Exploiting blackcurrant juice press residue in extruded snacks

lable at ScienceDirect

LWT - Food Science and Technology 57 (2014) 618e627

Contents lists avai

LWT - Food Science and Technology

journal homepage: www.elsevier .com/locate/ lwt

Exploiting blackcurrant juice press residue in extruded snacks

Leenamaija Mäkilä a, Oskar Laaksonen a, Jose Martin Ramos Diaz b, Marjatta Vahvaselkä c,Olavi Myllymäki c, Ilkka Lehtomäki c, Simo Laakso c, Gerhard Jahreis d, Kirsi Jouppila b,Petra Larmo a,1, Baoru Yang a, Heikki Kallio a,*

a Food Chemistry and Food Development, Department of Biochemistry, University of Turku, FI-20014 Turku, FinlandbDepartment of Food and Environmental Sciences, University of Helsinki, FI-00014 Helsinki, FinlandcDepartment of Biotechnology and Chemical Technology, School of Chemical Technology, Aalto University, FI-00076 AALTO, FinlanddDepartment of Nutritional Physiology, Friedrich Schiller University of Jena, D-07743 Jena, Germany

a r t i c l e i n f o

Article history:Received 8 August 2013Received in revised form11 November 2013Accepted 4 February 2014

Keywords:BlackcurrantExtrusionLikingPhysical propertiesPress residue

* Corresponding author. Tel.: þ358 2 333 6870; faxE-mail addresses: [email protected] (L. Mä

(O. Laaksonen), [email protected] ([email protected] (M. Vahvaselkä), [email protected] (I. Lehtomäki), [email protected] (G. Jahreis), [email protected] (P. Larmo), [email protected] (B.(H. Kallio).

1 Present address: Aromtech Ltd., Veturitallintie 1,

http://dx.doi.org/10.1016/j.lwt.2014.02.0050023-6438/� 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

Extrusion process was developed to exploit blackcurrant juice press residues from industrial side-streams.Press residues obtained from conventional enzymatic pressing, with high content of fiber and seed oil, andnovel non-enzymatic juice processing, with high content of sugars, fruit acids and anthocyanins, wereextrudedwith barley flour, oat flour or oat bran. The recipes consisted of blackcurrant press residues (30%),cereal materials (40%) and potato starch (30%) and small amount of sugar and salt. When compared toenzymatic press residue and oat bran, the novel non-enzymatic press residue extruded with barley or oatflour had higher expansion, lower hardness and density, higher redness (a*), lower pH, and higher contentsof fructose, glucose and fruit acids, all contributing positively to liking of texture, appearance, and flavor aswell as berry-like experience. These characteristics were obtained with more gentle processing parame-ters, consisting of a lower total mass flow, screw speed and barrel temperature. Female consumers gavelower ratings in flavor, appearance and overall pleasantness for blackcurrant snacks thanmales. The studypresented a sustainable way of utilizing industrial press residues from different processes of berry juicepressing for production of healthy snacks and breakfast cereals.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The food industry generates a vast quantity of by-products byprocessing fruits, berries and vegetables. Side streams, such asskins, seeds, stems and cores of the fruits and berries, are rich invarious phytochemicals. They include for example fiber and edibleoils with valuable polyunsaturated fatty acids important for humanhealth, and thus ought to be utilized more thoroughly (O’Shea,Arendt, & Gallagher, 2012). Currant seeds contain especially sig-nificant amounts of g-linolenic, a-linolenic and stearidonic acids(Johansson, Laakso, & Kallio, 1997), which have positive effects on

: þ358 2 231 7666.kilä), [email protected] Diaz), marjatta.

[email protected] (O. Myllymäki),aalto.fi (S. Laakso), b6jage@(K. Jouppila), petra.larmo@Yang), [email protected]

95410 Tornio, Finland.

skin health and on symptoms of atopic dermatitis (Linnamaa et al.,2010). These by-products are currently mainly used as animal feed,or are taken as landfill or for incineration (O’Shea et al., 2012).Potentially more efficient ways to exploit the side-streams wouldbe to use them as colorants and antibrowning additives, as anti-microbial agents to improve the shelf-life, or as flavoring in-gredients (Ayala-Zavala et al., 2011; Viuda-Martos, Ruiz-Navajas,Fernándes-López, & Pérez-Àlvarez, 2010). The extracts of phenolicor seed oil fractions are potentially applicable in the food andpharmaceutical industries (Sandell et al., 2009;Wijngaard, Hossain,Rai, & Brunton, 2012).

Extrusion cooking has been widely investigated for exploitingthe by-products of food manufacturing. The extrusion process ispreferred over conventional cooking due to the distinct texturalproperties of the end-products, such as a high expansion ratio, lowdensity, crispiness, crunchiness and hardness (Brennan, Brennan,Derbyshire, & Tiwari, 2011; Meng, Threinen, Hansen, & Driedger,2010), all important parameters for the acceptability of the prod-ucts for consumers (Patil, Berrios, Tang, & Swanson, 2007).Increasing consumer demand has arisen for nutritious ready-to-eatsnack products with enhanced bioactive compounds (Brennan,

Fig. 1. Five parallel pictures of the samples: enzymatic residue (ER) extrudates A. ER-barley, B. ER-oat and C. ER-oat bran; non-enzymatic residue (NR) extrudates D. NR-barley, E.NR-oat and F. NR-oat bran.

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627 619

Derbyshire, Tiwari, & Brennan, 2013). Traditional snacks consistmainly of starch, with low dietary fiber and protein contents. Thenutritional value of these snacks could be improved by usingwhole-grain flour with bioactive fruit and berry by-products (Altan,McCarthy, & Maskan, 2008a; Faraj, Vasanthan, & Hoover, 2004; Maet al., 2012; Potter, Stojceska, & Plunkett, 2013). High contents ofdietary fiber, lipids and protein introduce challenges in the extru-sion process (Ya�gcı et al., 2014).

The aim of this study was to investigate a sustainable way toutilize blackcurrant press residues from different industrial juiceprocesses by exploiting extrusion technologies, based on our previ-ous study by Tahvonen et al. (1998). Press residues of the presentstudy were obtained from different industrial pressing processesdescribed earlier (Laaksonen, Mäkilä, Tahvonen, Kallio, & Yang,2013). The purpose was to gain a high nutritional value from highcontent of press residue (30%) extrudedwith barley flour, oatflour oroat bran, but still retain the textural advantages unique to extrudedproducts. To studyandexplain thehedonic responses of extrudates, avariety of physico-chemical measurements were conducted.

2. Materials and methods

2.1. Extrusion ingredients

Blackcurrants of cultivar ‘Mortti’ grown in Finland in 2011 wereprocessed in 500 e 1300 kg batches for juice production by Saar-ioinen Oy (Huittinen, Finland) after the berries were separatedfrom leaves and stems (Toripiha Oy, Vesanto, Finland). The enzyme-

aided standard protocol with pectinase enzyme and a hydraulicpress yielded the press residue ER (Enzymatic Residue) for furtherextrusion tests. The process contained the steps of thawing,crushing, heating, enzyme addition, incubation, and pressing. Inaddition, juice and press residue NR (Non-enzymatic Residue) wereproduced by otherwise following the same protocol as above butomitting the enzyme treatment. Non-enzymatic juice process wasperformed in duplicate.

The batches of both ER and NR were air-dried at 40 �C (MTTAgrifood Research Finland, Piikkiö, Finland). The dried ER wasproperly milled to break down the seeds and sieved using a 0.5 mmscreen. Potato flour (7%) was used in the milling process to bind thereleased seed oil and to facilitate the milling. The NR residue wascrushed to coarse particles, without aiming to break down the seeds.The other ingredients barley, oat and potato flour, oat bran, andpotato flakes were obtained from Leipomo Rosten Oy (Turku,Finland). Sugar and salt were conventional commercial products.Four commercial, extruded breakfast cereals were selected asreference products based on their similarity to our samples in size,shape and texture. Reference 1 was a cocoa cereal (whole grainwheat flour), reference 2 was a spelt cereal, reference 3 wasmade ofwhole grain rye and oat flour with sea buckthorn juice, and refer-ence 4 was made of whole grain wheat, rye, oat and barley flour.

2.2. Extrusion process

Six different recipes were used to produce six extrudates (Fig. 1and Table 1.). The samples were extruded with a Clextral BC 21

Table 1The recipe and the process parameters of the extrudates.

Enzymatic press residue Non-enzymatic pressresidue

ER-barley ER-oat ER-oatbran

NR-barley NR-oat NR-oatbran

The recipePress residue (g/100 g) 27 27 27 28 28 28Barley flour (g/100 g) 39 e e 38 e e

Oat flour (g/100 g) e 39 e e 38 e

Oat bran (g/100 g) e e 39 e e 38Potato flour (g/100 g) 15 15 15 14 14 14Potato flake (g/100 g) 15 15 15 14 14 14Sugar (g/100 g) 4.8 4.8 4.8 4.7 4.7 4.7Salt (g/100 g) 0.5 0.5 0.5 0.5 0.5 0.5The process parametersMeal feed rate

(kg/h)4.9 5.6 7.0 5.2 5.2 5.2

Water feed rate(mL/h)

400 400 730 350 350 350

Screw speed (force � g) 400 410 420 400 400 400Barrel temperature (�C) 100 100 102 94 94 94Screw torque (Nm) 16 16 14 16 14 13SME (kWh/kg) 0.3 0.2 0.2 0.3 0.2 0.2Bearing pressure

(bar)100 120 75 55 94 100

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627620

twin-screw extruder (Clextral S.A., Groupe Clextral, Firminy,France) with a K-Tron Somer feeder (K-Tron International, Inc,Pitman, NJ). The basic recipes and the extruding parameters wereselected based on the results of a preliminary trial (data notshown). The extrudates were designed to resemble commerciallyextruded breakfast cereals in their appearance, shape and textureand also to be useful as a basis for different snack products. Thesamples consisted of 27e28% press residue (ER or NR) and 38e39%cereal material: barley flour, oat flour or oat bran, and 14e15%potato flour and flakes. In addition, 4.7e4.8% sugar and 0.5% saltwere added (Table 1.).

The screw speed of the extruder and feeders as well as the speedof the cutter were controlled. The barrel length was 400 mm,consisting of four 100 mm modules. The first feed section modulewas a transfer-only unit and the rest of the modules were equippedwith both controllable cooling and heating units. There were insidethe barrel two intermeshing co-rotating forward-screw elementsand one reverse screw in front of the die. The diameter of eachscrew was 21 mm. The combination of screw elements from feedopening towards the die section was:

1) Forward screws, length 100 mm, pitch 15 mm2) Forward screws, length 150 mm, pitch 11 mm3) Forward screws, length 125 mm, pitch 8 mm4) Reverse screws length 25 mm, pitch 8 mmwith three openings

of width 5 mm

After the last module, the die section consisted of a 50 mmchannel of the size of the extruder barrel element. In this elementtwo openings from the end of the screw shafts were combined intoone opening through which the plastic melt was distributed bydividing plate to flow through a 4 mm circular die into the atmo-spheric pressure. The IS Leroy Somer cutter (IS Leroy Somer, France,part from Emerson Electric Co, St Louis, MO, USA), was installed justafter the die plate.

The extrusion parameters were: meal feed rate 4.9e7.0 kg/h,water feed rate 350e730 mL/h and screw speed 400e420 rpm(Table 1.). Temperatures of the first two heating units were main-tained at a constant 95 �C throughout the experiments. The barreltemperature of the last heating module, located before the die,

varied from 94 to 102 �C. The screw torques were calculated as theactual screw torque e the base torque, values varying from 13 to16 Nm. The bearing pressures varied from 55 to 120 bar betweendifferent dough mixtures. The specific mechanical energy (SME)was calculated from the following relationship: SME (kWh/kg) ¼[screw torque (Nm) � screw speed (rpm) � 2 � p � number ofscrews / {meal feed rate (kg h�1)� 60}] / 1000 (Carvalho &Mitchell,2000). The values varied from 0.2 to 0.3 kWh/kg. The ER-oat brandiffered the most from the other snacks due to a high content of b-glucan and lipids from the oat bran and free seed oil, resulting fromthe grinding process, all known to decrease expansion(Bhattacharya & Hanna, 1988; Faubion, Hoseney, & Seib, 1982;Yanniotis, Petraki, & Soumpasi, 2007). Thereby the snack wasextruded with the roughest process parameters to increase theexpansion of the sample, and with the highest water feeding rate toprevent the extrudate being too dry because of the water bindingproperties of b-glucan (Yanniotis et al., 2007). The extrudates werecut before they started to bend due to temperatures above glasstransition temperature, that is, the temperature when the productis in amorphous state (Truong, Bhandari, Howes, & Adhikari, 2004).

Extrudates were cooled at room temperature. No drying processwas performed since the moisture of the extrudates was clearlyunder 10% after cooling. The snacks were sealed in plastic bags(250 mm by 200 mm) after approximately one week, using nitro-gen to generate a modified atmosphere in the plastic bags. Thesebags were of multiple layer laminate PA/PE (Wipak Oy, Nastola,Finland) with 40 cc/m2/day oxygen penetration. The oxygen traceremaining in the sealed bags was 0.2e0.6%.

2.3. Sensory evaluation

A panel of 77 voluntary subjects (49 females and 28 males; age20e67) was recruited in the Turku area, mainly from the studentsand employees of the University of Turku. The subjects rated theirliking of the appearance (shape and color), flavor, texture and theoverall pleasantness of the six samples (Fig. 1.) on a balanced 9-point hedonic scale (from 1 ¼ dislike extremely to 9 ¼ likeextremely). Additionally, the subjects ranked the six samples inorder of preference (ISO 8587:2006). The extrudates were rated inone session and presented as blind-coded in a randomized order.The data was collected using Compusense-five data collectionsoftware (version 5.2, Compusense, Guelph, Canada). Tests wereconducted in controlled sensory laboratory conditions in accor-dance with ISO8589:2007 standard.

2.4. Physical properties

2.4.1. Size, sectional expansion index (SEI), bulk density, waterabsorption index (WAI) and water solubility index (WSI)

Length and height were measured as longitudinal elongationand cross-sectional growth, respectively. The mass of every spec-imen (weight) was measured (Precisa Instrument AG analytic scale,Switzerland). Sectional expansion index (SEI) was calculated as theratio between the cross-sectional area of the extrudate and the die(Alvarez-Martinez, Kondury, & Harper, 1988). The cross-sectionalarea of NR samples was measured from the “projection ring”around them (Fig. 1.). Bulk density of samples was determinedusing the procedure described by Dansby and Bovell-Benjamin(2003) as was the water absorption index (WAI) and water solu-bility index (WSI).

2.4.2. Hardness, color and pHHardness was measured at 23 �C using a universal testing ma-

chine (Instron 4465, Instron Ltd., High Wycombe, UK). Sampleswere vacuum-dried at 54 �C for 72 h prior to the hardness

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627 621

measurement. The universal testing machine was fitted with acylindrically shaped aluminum probe: this probe had a length of74.7 mm with an end-tip diameter of 20.9 mm. Single specimenswere placed directly under the probe. Hardness was measured asthe slope of the forceedistance curve (N/mm) when penetrationwas perpendicularly enforced at a speed of 5 mm/min on anextrudate structure. The force resisting the movement of the probewas recorded. A linear trend-line was fitted to the raw data inwhich force (N) was shown as a function of distance (mm). Thevalue of the slope represented the value of the hardness (slope N/mm) of the sample, which was reported as an average of ten rep-licates as a modified method from Kirjoranta et al. (2012). From thehardness data, the area under the hardness curve (area N * mm)was calculated using the Origin 8 (OriginLab Corporation, North-ampton, USA); Mathematical Integral Area. The area under thehardness curves indicated the total amount of energy needed tothoroughly break down the structure of the extrudates.

Before color measurements were taken, the extrudates weremilled using an ultra-centrifugal mill (Retsch ZM 200, Haan, Ger-many) at 5219 � g. The milled sample (3e5 g) was then placed on apetri dish and spread on a planar surface. The color space para-meters L*, a*, b* were measured using a Minolta CR-400 chro-mometer (Konica Minolta Sensing, Inc., Osaka, Japan), using amodified method from Voutila, Perero, Ruusunen, Jouppila, andPuolanne (2009). The L* value represents the lightness of thesnack, 0 ¼ black, 100 ¼ white. A positive value of a* represents theredness/magenta of the snack, a negative value its greenness. Apositive value of b* represents the yellowness of the sample, anegative value its blueness. The color difference between theextrudates and reference products (DE) was determined asDE ¼ (DL2 þ Da2 þ Db2)0.5. The pH measurements were taken bymeasuring 70 mL of distilled water to a beaker and adding 1.81 g ofthe grounded sample with constant stirring, based on a modifiedmethod from AOAC Method No. 981.12 (AOAC, 1990).

2.5. Chemical properties

2.5.1. Nutritional value AnalysesThe contents of energy, carbohydrates, protein, crude fat,

moisture and ash were analyzed at MTT (Jokioinen, Finland), whichis a test laboratory T024 accredited by FINAS. Energy and carbo-hydrates (all the carbohydrates which are metabolized in humanbody) were defined computationally using Evira’s instruction17030/1 (Finnish Food Safety Authority Evira, 2012). Protein (Kjel-dahl method), crude fat, moisture and ash contents were deter-mined according to AOAC (1980) methods. The seeds in NR samplesremained unbroken in the crude fat analysis, whereas seeds in ERwere already milled before extrusion.

Total fiber was determined enzymatically with BIOQUANT�

Total Dietary Fiber Method (Merck, Darmstadt, Germany) in repli-cates (AOAC International, 1995). For fiber analysis also crude pro-tein analysis was carried out, applying the Kjeldahl method(N � 6.25) in replicates. NDF-fiber (neutral detergent fiber) andADF-fiber (acid detergent fiber) analyses were carried out accord-ing to the method described by Van Soest, Robertson, and Lewis(1991). Hemicellulose was calculated as NDF-ADF (Van Soestet al., 1991).

Amino acids were verified by ion exchange chromatographywith ninhydrin post column derivatization (Eppendorf-BiotronikLC 3000 Amino Acid Analyzer, Hamburg, Germany) as described byLandry, Delhaye, and Jones (1992) in single measurements. Cali-bration standards were purchased from Onken GmbH (Gründau,Germany). Amino acids were analyzed to calculate the proportionsof the essential amino acids of the daily requirements satisfied from100 g of a sample for a person of 70 kg body weight, according to

WHO’s present estimation (Joint WHO/FAO/UNU ExpertConsultation, 2007).

2.5.2. Sugars and organic fruit acidsSugars and organic acids were analyzed at MTT Jokioinen.

Sugars were analyzed with HPLC using an in-house method basedon the method of the Nordic committee on food analysis No. 148(1993). Fruit acids were analyzed with HPLC using an in-housemethod based on HPLC Application Organic acid std ID No.14749(Phenomenex Applications, 2013).

2.6. Statistical analyses

Differences in physico-chemical properties between extrudateswere analyzed with One-way ANOVA with a suitable post hoc test(Tukey’s HSD or Tamhane test). Principal component analysis (PCA)was applied to find correlations between the extrudates and thereference products in physical properties. Two-way ANOVA (sam-ple, gender) was used for the analysis of the hedonic response data.Friedman’s analysis of variance was used for the rank order test. Tofind relationships between liking and physico-chemical properties,the partial least squares regression (PLS) method was applied forstandardized data. X-variables (predictors) were the physico-chemical properties and Y-variables (responses) were the hedonicresponses. Cross validation was used to estimate the number offactors for a statistically reliable model. ANOVA models were per-formed using SPSS 16.0 (SPSS Inc. H, Chicago, IL) and PLS and PCAmodels using Unscrambler 10.1 (Camo Process AS, Oslo, Norway).The criterion for statistical significance in all tests was p < 0.05.

3. Results and discussion

3.1. Sensory evaluation

Among the six samples evaluated, the liking ratings of extru-dates were higher for snacks with NR compared to ER extrudates(Table 2.). The NR extrudates did not differ statistically from eachother in the liking ratings. Barley-flour-based snacks were given thehighest ratings, and oat bran the lowest, in both press residues. NR-barley was the most liked (slightly liked in the balanced 9-pointscale) in all four liking attributes. ER-oat branwas rated the lowest,being highly dislike in appearance and slightly dislike in textureand overall pleasantness. The lowest ratings in flavor were given tothe snacks of ER-oat. In the rank order test (Table 2.), NR extrudateswere the most preferred, whereas those of ER-oat bran were theleast preferred.

Based on a two-way ANOVA, a gendermain effect was significantin rating flavor, appearance and overall pleasantness, but not texture(Table 2.). Interestingly, as a trend females gave lower ratings inflavor, appearance and overall pleasantness for blackcurrant snacksthandidmales. This gendereffectwas especiallyclear in the two leastliked samples for appearance, ER-oat bran and ER-oat, for which fe-males gave the lowest ratings. No significant gender effectwas foundin the rankorder test. Generally, it hasbeen shownthat females showmore positive attitudes towards healthy diets than domales (Hearty,McCarthy, Kearney, & Gibney, 2007; Kearney, Kearney, Dunne, &Gibney, 2000; Roininen, Lähteenmäki, & Tuorila, 1999).

3.2. Physical properties of the extrudates

3.2.1. Size, sectional expansion index (SEI), bulk density, waterabsorption index (WAI) and water solubility index (WSI)

NR-based extrudates had a more round shape (length to heightratio 1.1.e1.3), whereas ER snacks had a more oblong shape (ratio1.6e1.8) resulting from the cutting process as described in

Table 2The hedonic responses of the extrudates.

Enzymatic press residue Non-enzymatic press residue

ER-barley ER-oat ER-oat bran NR-barley NR-oat NR-oat bran

Appearance 4.5 � 1.4b 3.1 � 1.1c 2.5 � 1.2c 5.9 � 1.3a 5.8 � 1.3a 5.3 � 1.3a

Male* 4.7 � 1.2 3.6 � 1.2* 3.1 � 1.1* 5.8 � 1.3 5.9 � 1.3 5.4 � 1.3Female 4.5 � 1.6 2.9 � 1.0 2.2 � 1.2 6.0 � 1.4 5.8 � 1.3 5.3 � 1.3

Flavor 4.1 � 1.7b 3.9 � 1.7b 4.2 � 1.7b 5.7 � 1.7a 5.7 � 1.7a 5.4 � 1.5a

Male* 4.3 � 1.4 4.3 � 1.6 4.5 � 1.6 6.0 � 1.5 5.9 � 1.3 5.6 � 1.1Female 4.0 � 1.8 3.7 � 1.7 4.0 � 1.7 5.5 � 1.8 5.6 � 1.8 5.3 � 1.7

Texture 5.7 � 1.7a 4.3 � 1.7b 3.7 � 1.9b 6.4 � 1.4a 6.1 � 1.5a 6.3 � 1.2a

Male 5.9 � 1.3 4.5 � 1.7 3.6 � 1.7 6.4 � 1.3 6.2 � 1.3 6.3 � 0.9Female 5.6 � 1.9 4.2 � 1.7 3.7 � 1.9 6.4 � 1.4 6.0 � 1.6 6.2 � 1.3

Overall pleasantness 4.3 � 1.6b 3.6 � 1.6bc 3.4 � 1.6c 5.7 � 1.6a 5.6 � 1.5a 5.5 � 1.4a

Male* 4.6 � 1.5 3.9 � 1.5 3.6 � 1.5 6.1 � 1.5 5.9 � 1.3 5.8 � 1.4Female 4.1 � 1.6 3.5 � 1.6 3.3 � 1.6 5.5 � 1.6 5.4 � 1.6 5.3 � 1.4

Rank order test 2 3 4 1 1 1

Means and standard deviations (scale 1e9) of 77 subjects, 28 males and 49 females. Significant differences between extrudates in each attribute (two-way ANOVA withTukey’s HSD test; p < 0.05) are marked with aec. Statistically significant difference between genders (italics) in general (all samples; two-way ANOVA) or within sample(Student t-test; p < 0.05) are marked with *. Significant differences between extrudates in rank order test (most preferred ranking 1) is based on Friedman’s analysis ofvariance.

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627622

paragraph 2.2 (Table 3.). At the same time, NR extrudates had ahigher sectional expansion index (SEI) than the ER samples, as thecross-sectional area of NR snacks was measured from the “projec-tion ring” around them. However, NR extudates were smaller in sizethan the ER samples (Fig. 1.) because of the amorphous state of themelt, which could not hold the expansion. NR extrudates had alsolower weight and bulk density compared to ER-based snacks. Oat-bran-based snacks had the lowest expansionwhereas barley-basedextrudates clearly had the lowest bulk density. ER with an oatingredient, either as flour or as bran, had the highest density. NR-barley expanded the most and had the lowest density, whichmight explain it receiving the highest ratings in the liking of texture(Table 2.). ER-oat bran had the lowest expansion and clearly thehighest density, 2.9 times higher than NR-barley, which seemed toaffect the hedonic response to texture. The higher porosity from theexpansion of NR samples and barley-based snacks are also seen inFig. 1 as a rougher surface compared to ER extrudates and oat-bran-based samples. Higher expansion and porosity with lower densityis a result of the lower contents of b-glucan and lipids (Beer,Arrigoni, & Amado, 1996; Bhattacharya & Hanna, 1988; Bhatty,1993; Faraj et al., 2004; Faubion et al., 1982; Yanniotis et al., 2007).

Table 3Physical properties of the extrudates and the commercial references.

Enzymatic press residue Non-enzymatic p

ER-barley ER-oat ER-oat bran NR-barley NR-

Length (cm) 1.3 � 0.1a 1.2 � 0.2b 1.0 � 0.1c 1.0 � 0.1c 0.9Height (cm) 0.7 � 0.1bc 0.8 � 0.1bc 0.6 � 0.0d 0.9 � 0.1a 0.8Weight (mg) 95 � 23b 160 � 36a 180 � 28a 61 � 5.2c 50Expansion (SEI) 3.5 � 0.5bc 3.7 � 0.7bc 2.3 � 0.4d 4.9 � 0.6a 3.9Bulk density (g/mL) 17 � 0.1d 33 � 0.5b 55 � 0.7a 14 � 0.2f 16WAI 250 � 14d 290 � 8.0c 350 � 4.8ab 330 � 15b 360WSI 65 � 6.4a 48 � 2.0a 27 � 0.2c 53 � 3.9a 50Hardness (slope N/mm) 12 � 3.9b 21 � 6.4b 140 � 50a 13 � 3.6b 5.7Hardness (area N* mm) 23 46 49 13 12ColorL* 39 � 2.0cd 41 � 1.5ab 43 � 1.0a 40 � 1.3bc 38a* 15 � 0.3d 15 � 0.3de 14 � 0.2f 19 � 0.5b 20b* 18 � 0.4a 17 � 0.3b 9.6 � 0.1e 15 � 0.3cd 15DER1 14 � 1.9bc 12 � 1.8c 15 � 1.0b 15 � 1.0b 17DER2 44 � 2.3b 41 � 1.7cd 40 � 1.1d 43 � 1.6bc 46DER3 46 � 1.9cd 45 � 1.2d 48 � 0.5b 48 � 1.2bc 50DER4 48 � 2.6bc 47 � 1.6c 49 � 1.1bc 50 � 1.6b 52

pH 4.2 � 0.1a 4.3 � 0.2a 4.3 � 0.1a 3.7 � 0.0b 3.6

Means and standard deviations of 20 replicates from length, height, weight and expansiments, 4 replicates from WAI and WSI and 2 replicates from pH analyzes. Significant(p < 0.05) are marked with aef.

Barley-based extrudates had the lowest water absorption index(WAI), that is, the lowest swelling power of starch when comparedto other cereal materials (Table 3.). Snacks with oat bran had thelowest water solubility index (WSI) indicating the total degradationof starch granules (Colonna, Doublier, Melcion, de Monredon, &Mercier, 1984). As a trend, high WAI is an indication of low WSIand vice versa. WAI was highest in NR-oat and lowest in ER-barley.WSI was highest in ER-barley and lowest in ER-oat bran.

3.2.2. HardnessThe hardness (slope N/mm) of the samples varied considerably

from 5.7 to 140 N/mm, ER-oat bran being the hardest and NR-oatthe most brittle (Table 3.). The curves of hardness were drawnfrom the hardness data, as described in Section 2.4.2, to illustratethe force peaks during the rupture of the extrudates (Fig. 2A). Themost distinct among all the extrudates according to the hardnesscurves was the ER-oat bran. The breaking of the snack occurred inseveral high peak forces, the maximum being almost 30 N, indi-cating that the rupture of the extrudate happened in hard splits. ER-oat bran was extruded with the highest water feed and barreltemperature, often contributing to a hard and dense product

ress residue Reference products

oat NR-oat bran Reference 1 Reference 2 Reference 3 Reference 4

� 0.1c 0.9 � 0.1c 1.1 � 0.1 1.1 � 0.1 0.9 � 0.1 0.5 � 0.0� 0.1b 0.7 � 0.0c 1.1 � 0.0 1.3 � 0.1 0.5 � 0.0 0.5 � 0.0� 11d 80 � 13b 210 � 22 130 � 11 27 � 2.7 45 � 4.2� 0.6b 3.3 � 0.4c e e e e

� 0.1e 25 � 0.5c 23 � 0.5 9.9 � 0.3 15 � 0.4 24 � 0.4� 9.7a 350 � 1.1ab 320 � 2.1 460 � 11 400 � 4.2 470 � 25

� 1.8a 36 � 0.2b 47 � 0.2 37 � 0.9 40 � 0.9 25 � 0.7� 3.8c 17 � 6.9b 65 � 20 8.7 � 3.0 9.7 � 2.1 22 � 6.2

21 30 34 18 22

� 1.3d 39 � 1.3abc 52 � 0.7 80 � 0.6 76 � 0.5 79 � 1.2� 0.4a 18 � 0.2c 12 � 0.1 2.3 � 0.2 2.4 � 0.2 �4.1 � 0.3� 0.2d 16 � 0.2c 22 � 0.3 15 � 0.3 43 � 0.4 36 � 0.8� 1.2a 15 � 1.4b e e e e

� 1.4a 44 � 1.4b e e e e

� 1.1a 48 � 1.1b e e e e

� 2.0a 50 � 1.6ab e e e e

� 0.0b 3.7 � 0.1b e e e e

on, 5 replicates from bulk density, 10 replicates from hardness and color measure-differences between extrudates based on one-way ANOVA with Tukey’s HSD test

Fig. 2. Hardness values (N/mm: averages of triplicate measurements, 9e10 replicates)of the extrudates: ( ) ER-barley, ( ) ER-oat, ( ) ER-oat bran, ( ) NR-barley, ( ) NR-oatand ( ) NR-oat bran (A) and reference products: ( ) Reference 1, ( ) Reference 2, ( )Reference 3, ( ) Reference 4 B). The force (compressive load; N) needed for probe topenetrate to certain depth (compressive extension; mm) into the extrudate. ER-oatbran had 2e9 replicates in compressive extensions 0.8e3.0 mm.

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627 623

(Bisharat, Oikonomopoulou, Panagiotou, Krokida, &Maroulis, 2013;Ding, Ainsworth, Tucker, & Marson, 2005; Meng et al., 2010; Menis,Milani, Jordano, Boscolo, & Conti-Silva, 2013).With other samples, alower area under the hardness curves indicated that these sampleswere more brittle and broke down without any higher peak force,except ER-oat with the similar value of the area under the hardnesscurve as in ER-oat bran. The maximum peak force for ER-barley andNR-oat bran was approximately 10 N, for NR-oat and NR-barley itwas 5e7 N. The slope before the first major fracture peak refers tothe crispiness of the extrudates: the lower the slope, the crispierthe product (Altan, McCarthy, & Maskan, 2008b; Jackson, Bourne, &Barnard, 1996). Here, the crispiest and most liked snack seemed tobe NR-oat, the least crispy and least liked ER-oat bran.

3.2.3. ColorNR extrudates had higher a* values (Table 3.) with more red/

magenta color than ER snacks, which can also be visually detectedfrom Fig. 1. This was most likely due to the higher content of an-thocyanins of NR and themore gentle extrusion process parameters(Table 1.), since berry anthocyanins are vulnerable to high tem-peratures during the extrusion process (Khanal, Howard,Brownmiller, & Prior, 2009). The lower pH (0.5 lower) of NR-based snacks, resulting from their higher fruit acid content, kept

the anthocyanins in a more stable red form, that is, in the form ofthe red flavylium cation (AHþ). The more berry-like reddish color ofthe NR-based samples seen from Fig. 1, may explain the higherliking ratings of appearance when compared with the brownish ERextrudates. ER-oat bran had the weakest red color (lowest a*),which may have been a result of the highest barrel temperature(102 �C), and also from the increased moisture content (Obatolu,Skonberg, Camire, & Dougherty, 2005). The most yellow snack(highest b* values) was ER-barley, with ER-oat bran the least yellow.The ratio of redness (a*) to yellowness (b*) of NR-based snacks was1.1e1.3, while the ratio for ER-based extrudates was 0.8e1.5. ER-oatbran had the lightest color (highest L* value) (Table 3.). Stojceska,Ainsworth, Plunkett, and _Ibano�glu (2009) found a higher mois-ture content to increase the lightness of the product. Surprisingly,ER-oat branwas the darkest of the snacks (Fig. 1.). This dark color ofER-oat bran might have been the result of the Maillard browningreaction due to the high barrel temperature and high content ofsugars and amino acids (Table 4.) (Altan & Maskan, 2012). NR-oathad the darkest color (lowest L* value). The samples in this studywere round in shape, making grinding necessary. Thus, the colorvalues do not match the visually observed surface colors of theextrudates.

The PCA bi-plot (Fig. 3.) shows the correlation in physicalproperties between the extrudates and the reference products.Reference 1 correlated strongly with the snacks, and had similar-ities in density, hardness, area under the hardness curve, WSI,shape (length-to-height ratio) and color (Table 3.). As reference 1was an extruded cocoa cereal, its brownish color resembled the ER-based sample, giving the lowest DE values (Table 3.). Other refer-ences on the right side of the plot, and along PC-1, were notablydifferent from the extrudates, which weremore aligned along PC-2.Reference 4 was the most distinguishable from the other refer-ences, with high yellowness (b*), lightness (L*) and WAI values andwas less hard and dense compared to the other reference productsand extrudates (Table 3.). The hardness curves of the references 1, 3and 4 were very similar to ER-barley and NR-oat bran (Fig. 2B.).

3.4. Chemical properties

The chemical properties of the samples are presented in Table 4.In general, ER-based extrudates contained more energy, protein,crude fat and fiber than the NR-based snacks. In addition, the ERsnacks were richer in total fiber, neutral detergent fiber (NDF-fiber)and acid detergent fiber (ADF-fiber) than the NR extrudates. Thesehigh fat and fiber contents resulted from enzyme-aided juicepressing which efficiently broke down the berry structure and thusincreased the proportion of the seed, seed oil and skin fractions inthe press residue, compared to non-enzymatic pressing. Addi-tionally, the milling of the ER-residue before extrusion broke downthe seeds to release oil into the extrudates. For the NR samples, thecrude fat analysis showed mainly the fat content of the cerealmaterial, since the milling process did not aim to break down theseeds. The barley snacks contained more fiber and ADF-fiber thandid the oat flour or oat bran samples. Oat bran extrudates seemed tobe richest in protein, crude fat, ash, NDF-fiber and hemicellulose, ofwhich the ER-oat bran showed the highest content of components.

The higher contents of glucose and fructose in the NR extrudateswere due to the remains of juice and soluble solids in the pressresidue of the non-enzymatic process (Laaksonen, Mäkilä, et al.,2013). Analogously, the NR snacks were richer in citric acid, suc-cinic acid, acetic acid and total organic acids compared to the ER-based samples. The elevated levels of sucrose (Table 4.) were theresults of added sugar, not from press residue (Sandell et al., 2009).All the snacks were good sources of essential amino acids (Table 4.).The daily requirement for cysteine is satisfied if 100 g of the

Fig. 3. PCA bi-plot of physical properties of the extrudates and reference products (References 1e4). The physical properties are presented in blue, reference products in red andextrudates in brown font. Abbreviations of physical properties refer to Table 3. (For interpretation of the references to color in this figure legend, the reader is referred to the webversion of this article.)

Table 4Chemical properties of the extrudates.

Enzymatic press residue Non-enzymatic press residue

ER-barley ER-oat ER-oat bran NR-barley NR-oat NR-oat bran

Nutritional value (g/100 g)Energy (kJ/100 g) 1710 1730 1730 1660 1670 1680Protein 9.1 � 0.0d 10 � 0.0b 11 � 0.1a 7.6 � 0.1f 8.3 � 0.1e 9.5 � 0.1c

Carbohydrates 78 75 74 82 79 77Crude fat 6.2 � 0.1b 8.0 � 0.0a 8.0 � 0.0a 3.7 � 0.1e 4.8 � 0.2d 5.6 � 0.3c

Moisture 4.1 � 0.1c 5.0 � 0.2ab 5.2 � 0.2a 4.9 � 0.1ab 4.8 � 0.1ab 4.7 � 0.1b

Ash 2.4 � 0.1ab 2.4 � 0.2ab 2.5 � 0.3ab 2.0 � 0.4b 2.7 � 0.2ab 2.8 � 0.3a

Total fiber 21 20 21 19 14 18NDF-fiber 17 � 3.0 21 � 4.3 29 � 8.3 15 � 2.1 18 � 3.7 19 � 2.1ADF-fiber 8.6 � 1.0ab 8.0 � 1.3ab 7.3 � 2.1ab 10 � 3.4a 6.6 � 1.0b 6.2 � 0.8b

Hemicellulose 8.7 13 22 5.1 11 12Sugars and acids (g/100 g)Fructose 2.1 � 0.0c 1.6 � 0.0d 1.5 � 0.0e 4.8 � 0.0b 5.0 � 0.0a 4.8 � 0.0b

Glucose 1.3 � 0.0c 0.9 � 0.0d 0.9 � 0.0d 3.6 � 0.0ab 3.7 � 0.1a 3.5 � 0.0b

Sucrose 3.1 � 0.0c 3.9 � 0.0b 5.4 � 0.0a 2.0 � 0.1f 2.7 � 0.0d 2.4 � 0.1e

Total sugars 6.4 � 0.0e 6.4 � 0.0e 7.8 � 0.0d 10 � 0.0c 11 � 0.1a 11 � 0.0b

Citric acid 2.6 � 0.1b 2.2 � 0.0bc 2.1 � 0.0c 5.1 � 0.1a 4.9 � 0.2a 4.9 � 0.0a

Succinic acid 0.8 � 0.0d 0.9 � 0.0cd 0.9 � 0.0c 2.6 � 0.0a 2.4 � 0.0b 2.4 � 0.0b

Acetic acid 0.8 � 0.0b 0.6 � 0.0c 0.6 � 0.0c 1.1 � 0.1a 1.0 � 0.0ab 1.1 � 0.1a

Total organic acids 4.2 � 0.1b 3.7 � 0.0c 3.6 � 0.0c 8.8 � 0.2a 8.3 � 0.2a 8.3 � 0.1a

Sugars/acids 1.5 � 0.0c 1.7 � 0.0b 2.1 � 0.0a 1.2 � 0.0e 1.4 � 0.0d 1.3 � 0.0d

The proportions of essential amino acids of the daily requirement (g/100 g of product/70 kg body weight)Met 16 16 21 14 16 21Leu 19 22 25 16 18 21Ileu 20 21 25 17 18 22Lys 5.9 8.6 15 7.2 7.8 9.6Phe þ Tyr 32 31 39 26 29 30Val 18 19 21 14 17 20Thr 28 29 33 22 26 28Cys 85 100 130 72 100 99His 21 23 28 17 20 22Try 35 44 47 32 36 39

Means and standard deviations of 2 replicates from protein and crude fat, 2e3 replicates from moisture and ash, 6 replicates from NDF-fiber and ADF-fiber, 2 replicates fromsugars and acids, no replicates from essential amino acids. Energy and carbohydrates were defined computationally. Total fiber was defined computationally by enzymaticallydefined fiber (2 replicates) - crude protein (Kjeldahl method, 2 replicates). Significant differences between extrudates based on one-way ANOVA with Tukey’s HSD test(p < 0.05) are marked with aef.

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627624

Fig. 4. PLS regression plot of the interactions between physico-chemical properties and liking ratings. Liking ratings are presented in red, physical properties in blue, processparameters in blue italics, chemical properties in green and extrudates in brown font. Abbreviations of process parameters, liking ratings, physical and chemical properties refer toTables 1e4.

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627 625

extrudate is consumed (for 70 kg body weight). The limitingessential amino acid was lysine, which is the typical limiting factorin diet proteins. Approximately 20% of the daily requirement ofbranched-chain amino acids is achieved when 100 g of snacks areconsumed (Joint WHO/FAO/UNU Expert Consultation, 2007). ERsnacks had somewhat higher contents of essential amino acidscompared to NR extrudates, which is in agreement with the higherprotein content of ER-based extrudates. Oat snacks, either as branor flour, had higher essential amino acid quality.

3.5. Properties contributing to hedonic responses

A PLS-regression model was applied to distinguish the physico-chemical properties contributing to the liking of extrudates. InFig. 4., 84% of the physical and chemical variables explained 98%(R2 ¼ 0.987; validated R2 ¼ 0.947) of the variation within the likingdata, with two factors. NR-based snacks were more preferred thanER extrudates. Barley- or oat-flour-based snacks were more likedwithin press residue thanwere oat-bran-based snacks. The liking offlavor correlated strongly with glucose and fructose content alongwith organic acids, succinic acid and citric acid content. On theother hand, sucrose and the sugar-to-acid ratio correlated nega-tively with the liking of flavor. A high sugar-to-acid ratio in berriesoften correlates positively with sweetness and negatively withsourness (Laaksonen, Mäkilä, et al., 2013; Laaksonen et al., 2012).However, sourness can be a positive feature to some extent inberry-related products (Laaksonen, Ahola, & Sandell, 2013). Theredness of color (a*) due to a low pH, and a low length-to-heightratio (roundness in shape) correlated positively with appearance,contributing to a berry-like first impression.

The liked texture of the extrudates was comprised of highexpansion, low density and hardness (Fig. 4.), which is also inaccordance with results published by Meng et al. (2010). Thesecharacteristics were obtained with lower total mass flow, screwspeed and torque, barrel temperature and bearing pressure butwith higher specific mechanical energy (SME) values. The lowerSME values of extrudates may indicate the lipid migration outsidethe dough and acting as a lubricant in the barrel during extrusion(De Pilli, Carbone, Derossi, Fiore, & Severini, 2008). Higher contents

of protein, crude fat, energy, sucrose, neutral detergent fiber (NDF-fiber) and hemicellulose correlated negatively with expansion andthe liking of texture (Fig. 4.) which was in accordance with thefindings of Ya�gcı et al. (2014). A challenge is to obtain the desiredtextural qualities in a snack which has high contents of fiber, pro-tein and lipids, but which is low in starch. Especially the high lipidcontent in the dough leads to formation of amyloseelipid com-plexes which prevent the starch from gelatinization and thusdecrease the expansion in extrusion. A solution would be to addenzymes or emulsifiers into the dough. The enzymes make thedough more preferable for expansion by degrading the starchmacromolecules (De Pilli, Legrand, Giuliani, Derossi, & Severini,2009) and by breaking down the gluten proteins (Ishida &Nagasaki, 1989). The emulsifiers encapsulate the oil and preventthe collapse of the air bubbles during expansion (Richardson,Bergenstahl, Langton, Stading, & Hermansson, 2004). The result-ing extruded products would have increased porosity anddecreased breaking strength with enhanced textural qualities (DePilli, Carbone, Fiore, & Severini, 2007; De Pilli, Severini, Carbone,Giuliani, & Derossi, 2004). Further, the possible degradation ofthe vulnerable bioactive berry compounds should be taken intoaccount, when optimizing the processing conditions.

4. Conclusions

This study presents alternatives for sustainable exploitation ofby-products from blackcurrant juice production in extruded snacks.The processes may be further developed for producing a range offoods; from ready-to-eat snacks to breakfast cereals, in larger scale.Further development is necessary also to improve the sensoryproperties of the products and acceptance by different groups ofconsumers. For the consumers, it is important to maintain theberry-like characteristics, such as fresh berry taste and color inproduct development process. As shown in this study, the exploi-tation of non-enzymatically processed press residue enabled toremain the characteristics in our extrusion product. The conven-tional industrial press residue as such may lack the wanted flavorunique to berry material. The characteristics could be restored byadding original juice or other berry fractions in the extrusion

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627626

process. All in all, the study demonstrates new opportunities toexploit by-products from various berry and fruit processes to meetthe growing demand for healthier ready-to-eat products.

Acknowledgments

We thank Mrs Ute Helms, Department of Nutritional Physiology,Institute of Nutrition, Friedrich Schiller University of Jena, Germany,for the work in laboratory with the fiber and amino acid analyses.We thank also MTT (Piikkiö, Finland), for drying the press residues.This study was funded by Tekes e The Finnish Funding Agency forTechnology and Innovation together with Finnish food companiesas partners of the project “Black Currant as Unique Source ofFunctional Ingredients of Food: Novel Processes and Innovations”.

References

Altan, A., & Maskan, M. (2012). Advances in food extrusion technology. InM. Maskan, & A. Altan (Eds.), Contemporary food engineering series (pp. 132e155). North West: CRC Press Taylor & Francis Group.

Altan, A., McCarthy, K. L., & Maskan, M. (2008a). Evaluation of snack foods frombarleyetomato pomace blends by extrusion processing. Journal of Food Engi-neering, 84, 231e242.

Altan, A., McCarthy, K. L., & Maskan, M. (2008b). Twin-screw extrusion of barleyegrape pomace blends: extrudate characteristics and determination of optimumprocessing conditions. Journal of Food Engineering, 89, 24e32.

Alvarez-Martinez, L., Kondury, K. P., & Harper, J. M. (1988). A general model forexpansion of extruded products. Journal of Food Science, 53, 609e615.

AOAC. (1980). Official methods of analysis of the AOAC (13th ed.). Washington: As-sociation of Official Analytical Chemists.

AOAC. (1990). Official methods of analysis of the AOAC (15th ed.). Washington: As-sociation of Official Analytical Chemists.

AOAC International. (1995). Official methods of analysis of AOAC international (16thed.). Washington: Association of Official Analytical Chemists.

Ayala-Zavala, J. F., Vega-Vega, V., Rosas-Domínguez, C., Palafox-Carlos, H., Villa-Rodriquez, J. A., Wasim Siddiqui, Md, et al. (2011). Agro-industrial potential ofexotic fruit byproducts as a source of food additives. Food Research International,44, 1866e1874.

Beer, M. U., Arrigoni, E., & Amado, R. (1996). Extraction of oat gum from oat bran:effects of process on yield, molecular weight distribution, viscosity and(1/ 3)(1/ 4)-b-D-glucan content of the gum. Cereal Chemistry, 73, 58e62.

Bhattacharya, M., & Hanna, M. A. (1988). Effect of lipids on the properties ofextruded products. Journal of Food Science, 53, 1230e1231.

Bhatty, R. S. (1993). Extraction and enrichment (1 leads to 3),(1 leads to 4)-beta-D-glucan from barley and oat brans. Cereal Chemistry, 70, 73e77.

Bisharat, G. I., Oikonomopoulou, V. P., Panagiotou, N. M., Krokida, M. K., &Maroulis, Z. B. (2013). Effect of extrusion conditions on the structural propertiesof corn extrudates enriched with dehydrated vegetables. Food Research Inter-national, 53, 1e14.

Brennan, C., Brennan, M., Derbyshire, E., & Tiwari, B. K. (2011). Effects of extrusionon the polyphenols, vitamins and antioxidant activity of foods. Trends in FoodScience & Technology, 22, 570e575.

Brennan, M., Derbyshire, E., Tiwari, B. K., & Brennan, C. (2013). Ready-to-eat snackproducts: the role of extrusion technology in developing consumer acceptableand nutritious snacks. International Journal of Food Science and Technology, 48,893e902.

Carvalho, C. W. P., & Mitchell, J. R. (2000). Effect of sugar on the extrusion of maizegrits and wheat flour. International Journal of Food Science and Technology, 35,569e576.

Colonna, P., Doublier, J. L., Melcion, J. P., de Monredon, F., & Mercier, C. (1984).Extrusion cooking and drum drying of wheat starch. I. Physical and macro-molecular modifications. Cereal Chemistry, 61, 538e543.

Dansby, M. Y., & Bovell-Benjamin, A. C. (2003). Physical properties and sixthgraders’ acceptance of an extruded ready-to-eat sweetpotato breakfast cereal.Sensory and Nutritive Qualities of Food, 68, 2607e2612.

De Pilli, T., Carbone, B. F., Derossi, A., Fiore, A. G., & Severini, C. (2008). Effects ofoperating conditions on oil loss and structure of almond snacks. InternationalJournal of Food Science and Technology, 43, 430e439.

De Pilli, T., Carbone, B. F., Fiore, A. G., & Severini, C. (2007). Effect of some emulsifierson the structure of extrudates with high content of fat. Journal of Food Engi-neering, 79, 1351e1358.

De Pilli, T., Legrand, J., Giuliani, R., Derossi, A., & Severini, C. (2009). Effect of pro-cessing variables and enzymatic activity on wheat flour dough extruded underdifferent operating conditions. Food Technology and Biotechnology, 47, 404e412.

De Pilli, T., Severini, C., Carbone, B. F., Giuliani, R., & Derossi, A. (2004). Improvingfatty extrudate structure with amylase and protease. Journal of Food Biochem-istry, 28, 387e403.

Ding, Q. B., Ainsworth, P., Tucker, G., & Marson, H. (2005). The effect of extrusionconditions on the physicochemical properties and sensory characteristics ofrice-based expanded snacks. Journal of Food Engineering, 66, 283e289.

Faraj, A., Vasanthan, T., & Hoover, R. (2004). The effect of extrusion cooking onresistant starch formation in waxy and regular barley flours. Food Research In-ternational, 37, 517e525.

Faubion, J. M., Hoseney, R. C., & Seib, P. A. (1982). Functionality of grain componentsin extrusion. Cereal Foods World, 27, 212e215.

Finnish Food Safety Authority Evira. (2012). Guidelines for controlling food labeling(instruction 17030/1). Product safety unit Evira, guide/Rev.17055/1/en http://www.evira.fi/files/attachments/en/evira/forms_and_instructions/food/pakkausmerkintojep_valvontaohje_versio_4_1_en.pdf Accessed 24.04.13.

Hearty, A. P., McCarthy, S. N., Kearney, J. M., & Gibney, M. J. (2007). Relationship be-tween attitudes towards healthy eating and dietary behaviour, lifestyle and de-mographic factors in a representative sample of Irish adults. Appetite, 48, 1e11.

Ishida, K., & Nagasaki, M. (1989). Effects of proteases on textural properties of wheatflour dough. Journal of Japan Society of Nutrition and Food Sciences, 36, 1003e1008.

Jackson, J. C., Bourne, M. C., & Barnard, J. (1996). Optimization of blanching forcrispness of banana chips using response surface methodology. Journal of FoodScience, 61, 165e166.

Johansson, A., Laakso, P., & Kallio, H. (1997). Characterization of seed oils of wild,edible Finnish berries. Zeitschrift für Lebensmittel-untersuchung und eForschung,204, 300e307.

Joint WHO/FAO/UNU Expert Consultation. (2007). Protein and amino acid re-quirements in human nutrition. In World health organization technical reportseries, 935 (pp. 1e265).

Kearney, M., Kearney, J. M., Dunne, A., & Gibney, M. J. (2000). Sociodemographicdeterminants of perceived influences on food choice in a nationally represen-tative sample of Irish adults. Public Health Nutrition, 3, 219e226.

Khanal, R. C., Howard, L. R., Brownmiller, C. R., & Prior, R. L. (2009). Influence ofextrusion processing on procyanidin composition and total anthocyanin con-tents of blueberry pomace. Journal of Food Science, 74, 52e58.

Kirjoranta, S., Solala, K., Suuronen, J.-P., Penttilä, P., Peura, M., Serimaa, R., et al.(2012). Effects of process variables and addition of polydextrose and wheyprotein isolate on the properties of barley extrudates. International Journal ofFood Science & Technology, 47, 1165e1175.

Laaksonen, O., Ahola, J., & Sandell, M. (2013). Explaining and predicting individuallyexperienced liking of berry fractions by the hTAS2R38 taste receptor genotype.Appetite, 61, 85e96.

Laaksonen, O., Mäkilä, L., Tahvonen, R., Kallio, H., & Yang, B. (2013). Sensory qualityand compositional characteristics of blackcurrant juices produced by differentprocesses. Food Chemistry, 138, 2421e2429.

Laaksonen, O., Sandell, M., Nordlund, E., Heiniö, R.-L., Malinen, H.-L., Jaakkola, M.,et al. (2012). The effect of enzymatic treatment on blackcurrant (Ribes nigrum)juice flavour and its stability. Food Chemistry, 130, 31e41.

Landry, J., Delhaye, S., & Jones, D. G. (1992). Determination of tryptophan in feed-stuffs: comparison of two methods of hydrolysis prior to HPLC analysis. Journalof the Science of Food and Agriculture, 58, 439e441.

Linnamaa, P., Savolainen, J., Koulu, L., Tuomasjukka, S., Kallio, H., Yang, B., et al.(2010). Blackcurrant seed oil for prevention of atopic dermatitis in newborns: arandomized, double-blind, placebo-controlled trial. Clinical & Experimental Al-lergy, 40, 1247e1255.

Ma, H., Pan, Z., Li, B., Atungulu, G. G., Olson, D. A., Wall, M. M., et al. (2012). Prop-erties of extruded expandable breadfruit products. LWT e Food Science andTechnology, 46, 326e334.

Meng, X., Threinen, D., Hansen, M., & Driedger, D. (2010). Effects of extrusionconditions on system parameters and physical properties of a chickpea flour-based snack. Food Research International, 43, 650e658.

Menis, M. E. C., Milani, T. M. G., Jordano, A., Boscolo, M., & Conti-Silva, A. C. (2013).Extrusion of flavored corn grits: structural characteristics, volatile compoundsretention and sensory acceptability. LWT e Food Science and Technology, 54,434e439.

Nordic committee on food analysis No 148. (1993). Fructose, glucose and saccharose.In Liquid chromatographic determination in fruit and vegetable products. http://www.nmkl.org/ABONNEMENT/METPDF/148fi.PDF Accessed 24.04.13.

Obatolu, V. A., Skonberg, D. I., Camire, M. E., & Dougherty, M. P. (2005). Effect ofmoisture content and screw speed on the physical chemical properties of anextruded crab-based snack. Food Science and Technology International,11, 121e127.

O’Shea, N., Arendt, E. K., & Gallagher, E. (2012). Dietary fibre and phytochemicalcharacteristics of fruit and vegetable by-products and their recent applicationsas novel ingredients in food products. Innovative Food Science and EmergingTechnologies, 16, 1e10.

Patil, R. T., Berrios, J. de J., Tang, J., & Swanson, B. B. (2007). Evaluation of methodsfor expansion properties of legume extrudates. Applied Engineering in Agricul-ture, 23, 777e783.

Phenomenex Applications. (2013). HPLC application organic acid std ID No.14749.http://www.phenomenex.com/Application/Detail/14749 Accessed 30.4.13.

Potter, R., Stojceska, V., & Plunkett, A. (2013). The use of fruit powders in extrudedsnacks suitable for Children’s diets. LWT e Food Science and Technology, 51, 537e544.

Richardson, G., Bergenstahl, B., Langton, M., Stading, M., & Hermansson, A. M.(2004). The function of a-crystalline emulsifiers on expanding foam surfaces.Food Hydrocolloid, 18, 655e663.

Roininen, K., Lähteenmäki, L., & Tuorila, H. (1999). Quantification of consumer at-titudes to health and hedonic characteristics of foods. Appetite, 33, 71e88.

Sandell, M., Laaksonen, O., Järvinen, R., Rostiala, N., Pohjanheimo, T., Tiitinen, K.,et al. (2009). Orosensory profiles and chemical composition of black currant

L. Mäkilä et al. / LWT - Food Science and Technology 57 (2014) 618e627 627

(Ribes nigrum) juice and fractions of press residue. Journal of Agricultural andFood Chemistry, 57, 3718e3728.

Stojceska, V., Ainsworth, P., Plunkett, A., & _Ibano�glu, S. (2009). The effect of extru-sion cooking using different water feed rates on the quality of ready-to-eatsnacks made from food by-products. Food Chemistry, 114, 226e232.

Tahvonen, R., Hietanen, A., Sankelo, T., Korteniemi, V.-M., Laakso, P., & Kallio, H.(1998). Black currant seeds as a nutrient source in breakfast cereals producedby extrusion cooking. Zeitschrift für Lebensmittel-Untersuchung und eForschung,206, 360e363.

Truong, V., Bhandari, B. R., Howes, T., & Adhikari, B. (2004). Glass transitionbehaviour of fructose. International Journal of Food Science and Technology, 39,569e578.

Van Soest, P. J., Robertson, J. B., & Lewis, B. A. (1991). Methods for dietary fiber,neutral detergent fiber, and nonstarch polysaccharides in relation to animalnutrition. Journal of Dairy Science, 74, 3583e3597.

Viuda-Martos, M., Ruiz-Navajas, Y., Fernándes-López, J., & Pérez-Álvarez, J. A. (2010).Effect of orange dietary fibre, oregano essential oil and packaging conditions onshelf-life of bologna sausages. Food Control, 21, 436e443.

Voutila, L., Perero, J., Ruusunen, M., Jouppila, K., & Puolanne, E. (2009). Muscle fiberproperties and thermal stability of intramuscular connective tissue in porcineM. semimembranosus. Journal of the Science of Food andAgriculture, 89, 2527e2534.

Wijngaard, H., Hossain, M. B., Rai, D. K., & Brunton, N. (2012). Techniques to extractbioactive compounds from food by-products of plant origin. Food Research In-ternational, 46, 505e513.

Yanniotis, S., Petraki, A., & Soumpasi, E. (2007). Effect of pectin and wheat fibers onquality attributes of extruded cornstarch. Journal of Food Engineering, 80, 594e599.

Ya�gcı, S., Gö�güs, F., Jouppila, K., Talja, R., Hyvönen, L., & Kristbergsson, K. (2014).Extrusion technology. In K. Kristbergsson, & S. Ötles (Eds.), Food processing: Vol.8. ISEKI food series (pp. 1e15). Springer. in press.