relative to 25.04, 28.13, 35.23 mEq/kg, respectively (Figure 5). Similarly, PV for 0PB, 3PB and 10PB was 14.04, 13.24,

15.43 mEq/kg, and increased significantly (P<0.05) at the end of the study to 28.97, 35.87

114

and 43.57 mEq/kg), respectively (Figure 6). PV is generally considered a measure of shelf life and these data indicate that increased addition of PS resulted in overall reduced shelf life of both products beginning at approximately 8 wks of storage.

Oil from six lots of peanut skins and corresponding peanuts had similar fatty acid composition to corresponding peanuts (Table 2). Palmitic acid was generally higher in skin samples compared to the corresponding peanut samples except for S6. The major fatty acids in peanut skins and peanuts are oleic acid and linoleic acid. Oleic acid ranged from 46.62 to

56.07 for both peanut skins and peanuts with the mean for skins being 49.18 and peanuts

53.97. Oleic acid was significantly higher (P<0.05) in the peanut samples, but there was no significant difference (P>0.05) between S1 and P1. Sobolev and Cole (2003) found similar results where oleic acid was higher in peanuts than the skin. A lower oleic content in peanut skins maybe an indicator of low stability of oil, as high oleic peanuts are reported to have improved shelf life (Norden and others 1987; Isleib and others 2006). Linoleic acid ranged from 21.89 to 31.60 for all samples with peanut skins being higher than the seed. These results are supported by Sobolev and Cole (2003), who also found that linoleic acid was greater in the skins of peanuts. Linoleic acid is less oxidatively stable than oleic acid and the higher concentration of this fatty acid will also contribute to lower shelf life. The results for longer chain saturated fatty acids (behenic and lignoceric) were similar to previous reports

(Isleib and others 2006; Shin and others 2010) for peanuts (Table 2). Sobolev and Cole

(2003) reported much higher behenic acid (1.13 to 8.81) in peanut skins and peanuts.

115

Overall, the fatty acid composition of both peanuts and skins are within the range reported for various varieties and production locations.

The O/L ratio for both peanut skins and peanuts ranged from 1.48 to 2.4. Bolton and

Sanders (2002) reported that the O/L ratio of non-high oleic peanuts ranges between 0.9 and

2.5 with higher ratios generally found in warmer production locations. The ratio of oleic to linoleic acid (O/L) has been commonly associated with potential shelf-life. Although the peanuts used in this study were not high oleic, the overall oleic content of skins was significantly lower than in peanuts. The OSI of oil from peanut skins ranged from 1.1 to 4.1 h, while oils from peanuts were significantly higher (P<0.05) ranging from 9.8-15.7 h (Table

3). The majority of the fatty acids found in peanut skin oil are unsaturated, with the skins having more linoleic acid and less oleic acid than peanuts (Table 2).

Tocopherols (vitamin E) occur in peanuts as , β,  and  forms. Tocopherols were measured by HPLC with a UV-VIS detector (Table 4). Alpha tocopherol was not generally confirmed in skin oil samples with the UV-VIS detector and are reported as trace amounts

(Table 4) only after confirmation of the presence and identification of -tocopherol with a different analytical system using a fluorescence detector (data not shown). In peanut oil samples, -tocopherol ranged from 13.74 to 22.29 mg/100g which is within the range normally reported for peanuts (Shin and others 2009; Shin and others 2010; Silva others

2010). According to the USDA National Database for Standard Reference, -tocopherol in peanuts is 8.33 mg/100g (USDA 2011).

116

In addition to peanuts, -tocopherol is the predominant tocopherol in other nuts including almond, hazelnut, and macadamia (Maguire and others 2004). Alpha tocopherol is a major antioxidant in nature, which can donate a hydrogen to the hydroxyl group on a peroxyl radical. Chemically, -tocopherol is generally considered as the most active form of vitamin E due to the substitution pattern of methyl groups on the chromanol ring making the hydrogen of the C-6 hydroxy group especially active (Burton and Ingold 1981). Alpha- tocopherol prevents the propagation step of lipid oxidation to slow down the rate of lipid oxidation (Chapman and others 2009; Schneider 2005). Alpha-tocopherol functions through the glutathione peroxidase pathway and protects cell membranes from oxidation by reacting with lipid radicals produced in the lipid peroxidation chain reaction. Some studies have reported that fatty acid composition of oil is important to -tocopherol degradation.

Verleyen and others (2002) and Jorge and others (1996) found that -tocopherol degraded faster in more saturated oils. In another study, Yoshida and others (1992) reported that - tocopherol loss was greatest followed by β, ,  - tocopherol in animal fats. Recent studies suggest that -tocopherol, may be as important as -tocopherol (Jiang and others 2001;

Fukuda and others 1986). The -tocopherol concentrations in skin oils (except for sample

S5) were approximately 25% of that found in peanut oil and cumulatively, the total concentration of tocopherols in peanut oil was much higher than that of skin oil.

We found no literature reporting the absence of -tocopherol in skin oil from nuts. In a study by Chun and others (2005), -tocopherol decreased in both raw and roasted peanuts

117

during storage. Silva and others (2010) reported tocopherol content of raw and roasted peanuts, and found that roasting caused a significant decrease in -tocopherol. Holownia and others (2001) reported that , β,  and -tocopherols in peanut oil decreased from 14.1 to

11.16 mg/100 g oil, from 0,3 to 0.18 mg/100 g oil, from 12.4 to 8.12 mg/100 g oil and from

0.75 to 0.62 mg/100g oil, respectively, 24 h after frying. Li and others (1996) observed a decrease in tocopherols in flax, palm and sunflower oils after heating at 110°C.

Antioxidants slow oxidation, in contrast prooxidant compounds, such as metals may increase oxidation. Cu in peanut skins (39.6-56.0 ppm) was significantly (P<0.05) higher than in peanuts (4.8-7.1 ppm) (Table 5). Fe in the skins ranged from 108.9-206.1 ppm, while peanuts ranged from 17.2-20.4 ppm and were significantly lower (P<0.05). Angelova et al.

(2004) studied heavy metal accumulation in oil crops, and documented peanuts as having the highest accumulation of metals out of several oil crops evaluated. Although the oil from peanut skins was not tested as a potential carrier of heavy metals, the possibility of oils containing heavy metals has been reported (Ivano and others 1990; Garrido and others 1994;

Zhu and others 2011). Some agricultural by-products such as peanut skins and hulls are considered as natural absorbents of metals through ion exchange as a chelating agent

(Randall and others 1978). Metals in peanut skins may come from the soil or fertilizers

(Garrido and others 1994; Zeiner and others 2005). A limiting factor for the usage of peanut skins could be potentially the high concentration of heavy metals, which should be further investigated.

118

Conclusion

The shelf-life of peanut products containing peanut skins decreased over time and was inversely related to the percentage of peanut skins added. Low OSI, trace amount of - tocopherol, and high levels of Cu and Fe in peanut skin oil suggest a cause for reduced shelf- life in peanut products containing peanut skins.

119

References

Angelova V, Ivanova R, Ivanov K. 2004. Heavy metal accumulation and distribution in oil crops. Commun Soil Sci Plant Anal 35:2551-66.

AOAC 1990. Official Methods of Analysis. 14th Ed. Association of Official Analytical Chemists, Washington, DC.

AOCS. 2004. Official methods and recommended practices of the American oil chemists’ society. Fireston, D. ed. Champaign IL, AOCS Press.

Bolton GE, Sanders TH. 2002. Effect of roasting oil composition on the stability of roasted high-oleic peanuts. J Anal Oil Chem Soc 79:129-32.

Burton GW, Ingold KU.1981. Autoxidation of biological molecules.: The antioxidant activity of vitamin E and related chain-breaking phenolic antioxidants in vivo. J Am Chem Soc 103:6472-7.

Chapman TM, Kim HJ, Min DB. 2009. Prooxidant activity of oxidized -tocopherol in vegetable oils. J Food Sci 74:C536-42.

Chun, J., Junsoo, L. and Eitenmiller, R.R. 2005. Vitamin E and oxidative stability during storage of raw and dry roasted peanuts packaged under air and vacuum. J Food Sci 70:C292- 7.

Davis JP, Dean LL, Price KM, Sanders TH. 2010. Roast effects on the hydrophilic and lipophilic capacities of peanut flours, blanched peanut seed and peanut skins. Food Chem 119:539-47.

Fukuda Y, Nagata M, Osawa T, Namiki M. 1986. Chemical aspects of the antioxidative activity of roasted sesame seed oil and the effect of using the oil for frying. Agri Biol Chem 50:857-62.

Garrido MD, Frías I, Díaz C, Hardisson A. 1994. Concentrations of heavy metals in vegetable edible oils. Food Chem 50:237-43.

Gawel R. 1998. Red wine astringency a review. Aust J Grape Wine Res 4:74-95.

Hale OM, McCormick WC. 1981. Value of peanut skins (Testa) as a feed ingredient for growing-finishing swine. J Anim Sci 53:1006-10.

120

Hashim IB, Koehler PE, Eitenmiller RR, Kyien CK. 1993. Fatty acid content and tocopherol content of drought stressed Florunner peanuts. Peanut Sci 20:21-4.

Hathorn CS, Sanders TH. 2012. Flavor and antioxidant capacity of peanut paste and peanut butter supplemented with peanut skins. J Food Sci doi10.111/j.1750-3841.2012.02953.x.

Hill GM. 2002. Peanut by-products fed to cattle. Vet Clin North Amer: Food Anim Prac 18:295-15.

Holownia KI, Erickson MC, Chinnan MS, Eitnmiller RR. 2001. Tocopherol losses in peanut oil during pressure frying of marinated chicken strips coated with edible films. Food Res Int 34:77-80.

Huang S-W, Hopia A, Schwarz K, Frankel EN, German JB. 1996. Antioxidant activity of - tocopherol and trolox in different lipid substrates: bulk oils-in-water emulsions. J Agric Food Chem 44:444-52.

Isleib TG, Pattee HE, Sanders TH, Hendrix KW, Dean LO. 2006. Compositional and sensory comparisons between normal and high oleic peanuts. J Agric Food Chem 54: 1759-63.

Ivano K, Popva M, Kratschanov C, Maneva D. 1990. Application or pectic acid for removing heavy metals from vegetable oils. Z Lebansm Unters Forsch 191:210-3.

Jiang Q, Christen S, Shigenaga MK, Ames BN. 2001. -Tocopherol, the major form of vitamin E in the US diet, deserves more attention. Am J Clin Nutr 74:714-22.

Johnson HM. 2007. Almond skins as a natural antioxidant. [MSc thesis]. Clemson, SC: Clemson Univ. 86 p. Available from: Clemson Libraries, Clemson, SC.

Jorge N, Marquez-Ruiz G, Martin-Polvillo M, Ruiz-Mendez MW, Dobarganes MC. 1996. Influence of dimethylpolysiloxane addition to edible oils: dependence of the main variables of the frying process. Gras Aceit 47:14-9.

Karchesy JJ, Hemingway RW. 1986. Condensed Tannins (4β → 8;2β → O→7)-Linked Procyanidins in Arachis Hypogea L. J Agric Food Chem 34: 966-70.

Li SX, Cherian G, Ahn DU, Hardin RT, Sim JS. 1996. Storage, heating, and tocopherols affect cholesterol oxide formation in food oils. J Agric Food Chem 44:3830-4.

Lou H, Yamazaki Y, Sasaki T, Uchida M, Tanaka H, Oka S. 1999. A-type proanthocyanidins from peanut skins. Phytochem 54: 297-308.

121

Lou H, Yuan H, Yamazaki Y, Sasaki T, Oka S. 2001. Alkaloids and flavonoids from peanut skins. Planta Med 67:345.

Maguire LS, O’Sullivan SM, Galvin K, O’Connor TP, O’Brien NM. 2004. Fatty acid profile, tocopherols, squalene and phytosterol content of walnuts, almonds, peanuts, hazelnuts and the macadamia nut. Intl Food Sci Nutr 55:171-8.

McBrayer AC, Utley PR, Lowrey RS, McCormick WC. 1983. Evaluation of Peanut Skins (Testa) as a feed ingredient for Growing-Finishing Cattle. J Anim Sci 56:173-83.

Nepote V, Grosso NR, Guzman CA. 2002. Extraction of antioxidant components from peanut skins. Grasas Y Aceites 54: 391–5.

Norden AJ, Gorbet DW, Knauft DA, Young CT. 1987. Variability in oil quality among peanut genotypes in the florida breeding progam. Peanut Sci 14:7-11.

Palozz P, Krinsky N. 1992. β-carotene and -tocopherol are synergistic antioxidants. Archives Biochem Biophysics 297:184-7.

Randall JM, Hautala E, McDonald G. 1978. Binding of heavy metals ions by formaldehyde- polymerized peanut skins. J Applied Polymer Sci 22: 379-87.

Sanders TH, McMichael RW, Hendrix KW. 2000. Occurrence of Resveratrol in Edible Peanuts. J Agric Food Chem 48:1243-6.

Sanders TH. 1979. Effect of variety, location and year on tannin content of peanut seed coats. Peanut Sci 6:62-4.

Schwartz SJ, von Elbe JH, Giusti MM. 1996. Colorants. In Fennema’s Food Chemistry. CRC Press., Boca Raton, FL.

Schwarz K, Frankel EN, German JB. 1996. Partition behavior of antioxidative phenolic compounds in heterophasic systems. Fett/Lipid 98:115-21.

Schnider C. 2005. Chemistry and biology of vitamin E. Mol Nutr Food Res 49:7-30.

122

Shin E-C, Huang Y-Z, Pegg RB, Phillips RD, Eitenmiller RR. 2009. Commercial runner peanut cultivars in the united states: tocopherol composition. J Agric Food Chem 57:10289- 95.

Shin E-C, Ronald B-P, Phillips RD, Eitenmiller RR. 2010. Interrelationships among tocopherols of commercial runner market type peanuts grown in the united states. Intl J Food Sci and Technol 45:2622-8.

Silva MP, Martinez M J, Casini C, Grosso NR. 2010. Tocopherol content, peroxide value and sensory attributes in roasted peanuts during storage. Intl J Food Sci Technol 45:1499-504.

Sobolev VS, Cole RJ. 1999. trans-Resveratrol content in commercial peanuts and peanut products. J Agric Food Chem 47:1435-9.

Sobolev VS, Cole RJ. 2003. Note on utilization of peanut seed testa. J Sci Food Agric 84:105-11

USDA. U.S. Department of Agriculture, Agricultural Research Service. 2011. USDA National Nutrient Database for Standard Reference, Release 24. Available at: http://ndb.nal.usda.gov/ndb/foods/show/4779. Accessed 17 January 2012.

Verleyen T, Kamal-Eldin A, Mozuraityte R, Verhe R, Dewettinck K, Huyghebaert A, Grey W. 2002. Oxidation at elevated temperatures: competition between -tocopherol and unsaturated triacylglycerol. Eur J Lipid Sci Technol 104:228-33.

Yoshida H, Kondo I, Kajimoto G. 1992. Effects of microwave energy on the relative stability of vitamin E in animal fats. J Sci Food Agric 58:531-4.

Yu J, Ahmedna M, Goktepe I. 2005. Effects of processing methods and extraction solvents on concentration and antioxidant activity of peanut skin phenolics. Food Chem 90:199.

Zeiner M, Steffan I, Cindric IJ. 2005. Determination of trace elements in olive oil by ICP- AES and ETA-AAS: a pilot study on the geographical characterization. Microchemical J. 81:171-6.

Zhu F, Fan W, Wang X, Qu L, Yao S. 2011. Health assessment of eight heavy metals in nine varieties of edible vegetable oils consumed in China. Food and Chem Toxicol 49:3081-5.

123

Table 1. Peanut skins and corresponding peanut designations

S1-S4 Blanched skins S5 Light roasted skins S6 Roasted skins P1-P4 Blanched peanuts P5 Light roasted peanuts P6 Roast roasted peanuts

124

Table 2. Fatty acid composition (%) of oil from peanut skins and peanuts

Samples C16:0 C18:0 C18:1 C18:2 C20:0 C20:1 C22:0 C24:0 O/L

S1 10.94 d 2.53 b 46.62 a 31.60 a 1.29 d 1.28 cde 3.03 cde 1.57 b 1.48 c S2 12.04 e 2.65 c 47.15 a 29.03 b 1.42 bc 1.38 b 3.57 ab 1.77 a 1.62 c S3 12.11 e 2.71 c 47.25 a 29.98 b 1.43 bc 1.37 bc 3.59 ab 1.81 a 1.63 c

S4 12.02 e 2.72 c 47.32 a 28.91 bc 1.43 bc 1.38 b 3.51 ab 1.79 a 1.64 c S5 10.51 cd 2.69 c 51.19 b 27.31 cd 1.39 c 1.30 bcde 3.33 bc 1.54 bc 1.87 b S6 9.08 a 2.65 c 55.57 cd 23.16 ef 1.48 b 1.60 a 3.65 a 1.93 a 2.40 a

P1 10.57 cd 2.38 a 48.73 a 31.21 a 1.28 d 1.23 efg 2.77 e 1.34 d 1.56 c P2 9.96 b 2.92 d 54.21 cd 24.63 e 1.42 bc 1.17 fg 3.05 cde 1.43 bcd 2.23 a P3 9.81 b 2.98 d 55.76 d 24.06 e 1.43 bc 1.15 g 2.97 de 1.42 bcd 2.32 a

P4 9.81 b 2.89 d 55.83 d 24.14 e 1.39 c 1.14 g 2.91 de 1.38 cd 2.31 a P5 10.11 bc 2.75 c 53.21 bc 26.08 d 1.38 c 1.25 def 3.17 cd 1.45 bcd 2.05 b P6 9.13 a 3.30 e 56.07 d 21.89 f 1.73 a 1.32 bcd 3.81 a 1.84 a 2.38 a a 16:0 – Palmitic acid; 18:0 –Stearic acid; 18:1 – Oleic acid; 18:2 – Linoleic acid; 20:0 –Arachidic acid; 20:1 –Eicosenoic acid; 22:0 –Behenic acid; 24:0 –Lignoceric acid b Means in the same column with different letters are significantly different (P<0.05)

125

Table 3. OSI of oil from peanut skins and peanuts

Samples OSI (hrs)

S1 1.5 c S2 4.1 e S3 1.3 b S4 1.3 b S5 1.1 a S6 1.7 d P1 11.9 h P2 14.5 j P3 15.7 l P4 15.0 k P5 9.8 g P6 13.9 i

Means with different letters are significantly different (P<0.05)

126

Table 4. Total tocopherol content (mg/100g) of oils from peanut skins and peanuts

Samples  Β   S1 Tr ND ND 1.77 a S2 Tr ND 1.03 b 4.69 b S3 Tr 0.44 a 1.79 h 5.18 d S4 Tr 6.15 i 1.68 g 4.74 c S5 Tr 2.38 g 3.19 j 15.44 j S6 Tr 5.30 h 1.33 e 5.71 e P1 18.36 e 1.15 c 1.85 i 17.83 l P2 15.89 d 2.29 f 1.44 f 14.65 i P3 15.39 c 2.06 e 1.22 c 14.25 h P4 13.74 a 2.40 g 1.42 f 16.67 k P5 22.29 f 1.74 d 1.28 d 12.74 g P6 15.14 b 0.99 b 0.90 a 9.80 f a Means in the same column with different letters are significantly different (P<0.05) Tr – trace; ND = Not detected

127

Table 5. Copper (ppm) and iron (ppm) content in peanut skins and peanuts

Samples Cu (ppm) Fe (ppm) S1 56.0 d 206.1 e S2 48.1 c 108.9 b S3 49.0 c 118.5 c S4 46.9 c 107.7 b S5 39.6 b 172.1 d S6 55.6 d 181.8 d P1 5.6 a 20.4 a P2 5.0 a 17.7 a P3 5.9 a 17.2 a P4 7.1 a 17.9 a P5 4.9 a 18.4 a P6 4.8 a 19.3 a a Means in the same column with different letters are significantly different (P<0.05)

128

Figure 1. Hydrophilic ORAC of peanut paste containing peanut skins stored at 30°C

129

Figure 2. Hydrophilic ORAC of peanut butter containing peanut skins stored at 30°C

130

Figure 3. Oxidative stability index (110°C) of peanut paste containing peanut skins stored at 30°C

131

Figure 4. Oxidative stability index (110 °C) of peanut butter containing peanut skins stored at 30°C

132

Figure 5. Peroxide value (meq/kg) of peanut paste containing peanut skins stored at 30°C

133

Figure 6. Peroxide value (meq/kg) of peanut butter containing peanut skins stored at 30° C

134

CHAPTER 4

Sensory and nutritional optimization of a peanut-sweetpotato ready to use therapeutic type food (RUTF)

135

Sensory and nutritional optimization of a peanut-sweetpotato ready to use therapeutic type food (RUTF)

Chellani S. Hathorn1, Van Den Truong1, and Timothy H. Sanders2,3

1Department of Food, Bioprocessing, Nutrition Sciences North Carolina State University Raleigh, NC 27695

2USDA-ARS-MQHRU Department of Food, Bioprocessing, Nutrition Sciences North Carolina State University Raleigh, NC 27695

3Corresponding Author: Timothy H. Sanders 236 Schuab Hall, Campus Box #7624 Raleigh, NC 27695 Telephone: (919) 515-6312; Fax: (919) 513-8023 E-mail: [email protected]

The content of this chapter will be submitted for publication in Journal of Food Science

136

Abstract

Malnutrition is the result of an unbalanced diet in humans or animals resulting in a disproportionate number of illnesses and deaths. Ready to Use Therapeutic Foods (RUTF) are used to address malnutrition due to lack of nutrients especially in children; however, similarly formulated Ready to Use Sustainable Foods (RUSF) may be used to feed limited resource populations in underdeveloped countries or in developed countries during national or local disasters. Nutritional and sensory characteristics of eleven mixtures of peanut paste fortified with sweetpotato flakes and milk powder were determined in this study. In the samples evaluated, β-carotene ranged from 230 to 3470 µg/100g and ORAC ranged from

4030 to 5025 µMol/100g. Milk powder inclusion and various ingredients used in this study resulted in a more complete amino acid profile in the formulations specifically lysine, methionine and threonine increased. A lexicon incorporating roasted peanut and sweetpotato flavor descriptors was developed which utilized 8 flavor, 4 taste, and 1 chemical feeling factor attributes. In mixtures containing more than 15 % of both sweetpotato and milk powder components, roasted peanutty decreased; and sweet aromatic, earthy/canned carrot, and sweet taste attributes significantly increased.

Keywords: peanuts, sweetpotatoes, β-carotene, descriptive sensory, RUTF

137

Practical Application

Fortified peanut paste is a significant source of energy and protein that can be used to treat malnutrition. These products are stable at ambient temperature because of low moisture content. The information from this study is useful to food industries concerned with malnutrition, sustainable food or healthier food alternatives for the general public.

138

Introduction

Malnutrition, a major public health problem, is a primary contributor to child mortality and total global disease burden (WHO 1999; Müller and Krawinkel 2005; WHO

2007). Individuals are malnourished or suffer from under nutrition if their diet does not provide adequate calories and protein for maintenance and growth, or they cannot fully utilize the food they eat due to illness (Müller and Krawinkel 2005). Several different nutrition disorders may develop depending on which nutrients are lacking.

From 2000-2002 more than 850 million people in developing countries were reported to be undernourished (FAO 2004a). Nearly 20 million children under the age of five suffer from severe acute malnutrition (SAM) and 1 million die each year (WHO 2007).

Children under the age of 5 are particularly at risk from malnutrition because of demanding dietary requirements. In addition, children living in rural areas are almost twice as likely to be underweight as children in urban households. In particular, children in Africa and Asia have higher incidence of stunting, being underweight and wasting.

SAM is classified by a very low weight for height, by severe wasting or by the presence of nutritional oedema. Protein-energy malnutrition in children is defined by measurements that fall below the normal weight for age (underweight), height for age

(stunting), and weight for height (wasting) (Pinstrup-Anderson and others 1993; WHO

2007). Milk and milk products are an excellent source of nutrients, especially proteins and has been successfully used to fortify foods to increase protein quality. Ready to Use

Therapeutic Foods (RUTF) are used to address malnutrition especially in children; however,

139

similarly formulated Ready to Use Sustainable Foods (RUSF) may be used to feed limited resource populations in underdeveloped countries or in developed countries during national or local disasters.

Another major public health endemic is vitamin A deficiency (VAD), which is a problem in more than half of all countries (WHO 2012). Vitamin A is essential for human health and VAD is the leading cause of childhood blindness in the world. Vitamin A can be obtained from plants as provitamin A carotenoids and also from animals in the form of retinoid (Valdez 2001). Carotenoids are commonly found in the tissue of orange, yellow and green plants. Carotenoids are natural, lipid-soluble pigments typically found in vegetables such as the sweetpotato.

Sweetpotato is a highly nutritious crop (FAO 2004b; Loebenstein 2009) and ranks as the third most important starchy food crop (Padda and Picha 2008). Sweetpotatoes are grown in many developing countries, while over 80% of the crop is grown in developed countries (Woolfe 1992). Sweetpotatoes grow well in warm climates and in the US naturally thrives best in the South Atlantic and Gulf Coast States (Grubb and Guilford 2008).

Processed sweetpotatoes in the form of purees, powders, flour and flakes may be added to foods to provide a variety of functional benefits (Grabowski and others 2008). Community based approaches are one way of combating issues such as SAM and VAD. The objective of this study was to optimize peanut-sweetpotato-milk powder formulations for potential use as

RUTF or RUSF.

140

MATERIAL AND METHODS

Peanuts (runner type roasted peanuts; L value 50±1) were obtained from a commercial roasting facility and stored under refrigeration until utilized. Prior to grinding, the skins on the peanuts were removed and peanuts were ground into a paste using a Robot

Coupe Blixer (Ridgeland, MS, USA) following a grind-cool protocol (Sanders and others

1989). Sweetpotato flakes were obtained from Bruce Foods and milled using a Wiley Mill fitted with a 0.5 mm sieve. Nonfat milk powder, peanut oil and sucrose were purchased locally. Formulations evaluated are described in Table 1.

Color, Moisture, Fat

Hunter L (0=black, 100=white), a (+value=red, -value=green) b (+value=yellow, - value=blue) value of all samples were determined using a HunterLab Colorimeter

(HunterLab DP-9000™ Reston, VA, USA). Moisture content was determined gravimetrically on peanuts that were dried at 130ºC for 6 hours in a LXD Series Despatch forced air oven (Despatch Industries, Minneapolis, MN). Fat content was measured by

Nuclear Magnetic Resonance (NMR) Spectroscopy using a Minipec mq-one analyzer

(Bruker Optics Inc. Billerica, MA). All samples were prepared and analyzed in triplicate.

β-Carotene Analysis

β-carotene was determined using a modified procedure described by Grabowski and others (2008). Five grams of each sample was mixed with 4 grams of calcium carbonate, 4

141

grams of celite and 50 mL of methanol was added. The slurry was then homogenized for one minute using a Brinkman (Heidolph North America, Elk Grove Village, IL, USA) overhead stirrer, and filtered squentially through Whatman # 1 and #42 filter paper. Addition of 100 mL acetone:hexane (1:1) preceded filteration through the same paper and the mixture was homogenized. The filtering step was repeated a second time. The filtrate was placed in a separatory funnel and the filtrate collection flask was rinsed with 10 mL deionized water which was added to the separatory funnel. A saturated solution of NaCl was added to the separatory funnel in 1 mL aliquots up to 10 mL, as necessary, to facilitate separation of the organic and aqueous phases. The aqueous phase was removed and the organic upper layer was transferred to a volumetric flask and bought to a final volume of 50 mL. Absorbance of the extracts were determined on a UV-1700 PharmaSpec spectrophotometer (Shimadzu,

1% Kyoto, Japan) at 450 nm and Beer’s law (A=E /cm *b*c) was used to calculate the concentration of β-carotene in the solution. All samples were prepared and analyzed in triplicate.

ORAC Sample Extraction

Samples for hydrophilic ORAC analyses were extracted using a Dionex (Sunnyvale,

CA) ASE® 200 Accelerated Solvent Extractor (Prior and others 2003 and Davis and others

2010). Approximately 1.0 g of sample was weighed analytically and mixed with 25 g of clean sand. Samples and sand were then extracted with 70:29.5:0.5 acetone:water:acetic acid

142

(AWA) and brought to 50 mL final volume with additional AWA in preparation for hydrophilic analysis.

Hydrophilic - ORAC

Milwaukee, WI) were used as the control standards. Approximately 130µL of standards and hydrophilic extracts were added to a Costar polystyrene flat-bottom black 96 microwell plate

(Corning; Acton, Massachusetts). Sixty µL of a 70 nM fluorescein (FL) solution was then added to the wells and then the plate was incubated in the SAFIRE2 for 15 min at 37 ºC. Next

60 L of a 153 mM 2,2’-azobis (2-amindino-propane) dihydrochloride (AAPH) (Wako;

Richmond, VA) solution was added rapidly. Fluorescence excitation and emission wavelengths were programmed to 485 and 535 nm. H-ORAC was calculated using a regression equation between the concentration of Trolox and the net area under the curve. H-

ORAC was reported as Trolox equivalents (µMol Trolox /100g). All samples were prepared and analyzed in triplicate.

Total Amino Acids Amino acid content was determined using a Hitachi L-8900 Amino Acid Analyzer

(Hitachi High-Technologies Corp., Schaumburg, IL). Peanut-sweetpotato samples (0.5 g)

143

and 4mL of 6 N HCl containing 1 % phenol were vortexed for 30s. The samples were hydrolyzed using a Discover® SP System (CEM, Matthews, NC) with Synergy™ software and an autosampler where the samples were heated for 10 minutes at 165ºC under nitrogen.

As a control, a standard reference sample of peanut paste, SRM 2387 (NIST, Washington,

DC) was treated as a sample with every sample set. All sample solutions were brought to volume (25mL) with 0.02 N HCl. All samples were additionally diluted 1:10 in 0.02 N HCl and the peanut reference was diluted 1:5 in 0.02 N HCl. Analysis was performed by an instrumental method. All samples were prepared and analyzed in triplicate.

Development of a Descriptive Sensory Peanut-Sweetpotato Lexicon Combination of some terms from the peanut lexicon (Johnsen and others 1988;

Sanders and others (1989)), and the sweetpotato lexicon by Leksrisompong and others (2012) were utilized by a highly trained panel to select and/or modify terms to describe a peanut- sweetpotato based product. Descriptors were generated across 6 30-45 min training sessions where panelists generated terms and definition.

Descriptive Sensory Analysis

A trained descriptive sensory panel (n=5, females; n=5, males) using the 0-15

Spectrum™ method point universal intensity scale (Meilgaard and others 1999) evaluated the peanut-sweetpotato samples. Each panelist had more than 500 h of experience with descriptive analysis of peanut flavor. The panel received additional indepth training to allow the panel to calibrate and consistently identify and score descriptors during 10 30 minute

144

sessions. A roasted peanut paste reference and sweetpotato reference was provided for each session. Peanut-sweetpotato samples were served at room temperature and deionized water and unsalted crackers were provided at all panels. All samples were coded with random three digit codes and 4 replications were presented randomly.

Statistical Analysis

Analysis of variance (ANOVA) was generated using PROC GLM and comparison of means were made using Duncan’s post-hoc test (SAS version 9.1, Cary, NC). Significance was established at P<0.05.

145

Results and Discussion

Eleven peanut-sweetpotato formulations and a roasted peanut paste were evaluated

(Table 1). Color ranged from L 45.0 to L 55.2 (Table 2) and became darker as the percentage of sweetpotato flakes increased. Sweetpotato flakes were dried with high heat to a moisture content of 2.0% and this drying resulted in darkening via Maillard reaction products which contributed to the darker color with higher sweetpotato concentrations.

Overall, the a value (red to green) increased (P<0.05) ranging from 8.6 to 11.6. The general trend was an increase in red, which can be attributed to the increase in sweetpotato flakes added. While this holds true for all samples, it does hold true for PS9, which contained 15% sweetpotato flakes and had a greater a value than PS10 and PS11, which maybe in part due to lower amounts of peanut paste and increased amounts of milk powder. The b value ranged from 21.73 to 25.50 for all samples, which indicated a decrease in yellow color. There was a slight but significant increase in moisture content which ranged from 1.1 to 1.3.

β-carotene was lowest in PP (230 µg/100g) and highest in PS11 (3470µg/100g), which was significantly higher (P<0.05) (Table 3). The β-carotene content of the sweetpotato flakes used in this study was 7490 µg/100g (Table 3). Other studies have reported that levels of β-carotene in sweetpotato flakes range from 3980 to 5750 µg/100g

(Emenhiser and others 1999; Valdez and others 2001). Sweetpotatoes have been reported to have β-carotene concentrations ranging from 4490 to 22600 µg/100g (Simonne and others

1993; Teow and others (2007). The β-carotene concentrations of sweetpotato flakes suggest

146

the degradation of β-carotene during processing. β-carotene increased as the concentration of milk powder increased. β-carotene was measured in raw and processed milk, with raw milk containing on average 18.8 µg/100g and semi-skimmed milk containing 7.8 µg/100g

(Hulshof and others 2006). Although the β-carotene content of nonfat milk powder is relatively low compared to SPF, Vavich and others (1954) indicated that milk improves the availability of β-carotene.

ORAC of the sweetpotato flakes used in these studies was 3285 µMol/100g. ORAC in formulations ranged from 4030 to 5025 µMol/100g. The general increase observed in

ORAC was related to the addition of β-carotene (an antioxidant) in the sweetpotato flakes.

Teow and others (2007) reported the ORAC of orange flesh sweetpotatoes ranged between

589-1030µMol/100g. Reports of ORAC for raw, boiled sweepotatoes without the skin and baked sweetpotatoes in the skin was 903, 766 and 2115 µMol/100g, respectively (USDA

2007). The differences between the ORAC of sweeetpotatoes and sweetpotato flakes may be due to the reduced moisture content, as well as an increase in Maillard products which have antioxidant activity (USDA 2007). Antioxidant activity in sweetpotatoes can come from phenolic compounds, anthocyanins and carotenoids.

Significant increases (p<0.05) in threonine, serine, cysteine, valine, methionine, isoleucine, leucine, lysine, arginine and proline were observed in the peanut-sweetpotato formulations (Table 4). Peanuts are an excellent source of protein (~30%) and the two major types are arachin and conarachin (Basha and Cherry 1976). Despite the protein content, peanuts lack adequate amounts of lysine, methionine and threonine. Pominski and others

147

(1991) added nonfat milk powder to peanut butter increasing lysine, threonine and reported only a slight increase in methionine. Threonine increased from 1.98 to 4.01 g/100g, methionine from 0.99 to 1.92 g/100g and lysine increased from 2.58 to 3.75 g/100g in the peanut-sweetpotato formulations (Table 4). Alid and others (1981) improved the protein efficiency ratio (PER) of peanut flour by 41% when 2% lysine and 0.2% methionine was added. Yeh and others (2002) reported that the addition of 19% roasted soybean and 14% non fat dried milk improved lysine in peanut spreads. Protein deficiency reduces the concentration of amino acids in plasma (Wu and others 1999) and compromises the immune system, which is currently a significant nutritional problem in developing countries

(Woodward 1998; Dasgupta and others 2005). Dietary supplementation with high-quality protein may be effective in improving protein nutritional status in malnourished individuals.

Twelve sensory descriptors were identified in peanut-sweetpotato formulations. The language included 8 flavor, 4 taste, and 1 chemical feeling factor. Roast peanutty ranged from 3.9 to 5.4 for all samples. PS9, PS10, PS11 and PS12, had the lowest roast peanutty intensity of all the formulations, which contained the least amounts of peanuts (Table 6).

Baked sweetpotato/dried apricot/floral intensity ranged from 1.5 to 2.7 for all samples except

PP. There was a significant (P<0.05) increase in baked sweetpotato/dried apricot/floral intensity in formulations containing increased amounts of SPF. The panel did not detect differences (P>0.05) in baked sweetpotato/dried apricot/floral intensity through 10% addition of SPF but differences were detected when 15% or more SPF were added. Baked sweetpotato/dried apricot/floral is described as aromatics associated with the flesh portion of

148

a sweetptoato with dried apricot and floral notes. In general, formulations that contained more SPF were more intense in sweet aromatic, with intensity ranging from 3.2 to 4.5.

Sweet aromatic was significantly lower for PP compared to the formulations because they did not contain any SPF. Peanuts typically have a distinctive sweet aromatic flavor with an intensity generally around 3.0 or higher (Johnsen and others 1988; Sanders and others 1989;

Bett and others 1994). Earthy/ canned carrot was detected (threshold of 1 or higher) in PS2,

PS3, PS4, PS6, PS8, PS9, PS10, and PS11. Orange flesh sweetpotatoes are distinctly characterized by earthy/canned carrot flavor (Lekrisompong and others 2011).

Earthy/canned carrot is described as an earthy aromatic associated with canned carrot flavor.

Leigthon and others (2010) reported earthy notes in sweetpotatoes as being associated with damp soil, wet foliage or slightly undercooked potatoes. Sweet taste ranged from 2.4 to 6.8 and generally was significantly (P<0.05) higher in samples that contained more SPF.

Sweetpotatoes, in general are naturally sweet and were expected to increase sweet taste.

Sucrose, glucose and fructose are present in raw sweetpotatoes (Kays and others 2005;

Horvat and others 1991) and the formation of maltose occurs during cooking via the action of

α-amylase and β-amylase (Kays and others 2005; Sun and others 1994). Individual sugars vary in sweetness intensity, for example fructose is 2.8 times sweeter than maltose. There was an increase (P<0.05) in bitter intensity in PS8, PS9, PS10 and PS11 which contained either 15% or 20% SPF. PS4 and PS6 contained 15% SPF, but a difference (P>0.05) in bitter intensity was not observed by the panel. These differences may be attributed to these

149

formulations containing a different ratio of peanuts and milk powder (more peanuts, less milk powder) than PS8, PS9, PS10 and PS11.

Conclusions

RUTF were developed with improved nutritional qualities, and descriptive sensory was used to provide useful information about the sensory characteristics of these formulations. Increased β-carotene and ORAC were observed in the formulations. The addition of milk powder improved protein quality, resulting in increased concentration of some specific amino acids. Selection of formulations based on nutritional and sensory characteristics can be used to further improve future RUTF in future work. Descriptive sensory and consumer acceptance studies can be powerful tools to understand the flavor of

RUTF or RUSF. Based on the data collected in this study, formulations PS9 and PS11 appear to the best combination of flavor and nutritional composition.

150

References

Alid G, Yanez E, Aguilera JM, Monckeberg F, Chichester CO . 1981. Nutritive value of an extrusion texturized peanut protein. J Food Sci 46: 948-9

Basha SM, Cherry JP. 1976. Optimizing protein- and vitamin-fortified peanut spreads containing soybean or milk powder. J Food Qual 26:243-56.

Bett KL, Vercellotti JR, Lovegren NV, Sanders TH, Stinsch RT, Rasmussen GK. 1994. A Comparison of flavor and compositional quality of peanuts from several origins. Food Chem 51: 21-7.

Dasgupta M, Sharkey JR, Wu G. 2005. Inadequate intakes of indispensable amino acids among homebound older adults. J Nutr Elderly 24:85-99.

Davis JP, Dean LL, Price KM, Sanders, TH. 2010. Roast effects on the hydrophilic and lipophilic capacities of peanut flours, blanched peanut seed and peanut skins. Food Chem 119: 539-47.

Emenhiser C, Watkins RH, Simunovic N, Solomons N, Bulux J, Barrows J, Shwartz SJ. 1999. Packaging preservation of β-carotene in sweet potato flakes using flexible film and an oxygen absorber. J Food Qual 22:63-73.

FAO. 2004a. Food and Agriculture Organization of the United Nations. Undernouishment around the world. In: The state of food insecurity in the world 2004. Rome: The Organixation.

FAO. 2004b.Food and Agriculture Organization of the United Nations. Agricultural data. Available at: http://faostat.fao.org

Grabowski JA, Troung VD, Daubert CR. 2008. Nutritional and rheological characterization of spray dried sweetpotato powder. LWT: 41:206-16.

Grubb EH, Guilford, WS. 2008. The Potato: A Compilation of Information from Every Available Source (1912). Kessinger Publishing. pp.1-632.

Horvat RJ, Arrendale RJ, Dull GG, Chapman GW, Kays SJ.1991. Volatile Constituents and Sugars of Three Diverse Cultivars of Sweetpotato. J Food Sci 56: 714-5.

151

Hulshof PJM, van Roekel-Jansen T, van de Bovenkamp P, West CE. 2006. Variation in retinol and carotenoid content of milk and milk products in The Netherlands. J Food Comp Anal 19:67-75.

Johnsen PB, Civille GV, Vercellotti JR, Sanders TH, Dus, CA. 1988. Development of a lexicon for the descripton of peanut flavor. J Sensory Stud 3: 9-17.

Kays SJ, Wang Y, McLaurin WJ. 2005. Chemical and Geographical Assessment of the Sweetness of the Cultivated Sweetpotato Clones of the World. J Amer Soc Hort Sci 130:541-97.

Leighton CS, Schönfeldt HC, Kruger R. 2010. Quantitative descriptive sensory analysis of five different cultivars of sweet potato to determine sensory and textural profiles. J Sensory Stud 25: 2-18.

Loebenstein G, Thottappilly G, Fuentes S, Cohen J.2009. Virus and phytoplasma diseases. In: The sweetpotato. Springer. p. 105-34.

Leksrisompong PP, Whitson ME, Truong VD, Drake MA. 2012. Sensory attributes and consumer acceptance of sweet potato cultivars with varying flesh colors. J Sensory Stud 27:59-69.

Meilgaard MM, Civille GV, Carr BT. 1999. Sensory Evaluation Techniques, 2nd Ed., CRC Press, Boca Raton, FL.

Müller O, Krawinkel M. 2005. Malnutrition and health in developing countries. Can Med Assoc J 173:279-86.

Padda MS, Picha DH. 2008. Phenolic Acid Composition and Antioxidant Capacity of Different Heat-Processed Forms of Sweetpotato cv. ‘Beauregard’. Inter J Food Sci Technol 43:1404–9.

Pominski, L. and Hamlet, C. A. 1991. Peanut butter quality as affected by inclusion of nonfat dry milk components. Peanut Sci. 18:16-18.

Prior RL, Hoang H, Gu L, Wu XL, Bacchiocca M, Howard L, Hampsch-Woodill M, Huang D, Ou B, Jacob R. 2003. Assays for hydrophilic and lipophilic antioxidant

152

capacity (oxygen radical absorbance capacity (ORAC(FL)) of plasma and other biological and food samples. J Agric Food Chem 51:3273-9.

Sanders TH, Vercellotti JR, Crippen KL, Civille, GV. 1989. Effect of maturity on roast color and descriptive flavor of peanuts. J Food Sci 54: 475-7.

Simonne AH, Kays SJ, Koehler PE, Eitenmiller RR. 1993. Assessment of β-carotene content in sweetpotatobreeding lines in relation to dietary requirements. J Food Comp Anal 6:336-45.

Sun JS, Severson RF, Kays SJ. 1994. Effect of heating temperature and microwave pretreatment on the formation of sugars and volatiles in Jewel sweet potato. J Food Qual 17:447-56.

Teow CC, Troung V-D, McFeeters RF, Thompson RL, Pecota KV, Yencho GC. 2007. Antioxidant activities, phenolic and β-carotene contents of sweet potato genotypes with varying flesh colours. Food Chem 103:829-38.

United States Department of Agriculture (USDA) 2007. Oxygen radical absorbance capacity (ORAC) of selected foods. pp. 1-34.

Valdez C, López Y, Schwartz S, Bulux J, Solomons N. 2001. Sweet potato buds: the origins of a “designer: food to combat hypovitaminosis A in Guatemala. Processing, vitamin A content and preservation characteristics. Nutr Res 21: 61-70.

Van Jaarsveld PJ, Marais DW, Harme E, Nestel P, Rodriguez-Amaya DB. 2006. Retention of β-carotene in boiled, mashed orange-fleshed sweet potato. J Food Comp Anal 19:321-9.

Vavich MG, Stull JW, Raica N, Kemmerer AR. 1954. Effect of nonfat milk on the utilization of carotene and vitamin A. Arizona agricultural experiment station technical paper No. 339. 310-4.

Woodward B. 1998. Protein, calories, and immune defenses. Nutr Rev 56:S84-S92.

Woolfe J. 1992. Sweepotato: an untapped food resource. Cambridge Univ. Press. Cambridge, UK. pp.41-117.

World Health Organization. 1999. Management of severe malnutrition: A manual for physcians and other senior health workers. WHO: Geneva. pp.1-67.

153

World Health Organization (WHO). 2012. Micronutrient deficiencies. Vitamin A deficiency. Accessed on September 29, 2012 http://www.who.int/nutrition/topics/vad/en/

Wu G, Flynn NE, Flynn SP, Jolly CA, Davis PK. 1999. Dietary protein or arginine deficiency impairs constitutive and inducible nitric oxide synthesis by young rats. J Nutr 129:1347-54.

Yeh, JY, Resurreccion, AVA, Phillips, RD, Hung, Y.C. 2002. Overall acceptability and sensory profiles of peanut spreads fortified with protein vitamins and minerals. J Food Sci 67:1979-85.

154

Table 1. Peanut-sweetpotato paste formulations

PP PS1 PS2 PS3 PS4 PS5 PS6 PS7 PS8 PS9 PS10 PS11 Peanut Paste 100 85 80 77 72 75 70 72 63 58 53 43 Sweetpotato Flakes --- 5 10 10 15 10 15 10 15 15 20 20 Milk Powder --- 7 7 10 10 12 12 15 15 20 20 30 Peanut Oil --- 3 3 3 3 3 3 3 5 5 5 5

155

Table 2. Color, moisture and fat content of peanut-sweetpotato formulations

L a b Moisture (%) Fat (%) PP 51.4 c 8.7 f 24.4 b 1.0 c 49.0 a PS1 55.1 a 8.6 f 25.3 a 1.3 ab 40.9 b PS2 53.7 b 9.8 e 25.4 a 1.2 ab 34.9 c PS3 53.4 b 9.6 e 25.2 a 1.1 bc 21.1 f PS4 53.0 b 10.2 d 25.1 a 1.2 ab 26.8 d PS5 54.9 a 9.6 e 25.2 a 1.3 a 27.1 d PS6 53.3 b 10.3 d 25.2 a 1.1 bc 24.6 e PS7 50.0 c 10.4 d 24.7 b 1.3 ab 27.1 d PS8 46.4 d 10.8 bc 22.3 c 1.2 ab 25.7 d PS9 45.0 e 11.6 a 21.7 d 1.2 ab 25.9 d PS10 45.6 e 11.0 b 22.1 c 1.3 ab 26.4 d PS11 47.8 d 10.8 bc 22.2 c 1.3 ab 25.7 d

Means in columns followed by different letters are significantly different (p<0.05)

156

Table 3. β-carotene and ORAC of peanut-sweetpotato formulations

β-carotene H-ORAC (µg/100g) (µMol/100g) PP 230 d 4030 h PS1 760 cd 4422fg PS2 1160c 4363 c PS3 1170 c 4562e PS4 1470 c 4722 d PS5 1250 c 4527ef PS6 1640 c 4839 bcd PS7 1270 c 4747 cd PS8 1890 c 4866 bc PS9 2280 b 4896 b PS10 2790 b 4890 b PS11 3470 a 5025 a

Means in columns followed by different letters are significantly different (p<0.05) *β-carotene of SPF used in study - 74.9 µg/g

157

Table 4. Total amino acid content (g/100g) of peanut-sweetpotato formulations

Asp Thr Ser Glu Gly Ala Cys Val

PP 13.00 bc 2.48 cd 4.51 b 20.78 c 6.25 a 4.14 a 0.01 d 4.39 d

PS1 12.81 bcd 2.72 bc 5.04 a 21.24 a 5.85 b 4.07 a 0.17 c 4.23 e

PS2 12.49 de 1.98 d 3.87 c 20.90 bc 5.86 b 4.26 a 0.08 cd 4.54 c

PS3 12.53 de 1.98 d 3.87 c 20.83 bc 5.86 b 4.28 a 0.10 cd 4.56 c

PS4 12.33 de 2.05 d 3.92 c 20.87 bc 5.66 c 4.26 a 0.03 d 4.59 c

PS5 12.45 de 2.04 d 3.91 c 20.82 bc 5.73 bc 4.25 a 0.11 cd 4.60 c

PS6 12.33 de 2.00 d 3.87 c 20.87 bc 5.77 bc 4.15 a 0.07 cd 4.54 c

PS7 13.48 a 3.88 a 4.95 a 20.66 c 5.15 e 3.50 b 0.40 b 4.77 b

PS8 13.20 ab 3.19 b 4.94 a 20.83 bc 5.35 d 3.37 b 0.57 a 4.73 b

PS9 12.63 cd 3.08 b 4.97 a 21.09 ab 5.09 e 3.19 b 0.677 a 4.85 b

PS10 12.33 de 2.00 d 3.87 c 20.87 bc 5.77 bc 4.25 a 0.07 d 4.54 c

PS11 12.06 e 4.01 a 4.93 a 19.81 d 2.69 f 3.22 b 0.59 a 5.35 a

Means in columns followed by different letters are significantly different (p<0.05) Aspartic acid (Asp); Threonine (Thr); Serine (Ser); Glutamic acid (Glu); Glycine (Gly); Alanine (Ala); Valine (Val); Methionine (Met); Isoleucine (Ile); Leucine (Leu); Tyrosine (Tyr); Phenylaline (Phe); Lysine (Lys); Histidine (His), Arginine (Arg); Proline (Pro)

158

Table 4. Continued

Met Lle Leu Tyr Phe Lys His Arg Pro PP 0.99 f 3.48 d 7.58 f 4.45 b 5.69 a 2.58 g 2.53 bc 12.53 a 4.59 g PS1 1.05 ef 3.34 e 8.04 e 4.33 bc 5.31 b 2.97 f 2.48 c 11.19 c 5.15 e PS2 1.12 de 3.59 cd 8.34 bc 4.72 a 5.56 a 3.18 e 2.59 ab 11.64 b 5.37 de PS3 1.06 ef 3.59 c 8.26 cd 4.74 a 5.56 a 3.16 e 2.59 ab 11.65 b 5.36 cd PS4 1.14 cd 3.62 c 8.44 c 4.73 a 5.61 a 3.41 c 2.57 ab 11.27 c 5.49 bc PS5 1.10 de 3.64 bc 8.36 bc 4.72 a 5.57 a 3.31 d 2.58 ab 11.30 c 5.48 bc PS6 1.09 de 3.59 cd 8.40 bc 4.71 a 5.67 a 3.29 d 2.57 ab 11.38 bc 5.60 b PS7 1.31 b 3.63 c 8.19 d 4.28 bcd 4.62 c 3.37 cd 2.58 ab 10.30 d 4.93 f PS8 1.20c 3.65 bc 8.43 b 4.11 d 4.60 c 3.49 b 2.57 ab 10.35 d 5.39 cd PS9 1.29 b 3.75 b 8.76 a 4.19 cd 4.59 c 3.75 a 2.60 a 9.88 e 5.59 b PS10 1.09 de 3.59 cd 8.39 bc 4.71 a 5.67 a 3.29 d 2.57 ab 11.38 bc 5.59 b PS11 1.92 a 4.44 a 8.72 a 3.62 e 4.09 d 3.49 b 2.35 d 11.30 c 7.41 a

159

Table 5. Descriptors and definition of peanut-sweetpotato lexicon

Descriptors Definition Aromatics Roasted peanutty The aromatic associated with medium roast peanuts Sweet aromatic The aromatics associated with sweet material such as caramel, vanilla, and non sulfur molasses Baked sweetpotato / Dried The aromatics associated with the internal portion of a baked orange-flesh apricot/Floral sweetpotato with dried apricot and floral notes Dark roast The aromatic associated with hark roasted peanuts Raw beany The aromatic associated with light roasted peanuts and having legume-like notes Woody/Hulls/Skins The aromatics associated with base peanut character (absence of fragrant top notes) and related to dry wood, peanut hulls, and skins Earthy/Canned Carrot The aromatics associated with moist soil and having root-like vegetable notes Cardboardy The aromatic associated with somewhat oxidized fats and oils and reminiscent of cardboard Tastes Sweet taste The taste on the tongue associated with sugars Salty taste The taste on the tongue stimulated by sodium salts Sour The taste on the tongue stimulated by acids Bitter The taste on the tongue associated with bitter agents such as caffeine or quinine Chemical Feeling Factors Astringency The chemical feeling factor on the tongue, described as puckering/dry and associated with tannins or alum

160

Table 6. Descriptive sensory analysis of peanut-sweetpotato formulations

Roast peanutty Baked Sweet Dark Raw Woody/ Earthy/ Sweet Bitter Astringency sweetpotato/ aromatic roast beany Hulls/Skins Canned carrot Taste Dried apricot/ Floral PP 5.4 a ND 3.2 e 2.9a 2.1 a 3.1 a ND 2.4 d 2.4 abc 1.1 a PS1 5.3 ab 1.5a 3.8 d 2.9 a 1.9 b 3.0 a 0.3 d 3.7 c 2.3 bc 1.0 a PS2 5.2 ab 1.7a 3.9 cd 2.9 a 2.0ab 3.0 a 1.0 bc 3.9 c 2.2 c 1.0 a PS3 5.2 ab 1.7a 3.9 cd 3.0 a 2.0 ab 3.0 a 1.0 bc 4.1 bc 2.2 c 1.0 a PS4 5.0 b 2.6b 4.1 bc 3.0 a 2.0 ab 3.0 a 1.5 a 4.6 b 2.2 c 1.0 a PS5 5.2 ab 1.9a 3.9 cd 3.0 a 2.0 ab 3.0 a 0.7 c 4.0 bc 2.2 c 1.0 a PS6 5.0 b 2.5b 4.0 cd 3.0 a 2.0 ab 3.0 a 1.4 ab 4.6 b 2.2 c 1.0 a PS7 5.2 ab 1.9a 3.9 cd 3.0 a 2.0 ab 3.0 a 0.7 c 4.0 bc 2.2 c 1.0 a PS8 4.3 c 2.6b 4.5 a 3.1 a 2.0 ab 3.0 a 1.5 a 6.8 a 2.5 ab 1.0 a PS9 4.2 c 2.7b 4.3 ab 3.0 a 2.1 a 3.0 a 1.3 ab 6.8 a 2.5 a 1.0 a PS10 4.2 c 2.7b 4.4 a 3.1 a 1.9 b 3.0 a 1.0 bc 6.6 a 2.5 ab 1.0 a PS11 3.9 d 2.6b 4.3 ab 3.0 a 2.0 ab 3.0 a 1.3 ab 6.6 a 2.5 ab 1.0 a Means in columns followed by different letters are significantly different (p<0.05) ND=not detected

161

CHAPTER 5

Nutritional composition, descriptive sensory analysis, and consumer acceptance of a peanut-sweetpotato based ready-to-use therapeutic food

162

Nutritional composition, descriptive sensory analysis, and consumer acceptance of a peanut-sweetpotato based ready-to-use therapeutic food

Chellani S. Hathorn1, MaryAnne Drake1, Van Den Truong1, Timothy H. Sanders2,3

1Department of Food, Bioprocessing, Nutrition Sciences

North Carolina State University

Raleigh, NC 27695

2USDA-ARS-MQHRU

Department of Food, Bioprocessing, Nutrition Sciences

North Carolina State University

Raleigh, NC 27695

3Corresponding Author: Timothy H. Sanders

236 Schuab Hall, Campus Box #7624

Raleigh, NC 27695

Telephone: (919) 515-6312

Fax: (919) 513-8023

E-mail: [email protected]

The content of this chapter will be submitted for publication in Journal of Food

Science

163

Abstract

Ready-to-use therapeutic foods (RUTF) are nutritionally enhanced products which have been successfully used to treat severe malnutrition. Ready-to-use sustainable foods

(RUSF) may have different composition and may be used to address less severe nutritional issues. Addition of sweetpotato flakes (SPF) as a source of beta carotene (pre Vitamin A) to products that may be used in situations of malnourishment may contribute to overall health of individuals especially related to prevention of blindness. The objective of this study was to evaluate nutritional composition, viscosity, color, descriptive sensory analysis and consumer acceptance of three formulations which contained 49/15/20, 28/20/30 and 56/0/30 percentages of peanuts/SPF/milk powder, respectively and other standard ingredients. A commercial RUTF, MANA, was used in descriptive and consumer studies. β-carotene concentrations in the formulations containing SPF were 2420 and 3070 µg/100g. Descriptive sensory analysis characterized the SPF containing formulations as high in roasted peanut, sweet aromatic, baked sweetpotato/dried apricot/floral aromatics and high in sweet taste.

The formulation with the 15% SPF was one of the most overall liked by consumers and overall flavor liking was highest for this formulation.

Keywords: consumer acceptance, descriptive sensory, RUTF, peanut, sweetpotato

164

Practical Application

Ready to use therapeutic foods (RUTF) can be used to treat severe cases of malnutrition and may be beneficial in other nutritional deficiencies when appropriate ingredients are added to address those deficiencies. RUTF typically contain peanuts as a base and the incorporation of sweetpotatoes may decrease Vitamin A deficiency in malnourished populations resulting in prevention of blindness in children and overall improved health.

165

Introduction

Severe malnutrition (SAM) affects 20 million children under the age of five and approximately 1 million die each year (WHO 2007). SAM is defined by a very low weight for height, by severe wasting or by the presence of nutritional oedema. Protein-energy malnutrition in children is defined by measurements that fall below the normal weight for age

(underweight), height for age (stunting), and weight for height (wasting) (Pinstrup-Anderson and others 1993; WHO 2007). A child between the ages of 6-59 months with an arm circumference of less than 110 mm is also a sign of SAM (WHO 2007). The burden of child mortality due to SAM primarily remains absent from the international health agenda and there are few national policies aimed at addressing the issue comprehensively.

SAM was originally treated with therapeutic milk (Formula-75 and Formula-100) developed in the 1980s by Nutriset, a private French company specializing in therapeutic foods (Guimón and Guimón 2012). Disadvantages of F-75 and F-100 were specific requirements for reconstitution with drinking water, heat, clean utensils, precise measurement and hospital administration by trained individuals. The finished product only retained therapeutic properties for a few hours and could not be left at room temperature

(Guimón and Guimón 2012).

In the 1990s, ready-to-use-therapeutic foods (RUTF) were developed as a solution to eradicate SAM and often contained peanuts, milk powder, vitamins and minerals. RUTFs are soft enriched foods that do not require additional water, can be consumed easily by children from the age of six months, do not require specialize storage nor any special training

166

to administer (WHO 2007). Several forms of RUTFs were developed by Nutriset scientists and included enriched pancakes, doughnuts and biscuits. Attempts to create RUTF in the form of a chocolate bar with similar composition to F-100 were unsuccessful since the bar melted easily and had an unacceptable taste when vitamins and proteins were added (Guimón and Guimón 2012).

Plumpy’nut, a peanut-base RUTF used to treat SAM is manufactured by Nutriset.

Plumpy’nut has a two year shelf life and requires no water, preparation, or refrigeration

(Lennema and others 2007). Unlike therapeutic milk, Plumpy'nut can be administered at home without medical supervision. The ingredients in Plumpy'nut include peanut paste, sucrose, vegetable fat, skimmed milk powder, vitamins, and minerals. Mother-administered- nutritive-aid (MANA), similar to Plumpy’nut, is a peanut-base RUTF developed by MANA

Nutrition. Maize, soybeans (Unimix) and chickpeas (Wawa Mum) are examples other commodities incorporated into RUTFs.

Sweetpotatoes are an excellent source of carbohydrates and β-carotene, a vitamin A precursor. β-carotene is an antioxidant which is capable of quenching singlet oxygen, a reactive molecule that can trigger free radical chain reactions which may be related to several diseases. β-carotene is readily available in foods such as, sweetpotatoes, carrots, kale and spinach and is converted in the body as required to maintain appropriate physiological concentrations. Vitamin A deficiency (VAD), affects people in more than half of all countries (WHO 2012), and is the leading cause of childhood blindness in the world. An estimated 250,000 to 500,000 vitamin A-deficient children become blind every year, half of

167

them dying within 12 months of losing their sight (WHO 2012). Vitamin A is necessary for a large number of human metabolic functions and can be obtained from plants as provitamin

A carotenoids or from animals in the form of retinoids (Valdez 2001). Traditionally, vitamin

A activity of beta-carotene has been expressed in International Units (IU); 1 IU = 0.60 µg of all-trans beta-carotene. However, this conversion factor does not take into account the bioavailability of carotenoids in humans. Thus, a FAO/WHO Expert Committee proposed that vitamin A activity be expressed as retinol equivalents (RE) where1 RE = 6 µg beta- carotene (FAO/WHO 1967). RUTFs are typically used in treating malnutrition, but the potential exists for use of similar products as sustainable foods, in aging populations, or in relief efforts in disaster situations. Other ingredients such as sweetpotato broaden the nutritional character of RUTFs and thus broaden the potential use in different situations that may involve adults as well as children. Flavor is an important attribute to the acceptance of food products, and is often evaluated by descriptive sensory analysis (DSA) and consumer acceptance testing. Testing children under 5 can be challenging as their language skills, memory and ability to reason is less developed (Popper and Kroll 2005); however, some researchers suggest that children as young as 3 years of age have formed food preferences and can provide reliable data about their food preferences (Birch 1979; Alles-White and

Welch 1985). Descriptive sensory and consumer preference studies of RUTFs are very limited and studies to incorporate sweetpotatoes in a RUTF were not found in the literature.

The objective of this study was to evaluate nutritional composition, flavor using descriptive

168

sensory analysis and consumer acceptance of a commercial RUTF, two optimized peanut- sweetpotato formulations (Chapter 4) and one peanut only formulation.

169

Materials and Methods

Runner type roasted peanuts (L value = 50±1) were obtained from a commercial roaster and stored until analysis. Peanut skins were removed and peanuts were ground into a paste using a Robot Coupe Blixer (Ridgeland, MS, USA) following a grind-cool protocol by

Sanders and others (1989), which included grinding for 1-min followed by 30s rest for a total of 3-min grind. Sweetpotato flakes (SPF) were obtained from Bruce Foods (Wilson, NC,

USA) and milled using a Wiley Mill (Paul N, Gardner Co., Inc., Pompano Beach, FL, USA) fitted with a 0.5 mm sieve. Nonfat milk powder, peanut oil and sucrose were purchased locally. Stabilizer (BPX-50) was obtained from Caravan Ingredients (Lenexa, KS, USA).

Vitamin/mineral mix was obtained from Fortitech (Schenectady, NY, USA). Based on previous optimization studies (Chapter 4) using peanut, sweetpotatoes, milk powder and with the addition of ingredients that are commonly found in RUTFs, three test RUTF formulations

(Table 1) and MANA, a commercially produced RUTF (MANA Nutrition, Matthews, NC,

USA) were used in this study.

Color, Moisture, Fat

Hunter L (0=black, 100=white), a (+value=red, -value=green), b (+value=yellow, - value=blue) value of all samples were determined using a HunterLab Colorimeter

(HunterLab DP-9000™ Reston, VA, USA). Moisture content was determined gravimetrically by drying samples at 130ºC for 6 hours in a LXD Series Despatch forced air oven (Despatch Industries, Minneapolis, MN, USA). Fat content was measured by Nuclear

170

Magnetic Resonance (NMR) Spectroscopy using a Minispec mq-one analyzer (Bruker Optics

Inc. Billerica, MA, USA). All samples were prepared and analyzed in triplicate.

Viscosity

Viscosity (Pa.s) measurements were performed using the Brookfield Digital

Viscometer HB (Brookfield Engineering Laboratories, Stoughton, MA) equipped with T spindle-D mounted on a helipath stand. Measurements were conducted on 200 g samples in jars at a shear rate of 50 rpm at 22±2 °C. All samples were prepared and analyzed in triplicate.

β-Carotene Analysis

Determination of β-carotene was accomplished using a modified procedure described by Grabowski and others (2008). Five grams of sample were mixed with 4 grams of calcium carbonate, 4 grams of celite and 50 mL of methanol. The slurry was homogenized for one minute using a Brinkman (Heidolph North America, Elk Grove Village, IL, USA) overhead stirrer, and filtered through Whatman # 1 and # 42 filter paper. The solid material was removed and 100 mL acetone:hexane (1:1) was added, homogenized and then filtered on the previously used filter paper. The filtering step was repeated once again. The filtrate was added to a separatory funnel and the collection flask was rinsed with 10 mL deionized water and added back to the funnel. A saturated solution of NaCl was added in 1 mL aliquots up to

10 mL if necessary to facilitate a separation of the phases. The aqueous phase was removed and the upper layer was transferred to a volumetric flask and bought to a final volume of 50

1% mL. The extracts were read on a spectrophotometer at 450 nm and Beer’s law (A=E /cm

171

*b*c) was used to calculate the concentration of the β-carotene in the solution. β-carotene is reported as ug/100g RUTF. All samples were prepared and analyzed in triplicate.

Amino Acid Content

The amino acid compositions were determined using a Hitachi L-8900 Amino Acid

Analyzer (Hitachi High-Technologies Corp., Schaumburg, IL). Formulation samples (0.5g) and 4mL of 6 N HCl containing 1 % phenol were vortexed for 30s. Then the samples were hydrolyzed using a Discover® SP System (CEM, Matthews, NC) with Synergy™ software, where samples were heated for 10 minutes at 165ºC. As a control, a standard reference sample of peanut paste, SRM 2387 (NIST, Washington, DC) was evaluated along with the samples. The hydrolyzed samples were brought to volume (25mL) with 0.02 N HCl. Before analysis, the samples were diluted 1:10 in 0.02 N HCl and the peanut reference was diluted

1:5 in 0.02 N HCl. Analyses were performed on a Hitachi L-8900 Amino Acid Analyzer

Method C&M-L89-PH. All samples were prepared and analyzed in triplicate.

Descriptive Sensory Analysis (DSA)

Test samples were evaluated by an experienced descriptive sensory panel (n=4, females; n=5, males) using the Spectrum™ universal 15-point intensity scale (Meilgaard and others 1999). The peanut-sweetpotato lexicon was developed using a combination of the peanut lexicon described by Johnsen and others (1988), and the sweetpotato lexicon by

Leksrisompong and others (2012). These lexicons were used to help generate terms to describe a peanut-sweetpotato based product. Each peanut-sweetpotato sample was served at

172

room temperature with deionized water and unsalted crackers. The order of sample presentation was randomized for 4 replications and all samples were coded with 3 digit randomly generated numbers.

Consumer Acceptance Testing

Consumer testing was approved by the North Carolina State University Institutional

Review Board (IRB). Panelists were recruited by notices and e-mail for participation.

Individuals with peanut, sweetpotato or milk allergies were not allowed to participate in the panel.

Samples (n=4) were selected for consumer acceptance testing based on previous optimization of flavor and nutritional attributes. A commercially available RUTF, MANA, was also included in the study. Samples were evaluated with a minimum of one minute between samples accompanied with cracker and water use. Twenty gram samples were served in lidded soufflé cups. Consumer acceptance testing (n=121) was conducted at the

NCSU Sensory Service Center and consent forms were provided consistent with human subjects approval followed by a ballot. Evaluations were conducted individually in sensory booths using Compusense ® Five version 5.2 (Compusense, Guelp, ON, Canada) in an enclosed room with positive air flow and free from external aromas and distractions.

Consumers first answered demographic questions and were then instructed to a read concept statement describing the test samples. Questions related to the concept statement were followed by evaluation of overall appearance. The appearance of the sample was

173

scaled using a 9 pt hedonic scale where 1=dislike extremely and 9=like extremely.

Consumers then tasted the sample and were then asked to evaluate overall liking, overall flavor, sweetness, mouthfeel and smoothness using the 9 pt hedonic scale.

Statistical Analysis

Analysis of variance (ANOVA) was generated for color, moisture, fat, viscosity, β- carotene and amino acid using PROC GLM and comparison of means were made using

Duncan’s post-hoc test (SAS version 9.1, Cary, NC). Significance was established at

P<0.05. Descriptive analysis was analyzed using (ANOVA) Fisher’s least significant difference test at P < 0.05 significance level. All consumer acceptance analyses were conducted with XLSTAT (Addinsoft, New York, NY). Liking scores were evaluated by

ANOVA with Fisher’s post hoc test. Just About Right scores were evaluated using Penalty

Analysis and Chi-Square. Consumer consumption interest and product fit questions were evaluated using Kruskal-Wallis with Dunn’s post hoc test. Consumer clusters were evaluated using K-means. Clusters were validated using discriminate analysis. Partial least squares regression was conducted on consumer segmentation, descriptive and consumer data.

174

Results and Discussion

Moisture content was not significantly different (P>0.05) among the three RUTFs formulations and commercial RUTF (Table 2). MANA and 56/0/30 were significantly

(P<0.05) lighter in color, while 49/15/20 and 28/20/30 were darker due to the addition of

SPF. 49/15/20 and 28/20/30 exhibited more red hues than the other samples but b values were similar. Fat contents among all samples were significantly (P<0.05) different except for

28/20/30 the numerical differences were 0.6 percent or less. 28/20/30 contained more added peanut oil since it contained a low percentage of peanuts and more dry ingredients (Table 2).

Fats of vegetable origin are preferred in RUTFs because they are more digestible (Briend and others 2002).

The variation in viscosities among the samples was highly affected by the varied concentrations of ingredients (Table 2). For example, peanut paste is composed of peanut particles suspended in oil, and the individual concentrations added to a formulation can affect viscosity. All formulations contained a commercial stabilizer, composed of 52% monoglycerides and most commercial RUTFs contain a monoglyceride stabilizer. Totlani and Chinnan (2007) concluded that stabilizer concentration is an important factor that affects consistency of peanut products. RUTFs should have a low enough viscosity for the products to be flowable when little pressure is exerted on the packet, and viscous enough for the product to not spill instantly under its own weight (Briend and others 2002). MANA (10.8

Pa.s) and 56/0/30 (10.8 Pa.s) had viscosity similar to honey (10 Pa.s), which is flowable with little pressure applied to a squeezable container but will not spill immediately (Table 2). The

175

viscosity of 49/15/20 (7.7 Pa.s) and 28/20/30 (5.5 Pa.s) was significantly lower (P<0.05) than

MANA and 56/0/30, which indicates that these formulations would flow more readily than honey (Table 2).

Vitamin A is often expressed in retinol equivalents (RE) because retinol is the form of

Vitamin A derived from animal based foods. One RE is equal to 1 µg of Vitamin A (as retinol) or 6 µg β-carotene. β-carotene in formulations 49/15/20 and 28/20/30 was 2040

µg/100g (403 µg RE) and 3070 µg/100g (511 µg RE), respectively (Table 2). MANA and the 56/0/30 formulation contained 350 µg/100g (53 µg RE) and 420 µg/100g (75 µg RE), respectively. There are many factors that influence vitamin A activity, and much is yet to be learned. It is known that processing can promote the release of beta-carotene from the food matrix (Castenmiller and others 1999) and that dietary oils enhance its absorption (beta- carotene is a fat soluble molecule) (Brown and others 2004).

β-carotene content of formulations containing SPF were significantly different

(P<0.05) from each other and much higher than the formulations without SPF (Table 2). The

β-carotene content of orange flesh sweetpotatoes range between 4490 to 22600 µg/100g

(Teow and others 2007; Simonne and others 1993). β-carotene of SPF has been reported to range from 3980 to 5750 µg/100g (Emenhiser and others 1999; Valdez and others 2001).

The SPF used in this study had a β-carotene content of 7490 µg/100g. β-carotene is absorbed in the intestines and stored in the liver where the molecule is cleaved to produce vitamin A.

The conversion of beta-carotene to vitamin A is influenced by the vitamin A status of the individual. The Recommended Dietary Allowance (RDA) for vitamin A is 300-500 µg/d for

176

children ages 6 months to 8 years of age (Food and Nutrition Board 2004). A 100g serving of 28/20/30 would thus provide 100% of the recommended RDA of vitamin A (from β- carotene) for children up to 8 years of age. In cases of SAM the requirement for vitamin A is increased and maybe as high as 1100 µg/100g in RUTFs (WHO 2007). A vitamin and mineral mix containing vitamin A in the form of palmitate ester was added (1.34 g vitamin and mineral mix/100 g formulation) to the formulations and contributed 475 µg of Vitamin

A. In the formulations, the vitamin and mineral mix added 475 µg (475 µg RE) of vitamin A as palmitate ester and β-carotene contributed 53 to 511 µg RE vitamin A. These values equate to 986 µg RE vitamin A in the 28/20/30 formulation which approaches the WHO potential requirement in severe cases of malnutrition (WHO 2007).

The addition of milk powder to peanut paste results in an enhanced supply of essential amino acids and milk powder is included in most RUTF formulatons. The concentrations of threonine, serine, glutamic acid, glycine, methionine, isoleucine, leucine, tyrosine, histidine, arginine, and proline in 28/20/30 were not significantly different (P>0.05) from MANA which also includes milk powder (Table 3). Plants are a good source of protein; however, no plant food provides adequate levels of all 9 essential amino acids.

Peanuts are an excellent source of protein but they lack adequate amounts of lysine, methionine and threonine (Yeh and others 2003; Pominski and Hamlet 1991). Similarly, there were no differences between the concentration of methionine for MANA and 28/20/30, while the other formulations contained significantly less (P<0.05). The lysine concentration in 28/20/30 (5.3 g/100 g) was significantly higher than other formulations. Milk powder has

177

been used extensively to increase certain amino acids that may be insufficient in some foods

(Alid and others 1981; Pominski and Hamlet 1991; Yeh and others 2003). Findings indicate positive associations between intake of relatively complete animal based proteins and weight gain and overall recovery of malnutrition (Marquis and others 1997; Ciliberto and others

2005; Amthor and others 2009; Oakley and others 2010). The arginine content of peanuts is high and 49/15/20 and 56/0/30, the two formulations with highest peanut percentages had significantly higher arginine contents. The combination of the various ingredients used in this study resulted in a more complete amino acid profile than peanuts alone.

The peanut-sweetpotato lexicon consists of 8 flavor, 4 taste, and 1 chemical feeling factor (Table 4). Roast peanutty ranged from 4.2 to 4.1, and was similar in intensity for

49/15/20, 28/20/30 and 56/0/30 (Table 5). These results suggest that SPF had little effect on roast peanutty flavor intensity. MANA contained peanuts along with milk powder but roast peanutty flavor intensity was below the threshold level of 1.0. Cardboardy, an attribute related to lipid oxidation, had an intensity of 1.7 and may have resulted from lipids in peanuts or milk powder (Table 5). A decrease in intensity of roast peanut flavor in peanuts is often associated with an increase in intensity of cardboardy (Bett and Boylston 1992; Nepote and others 2006). Samples 49/15/20 and 28/20/30 had baked sweetpotato/dried apricot/ floral and earthy/canned carrot notes, which is associated with sweetpotato flavor (Leksrisompong and others 2012; Leigthon and others 2010). Baked sweetpotato/dried apricot/ floral is similar to several of the defining characteristics of sweet aromatic in roasted peanuts.

Sensory separation of baked sweetpotato/dried apricot/ floral was one of the most difficult

178

for the trained panel. Sweet aromatic for 56/0/30 was 3.6 which is the intensity commonly found in roasted peanuts (Sanders and others 1990; Greene and others 2008) and was not significantly different from MANA. Sweet aromatic intensity in SPF formulations were not significantly different (P<0.05). In previous optimization studies (Chapter 4) of RUTFs composed of peanut/SPF/milk powder, significant increases in sweet aromatic were detected in samples containing increasing amounts of SPF and a similar trend was observed in this study. Sweet aromatic in peanuts is associated with flavor of caramel, vanilla, molasses and fruit (Johnsen and others 1988) and baked sweetpotatoes were described as having brown sugar aromatics (Lekrisompong and others 2011). As such, sweetpotato aromatics were similar to peanut sweet aromatics and the combination resulted in higher intensity of sweet aromatics in formulations with SPF.

Dark roast intensities were similar in 49/15/20, 28/20/30 and 56/0/30. MANA was significantly higher (P<0.05) in sweet taste (11.2) than all other samples, which ranged from

5.0 to 7.0. Although both peanut-sweetpotato formulations and peanut only formulation contained the same amount of sugar (5%) there were significant (P<0.05) increases in perception of sweet flavor in formulations containing SPF (49/15/20 and 28/20/30).

Sweetpotatoes and peanuts contain sucrose, glucose and fructose (Kays and others 2005;

Horvat and others 1991; Isleib and others 2006). During the cooking of sweetpotatoes, maltose is formed during cooking via the action of α-amylase and β-amylase (Kays and others 2005; Sun and others 1994).

179

One hundred twenty-one consumers participated in the consumer test. Table 6 presents demographic information and consumer characteristics of test participants. As a part of the demographic questions, consumers were asked about personal health, and 80.2% of the participants classified themselves as “moderately health conscious”, 14.1% “extremely health conscious”, and 5.8% “not very concerned about my health (Table 6).” Prior to beginning the test, participants were given a concept statement describing the products, which included phases such as “product is intended to be smooth in texture and sweet”, and

“this or similar products have application in treating malnutrition, providing a healthy high energy food for aging populations, disaster relief and health to overall populations.”

Following the concept statement, when asked “how interested would you be in consuming a nutritionally enhanced peanut-butter like product”, 33.9% responded definitely interested, when the same question was asked about giving the product to an aging parent, 47.2% responded “definitely interested (Table 7).”

For appearance, MANA and 56/0/30 were least liked by consumers, while the appearance of 49/15/30 and 28/20/30 was liked more favorably by consumers (Table 8).

Consumers preferred 49/15/20 and 56/0/30 over MANA and 28/20/30 for overall liking.

These results may suggest that consumer expectation of a peanut based product resulted in overall liking of products with the highest percentage of peanuts. Alternatively, as sweetpotatoes are dried the flavor profile intensifies which may have resulted in lower overall liking of the highest sweetpotato percentage. Overall liking for MANA with a score of 3.9 was the least liked by consumers. For sweetness liking, 56/0/30 was liked best

180

followed by 49/15/20 which suggests that the very high sweet intensity in MANA was not favorable to consumers. These results correlate well with sweet taste evaluated by descriptive analysis, where 56/0/30 had the lowest intensity for sweet taste followed by

49/15/20 contrasted against a very high sweet taste intensity in MANA. In penalty analysis, the overall liking score is penalized for not being categorized as just about right (Lawless and

Heymann 2010). This scale is used to measure a consumer response to a particular attribute and is anchored by the words “Too little” and “Too much” on opposites ends, with “Just About

Right” in the center of the scale (Lawless and Heymann 2010). Penalty analysis, revealed that the flavor of 56/0/30 was just about right (P<0.05), while the flavor of MANA, 49/15/20 and

28/20/30 were penalized as being too strong (Table 9). The expectation of a peanut based formulation by consumers may have resulted in these consumer responses. MANA was penalized for being too sweet (P<0.05), while 56/0/30 had just about right sweetness.

MANA, 49/15/20 and 28/20/30 were perceived as not smooth enough, while 56/0/30 was perceived as having an appropriate level of smoothness. According to consumer responses,

MANA was the RUTF that fit the least among all other RUTFs. Consumers indicated that

MANA was “too sweet”, which is reflected in the lower liking score of MANA.

Partial least squares (PLS) was performed to determine the drivers of liking or major contributors for each consumer segmentation (Figure 1). Variable importance projection

(VIP) scores estimate the importance of each variable in the projection used in a PLS model.

Major contributors for cluster 1 liking were cardboard and sweet taste, while drivers of dislike were roasted peanutty, woody/hulls/skins and raw beany (Figure 1). Cluster 1 (n=24)

181

rated MANA with an overall liking mean score of 7.2, which indicates that they preferred this sample over the others (Table 10). MANA was less characterized by any distinctive flavor and this may explain why all other clusters (2 and 3) which make up the majority of the consumers had low overall liking scores for MANA. Descriptive data further indicates this lack of flavor in MANA due to the absence of key flavor attributes, roast peanutty and baked sweetpotato/dried apricot/floral. Cluster 2 (n=59) drivers of liking consisted of sweet aromatics, earthy/canned carrot, roast peanutty, baked sweetpotato/dried apricot/floral, and dark roast (Figure 1). Cluster 2 scored 56/0/30, 49/15/20 and 28/20/30 with overall liking mean scores of 5.8, 6.6 and 6.2, respectively and they were not significantly different P<0.05

(Table 10). Drivers of dislike were cardboardy and sweet taste. MANA was the least liked by cluster 2 consumers. Drivers of liking for cluster 2 were roasty peanutty, dark roast, raw beany and woody/hulls/skins. Drivers of dislike were sweet aromatic, baked sweetpotato/dried apricot/floral, earth/canned carrot, cardboardy and sweet taste. This explains why cluster 3 scored 56/0/30 as their top choice.

Conclusion

RUTFs formulated with peanuts, sweetpotatoes, milk powder, and other standard ingredients contained very high concentrations of β-carotene which may contribute to prevention of blindness in certain locations. The formulations were more liked by consumers than a currently available RUTF, MANA. The developed formulations may serve as the basis for products that address not only malnutrition but also sight in various populations.

182

References

Alid G, Yanez E, Aguilera JM, Monckeberg F, Chichester CO. 1981. Nutritive value of an extrusion texturized peanut protein. J Food Sci 46: 948-9

Alles-White ML, Welch P. 1985. Factors affecting the formation of food preferences in preschool children. Early Child Dev Care 21:256-76.

Amthor RE, Cole SM, Manary MJ. 2009. The use of home-based therapy with ready-to-use therapeutic food to treat malnutrition in a rural area during a food crisis. J Am Diet Assoc 109:464-67.

Bett KL, Boylston T. 1992. Effect of storage on roasted peanut quality. In A. St. Angelo, Lipid Oxidation in Food. Washington DC: American Chemical Society. pp. 322–343.

Birch LL. 1979. Dimensions of preschool children’s food preference. J Nutr Educ 11:77-80.

Briend A, Lescanne M 2002. Food or nutritional supplement, preparation method and uses. Patent No. US 6,346, 284. pp. 1-10.

Brown MJ, Ferruzzi MG, Nguyen ML, Cooper DA, Eldridge AL, Schwartz SJ, White, WS. 2004. Carotenoid bioavailability is higher from salads ingested with full-fat than with fat- reduced salad dressings as measured with electrochemical detection. Am J of Clin Nutr 80:396-403.

Castenmiller, JJM, West CE, Linssen JPH, van het Hof KH, Voragen AGJ. 1999. The food matrix of spinach is a limiting factor in determining the bioavailability of β-carotene and to a lesser extent of lutein in humans. J Nutr 129:349-55.

Ciliberto MA, Sandige H, Ndekha MJ, Ashorn P, Briend A, Ciliberto HM, Manary MJ. 2005. Comparison of home-based therapy with ready-to-use therapeutic food with standard therapy in the treatment of malnourished Malawian children: a controlled, clinical effectiveness trial. Am J Clin Nutr 81:864-70.

Emenhiser C, Watkins RH, Simunovic N, Solomons N, Bulux J, Barrows J Schwartz SJ. 1999. Packaging preservation of β-carotene in sweet potato flakes using flexible film and an oxygen absorber. J Food Qual 22:63-73.

183

FAO/WHO. 1967. Requirements of vitamin A, thiamine, riboflavine and niacin. Report of a Joint FAO/WHO Expert Group. WHO technical report series. 362, World Health Organization, Geneva.

Food and Nutrition Board. 2004. Institute of Medicine. National Academics. Dietary Reference Index (DRI). Recommended Intakes for Individuals. Washington, DC: National Academy Press.

Grabowski JA, Troung VD, Daubert CR. 2008. Nutritional and rheological characterization of spray dried sweetpotato powder. LWT: Food Sci Technol 41:206-16.

Greene JL, Sanders TH & Drake MA. 2008. Characterization of volatile compounds contributing to naturally occurring fruity fermented flavor in peanuts. J Ag Food Chem 56:8096-102.

Guimón J, Guimón P. 2012. How ready-to-use therapeutic food shapes a new technological regime to treat children malnutrition. Technological Forecasting and Social Change 79:1319- 27.

Horvat RJ, Arrendale RJ, Dull GG, Chapman GW, Kays SJ.1991. Volatile constituents and sugars of three diverse cultivars of sweetpotato. J Food Sci 56: 714-5.

Isleib TG, Pattee HE, Sanders TH, Hendrix KW, Dean LO. 2006. Compositional and sensory comparisons between normal and high oleic peanuts. J Agric Food Chem 54: 1759-63.

Johnsen PB, Civille GV, Vercellotti JR, Sanders TH, Dus CA. 1988. Development of a lexicon for the descripton of peanut flavor. J Sensory Stud 3:9-17.

Kays SJ, Wang Y, McLaurin WJ. 2005. Chemical and Geographical Assessment of the Sweetness of the Cultivated Sweetpotato Clones of the World. J Amer Soc Hort Sci 130:541- 97.

Lawless HT, Heymann H. 2010 Sensory Evaluation of Food: Principles and Practices. Springer, NY: Springer Science Business Media. 2nd Edition.

Leighton CS, Schönfeldt HC, Kruger R. 2010. Quantitative descriptive sensory analysis of five different cultivars of sweet potato to determine sensory and textural profiles. J Sensory Stud 25: 2-18.

184

Leksrisompong PP, Whitson ME, Truong VD, Drake MA. 2012. Sensory attributes and consumer acceptance of sweet potato cultivars with varying flesh colors. J Sensory Stud ISSN 0887-8250 1-11.

Lennema Z, Matilsky D, Hdekha M, Manary MJ, Maleta K, Manary MJ. 2007. A large scale operational study of home-based therapy with ready-to-use therapeutic food in childhood malnutritionin Malawi. Maternal Child Nutr. 3:206-15.

MANA Nutrition. 2012. Mother Administered Nutritive Aid. http://mananutrition.org. Accessed on September 15, 2012.

MANA Nutrition. 2011. Administration of MANA Ready-to-use therapeutic food of children suffering from severe acute malnutrition without complications. pp.1-21.

Marquis GS, Habicht JP, Lanata CF, Black RE, Rasmussen KM. 1997. Breast milk or animal-product foods improve linear growth of Peruvian toddlers consuming marginal diets. Am J Clin Nutr. 66:102-9.

Meilgaard M, Civille GV, Carr BT. 1999. Spectrum terminology for descriptive analysis. 3rd Ed. CRC Press, Boca Raton, FL. pp. 177-94.

Moskowitz HR. 1994. Food Concepts and Products. Just-in-time Development. Food and Nutrition Press, Trumbull, CT. pp. 293-31.

Nepote V, Mestrallet MG, Ryan L, Conci S and Grosso NR. 2006. Sensorial and chemical changes in honey roasted peanuts and roasted peanuts stored under different temperatures. J Sci Food Agri 86:1057-63.

Oakley E, Reinking J, Sandige H, Trehan I, Kennedy G, Malets K, Manary M. 2010. A ready-to-use therapeutic food containing 10% milk is less effective than one with 25% milk in the treatment of severely malnourished children. J Nutr 140:2248-52.

Pominski L and Hamlet CA. 1991. Peanut butter quality as affected by inclusion of non-fat dry milk components. Peanut Sci. 18:16-8.

Popper R, Kroll JJ. 2005. Issues and viewpoints conducting sensory research with children. J Sensory Stud 20:75-87.

185

Sanders TH, Vercellotti JR, Crippen KL, Civille GV. 1989. Effect of maturity on roast color and descriptive flavor of peanuts. J Food Sci 54:475-7.

Sanders TH, Blankenship PD, Vercellotti JR, Crippen KL.1990. Interaction of curing temperature and inherent maturity distributions on descriptive flavor of commercial grade sizes of florunner peanuts. Peanut Sci 17:85-9.

Simonne AH, Kays SJ, Koehler PE. Eitenmiller RR. 1993. Assessment of β-carotene in sweetpotato breeding lines in relation to dietary requirements. Food Comp Anal 6:336-45.

Sun JS, Severson RF, Kays SJ. 1994. Effect of heating temperature and microwave pretreatment on the formation of sugars and volatiles in Jewel sweet potato. J Food Qual 17:447-56.

Teow CC, Troung Van-Den, McFeeters RF, Thompson RL, Pecota KV, Yencho GC. 2007. Antioxidant activities, phenolic and β-carotene contents of sweet potato genotypes with varying flesh colours. Food Chem 103:829-38.

Totlani VM, Chinnan MS. 2007. Effect of stabilizer levels and storage conditions on texture and viscosity of peanut butter. Peanut Sci 34:1-9.

Woodroof JG.1983. Peanut Butter. In J.G. Woodroof ed. Peanuts: Production, Processing, and Products. 3rd ed. Westport, CT. AVI Publishers, pp. 181-227.

WHO. World Health Orgnaization. 2012. Micronutrient deficiencies: Vitamin A deficiency. Accessed on October 17, 2012; http://www.who.int/nutrition/topics/vad/en/

WHO. World Health Organization/World Food Programme/United Nations System Standing Committee on Nutrition/The United Nations Children’s Fund. 2007. Community-based management of severe malnutrition. pp. 1-8.

Yeh J-Y, Phillips RD, Hung Y-C. 2003. Optimizing protein and vitamin-fortified peanut spreads containing soybean or milk powder. J Food Qual 243-56.

186

Acknowledgments

The authors thank the USDA-ARS-MQHRU “peanut” lab for participation on descriptive sensory panels, help with prepping samples, and NCSU Sensory Service Center for their assistance with execution of the consumer study.

187

Table 1. Percentage of ingredients used in optimized peanut-sweetpotato formulations

49/15/20 28/20/30 56/0/30 Peanut Paste 49.16 28.16 56.16 Sweetpotato Flakes 15 20 0 Milk Powder 20 30 30 Peanut Oil 9 15 7 Sucrose 5 5 5 Vitamins and Mineral Mix 1.34 1.34 1.34 Stabilizer 0.5 0.5 0.5

MANA – Commercial RUTF * Commercial stabilizer - BFP 65 PLM

188

Table 2. Color, moisture, fat, viscosity and β-carotene of commercial and formulated RUTFs

MANA 49/15/20 28/20/30 56/0/30 L 69.2 a 48.2 b 48.1 b 69.2 a a 3.6 b 11.0 a 11.2 a 4.4 b b 23.3 a 22.6 a 22.2 a 24.0 a Moisture (%) 2.6 a 2.4 a 2.7 a 2.5 a Fat (%) 31.6 b 31.0 d 35.7 a 31.2 c Viscosity (Pa.s) 10.8 a 7.7 b 5.5 c 10.8 a β-carotene (µg /100g SPF) 350 d 2420 b 3070 a 420 c

MANA is a commercial product and numbers for formulations are percentage of peanuts/sweetpotato flakes/milk powder, respectively. β-carotene concentration in sweetpotato flakes was 7490 µg/100g Means in the same row with different letters are significantly different (P<0.05).

189

Table 3. Amino acid content (g/100 g) in RUTF

Asp Thr Ser Glu Gly Ala Cys Val Met Ile Leu Tyr Phe Lys His Arg Pro MANA 12.6a 4.0a 5.0a 20.0b 2.7b 3.1c 5.2a 4.8bc 1.8a 4.1a 9.7a 3.3b 3.7c 4.2c 2.3a 6.5b 6.8ab 49/15/20 12.5a 3.5b 4.7b 19.5c 4.4a 3.4a 4.2b 4.7c 1.6b 3.9b 8.9b 3.8ab 3.9b 4.2b 2.1b 8.2a 6.2bc 28/20/30 12.0b 3.9a 4.9a 19.9b 2.6b 3.2b 4.4b 5.2a 1.8a 4.3a 9.8a 3.8ab 3.9b 5.3b 2.3a 5.4b 7.1a 56/0/30 11.4c 3.3c 4.6b 20.5a 4.3a 3.4a 3.3c 4.9b 1.5b 3.9b 9.1b 4.6a 4.3a 4.5a 2.1b 8.5a 5.6c

Means in the same column with different letters are significantly different (P<0.05) Aspartic acid (Asp); Threonine (Thr); Serine (Ser); Glutamic acid (Glu); Glycine (Gly); Alanine (Ala); Valine (Val); Methionine (Met); Isoleucine (Ile); Leucine (Leu); Tyrosine (Tyr); Phenylaline (Phe); Lysine (Lys); Histidine (His); Arginine (Arg); Proline (Pro)

190

Table 4. Descriptors and definition of peanut - sweetpotato lexicon Descriptor Definition Aromatics Roast Peanutty The aromatic associated with medium roast peanuts Sweet Aromatic The aromatics associated with sweet material such as caramel, vanilla, and non sulfur molasses Baked/Sweetpotato The aromatics associated with the internal portion of an / Dried Apricot/ Floral orange-flesh sweetpotato with dried apricot and floral notes Dark Roast The aromatic associated with dark roasted peanuts Raw Beany The aromatic associated with light roasted peanuts and having legume-like notes Woody/Hulls/Skins The aromatics associated with base peanut character (absence of fragrant top notes) and related to dry wood, peanut hulls, and skins Earthy/Canned Carrot The earthy aromatics associated with canned carrot Cardboardy The aromatic associated with somewhat oxidized fats and oils and reminiscent of cardboard Tastes Sweet The taste on the tongue associated with sugars Salty The taste on the tongue stimulated by sodium salts Sour The taste on the tongue stimulated by acids Bitter The taste on the tongue associated with bitter agents such as caffeine or quinine Chemical Feeling Factors Astringency The chemical feeling factor on the tongue, described as puckering/dry and associated with tannins or alum

191

Table 5. Descriptive sensory analysis of MANA and RUTF formulations

Roast Baked Sweet Dark Raw Woody/ Earthy/ Cardboardy/ Sweet Bitter Astringency Peanutty Sweetpotato/ Aromatic Roast Beany Hulls/Skins Canned Stale Taste Dried Carrot Apricot/ Floral MANA ND NP 3.3 a ND 1.3 b ND NP 1.7 11.2 a 2.5 a 1.2 a 49/15/20 4.2 a 2.7 a 4.4 b 2.6 a 2.1 a 2.7 a 1.2 a ND 6.3 b 2.6 a 1.0 a 28/20/30 4.1 a 2.8 a 4.6 b 2.7 a 1.8 ab 2.1 b 1.5 a ND 7.0 b 2.5 a 1.5 a 56/0/30 4.2 a NP 3.6 a 2.6 a 2.0 a 2.8 a NP ND 5.0 c 2.4 a 1.0 a

Means in the same column with different letters are significantly different (P<0.05). ND = Below the threshold mean of 1.0 or not detected NP = Not present

192

Table 6. Demographic information and consumer characteristics for RUTF consumer studies

Male 37.2% Gender Female 62.8% 18 or younger 1.7% 19-25 49.6% 26-35 23.1% Age (y) 36-45 6.6% 46-55 9.9% 56-64 8.3% 65 and older 0.8% Yes 16.5% Do you care for children in your home? No 83.5% Yes 5.8% Do you care for an aging parent? No 94.2 Do you make contribution (i.e. monetary, send Yes 30.6% supplies, mission trips) to world humanitarian organizations such as: Compassion International, Feed the Children, World Vision, or Samaritan’s Purse? No 69.4% Extremely health 14.1% conscious Moderately health 80.2% conscious Not very concerned 5.8% Would you describe yourself as being: about my health Not concerned at ----- all

Data represents n=121 consumers

193

Table 7. Consumer consumption interest for RUTF

Definitely not interested 1.7% How interested would you be in Not very interested 6.6% consuming a nutritionally Maybe/Maybe not 19.8% enhanced “peanut butter-like interested product?’ Slightly interested 38.0% Definitely interested 33.9% Definitely not interested 1.7% Not very interested 2.5% How interested would you be in Maybe/Maybe not 15.6% giving a nutritionally enhanced interested “peanut butter-like product” to an Slightly interested 33.1% aging parent? Definitely interested 47.1%

194

Table 8. Liking attribute means from consumer acceptance testing of RUTF

MANA 49/15/20 28/20/30 56/0/30 Appearance liking 4.7 b 6.3 a 6.0 a 4.8 b Overall liking 3.9 c 5.7 a 4.8 b 5.8 a Overall flavor liking 4.1 c 6.0 a 4.7 b 6.2 a Sweetness liking 4.2 c 5.2 b 4.8 b 5.9 a Mouthfeel liking 4.3 c 4.9 ab 4.5 bc 5.3 a Smoothness liking 4.3 c 4.9 ab 4.7 bc 5.7 a

Date represents n=121 consumers Liking attributes were scored on a 9-point hedonic scale where dislike extremely=1 and like extremely=9 Means in rows followed by different letters are significantly different (p<0.05)

195

Table 9. Consumer just about right (JAR) scores for RUTF

CHARACTERISTICS MANA 49/15/20 28/20/30 56/0/30 Too little 26.4% a 10.7% b 15.7% b 29.8% a FLAVOR JAR JAR 28.1% b 47.9% b 38.8% b 57.9% a Too much 45.5% a 41.3% a 45.5% a 12.4% b Too little 8.3% c 35.5% a 32.2% a 21.5% ab SWEETNESS JAR JAR 31.4% b 40.5% b 36.4% b 62.0% a Too much 60.3% a 24.0% b 31.4% b 16.5% b Too little 59.5% a 61.2% a 63.6% a 37.2% b SMOOTHNESS JAR JAR 36.4% b 34.7% b 33.9% b 56.2% a Too much 4.1% a 4.1% a 2.5% a 6.6% a

JAR scales were scored on a 5-point scale where too little =1or2, just about right=3 and too much = 4 or 5; percentage of consumers that selected these options is presented Data represents n=121 consumers Means in rows for individual characteristics followed by different letters are significantly different (p<0.05)

196

Table 10. Overall liking means for cluster segmentation

Formulation Cluster 1 (n=24) Cluster 2 (n=59) Cluster 3 (n=38) 56/0/30 5.7 b 5.8 a 6.8 a 49/15/20 4.1 c 6.6 a 4.3 b 28/20/30 4.5 c 6.2 a 3.1 c MANA 7.2 a 2.6 b 4.1 b

Means in columns followed by different letters are significantly different (p<0.05) Liking attributes were scored on a 9-point hedonic scale where dislike extremely =1 and like extremely =9

197

92.3%

1 Bitter 0.75 Sweet Aromatic Astringency Baked Sweetpotato/Dried28/20/30 0.5 Apricot/Floral Sweet Taste Earthy/ Canned Cardboardy/Stale

0.25 Carrot MANA Woody/Hulls/Skins 49/15/20 Cluster 1

0 t2(31.3%) ClusterDark 2 Roast -0.25 Roast Peanutty A ttributes Raw Beany -0.5

C lusters -0.75 56/0/30Cluster 3

-1 -1 -0.75 -0.5 -0.25 0 0.25 0.5 0.75 1 T reatments t1 (61.0%)

Figure 1. Partial least squares correlation biplot of clusters for RUTF Cluster 1 (n=24), cluster 2 (n=59), cluster 3 (n=38)

198

APPENDICES

199

Appendix 1. Sensory language for skim milk powder and dried dairy ingredients (Drake et al. 2002)

Descriptor Reference Preparation

Cooked/sulfurous -Heated Milk -heat pasteurized skim milk to 85ºC for 45 min

Caramelized/butterscotch -Autoclaved milk, -Caramel syrup -Autoclave whole milk at 121ºC for 30 min

Sweet aromatic/cake mix -Pillsbury-White cake mix, -Vanillin -Dilute a tablespoon of caramel syrup in 400 mL skim milk

-dilute 5 mg of vanillin in skim milk

Cereal/grass-like -Breakfast cereals (corn flakes, oat and wheaties) -soak one cup cereal into three cups milk for 30 min and filter to remove cereals

Barny -p-cresol -20 ppm in skim milk

Brothy/potato-like -Kroger-canned white potato slices -remove the sliced potatoes from the broth

-methional -few drops of 20 ppm in methanol in sniffing jars

Animal/gelatin-like/ wet dog -Knox-unflavored gelatin -dissolve one bag of gelatin (28g) in two cups of distilled water

Milkfat/lactone Heavy cream, Delta dodecalactone 40 ppm on filter paper

Fried fatty/painty - (E,E)-2,4-decadienal -2 ppb in skim milk

Fishy Fresh fish with skin, Canned tuna juice

Mushroom/metallic - fresh mushroom -slice fresh mushroom in skim milk for 30 min and filter to remove mushroom slices

Papery/cardboard -cardboard paper -soak pieces of cardboard paper in skim milk overnight

Burnt/charcoal -oven toasted bread slice

Vitamin/rubber -Enfamil – liquid Polyvisol vitamins

Diacetyl Diacetyl Diacetyl, 20 ppm on filter paper

Earthy/musty Potting soil, odor reminiscent of damp basement

Sweet taste -sucrose -5% sucrose solution

Salty -NaCl -2% NaCl solution

200

Appendix 2. Continued

Sour -citric acid -1% citric acid solution

Bitter -caffeine -0.5% caffeine solution

Umami Monosodium glutamate -1% MSG in water

Astringent -tea -soak 6 tea bags in water for 10 min

201