A kinetic study on extraction and transformation phenomena of phenolic compounds during red wine fermentation

Original articleDevelopment of a fortified peanut-based infant formula for recovery of severely malnourished children Nimsate Kane,1Mohamed Ahmedna2* & Jianmei Yu2 1 Institut de Technolo

Original article

Development of a fortified peanut-based infant formula for

recovery of severely malnourished children

Nimsate Kane,1Mohamed Ahmedna2* & Jianmei Yu2

1 Institut de Technologie Alimentaire, Route des Pe´res Maristes - Dakar Hann, Senegal

2 Food and Nutritional Sciences, North Carolina A&T State University, 1601 East Market Street, Greensboro, Greensboro, NC 27410, USA

(Received 3 December 2009; Accepted in revised form 4 June 2010)

temperature on the yield and protein content of spray-dried peanut milk were evaluated Peanut-based infantformulas (PBIF-75) was developed using spray-dried peanut milk and a premix of vitamins and minerals.Physical properties, approximate composition, minerals, vitamins and amino acid composition, and caloricvalue of PBIF-75 were evaluated and compared to those of soya-based infant formula (SBIF) and WorldHealth Organization (WHO) F-75 Spray-dried peanut milk yield was 15–18% with a protein content of30–45%, depending on the extraction pH and temperature PBIF-75 was nearly identical to WHO F-75 interms of amino acid profile, most vitamins and minerals, proximate composition, caloric value, andphysicochemical characteristics such as water activity and colour However, few of the vitamins and minerals

in PBIF-75 will require further adjustment to fully meet WHO’s requirements of a recovery formula forundernourished infants

Introduction

Malnutrition represents the direct cause of about

300 000 child deaths per year in developing countries

(Black et al., 2003; Muller et al., 2003) The incidence of

stunting in some African countries is very high among

children, reaching 50% in some areas (Enwonwo et al.,

2004) The most common form of malnutrition

encoun-tered in African countries is severe malnutrition (PEM),

which is attributed to the lack of protein rich foods such

as meat and dairy products, and the low buying power

of people Internationally, the World Health

Organiza-tion (WHO) has recommended the F-75 and F-100

(fortified-high-energy milk containing 75 or 100 Kcal,

respectively) formula for the treatment of severely

malnourished children The F-75 and F-100 consists of

dried-skim milk, sugar, oil, as well as vitamins, and

mineral supplements The F-75 differs from F-100 in

that it contains dextrin maltose and cooked rice or corn

flour The F-75 is given to children at the beginning of

nutritional recovery regimen to cover their basic needs

in protein and energy After the initial recovery using

F-75, F-100 is administrated for promotion of weight

gain (Briend, 2003) It is important to note that these

WHO-recommended formulas do not contain ironbecause severely undernourished children are known

to have an excess of iron, which leads to a higher rate ofdeath in this group (Ramdath & Golden, 1989) Ironsupplementation is, therefore, only recommended afterthe children have recovered from severe nutritionaldeficiency The F-75 and the F-100 formulas have beentested in many areas in Senegal and around the worldand their efficacy in promoting weight gain has beenproven F-75 and F-100 have, however, some limita-tions They can be used only in recovery centres withstrict nutritional and medical supervision to control thequality of the formula as to prevent microbial contam-ination that may harm the children (Briend, 2003) Formore convenience, F-100 formula has been replaced by

a solid formula made of peanut butter and skim milk.The product is made up of 30% peanut butter, 20%skim milk, 28% sugars, 20% vegetable oil, and 2%vitamin and mineral supplement This new formula hasbeen successfully tested in Senegal with a reportedweight gain significantly higher than that of F-100 (Diop

et al., 2003) Another advantage of this solid formula isthat it can be used at home without nutritional ormedical supervision Furthermore, the risk of contam-ination is reduced because the product is dry enough toprevent microorganism growth and the formulation isconsumed without mixing with water Because of its

*Correspondent: Fax: +336 334 7239;

e-mail: Ahmedna@ncat.edu

convenience, this formulation has been produced

com-mercially by Plumpynut, Nutriset, and Malaunay

France for emergency child nutritional recovery

throughout the world (Briend, 2003)

Peanut (Arachis hypogeae) is a legume widely grown

and is abundant in some African countries such as

Senegal It contains 26% of protein, 49.5% of lipids,

and about 16% of carbohydrates of which 9% is dietary

fibre (Khalil & Chunghtai, 1983) Therefore, peanut is a

good source of protein and energy Peanut protein

contains all essential amino acids that are indispensable

for health maintenance, although methionine, cystine,

and tryptophan are in relatively low amount (Kholief,

1987) Among plant proteins, the essential amino acids

content of peanut protein is relatively high (Ahmed &

Young, 1982; Andersen et al., 1998) Depending on the

variety, peanut oil contains 54–82% of

monounsatu-rated fatty acid and 4.6–28.4% polyunsatumonounsatu-rated fatty

acids, and a low level of saturated fats (Andersen et al.,

1998) Such a desirable fat profile is known to lowers

LDL cholesterol, total cholesterol, and triglycerides that

are associated with heart disease, diabetes, and obesity

(Kris-Etherton, 1999; The Peanut Institute, 2005)

The high nutritional value of peanut has made it one

of the most important crops in the developing world

Peanut has the potential to be used as raw material for

peanut-based milk While soya milk has gained

increas-ing popularity worldwide, the consumption of peanut

milk is limited, particularly, in the developed countries,

because of peanut allergy issue and unpleasant beany

flavour (Lee & Beuchat, 1992) In developing countries,

however, peanut allergy is uncommon and consumers

are familiar with and actually like the beany flavour of

peanuts Therefore, peanut milk may represent a

nutri-tionally balanced beverage that can be used as a

substitute of milk in areas where dairy products are

scarce and⁄ or prohibitively expensive (Schmidt et al.,

1980; Rubico et al., 1987; Lee & Beuchat, 1992) The

most recent body of knowledge on peanut milk includes

studies in the 1990s on the use of peanut milk as

buttermilk for the preparation of salad dressing (Lee &

Beuchat, 1991) and in the formulation of coffee whitener

(Abdullah et al., 1993) In other studies,

chocolate-flavoured and strawberry-chocolate-flavoured peanut milk

pro-cessed in a pilot plant was UHT-sterilised and studied

for shelf stability (Ismail et al., 1995, 1996)

Traditionally, oriental consumers have used mild

alkali such as sodium bicarbonate (NaHCO3) to

improve the flavour and mouth feel of common dry

beans (Bourne et al., 1976) Similar process (e.g sodium

bicarbonate soaking) can be used in the preparation of

peanut milk with reduced beany flavour The production

of spray-dried peanut milk may represent a new

value-added use of peanut while addressing the nutritional

needs of undernourished children The objectives of this

study are to (i) investigate the combined effect of pH and

temperature on the yield and protein content of drypowdered peanut milk and, (ii) using powered peanutmilk to develop a shelf-stable infant formula that meetsthe nutritional requirement of the WHO F-75 formulafor the recovery of malnourished children

Materials and methodsPreparation of peanut milkPeanut kernels (Virginia type) purchased from GoodEarth Peanuts, Inc (Skippers, Virginia, USA) were used

in peanut milk production Peanut milk was preparedfollowing a modified procedure of Lee & Beuchat(1992) Peanuts were visually inspected to removediscoloured kernels that might be moulded or poten-tially contaminated with aflatoxin Screened peanutkernels were then rinsed with water to remove anyaflatoxin residues on the surface of kernels Peanutkernels were then soaked overnight in a 0.5% NaHCO3solution at a kernel to solution ratio of 1:2 Water wasthen drained, and peanuts were washed with tap waterthen mixed with water at a kernel to water ratio of 1:5 asdescribed by Abdullah et al (1993) The kernel⁄ watermixtures were allowed to soak for 5 min at treatmenttemperatures of 25, 50, and 100C before they wereground using an Oster-14 speed blender (Blue Chill,Inc., Boca Raton, FL, USA) The resulting slurry wasfirst filtered using a double layer of cheese cloth,followed by filtration through a Whatman No 1 filterpaper The resulting peanut milk was then homogenisedfor 10 min using a Brinkman PT 2100 Polytronhomogenizer (Westbury, NY, USA) The pH of peanutmilk was adjusted with NaOH (0.1N) or HCl (0.1N) tothe desired value (6, 7, and 8) A BU¨CHI Mini SprayDryer B-191 (Westbury, NY, USA) was used to dryaqueous peanut milk The spray drying parameters such

as temperature, aspiration, and flow rate were set at

of 9 water-soluble vitamins, 4 fat-soluble vitamins, andseven minerals) The latter was custom-formulated byFortitech (Schenectady, New York, USA) The detailedcomposition of the micronutrient premix used to fortifypeanut milk is shown in Table 1 An adequate amount

of spray-dried peanut milk was mixed with the priate amount of carbohydrates and micronutrientpremix to mimic the composition of WHO F-75.Specifically, the PBIF-75 formula contained 24 g of

appro-dry full fat peanut milk, 60 g of sucrose, 12.5 g of corn

starch, and 3.5 g of micronutrient premix These

ingre-dients were mixed thoroughly to ensure the uniformity

of ingredients within the dry powder matrix The

peanut-based infant formula was analysed for

physico-chemical properties and proximate composition

Soya-based infant formula ‘‘Isomil’’ was used as reference for

the physicochemical characteristics, and F-75 formula

were used as references for the desired target nutritional

values

Measurement of physical properties

The physical properties including water activity and

spectral properties (Hunter’s L-, a- and b-colour scale

value) of peanut milk powder and PBIF-75 were

determined using a Minolta CM-3500d

Spectrophoto-meter (Ramsey, new Jersey, USA) and an Aqua Lab

Water activity-meter (Pullman, Washington, USA),

respectively Nonfat dry cow milk (Nestle) and a based infant formula (SBIF) were evaluated for thesame physical properties and used as references

soya-Proximate composition analysis of dry peanut milks andPBIF-75

Crude protein was analysed using a Truspec CNElemental Analyzer (LECO Corporation, Warrendale,

PA, USA) and a conversion factor of 6 25 The totalfat⁄ lipid was determined using a Sotex Avanti 2050automated fat analyzer (Foss, Hoganas, Sweden).Briefly, 2 g of spray-dried PEANUT MILK wereextracted with 85 mL of petroleum ether (Fisher Scien-tific, New Jersey, USA) at 155 C for 73 min Thedifference in weight of the extraction cups before andafter lipid extraction served as a measure of the lipidcontent of each sample The moisture was determinedusing a HG63 Mettler-Toledo Moisture Analyzer (Grei-fensee, Switzerland) The ash content was determinedaccording to AOAC method 923.03 (AOAC, 2003)using a 30400 Fisher Scientific Furnace, while thecarbohydrate contents of samples were determined bydifference

Mineral quantificationThe mineral content was determined using an Optima

3300 ICP (Perkin Elmer, Norwalk, CT, USA) Samples

of 0.25 g PBIF were digested in 10 mL of HNO3 and

2 mL of hydrogen peroxide in a Marsx microwave oven(Matthews, NC, USA) for 25 min at 210C Thedigested samples were analysed by ICP along withstandard mineral references The run time for thequantification of minerals was 30 min, at an operatingtemperature of 210C Concentrations of minerals inpeanut milk samples were calculated using standardcurves developed using known concentrations of eachtest mineral

Determination of vitamins in peanut milk and PBIF-75The extraction of water-soluble vitamins was performedwith distilled water and that of fat-soluble vitamins withhexane The low detection limit of certain vitamins didnot allow all the vitamins to appear in the samechromatogram Thus, different methods were used toquantify vitamins in peanut-based infant formula,separately Specifically, Vitamins A and E were deter-mined by the method of DeVries & Silvera (2002);vitamin C was determined by method of Deutsch &Weeks (1965); vitamin D was determined by AOACmethod 2002.05; vitamins B1, B2, B6, pantothenic acid,and niacin were determined by AOAC methods 942.23,970.65⁄ 981.15, 961.15, 945.74 ⁄ 960.46, and 944.13,respectively (AOAC, 2000) Folic acid was determined

Table 1 Composition of premix used in fortification of peanut-based

infant Formula*

Components Level ⁄ 5.2 g Mix % RDI

Vitamin A (as Palmitate, USP-FCC) 5000 IU 15.0

Vitamin D3 (as Cholecalciferol, USP-FCC) 1200 IU 15.0

Vitamin E (as acetate, USP-FCC) 22 IU 15.0

Biotin (USP) 0.1 mg 20.0

Folic Acid (USP) 0.35 mg 20.0

Niacin (as Niacinamide, USP-FCC) 10 mg 15.0

Pantothenic Acid (as D-Calcium

Pantothenate, USP)

3 mg 15.0 Vitamin B1 (as Thiamin Mononitrate,

USP-FCC)

0.7 mg 15.0 Vitamin B12 (as Cyanocobalamin, USP) 1 mcg 20.0

Vitamin B2 (as Riboflavin, USP-FCC) 2 mg 15.0

Vitamin B6 (as Pyridoxine HCl, USP) 0.7 mg 15.0

Vitamin C (as Ascorbic Acid, USP-FCC) 100 mg 15.0

Vitamin K1 (as Phytonadione, USP) 40 mcg 15.0

Calcium (as Tricalcium Phosphate, FCC) 317.4 mg 7.0

Copper (as Copper Amino Acid Chelate

(Cu 10%))

2.7 mg 10.0 Iodine (as Potassium Iodide, USP-FCC) 77 mcg 10.0

Magnesium (as Magnesium Oxide, USP) 90.7 mg 7.0

Phosphorous (as Dipotassium Phosphate,

Sodium (as Sodium Chloride, FCC) 42 mg 7.0

Zinc (as Zinc Oxide, USP) 20.3 mg 10.0

PBIF-75 = Peanut-Based Infant Formula mimicking the World Health

Organization (WHO) F-75 formula for nutritional recovery of

malnour-ished children.

*Values are actual specifications certified by Fortitech (Schenectady,

New York, USA), the manufacturer of the premix.

by the method of Phillips et al (2006); biotin was

determined by a microbiology method described by

Augustin et al (1985)

Determination of amino acid profile of PBIF-75

Amino acids such as Alanine, Arginine, Aspartic

Acid, Cystine, Glutamic Acid, Glycine, Histidine,

Isoleucine, Leucine, Lysine, Methionine, Phenylalanine,

Proline, Serine, Threonine, and Valine were

determina-tion by AOAC method 994.12 (AOAC 2003)

Trypto-phan and Tyrosine were determined by AOAC method

982.30 (AOAC, 2000)

Statistical analysis

Data was analysed statistically using SAS software

Version 8 (SAS Institute; Cary, NC, USA) Means and

standard deviations of replicated data (3–5 replications)

were used in summary statistics Analysis of variance

and t-test were used to evaluate significance of difference

between means for conditions used in peanut milk

extraction and between PBIF-75 and WHO F-75 infant

formula, respectively Differences between means were

judged significant at the 5% significance level

Results and discussions

Effect of extraction pH and temperature on peanut milk

yield and protein content

Figure 1 shows that the highest dry milk yield (17.49%

w⁄ w) was obtained at room temperature and native pH

(pH 7.0) Statistical analysis indicated that there was no

significant difference in yield among dry milk extracted

at different pH–temperature conditions Therefore, the

yield of dry peanut milk was not significantly affected by

the processing conditions used in this study However,

room temperature extraction is advantageous given the

relatively high yield and saving of energy⁄ time required

for heated extraction Figure 1 also shows higher

extraction temperature and pH yielded peanut milk

powder with higher protein content (46.2%, at 100C

and pH 8) This is contrary to the results reported by

Lee & Beuchat (1992) who found that the protein

content in aqueous peanut milk decreased significantly

as the temperature increased The high protein content

reported in this paper may be explained by discarding of

fat layer that formed in heated sample The latter would

have reduced relative concentration of protein in the

extract prior to spray drying Overall, peanut milk

powders produced at all temperature–pH combinations

used in this study had higher protein content than

nonfat dry milk (33.4%) and milk powder made from

cowpea (22.8–26.8%) as reported by Akinyele & Abudu

(1990) Considering yield and potential production cost,

room temperature and native pH were selected as thepotentially most cost effective conditions to preparepeanut milk powder for use in infant formulas

Physical properties of peanut milk and PBIF-75The physical properties of dry peanut milk and peanut-based infant formula (PBIF-75) were evaluated andcompared to Nestle’s nonfat dry milk and soya-basedinfant formula (SBIF), respectively (Table 2) As shown

in Table 2, the water activity of dry peanut-based milkwas slightly higher than that of Nestle’s nonfat dry milk.However, change in spray drying parameters shouldenable adjustment of the final moisture level in peanutmilk to the desired value Peanut milk showed whiteness(L-values) similar to that of the commercial cow milksample Furthermore, there was no significant difference

in a-values between peanut-based milk and the ences of dry cow milk The negative a-values of peanut-based milk indicate a slight greenish colour of peanutmilk consistent with the study of Lee & Beuchat (1992).The b-values indicated that the reference cow milk tend

refer-to be more yellowish than peanut milk These opticalcharacteristics indicate that the colour and appearance

of peanut-based milk would be highly acceptable givenits close similarity to that of dried cow milk

0 5 10 15 20 25

The physical properties of PBIF-75 were evaluated

and compared to those of commercial soya-based infant

formula (SBIF) used as reference (Table 2) The water

activity and L-value of PBIF-75 were identical to those

of SBIF PBIF-75 had a slightly higher negative a-value

but a lower b-value than SBIF This difference was

easily noticeable because of the pronounced yellow

colour of SBIF Overall, the colour characteristics of

PBIF-75 compared favourably to those of SBIF and in

some cases PBIF displayed better colour (less

yellow-ness) characteristics than the commercial SBIF

Proximate composition of peanut milk powder and infant

formula PBIF

Peanut milk had similar protein and moisture contents

to nonfat dry milk, significantly higher fat content, and

lower ash and carbohydrate contents (Table 3)

How-ever, as expected, the fat and carbohydrate contents of

peanut milk were significantly higher and lower, tively, than nonfat dry milk because of the defatting andhigh lactose level in the latter Powered peanut milkexhibited lower ash content than nonfat dry milkbecause of the richness of milk in minerals such ascalcium However, the low mineral and carbohydratecontent of dry peanut milk can be easily compensatedthrough mineral and carbohydrate fortification Peanut-based milk offers advantage of being both protein andenergy rich, and therefore, has the potential to respond

respec-to the protein and energy needs of children in Senegaland other areas where animal proteins and dairyproducts are scarce Following fortification, the proteinand ash contents of the peanut-based infant formula(PBIF-75) were significantly higher than those of WHO-

75, while fat and moisture content were same as that ofWHO-75 (Table 3) Only carbohydrates were slightlylower in PBIF-75, a deficiency that can be correctedthrough adjustment of carbohydrates in the fortificationmix

Mineral content of peanut-based infant formula (PBIF-75)The mineral contents of dry full fat peanut milk,PBIF-75 and reference WHO F-75 are included inTable 4 Peanut milk base exhibited very low Ca, Cu,and Zn contents but relatively high K, Mg, Na, and Pcontents Following fortification, minerals such as Ca,

Mg, and P were present in PBIF at levels higher thanthose of WHO’s F-75, while K and Zn were lower thanthe requirements of F-75 The Na levels in the twoformulas were identical While potassium (K) was high

in the micronutrient premix (Table 1), it was low inPBIF-75 (Table 4) where its final concentration inPBIF-75 was lower than WHO’s target This observeddeficiency might be caused by an incomplete extraction

of K from PBIF-75 and⁄ or an analytical tion Our data also shows that phosphate (P) content in

underestima-Table 2 Physical ⁄ spectral properties of peanut milk powder and

peanut-based infant formula (PBIF-75) in comparison with Nestle’s

nonfat dry milk and Soya-based Infant formula (SBIF)

*Product property means with different superscript are significantly

different at 5% significance level.

PBIF-75 = Peanut-Based Infant Formula mimicking the World Health

Organization (WHO) F-75 formula for nutritional recovery of

33.5 ± 0.2 a

8 ± 0.016 a

5.53 b

2.47 Moisture (%) 6.00 ± 2.50 a

5.30 ± 2.50 a

3.16 ± 0.069 a

2.50 b

0.66 Fat (%) 33.5 ± 2.17 a

2.1 ± 1.2 b

13 ± 0.577 a

12.3 a

0.7 Ash (%) 2.4 ± 0.17 a

7.3 ± 0 a

5.64 ± 0.07 a

3.13 b

2.53 Carbohydrates (%) 24.00 51.00 70.20 a

82.00 b

)11.80 PBIF-75 = Peanut-Based Infant Formula mimicking the World Health Organization (WHO) F-75 formula for nutritional recovery of malnourished

the peanut milk base was as high as 444.4 mg⁄ 100 g

powder This content yielded a P level of 170 mg⁄ 100 g

in the formula PBIF-75, still higher than the WHO

requirement (147 mg⁄ 100 g) Hence, no P fortification

was needed It is noteworthy to point out that F-75

micronutrient supplement did not contain Fe because

severely undernourished children tend to have an excess

of Fe and are at risk of death from excess of Fe (Briend,

2003) Likewise, PBIF-75 was not fortified with Fe Zinc

and potassium content, however, should be maintained

high because these two elements are essential for muscle

building Therefore, the mineral composition of the

premix used to fortify peanut milk needs to be

reformulated to incorporate higher levels of Zn and K

to meet the WHO requirements for F-75

Vitamin content of peanut-based infant formulas (PBIF)

Table 5 shows that peanut milk contained much less

vitamins than F-75 except for Biotin and Niacin After

fortification with the vitamin–mineral premix, all of the

vitamins required in therapeutic milk⁄ formula were

present in PBIF-75 at significantly higher concentrations

Vitamins A, C, D3, and K1 were present in PBIF-75 at

lower than the recommended levels as indicated by the

negative differences on the right column of Table 5

These vitamins need to be added at higher level in the

fortifying premix used to formulate PBIF-75 The

negative differences indicate deficits that might be

because of losses during processing as these vitamins

are unstable On the other hands, vitamins B6, Biotin,

and Niacin were present in PBIF-75 at the levels130–300% higher than those recommended in F-75 Thisexcess may be caused by higher level of these vitamins inpeanut milk used as base ingredient, which was notconsidered when the premix was formulated Therefore,the vitamin composition of the premix needs to beadjusted to account for the native B6, biotin, and niacincontents of dry peanut milk base Reformulation shouldalso boost the levels of vitamin A, C, D, K1, and B12

Amino acid profile of peanut milk and peanut-based infantformula (PBIF-75)

The amino acid profile in Table 6 reveals that peanutmilk contained all amino acids at levels higher than theWHO-recommended levels for F-75 Mixing of peanutmilk with other ingredients resulted in PBIF-75 with amajority of amino acids at or above the desired levels forWHO F-75 primarily because of the high proteincontent of peanut milk Among the 18 amino acidsanalysed, 10 (aspartic acid, threonine, serine, glutamicacid, glycine, alanine, leucine, phenylalanine, histidine,and cystine) were present in PBIF-75 at levels higherthan the reference F-75, while four (valine, isoleucine,arginine, and tryptophan) were slightly below therecommended levels and 4 amino acids including pro-line, lysine, tyrosine, and methionine were significantlylow in PBIF-75 Therefore, the limiting amino acids inPBIF-75 are lysine, tyrosine, and methionine This is

Table 4 Mineral contents of peanut milk, PBIF-75 compared to that of

milk PBIF-75 à

WHO F-75 à

PBIF-75 = Peanut-Based Infant Formula mimicking the World Health

Organization (WHO) F-75 formula for nutritional recovery of

malnour-ished children.

*

Data are means concentrations of minerals with standard deviation

<10% mean values Values for WHO F-75 are single values obtained from

specification sheet.

†

The positive differences indicate excess of mineral in PBIF-75, while

negative difference indicate deficit in PBIF-75 compared to WHO F-75.

milk PBIF-75 WHO F-75 Vitamin A Retinol (IU) <100 1530 3125 )1595 Vitamin C (mg) <1.00 57.0 62.5 )5.5 Vitamin E (IU) <0.100 14.9 13.75 1.15 Vitamin B1 (mg) 0.270 0.69 0.437 0.253 Vitamin B2 (mg) 0.130 1.27 1.25 0.02 Vitamin B6 (mg) 0.140 1.02 0.437 0.583 Folic acid (lg) 29.10 287 218 69 Vitamin B12 (lg) <0.100 0.42 0.625 )0.205 Biotin (lg) 38.90 75.2 22 53.2 Panthotenic acid (mg) 0.880 3.19 1.875 1.315 Niacin (mg) 7.60 26.9 6.25 20.65 Vitamin D3 (IU) 99.10 730 750 )20 Vitamin K1 (lg) <10.0 <10.0 25 15 PBIF-75 = Peanut-Based Infant Formula mimicking the World Health Organization (WHO) F-75 formula for nutritional recovery of malnour- ished children.

expected because these essential amino acids are scarce

in plant proteins It is noteworthy, however, that the

amino acid profile of PBIF-75 does not reflect any

amino acid fortification A reformulated PBIF-75 could

be easily fortified by including lysine, tyrosine, and

methionine in the fortification premix as to meet or

exceed the WHO recommendation for F-75

Evaluation of the caloric value of peanut-based infant

formula (PBIF-75)

WHO protocol recommends that children suffering

from severely protein-energy deficiency be treated with

two different recovery formulas, F-75 and F-100 The

first treatment formula (F-75) provides 75 Kcal kg)1of

body weight per day, while the second treatment

formula (F-100) provides 100 Kcal kg)1day)1for rapid

weight gain Based on current formulations, 17.5 g

PBIF-75 formula is needed to yields 75 kcal (Fig 2),

which is slightly higher than the 16.2 g of WHO F-75

formula needed to provide equivalent energy level The

first step of treatment will cover the basic needs in

protein and energy at the lowest protein content WHO

recommends that the beginning of the treatment behypoproteinic because starting with a high proteintreatment in children who have kwashiorkor may induceanorexia because of disturbance of hepatic metabolismand urea synthesis Treatment with F-75 provides lowsodium for feeding children who have oedema (Briend,2003) The second treatment consists of administering ahigh protein and high calorie diet to promote rapidweight gain for a rapid recovery of protein-energydeficient children

ConclusionPeanuts were successfully used to develop spray-driedfull fat dried peanut milk powder with physicochemicaland nutritional properties that compare favourably tothose of commercially available dry cow milk, specifi-cally, nonfat dry milk Among nine combinations oftemperatures and pH, room temperature and native pHwere selected as the best condition to produce peanutmilk Full fat peanut milk exhibited high nutritionalvalue because of the high protein and fat contents inpeanuts Peanut-based infant formula (PBIF-75) devel-oped by fortifying peanut milk according to the spec-ifications of the World Health Organization for its F-75nutritional recovery formula, was nearly identical toWHO F-75 in terms of amino acid profile, vitamin–mineral content, energy content, proximate composi-tion, and physicochemical characteristics such as wateractivity and colour However, as formulated, the proteincontent of PBIF-75 was higher than that of WHO-75,which may be undesirable in the first phase of recovery

of severely malnourished children where excess proteinaggravates the metabolic imbalance (Bhan et al., 2003).Hence, future reformulation of PBIF-75 should targetlower level of protein through addition of higher

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00

PBIF-Table 6 Amino acid profile of peanut milk and PBIF-75 compared to

(g ⁄ 100 g)

PBIF-75 (g ⁄ 100 g)

WHO F-75 Aspartic acid 2.79 1.06 0.428 0.632

Data shown are means with standard deviations <5% of the mean

values as certified by Medallion Labs, Minneapolis MN, USA.

†

The negative differences indicate that the amino acid of PBIF-75 is

deficient compared to WHO F-75, while positive differences indicate that

PBIF-75 has excess amino acids.

PBIF-75 = Peanut-Based Infant Formula mimicking the World Health

Organization (WHO) F-75 formula for nutritional recovery of

malnour-ished children.

amount of carbohydrates for energy In addition, the

premix used for micronutrient fortification should be

adjusted to increase the contents of vitamins A, C, D, K,

and B12in PBIF-75 to the levels in WHO-75 as well as

refinement of the levels of some of the minerals to

closely match WHO’s F-75 Therefore, future R&D

efforts will further optimise the protein content and

micronutrient composition of PBIF-75 and evaluate its

shelf-life and sensory quality

The peanut-based infant formula (PBIF-75) proposed

in this study can be a competitive alternative to the

milk-based WHO F-75 formula for recovery of malnourished

children The optimised formulation will contain less

protein through the dilution effect from adding more

carbohydrates The dry formula is designed to be shelf

stable for use in environment where refrigeration is

unavailable However, to ensure safety, boiled water is

to be used for reconstitution of the formula This should

further help reduce potential osmolarity effect The

advantages of a peanut-based infant formula are (i)

shelf-stability as it is made of dried peanut milk; (ii) high

micronutrient contents as it is fortified with vitamins

and minerals to the levels of the WHO

recommenda-tions; (iii) potentially low cost because the base

ingre-dient is a local crop

Acknowledgment

This project was financially supported by USAID

Peanut CRSP NCA32U, and Agricultural Research

Program at North Carolina A&T State University

References

Abdullah, A., Malundo, T.M.M., Rsurreccion, A.V.A & Beuchat,

L.R (1993) Descriptive sensory profiling for optimizing the formula

of a peanut milk based liquid coffee whitener Journal of Food

Science, 58, 120–123.

Ahmed, E.M & Young, C.T (1982) Composition, quality and flavour

of peanuts In: Peanut Science and Technology (edited by H.E Pattee

& C.T Young) Pp 655–688 Yoakum, Texas 77995: American

Peanut Research and Education Society, Inc.

Akinyele, I.O & Abudu, I.A (1990) Acceptability and nutrient

content of milk substitutes from four cultivars of cowpeas (Vigna

unguiculata) Journal of Food Science, 55, 701–703.

Andersen, P.C., Hill, K., Gorbet, D.W & Brodbeck, B.V (1998).

Fatty acid and amino acid profiles of selected peanut cultivars and

breeding lines Journal of Food Composition Analysis, 11, 100–111.

AOAC (2000) American Organization of Analytical Chemists

(AOAC) International Official Methods 942.23, 970.65 ⁄ 981.15,

961.15, 945.74 ⁄ 960.46, 944.12, 944.13 and 982.30.

AOAC (2003) American Organization of Analytical Chemists

(AOAC) International Official Methods 923.03 and 944.12.

Augustin, J., Klein, B.P., Becker, D & Venugopal, P.B (1985)

‘‘Methods of Vitamin Assay’’, 4th edn New York: John Wiley and

Sons, Pp 43, 50–58 and 541–549.

Bhan, M.K., Bhandari, N & Bahl, R (2003) Management of the

severely malnourished child: perspective from developing countries.

British Medical Journal, 326, 146–151.

Black, R., Morriss, S & Bryce, J (2003) Where and why are 10 million children dying every year? Lancet, 361, 2226–2234 Bourne, M.M., Escueta, E.E & Banzon, J (1976) Effect of sodium alkalis and salts on pH and flavour of soy milk Journal of Food Science, 41, 62.

Briend, A (2003) Prise en charge de la malnutrition severe de l’enfant

en milieu tropical Medicine Tropicale, 63, 70.

Deutsch, M.J & Weeks, C.E (1965) Vitamin C (Total) in Vitamin Preparations Journal of AOAC, 48, 1248–1256.

DeVries, J.W & Silvera, K.R (2002) Determination of Vitamins A (Retinol) and E (alpha-Tocopherol) in Foods by Liquid Chroma- tography: Collaborative Study Journal of AOAC International, 85, 424–434.

Diop, H., Dossou, N., Ndour, M., Briend, A & Wade, S (2003) Comparison of the efficacy of a solid ready to use food and a liquid milk-based diet for the rehabilitation of severely malnourished children: a randomized trial American Journal of Clinical Nutrition,

78, 302–307.

Enwonwo, C., Phillips, R.S & Ibrahim, C.D (2004) Nutrition and Oral Health in Africa International Dental Journal, 54, 344–351 Ismail, Y.S., Rustom, M.H., Lopez, L & Baboo, M.N (1995) UHT- Sterilized peanut beverages: changes in physicochemical properties during storage Journal of Food Science, 60, 378–383.

Ismail, Y.S., Rustom, M.H., Lopez-Leiva, M.M & Nair, B.M (1996) UHT-Sterilized peanut beverages; kinetics of physicochemical changes during storage and shelf-life prediction modeling Journal

of Food Science, 61, 198–203.

Khalil, J.K & Chunghtai, M.I.D (1983) Chemical composition and nutritional quality of five peanut cultivars grown in Pakistan Plant Foods for Human Nutrition, 33, 63–70.

Kholief, T.S (1987) Chemical composition and protein properties of peanuts Zeitschrift fu¨r Erna¨hrungswissenschaft, 26, 56–61 Kris-Etherton, P (1999) High monounsaturated fatty acid diet lowers both plasma cholesterol and triacyl glycerol concentrations Amer- ican Journal of Clinical Nutrition, 70, 1009–1015.

Lee, C & Beuchat, L (1991) Functional and sensory properties of salad dressing containing fermented peanut milk Journal of Food Science, 56, 1665–1666.

Lee, C & Beuchat, L (1992) Chemical, physical, and sensory characteristics of peanut milk as affected by processing conditions Journal of Food Science, 57, 401–405.

Muller, O., Gurrenne, M., Kouyate, B & Becher, H (2003) he association between protein-energy malnutrition, malaria morbidity, and all cause of mortality in West African children Tropical Medicine and International Health, 8, 507–511.

Phillips, K.M., Ruggio, D.V., Ashraf-Khorassani, M & Haytowitz, D.B (2006) Difference in Folate Content of Green and Red Sweet Peppers (Capsicum annuum) Determined by Liquid Chromatogra- phy-Mass Spectrometry Journal of Agricultural and Food Chemistry,

Schmidt, R.H., Sistrunk, C.P., Rchter, R.L & Cornell, J.A (1980) Heat treatment and storage effects on texture characteristics of milk and yogurt systems fortified with oil seed proteins Journal of Food Science, 45, 471.

The Peanut Institute (2005) Protein, good fats in peanuts reduce blood pressure and heart disease http://www.peanut-institute.org/ 111705_PR.htlm

Original article

Effect of different precooking methods on chemical composition

and lipid damage of silver carp (Hypophthalmichthys molitrix)

muscle

Mahmood Naseri,1Masoud Rezaei,1* Sohrab Moieni,2Hedayat Hosseni3& Soheil Eskandari4

1 Department of Fisheries, Tarbiat Modares University, Noor, P.O Box 46414-356, Mazandaran, Iran

2 Department of Food Science and Technology, School of Agriculture, Tehran University, Tehran, Iran

3 National Nutrition and Food Technology Research Institute, Shaheed Beheshti University, Tehran, Iran

4 Bureau of Food and Drug Laboratories, Tehran, Iran

(Received 3 March 2010; Accepted in revised form 11 June 2010)

composition and lipid quality of silver carp fillets was evaluated The changes in protein, fat and moisturewere found to be significant for all the treatments (P£ 0.05) The iron content in the samples subjected tosteam-cooking increased; however, the other precooking methods did not change the mineral contents(P‡ 0.05) The free fatty acid content of the fillets did not change by the different precooking methods, whilethiobarbituric acid (TBA) values increased for oven- and microwave-cooked fillets and remained constant inthe steam-cooked samples Conjugated diene and browning colour formation levels significantly increased inthe oven-baked fillets Oven-baking and microwave-cooking marginally affected the fatty acid composition

of the silver carp On comparing the raw and precooked fillets, steam-cooking was found to be the bestprecooking method on retaining nutritional constituents

Introduction

Seafood products have attracted considerable attention

as a source of high amounts of important nutritional

components to the human diet (Ackman, 1989)

Seafo-ods have high protein content, low saturated fat and

also contain vitamins and minerals Mineral

compo-nents such as magnesium, calcium, zinc, iron and

phosphorus are essential for human nutrition (Erkan

& O¨zden, 2007)

However, in recent years, the fishing sector has

suffered from dwindling stocks of traditional species as

a result of dramatic changes in their availability This

has prompted fish technologists and the fish trade to pay

more attention to aquaculture techniques as a source of

fish and other seafood products (Aubourg, 2001) Silver

carp (Hypophthalmichthys molitrix) is an extensively

cultured species Aquaculture production of silver carp

is the highest of any finfish species in the world;

especially important in the Asia-Pacific region that has

an annual global production of nearly 4.2 million metric

tons (Gheyas et al., 2009) Accordingly, the

consump-tion of value-added products of this species has recentlyincreased

Assurance of both the quality and safety of seafoodwill be a major challenge faced by humans in this newcentury (Aubourg et al., 2005) In this sense, wild andfarmed fish species are known to deteriorate after deathbecause of the action of different factors that can besummarised as microbiological growth, endogenousenzyme activity, nonenzymatic lipid oxidation andbrowning The relative incidence of each damage mech-anism will depend on the kind of technological processapplied (Aubourg, 2001; Horner, 1997; Pigott & Tucker,1987)

Canning belongs to the most important means of fishpreservation (Horner, 1997; Aitken & Connell, 1979)

As with any other treatment, canning should bedesigned to retain as much as possible of all thenutritional constituents present in the initial matter toserve human nutrition (Aubourg, 2001) The extensiveheat treatment involved in canning steps substantiallyalters the nature of the raw material so that, in effect, aproduct with different characteristics is formed

Precooking is a critical thermal process before ing and designated to destroy pathogenic micro-organisms, endogenous enzymes, secure certain

retort-*Correspondent: Fax: +98 122 6253499;

e-mail: rezai_ma@modares.ac.ir

sensory properties and product stability for canning

(Aubourg, 2001) In canneries, fish is usually precooked

by steam, but sometimes it is treated by other cooking

methods (FAO, 1988) The fish species and the cooking

method used may be determinant factors for the content

of nutritional constituents (lipids, proteins, minerals and

vitamins) in the final product Some of the major

changes that occur during processing and final

prepa-ration of heated food are because of oxidation The

polyunsaturated fatty acids (PUFA), such as

eicosapen-taenoic acid (EPA) and docosahexaenoic acid (DHA),

are considered to be especially susceptible to oxidation

during heating and other culinary treatments (Sant’Ana

& Mancini-Filho, 2000)

However, research concerning the quality changes

that might occur during canning and thermal treatment,

such as the mechanisms of damage taking place in silver

carp during precooking relatively unknown In the light

of this situation, this study was conducted to determine

the influence of three precooking methods (steaming,

oven-baking and microwave-cooking) on the

composi-tion (proximate, mineral and fatty acid profile) and lipid

oxidation of silver carp

Materials and methods

Sample procedures

Seventeen silver carp (H molitrix) weighing between 2.3

and 3.0 kg were obtained from a local fish farm

(Khuze-stan, Iran) during the winter 2009 They were eviscerated,

washed and immediately transported to the laboratory in

ice-containing boxes Fresh fish were washed with water

several times to remove adhering blood and slime, they

were then prepared using common household practices,

such as removing head, tail and fin yielding two fillets

The fillets were randomly divided into thirty

homo-genous portions of 900 g, which were assigned to three

repetitions of each one of the three cooking methods and

to the raw group that was used as a reference

Precooking methods

For purposes of this study, fillets were divided into three

equal lots representing the following precooking

treat-ments: steam-cooked, oven-cooked and

microwave-cooked For all precooking treatments, fillets were

placed belly-side down on perforated trays and heated

At the time of processing, mean core temperature of

fillets were monitored by thermocouple (Aidin Scientific

trade mark, Tehran, Iran), when the mean core

temper-ature of fillets reached almost to 65C, process was

complete and the equipments turned off (FAO, 1988)

To prepare oven-baked fillets, the oven temperature

was set at 175C for 60 min Microwave-baked fillets

were prepared in a domestic microwave oven (LG

model Solar Dom, Seoul, Korea) at potency 900 w, for

7 min Steam-cooked fillets were prepared in a tal retort at 102–103C for 48 min

horizon-Samples of raw or cooked fish fillets were ised and used to determine proximate composition,mineral and fatty acid profile as well as the level of freefatty acids (FFAs), conjugated dienes, brown colourformation, fluorescence shift and thiobarbituric acidindex (TBA-i)

homogen-Proximate compositionThe moisture content of raw and cooked fillets weredetermined by drying in an oven at 105C until aconstant weight was obtained (AOAC, 1995) Crudeprotein content was calculated by converting the nitro-

(6.25· N) Fat was determined by the method described

by the Bligh & Dyer (1959) Ash content was determined

by dry ashing in a furnace at 525C for 24 h (AOAC,1995)

Mineral analysesMineral contents of raw and precooked samples (cal-cium, copper, iron, zinc and sodium) were analysed bymeans of atomic absorption spectrophotometer using aShimadzu Spectra atomic absorption (AAS) model AA-

680 by AOAC method (1995)

Analysis of lipid damageThe FFA content was determined by the method ofEgan et al (1981) Results are expressed as percentage

of oleic acid The TBA-i (mg malondialdehyde per kgfish flesh) was determined by the method described inthe Pearson’s composition and analyses of food (1991).Conjugated diene (CD) formation was measured at

233 nm (Kim & Labella, 1987) The results were

CD = (B· V) ⁄ w, where B is the absorbance reading

at 233 nm, V denotes the volume (mL) of the sample,and w is the mass (mg) of lipid extract measured.Browning development was determined by spectropho-tometer at 420 nm in the lipid extract of the edible flesh.The results were calculated using the equation: Brow-ning = A· V ⁄ w, where A is the absorbance reading at

420 nm, V is the volume (mL) of the sample, and w isthe amount (mg) of the lipid sample (Smith et al.,1990).Formation of fluorescent compounds was determinedwith a Perkin-Elmer LS 5B fluorimeter (Perkin-Elmer,Norwalk, CT, USA) by measurements at 393⁄ 463 and

327⁄ 415 nm, as the method described by Aubourg et al.(1998) Relative fluorescence (RF) was calculated as:

RF = F⁄ Fst, where F is the fluorescence measured ateach excitations⁄ emission maximum, and Fst is the

fluorescence intensity of a quinine sulphate solution

(1 lg mL)1 of 0.05 m H2SO4) at the corresponding

wavelength The fluorescence ratio (FR) was calculated

as the ratio between both RF values: FR = RF393–

463 nm⁄ RF327–415 nm The aqueous phase resulting

from the lipid extraction (Bligh & Dyer, 1959) was used

to evaluate the FR value

Fatty acid composition

Fatty acid methyl esters (FAME) were prepared by

methylation of the triacylglycerols, as described by

Cronin et al (1991), and analysed using a Shimadzu

17A (Kyoto, Japan) gas chromatograph equipped with

flame ionisation detector and fused silica capillary

column (50 m – 0.32 mm and 0.20 mm of Carbowax

20M) The column temperature was programmed at

2C min)1from 150 to 240C The injection port and

detector were maintained at 220 and 245C,

respec-tively The carrier gas was hydrogen (1.2 mL min)1), the

make-up gas was nitrogen (30 mL min)1) and the split

used was 1:100 The identification of normal fatty acids

was carried out by comparing the relative retention

times of FAME peaks from samples with standards

from Restek and the main fatty acids, in order of

abundance, were confirmed using another Shimadzu

17A (Japan) gas chromatograph

Statistical analysis

Data from the different chemical measurements were

subjected to one-way analysis of variance (P < 0.05)

Comparison of means was performed using Duncan

method

Results and discussion

Proximate composition

The changes in moisture, ash, protein and fat content of

samples after precooking processes are shown in

Table 1 The proximate composition of raw fillets is

comparable to that of observed by Siddaiah et al (2001)

and Ali et al (2005) for silver carp

The moisture content of the fish fillets ranged from74% to 67%, which decreased after cooking (Table 1).The ash content decreased after cooking, except for thesteam-cooked fillets, the protein content increased aftercooking in all evaluated methods and fat contentdecreased after cooking processes (P £ 0.05) (Table 1).The decrease in the moisture content has been described

as the most prominent factor that causes protein, fat andash contents alter significantly in cooked fish fillets(Garcı¢a-Arias et al., 2003) Accordingly, the increase inprotein content of cooked silver carp fillets could beexplained by the reduction of moisture (Table 1)

Mineral compositionTable 2 shows the composition of the mineral elements

in raw and precooked fillets The contents of gated mineral elements in raw and precooked fishsamples were found to be in the range of 3.05–4.19 mg kg)1for copper, 32.18–40.70 mg kg)1for iron,71.45–82.85 mg kg)1 for zinc, 425.6–529.4 mg kg)1 forsodium and 315.5–534.7 mg kg)1 for calcium Accord-ing to these data, calcium had the highest concentration,followed by sodium, zinc, iron and copper The highconcentration of calcium could be because of the bonynature of silver carp This result is in accordance withthe finding that was reported by Steiner-Asiedu et al.(1991) for fresh water bony fish

investi-The applied precooking methods had little or no effect

on the concentration of sodium, zinc, calcium and copper(P‡ 0.05) In the previous studies, it was found that theprocessing and cooking methods had no effect on themineral contents of fish (Gall et al., 1983; Steiner-Asiedu

et al., 1991) In this investigation, after processing theiron values were the highest for microwave-cookedsamples when compared to the other precookingmethods However, it might be because of the individualdifference of samples

Lipid damageLipid hydrolysisComparison of the initial raw fish before and afterprecooking (all methods) showed that the thermal

Table 1 Proximate composition of raw and cooked silver carp

Results are means ± SD Means within the same row that have no

common letters are significantly different (P £ 0.05).

Table 2 Mineral composition of raw and cooked silver carp (mg kg)1)

Raw Steam-cooked Oven-baked Microwave-cooked

process did not lead to a significant variation in the FFA

content of the fish muscle (P‡ 0.05) This result is in

agreement with those of Rodrı¢guez et al (2008) FFA

formation often occurs as a result of catalysis by

endogenous enzymes, and only microbial effects would

be significant after the end of the lag phase (Whittle

et al., 1990) However, deactivation of enzymes, as a

result of the heating process, would prevent the release

of FFAs owing to lipase activity in the cooked samples

(Weber et al.,2008) Furthermore, during a thermal

treatment, breakdown of high-molecular weight

(trigly-cerides and phospholipids, namely) lipids would be

likely to occur and be the source of new FFA formation

(Gallardo et al., 1989; Yamamoto & Imose, 1989) On

the one hand, Weber et al (2008) reported that the loss

of volatile FFA occurred during heating, leading to a

decreased FFA content Insignificant change in FFA in

this study can be explained by deactivation of enzymes

or FFA volatilisation

The formation of FFA itself does not lead to

nutritional losses However, accumulation of FFA has

been related to some extent to lack of acceptability,

because FFA is known to have detrimental effects on

protein solubility and causes texture deterioration by

interacting with proteins (Sikorski & Kolakowska, 1994)

and oxidises faster than higher-molecular weight lipid

classes by providing a greater accessibility (lower steric

hindrance) to oxygen and other pro-oxidant molecules

(Labuza, 1971)

Lipid oxidation

Results concerning the conjugated diene formation are

shown in Table 3 As a result of oven precooking,

significant increase in diene contents was detected

(P£ 0.05) However, this damage index did not show

significant difference between raw fish and steam- or

microwave-precooked fillets (P‡ 0.05)

The measurement of the diene absorbance at 233 nm

is considered as a very sensitive method of detecting the

beginning of lipid oxidation and has often beenemployed (White, 1995) Conjugated diene levels inoven-baked fillets increased But this index did not showsignificant difference between steam- and microwave-cooked samples with raw fillets (P‡ 0.05) However,when a thermal treatment such as precooking isinvolved, an important thermal breakdown of conju-gated diene is likely to occur so that their assessmentwould not allow an accurate tool for assessing lipiddamage progress (Aubourg et al., 1995; Lubis & Buckle,1990)

The formation of secondary oxidation products wasmeasured by means of the TBA-i Marked increases insuch compounds were obtained as a result of microwaveand oven precooking No significant difference wasobtained between steam-precooked and raw fish(P‡ 0.05) The secondary oxidation compound forma-tion resulted in an interesting tool to assess the chemicalchanges produced as a result of the cooking process Inthis sense, previous research already accounts forcarbonyl formation during cooking in tuna, salmonand sardine fishes (Aubourg et al., 1995; Rodrı¢guez

et al., 2008; Yamamoto & Imose, 1989)

Formation of interaction compoundsInteraction compound formation was measured bymeans of the fluorescent compound and browningdevelopment in the silver carp muscle before and afterprecooking (Table 2) The browning measurement formuscle lipid was increased after oven cooking(P£ 0.05) The elevation of interaction compoundscould be explained by the obtained results of CD andTBA These results agree with previous work carried out

on cooking of salmon (Rodrı¢guez et al., 2008) Nodevelopment of browning colour was observed in silvercarp muscle as a result of cooking by steam ormicrowave

The formation of fluorescence compounds also calledtertiary oxidation compounds (Aubourg, 1999) is theresult of the interaction between lipid oxidation prod-ucts (primary and secondary) and protein-like moleculespresent in fish muscle The detection of fluorescencecompounds (dF) in the silver carp muscle varied little asresult of steam and oven cooking However, microwave-cooking has increased in dF content (P£ 0.05) We didnot find studies evaluating the effect of oven- ormicrowave-cooking on the dF content of fish fillets.Although there is a report of significant increase in dFcontent after steam-cooking of two tuna species (Au-bourg et al., 1995)

Fatty acid compositionThe profile of the most important fatty acids of the silvercarp and precooked fillets are shown in Table 4 The

Table 3 Assessment of lipid damage and interaction compound

Results are means ± standard deviation Means within the same row that

have no common letters are significantly different (P £ 0.05).

FFA, free fatty acid (as g oleic acid ⁄ 100 g)1lipid); TBA, thiobarbituric acid

(as mg MDA kg)1muscle); CD, conjugated diene; BFC, browning colour

formation; dF, fluorescence shift were measured as expressed in

Materials and methods.

most abundant fatty acids found in raw and precooked

silver carp fillets were oleic acid (C18:1 x9), palmitic

acid (C16:0) and palmitoleic acid (C16:1 x7) These

findings are in agreement with those obtained by Mieth

et al.(1989) and Vujkovic et al (1999) Silver carp fillets

also showed considerable amounts of palmitoleic acid

(C16:1 x7), stearic acid (C18:0), linoleic acid (C18:2 x6)

and DHA (C22:6 x3)

Steam-cooking had no effect on the fatty acid content

of silver carp, although microwave- and oven-cooking

marginally affected some fatty acids content The

changes were not homogeneous for the different fatty

acids because some fatty acids decreased, some

in-creased, while the others had no change (Table 4) In

microwave-cooked samples, the content of palmitoleic

(C16:1 x7) and linoleic acid (C18:2 x6) was decreased

and DHA (C22:6 x3) was increased, likewise in

oven-cooked fillets the content of palmitoleic (C16:1x7) and

DHA (C22:6 x3) was decreased and the amount of oleic

acid (C18:1x9) was increased (P£ 0.05)

In all four groups (raw, steam-cooked, oven-cooked

and microwave-cooked samples) of silver carp fillets,

monounsaturated FAs were the dominant class of fatty

acids and PUFs were the least one This result is similar

to the finding that was reported by Vujkovic et al

(1999) Comparison of raw fillets with all groups of

precooked samples showed that the amounts of unsaturated and monounsaturated fatty acids did notalter by processing Similar results were obtained byGarcı¢a-Arias et al (1994) during steaming of whitetuna and Aubourg et al (1990) during processing ofalbacore

poly-The x3⁄ x6 ratio has been suggested to be a usefulindicator for comparing relative nutritional values offish oils It was suggested that a ratio of 1:1–1:5 wouldconstitute a healthy human diet (Osman et al., 2001).Raw and precooked silver carp had the x3⁄ x6 ratiowithin the recommended ratio However, this indexshowed no significant difference between raw fish andprecooked samples (steam-, microwave- and oven-cooked fillets)

ConclusionsAll of the three evaluated precooking methods changedproximate composition and lipid oxidation parameters

of silver carp (H molitrix) fillets Changes in proximatecomposition were more prominent in oven- and micro-wave-cooked fillets Oven- and microwave-cooked filletshad increased levels of conjugated diene, TBA-i,browning colour formation and fluorescence compoundindicating oxidative changes, but these indices in

Table 4 Fatty acids composition of raw and precooked silver carp

Fatty acids Raw Steam-cooked Oven-baked Microwave-cooked 14:0 1.51 ± 0.1 a

Values are percentage of total fatty acid expressed as mean ± SD of three separate determinations.

Means within the same row that have no common letters differ significantly (P £ 0.05).

SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid.

samples subjected to steam-cooking were constant The

changes in mineral contents of precooked samples (all

methods) were found to be insignificant with the

exception of increases in iron content of microwave

precooked samples Steam precooking had no effect on

the fatty acid content of silver carp although microwave

and oven precooking marginally affected some fatty

acids content Steaming appeared to be the best

precooking method concerning proximate composition,

oxidative stability and the fatty acid profile It seems this

method retains as much as possible of all the nutritional

constituents present in the initial silver carp fillets

Acknowledgments

The authors thank Ms Mahdiyeh Abbasi and Mrs

Masoumeh Fekri for their great assistance in chemical

testing The help of Ms Farahnaz Gaffari is also greatly

appreciated

References

Ackman, R.G (1989) Fatty acids In: Marine Biogenic Lipids, Fats

and Oils (edited by R Ackman) Pp 103–137 Boca Raton, USA:

CRC Press.

Aitken, A & Connell, J (1979) Fish In: Effects of heating on food

stuffs (edited by R Priestley) Pp 219–254 London: Applied Science

Publishers Ltd.

Ali, M., Iqbal, F., Salam, A., Sial, F & Athar, M (2005) Comparative

study of body composition of four fish species in relation to pond

depth Journal of Environmental Science and Technology, 2, 359–364.

AOAC (1995) Official Methods of Analyses of Association of

Analytical Chemist, 15th edn.Washington, DC: AOAC.

Aubourg, S (1999) Review: recent advances in assessment of marine

lipid oxidation by using fluorescence Journal of the American Oil

Chemists Society, 76, 409–419.

Aubourg, S (2001) Loss of quality during the manufacture of canned

fish products Food Science and Technology International, 7, 199–

215.

Aubourg, S., Carmen, G.S & Gallardo, J (1990) Change in flesh

lipids and fill oils of albacore during canning and storage Journal of

Agricultural and Food Chemistry, 38, 809–812.

Aubourg, S., Medina, I & Perez-Martin, R (1995) A comparison

between conventional and fluorescence detection methods of

cook-ing-induced damage to tuna fish lipids Zeitschrift fu¨r

Lebensmitte-luntersuchung und-Forschung A, 200, 252–255.

Aubourg, S., Medina, I & Gallardo, J (1998) Quality assessment of

blue whiting (Micromesistius poutassou) during chilled storage by

monitoring lipid damages Journal of Agricultural and Food

Chem-istry, 46, 3662–3666.

Aubourg, S., Pineiro, C., Gallardo, J & Barros-Velazquez, J (2005).

Biochemical changes and quality loss during chilled storage of

farmed turbot (Psetta maxima) Food Chemistry, 90, 445–452.

Bligh, E.G & Dyer, W.J (1959) A rapid method of total lipid

extraction and purification Canadian Journal of Biochemistry and

Physiology, 37, 911–917.

Cronin, D., Powell, A.R & Gormley, R (1991) An examination of the

(n-3) and (n-6) polyunsaturated fatty acid and status of wild and

farmed Atlantic salmon (Salmo salar) Irish Journal of Food Science

& Technology, 15, 53–62.

Egan, H., Kirk, R.S & Sawyer, R (1981) Oil and Fat in Pearsons

Chemical Analysis of Foods New York: Longman Scientific and

Technical Press, 536 pp.

Erkan, N & O¨zden, O¨ (2007) Proximate composition and mineral contents in aqua cultured sea bass (Dicentrarchus labrax), sea bream (Sparus aurata) analyzed by ICP-MS Food chemistry, 102, 721–725 FAO (1988) Manual on fish canning Rome, Italy: FAO Techni- cal paper Available at: http://www.fao.org/docrep/003/t0007e/ t0007e00.htm.

Gall, K.L., Otwell, W.S., Koburger, J.A & Appledorf, H (1983) Effects of four cooking methods on proximate, mineral and fatty acid composition of fish fillets Journal of Food Science, 48, 1068– 1074.

Gallardo, J.M., Aubourg, S & Pe¢rez-Martı¢n, R (1989) Lipid classes and their fatty acids at different loci of albacore (Thunnus alalunga): effects of the pre-cooking Journal of Agricultural and Food Chemistry, 37, 1060–1064.

Garcı¢a-Arias, T., Sa¢nchez-Muniz, F.J., Castrillo¢n, A & Navarro, P (1994) White tuna canning, total fat, and fatty acid changes during processing and storage Journal of Food Composition and Analysis, 7, 119–130.

Arias, T., Pontes, E.A´., Linares, M.C., Ferna´ndez, M.C & Sa´nchez-Muniz, F.J (2003) Cooking-freezing reheating (CFR) of sardine (Sardina pilchardus) fillets Effect of different cooking and reheating procedures on the proximate and fatty acid compositions Food Chemistry, 83, 349–356.

Garcı´a-Gheyas, A.A., Woolliams, J.A., Taggart, J.B et al (2009) Heritability estimation of silver carp (Hypophthalmichthys molitrix) harvest traits using microsatellite based parentage assignment Journal of Aqua- culture, 294, 187–193.

Horner, W (1997) Canning fish and fish products In: Fish Processing Technology (edited by G Hall) Pp 119–159 London: Blackie Academic and Professional, Chapman and Hall.

Kim, R & Labella, F (1987) Comparison of analytical methods for monitoring autoxidation profiles of authentic lipids Journal of Lipid Research, 28, 1110–1117.

Labuza, T (1971) Kinetics of lipid oxidation in foods CRC Critical Reviews in Food Technology, 2, 355–405.

Lubis, Z & Buckle, K (1990) Rancidity and lipid oxidation of salted sardines International Journal of Food Science and Technol- ogy, 25, 295–303.

dried-Mieth, G., Wirth, M., Friedrich, M., Steffens, W & Lieder, U (1989) Cyprinida lipid content, 1st Report, lipid content and fatty acids composition of silver carp (Hypophthalmichthys molitrix) Die Nahrung, 33, 91–93.

Osman, H., Suriah, A.R & Law, E.C (2001) Fatty acid composition and cholesterol content of selected marine fish in Malaysian waters Food Chemistry, 73, 55–60.

Pigott, G & Tucker, B (1987) Science opens new horizons for marine lipids in human nutrition Food Reviews International, 3, 105–138.

Rodrı¢guez, A., Carriles, N., Cruz, J.M & Aubourg, S (2008) Changes in the flesh of cooked farmed salmon (Oncorhynchus kisutch) with previous storage in slurry ice Journal of Food Science and Technology, 41, 1726–1732.

Sant’Ana, L.S & Mancini-Filho, J (2000) Influence of the addition of antioxidants in vivo on the fatty acid composition of fish fillets Food Chemistry, 68, 175–178.

Siddaiah, D., Vidya Sagar Reddy, G., Raju, C.V & Chandrasekhar, T.C (2001) Changes in lipids, proteins and kamaboko forming ability of silver carp (Hypophthalmichthys molitrix) mince during frozen storage Food Research International, 34, 47–53.

Sikorski, Z & Kolakowska, A (1994) Changes in protein in frozen stored fish In: Seafood Proteins (edited by Z Sikorski, B Sun Pan &

F Shahidi) Pp 99–112 New York, USA: Chapman and Hall Smith, G., Hole, M & Hanson, S (1990) Assessment of lipid oxidation in Indonesian salted –dried Marine catfish (Arius thalas- sinus) Journal of the Science of Food and Agriculture, 51, 193–205 Steiner-Asiedu, M., Julshamn, K & Lie, Q (1991) Effect of local processing methods (cooking, frying and smoking) on three fish

species from Ghana: part I Proximate composition, fatty acids,

minerals, trace elements and vitamins Food Chemistry, 40, 309–

321.

Vujkovic, G., Karlovic, D., Vujkovic, I., Vorosbaranyi, I & Jovanovic, B.

(1999) Composition of muscle tissue lipids of silver carp and

bighead carp Journal of the American Oil Chemists’ Society, 6, 475–

480.

Weber, J., Bochi, V.C., Ribeiro, C.P., Victo¢rio, A.M & Emanuelli, T.

(2008) Effect of different cooking methods on the oxidation,

proximate and fatty acid composition of silver catfish (Rhamdia

quelen) fillets Food Chemistry, 106, 140–146.

White, P (1995) Conjugated diene, anisidine value and carbonyl value analyses In: Method to Assess Quality and Stability of Oils and Fat Containing Foods (edited by K Warner & N.A.M Eskin) Pp 159–

178 Champaign, IL, USA: AOCS Press.

Whittle, K., Hardy, R & Hobbs, G (1990) Chilled fish and fishery products In: Chilled Foods (edited by T Gormley) Pp 87–116 The State of the Art New York, USA: Elsevier Applied Science.

Yamamoto, Y & Imose, K (1989) Changes in fatty acid sition in sardines (Sardinops melanosticta) with cooking and refrigerated storage Journal of Nutritional Science and Vitaminol- ogy, 35, 39–47.

compo-Original article

Fat, wheat bran and salt effects on cooking properties of meat patties studied by response surface methodology

Hasibe Tekin,1Cemalettin Saricoban1* & Mustafa Tahsin Yilmaz2

1 Selcuk University, Agriculture Faculty, Food Engineering Department, Konya, 42049, Turkey

2 Erciyes University, Safiye Cikrikcioglu,Vocational College, Food Technology Division, Kayseri, 38039, Turkey

(Received 4 February 2010; Accepted in revised form 21 June 2010)

(processing) variables such as fat (10–30%), wheat bran (5–15%) and NaCl (0–2%) on cooking properties ofbeef patties In addition, the ridge analysis was conducted to find the values of processing variables thatmaximise and minimise the cooking parameters (moisture retention, fat retention, reduction in thickness,reduction in diameter, cooking yield, shrinkage and water-holding capacity) It was found that the moistureand fat retention, reduction in thickness and cooking yield values decreased; however, reduction in diameterand shrinkage values increased, respectively, as the amount of fat increased However, wheat bran additionincreased fat retention, moisture retention, cooking yield and water-holding capacity values of the patties.Increasing NaCl levels decreased water-holding capacity value by its quadratic effect and moisture and fatretention value by its interaction effect with wheat bran

Introduction

Fat and salt have recently been partially replaced with

some non-meat ingredients in the formulations of

ground meat products because of their potential health

risks However, such replacements for reducing the fat

and salt content result in some technological problems

such as increased cooking loses and purge because of

poor fat and water binding as well as undesired change

in colour, texture and flavour of the product (Girard

et al., 1990; Crehan et al., 2000; Yilmaz, 2004; Huang

et al., 2005)

The compositional change in the meat products by

means of incorporation of non-meat ingredients into

their formulations are among the solutions used to

minimise problems related to fat reduction (Claus &

Hunt, 1991; Gregg et al., 1993; Keeton, 1994) and salt

reduction (Ruusunen et al., 2003, 2005) Several dietetic

fibres are among these ingredients Fibre has been

recently added into meat products and is on the increase

nowadays because of its functional properties and

benefits to human health (Vendrell-Pascuas et al.,

2000) Plant origin foods contain fibres; however, dairy

and animal foods do not contain fibre Wheat bran is the

best known source of insoluble dietary fibre In addition,several dietetic fibres have been used in meat products todetermine their possible beneficial effects on cookingproperties of meat products (Yilmaz, 2004, 2005).Because of these functional properties, several grainfibres have been extensively used in the ground meatproducts as a replacement of fat and some other mainconstituents of these products

The use of dietetic fibre in the meat products may not

be the only solution for the aforementioned problemscaused by fat and salt reduction It should be incorpo-rated into the meat products with other main ingredientssuch as fat and salt at optimum levels, which couldachieve the best cooking performance To overcome thechallenge and obtain ideal combination levels, theimportance of determining the optimum levels of thesereplacements in these products comes into prominence

It is well known that sensory properties of the groundmeat products are closely related with high fat andmoisture retention within the matrix of meat products(Anderson & Berry, 2001), which could affect thecooking performance of these products Therefore,finding optimum values of some processing variablesthat optimise the cooking properties of product willprovide more qualified knowledge to obtain betterproducts that have desired technological properties Toachieve this, finding the optimum critical values of the

*Correspondent: Fax: +90 332 241 0108;

e-mail: cscoban@selcuk.edu.tr

processing variables that maximise sensory properties

should be a high priority process Response surface

methodology (RSM) is an effective tool to find these

optimum levels of the processing variables for the

parameters studied (Hunter, 1959) Ridge analysis

involved with RSM computes the levels of these

processing variables that maximise and minimise the

values of textural and sensory parameters (Chwen-Jen

et al., 1996)

Previous researches have focused on the effects of

non-meat ingredients on sensory, cooking and

compo-sitional properties; however, they have not been dealt

with the effects of wheat bran as fat replacers and

extenders on these properties (Dzudie et al., 2004;

Serdaroglu & Degirmencioglu, 2004; Yilmaz, 2004,

2005; Serdaroglu et al., 2005; Serdaroglu, 2006) In

addition, the objectives of these studies were to

determine fat binding and retention ability of these

added ingredients; however, they did not address how

to increase fat and water retention, but still keeping a

satisfactory cooking quality In other words, these

studies have focused on the effect of individual

processing factors with a one-at-a-time approach on

statistical analysis Therefore, information on

com-bined effects and interactions of the major processing

factors on cooking properties of the products is

unclear In addition, no study has appeared to examine

the levels of the processing variables to maximise and

minimise the cooking parameters Therefore, the

objec-tive of this research was to study the effect of

processing variables such as fat (10–30%), wheat bran

(5–15%) and NaCl (0–2%) on cooking properties of

cooked beef patties and to find the levels of processing

variables to maximise and minimise the cooking

parameters

Materials and methods

Patty preparation

The beef samples were prepared as described

(Sarico-ban et al., 2009) Beef was obtained from a local

market in Konya The M longissimus dorsi and

M psoas major muscles were removed from the right

side of the carcass Muscles were then vacuum-packed

and frozen at )20 C and stored in dark for a week

Then, the frozen muscles were placed in a refrigerator

at 4 ± 1C for 12 h to facilitate ease of cutting and

grinding All subcutaneous fat was removed from the

muscles and used as a fat source in the patty

formulation prior to cutting the muscles into cubes

(approximately 2 cm3) The obtained beef cuts and

obtained fat were ground through a 3-mm plate

grinder The spice mix (ground black pepper 0.1%,

red pepper 2% and cumin 0.4%) and onion (1.5%)

was prepared and added into the ground beef The mix

was kneaded for 15 min by hand to obtain patty doughdivided into fifteen experimental batches (Table 1) thatcontained 120 g of ground beef each, with spice mix.Relevant proportions of ground subcutaneous fat(10%, 20% and 30%), wheat bran (5%, 10% and15%) and NaCl (0%, 1% and 2%) were added intoeach batch on the top of total weight (120 g) aspresented in Table 1 Each experimental batch wasseparated into three equal parts to obtain three pattysamples (weighing approximately 40 g) to conductcooking measurements in three replicates for eachexperimental batch Each batch was mixed andkneaded for additional 15 min to obtain homogeneousdough batches Patty dough batches were shaped intopatties with 62.5 mm diameter and 11 mm thicknessusing a metal shaper Then, they were placed on plastictrays, wrapped with polyethylene film and frozen at)18 C until further analysis For cooking procedure,patties were thawed at 4C overnight in a refrigeratorand cooked in a preheated electric grill set at 170–

190 C for 4 min on one side, turned over and cookedfor a further 3 min During cooking, core temperature

of patty batches was monitored by a thermocouple.The final internal temperature reached over all pattybatches was determined to change between 77 and

82C After cooking, patty samples were allowed tocool to 25C in the room conditions before beingplaced in oxygen-permeable bags (low-density poly-ethylene) and cooking measurements were carried out

in these samples Wheat bran (moisture 14.1%, protein13.2%, fat 4.9%, dietary fibre 42.5%, carbohydrate

Table 1 Second-order design matrix used to evaluate the effects of process variables on cooking properties of patties

Runs

Coded variables Uncoded variables

X 1 X 2 X 3

Subcutaneous fat (%)

Wheat bran (%)

NaCl (%)

20.8% and ash 4.5%) was obtained from a local

market in Konya The chemical composition of wheat

bran was determined by AOAC (2000) methods:

protein (920.152), fat (920.39), dietary fibre (985.29),

carbohydrate (997.08) and ash (940.26) and by AACC

(1998) method: moisture (44–40)

Cooking measurements

Moisture retention

Moisture retention value indicates the amount of

moisture retained in the cooked product per 100 g

of sample and was determined using an equation of

Fat retention was calculated according to the following

equation by Murphy et al (1975)

Fat retention (%)¼

ðCooked weight % Fat in cooked patty)

ðRaw weight  %Fat in raw pattyÞ 100

Reduction in thickness and diameter

The reduction in patty thickness and diameter was

determined with a digital calliper (Mitutoyo, Japan)

using the following equations, respectively, as indicated

(Serdaroglu & Degirmencioglu, 2004)

Reduction in thickness (%)¼

ðUncooked patty thicknessCooked patty thicknessÞ

Reduction in diameter (%)¼

ðUncooked patty diamaterCooked patty diameterÞ

Cooking yield

Cooking yield of the patty samples were calculated using

the equation by Murphy et al (1975)

Cooking yieldð%Þ ¼ Cooked patty weight

Uncooked patty weight100

ShrinkageDimensional shrinkage of the patty samples was deter-mined using the following equation (El-Magoli et al.,1996)

Water-holding capacityThe method reported by Ockerman (1985) was used tomeasure the water-holding capacity (WHC) of the rawbeef patties Of patty sample, 0.5 g was placed on thefilter paper (Whatman no 1, 90 mm Ø, stored oversaturated KCl), which was placed between two plexiglassheets and pressed for 20 min by a 1- kg weight Thearea of pressed meat and a spread juice was measured by

a polar planimeter (Placom, Koizumi, Digital eter KP-90 N, Japan), and the WHC was calculated asfollows:

Planim-Free water (%)¼ðTotal surface area  meat film areaÞðmm2Þð6:11ÞTotal moistureðmgÞ in patty sample 100WHC (%) = 100 – Free water

Experimental design and data analysis

A 3-factor-3-level Box–Behnken experimental design(Box & Behnken, 1960) with three replicates at thecentre point was used to study the simultaneous effects

of three compositional variables namely, subcutaneousfat (10–30%), wheat bran (5–15%) and NaCl (0–2%)(Table 1) Three levels of each factor or processingvariable (subcutaneous fat, wheat bran and NaCl) wereprovided in accordance with the principles of the Box–Behnken design Assessment of error was derived fromthree replications of one treatment combination (20%fat, 10% wheat bran and 1% of NaCl) as suggested bythe design (runs 7, 9 and 11, Table 1) Variance for eachfactor assessed was partitioned into linear, quadraticand interactive components to determine the suitability

of the second-order polynomial function and the relativesignificance of these components (Lyons et al., 1999).Shrinkage (%)¼ðRaw thickness  Cooked thicknessÞ þ ðRaw diameter  Cooked diameterÞ

These components (processing variables), levels and

experimental design in terms of coded and uncoded are

presented in Table 1 The following second-order

poly-nomial equation of function Xiwas fitted for each factor

assessed:

Y¼ b0þX3

i¼1

biXiþX3 i¼1

biiX2iiþX3

i¼1 i<j

X3 j¼1

bijXiXj

where Y is the estimated response; B0, Bi, Bii, Bij are

constant coefficients Xi, Xj, which are defined as the

coded independent variables, are the per cent

concen-trations of fat, wheat bran and NaCl For each

parameter assessed, the compositional variables were

divided into linear, quadratic, interactive, lack of fit and

error components to determine the suitability of the

second-order polynomial function and the significance

of variables being assessed The significance of the

equation parameters for each response variable was

assessed using the F test (Table 2) The analysis was

performed using uncoded units

The majority of generated models adequately

explain the variation of the responses with satisfactory

R2 values (R2> 0.90) and non-significant lack of fit,

which indicated that most variations could be well

explained by the quadratic models and can be

considered adequate, because the probability level of

F was P < 0.01 (Thompson, 1982) The

computa-tional work, including the surface and contour

graphical presentations of the response surface

mod-els, was performed using a Statistica for Windowssoftware package (Statsoft, Tulsa, OK, USA) JMPstatistical package software (Version 5.0.1.a; SASInstitute Inc., Cary, NC, USA) was used to plot thebar graphs indicating the scaled estimates for cookingparameters for representing direction of interactionand quadratic effects of the processing variables and

to compute the estimated ridges of maximum andminimum response for increasing radii from the centre

of the original design Minitab (2000) was used toanalyse the Pearson correlations between cookingparameters of patty samples

Results and discussionTable 2 shows the effects of added fat, wheat bran andNaCl levels on the cooking properties of the cookedpatties In addition, Figs 1–4 illustrate these effects asthree-dimensional graphs where direction of the effects

of the processing variables on cooking properties could

be seen The second-order regression model equationspredicting effects of processing variables are alsoincluded in the figures

Moisture retentionTable 2 indicates that the linear effects of fat and wheatbran were significant (P < 0.01 and 0.05, respectively).Figure 1(a, b) indicates that increasing fat leveldecreased (P < 0.01) moisture retention Similar results

Table 2 Significance of the regression models (F-values) and the effects of processing variables on cooking properties of patties

Source of

variance DF

Moisture retention

Fat retention

Water-holding capacity Linear

Figure 2 Effect of (a) subcutaneous fat and

wheat bran, (b) subcutaneous fat and NaCl,

(c) wheat bran and NaCl on reduction in

thickness; (d) subcutaneous fat and wheat

bran, (e) subcutaneous fat and NaCl,

(f) wheat bran and NaCl on reduction in

diameter along with the second-order

poly-nomial model equations predicting effects of

the variables.

were found by Serdaroglu & Degirmencioglu (2004)

and Serdaroglu (2006) in cooked meatballs, who

determined that decreasing fat in meatball formulations

resulted in higher moisture retention On the other

hand, wheat bran addition increased (P < 0.01) the

moisture retention values (Fig 1a, c) Yilmaz (2005)

reported that the lowest weight loss, a partial indicator

of moisture loss during cooking, was obtained from

20% (the highest amount tested) wheat bran-added

meatball sample No significant linear effect of NaCl

on the moisture retention values of the patty samples

was observed (Table 2) Although fat alone had a

decreasing effect on the moisture retention (Fig 1a, b),

an inverse effect was observed by the

interac-tion between fat and wheat bran, which increased

the moisture retention (Table 2, Fig 5) However, the

interaction of wheat bran with NaCl decreased the

moisture retention (Fig 5), which could be because of

the increasing salt levels in its interaction effect At

higher salt concentration, the protein denaturised,

unfolded and exposure of hydrophobic areas in the

proteins increased, leading to aggregation and loss of

water from the muscle (Thorarinsdottir et al., 2004)

The levels of independent variables that minimise and

maximise the moisture retention were determined by

ridge analysis Results from ridge analysis, which

computes the estimated ridge of optimum response

for increasing radii from the centre of the originaldesign (Chwen-Jen et al., 1996), indicated that mini-mum moisture retention (39.94%) would be at fat =30.00%, wheat bran = 5.00% and NaCl = 0.77%

w⁄ w at the distance of coded radius 1.0 Maximummoisture retention (51.10%) would occur at fat =10.00%, wheat bran = 15.00% and NaCl = 0.20%

w⁄ w at the distance of coded radius 1.0 These resultsindicate that maximum level of wheat bran butminimum levels of fat and moderate levels of saltshould be used in the patty formulations to obtainmaximum moisture retention

Fat retentionSimilar results were also determined for fat retention.Namely, the fat retention of patties decreased withincreases in fat level (P < 0.01) (Fig 1d, e), which wasconsistent with the literature Serdaroglu & Degirmen-cioglu (2004) determined that fat retention increasedwith decreased fat level in meatball formulation, andMansour & Khalil (1999) observed that low-fat pattiesretained more fat than high-fat samples during cooking

In addition, Tornberg et al (1989) concluded that denseprotein matrix of low-fat ground beef prevented fatmigration On the other hand, fat retention of pattiesincreased (P < 0.05) with wheat bran level (Fig 1d, f)

Water-holding capacity

(c)

Figure 4 Effect of (a) subcutaneous fat and

wheat bran, (b) subcutaneous fat and

NaCl, (c) wheat bran and NaCl on

water-holding capacity along with the second-order

polynomial model equations predicting

effects of the variables.

Fat retention is a complex phenomenon and probably

the result of several chemical and physical mechanisms

Proteins are considered to be excellent fat binders

because they have dual functionality in respect of fatinteractions in which non-polar side chains of proteinsfurnish sites for lipid–protein interactions and interfacial

Figure 5 Scaled estimates for cooking parameters showing the direction of interaction and quadratic effects of the processing variables;

X 1 = subcutaneous fat, X 2 = wheat bran, X 3 = NaCl Positive- and negative-scaled estimates values indicate the direction of the effects, increasing and decreasing, respectively.

film formation In addition, the three-dimensional

matrix formed by myofibrillar protein gelation

physi-cally entraps fat (Zayas, 1997; Anderson & Berry, 2001)

However, it is known that wheat bran had a

consider-able source of dietary fibre, as reported earlier (42.5%

dietary fibre content) Accordingly, Sosulski & Cadden

(1982) determined that fibres rich in hemicelluloses and

lignin such as wheat bran possess some fat-holding

properties Also, Ang (1991) reported that what bran

fibre had a considerable ability to hold fat because of

having longer length cellulose fibres that could retain

more oil than shorter length cellulose fibres Even

though fat alone had a decreasing effect on the fat

retention, two-factor interaction effects showed that fat

retention was significantly (P < 0.01) increased by the

interaction of fat with wheat bran (Fig 5) Of the

quadratic effects, only the quadratic effect of wheat bran

on the fat retention was found significant (P < 0.01)

(Table 2), which indicated that further increase in the

wheat bran level also increased the fat retention of the

patty samples (Fig 5) The ridge analysis indicated that

minimum fat retention (36.64%) would be at fat =

30.0%, wheat bran = 5.00% and NaCl = 0.00% w⁄ w

at the distance of coded radius 1.0 Maximum fat

retention (83.61%) would occur at fat = 30.00%, wheat

bran = 15.00% and NaCl = 0.00% w⁄ w at the

dis-tance of coded radius 1.0 From these optimisation

results, wheat bran may, at first, appear to be the only

processing variable influencing fat retention of patty

samples; however, fat alone had a decreasing effect on

the fat retention as mentioned earlier This was thought

to be because of the prevalent effect of wheat bran to

bind fat even up to the 30% level of fat incorporation,

suppressing the effect of fat to decrease fat retention

Briefly, it could be concluded that the effect of fat was

strongly dependant on bran level It could also be

concluded that wheat bran affected the fat retention

regardless of the added NaCl level as well, which could

be attributed to the effect of salt more on moisture

retention than on fat retention Accordingly, the effects

of salt level were mostly documented on the moisture

retention and WHC of the meat products in the

literature (Fennema, 1990; Akse et al., 1993)

Reduction in thickness and diameterFat had significant effect on reduction in thickness(P < 0.05) and reduction diameter (P < 0.01), andwheat bran had significant (P < 0.01) effect on thereduction in diameter parameter of cooked patties(Table 2) Figure 2(a, b, c, d, e and f) showed thedirection of these effects clearly Fat decreased reduc-tion in patty thickness; however, it increased thereduction in patty diameter, which was consistent withthe findings of Serdaroglu & Degirmencioglu (2004)who determined that meatballs formulated with 20%fat had the higher reduction in diameter On the otherhand, wheat bran decreased the reduction in diameter.This could be expected because of the fact that whatbran fibre had ability to hold fat and water, whichcontributed to a decrease in the reduction in diameter

of the patty samples in this study The quadratic effect(P < 0.01) of wheat bran on the reduction in thick-ness was negative (Table 2), which indicated that theeffect of wheat bran to decrease the reduction inthickness (Fig 5) could be observed with the higherwheat bran levels Namely, reduction in thicknessincreased up to 10.28% level of wheat bran; however,

it decreased after this level as the wheat bran levelfurther increased (Fig 2a, c) Minimum reduction inthickness (20.42%) determined by ridge analysis wouldoccur at fat = 20.75%, wheat bran = 15.00% andNaCl % 2.00 w⁄ w Minimum reduction in diameter(7.52) would be at fat = 10.0%, wheat bran =15.00% and NaCl = 0.00% Based on these optimi-sation results, maximum level of wheat bran incorpo-ration into patties could be recommended to achievedesired patty appearance irrespective of fat and saltlevels

Cooking yieldEffect of processing variables on the cooking yield ofpatties is illustrated in Fig 3(a, b and c) Fat was thevariable, which had a linear effect on cooking yield(Table 2, P < 0.05) Fat caused a decrease in cookingyield of patties (Fig 3d, e) It was reported that the

Table 3 Correlation coefficients of the cooking parameters

Cooking

parameters

Moisture retention

Fat retention

Reduction in thickness 0.687 **

0.209 Reduction in diameter )0.835 ** )0.675 ** )0.619 *

mean free distance between fat droplets decreases as the

fat content increases, which causes fat coalescing and

then leaking from the product (Tornberg et al., 1989)

Therefore, this might be attributed to the excessive fat

separation and water release in the higher fat including

meatballs during cooking The addition of wheat bran,

however, increased (P < 0.01) the cooking yield of the

patty samples (Fig 3a, c), which could be because of

the ability of wheat bran to keep the moisture and fat in

the matrix It could be hypothesised that the mechanism

responsible for moisture and fat retention might be

prevalently related with the swelling of fibre in the wheat

bran By swelling, the fibre could have absorbed some

fat and interacted with the protein of ground beef to

form a matrix, which acts to prevent the coalescence and

migration of fat out of the product (Anderson & Berry,

2001) The effect of interaction between fat and wheat

bran increased (P < 0.05) the cooking yield of the

patties (Fig 5), although fat alone decreased this

cooking parameter This could be explained by the fact

that wheat bran has high fat retention ability (Ang,

1991) as seen by the interaction effect that inversed the

linear effect of fat to decrease cooking yield Namely, the

more wheat bran retained fat, the less fat loss occurred,

increasing cooking yield Furthermore, the parallel

result was obtained for the fat retention parameter as

affected by the aforementioned linear and interaction

effects of fat with wheat bran Results from ridge

analysis indicated that maximum cooking yield

(95.11%) would be at fat = 10%, wheat bran =

15.00% and NaCl = 0.00% w⁄ w Therefore, it could

be concluded that wheat bran could be better successful

to get maximum product yield at low fat levels

Shrinkage

Table 2 indicates that fat and wheat bran were the

processing variables having significant (P < 0.05; 0.01,

respectively) effects on the shrinkage Figure 3(d, e)

shows that fat increased (P < 0.05) the shrinkage of

patties During cooking process, patty samples shrunk

because of denaturation of meat proteins by heat; as a

consequence, loss of water and fat contributed to the

shrinkage process Similarly, Serdaroglu &

Degirmen-cioglu (2004) found that fat level affected meatball

shrinkage, reducing the fat level from 20% to 5%

decreased shrinkage Figure 3(d, f) shows that wheat

bran decreased shrinkage of patties Shrinkage is a

process in which some loss of water and fat occur

However, the effect of wheat bran to decrease shrinkage

could be expected because of the ability of wheat bran to

hold water and fat, as mentioned earlier According to

ridge analysis, minimum shrinkage (10.40%) would

occur at fat = 10.0%, wheat bran = 15.00% and

NaCl = 0.00% w⁄ w at the distance of coded radius

1.0 Shrinkage is one of the undesired appearance

defects resulting from cooking process; therefore, meatindustry aims at decreasing shrinkage of the product asmuch as possible By taking into account these optimi-sation results, however, minimum shrinkage could beachieved by maximum wheat bran, minimum fat andsalt addition into the product

Water-holding capacityWheat bran was the only processing variable showing asignificant (P < 0.01) linear effect on WHC of patties.Figure 4(a, b and c)shows the effects of processingvariables Wheat bran increased the WHC of patties(Fig 4a, c) Similar explanations could be given forincreasing effect of wheat bran to increase WHC ofpatties; namely, this result could be because of theability of wheat bran to keep the moisture in the matrix.Figure 4(b, c) illustrates that NaCl initially increased theWHC up to the 1.15% of NaCl concentration; however,

it decreased the WHC at further NaCl concentrations,which could also be seen by its quadratic effect in Fig 5.This result was consistent with the information given inthe literature Fennema (1990) suggested that at arelatively low salt concentration, salt anions bind tothe filaments and thereby increase repulsive forcesbetween the filaments Salt was also believed to lowerthe structural constraints to swelling At higher saltconcentration, the protein denaturised, unfolded andexposure of hydrophobic areas in the proteins increased,leading to aggregation and loss of water from themuscle Cross-linking between proteins and shrinkage ofthe muscle might have resulted in less space for waterand therefore decreased WHC (Thorarinsdottir et al.,2004) Based on these reports, the protein in the pattysamples could be assumed to swell more and have abetter WHC if processed at an initially lower saltconcentration This might indicate that the distancebetween the myofibrils is too long for a protein–proteinbinding to occur; resulting in less shrinkage of themuscle (Akse et al., 1993) Ridge analysis indicated thatmaximum WHC (82%) would be at fat = 10%, wheatbran = 15.00% and NaCl = 1.15% w⁄ w at the dis-tance of coded radius 1.0 These results indicate thatmaximum wheat bran addition could achieve the bestWHC values up to certain levels of salt (1.15%) Based

on these results, the muscle should be assumed to swellmore and have a better WHC if processed at an initiallylower salt concentration (Thorarinsdottir et al., 2004)

Correlations between cooking parametersThe results of correlation analysis showed significantcorrelations between cooking parameters with theexception of that between WHC and the others(Table 3).The correlations between the results obtainedare important with respect to confirmation of the

results; accordingly, the trends between the cooking

parameters in this study were confirmed by the

corre-lation estimates For example, positive correcorre-lations

between moisture retention, fat retention, cooking yield

and negative correlations of moisture retention with

reduction in diameter and shrinkage confirmed the

results of this study, as expected Namely, a product

with good moisture and fat retention could be expected

to show high cooking yield and low reduction in

diameter and shrinkage The same interpretations could

be done for the remaining correlations Furthermore,

similar correlations were determined by Serdaroglu

et al (2005) who found that a significant correlation

between cooking yield and fat retention and moisture

retention and negative correlation between cooking

yield and diameter changes for meatballs

Conclusions

Wheat bran showed very good performance for each

cooking parameter examined in study Therefore, wheat

bran could be a promising ingredient not only for the

development of high-fat specialty foods that are beef

based and required to improve the cooking properties of

patties at high-temperature processing but also for

obtaining more healthy products because of its high

fibre content In this study, the levels of the processing

variables (fat, wheat bran and NaCl) that maximise and

minimise the cooking parameters as requested by meat

industry were calculated and regarded as the optimum

levels Therefore, the results of this optimisation study

would be useful for meat industry that tends to increase

the product yield for patties using the optimum levels of

fat, wheat bran and salt determined by the ridge analysis

Acknowledgments

Authors thank the Scientific Research Projects

(SU-BAP-Konya, Turkey) of Selcuk University

Coor-dinating Office for financial support

References

AACC (1998) American Association of Cereal Chemists (AACC)

1998 Approved Methods, 9th edn, Vol 2 St Paul: AACC Publishers.

Akse, L., Gundersen, B., Lauritzen, K., Ofstad, R & Solberg, T.

(1993) Saltfisk: Saltmodning, Utproving av Analysemetoder,

Misfar-get Saltfisk (In Norweigan) Pp 1–61 Tromso¨: Fiskeriforskning.

Anderson, E.T & Berry, B.W (2001) Effects of inner pea fiber on fat

retention and cooking yield in high fat ground beef Food Research

International, 34, 689–694.

Ang, J.F (1991) Water retention capacity and viscosity effect of

powdered cellulose Journal of Food Science, 56, 1682–1684.

AOAC (2000) Association of Official Analytical Chemists (AOAC)

2000 Official Methods of Analysis, 17th edn, Vol 2 AOAC

International, Gaithersburg.

Box, G.E.P & Behnken, D.W (1960) Some new three level designs for

the study of quantitative variables Technometrics, 7, 455–475.

Chwen-Jen, S., Koehler, P.E & Akoh, C.C (1996) Optimization of sucrose polyester synthesis using response surface methodology Journal of Food Science, 61, 97–100.

Claus, J.R & Hunt, M.C (1991) Low-fat, high added water bologna formulated with texture-modifying ingredients Journal of Food Science, 56, 937–940.

Crehan, C.M., Hughes, E., Troy, D.J & Buckley, D.J (2000) Effects

of fat level and maltodextrin on the functional properties of frankfurters formulated with 5, 12 and 30% fat Meat Science, 55, 463–469.

Dzudie, T., Kouebou, C.P., Essia-Ngang, J.J & Mbofung, C.M.F (2004) Lipid sources and essential oils effects on quality and stability of beef patties Journal of Food Engineering, 65, 67–72.

El-Magoli, S.B., Laroia, S & Hansen, P.T.M (1996) Flavour and texture characteristics of low fat ground beef patties formulated with whey protein concentrate Meat Science, 42, 179–193.

Fennema, O.R (1990) Comparative water holding properties of various muscle foods A critical review relating to definitions, methods of measurement, governing factors, comparative data and mechanistic matters Journal of Muscle Foods, 1, 363–381.

Girard, J.P., Culioli, J., Maillard, T., Denoyer, C & Touraille, C (1990) Influence of technological parameters on the structure of the batter and the texture of frankfurter type sausages Meat Science, 27, 13–28 Gregg, L.L., Claus, J.R., Hackney, C.R & Marriott, N.G (1993) Low-fat, high added water bologna from massaged minced batter Journal of Food Science, 58, 259–264.

Huang, S.C., Shiau, C.Y., Liu, T.E., Chu, C.L & Hwang, D.F (2005) Effects of rice bran on sensory and physico-chemical properties of emulsified pork meatballs Meat Science, 70, 613–619.

Hunter, J.S (1959) Determination of optimum condition by imental methods Industrial Quality Control, 15, 6–15.

exper-Keeton, J.T (1994) Low-fat meat products-technological problems with processing Meat Science, 36, 261–276.

Lyons, P.H., Kerry, J.F., Morrissey, P.A & Buckley, D.J (1999) The influence of added whey protein ⁄ carrageenan gels and tapioca starch

on the textural properties of low fat pork sausages Meat Science,

51, 43–52.

Mansour, E.H & Khalil, A.H (1999) Characteristics of low fat beef burgers as influenced by various types of wheat fibers Journal of the Science of Food and Agriculture, 79, 493–498.

Minitab (2000) Computer Program, MINITAB Releases 13.0 for Windows USA: Minitab Inc.

Murphy, E.W., Criner, P.E & Grey, B.C (1975) Comparison of methods for calculating retentions of nutrients in cooked foods Journal of Agricultural and Food Chemistry, 23, 1153–1157.

Ockerman, H.W (1985) Quality Control of Post-mortem Muscle Tissue, 13th edn Columbus, OH, USA: The Ohio State Univ.

Ruusunen, M., Vainionpa¨a¨, J., Puolanne, E et al (2003) Physical and sensory properties of low-salt phosphate-free frankfurters composed with various ingredients Meat Science, 63, 9–16.

Ruusunen, M., Vainionpa¨a¨, J., Lyly, M et al (2005) Reducing the sodium content in meat products: the effect of the formulation in low-sodium ground meat patties Meat Science, 69, 53–60.

Saricoban, C., Yılmaz, M.T & Karakaya, M (2009) Response surface methodology study on the optimisation of effects of fat, wheat bran and salt on chemical, textural and sensory properties of patties Meat Science, 83, 610–619.

Serdaroglu, M (2006) Improving low fat meatball characteristics by adding whey powder Meat Science, 72, 155–163.

Serdaroglu, M & Degirmencioglu, O (2004) Effect of fat level (5%, 10%, 20%) and corn flour (0%, 2%, 4%) on some properties of Turkish type meatballs (koefte) Meat Science, 68, 291–296.

Serdaroglu, M., Turp, G.Y & Abrodı´mov, K (2005) Quality of fat meatballs containing Legume flours as extenders Meat Science,

low-70, 99–105.

Sosulski, F.W & Cadden, A.M (1982) Composition and ical properties of several sources of dietary fiber Journal of Food Science, 47, 1472–1477.

physiolog-Thompson, D.R (1982) Response surface experimentation Journal of

Food Processing and Preservation, 6, 155–162.

Thorarinsdottir, K.A., Arason, S., Bogason, A.G & Kristbergsson, K.

(2004) The effects of various salt concentrations during brine curing

of cod (Gadus morhua) International Journal of Food Science &

Technology, 39, 79–89.

Tornberg, E., Olsson, A & Person, K (1989) A comparison in fat

holding between hamburgers and emulsion sausages in: Proceedings

of the 35th international congress on meat science and technology (pp.

753–757), 24–26 September 1989, Copenhagen, Denmark.

Vendrell-Pascuas, S., Castellote-Bargallo, A.I & Lopez-Sabater, M.

(2000) Determination of inulin in meat products by high

performance liquid chromatography with refractive index detection Journal of Chromatography A, 881, 591–597.

Yilmaz, I (2004) Effects of rye bran addition on fatty acid composition and quality characteristics of low-fat meatballs Meat Science, 67, 245–249.

Yilmaz, I (2005) Physicochemical and sensory characteristics of low fat meatballs with added wheat bran Journal of Food Engineering,

69, 369–373.

Zayas, J.F (1997) Functionality of Proteins in Foods New York: Springer.

Original article

Antioxidative and reducing capacity, macroelements content and

sensorial properties of buckwheat-enhanced gluten-free bread

Małgorzata Wronkowska,1* Danuta Zielin´ska,2Dorota Szawara-Nowak,1Agnieszka Troszyn´ska1&

(Received 22 April 2010; Accepted in revised form 21 June 2010)

component of a gluten-free bread formula, to make buckwheat-enhanced gluten-free breads The 40% enhanced gluten-free bread showed the highest antioxidant capacity against ABTS+·and DPPH·radicals(4.1 and 2.5 lmol Trolox g)1 DM, respectively) and reducing capacity measured by cyclic voltammetry(1.5 lmol Trolox g)1 DM) The antioxidant and reducing capacity of buckwheat-enhanced gluten-freebreads were positively correlated with their total phenolic contents (r = 0.97) The 40% BF-enhancedgluten-free bread showed the highest overall sensory quality (7.1 units) when compared to control gluten-freebread (1.8 units) The linear relationship between applied increasing BF doses in gluten-free bread formulaand magnesium, phosphorus and potassium content in breads was noted It was concluded that 40% BF-enhanced gluten-free bread could be developed and dedicated to those people suffering from coeliac disease

Introduction

Buckwheat is an important crop in some areas of the

world and at present is considered as food component of

high nutritional value (Li & Zhang, 2001) Buckwheat

has attracted increasing attention from food scientists

for its healing effects over chronic diseases, and it is of

increasing popularity in many countries as a healthy

food (Li & Zhang, 2001; Gabrovska et al., 2002; Lin

et al., 2009) Buckwheat is a rich source of proteins with

a low content of a-gliadin fraction (Kreft et al., 1996), it

has balanced amino acid composition Buckwheat

proteins are rich in lysine and arginine (Watanabe,

1998) Besides its high-quality proteins, buckwheat is

rich in many rare components with healing effects to

some chronic diseases Among them, the most

attrac-tive ones are phenolic acids, flavonoids, phytic acid,

vitamin B1, B2and E, glutathione, D-chiro-inositol and

fagopyritols, carotenoids, phytosterols, melatonin, and

thiamin-binding proteins (Li & Zhang, 2001; Wijngaard

& Arendt, 2006; Christa & Soral-S´mietana, 2008;

Zielin´ska & Zielin´ski, 2009) Buckwheat grain is milled

into flour or dehulled to produce groats Two types ofmilling are used to produce buckwheat flour (BF) One

is similar to wheat milling, in which the grain is milledinto flour, and second type of milling involves millingthe dehulled buckwheat groats At present, BF isconsidered as an interesting ingredient for the gluten-free formulations, which can improve nutritional andfunctional properties of gluten-free products desired forpeople suffered from coeliac disease

Coeliac disease is a lifelong inflammatory condition ofthe gastrointestinal tract that affects the small intestine

in genetically susceptible individuals The inflammationclassically produces a malabsorption syndrome, withdiarrhoea, steatorrhoea, and loss of weight or failure tothrive (Kennedy & Feighery, 2000) Also commonsymptoms are deficiencies of iron, calcium, folic acidand fat-soluble vitamins D, E, A (Murray, 1999).Serologic screening studies have shown the worldwideprevalence in human to be 1 in 266 (Fasano & Catassi,2001) Environmental, genetic and immunological fac-tors, some of which are still unknown, are importantplayers in the pathogenesis of coeliac disease The mainenvironmental factors triggering coeliac disease arespecific peptides that are present exclusively in dietarygluten proteins from wheat, rye and barley These

*Correspondent: Fax: (48 89) 5240124;

e-mail: m.wronkowska@pan.olsztyn.pl

peptides are particularly rich in proline, a structure that

confers resistance to digestion by gastric, pancreatic and

brush border enzyme The mainstay of treatment is a

strict lifelong adherence to a gluten-free diet, in which

the patient must carefully avoid all kinds of food

products containing wheat, rye and barley (Fasano &

Catassi, 2001) The diet based on gluten-free products is

characterised by a low content of some nutritional

components such as proteins and mineral components,

as well as nonnutritional but physiologically important

component such as dietary fibre The global poor

quality of gluten-free products has lead technologists

to investigate new ingredients and formulations to

obtain gluten-free products (Gallagher et al., 2004) In

recent years, some studies have been done, mainly

involving the approach of incorporation of starches,

dairy proteins and hydrocolloids into a gluten-free flour

base (rice, buckwheat and soy) (Ribotta et al., 2004;

Sivaramakrishnana et al., 2004; Lazaridou et al., 2007)

Torbica et al (2010) studied the rheological, textural

and sensory properties of gluten-free bread formulations

based on rice and BF They found that the increase in

the amount of BF resulted in increase in dough

development time, decrease in the starch retrogradation

degree and the breads were acceptable according to

results of the sensory analysis

In this study, flour originated from common

buck-wheat (Fagopyrum esculentum Moench) (BF) was used

to substitute 10%, 20%, 30% and 40% corn starch,

the main component of gluten-free bread formula, to

make buckwheat-enhanced gluten-free breads The

first aim of this study was to develop a new formula

for buckwheat-enhanced gluten-free breads The

sec-ond aim was to select BF-enhanced gluten-free bread

of high sensory and antioxidative properties The

antioxidant and reducing capacity, sensorial properties

and macroelements content in buckwheat-enhanced

gluten-free breads have been studied to achieve these

objectives

Materials and methods

Chemicals

Ferulic acid,

2,2¢-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) diammonium salt (ABTS+·),

2,2-diphe-nyl-1-picrylhydrazyl (DPPH+) and

6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) were

purchased from Sigma (Sigma Chemical Co., St Louis,

MO, USA) Other reagents of reagent-grade quality were

from POCh, Gliwice, Poland

Materials

Ingredients used for gluten-free bread formulations were

as follows: corn and potato starch (Cargill Sp z o.o.,

Poland), pectin (Pectowin SA, Poland), sunflower oil,white sugar and baker¢s yeast (local market, Olsztyn,Poland), and tap water Buckwheat flour was producedfrom commercially available dehulled buckwheat grains(groat) after milling in laboratory conditions usinglaboratory mill W _Z-1 (ZBPP, Bydgoszcz, Poland) thenwere sieved through a 60-mesh screen

Bread-making processThe gluten-free formulation was prepared according tothe procedure described in Polish patent specification(Wronkowska et al., 2008a) The BF substituted 10%,20%, 30%, 40% w⁄ w of gluten-free formulas basis, cornstarch was exchanged for BF The amount of addedwater was 80 g for 100 g of gluten-free formulation Thesunflower oil (2%), fresh yeast (8%), salt (1%) andsugar (5%) were also added The mixture was blendedwith a planetary rotation of mixing within 5-speed mixer(Kitchen Aid, St Joseph, MI, USA) for 12 min Thedough was proofed at 35–40C for 40 min and baked at

215C for 25–35 min The baking tests were carried out

in an electric oven with an incorporated proofingchamber (ZBPP, Bydgoszcz, Poland) The bread sampleswere freeze-dried (dry matter of samples was about 95 gper 100 g of sample), ground and sieved through a60-mesh screen to obtain powdered bread The followingsample abbreviation was used: control (100% formula);10% BF (10% BF and 90% gluten-free formula); 20%

BF (20% BF and 80% gluten-free formula); 30% BF(30% BF and 70% gluten-free formula); 40% BF (40%

BF and 60% gluten-free formula)

Preparation of gluten-free bread crude extractsThe powdered gluten-free breads were extracted intriplicate with 80% aqueous methanol (1⁄ 10; w ⁄ v) for

2 h of shaking at 37C Samples were then centrifuged

at 2600 g at 4C for 15 min in a Beckman GS-15Rcentrifuge (Beckman Instruments, Inc., Palo Alto, CL,USA) The fresh 80% methanolic extracts were used forthe determination of their ability to scavenge ABTS+·and DPPH· radicals, reducing capacity by cyclic vol-tammetry (CV) method and content of total phenoliccompounds (TPC)

Determination of total phenolic compoundsThe content of TPC was determined according toShahidi & Naczk (1995) Exactly 0.25 mL of eachextract was mixed with 0.25 mL of Folin-Ciocalteureagent (previously diluted with water 1:1 v⁄ v) and0.5 mL of saturated sodium carbonate (Na2CO3)solution and 4 mL water The mixture was allowed

to stand at a room temperature for 25 min andthen was centrifuged at 2000 g for 10 min at room

temperature Absorbance of the sample was measured

at 725 nm using a spectrophotometer (UV-160 1PC,

Shimadzu, Kyoto, Japan) The data were expressed as

mg of ferulic acid equivalents (FAE) per g of bread dry

matter (DM)

Measurement of the antioxidant capacity of

buckwheat-enhanced gluten-free breads

The antioxidant capacity of gluten-free breads was

measured against stable, nonbiological radicals such as

2,2¢-azinobis-(3-ethylbenzothiazoline-6-sulphonate)

rad-ical cation (ABTS+·) and 2,2-diphenyl-1-picrylhydrazyl

radical (DPPH·) using a spectrophotometric assay

The ABTS+· scavenging assay, based on the

reduc-tion of the ABTS+· radical cation by antioxidants

present in 80% methanolic extracts, was applied The

ABTS+· radical cation was prepared by mixing

ABTS+·stock solution (7 mm in water) with 2.45 mm

potassium persulphate This mixture had to remain for

12–24 h until the reaction was complete, and the

absorbance was stable Antioxidant capacity was

deter-mined following the procedure described by Re et al

(1999) with a minor modification For measurements,

the ABTS+·solution was diluted with 80% methanol to

the absorbance of 0.700 ± 0.020 at 734 nm For the

photometric assay, 1.48 mL of the ABTS+· solution

and 20 lL of the extracts or Trolox standards were

mixed and measured immediately at 30C after 6 min at

734 nm using a spectrophotometer (UV-160 1PC;

Shi-madzu) Appropriate solvent blanks were run in each

assay The antioxidant capacity was calculated on the

basis of percentage inhibition of absorbance at 734 nm

using Trolox standard curve and was expressed as lmol

Trolox g)1of bread DM

The DPPH·radicals scavenging assay was based on a

modified method of Brand-Williams et al (1995) In this

assay, antioxidants present in 80% methanol extracts

reduce the free radicals 2,2-diphenyl-1-picrylhydrazyl,

which have an absorption maximum at 515 nm The

DPPH· radical solution was prepared by dissolving

10 mg DPPH· in 25 mL of 80% methanol First, the

extinction of the disposable cuvette with 250 lL of the

methanol DPPH·solution and 2.1 mL of 80% methanol

was measured as blank Then, the 80% methanol bread

extracts (100 lL) were added to 250 lL of the

methan-olic DPPH· solution and 2 mL of 80% methanol The

mixture was shaken vigorously and allowed to stand at

room temperature in the dark for 20 min The decrease

in absorbance of the resulting solution was monitored at

517 nm for 20 min using a spectrophotometer (UV-160

1PC, Shimadzu) The Trolox standard solutions

(con-centration 0.1–2.0 mm) in 80% methanol were prepared

and assayed under the same conditions Antioxidant

capacity against DPPH·radicals was expressed as lmol

Trolox g)1of bread DM

Measurement of the reducing capacity of enhanced gluten-free breads by cyclic voltammetry methodThe cyclic voltammetry experiments were performed in80% methanol extracts mixed with 0.2 m sodiumacetate–acetic buffer (pH 4.5) at ratio 1:1 (v⁄ v) accord-ing to Zielin´ska et al (2007) The sodium acetate–aceticbuffer acted also as a supporting electrolyte for CVmeasurements During the course of this experiment, amicroelectrochemical cell (with total volume of 200 lL)made of Teflon was used The cell comprised threeelectrodes: a glassy carbon (GC) working electrode(BAS MF-2012, 3 mm diameter), an Ag⁄ AgCl (3.5 mKCl) reference and a Pt (0.5 mm diameter coiled Ptwire) counter electrode Prior to each CV experiment,

buckwheat-GC working electrode was carefully hand-polished with0.05- lm alumina–water paste (BAS CF-1050), usingBAS (MF-1040) polishing cloth Then, the electrode wasrinsed with ultra-pure water and finally with methanol.The CV experiment was performed in the range of100–1100 mV at a potential sweep-rate of 100 mV s)1atroom temperature using a potentiostat⁄ galvanostat

G 750 (Gamry Ins., Warminter, PA, USA) For the testpurpose, the total charge below the anodic wave curve

of the voltammogram was calculated by instrumentsoftware The cyclic voltammograms of Trolox solutionsover the concentration range of 0.05–2.5 mm was alsoacquired, and then the reducing capacity of gluten-freebreads was expressed as lmol Trolox g)1of bread DM

Determination of macroelementsThe measurement of macroelements content in buck-wheat-enhanced gluten-free breads was carried out usingthe atomic absorption spectroscopy (AAS) method by aUnicam 939 spectrometer equipped with data baseADAX, background correction and cathode lamps(Soral-S´mietana et al., 2001) Before macroelementsdetermination, each bread sample was wet mineralisedwith a mixture of nitric and perchloric acids (3:1; v⁄ v).Potassium was determined with the photometric flamemethod, while phosphorus with the colorimetric method

by molybdate with hydroquinonate and sodium phate (IV) For the validation of calcium measurement,the solution of lanthanum chloride was added to allsamples in the amounts assuring 0.5% concentration

(seven females and one male, ranging in age between 26–

39 years) previously selected and trained according to

ISO guidelines (1993) Prior to their participation in the

experiments, the subjects were trained on sensory

descriptors for the commercial wheat breads purchased

from a local supermarket After that, vocabularies of the

sensory attributes of buckwheat-enhanced gluten-free

bread were developed by the panel in a round-table

session, using a standardised procedure (ISO 1998a)

Twenty attributes related to the appearance, odour,

taste and texture of breads were selected and thoroughly

defined for profiling Definition and description of these

sensory attributes are summarised in Table 1 The

panellists evaluated the intensity perceived for each

sensory attribute on unstructured graphical scales The

scales were 10 cm long and verbally anchored at each

end, and the results were converted to numerical values

(from 0 to 10 units) by a computer Loaves were sliced

(15 mm thickness) on slicing machine and served to the

assessors in transparent plastic boxes The samples were

coded with a three-digit number and presented to

panellists in random order Mineral water was offered

between the samples The assessments were carried out

in sensory laboratory room, which fulfils the

require-ments of the ISO standards (1998b) The results were

collected using a computerised system ANALSENS

(IAR&FR PAS, Olsztyn, Poland) Each bread samplewas tested in two replications

Consumer test

A semi-consumer panel of 30 members (including staff,graduate, and undergraduate students of the Institute)has made hedonic evaluation of the samples In the test,each panellist was asked to assess the breads forpalatability, based on the overall colour, odour, tasteand texture An unstructured graphical scale was 10 cmlong and anchored on both ends: disliked (0) – extremelyliked (10)

Statistical analysisThe results of the chemical analyses are given as themeans and the standard deviation of three independentmeasurements The results were subjected to one-wayanalysis of variation anova using least significantdifference (LSD) test, at P < 0.05 The correlationanalysis was performed, and the Pearson correlationcoefficient was calculated

anova was used to test statistical differences insensory attributes between the breads Means werecompared using Fisher’s protected LSD (at P < 0.05)

Table 1 The attributes and definitions used for descriptive analysis of buckwheat-enhanced gluten-free breads

Attribute Definition

Appearance

1 Colour Visual impression of the brad colour (from light to dark)

2 Porosity Visual impression of the bread crumb porosity (poreless – porous)

Odour

3 Rancid Odour typical of rancid nut oil (none – very intensive)

4 Sweet Odour characteristic to bun produced from wheat flour (none – very intensive)

5 Yeast Odour characteristic to yeast-raised bread produced from wheat flour (none – very intensive)

6 Buckwheat Odour typical of boiled buckwheat (none – very intensive)

7 Acidulous The intensity of the acidulous odour (none – very intensive)

Taste

8 Sweet Basic taste illustrated by sucrose diluted in water 1.5% (none – very intensive)

9 Rancid Taste typical of rancid nut oil (none – very intensive)

10 Bitter Basic taste illustrated by caffeine diluted in water 0.5% (none – very intensive)

11 Buckwheat Taste characteristic to boiled buckwheat (none – very intensive)

12 Acidulous The intensity of the acidulous taste (none – very intensive)

13 Yeast Taste characteristic to yeast-raised bread produced from wheat flour (none – very intensive)

14 Aftertaste Aftertaste which continued after the removal of sample (none – very intensive)

Aftertaste

15 Springiness Degree of springiness in bread crumb by pressing with finger (not springy – springy)

16 Elasticity Response to stretching (not elastic – elastic)

Texture (mouth feel)

17 Mastication Degree of perceived resistance while chewing the sample ten times (not mastic – mastic)

18 Texture (mouth feel) Degree of adhesiveness perceived while chewing the sample ten times (low – high)

19 Gumminess Degree of gumminess perceived while chewing the sample ten times (not gummy – gummy)

20 Moistness Degree of moistness perceived while chewing the sample ten times (dry – moist)

Anchoring points: odour ⁄ taste, none – very intensive; Texture, low – high.

Principal component analysis (PCA) was performed to

describe the variance among the whole sensory data

obtained Statistical analysis was performed using

soft-ware package (v 8.0; StatSoft Inc., Tulsa, OK, USA)

Results and discussion

Antioxidant capacity and total phenolics of

buckwheat-enhanced gluten-free breads

The 80% methanol extracts of the gluten-free breads

and BF were examined for their Trolox equivalent

antioxidant capacity and total phenolic content (TPC)

(Table 2) BF was characterised by the very high value

of antioxidant capacity against ABTS+· radicals

(18.5 lmol Trolox g)1 DM), against DPPH· radicals

(15.7 lmol Trolox g)1DM) and TPC (5.1 mg FAE g)1

DM) The antioxidant capacity of buckwheat-enhanced

gluten-free breads showed dose-depend – BF substitute

effect The statistically significant increase (P < 0.05) in

antioxidant capacity and TPC was noted with the

increasing amount of BF The highest value of

antiox-idant capacity, determined against ABTS+·and DPPH·

radicals was noticed for gluten-free bread with 40% of

BF No statistically significant differences in antioxidant

capacity of gluten-free breads with 30%, and 40% of BF

was noticed when determined against DPPH· radicals

(Table 2) The similar trend was found for TPC in

buckwheat-enhanced gluten-free breads Lin et al

(2009) reported that buckwheat-enhanced wheat bread

contained more rutin and quercetin than typical wheat

bread They showed that addition of BF, especially

dehulled, into wheat bread greatly enhanced bread’s

antioxidant properties Moreover, Zielin´ski et al (2008)

found that rye bread based on dark flour rich in

phenolic compounds showed higher antioxidant

capac-ity than bread based on brown flour produced after

partially removing of the outer layer of grain

Reducing capacity of buckwheat-enhanced gluten-free

breads by CV assay

The reducing capacity of buckwheat-enhanced

gluten-free breads without specific determination of the

contribution of each individual reducing componentwas determined by cyclic voltammetric methods (CV)(Zielin´ska et al., 2008) The total charge below anodicwave of the background signal (solvent + supportingelectrolyte) was subtracted from the total chargeobtained for each Trolox concentration within the range

of 100–1100 mV, and then the standard curve wasconstructed for calculating the antioxidant capacity ofthe samples The cyclic voltammograms of 80% MeOHextracts from gluten-free breads were recorded as itshown on Fig 1 The observed anodic wave wasbroadened because of the response of several antioxi-dant with different oxidation potentials (Oomah &Mazza, 1996; Zielin´ska & Zielin´ski, 2009) The totalcharge below anodic wave recorded for BF was 41.9 lCand reducing capacity was 6.9 lmol Trolox per g offlour DM The highest reducing capacity of the buck-wheat-enhanced gluten-free breads was found for breadswith 30% and 40% of BF (Table 2) The reducingcapacity of gluten-free breads provided by CV assay waslower than that determined against ABTS+· andDPPH·radicals The results from CV experiments werepositively correlated with antioxidant capacity againstABTS+·and DPPH·radicals (r = 0.97) The TPC wasalso positively correlated with antioxidant capacity ofbuckwheat-enhanced gluten-free breads determined by

DM, respectively The positive correlation betweenmacroelements content and increasing amount of BF ingluten-free breads was observed in this study (Table 3).Special attention should be paid for beneficial increasing

of calcium level in the BF-enhanced breads Recentevidence of analysis of 3-day food records of women ongluten-free diets in United States suggested that about50% have inadequate intakes of calcium, iron and fibre(Thompson et al., 2005) Moreover, the content of

Table 2 The antioxidant and reducing capacity of buckwheat-enhanced gluten-free breads

Sample

TEAC

(lmol Trolox g)1DM)

DPPH RSA (lmol Trolox g)1DM)

Total phenolics content (TPC) (FAE g)1DM)

Total charge below anodic wave (lC)

Reducing capacity (lmol Trolox g)1DM) Control 1.07 ± 0.01d 0.03 ± 0.01d 0.25 ± 0.01e 0.76 ± 0.01c 0.21 ± 0.00c

10% BF 1.70 ± 0.06c 0.76 ± 0.09c 0.42 ± 0.02d 1.71 ± 0.42bc 0.36 ± 0.07bc

20% BF 2.02 ± 0.30c 1.34 ± 0.14b 0.62 ± 0.01c 2.97 ± 0.52b 0.58 ± 0.09b

30% BF 3.57 ± 0.06b 2.34 ± 0.21a 0.93 ± 0.04b 7.75 ± 0.11a 1.38 ± 0.02a

40% BF 4.12 ± 0.06a 2.56 ± 0.08a 1.22 ± 0.04a 8.64 ± 1.24a 1.52 ± 0.21a

Data expressed as mean ± standard deviation (n = 3) Means in each column followed by the different letter are significantly different (P < 0.05).

magnesium noted in buckwheat-enhanced gluten-free

bread has been also of interest because of its regulatory

function in distribution of calcium ions in the cells

Magnesium ions are natural blockers of the calcium

channels; in human cells, a strong competition between

calcium and magnesium ions exists (Hordyjewska &

Pasternak, 2004) The addition of BF into the

experi-mental gluten-free dough affected the improvement of

technological quality of bread In comparison with the

control gluten-free bread (2.3 cm3g)1), the volume of all

breads supplemented with BF was considerably higher:

2.9 cm3g)1 for 10% BF; 3.2 cm3g)1 for 20% BF;

3.3 cm3g)1for 30% BF and for 40% BF – 3.1 cm3g)1

In our earlier study (Wronkowska et al., 2008b), the

loaf-specific volume of breads decreased with increasing

amounts of BF used as a ingredient of commercial

gluten-free formulation Common buckwheat that stitute 15% of wheat flour in buckwheat-enhanced wheatbreads did not interfere with bread-specific volume aspresented Lin et al (2009)

sub-Sensory evaluation of buckwheat-enhanced gluten-freebreads

The effect of the addition of BF for gluten-free formula

on the overall quality of breads is shown in Table 4 Theresults indicated that palatability of all investigatedbreads with BF was significant higher than those of thecontrol gluten-free bread The average overall quality ofscores for BF breads ranged from 2.7 units (10% BF) to7.1 units (40% BF), whereas the control obtained1.8 units (in the scale of 10 units) It indicated that BFmight contribute to improve the sensory properties ofgluten-free bread To find attributes that influenced thesensory quality of breads QDA was used in the study.QDA is the most sophisticated method in sensoryevaluation and involves the discrimination and descrip-tion of both the qualitative and quantitative sensoryattributes of a product by trained panels The meansensory ratings for the samples and the analysis ofvariance are presented in Table 4 The results show thatamong twenty attributes, twelve of them were statistically

Figure 1 Cyclic voltammograms of buckwheat-enhanced gluten-free

Data expressed as mean ± standard deviation (n = 3) Means in each

column followed by the different letter are significantly different

(P < 0.05).

Table 4 The sensory attributes of buckwheat-enhanced gluten-free breads Mean descriptive analysis ratings was performed on 0–10 unstructured intensity scale (n = 8, two replicates) a

Attributes b

Control 10% BF 20% BF 30% BF 40% BF Overall quality 1.8a 2.7b 3.9c 5.9d 7.1e Colour 1.7a 0.6a 6.1b 6.3b 7.7c Porosity 6.3c 0.9a 4.1b 2.8b 4.2b O-rancid 7.7c 5.1b 4.0b 2.0a 2.0a O-sweet 2.1a 3.1a 3.1a 2.5a 2.6a O-yeast 0.6a 3.3b 2.3ab 1.3ab 2.3ab O-buckwheat 0.7a 0.1a 2.5b 2.8b 5.7c O-acidulous 1.2a 2.0a 1.9a 2.9a 2.1a F-sweet 1.2a 2.1a 3.7b 2.6ab 3.8b F-rancid 3.9c 2.9bc 1.6ab 0.9a 0.6a F-bitter 0.6ab 0.2a 1.0ab 2.2bc 3.4c F-buckwheat 0.1a 0.1a 3.4b 3.7b 6.5c F-acidulous 1.5a 1.3a 1.3a 1.9a 1.2a F-yeast 0.7a 5.2b 2.7a 1.6a 2.4a Aftertaste 3.5a 2.7a 3.4a 3.4a 3.9a T-springiness 4.0ab 2.3a 5.4bc 5.9c 6.2c T-elasticity 1.1a 1.8a 4.2b 5.4b 5.5b T-mastication 2.4a 2.8a 1.7a 2.2a 2.0a T-adhesiveness 1.2a 1.8a 0.9a 1.1a 1.1a T-gumminess 1.7b 1.2ab 0.6a 0.5a 0.5a T-moistness 1.7ab 1.2a 3.7c 3.0bc 4.3c

significant Highly significant differences (P < 0.001)

were observed in the intensity of attributes such as:

colour, porosity, rancid odour, buckwheat odour, ‘rancid’

taste, ‘buckwheat’ taste, springiness and elasticity what

was connected with the kind of bread For the consumers,

in the profile of control bread, dominating desirable

sensory attributes were rancid odour and ‘rancid’ taste It

should be emphasised that the control bread

demon-strated approximately six times as high ‘rancid’ taste as

the bread with 40% BF This note probably did affect the

sensory overall quality In contrast, in the breads with

BF, the ‘buckwheat’ attribute dominated, being

accom-panied by bitterness Moreover, the results proved that

the colour, springiness, elasticity and moistness of breads

increased with the addition of BF The sample with an

increasing addition of BF was characterised by a higher

moistness It suggests that BF might kept the softness of

gluten-free bread during storage

To observe the above-mentioned differences in the

breads more clearly, the PCA was performed This

statistical method allows us to see a graphic

represen-tation of the data so that the variations between the

samples can be more easily interpreted Four principal

components (PC1–PC4) were extracted among which

the first two principal components (PC1 and PC2)

accounted for 94.10% of the total variance PC1

explained the majority of the variations, comprising

75.98%, while 18.12% was because of PC2 The first

two principal components were plotted in Fig 2a,b It

can be seen that PCA technique differentiated the

samples by kind of breads (Fig 2a) The samples

formed distinctly separate cluster found along the first

and second principal component The control and 10%

BF-enhanced breads were located oppositely to 20%

BF, 30% BF and 40% BF respective breads on the PCA

chart, which indicated their different sensory

character-istics The distribution of sensory attributes in space

defined by the first and second PCA dimensions is

presented in Fig 2b The attributes 1 (colour), 3 (rancid

odour), 6 (buckwheat odour), 9 (rancid ‘taste’), 10

(bitter taste), 11 (‘buckwheat’ taste), 15 (springiness), 16

(elasticity), 19 (gumminess) and 20 (moistness) had a

high loading (‡0.9) in PC1 Whereas the second PC2 was

differentiated by the attributes such as 2 (porosity), 5

(yeast odour) and 13 (yeast taste) It is a common

knowledge that the PC1 contains the most important

information and includes more important

characteris-tics Thus, these attributes had a decisive effect on the

variation in the sensory quality of the samples This was

also indicated by the length of plotted the vectors

describing the sensory attributes

Concluding remarks

Buckwheat flour as a natural source of minerals and

antioxidant activity, and also as a structure-forming

factor improving the sensory quality, can be used forpreparation of new buckwheat-enhanced gluten-freebreads The positive correlation coefficient betweenthe overall quality of buckwheat-enhanced gluten-free breads and their ability to scavenge ABTS+· andDPPH·radicals and their reducing capacities by CV wasfound (r = 0.98) Buckwheat flour could be incorpo-rated into gluten-free formula and then offers buck-wheat-enhanced gluten-free breads with more functionalcomponents and higher antioxidative and reducingcapacities The results proved that the addition of the

17 18

BF to gluten-free bread formula improved the sensory

quality The overall sensory quality of

buckwheat-enhanced breads was significantly higher than that

obtained for gluten-free bread It indicates that these

kinds of breads can be offered for the consumers,

especially when gluten-free formula is substituted by

40% of BF

References

Brand-Williams, W., Cuvelier, E & Berset, C.M (1995) Use of free

radical method to evaluate antioxidant activity

Lebensmittel-Wissenschaft und-Technologie, 28, 25–30.

Christa, K & Soral-S´mietana, M (2008) Buckwheat grains and

buckwheat products – nutritional and prophylactic value of their

components – a review Czech Journal of Food Sciences, 26, 153–162.

Fasano, A & Catassi, C (2001) Current aproaches to diagnosis and

treatment of celiac diasease: an envolving spectrum

Gastroenterol-ogy, 120, 636–651.

Gabrovska, D., Fiedlerova, V., Holasova, M et al (2002) The

nutritional evaluation of underutilized cereals and buckwheat Food

and Nutrition Bulletin, 23, 246–249.

Gallagher, E., Gormley, T.R & Arendt, E.K (2004) Recent advances

in the formulation of gluten-free cereal-based products Trends in

Food Science and Technology, 15, 143–152.

Hordyjewska, A & Pasternak, K (2004) Magnez a ukad

sercowo-naczyniowy (eng Magnesium and cardiovascular system) Journal of

Elementology, 9, 71–80 (English abstract).

ISO 8586–1 (1993) Sensory analysis – General guidance for the

selection, training and monitoring of assessors – Part 1: Selected

assessors.

ISO ⁄ DIS 13299:199 (1998) Sensory analysis – Methodology – General

guidance for establishing a sensory profile ISO 8589:1998b Sensory

analysis – General guidance for the design of test rooms.

Kennedy, N.P & Feighery, C (2000) Clinical features of celiac disease

today Biomedicine and Pharmacotherapy, 54, 373–380.

Kreft, I., Sˇkrabanja, V., Ikeda, S., Ikeda, K & Bonafaccia, G (1996).

Dietary value of buckwheat Research Reports of the Biotechnical

Faculty of the University of Lubljana, 67, 73–78.

Lawless, H.T & Heymann, H (1999) Sensory Evaluation of Food:

Principles and Practices Pp 341–372 Gaithersburg, MD: Chapman

and Hall, Aspen Publication.

Lazaridou, A., Duta, D., Papageorgiou, M., Belc, N & Biliaderis,

C.G (2007) Effects of hydrocolloids on dough rheology and bread

quality parameters in gluten-free formulations Journal of Food

Engineering, 79, 1033–4714.

Li, S & Zhang, Q (2001) Advances in the development of functional

foods from buckwheat Critical Reviews in Food Science and

Nutrition, 41, 451–464.

Lin, L.-Y., Liu, H.-M., Yu, Y.-W., Lin, S.-D & Mau, J.-L (2009).

Quality and antioxidant property of buckwheat enhanced wheat

bread Food Chemistry, 112, 987–991.

Murray, J.A (1999) The widening spectrum of celiac disease.

American Journal of Clinical Nutrition, 69, 354–365.

Oomah, B.D & Mazza, G (1996) Flavonoids and antioxidative

activities in buckwheat Journal of Agriculture and Food Chemistry,

44, 1746–1750.

Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M & Evans, C (1999) Antioxidant activity applying an improved ABTS Radical cation decolorization assay Free Radical Biology and Medicine, 26, 1231–1237.

Rice-Ribotta, P.D., Ausar, S.F., Morcillo, M.H., Perez, G.T., Beltramo, D.M & Leon, A.E (2004) Production of gluten-free bread using soybean flour Journal of the Science of Food and Agriculture, 84, 1969–1974.

Shahidi, F & Naczk, M (1995) Methods of analysis and fication of phenolic compounds In: Food Phenolic: Sources, Chemistry, Effects and Applications (edited by F Shahidi &

quanti-M Naczk) Pp 287–293 Lancaster, Pennsylvania: Technomic Publishing Company.

Sivaramakrishnana, H.P., Sengeb, B & Chattopadhyay, P.K (2004) Rheological properties of rice dough for making rice bread Journal

of Food Engineering, 62, 37–45.

Soral-S´mietana, M., S´wigon´, A & Markiewicz, K (2001) mental products from wheat flour of decreased sodium content and supplemented with calcium and magnesium Biuletyn Magne- zologiczny, 6, 165–175 (in Polish, summary in English).

Experi-Stone, H & Sidel, J.L (1993) Sensory Evaluation Practices, 2nd edn San Diego, California: Academic Press.

Thompson, T., Dennis, M., Higgins, L.A., Lee, A.R & Sharrett, M.K (2005) Gluten-free diet survey: are Americans with coeliac disease consuming recommended amounts of fibre, iron, calcium and grain foods? Journal of Human Nutrition and Dietetics, 18, 163–169 Torbica, A., Hadnadev, M & Dapcˇevic´, T (2010) Rheological, textural and sensory properties of gluten-free bread formulations based on rice and buckwheat flour Food Hydrocolloids, 24, 626–632 Watanabe, M (1998) Catechins as antioxidants from buckwheat (Fagopyrum esculentum Moench) groats Journal of Agricultural and Food Chemistry, 46, 839–845.

Wijngaard, H.H & Arendt, E.K (2006) Buckwheat Cereal try, 83, 391–401.

Chemis-Wronkowska, M., Soral-S´mietana, M & Krupa, U (2008a) anka wypiekowa bezglutenowa (eng Gluten-free baking formula) Polish Patent Specification No P 386254 (in Polish).

Miesz-Wronkowska, M., Troszyn´ska, A., Soral-S´mietana, M & Woejszo, A (2008b) Effects of buckwheat flour (Fagopyrum esculentum Moench)

on the quality of gluten-free-bread Polish Journal of Food and Nutrition Sciences, 58 ⁄ 2, 213–216.

Zielin´ska, D & Zielin´ski, H (2009) Low molecular weight oxidants and other biologically active components of buckwheat seeds The European Journal of Plant Science and Biotechnology, 3, 29–38.

anti-Zielin´ska, D., Szawara-Nowak, D & Zielin´ski, H (2007) Comparison

of spectrophotometric and electrochemical methods for the ation of antioxidant capacity of buckwheat products after hydro- thermal treatment Journal of Agriculture and Food Chemistry, 55, 6124–6131.

evalu-Zielin´ska, D., Frias, J., Piskua, M.K., Kozowska, H., Zielin´ski, H & Vidal-Valverde, C (2008) Evaluation of the antioxidant capacity of lupin sprouts produced with presence of selenium European Food Research and Technology, 227, 1711–1720.

Zielin´ski, H., Michalska, A., Ceglin´ska, A & Lamparski, G (2008) Antioxidant properties and sensory quality of traditional rye bread

as affected by the incorporation of flour with different extraction rate in the formulation European Food Research and Technology,

226, 671–680.

Original article

Brining kinetics of different cuts of anchovy (Engraulis anchoita)

Marina Czerner* & Marı´a I Yeannes

Grupo de Investigacio´n Preservacio´n y Calidad de Alimentos, Departamento de Ingenierı´a Quı´mica, Facultad de Ingenierı´a, Universidad Nacional

de Mar del Plata-CONICET, Juan B Justo 4302, (7600) Mar del Plata, Argentina

(Received 1 March 2010; Accepted in revised form 23 June 2010)

fish is immersed in saturated brine (osmotic solution) until equilibrium The aim of this work was to studythe mass transfer kinetics in three different cuts of anchovy (whole fish, gutted fish and fillet) and to model itusing the Zugarramurdi and Lupı´n and the Peleg equations Fillet reached equilibrium after 5 h, gutted fishafter 10.42 h and whole fish after 19.75 h Equilibrium constants were 1.054, 0.706 and 0.603 for fillet, guttedfish and whole anchovy, respectively These results indicate that the presence of skin affects both the initialrate of mass transfer and the equilibrium water and salt content The Peleg model was more suitable todescribe the salt gain kinetics, and the Zugarramurdi and Lupı´n model results in a more accurate prediction

of water loss and equilibrium conditions

Introduction

Engraulis anchoitais the most abundant pelagic species

in the south-western Atlantic Ocean, distributed from

Brazilian to Argentinean waters Argentina is a pioneer

country in the exploitation and manufacture of this

species for human consumption, being salted ripened

anchovy the main product This is a traditional product

with typical sensorial characteristics and strongly

posi-tioned in the international market It is exported as a

commodity in barrels mainly to Spain, Peru, the United

States, Italy and Morocco, and a small amount is locally

processed to supply the domestic market (Madureira

et al., 2007)

The conventional process of salting–ripening of

anchovy is based on empirical knowledge developed in

European countries, where the raw material for this

product is Engraulis encrasicolus Other anchovy-type

products are also produced from herring (Clupea

harengus), sprats (Sprattus sprattus) and sardine

(Sardi-na pilchardus) (Steffa´nson & Guðmundsdo´ttir, 1995) In

Argentina, the procedure was adapted to the species

E anchoita The process involves a preliminary

opera-tion of wet salting (brining), where whole fish is

immersed in saturated brine until equilibrium of water

and salt content is reached in the muscle During this

stage, water activity (aw) is reduced from 0.99,

corre-sponding to raw fish, to 0.80–0.84 (Filsinger, 1987).Following this, anchovies are handled beheaded andpartially gutted (leaving gonads and pyloric caeca),placed in barrels alternating layers of fish and salt andpressed The ripening process may take from 8 to

12 months and implies several transformations, whichinclude proteolysis and lipolysis As a result of thesechanges, the product acquires firm consistency, reddishcolour, juicy texture and a characteristic odour andflavour (Filsinger et al., 1982)

The stage of brining can be described as an osmoticdehydration process (OD), in which driving force forwater removal is set up owing to a dissimilar osmoticpressure between the food and the surrounding solution.The internal structure of the muscle is not a perfectsemipermeable membrane; thus, mass transport in the

OD is actually a combination of simultaneous water andsolute transfer During brining, the major transportedsolutes are salt from brine to fish flesh and proteins inthe opposite direction (Barat et al., 2003; Gallart-Jornet

et al., 2007a) The rate of diffusion depends on operativefactors such as temperature and concentration of theosmotic solution, size and geometry of the material, thesolution to material mass ratio and the level of agitation

of the solution (Rastogi et al., 2002, 2005; Barat et al.,2003; Ochoa Martinez & Ayala Aponte, 2005)

Different theoretical and empirical approaches havebeen employed to study and model the OD process.Models based on solution of Fick’s second law have

*Correspondent: E-mail: mczerner@fi.mdp.edu.ar

been broadly used to describe OD process in different

fish and meat products (Wang et al., 2000; Gou et al.,

2003; Telis et al., 2003; Graiver et al., 2006; Corzo &

Bracho, 2007; between others) On the other hand,

empirical models are preferred in some cases because of

their relatively simple applicability Zugarramurdi &

Lupı´n (1980) established a general model that allows the

achievement of general relationships for fish salting

kinetics and process equilibrium Several authors

showed that the empirical model of Zugarramurdi and

Lupı´n adequately describes water and salt kinetics of

osmotically dehydrated fish, for which yellowtail

(Berhimpon et al., 1990) and sardine fillets under

various conditions (Corzo & Bracho, 2005, 2006a;

Bellagha et al., 2007) are good examples The Peleg

(1988) model also describes sorption curves that

approaches to the equilibrium asymptotically, and it

have been used to describe OD in chickpea (Turhan

et al., 2002), sardine sheets (Corzo & Bracho, 2006b),

potato (Khin et al., 2006), apricot (Khoyi & Hesari,

2007) and chicken breast (Schmidt et al., 2009)

The possibility to modify the traditional process of

salting–ripening was explored in a previous work

(Czerner & Yeannes, 2008) Salting and ripening stages

were modified using nontraditional fish cuts in this

species (fillet and beheaded-partially gutted) Products

obtained showed sensory characteristics similar to those

obtained by traditional methods, indicating that the

modifications proposed were potentially applicable from

the sensorial point of view The analysis of mass transfer

behaviour in these cuts during OD is of great

techno-logical importance because it will eventually allow for

the estimation of immersion time and composition at

equilibrium There is scarce literature related to whole

fish salting, and up to our knowledge, no attempts have

been made to study OD process in whole or

beheaded-partially gutted fish by applying the Peleg model The

studies have been oriented to process of mass transfer on

fillet or pieces with defined shape and dimensions, but

not with the presence of skin, head or viscera As

regards the Zugarramurdi and Lupı´n model, it has been

applied to study wet salting in whole anchovy and dry

salting in gutted anchovy (Zugarramurdi & Lupı´n,

1980)

The objectives of this study were to analyse the mass

transfer process in different cuts of E anchoita during

brining stage and to model the kinetics of OD by two

empirical models: the Zugarramurdi & Lupı´n (1980) and

the Peleg (1988) models

Materials and methods

Raw material

Anchovy used for the experiment was caught near Mar

del Plata, Argentina, during November Following

catch, fish was placed in bins with ample ice andmaintained in this condition until they arrived to thelaboratory

Sample preparation and experimental designFish was arranged into three lots depending on the guttingmethod applied: W, whole anchovy (traditional method)(length: 136 ± 5 cm, thickness: 11 ± 1 mm); H&G,beheaded-partially gutted (length: 112 ± 9 cm, thick-ness: 11 ± 1 mm); and F, fillet (length: 112 ± 9 cm,thickness: 5 ± 1 mm) Cuts were performed by handaccording to standard industrial procedures

Saturated brine (26% NaCl) was prepared with salt(NaCl) Anchovies were immersed in the brine at a 1:1ratio to maintain the relationship used in industry andkept immersed by a screen cover Brine was maintained

at saturation by adding extra salt During the ence, the NaCl content of brine was checked followingMohr’s method (Kirk et al., 1996) The three lots weremaintained at 15C in an adiabatic chamber, andsamples of approximately 200 g (15–20 pieces for lot

experi-W,40–60 pieces for lot H&G, 80–100 pieces for lotF) were periodically taken until equilibrium wasreached An equal mass of brine was extracted at eachsample time to maintain the brine-to-fish ratio

Chemical analysesFresh anchovy was analysed to determine chemicalproximal composition Minced anchovy flesh was anal-ysed for water content by oven-drying at 105 ± 1Cuntil constant weight (AOAC, 1990); fat content by acidhydrolysis method (AOAC, 1990); protein content byKjeldhal (AOAC, 1993); and ashes by incineration at

500C (AOAC, 1993)

During brining, samples were analysed for their water(AOAC, 1990) and NaCl content (Kirk et al., 1996) ForNaCl determination in brined anchovy, the dry residuewas boiled with distilled water during 5 min and thenfiltered and made up to 250 mL In fresh anchovy, thedry residue was calcined at 500C (AOAC, 1993) andmade up to 100 mL Titration was performed onaliquots of these extracts At the end of brining, proteincontent was determined in fish muscle, and wateractivity (aw) was measured in muscle and brine The awwas measured by a digital hygrometer Aqualab, modelCX-2T (Decagon, Pulman, WA, USA) Analyses wereconducted in triplicate for water content and inquadruplicate for salt content and aw

Mathematical modelsThe Peleg modelMass transfer kinetics of salt and water were modelledusing the Peleg equation as follows:

Xi¼ X0

i  t

where Xiand X0

i are the water or salt content (dry basis,

g⁄ gdb) at dehydration time t (h) and at instant 0,

respectively In Eqn (1), ‘±’ becomes ‘+’ if the process

is salt gain and ‘)’ if the process is water loss The

constant k1 (h (g⁄ gdb))1) (the Peleg rate constant) is

related to mass transfer rate at the beginning of the OD

process: t = t0, according to Eqn (2) The constant k2

((g⁄ gdb))1) (capacity constant) is related to moisture and

salt content attainable at tfi ¥, being Xi¼ Xeqi (Eqn 3)

The Zugarramurdi and Lupı´n model (Z&L model)

The following mathematical model, with an exponential

approach to the equilibrium value of salt and water

concentrations, was proposed by Zugarramurdi &

Lupı´n (1977, 1980)

dXi

where Xi and Xeqi are the water or salt content (dry

basis, g⁄ gdb) at dehydration time t (h) and at

equilib-rium, respectively, and k ((g⁄ gdb) h)2) the corresponding

specific rate constant

Integration of Eqn (4) with the initial condition

Xið0Þ ¼ X0

i results in:

Xi¼ X0i:ek:tþ Xeqi :ð1  ek:tÞ ð5Þ

The fitting of the two models to experimental data was

performed by nonlinear regression analyses using the

software OriginPro 7.5 (OriginLab Corporation,

Northampton, MA, USA)

Statistical analysis

The analysis of variance (anova) was carried out to find

effects of the gutting method and time on water and salt

transfer Differences between means were analysed using

the Tukey’s test for post hoc comparison Analyses were

performed using statistica 6.0 (Statsoft, Inc., Tulsa,

OK, USA)

For each case, the goodness of fit was evaluated by the

correlation coefficient (R2) and the root mean square

error (RMSE, Eqn 6)

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1

to analysis of variance and means comparison acrossgutting method and time effects The effect of the guttingmethod on both water loss and salt gain was highlysignificant (P < 0.001) Interaction between guttingmethod and brining time was detected (P < 0.001)

The equilibrium condition was defined as not cant variations in water and salt content In lot W,equilibrium was achieved after 19.75 h; in lot H&G, after10.42 h and in lot F, after 5 h (P < 0.01) Equilibriumtime obtained for lot W agrees with the results obtained

signifi-by Zugarramurdi & Lupı´n (1977) for whole anchovy Thedynamics of salt gain and water loss showed interestingdifferences between lots As it can be seen in Fig 1, lot Fshows a higher slope at the beginning of the briningprocess indicating high initial rate of salt and watertransfer As pointed out in earlier works, various factorsmust be considered to explain this behaviour: (i) lot F ishalf the thickness of H&G and W; therefore, the length ofthe diffusion pass is shorter, resulting in more rapid masstransfer; (ii) the ratio of the surface area to the charac-teristic length is bigger in lot F; thus, the water loss andsolid gain increase (Rastogi et al., 2005); and (iii) lots Wand H&G had skin in both surfaces offering andadditional resistance to mass transfer between fish muscleand brine (Zugarramurdi & Lupı´n, 1980), whereas in lot

F the complete internal surface of the fillet is directlyexposed to brine facilitating the diffusion process

The results of anova showed that salt contentincreased significantly from the beginning of brininguntil equilibrium in the three lots However, the reduc-tion in water content at the beginning of OD showeddifferent behaviour between lots During the first 1.5 h

of treatment, the reduction in water content was