Post on 28-Oct-2021
ISSN 0355-1180
UNIVERSITY OF HELSINKI
Department of Food and Nutrition
EKT Series 1860
DEGRADATION OF OAT LIPIDS: IMPACT ON THEFUNCTIONALITY OF β-GLUCAN RICH INGREDIENTS IN
ACID MILK GELS
Hanna Mahne
Helsinki 2018
ABSTRACTS
HELSINGIN YLIOPISTO ¾ HELSINGFORS UNIVERSITET ¾ UNIVERSITY OF HELSINKI
Tiedekunta/Osasto ¾ Fakultet/Sektion ¾ Faculty
Faculty of Agriculture and ForestryLaitos ¾ Institution ¾ Department
Department of Food and NutritionTekijä ¾ Författare ¾ Author
Hanna MahneTyön nimi ¾ Arbetets titel ¾ Title
Degradation of oat lipids: impact on the functionality of β-glucan rich ingredients in acidmilk gelsOppiaine ¾Läroämne ¾ Subject
Food ChemistryTyön laji ¾ Arbetets art ¾ Level
Master’s thesisAika ¾ Datum ¾ Month and year
October 2018Sivumäärä ¾ Sidoantal ¾ Number of pages
73Tiivistelmä ¾ Referat ¾ Abstract
The present literature was reviewed to understand the recent indications of lipid degradationaffecting β-glucan molecular weight in aqueous solutions, the main topics being oat lipidreactions, oat β-glucan, and oat bran/β-glucan fortified yogurts, as well as the possibleinterplay of these three. A non-heat treated, a heat treated, and a heat treated and defattedoat bran concentrate (OBC) were produced, all with ~20% β-glucan. OBCs were subjectedto a storage trial in dark for 0–12 weeks at 4, 22 and 40 °C. Lipid degradation was followedduring storage by assessing volatile compounds (HS-SPME-GC-MS), neutral lipid classes(NP-HPLC-ELSD) and tocopherol and tocotrienol levels (NP-HPLC-FLD) as well asβ-glucan molecular weights (HP-SEC). Acid milk gels were prepared as a model system foryogurts with added OBCs, and analysed for spontaneous and forced syneresis and texturalchanges. Lipid oxidation was observed in all OBCs, and hydrolysis in the non-heat treatedOBC, but no changes in β-glucan molecular weight were observed in the dry stored OBCs.OBC addition caused structural changes in an acid milk gels, the water retaining propertieswere improved, and the gels were softer and smoother than without added OBCs. Time andtemperature affected OBC lipid oxidation and hydrolysis behaviours. Adding OBCs to gelschanged gel textures compared to both without vs. with OBCs and adding fresh vs. storedOBCs.Avainsanat ¾ Nyckelord ¾ Keywords
Beta-glucan, oat bran concentrate, oat lipids, lipid degradation, yogurts, acid milk gelsSäilytyspaikka ¾ Förvaringsställe ¾ Where deposited
The Digital Repository of University of Helsinki, HeldaMuita tietoja ¾ Övriga uppgifter ¾ Further information
EKT Series 1860
HELSINGIN YLIOPISTO ¾ HELSINGFORS UNIVERSITET ¾ UNIVERSITY OF HELSINKI
Tiedekunta/Osasto ¾ Fakultet/Sektion ¾ Faculty
Maatalous-metsätieteellinen tiedekuntaLaitos ¾ Institution ¾ Department
Elintarvike- ja ravitsemustieteiden osastoTekijä ¾ Författare ¾ Author
Hanna MahneTyön nimi ¾ Arbetets titel ¾ Title
Kauran lipidien hajoamisen vaikutus β-glukaanipitoisten raaka-aineidenfunktionaalisuuteen hapanmaitogeeleissäOppiaine ¾Läroämne ¾ Subject
ElintarvikekemiaTyön laji ¾ Arbetets art ¾ Level
MaisterintutkielmaAika ¾ Datum ¾ Month and year
Lokakuu 2018Sivumäärä ¾ Sidoantal ¾ Number of pages
73
Tiivistelmä ¾ Referat ¾ Abstract
Kirjallisuuskatsaus rakennettiin työkaluksi ymmärtämään viimeaikaisia tutkimustuloksiarasvojen hajoamisen vaikutuksista β-glukaanin molekyylipainoon vesipitoisessamateriaalissa. Kirjallisuuskatsauksen pääaiheet kattoivat kauran lipidien reaktiot, kauranβ-glukaanin ja kaurakuiduilla rikastetut jogurttimallit. Kolme kauralesekonsentraattiavalmistettiin noin 20 %:n β-glukaanipitoisuuksilla: kuumennuskäsittelemätön, kuumennus-käsitelty ekä kuumennuskäsitelty ja rasvauutettu. Lesekonsentraatteja säilytettiin pimeässä0–12 viikkoa 4, 22 ja 40 °C:ssa. Lipidien hajoamista seurattiin määrittämällä muodostuvathaihtuvat yhdisteet (HS-SPME-GC-MS), neutraalilipidiluokat (NP-HPLC-ELSD),tokoferoli- ja tokotrienolipitoisuudet (NP-HPLC-FLD) sekä β-glukaanin molekyylipainot(HP-SEC). Kauralesekonsentraatteja käytettiin jogurtteja mallintavien hapanmaitogeelienainesosina. Geeleistä analysoitiin spontaani ja pakotettu synereesi sekä rakenteellisetmuutokset. Kaikissa kauralesekonsentraateissa havaittiin lipidien hapettumista jakuumennuskäsittelemättömässä kauranäytteessä lipidien hydrolysoitumista. β-glukaaninmolekyylipaino ei kuitenkaan muuttunut kuivissa kauralesekonsentraateissa.Kauralesekonsentraattien käyttäminen hapanmaitogeeleissä aiheutti rakenteellisiamuutoksia. Kauralesekonsentraattien lisäys paransi geelien veden sitomiskykyähuomattavasti ja muutti niiden rakennetta pehmeämmiksi ja sileämmiksi verrattuna ilmanlisättyä kuitua valmistetuttuihin kontrolligeeleihin. Aika ja lämpötila vaikuttivatkauralesekonsentraattien lipidien hapettumiseen ja hydrolysoitumiseen. Kauralese-konsentraattien lisääminen muutti geelien koostumusta ja rakennetta kontrolligeeleihinverrattuna. Erot geelien rakenteissa näkyi myös tuoreen ja säilötyn lesekonsentraatti-lisäysten välillä.Avainsanat ¾ Nyckelord ¾ Keywords
Beetaglukaani, kauralesekonsentraatti, kauran lipidit, lipidien hajoaminen, jogurtit,hapanmaitogeelitSäilytyspaikka ¾ Förvaringsställe ¾ Where deposited
Helsingin yliopiston digitaalinen arkisto, HeldaMuita tietoja ¾ Övriga uppgifter ¾ Further information
EKT-sarja 1860
PREFACE
This Master’s thesis was completed under the Business Finland funded project
OATyourGUT. The experimental work, writing, and instruction were performed as a
collaboration between VTT Technical Research Centre of Finland Ltd (VTT) and University
of Helsinki (UH), from December 2017 to early October 2018. The professor in charge was
Vieno Piironen from UH, and other instructors included Anna-Maija Lampi (UH), and
researchers Natalia Rosa-Sibakov and Martina Lille (VTT). A great big thank you to all my
instructors, for your flexibility and your expertise, we got the puzzle done! I am also grateful
to all the other team members at VTT and UH that helped and contributed to the work. Anni,
you know. Truly invaluable, lunches an all, #graduystävä. Lastly, my fambam and Jaakko.
Thank you for always believing in me, pushing me forward and helping me rest in between.
Hanna Mahne
Helsinki, 12.10.2018
LIST OF ABBREVIATIONS
DAG Diacylglycerol
ELSD Evaporative light scanning detector
FA Fatty acid
FFA Free fatty acid
FLD Fluorescence detector
GDL Glucono delta-lactone
H Non-heat treated OBC sample
HPLC High performance liquid chromatography
HS-SPME-GC-MS Headspace solid phase microextraction gas chromatographycoupled with mass spectrometry
L Heat treated, defatted OBC sample
M Heat treated OBC sample
MUFA Monounsaturated fatty acid
MW Molecular weight
OBC Oat bran concentrate
PUFA Polyunsaturated fatty acid
ROOH (Lipid) hydroperoxide
SC-CO2 Supercritical carbon dioxide
SFA Saturated fatty acid
TA Texture analyser
TAG Triacylglycerol
Tocols Tocopherols and tocotrienols
WHC Water holding capacity
TABLE OF CONTENTS
ABSTRACTS
PREFACE
LIST OF ABBREVIATIONS
1 INTRODUCTION
2 LITERATURE REVIEW 10
2.1 Oat bran, β-glucan, and its enrichment 102.1.1 Oat bran and β-glucan 102.1.2 Dry-fractionation for β-glucan enriched ingredients 13
2.2 Oat lipids 142.2.1 Lipid composition of oat bran 142.2.2 Antioxidants in oat bran 162.2.3 Lipid reactions in oats 18
Chemical oxidationEnzymatic hydrolysis and oxidationLipid reactions during oat processing and storageCo-oxidation of β-glucan
2.3 Yogurts 252.3.1 Milk proteins 252.3.2 Acid milk gels 272.3.3 Fibre fortified yogurts 29
3 EXPERIMENTAL RESEARCH 31
3.1 Background and aims 32
3.2 Overview of the storage trial and yogurt food model experiments 33
3.3 Test materials 343.3.1 Oat bran concentrates (OBCs) 343.3.2 Reagents and reference materials 35
3.4 Methods 363.4.1 Extraction of lipids and analysis of total FA content 363.4.2 Analysis of neutral lipid profiles with NP-HPLC-ELSD 383.4.3 Analysis of tocopherol and tocotrienol contents with NP-HPLC-FLD 393.4.4 Analysis of volatile compounds with HS-SPME-GC-MS 403.4.5 Preparation of acid milk gels 413.4.6 Analysis of spontaneous syneresis and water holding capacity 42
3.4.7 Analysis of gel firmness with TA 43
3.5 Results 433.5.1 OBC characteristics 433.5.2 Lipid degradation 44
Neutral lipidsTocopherols and tocotrienolsVolatile compounds
3.5.3 Properties of acid milk gels 51AcidificationSpontaneous syneresis and water holding capacity (WHC)Firmness of acid milk gels
3.6 Discussion 573.6.1 Lipid degradation behaviour of OBC samples 573.6.2 Effects of OBC addition on acid milk gel functionality 603.6.3 Limitations of study 62
4 CONCLUSIONS 64
REFERENCES 65
APPENDIX 73
Appendix 1. An exemplary graph for the analysis of gel firmness with TA 73
1 INTRODUCTION
It is widely recognized that a typical western diet does not provide enough dietary fibre.
According to the Nordic recommendations, 25–35 g, or approximately 3 g/MJ, of dietary
fibre should be consumed daily (NNR 2012; VRN 2014). Having insufficient fibre intake
coupled with the fact that the West is also plagued with cardiovascular disease and with other
diseases, for example many gastrointestinal disorders, that are all related to insufficient fibre
intake (see e.g. Makki et al. 2018), there seems to be a true need for nutraceuticals and easy-
to-use food products that conveniently add fibre to the Western diet. One interesting
ingredient used for functional foods is oats, and especially its fibre β-glucan.
Oats (Avena sativa) are a resilient crop still suiting Nordic growth conditions (Juhola et al.
2017). Oats have traditionally been used as high-energy feed material, however, the use of
oats as human food has been on a steady rise, especially in the Western world for the past
dozen years. According to the Natural Resources Institute Finland (2017), 92,000 tons of
oats was consumed as food in Finland during crop year 2016–2017, its highest amount for
oats ever recorded. Furthermore, of all cereals exported from Finland last crop year, oats had
the highest export volume of 326,000 tons (LUKE 2017). On their cereal’s balance sheet, it
is seen that a mere 9% of the harvested oats go toward food consumption and most are still
used for feed, indicating there is room to increase the food consumption portion.
The cell wall structures in oat aleurone and endosperm have 3–7% of β-glucan (Cui and
Wood 2000). β-glucan molecules are the main carbohydrates that form the familiar viscous
oat “slime” when preparing oat porridge, and they are also behind the healthy image of oats.
This healthy image has been scientifically supported, as oat β-glucan has three scientifically
proven, EFSA approved health claims: oat β-glucans attribute to (1) maintaining normal
blood cholesterol, (2) lowering blood cholesterol, and (3) reducing post-prandial glycaemic
responses (EFSA 2010, 2011). The beneficial effects are achieved when the daily
consumption of β-glucan reaches 3 grams.
Oats are typically considered as the fattiest cereal, with lipid contents ranging between 2 and
11% in oat grains, depending on the cultivar, growth environment, as well as extraction
method (Zhou et al. 1999a). Oat lipids contain a notable amount of nutritionally interesting
unsaturated fatty acids, such as oleic acid (C18:1) and linoleic acid (C18:2), each comprising
approximately 40% of the fatty acids in oats (Givens et al. 2004). When oats are processed
to flakes or flour, previously secured oat lipids are exposed to the environment, namely pro-
oxidants and oxygen, and it is likely for the lipid autoxidation chain reactions to start,
resulting in loss of lipids and formation of desired and un-desired oxidation products such
as free radicals or volatile compounds (Zhou et al. 1999b). In addition to oxidative reactions,
due to endogenous enzymes, oat lipids also degrade by hydrolysis (Yang et al. 2017). This
means that the possible lipid degradation changes the lipid composition and diminishes the
nutritional status of the oats, whilst forming undesired off-flavours and unpleasant aromas,
to begin with.
The effects of lipid degradation seem to also cause collateral damage, as discovered by Wang
et al. (2016): an oil-in-water emulsion with oxidized lipids induced β-glucan oxidative
degradation, which lead to diminished molecular weight of β-glucan and decreased the
viscosity of the solution. β-glucan amount and molecular weight seem to have significant
roles concerning the physiological actions of β-glucans in humans (Åman et al. 2004;
Andersson and Börjesdotter 2011). The degrading effect of oxidized lipids on β-glucan
might pose a problem in commercial functional foods with oat ingredients high in β-glucan,
as both the food matrix and the oat ingredients in themselves include potentially degradable
lipids. As this interaction between oxidized lipids and β-glucan molecular properties is a
fairly new finding, it is important to research the phenomenon more to understand it fully.
The previous studies have been made in oil-in-water emulsions, which makes it important
to investigate whether the same phenomenon happens in dry oat materials or in aqueous
foods containing oat. Degrading β-glucans in aqueous foods could propose technological
challenges as well as nutritional losses.
The object of this literature review was to gain knowledge and understanding of lipid
degradation in oats, and its impact on the functionality of oat ingredients, especially on the
physicochemical properties of β-glucan within the oat ingredients. The practical work was
designed as a three-month storage trial of β-glucan rich oat ingredients to experiment the
connection between lipid degradation and β-glucan functionality. Firstly, the lipid
degradation state in oat ingredients was analysed by chromatographic means at the
University of Helsinki. Secondly, the stored oat ingredients were used to prepare acid milk
gels, which modelled a yogurt-type semi-solid food, at VTT Technical Research Centre of
Finland. This information on oat lipids and β-glucan was crucial to obtain to be able to create
functional nutraceuticals with β-glucan to aid human health.
10
2 LITERATURE REVIEW
2.1 Oat bran, β-glucan, and its enrichment
2.1.1 Oat bran and β-glucan
According to AACC (1989), oat bran is “the food which is produced by grinding clean oat
groats or rolled oats and separating the resulting oat flour by sieving, bolting, and/or other
suitable means into fractions such that the oat bran fraction is not more than 50% of the
original starting material, and has a total β-glucan content of at least 5.5% (dry weight basis)
and a total dietary fibre content of at least 16.0% (dry weight basis), and such that at least
one-third of the total dietary fibre is soluble fibre.” Oat bran is thus a technical term used for
the outer layer milling fraction consisting of the coarse outer parts of the oat kernel: pericarp,
seed coat, aleurone, sub-aleurone, and starchy endosperm. In addition to starch (40–61%),
proteins (10–21%), and fat (8–12%), the bran is rich in minerals (2–4%) and cell wall
polysaccharides such as cellulose, arabinoxylan, and β-glucan (total dietary fibre 11–19%)
(Luhaloo et al. 1998). The main reason for high variations are the effects of oat variety,
agronomic as well as environmental factors (Givens et al. 2004). Additionally, there are
significant amounts of antioxidants present, such as tocopherols and tocotrienols, phytic
acid, avenanthramides and other phenolic compounds (Peterson 2001; Bryngelsson et al.
2002).
β-glucan is a water-soluble non-starch fibre component found in oat aleurone and endosperm
cell wall structures (Figure 1). It counts for approximately 3–11% (Cui and Wood 2000) of
the bran and can be described as a linear glucose homopolymer with β-D-glucopyranosyl
units (Figure 2a). Units are linked with β-(1à4) and β-(1à3) linkages at an approximate
ratio of 7:3. (1à3) linkages stand alone and are usually separated by either two (cellotriose
unit) or three (cellotetraose unit) (1à4) linkages, comprising a polysaccharide group called
mixed linkage β-glucans (Figure 2b).
11
Figure 1. Oat grain and its tissues, cell layers and nutrient distribution (e.g. β-glucan). Scheme by Grundy etal. (Grundy et al. 2018).
Figure 2. Oat β-glucan molecular structures: (a) A basic schematic structure of β-glucans as glucose polymers.(b) Approximate distribution of cellotrioses (one 3G1 and two 4G1s) and cellotetraoses (one 3G1 and three 4G1s)with occasional longer oligomers. 3G1 = (1à3) linkage, 4G1 = (1à4) linkage. Scheme b modified fromLazaridou and Biliaderis (Lazaridou and Biliaderis 2007).
The structure of the β-glucan molecule seems to be a critical aspect when considering its
bioaccessibility as well as its physicochemical properties and functionality, such as solubility
in water, viscosity and gelation properties (Lazaridou and Biliaderis 2007). It is believed that
the ability of β-glucans to form viscous solutions in the intestine helps to maintain normal
blood cholesterol, lower blood cholesterol, as well as reduce post-prandial glycaemic
responses (Regand et al. 2011; Whitehead et al. 2014). The scientifically proven beneficial
12
effect is achieved when the daily consumption of β-glucan reaches 3 grams (EFSA 2010,
2011). To use the EFSA approved health claims, the product must provide 1 g of β-glucan
per portion.
There has been much speculation on the mechanisms by which β-glucan positively affects
human health. In some studies, the desired effects have not been acquired even though the
consumption levels greatly exceeded the required daily amounts (Andersson and
Börjesdotter 2011; Whitehead et al. 2014). Two molecular aspects that seem to be crucial
for β-glucan’s physiological actions are the amount of solubilized β-glucan (in the intestine)
and its molecular weight (MW), both of which need to be sufficiently high (Åman et al.
2004; Andersson and Börjesdotter 2011). Furthermore, structural properties and MW greatly
affect the water solubility, viscosity and gelation properties of oat β-glucans (Cui and Wood
2000). The lack of physiological efficacy is most often connected to the β-glucan MW, and
it seems that medium to high MW β-glucan (530–2210 kDa) would be more efficient in, for
example, actively lowering serum cholesterol levels in humans than low MW β-glucan
(21 kDa) (Wolever et al. 2010). On the other hand, as discussed by Wang et al. (2017), low
MW polysaccharides have been thought to have higher mobility than larger β-glucans, and
thus have a greater antioxidant activity than high MW β-glucans as well as a better ability to
effectively improve the oxidative parameters, for example in the spleen (Błaszczyk et al.
2015). Also, technological properties of β-glucan are dependent on MW, and beverages, for
example, need to obtain low MW β-glucan to avoid aggregation and sedimentation (Sibakov
et al. 2013).
Whitehead et al. (2014) found that the solubility of β-glucans is affected by their
bioaccessibility, in other words by the food matrix. The food matrix may also affect the MW
of β-glucans as the food matrix and its modification (food processing) might induce
degradation of the β-glucan molecules resulting in lowered MW. The effect of commercial
processing and the source and type of extraction of β-glucan are considered in more detail,
for example, in the research of Keogh et al. (2003). Depending on the extraction method and
conditions, solvents, and sample history, the molecular weight of β-glucans can range
anywhere between 65 to over 3000 kDa (Lazaridou and Biliaderis 2007). Thus it should be
noted, as stated in the meta-analysis of Whitehead et al. (2014), that the lack of a
standardized method for β-glucan MW analysis makes it challenging to accomplish a
confident relation between the β-glucan MW and its functionality.
13
Other aspects effecting β-glucan molecular integrity are the presence of enzymes, heat,
mechanical forces, and acids as well as oxidative degradation by ascorbic acid and iron
cations, a pro-oxidant forming hydroxyl radicals that attack β-glucan (Kivelä et al. 2009a,
2009b, 2012; Mäkelä et al. 2015). However, there are differences on degradation behaviour
of β-glucans depending on their sources and levels of residual phytates: oat β-glucan has
been found to have a higher iron binding capacity, quenching and retarding the iron cations
as pro-oxidants in the formation of hydroxyl radicals through the Fenton reaction (Wang et
al. 2017). Therefore, when designing functional foods with β-glucan, it is important to
consider the processes of producing the β-glucan, the food matrix type (solid, semi-solid or
liquid), and the MW of the β-glucan in the final product, to make sure the product is
acceptable and the product’s role as a nutraceutical is not compromised.
2.1.2 Dry-fractionation for β-glucan enriched ingredients
Oat β-glucan can be isolated or enriched by dry milling, wet milling sieving, air classification
and solvent extraction (Tiwari and Cummins 2009). The process usually includes dehulling,
sieving, milling, more sieving and/or air classifying at varying repetitions together with
common heat treatments to remove part of the starch and obtain oat bran and finally β-glucan
enriched ingredients with varying particle sizes (Figure 3). Dry fractionation methods are
often more favoured than wet-fractionation methods, because they are more ecological and
cost-effective as there is no need for drying steps. Additionally, aspects favouring dry-
fractionation include much higher fraction mass yields than in wet-extraction, the ready
usability of all residual fractions, low microbiological risks in dry ingredients, and the fact
that the produced partially purified or enriched oat ingredients are pure enough for most food
products (Sibakov 2014). Dry milling concentration of β-glucan is also nutritionally
interesting, as the oat bran concentrate includes both soluble and insoluble material and thus
has combined beneficial physiological effects of the two dietary fibre classes (Lazaridou and
Biliaderis 2007). Patented dry-fractionation methods to produce oat fractions with high
β-glucan concentrations, have been designed for example by Mälkki et al. (2001). Separation
of oat grain cells can be improved by removing oat lipids with supercritical carbon dioxide
(SC-CO2) (Sibakov et al. 2011; Konak et al. 2014) and it has been found that this type of
processing also improves the foaming properties of oat-derived products or ingredients
(Kaukonen et al. 2011).
14
Figure 3. The structural levels when extracting β-glucan (Grundy et al. 2018).
Because dry-fractionation is a concentration method using only physical and technological
methods to modify β-glucan content, it should be noted that dry-fractioned ingredients are
hybrid ingredients that may contain proteins, lipids, minerals and other possible components,
such as phytic acid. These components will affect the behaviour of the β-glucan rich oat
ingredient in food applications, and should be considered when designing products with oat
bran concentrates.
2.2 Oat lipids
2.2.1 Lipid composition of oat bran
Oil storages in a grain are close to the living tissue, close to lipid degrading enzymes, so that
they can be used during germination (Heneen et al. 2009). Usually the germ of any cereal
grain is considered the main storage place containing these grain’s valuable energy bundles,
lipids. However, relative oat lipid concentration increases from the inner starchy endosperm
to the outer layers of the kernel (Miller and Fulcher 2011). In this way the aleurone layer has
the highest relative lipid concentration within the oat grain (Figure 4), thus making lipids a
significant constituent of oat bran.
15
Figure 4. Oil droplets (bright/yellow) in oat aleurone layer cells visualised by staining with Nile Blue in cryosections of oat grains. Microscopy image by courtesy of VTT Technical Research Centre of Finland Ltd.
Both lipid content and fatty acid composition of oats vary depending on the cultivar,
environment, and agronomic factors, as well as the extraction method of the lipids (Zhou et
al. 1999a; Givens et al. 2004). Luhaloo et al. (1998) explain that crude fat contents in
different oat brans range between 8 and 12%. Fatty acids are chiefly stored as
triacylglycerols (TAGs), and TAGs count for around 80% of all the lipid classes in oats.
TAGs consist of a glycerol and three fatty acids, and in oats, 80% of these fatty acids are
unsaturated. Zhou et al. (1999a) report that the lipid composition of oat brans varies
depending on processing conditions, but in oats in general, around 40% of the fatty acids are
oleic acid (C18:1) and another 40% are linoleic acid (C18:2). The rest of the fatty acids are
palmitic acid (C16:0, 15–20%) with smaller amounts of stearic acid (C18:0), α-linolenic acid
(C18:3), and myristic acid (C14:0). Concerning other lipid classes in oats, around 20% are
polar lipids (phospholipids and glycolipids) and the final percentages mainly sterol esters
(Zhou et al. 1999a). The high amount of unsaturated fatty acids in the oats, including an
essential fatty acid linoleic acid, add to the nutritional value of oats. The high lipid content
is also interesting from a technological aspect. The lipids can have a modifying effect on oat
starch pasting properties, which in turn affect oil stability and instability, and applications as
fat replacers can be made both in food and cosmetic industries (Zhou et al. 1999a).
16
Lipids in oat can be analysed as a whole fraction or as individual lipid classes. The pre-
treatments of the material, the oat fraction and the chosen method, from extraction to
characterisation and quantification, significantly affect the perceived total lipid content.
Most often, the method includes some kind of an extraction step. One of the most popular
extraction solvents used to be hexane, which gives most conservative estimations on the total
lipid content compared to e.g. 85% methanol or ethanol (Zhou et al. 1999a). The more polar
the extraction solvent is, the more it also extracts polar lipids and sterol esters. However, due
to its harmful nature, hexane is now successfully substituted with heptane in lipid analyses,
especially when mixed with more polar solvents, such as 2-propanol (Damerau et al. 2014).
As fatty acids count for most lipids in oats, sometimes it is sufficient enough to add all the
fatty acids together to represent lipid levels in the oat sample. If done so, the lipid content
should be expressed as the total sum of fatty acids, typically mg/g sample or as a percentage.
When extracting and analysing total fatty acids, the process usually includes acid hydrolysis,
saponification and methylation steps to form fatty acid methyl esters detectable by gas
chromatography (GC), as in the AOCS Ce 2c-11 method (Srigley and Mossoba 2017).
Instead of acid hydrolysis, acetone can also be used as an extraction solvent when used with
high enough temperature and pressure with an Accelerated Solvent Extractor (ASE) (Lampi
et al. 2015). Final quantification is most often performed using an internal standard, a fatty
acid not inherently found in the oat lipids.
2.2.2 Antioxidants in oat bran
In addition to lipid classes, the oat bran is also rich in lipid soluble antioxidants. Antioxidants
prevent or inhibit lipid oxidation, and the most abundant fat-soluble antioxidant in the oat
bran is α-tocotrienol (a-T3). Together with α-tocopherol (a-T), β-tocopherol (b-T), and
β-tocotrienol (b-T3), α-tocotrienol is one of four tocopherols and tocotrienols (tocols)
present in the oat bran (Bryngelsson et al. 2002). Tocols are phenolic compounds with
varying amounts and positions of methyl groups within their chromanol ring (Figure 5).
Tocopherols and tocotrienols also differ by the unsaturation level of their phytyl tails.
17
Figure 5. Tocopherol (a) and tocotrienol (b) molecular structures consisting of a chromanol ring and a phytyltail. a, b, g, d = α, β, γ, δ, respectively. T = tocopherol, T3 = tocotrienol.
When oat cell walls are broken during processing, tocols are exposed. This increases the
bioavailability of the antioxidants as well as the chances for their degradation (Bryngelsson
et al. 2002). The radical quenching antioxidant activity is based on the ability of tocols to
donate their phenolic hydrogens to lipid derived free-radicals, making the phenolic
chromanol ring the main location where lipid oxidation is inhibited. As a phenolic hydrogen
is donated to a lipid peroxyl radical, a resonance stabilized chromanoxyl radical is formed
(Kamal-Eldin and Appelqvist 1996). Degradation products of tocols include lactones and
quinones. Depending on the tocol, the environmental factors, as well as the food matrix,
dimers and trimers may also be formed. Because of their ability to act as free radical
scavengers, tocopherol and tocotrienol levels can be used to indirectly follow and assess
lipid oxidation states in oats.
Tocols are usually analysed by chromatographic methods. High performance liquid
chromatography (HPLC) offers the most advantageous methods, and Kamal-Eldin et al.
(2000) note that HPLC methods were fast, simple, sensitive, and selective, superior to GC
methods. The authors continued to explain that although both normal and reversed phase
columns can be used, and although reversed phase (RP) columns offer better stability and
longer durability, the normal phase (NP) columns are better able to separate β- and γ-isomers
of tocopherols and tocotrienols. Due to their heterocyclic nature, tocols can be detected based
on ultraviolet absorbance (UV detector) or fluorescence (fluorescence detector, FLD).
18
Besides tocols, additional oat bran antioxidants include phytic acid, avenanthramides and
other phenolic compounds (Peterson 2001; Bryngelsson et al. 2002). Avenanthramides have
been under vigorous research for the past decade or so, and they are unique for oats.
Avenanthramides are phenolic alkaloids and at least 25 avenanthramide types can be
identified, although most of the avenanthramides consist of three main types (Peterson 2001;
Martínez-Villaluenga and Peñas 2017). They are abundant in the bran fraction, although
measured concentrations depend on processing and analysis methods (Hitayezu et al. 2015).
Avenanthramide synthesis is enzyme regulated and these polyphenols are believed to be a
part of a plant defence mechanism (Peterson 2001). Avenanthramides are believed to greatly
add to the health benefits of oats. It is said that these polyphenols exhibit anti-inflammatory,
antiproliferative, and anti-itching activity, which may have bioactive effects alongside
β-glucans, such as contributing to protecting against coronary heart disease and colon cancer,
as well as calming skin irritations (Meydani 2009). As it is with tocols, avenanthramides
might degrade during heat and processing of oats (Bryngelsson et al. 2002), which could be
compensated by the possible release of bound avenanthramides during storage at high
temperatures (Kaur et al. 2018).
2.2.3 Lipid reactions in oats
Chemical oxidation
Lipid oxidation means predominantly the deterioration of double bond containing
unsaturated fatty acids (Schaich et al. 2013). Some of the main factors in affecting the overall
oxidation rate of lipids include the reaction environment, that is e.g. temperature,
the presence of inhibitors and catalysts, and the nature of the compounds involved (Kim and
Min 2008). As oats are high in unsaturated fatty acids, they are susceptible to lipid
degradation (Zhou et al. 1999a).
Chemical lipid oxidation can be described as a series of autocatalytic reactions (Reische et
al. 2008). In this autoxidation process, a free radical chain reaction is generated and
maintained as long as there are substrates and radicals present (Schaich et al. 2013). The
commonly accepted theory has been the free radical chain reaction theory (Ingold 1969;
Schaich et al. 2013). According to this theory, autoxidation can be divided into three main
phases: the initiation, propagation and termination stages (Ingold 1969). Together their
kinetic behaviour can be describes as a sigmoidal curve, where the initiation phase seems
19
long and slow, until the exponential propagation phase begins. The termination phase can be
seen as the final plateau of the curve, and slowing down of reactions.
Because normal oxygen is a triplet according to its energy state, an external initiator or a
catalyst is needed for the autoxidation to begin. Spin states of the reacting molecule parts
need to be corresponding for a reaction to occur, which is why the common triplet oxygen
cannot react spontaneously with unsaturated fatty acids that have singlet state double bonds.
Therefore, the start of the chain reaction, the initiation, most often requires an initiator or a
catalyst that can remove an electron and creates a radical molecule of one of the compounds.
Some of the most typical initiators include preformed or primary radicals, pro-oxidative
metals, enzymes, hemes, and pigments, like chlorophylls. Another considerable group of
initiators includes more physical factors, such as UV-radiation, visible light, and heat, which
all add energy to the lipid system (Shahidi and Wanasundara 2008). These often work
indirectly, for example visible light does not have the required energy to produce radicals
but it can transfer energy and affect oxidation through photosensitizers such as chlorophylls
(Schaich et al. 2013). After a radical has been formed, the radical reacts with a bisallylic
hydrogen of an unsaturated fatty acid and a lipid alkyl radical R• is formed, kick-starting the
free radical chain reaction (Ingold 1969, Kim and Min 2008; Schaich et al. 2013).
Propagation can be described as “the heart of lipid oxidation” (Ingold 1969). In
propagation, oxygen spontaneously converts the alkyl radical to a more reactive peroxyl
radical (ROO•). To quenc its instability, the lipid peroxyl radical will attack an adjacent fatty
acid, abstract a hydrogen, and thus turn into a lipid hydroperoxide (ROOH). In tandem,
another lipid alkyl radical R• is produced. As this process is looped, autocatalysis of lipid
oxidation is enabled (Shahidi and Wanasundara 2008; Schaich et al. 2013). Lipid oxidation
may be extended by excited oxygen atoms, singlet oxygens 1O2. Singlet oxygens are not
radicals, but they are able to react with non-radical, singlet state double bonds and form
hydroperoxides without the intermediate peroxyl radical, because of their high-energy state,
leading to chain branching and supercharging lipid oxidation even more (Kim and Min
2008).
Autocatalytic reaction chains are stopped only if (1) there are no more substrates (fatty acids)
or (2) two free radicals react together to form a more stable non-radical secondary product
(Schaich et al. 2013). Quenching of the radical reaction chains does happen throughout
autoxidation (Schaich et al. 2013). However, Schaich et al. (2013) continue to remark that
20
when the rate of quenching radicals is greater than the rate by which new free radical chains
are produced, the oxidation rate can and will slow down into the termination phase.
Termination has a cumulative effect as individual radicals react together and produce more
stable and non-radical products (Ingold 1969). For an overview on the classical chain
reactions during initiation, propagation, and termination, please refer to Figure 6.
Figure 6. The commonly accepted understanding of how the free radical chain reactions of lipids advance.R•=lipid alkyl radical, ROO•=peroxyl radical, ROOH=hydroperoxide, RR/ROOR=non-radical products.Scheme by Sahahidi and Wanasundara (2008).
The apparent loss of substrates (fatty acids) and possible deterioration on nutritional quality
are not the only changes the oil system undergoes. The increase of the primary oxidation
product, hydroperoxide, contributes to most of the changes in the oil phase. Hydroperoxides
are closely involved in the formation of various secondary products, since hydroperoxides
are transitory in nature and decompose further to other products (Shahidi and Wanasundara
2008). Hydroperoxides are colourless and odourless as such but the secondary products are
often associated with changes in flavour properties because of their highly volatile and
odour-active nature (Kim and Min 2008).
The secondary products are also usually product specific, which stems from the differences
of the fatty acids. Different fatty acids have different activation energies needed to start the
free radical chain reactions (Kim and Min 2008). Based on activation energies, reactions are
most probable in places where bond-energies are at their lowest. As explained by Kerr
(1966), fatty acids have these low bond-energy domains at their allylic C-H bonds.
Furthermore, Cosgrove et al. (1987) showed that the oxidizability of polyunsaturated fatty
acids (PUFA) is linearly dependent on the number of doubly allylic positions, which results
21
in easier abstractions of hydrogens in between double bonds: When a hydrogen is abstracted,
the alkyl is left with an imbalance of electrons. The pentadiene structure compensates the
unevenness by resonance stabilization. This resonance spreads the electrons on a larger area
and exposes the now electron deficient outside positions, which can result in further oxygen
addition and hydroperoxide formation (Schaich et al. 2013). In this manner, PUFA are more
susceptible to oxidation than monounsaturated fatty acids (MUFA) (Schaich et al. 2013).
Depending on the structure of the fatty acid, the creation of these enhanced hydrogen-targets
can enable chain branching and the formation of specific secondary products (Hyde and
Verdin 1968).
Alongside hydrogen abstraction, there is a plethora of other competing chemical reactions
producing secondary lipid oxidation products. These mechanisms include isomerization,
β-scission, and scission, cyclization, recombination and addition, producing more radicals
or polymers, aldehydes, alkanes, oxo compounds, epoxides, and ketones, to name a few
(Schaich et al. 2013). As oats are abundant in linoleic acid, one of the most typical
autoxidation products in oats is the volatile aldehyde hexanal (Zhou et al. 1999a; Lehtinen
and Laakso 2004). However, there are both primary oxidation products (fatty acid
hydroperoxides) as well as other secondary products, such as the aforementioned hexanal.
Enzymatic hydrolysis and oxidation
Lipids are a source of energy for the potentially growing oat seed, and lipid degrading
enzymes in oat are meant to release this energy for germination of the seed (Lascorz and
Drapron 1987). Oat lipids are close to the enzyme production sites that are related to
germination and mobilization (Heneen et al. 2009). In a developing grain, oil bodies from
the embryonic axis and the aleurone layer fuse and accumulate into the endosperm (Heneen
et al. 2008). During germination, enzymes synthetized and secreted from the aleurone and
scutellum tissues mobilize lipid reserves (TAGs) as transportable nutrients (FFAs), which
results in oil droplets accumulating close to the aleurone and scutellar epithelium (Lenova
et al. 2010). For this reason, it is likely for endogenous lipid degrading enzymes to be present
in oat fractions, especially milling fractions containing the aleurone layer.
The main lipid degrading enzymes in oats are lipase, lipoxygenase (LOX), and peroxidase
(POX), and at least lipase is known to reside close to the outer layers of the kernel, or the
bran (Ekstrand et al. 1992). Lipase hydrolyses the ester bound fatty acids from
triacylglycerols (TAGs) first partially into di- and monoacylglycerols, and then soon into
22
free fatty acids (FFAs), which are highly susceptible to autoxidation (Yang et al. 2017).
Accordingly, hydrolysis of oat lipids can be followed by the formation of FFA, which can
start immediately after the cell wall structures are broken in milling, promoting oxidation.
LOX oxidizes fatty acids with pentadiene structures and produces hydroperoxides. This
activity is held to be responsible especially for volatile secondary products contributing to
off-flavours, such as the volatile aldehyde, hexanal and 2-pentylfuran (Molteberg et al. 1996;
Heiniö et al. 2002). LOX activity has also potential to co-oxidate lipid soluble vitamins (e.g.
tocols) (Robinson et al. 1995), which prohibits the vitamins’ antioxidant activity and further
promotes oxidation. However, LOX activity in oats is found to be low, leaving the oxidation
of oat lipids to the non-enzymatic autoxidation (Yang et al. 2017). POX activity produces
fatty acid epoxides and hydroxy fatty acids by transferring oxygen intra- and
intermolecularly from fatty acid hydroperoxides to unsaturated fatty acids (Piazza et al.
2003). These non-volatile compounds are reported to contribute to the bitter taste of stale
oats (Biermann et al. 1980). To prevent enzymatic lipid degradation, enzymes are typically
inactivated by hydrothermal processes.
Lipid reactions during oat processing and storage
The first stage where unwanted lipid hydrolysis can occur, is during dehulling, as there is
the risk of kernel breakage (Doehlert and McMullen 2001). However, unless there is
excessive moisture, this is considered a minor problem with low correlations between kernel
breakage and lipid hydrolysis (Peltonen-Sainio et al. 2004). In milling and fractionation, cell
wall structures are broken, and both lipids and antioxidants are exposed form the protection
of the structures and rendering them to the effects of hydrolysis (if enzymes have not been
inactivated) and oxidation. As the oat bran is high in trace minerals, they too are exposed,
providing initiation sites for autoxidation. If fractionation is done in wet conditions, there is
a risk for lipolytic enzyme activity. However, this kind of enzyme activity can be suppressed
by using, for example, alkaline water (Liukkonen et al. 1993).
To inactivate enzymes, hydrothermal processing (steaming and kilning) is typically applied
to oat kernels. However, heat and added humidity induce degradation of antioxidants, such
as tocols (Bryngelsson et al. 2002), which in turn can promote autoxidative reactions. Heat
treatment has also been found to promote lipid oxidation: e.g. the more sufficient lipase
inactivation is, the more lipid oxidation related products are formed and observed (Lehtinen
et al. 2003). This is especially emphasised in long-term storage, which was the main focus
of the study by Head et al. (2011), in which commercially heated and superheated steam
23
treated oat groats were stored for up to 26 weeks. In their study, the hexanal amounts
increased substantially: more for commercially treated and less for superheated steam treated
oats. However, there were no changes in the colour or content of free fatty acids, indicating
little or no Maillard reaction or hydrolysis in either of the differently processed oat groats.
Extrusion has also been presented as a viable alternative to conventional heat treatments and
oat processing (Lampi et al. 2015). Parker et al. (2000) extruded oat flours in different
conditions and they concluded that extruded debranned oat flour had the highest amounts of
lipid oxidation related alkanals, 2-alkenals, and 2,4-alkedienals. Also, when the extrusion
temperature was at its highest and moisture content at its lowest, the levels of Maillard
reaction products were elevated, such as pyrazines, pyrroles, furans, and compounds
containing sulphur.
Processing also affects the flavour of oats. Initially the formed (volatile) flavour compounds
are considered positive and characteristic to oats. These can include raw (hay-like), oaty
(normal for oats) and toasted notes (derived from Maillard reaction) (Klensporf and Jeleń
2008). However, possible enzymatic hydrolysis and chemical oxidation products can soon
be experienced as bitter and rancid off-flavours, respectively. Hexanal is a typical volatile
lipid oxidation product that gives a grassy off-flavour and has a particularly low aroma
threshold (Heiniö et al. 2002). In fact, hexanal formation is one of the most abundant of the
prominent features of oat lipid oxidation. Interestingly though, nitrogen containing
compounds, such as amino acids, dipeptides, and polyamines, have been found to have
hexanal quenching abilities (Zhou and Decker 1999), which could add to the findings of
Lehto et al. (2003), who found aldehyde dehydrogenase type activity in oats when hexanal
was absorbed into the oat matrix and oxidized to hexanoic acid.
When oat products are stored for a prolonged time, both lipid hydrolysis and oxidation
reactions have time to branch out, yielding more and more degraded lipid-derived products
with high risk of co-oxidation (Schaich et al. 2013). Oat is a hybrid ingredient containing
other components such as proteins, fibres, starch and antioxidants in addition to its lipids.
These compounds will experience degradative reactions, such as co-oxidation, alongside
lipids during storage (Rakić et al. 2014). It is not always clear whether lipid reactions and
certain reactions of these other compounds have a cause and effect relationship. However, it
is very likely that the free radicals from lipid oxidation promote and cause the oxidation of
adjacent compounds (Schaich et al. 2013). Additionally, it is also very important to consider
what type of food is in question. An oat containing emulsion with high water activity and
24
lots of interfacial area will behave very differently compared to dry oat flour in storage. Even
though fine oat flour will have a lot of interfacial area with air, an emulsion will have
technological challenges (emulsion stability (Rousseau 2000)) and chemical challenges
(Laguerre et al. 2009). Time seems to be the one and most important factor in lipid stability
in storage. To slow down and control the inevitable degradation reactions during this time,
light, packaging materials, and storage conditions (temperature, moisture, and oxygen) have
the most considerable effects on lipid stability in stored oat products (Lehtinen and Laakso
2004; Kaur et al. 2018).
In conclusion, hydrolysis of oat lipids can be prevented by inactivating lipolytic oat enzymes
whereas chemical oxidation of oat products is most effectively prevented by controlling
storage temperature, light, and the nature of the container. A rancid oil that has gone through
all three stages of autoxidation is problematic in many ways. Not only can there be toxic
compounds such as epoxides present, but the sensory qualities of the oil suffer, resulting in
changes at least in odour and flavours, as discussed previously. The chemical changes in the
stored oat rise also technological and economical challenges as the functionality of the oat
product might change.
Co-oxidation of β-glucan
It was recently discovered by Wang et al. (2016) that in an emulsion, barley β-glucan
degradation was induced when oxidized lipids were present. The breakdown of β-glucan
molecules was observed as a significant decrease in both emulsion viscosity and β-glucan
molecular weight (MW). These findings raised a lot of questions, as the interactions of
oxidized lipids and β-glucan molecules have not yet been vigorously researched.
Oxidative degradation of β-glucans can happen in food processing and in aqueous food
models (Kivelä et al. 2009a, b, 2012). And as pointed out by Wang (2018), there are
indications of lipid co-oxidation of polysaccharides, where, for example, β-cyclodextrin was
degraded because of linoleate autoxidation and corn starch molecular characteristics
changed because of methyl linoleate. Even so, literature on β-glucan co-oxidation with lipids
is scarce. What can be said is that lipid oxidation induced degradation of β-glucan is most
probably linked to the termination phase of lipid oxidation, as is the case with many other
non-lipid molecules, such as proteins and carbohydrates are co-oxidated (Schaich et al.
2013). Wang (2018) suggests that although lipids are hydrophobic, lipid oxidation products
25
can affect the surrounding molecules through the interface, making lipid oxidation a source
of radicals, which would then degrade β-glucans.
2.3 Yogurts
2.3.1 Milk proteins
Milk is a complex colloidal system of globular fat (mainly TAGs) and proteins in a water-
based solution meant for its primary natural function: nutrition to the young mammals
(Figure 7). Bovine milk contains, on average 87% water, 3.3% protein, 4% fat together with
carbohydrates, vitamins and minerals (Walstra et al. 2006).
Figure 7. Schematic shapes of structural elements in milk illustrating the elements’ size differences, viewed ata magnification of 50 000. Fat globules and casein micelles float in a cloudy or opalscent liquid called serum.Image obtained from Walstra et al. (2006).
Milk proteins are typically divided into two types of proteins. The proteins are grouped based
on their solubility and structure/folding properties into caseins and whey proteins: (1)
Caseins (αS1-, αS2-, β-, and κ-casein) constitute around 80% of total nitrogen in milks.
Caseins have some helical and β-structures in their hydrophobic domains. As their primary
structure is rich in well-conserved prolines, caseins do not have a well-defined tertiary
structure. Consequently, caseins have an open rheomorphic conformation and are
outstandingly heat-stable (Holt and Sawyer 1993). Caseins are also characterised by their
precipitation properties at their isoelectric point (pI) at around pH 4.6. αS1-, αS2-, and β-
caseins are highly phosphorylated, and thus bind calcium cations (Ca2+) effectively making
them precipitate in the presence of Ca2+. κ-caseins, on the other hand, have their C-terminals
glycosylated, do not bind Ca2+ cations and are soluble in the presence of these cations
26
abundant in milk. As the glycosylated part of the C-terminal is a very hydrophilic domain,
and the N-terminals of the κ-caseins are highly hydrophobic, κ-caseins are a
characteristically amphiphilic group of proteins.
Milk is naturally rich in calcium phosphate, which is why caseins organize themselves as
association colloids by hydrophobic interactions and calcium phosphate bridges. These
association groups are often referred to as casein micelles (Figure 8). They vary in sizes (50–
600 nm), are hydrated, stable and polydispersed (Marchin et al. 2007). Because of the
amphiphilic properties of κ-caseins (protruding peptide chains in Figure 8), κ-caseins form
the negatively net charged outer layer of the micelles and are able to associate with the more
hydrophilic serum phase around the micelles (Dalgleish 2011). The inner parts of the
micelles hold calcium-sensitive α- and β-caseins protected. According to the review done by
Dalgleish (2011) caseins are thought to be held tightly together by colloidal calcium
phosphate, yet, micelle-micelle interactions are prevented by the “hairy layer”, the outer
layer of κ-caseins which makes the surface rough. However, there is some proof that the
interior parts of the micelles can be accessed through pores in the outer κ-casein layer
(Marchin et al. 2007; Knudsen and Skibsted 2010; Dalgleish 2011).
Figure 8. Cross-section of a casein micelle model where caseins are thought to be a somewhat organizedassociation colloid with smaller nanoclusters and submicelles making up the inner parts of the casein micelle.Image obtained from Walstra et al. (2006).
After the precipitation of caseins at their isoelectric point, (2) whey proteins are left,
counting up to 15% of the total nitrogen content in milk. Whey proteins are globular proteins
with compact and well-defined secondary and tertiary structures. 50% of whey proteins are
27
β-lactoglobulins (BLG, pI=5.2) and another 20% of the whey proteins are α-lactalbumin
(ALA, pI=4.8) (Haginaka 2000; Chen et al. 2016). Whey proteins have cysteines in their
primary structure which enable them to form intramolecular disulphide bonds. BLG has one
free cysteine that is hidden in its β-barrel structure (Carter and Ho 1994). When the
conformation of the protein changes and the tertiary structure is opened, the free cysteine
domain becomes exposed and reactive (Brownlow et al. 1997; Considine et al. 2007).
2.3.2 Acid milk gels
Traditionally yogurts are prepared using lactic acid bacteria (LAB) to ferment milk lactose
to lactic acid, which slowly acidifies the milk. Nonetheless, to provide control and
reproducibility for both food industry and research purposes, milk can be acidified by adding
an acidifier, such as glucono-δ-lactone (GDL), which gradually hydrolyses into gluconic
acid when in contact with water (Lucey et al. 1998; Lee and Lucey 2004; Marchin et al.
2007; Martin et al. 2009; Liu et al. 2017). When using GDL, the pH drop is similarly gradual
as with LAB fermentation. However, there are always structural differences between GDL
acidified and LAB fermented yogurts, as the initial acidification rate is faster with GDL
(Lucey et al. 1998; Martin et al. 2009). An example of the GDL induced protein network is
presented in Figure 9.
Figure 9. Confocal laser scanning micrographs of acidified acid milk gels at (a) 1 min, (b) 30 min, (c) 60 min,and (d) 120 min after the addition of glucono-δ-lactone, with the target pH of 4.6. Protein particles (bright)were labelled with Rhodamine B. Scale = 25 μm. Images obtained from Auty (2011).
28
Yogurts and acid milk gels can be described as a coarse network of linked protein clusters
and strands (Kaláb et al. 1983). To obtain this protein network, the proteins need to be
modified both with heat and the pH. The heat treatment partially unfolds caseins causing
changes in secondary structures and significant losses of the tertiary structures. This in turn,
causes hidden domains to be exposed which induces further conformational changes in the
proteins that are vital for the gel to form. Whey proteins are also partially unfolded: the heat
treatment (>70 °C) causes dissociation and aggregation of BLG proteins as well as thiol-
disulphide exchange reactions (Considine et al. 2007; Auty 2011). The free thiols will react
with other whey proteins and κ-caseins on the surfaces of micelles, which enables the
formation of micelle-bound and soluble κ-casein/whey protein complexes in milk.
Despite the role of the heat treatment, yogurts are most importantly acidified milk products.
The actual gelation happens by means of acidification even though the heat treatment is
important for the success of the gel. As the pH decreases and the matrix acidifies, micellar
calcium phosphate gradually dissolves and becomes fully soluble around pH 5.2 (Law and
Leaver 1998). This changes the charge-balance within the milk causing the negative surface
charges of the micelles to be balanced. The charge-balancing inflicts the outer hairy layer of
the micelles to collapse, bringing about a loss of steric and entropic effects between micelles
(Roefs and Van Vliet 1990; De Kruif 1997) as well as the increase of electrostatic repulsion
and weak physical interactions (hydrophobic and van der Waals attractions). Moreover, as
the concentration of dissolved Ca2+ increases and the pH approaches 4.6 (the pI of caseins),
the now exposed α- and β-caseins are precipitated. Thus, the pH drop increases micelle-
micelle interactions, which progressively creates a three-dimensional protein network
(Lucey and Singh 1997; Lucey et al. 1998).
The temperature of the heat treatment and the acidity of the milk highly affect the degree of
whey protein denaturation and the distribution of the bound and soluble complexes, thus
effecting how extensive the aggregation formation will be (Lucey and Singh 1997). The lack
of this extensity can be visibly seen on and around the gel as spontaneous syneresis, a
phenomenon where the protein gel network shrinks, becomes firmer, and releases some of
its serum phase (Ercili-Cura et al. 2013). Syneresis is commonly thought as a product defect
and is pronounced in low-fat products and when acidification time is reduced (see e.g.
Aryana and Olson 2017; Kumar et al. 2018).
29
2.3.3 Fibre fortified yogurts
The dietary fibre intake is usually low in the Western diet (Sonnenburg and Sonnenburg
2014). Even though it has been a controversial topic, it is known that dietary fibre has a
significant role in maintaining good human health (see e.g. Bingham et al. 2003; Makki et
al. 2018). To increase dietary fibre intakes, different dietary fibres have been added too
foodstuff as functional ingredients (Dhingra et al. 2012). As yogurts and other cultured dairy
products are expected to continue in their popularity among consumers, adding dietary fibres
to yogurts is an interesting and a convenient way of intervention (Gahruie et al. 2015; Aryana
and Olson 2017).
Dietary fibres used in yogurt fortification follow the definition of dietary fibres, where the
fibres are more often insoluble than soluble in water, are resistant to digestion and absorption
in the small intestine, and are usually polysaccharides, such as cellulose, hemicellulose and
pectins, and non-starch polysaccharides such as arabinoxylans (DeVries et al. 2001). These
fibres are usually extracted from the cell walls of fruits, vegetables, or cereals (Dello Staffolo
et al. 2004).
According to Dello Staffolo et al. (2004), the type of fibre and the storage time of yogurts
significantly affect rheological parameters of the yogurt. The authors continue to explain
that even though different fibres may or may not affect (spontaneous) syneresis and pH, they
do usually affect textural properties and sensory acceptability of yogurts: at their best, added
(dietary) fibres might add to the overall acceptability of fortified yogurt products in relation
to colour, flavour, and texture, and thus establish a functional food with commercial
applications. Some common (dietary) fibres used with yogurts include apple, inulin, wheat,
and bamboo fibres (Dello Staffolo et al. 2004), pineapple peel fibre (Sah et al. 2016), and
oat fibres (Fernández-García et al. 1998; Tomic et al. 2017). Adding oats to yogurts has been
associated with improved body and texture of plain yogurts (Fernández-García et al. 1998).
As the health benefits of oats have been contributed to its β-glucan fibres, it is feasible to not
only add parts of oats, but specifically increase the amount of oat β-glucan in yogurts.
The effect of adding β-glucans to yogurts have generally been associated with improved
water retaining properties observed by increased water holding capacity (WHC) and
decreased spontaneous syneresis (Table 1). β-glucans can be sourced from oats (Andersson
and Börjesdotter 2011) as well as form barley (Regand et al. 2011), spent brewer’s yeast
30
Table 1. Research on yogurts and acid milk gels with added β-glucan.
Reference Yogurt type β-glucan type Fibre inproduct Effects on yogurt quality and acceptability
Brennan andTudorica(2008)
Stirred yogurt Barley β-glucan (86%,Glucagel™)
β-glucan0.5–2.5%,
0.5% β-glucan was enough to improve serum retention and rheological properties. Overallimproved acceptability. Could be used as a fat-replacer.
Lazaridou etal. (2008)
Set acid skim milkgel
Barley β-glucans: 250, 210,180, 140, 70, and 40 kDa 0.5–2%
Weakened protein network, which could be compensated by high viscosity or a secondary gelnetwork with specific β-glucans. Low concentration increased and high concentrationsreduced spontaneous syneresis. Can influence textural attributes to form less cohesive butfirmer, more elastic, adhesive, and gummy products.
Sahan et al.(2008) Non-fat yogurt β-glucan composite (5.5%
soluble β-glucan) 0–1%Gel firmness and WHC1 were not influenced. Viscosity increased and spontaneous syneresisdecreased. In storage, firmness and WHC1 decreased while viscosity and syneresis furtherincreased. Lower levels of β-glucan composite gave satisfactory sensory results.
Elizaquívelet al. (2011)
Yogurt, reconstitutedsemi-skimmed milk
(1,3)-β-D-glucan producedby incorporated LAB2
strains [Pediococccusparvulus (strains 2.6 andCUPV22) and Lactobacillussuebicus CUPV221]
NA3 Did not significantly affect physical and rheological properties
Lazaridou etal. (2014)
Set yogurt,reconstituted milk
Oat β-glucan isolate,high solubility, 117 kDa 1.4%
Resulted in a liquid-like structure: a transient gel structure at an early stage due to phaseseparation of proteins and β-glucan, later retardation of protein aggregation and acidificationkinetics. β-glucan enhanced viability of probiotics in cold storage.
Rinaldi et al.(2015)
Non-fat yogurt,reconstituted milk
Barley β-glucan (75.6%,Glucagel™) 0.4%
Improved viscosity and delayed decrease of viscosity at high shear rate. Observations ofpossible phase separation might explain fast protein digestion for β-glucan enriched yogurts.
Fu et al.(2018) Set yogurt Linear (1,3)-β-D-glucan
(94%, Salecan) 0.5% (w/v)Stabilized product and fortified microstructure: improved anti-shear ability, temperaturestability, and water holding capacity, stabilized fat granules, and anchored casein micelles dueto an additional three-dimensional network.
Raikos et al.(2018)
Stirred skim yogurt,reconstituted milk
Spent brewer’s yeast(Saccharomyces cerevisiae)β-glucan, (Yestimun)
0.2–0.8%Reduced fermentations time, formed small spherical β-glucan clusters within the matrix. Didnot change syneresis, viscosity, colour, or acidity, but did increase hardness and adhesiveness.High concentrations showed adverse flavour and aftertaste, otherwise a potential thickener.
Vanegas-Azuero andGutiérrez(2018)
Yogurt, reconstitutedmilk
Ganoderma lucidumβ-glucan (Ganogen) andsacha inchi seeds(Plukenetia volubilis L.)
0-1.5%β-glucan and
4% sachainchi seeds
Did not affect fermentation kinetics. Significantly increased nutritional value. Increase ofβ-glucan content increased WHC and decreased syneresis. Firmness, consistency,cohesiveness and viscosity decreased. Sensorial acceptance >70%.
1 = Water holding capacity, 2 = Lactic acid bacteria, 3 = Not analysed
31
(Raikos et al. 2018), or mushrooms, such as Ganoderma lucidum (Vanegas-Azuero and
Gutiérrez 2018). The added β-glucan concentrations commonly range from 0.5 to 2%, which
is usually enough to get the 1 g β-glucan per portion required for the use of the health claims
(EFSA 2010, 2011).
Molecular aspects, such as molecular weight, ratio of linkages, and concentration highly
affect β-glucan functionality and physicochemical properties in yogurts (Lazaridou and
Biliaderis 2007; Lazaridou et al. 2008). It is believed that the added polysaccharides induce
phase separation of proteins and β-glucan and thus disturb the protein structure and break
the cohesiveness of the gel (Lazaridou et al. 2014; Rinaldi et al. 2015). However, β-glucans
may compensate for their interference by forming an alternative, water entrapping secondary
gel structure within the protein network, with long polysaccharides among casein micelles
or small spherical clusters within the yogurt matrix (Mårtensson et al. 2001; Fu et al. 2018;
Raikos et al. 2018). This secondary three-dimensional network increases the product’s
viscosity and adhesiveness (Sahan et al. 2008). The extent to which the β-glucans improve
or degenerate the yogurt structure seems to largely be dependent on the source of origin,
extraction and purification method, which in turn determine the molecular attributes of the
β-glucan in use (Lazaridou and Biliaderis 2007; Zhu et al. 2016).
32
3 EXPERIMENTAL RESEARCH
3.1 Background and aims
Health effects of β-glucan are believed to be linked with their physicochemical properties,
such as viscosity, which is linked to the concentration and molecular weight of β-glucan
(Regand et al. 2011). Recent research has shown that oxidized lipids may induce the
degradation of β-glucan molecules in aqueous food systems where β-glucan and lipids
co-exist. Wang et al. (2017) designed an oil-in-water emulsion system that contained
β-glucan and oxidized lipids, which they stored for 1 week at room temperature. During this
time, both viscosity of the emulsion, and molecular weight of β-glucan were reported to have
decreased.
This phenomenon needed to be further investigated, which was done as a part of a Business
Finland funded project, OATyourGUT. To further investigate the impact of water, fat, and
the food matrix on β-glucan properties during storage, a yogurt type food model was
designed. Oat β-glucan was to be added to these acid milk gels, with the aim to later link the
molecular level changes of β-glucan to human physiological behaviour and consumption of
oats.
The aim of the experimental part of this Master’s thesis work was to evaluate the impact of
hydrolysis and oxidation of intrinsic oat bran lipids on the functionality of β-glucan in oat
bran concentrate (OBC) ingredients and in acid milk gels (yogurt-type food model) during
storage. The hypothesis was that as the oat ingredients are stored and lipid oxidation is
induced, the β-glucan molecules would be affected and their molecular weight would
decrease, which would then be observed as changes in yogurt texture and stability. The
research questions to be answered were: (1) Does lipid degradation (hydrolysis and
oxidation) affect β-glucan functionality in dry oat ingredients and/or in yogurt models?
(2) Do different oat bran concentrate (OBC) ingredients cause differences in yogurt texture
and stability? This Master’s thesis was conducted as a collaboration between VTT Technical
Research Centre of Finland (VTT) and University of Helsinki (UH) during spring and
summer of 2018.
33
3.2 Overview of the storage trial and yogurt food model experiments
Three types of OBCs with approximately 20% of β-glucan were prepared with different
processing methods in order to induce three levels of lipid degradation. OBC with HIGH
level of lipid degradation (H) was prepared from non-heat treated oat bran; OBC with
MEDIUM lipid degradation (M) was prepared from heat treated oat bran; OBC with LOW
lipid degradation (L) was prepared from heat treated and defatted oat bran. To investigate
the effect of degraded lipids to the degradation and molecular integrity of β-glucan, a storage
trial was conducted: dry OBCs were stored in glass bottles for 0, 1, 2, 4, 8, and 12 weeks at
4, 22, and 40 °C. Separate glass vials with 1 g of sample were prepared and stored for volatile
compound analysis. Lipid degradation states were determined by analysing and following
formed volatile compounds, changes in tocopherol and tocotrienol concentrations, as well
as neutral lipid profiles at UH. At VTT, stored OBCs were used as fibre ingredients in
preparing acid milk gels at each time point of OBC storage test. Chilled acid milk gels were
analysed for spontaneous syneresis, water holding capacity, and texture (deformation test)
the following day from incubation. Spontaneous syneresis and deformation tests were
repeated after refrigerating (4 °C) a set of aliquots for a week, as depicted in Figure 10.
Figure 10. An overview of the experimental research. The experiment was started with lipid analyses (A), afterwhich acid milk gels were prepared (B).
34
3.3 Test materials
3.3.1 Oat bran concentrates (OBCs)
To produce the H sample, oat kernels were dehulled (no heat treatment), milled twice with
a Hosokawa Alpine 100UPZ pin disc mill (17800 rpm) and air-classified using a Minisplit
Air Classifier (British Rema Manufacturing Company Ltd., UK) with classifier wheel speed
of 2900 rpm and air flow of 220 m3/h to separate the bran into the coarse fraction. The coarse
fraction was further milled (2x17800 rpm, 100UPZ pin disc mill) and air classified
(Minisplit Air Classifier, 4500 rpm, 220 m3/h). Powdered concentrate was collected into the
coarse fraction (hereafter referred to as oat bran concentrate or OBC). For M sample, regular
oat bran produced from heat treated oat kernels was milled twice with a pin disc mill
(17800 rpm) and air-classified (Minisplit air Classifier, 3500 rpm, 220 m3/h) to obtain the
OBC in the coarse fraction. For L sample, regular oat bran from heat treated oat kernels was
defatted with supercritical carbon dioxide (SC-CO2) using a Nova Swiss extraction vessel
(Nova Werke AG, Effretikon, Switzerland) with a compressor Chematur Ecoplanning
(Chematur Engineering Ltd., Pori, Finland). Extraction pressure varied between 295 and
305 bar and the temperature in extraction container was 4 °C and in separation container
48 °C. The main points of producing oat bran concentrates with 20% β-glucan can be found
in Figure 11. Additionally, team members at VTT and UH have provided analyses of the
chemical characteristics of the OBC powders, such as β-glucan content, β-glucan molecular
weight (MW), lipase activity and moisture content. β-glucan content was analysed using
Megazyme’s Mixed linkage beta-glucan assay kit (K-BGLU 02/17; McCleary & Codd
1991), β-glucan molecular weight using a high performance size exclusion chromatography
(HP-SEC) method (Suortti 1993), lipase activity according to Yang et al. (2017), and
moisture content by evaporating water from the sample in a 105 °C oven overnight and
counting the moisture content as a percentage based on loss of sample weight. β-glucan
molecular weight was also analysed from fresh and stored (7 days at 4 °C) acid milk gels
with fresh M OBC and M OBC stored at 40 °C for 12 weeks.
35
Figure 11. Enrichment of β-glucan from fresh oat bran. For H and M oat bran concentrates (OBCs), millingand air-classification were performed twice (A and B), before subjecting the H and M OBCs to a storage trial(C). Oat bran for L OBC was milled and air-classified only once (A), since defatting (SC-CO2) the first OBC(β-glucan 15%) was enough to reach target β-glucan concentration before storage (C). Target β-glucanconcentration for all OBCs was 20%.
3.3.2 Reagents and reference materials
All solvents were of HPLC grade and chemicals were of pro-analysis grade and were
purchased from Merck (Darmstadt, Germany), Riedel-de Häen (Seelze, Germany), Sigma-
Aldrich (St. Louis, MI, USA), and Fisher Scientific (Loughborough, UK). Water was
purified by Milli-Q system (Millipore Corp., Bedford, MA, USA) and distilled water was
used for acid milk gel analyses. In acid milk gel preparation, glucono-δ-lactone (GDL)
(Sigma Chemicals, St. Louis, MO, USA) was used as an acidifier, and sodium azide 0.01 %
was prepared as an anti-microbial agent.
To identify and quantify fatty acid methyl esters, GLC-63 mixture (Nu-Check Prep, Inc.,
Elysian, MN, USA) and an internal standard C19:0 methyl ester were used, respectively.
Methylation reagent BF3-CH3OH was purchased from Sigma-Aldrich (St. Louis, MI, USA).
To identify neutral lipid classes, TLC-reference standard 18-6 A (Nu-Check Prep, Inc.,
Elysian, MN, USA) was used. To quantify neutral lipids, calibration mixtures were prepared
in heptane and consisted of tripalmitin for TAGs (>99%, Sigma-Aldrich, St. Louis, MI,
USA), dipalmitin for DAGs (>99%, Nu-Chek Prep, Inc., Elysian, MN, USA) and oleic acid
for FFAs (>99%, Nu-Chek Prep, Inc., Elysian, MN, USA). To identify tocols, tocopherol
reference substances DL-α-, β-, γ-, and δ-Tocopherol (Art. no 15496, stored in a freezer)
were obtained from Merck (Darmstadt, Germany).
To ensure method stabilities, the effectiveness of the ASE extraction and the accurate levels
of tocols and neutral lipids (NP-HPLC methods), in-house references were used and stored
36
in the freezer as follows: NP-HPLC methods, commercial whole wheat flour (Sunnuntai,
Raisio Ltd., Finland), and additionally for tocol analysis, commercial rapeseed oil (used as
an in-house reference in the Food Chemistry research group (UH) since 2011). For the
volatile compounds, an oat bran concentrate with excessive lipid degradation (obtained from
the grain research group (UH)).
3.4 Methods
The oat ingredients (β-glucan enriched OBCs) were stored in different temperatures up to
12 weeks, to study heat-induced lipid degradation (oxidation and hydrolysis). Lipid
degradation was analysed chemically by following three indicators: increase in specific
volatile compounds, tocopherol and tocotrienol degradation and the formation of free fatty
acid. Figure 12 presents an overview of the lipid degradation experiments.
Figure 12. A summary of the experimental design and chosen methods for lipid degradation analysis. a-T3 =α-tocotrienol, FFA = free fatty acids.
3.4.1 Extraction of lipids and analysis of total FA content
Lipid contents and fatty acid compositions of the OBC samples were determined by a gas
chromatographic (GC) method used for fatty acid methyl ester analysis (Lampi et al. 2015).
Lipids were first extracted from OBC samples with acetone (≥99.8% Sigma-Aldrich, St.
Louis, MI, USA) using an accelerated solvent extraction method (ASE) (Dionex ASE 200
Accelerated Solvent Extraction system, Thermo Fisher Scientific). The specifics of the
ASE method were: temperature 100 °C, pressure 1000 psi, heating time 5 min, two times
static cycle time 10 min, flush volume 60%, nitrogen gas purge time 60 s, and rinsing time
1 min. For the extraction, approximately 1 g of the sample was precisely weighed into the
ASE-extraction cell together with an equal amount of drying agent (Sand, General purpose
37
grade, Thermo Fisher Scientific). The cell was then filled with the drying agent, and before
closing the extraction cell, 1 ml of ethanol (≥99.5%, Altia, Rajamäki, Finland) was added on
top of the sand for better extraction of lipids. Lipid extracts were moved to 10 ml KIMAX
Test Tubes with Teflon Liner Caps. 5 mg of C19:0 in heptane was added in each extract as
an internal standard (ISTD). The extracts were transferred quantitatively to 50–100 ml
round-bottom flasks with ethanol, which also aided in evaporating trace water. The acetone
and ethanol were evaporated in a rotary evaporator at ≤37 °C, and redissolved in
3:2 heptane:isopropanol (heptane ≥98%, Merck, isopropanol ≥99.9%, Riedel-de Hän).
For methylation of the extracted lipids, the solvent was first evaporated at ≤40 °C under a
nitrogen flow. Then 1 ml of 0.5 M NaOH in methanol was added to each test tube, and
carefully closed tubes were placed in boiling hot water for 5 minutes. Once the tubes cooled
down, 2 ml of methylation reagent BF3-CH3OH (~10% in methanol, Sigma-Aldrich, St.
Louis, MI, USA) was added and the tubes were heated in boiling hot water for another
5 minutes. 5 ml of heptane and 2 ml of supersaturated NaCl solution were added to cooled-
down extracts, after which the extracts were agitated vigorously for approximately 1 minute.
Using glass Pasteur-pipettes, the upper heptane layers were transferred to clean KIMAX
tubes with some drying agent (NasSO4) at the bottom. After a minimum of 30 minutes,
heptane extracts were loaded to vials for GC analysis as described by Soupas et al. (2005)
and Damerau et al. (2014): The equipment used was an Agilent 6850 Network GC system
coupled with a flame ionization detector (FID). The column in use was a fused-silica
capillary column (Omegawax 250: 30m x 0.25 mm, 0.25 μm, Supelco, Bellefonte, PA,
USA). Helium acted as the carrier gas at a flow of 1 ml/min. The oven temperature program
started at 160 °C and was held there for 1 min, after which the temperature was raised
4 °C/min to 240 °C, and was finally held there for 5 min. The detector temperature was
250 °C, with detector gases helium and air. Injector temperature was 240 °C, and injection
was performed with a split ratio of 1:28 (split injection technique). Agilent ChemStation was
used to operate the instrument, sampling and in result handling.
The identification of fatty acid methyl esters was done based on a standard GLC-63 mixture
and the quantification was conducted using C19:0 methyl ester. Total lipid contents were
calculated as sums of the major fatty acid methyl esters, i.e. C14:0, C16:0, C18:0,
C18:1 (n-9), C18:1 (n-7), C18:2 (n-6), C18:3 (n-3), and C20:1 (n-9), and reported as mg/g
sample on fresh weight basis. Experiments were performed as triplicates. Coefficients of
variation for fatty acid methyl esters were 4 % or lower.
38
3.4.2 Analysis of neutral lipid profiles with NP-HPLC-ELSD
Lipids were extracted from 1.0 g of sample with acetone as described in Section 3.4.1
Extraction of lipids and analysis of total FA content, except no ISTD was used and the
redissolving in 10 ml was performed with HPLC-grade heptane only. After redissolving, the
extracts were further diluted 1:10 in heptane and placed in HPLC vials. To analyse neutral
lipid profiles of the oat ingredients, a method described by Lampi et al. (2015) was used: a
normal-phase HPLC (Agilent 1200 HPLC system, Agilent Technologies, Santa Clara, CA,
USA) with an evaporative light scanning detector (ELSD) (Waters 2420 ELSD, Waters®,
Milford, MA, USA) was used to analyse neutral lipid classes (triacyl, diacyl and monoacyl
glycerols and free fatty acids; TAGs, DAGs, MAGs, FFAs, respectively). A LiChrosorb diol
column (5 μm, 3 x 100 mm, VDS optilab Chromatographie Technik GmbH, Berlin,
Germany) was used to separate neutral lipid classes. A linear gradient elution was a mixture
of heptane and 0.1% acetic acid and an increasing proportion of isopropanol (from 0.09%
isopropanol from 0–5 min, to 3% from 5–15 min, at 3% from 15–35 min, and back to 0.09%
from 35–40 min) with a flow rate of 0.5 ml/min at 25 °C. The stabilisation time between runs
was 10 minute. The ELSD was set to drift tube temperature 60 °C, nebuliser to 42 °C and
the gain to 10. Filtered air was used for nebulisation at a flow rate of 1.4 l/min (i.e. 20 psi).
Quantification was performed by using an external standard method. Calibration mixtures
were prepared in heptane and consisted of tripalmitin, dipalmitin, and oleic acid, as
explained in Section 3.3.2. Calibration mixtures were analysed both at the beginning and the
end of each sequence at six amounts in the following ranges: 100–4000 ng/injection for
TAGs and FFAs, and 100–5000 ng/injection for DAGs. Calibration curves were computed
with a second order regression separately for the three acylglycerol classes and oleic acid.
The regression coefficients for calibration curves were ≥ 0.95. Each OBC lipid extract was
injected in two volumes, 5.0 and 50.0 μl, to meet the calibration ranges. Experiments were
performed as triplicates and the results were reported as mg/g sample on fresh weight basis.
To ensure the performance of the equipment and the stability of the HPLC method, a mixture
of oleyls (TLC-reference standard 18-6 A) was analysed with each HPLC sequence. Oleyls’
retention times remained stable throughout the experiment. To ensure the stability of the
whole method, starting from lipid extraction, a whole wheat flour (Sunnuntai, Raisio Ltd.,
Finland) was used as an in-house reference. Neutral lipid amounts in the in-house reference
stayed stable throughout the study: TAGs remained at 6.37 ± 0.45 mg/g sample, FFAs at
39
4.32 ± 0.26 mg/g sample, and DAGs at 0.78 ± 0.35 mg/g sample, with a total of
11.47 ± 0.51 mg/g sample.
3.4.3 Analysis of tocopherol and tocotrienol contents with NP-HPLC-FLD
Tocopherol and tocotrienol contents were analysed with a normal phase high performance
liquid chromatography method with a fluorescence detector (NP-HPLC-FLD) as described
by Schwartz et al. (2008): Lipids were extracted from 1.0 g of sample with acetone as
described in Section 3.4.1, except no ISTD was used and the redissolving in 10 ml was
performed with HPLC-grade heptane only. Approximately 1 ml of lipid extracts were then
filtered (Acrodisc GHP, 0.54 μm, 13 mm) into HPLC vials, ready to be analysed. To ensure
the performance of the HPLC equipment, rapeseed oil (50 mg in 10 ml heptane) was used
as an in-house reference.
Standard stock solutions were prepared from tocopherol reference substances DL-α-, β-, γ-,
and δ-tocopherol (Art. no 15496, Merck, Darmstadt, Germany, stored in a freezer) to a
concentration of approximately 500 mg/l in ethanol (AAS) and stored at -20 °C. Suitable
amount of each tocopherol were combined and diluted with heptane to create a reference
mixture with 1.5-2 mg/l of each tocopherol. Exact tocopherol concentrations were confirmed
spectrophotometrically (Ryynänen et al. 2004) using reported absorption coefficients for
each vitamer in ethanol E1% at a length of 1 cm. Tocotrienol vitamers were quantified with
each respective tocopherol (Ryynänen et al. 2004).
The HPLC-system consisted of a Waters 515 HPLC pump (Waters Corporation, Milford,
MA, USA), a Waters 717 plus Autosampler, a Guard-Pak Inserts Resolve Silica precolumn
(Waters), a Varian Inertsil 5 SI column (250 x 4.6 mm, 5 μm particle size, Varian Inc., Palo
Alto, CA, USA) and a Waters scanning fluorescence detector model 474. The system
software was Millenium 32 software (Version 4.00, Copyright® 2001 Waters Corporation).
Separation of tocols took 25 minutes, they were separated isocratically using a mobile phase
of 3% 1,4-dioxane and 97% heptane (v/v) with a flow rate of 2 ml/min, and detected
fluorometrically (λex: 290 nm, λem: 325 nm) (Kamal-Eldin et al. 2000). Dissolved air in the
mobile phase was removed first by sonication before analysis and by an online-helium
degasser. The column oven was set at 30 °C.
40
Tocopherols were identified by comparing retention times to those of the standards. The
elution order of the tocopherols and tocotrienols was α-tocopherol first, which was followed
by α-tocotrienol, β-tocopherol and β-tocotrienol, having the polar column retain more polar
molecules for a longer time. Quantification was performed by using external calibration
functions that were established by combining 2x7 calibration points obtained by injecting
tocopherol standard mixture at different volumes (2–100 μl) at the beginning and the end of
each sequence. Quantification of each tocotrienol was done by using the corresponding
tocopherol calibration functions (Ryynänen et al. 2004), and reported as μg/g sample on
fresh weight basis. Experiments were performed as triplicates. Regression coefficients for
the calibration curves were all >0.99.
To ensure the stability of the whole method, starting from lipid extraction, a commercial
whole wheat flour (Sunnuntai, Raisio Ltd., Finland) was used as an in-house reference.
Tocopherol and tocotrienol amounts in the in-house reference stayed stable throughout the
study: a-T remained at 6.95 ± 0.32 μg/g, a-T3 at 3.11 ± 0.12 μg/g, b-T at 4.37 ± 0.11 μg/g,
and b-T3 at 17.45 ± 0.47 μg/g. The tocopherol amounts in the rape seed oil remained within
the standard deviations of the in-house data since 2011.
3.4.4 Analysis of volatile compounds with HS-SPME-GC-MS
Six volatiles typical for lipid oxidation were followed, and changes in concentrations
recorded. The volatiles were, in retention order, 2-ethylfuran, 1-pentanol, hexanal,
2-pentlyfuran, benzaldehyde, and nonanal. Volatiles were measured with a headspace solid
phase microextraction gas chromatography method coupled with a mass spectrometer
detector (HS-SPME-GC-MS) (GC Perkin Elmer, USA) using 1 g (0.99-1.01 g) of the dry
OBC samples in each 20 ml amber headspace vial in triplicates for each time point. Tightly
sealed vials were stored at different temperatures (4, 22, 40 °C) for varying amounts of time
(0–12 weeks) to induce an array of lipid degradation levels. Volatile compounds were
analysed using an HS-SPME injector (combiPAL, CTC Analytics, Zwingen, Switzerland)
with a CAR/PDMS-fibre (85 μm film thickness; Supelco, Bellefonte, PA, USA) together
with a GC (HP 6890 series, Agilent Technologies Inc.), including a capillary column SPB-
624 (30 m x 0.25 mm i.d., 1.4 μm film thickness; Supelco, Bellefonte, PA, USA), coupled
to an MS detector (Agilent 5973 Network, Agilent Technologies Inc.). MSD Chemstation
(Agilent Technologies Inc.) was used for data processing.
41
The analysis was performed as reported earlier by Damerau et al. (2014): Samples were
incubated and extracted for 20 min at 50 °C with an agitation speed of 250 rpm to equilibrate
the volatile compounds between the solid and gas phase. Volatile compounds were identified
by matching their total ion mass spectra (m/z from 50 to 300) with the database Wiley 7N
(Wiley Registry TM of Mass Spectral Data, 7th Edition, Hoboken, NJ, USA) and by
comparing retention times and mass spectra to previously acquired data on oat volatile
compounds.
To ensure the stability of the HS-SPME-GC-MS method, the peak areas of five volatiles
were chosen and followed from an in-house-reference (freezer-stored OBC) in each
sequence, however, the freezer-stored OBC was old and unstable, and did not perform in a
repeatable manner. To monitor the performance of the equipment, air and water checks were
performed daily, and tune evaluation monthly. Experiments were performed as triplicates
and the results are presented as peak areas. The variation coefficients of the peak areas
ranged between 2.5 and 10% within a sequence.
3.4.5 Preparation of acid milk gels
Figure 13. The stages of preparing acid milk gels with added β-glucan (oat bran concentrates).
The preparation and acidification of the acid milk gels was performed as described by
Ercili-Cura et al. (2013) and Rosa-Sibakov et al. (2016) (overview in Figure 13).
Acidified milk gels were prepared in 1000 ml glass bottles with sealable caps to contain
9.57% (w/w) milk solids, 3.3% protein (w/w) and 1 g of β-glucan per portion (portion chosen
was 150 g of yogurt) to achieve the level of β-glucan specified in the health claims (EFSA
2010, 2011) and minimize the added dry matter. This meant an addition of 2.8–3.6% (w/w)
42
of OBC ranging with the assessed β-glucan levels of each sample. The OBC samples, each
with different levels of lipid degradation (stored in varying conditions: 4, 22, or 40 °C for
0–12 weeks), were mixed with distilled water by constant magnetic stirring (190 rpm) at
50 °C for 30 minutes. Skimmed milk powder (34.5% protein, 53% lactose, 8% ash, 3.8%
moisture, and 0.5% fat) (Valio Ltd., Helsinki, Finland) was reconstituted by adding the milk
powder to the OBC-water mixture, and mixed for another 30 minutes by constant magnetic
stirring (190 rpm) at 50 °C. The new mixture (reconstituted skim milk + OBC) was heat
treated by heating the sample ≥86 °C for 5 minutes and cooled down in ice to 40 °C. Sodium
azide (0.01% (w/w)) was added to an aliquot to prevent microbial growth in gels that were
to be stored for 1 week. The total volume of each acid milk gel sample prepared was 500 ml.
The acidification was started by adding 1.3% (w/w) glucono-δ-lactone (GDL) (Sigma
Chemicals, St. Louis, MO, USA) at a constant dosage to the mixture at 40 °C. The mixtures
were then stirred for 10 minutes with magnetic stirring (250 rpm) until GDL was totally
dissolved. Control samples were prepared in a similar manner without the added β-glucan.
The gels were then divided into 40 or 20 g aliquots in tightly sealed plastic cups (ø 3.5 cm)
or sealable 50 ml centrifuge tubes, respectively. Incubation was performed at 40 °C for
approximately 4.5 hours or until the pH reached 4.5–4.6. The acidification process and
changes in pH were recorded during incubation at 10 minute intervals. Each acid milk batch
produced four replicate samples for fresh gels, four replicate samples for stored gels, and
three replicate samples for water holding capacity.
3.4.6 Analysis of spontaneous syneresis and water holding capacity
The water retention ability of an acid milk gel depends on the microstructure of the gel that,
in turn, depends on the acidification temperature and pH combination (Ercili-Cura et al.
2013). Water retention can be assessed by measuring spontaneous syneresis, commonly
thought to be a product defect among consumers, and water holding capacity (WHC) of the
gels. Measuring spontaneous syneresis provides insight into the acceptability of a
prospective yogurt-type product.
After the gels (40 g) had reached the targeted pH (4.5-4.6) and had cooled down in the fridge
overnight, the gels were placed at room temperature for 10 min to condition. The water
released on top of the gel was poured out carefully and weighed. Spontaneous syneresis
43
analyses were performed in quadruplicates, and expressed as the percentage of the amount
of water released on top of the gel in relation to the initial weight of the gel.
For the WHC analysis, the formation of acid milk gels was performed in 50 ml centrifuge
tubes as previously described by Ercili-Cura et al. (2013): 20 g of sample was acidified in
the centrifuge tubes at 40 °C until pH arrived at 4.5–4.6. The gels were centrifuged at 3000g
for 15 minutes at 4 °C. Supernatant was poured out carefully and weighed. WHC analyses
were performed in triplicates, and expressed as the percentage of the weight of the pellet in
relation to the initial weight of the gel.
3.4.7 Analysis of gel firmness with TA
The firmness of the acid milk gels was analysed as previously reported by Ercili Cura et al.
(2009) and Rosa-Sibakov et al. (2016). A large deformation test was done using a
TA.XTPlus Texture Analyser (TA, Stable Microsystems Ltd., Godalming, UK) equipped
with a 5 kg load cell. The samples (40 g) were incubated (40 °C) in plastic cups (ø 3.5 cm)
until pH was 4.5–4.6, after which the gels were refrigerated overnight or for 1 week. Before
the deformation test, the chilled gels were left at room temperature for 10 minutes. Gel
deformation was done by penetrating the gel with a ø 25 mm cylinder probe at a constant
speed of 1.0 mm/s to a distance of 75% of the sample height. Deformation tests were
performed on three to four individual gels for each gel batch. The area under the force vs.
distance curve (up to 15 mm) was determined and taken as a measure of gel firmness. The
reported values are the means of two acid milk gel batches.
3.5 Results
3.5.1 OBC characteristics
The three OBCs contained 18–23% β-glucan (Table 2). OBC samples showed differences in
lipid degradation levels ranging from no degradation to higher levels of degradation. H OBC,
which was not heat treated, had the highest content of total fatty acids, 9.0±1.2%. M OBC,
which was heat treated, was not far off, and had nearly as much fatty acids as H OBC,
7.8±1.0%. L OBC was defatted, and had 1.3±0.2% of fatty acids left. H OBC was the only
oat ingredient to exhibit lipase activity. The MW of β-glucan in all OBCs did not change and
44
remained at >1000 kDa during the storage time. The MW of β-glucan remained >1000 kDa
also in fresh and stored (7 days at 4 °C) acid milk gels with fresh M OBC and M OBC stored
at 40 °C for 12 weeks.
Table 2. β-glucan, fat, and moisture contents of differently processed oat bran concentrates (OBCs)and their lipase activities.
Sample β-glucan(%, dm)
Total FA2
(%)Moisture(%)
Lipase activity(μmol/min/g)
Non-heat treated(H OBC) 18.1±0.4 9.0±1.2 5.7 0.9±0.05
Heat treated(M OBC) 23.2±0.3 7.8±1.0 4.7 0
Heat treated and defatted(L OBC) 19.9±0.2 1.3±0.2 5.2 NA1
1NA = not analysed, 2FA = fatty acid
3.5.2 Lipid degradation
Neutral lipids
The FFAs were eluted as two peaks, which were added together for the total FFA count.
H OBC was the only sample of the three OBCs, where FFAs were formed. When stored at
4 °C, FFAs started to form only after 4 weeks, whereas at the higher temperatures, FFAs
could be seen already at 2 weeks. The FFA amounts increased with time at all the storage
temperatures: at the slowest rate at 4 °C, steadily at 22 °C, and at the fastest rate at 40 °C
(Figure 14). The highest amount of FFAs formed after 12 weeks at 40 °C was
3.8 mg/g sample, which was over 2.5 times more than for the OBC stored at 4 °C (1.5 mg/g).
The total amount of neutral lipids decreased at a similar rate at all storage temperatures
(Figure 15). The share of TAGs to the amount of total neutral lipids was the greatest for
samples stored at 4 °C (91.6%), and the lowest for samples stored at 40 °C (83.7%). As in
compensation, there were more DAGs and FFAs at higher storage temperatures.
45
Figure 14. Free fatty acids (FFA) were formed during a storage trial of a non-heat treated oat bran concentrate(H OBC) with a β-glucan content of 23.2±0.3% and a fat content of 7.8±1.0%. H OBC was stored at 4, 22, and40 °C for up to 12 weeks. N = 3.
Figure 15. Neutral lipid classes (triacyl glycerols, diacyl glycerols, and free fatty acids; TAG, DAG, FFA,respectively) were observed in a storage trial of a non-heat treated oat bran concentrate (H OBC) with aβ-glucan content of 23.2±0.3% and a fat content of 7.8±1.0%. H OBC was stored at (a) 4 °C, (b) 22 °C, and(c) 40 °C for up to 12 weeks. N=3.
M OBC had less neutral lipids to start with, 49.8±0.8 mg/g sample, when compared to
H OBC’s 58.4±0.3 mg/g sample. A trend of slight decrease of TAGs, increase of DAGs, and
46
decrease of total neutral lipids is visible as time and storage temperature increase (Figure 16).
Fresh L OBC had only possible traces of neutral lipids (data not shown), and thus neutral
lipid classes were not followed during the storage trial.
Figure 16. Neutral lipid classes (triacyl glycerols, diacyl glycerols, and free fatty acids; TAG, DAG, FFA,respectively) were observed in a storage trial of a heat treated oat bran concentrate (M OBC) with a β-glucancontent of 23.2±0.3% and a fat content of 7.8±1.0%. M OBC was stored at (a) 4 °C, (b) 22 °C, and (c) 40 °Cfor up to 12 weeks. N=3.
Tocopherols and tocotrienols
H OBC had the highest amounts of tocopherols and tocotrienols (hereafter referred as tocols)
compared to M OBC and L OBCs. The amounts of tocols in fresh H OBC were
7.5±0.08 μg/g sample for a-T, 66.7±0.9 μg/g sample for a-T3, 0.9±0.01 μg/g sample for b-T,
and 8.2±0.2 μg/g sample for b-T3 (Figure 17). All the tocopherols degraded during time, and
more intensively with increased temperature: the final amounts for a-T, b-T, and b-T3 neared
zero at all temperatures, and a-T3 persisted at a range of 36.7±0.3 to 12.7±0.2 μg/g for
samples stored at 4 and 40 °C, respectively.
47
Figure 17. α-tocopherol (alpha), α-tocotrienol (aT3), β-tocopherol (beta), and β-tocotrienol (b3T)concentration levels were observed during a storage trial of a non-heat treated oat bran concentrate (H OBC)with a β-glucan content of 18.1±0.4% and a fat content of 9.0±1.2%. H OBC was stored at (a) 4 °C, (b) 22 °C,and (c) 40 °C for up to 12 weeks. N = 3.
Tocols in M OBC decreased with increased time and temperature (Figure 18) in a similar
manner as in H OBC. Higher temperature also accelerated the degradation of all tocols.
L OBC had only possible traces of tocols (data not shown), and thus were not followed for
during the storage trial.
48
Figure 18. α-tocopherol (alpha), α-tocotrienol (aT3), β-tocopherol (beta), and β-tocotrienol (b3T)concentration levels were observed during a storage trial of a heat treated oat bran concentrate (M OBC) witha β-glucan content of 23.2±0.3% and a fat content of 7.8±1.0%. M OBC was stored at (a) 4 °C, (b) 22 °C, and(c) 40 °C for up to 12 weeks. N = 3.
Volatile compounds
H OBC showed resilience towards storage and heat. There were no detectable volatile
compounds typical for lipid oxidation in samples stored at 4 or 22 °C during 12 weeks. Some
enzyme activity related 2-pentylfuran could be detected in samples stored at 40 °C, however,
in a small measure (Figure 19).
49
Figure 19. 2-pentylfuran formation was observed in a storage trial of a non-heat treated oat bran concentrate(H OBC) with a β-glucan content of 18.1±0.4% and a fat content of 9.0±1.2%. H OBC was stored at 40 °C forup to 12 weeks. N = 3.
M OBC portrayed more typical volatile profiles for heat treated and stored oat ingredients.
Autoxidation product hexanal was the dominant volatile compound together with a little
nonanal. Hexanal amounts increased steadily during 8 weeks of storage until it decreased
towards week 12. The higher the storage temperature was, the more there was hexanal
(Figure 20).
Figure 20. Hexanal formation was observed in a storage trial of a heat treated oat bran concentrate (M OBC)with a β-glucan content of 23.2±0.3% and a fat content of 7.8±1.0%. M OBC was stored at 4, 22, and 40 °Cfor up to 12 weeks. N = 3.
Similarly, nonanal peaked at 8 weeks, however, the highest amounts of nonanal seemed to
be formed at lower temperatures (Figure 21). The samples stored in 40 °C showed also some
formation of 2-pentylfuran (Figure 22), approximately double the amount found in H OBC
50
(Figure 19). All L OBC samples formed traces of hexanal, possibly showing slightly higher
amounts in samples stored at 40 °C.
Figure 21. Nonanal formation was observed in a storage trial of a heat treated oat bran concentrate (M OBC)with a β-glucan content of 23.2±0.3% and a fat content of 7.8±1.0%. M OBC was stored at 4, 22, and 40 °Cfor up to 12 weeks. N = 3.
Figure 22. 2-pentylfuran formation was observed in a storage trial of a heat treated oat bran concentrate(M OBC) with a β-glucan content of 23.2±0.3% and a fat content of 7.8±1.0%. M OBC was stored at 4, 22,and 40 °C for up to 12 weeks. N = 3.
As H OBC showed stability to elevated temperatures and time, and produced only little
amounts of volatiles, an extra experiment was performed in order to understand the
phenomena and exclude the possibility of an error in the sampling. Raw materials for H and
M OBC were subjected to a similar storage test as for the OBCs storage conditions being 8
weeks at 40 °C, and then analysed for volatile compounds. For H OBC, milled non-heat
treated oat kernels (H RAW), i.e. material used before obtaining oat bran and before the air-
51
classification procedures (Figure 11) to produce H OBC, was used. For M OBC, the basic
heat treated commercial type OBC (M RAW, β-glucan content ~15%), was used. Some
2-pentylfuran was formed in H RAW at 4 and 8 weeks of storage, which counted for double
the amount H OBC had at its highest at 8 weeks. Additionally H RAW showed signs of
oxidation as there was some formation of hexanal at weeks 4 and 8, although low in amounts
(Figure 23a). M RAW was more stable than M OBC, and showed some hexanal and
2-pentylfuran formation (Figure 23 b). However, it should be noted that e.g. the
concentration of hexanal was minor, four times smaller than the amounts formed in M OBC
stored at 40 °C. The main finding from analysing the RAW materials was their different
profiles with different volatile compounds.
Figure 23. Volatile compounds were observed in a storage trial of (a) a non-heat treated oat flour and (b) heattreated oat bran concentrate (β-glucan content ~15%), both raw materials for oat bran concentrates (H OBCand M OBC). Samples were stored at 40 °C for up to 8 weeks. N = 3.
3.5.3 Properties of acid milk gels
Acidification
Control acid milk gels (without added OBC) had initial pH values around 6.5, and were
acidified to pH 4.6 in 4–5 hours as shown in the acidification curve in Figure 24. When
H OBC was added to milk gels, the acidification time was longer and it would have taken
over 20 h to get close to pH 4.6 (Figure 24). For practical reasons, a pH value of 4.6±1.0
with an acidifying time of 5-6 h was settled upon. M and L OBCs arrived to the target pH in
a shorter time, and after approximately 5 h of acidifying, the pH was 4.5-4.6. The effect of
storage of the OBCs on acidification times was most pronounced in acid milk gels with
H OBC. For example, when using fresh H OBC, it took just over 6 hours for the pH to reach
4.7. However, when H OBC was stored 12 weeks at 40 °C and then used as an ingredient,
in the same amount of time, the pH reached 4.6 (Figure 25, shown with arrows).
52
Figure 24. Acidification of acid milk gels without (control) or with fresh oat bran concentrates (OBC). H OBCwas non-heat treated, M OBC heat treated, and L OBC heat treated and defatted.
Figure 25. Acidification of acid milk gels with added non-heat treated oat bran concentrate (H OBC). H OBCwas stored in dark at 40 °C for 0–12 weeks prior to use as an ingredient in the acid milk gels. Storage of oatingredients shortened the acidification time, indicated with arrows.
Spontaneous syneresis and water holding capacity (WHC)
Control acid milk gels (without added OBC) had 9±3% spontaneous syneresis (Figure 26c),
and a water holding capacity of 39±10% (Figure 27c). None of the gels with added OBCs
showed spontaneous syneresis (Figure 26ab). When gels with M and L OBCs were subjected
53
to centrifuging, water retention was a perfect 100% (Figure 27ab). However, the water
holding capacity of gels with fresh and stored H OBCs were not far and ranged between 96%
and 99% (Table 3).
Figure 26. Acid milk gels (40 g) (a, b) with and (c) without added oat bran concentrates (OBCs). The additionof OBCs improved the water retention properties of the gels leading to having gels with no visible spontaneoussyneresis on top of the gels (a, b).
Figure 27. Acid milk gels (20 g) (a, b) with and (c) without added oat bran concentrates (OBCs) were placedin individual 50 ml tubes and centrifuged in 3000g at 4 °C for 15 min. The addition of OBCs substantiallyimproved the water retention properties of the gels: from (c) 39% to (a, b) 96–100%.
Table 3. Water holding capacity (WHC) of acid milk gels (a) without and (b) with added oat fibre. Added oatfibre was a non-heat treated oat bran concentrate (H OBC) with a β-glucan content of 23.2±0.3% and a fatcontent of 7.8±1.0%. H OBC was stored at 4, 22, and 40 °C for up to 12 weeks before being used as aningredient in the acid milk gels. N =3. 1NA = not applicable.
SampleStorage
temperature(°C)
WHC (%)
0 wks 2 wks 4 wks 8 wks 12 wks
a) Control NA1 38.8±10.4 NA1 NA1 NA1 NA1
b) H OBC(non-heattreated)
4 95.8±3.74 98.1±0.17 99.4±0.07 99.3±0.08 98.0±0.54
22 95.8±3.75 96.4±1.12 99.3±0.02 99.3±0.06 98.9±0.15
40 95.8±3.76 98.0±0.19 99.5±0.06 99.7±0.27 97.2±0.23
54
Firmness of acid milk gels
Control acid milk gels (without added OBC) were firm but brittle, and would immediately
release serum when deformed. Acid milk gels with OBCs were softer and smoother than
control gels (Figure 28a). Small oat specs could be seen in all OBC gels, however, having
H OBC as an ingredient seemed to produce the grainiest gels. The control sample had the
largest area under the force vs. distance curve, 15.3±1.7 N.mm, indicating a very firm acid
milk gel. Compared to the control, the gels with fresh OBCs were not as firm: 5.1±0.4,
7.9±4.2, and 5.2±1.1 N.mm for H, M and L OBC gels, respectively. All gels with added
fresh OBCs were stored 7 days at 4 °C. After this storage time in the fridge, gels with H OBC
were softer than they were as freshly prepared gels, gels with M OBC were similar to what
they were as freshly prepared gels, and gels with L OBC were harder than what they were
as freshly prepared gels. An example of the force vs. distance curve is provided in Appendix
1.
Figure 28. The firmness of acid milk gels changed depending on the types of oat bran concentrates (OBCs):(a) a fresh gel with fresh OBC, (b) a fresh gel with OBC stored for 12 weeks at 22 °C, and (c) a gel stored at4 °C for 7 days with OBC stored for 12 weeks at 22 °C. Control sample was prepared without added OBC.H OBC was non-heat treated, M OBC heat treated, and L OBC heat treated and defatted.
All H OBC gels were firmer than or at the same firmness level as gels with fresh H OBC
(week 0), regardless of the temperature H OBCs were stored at (Figure 29). Between 2 and
12 weeks of storage of H OBCs, gels with H OBCs stored both at 4 and 40 °C decreased in
firmness. There were no differences between gels of weeks 2 and 12 with H OBCs stored at
22 °C (Figure 29). Gels stored for 7 days at 4 °C were generally firmer than fresh gels (Figure
28bc (H OBC) and Figure 29).
55
Figure 29. Firmness of acid milk gels with a non-heat treated oat bran concentrate (H OBC, 18.1±0.4%β-glucan and 9.0±1.2% fatty acids). H OBC was stored at 4, 22, and 40 °C for up to 12 weeks, before used inacid milk gels. A control sample without added fibre was also prepared (on the left). N = 4.
M OBC gels gained firmness after M OBCs were stored 8 weeks at any storage temperature,
only to decrease in firmness with ingredients stored for 12 weeks (Figure 30). Although,
standard deviations were notably high. For gels with M OBC, it was not clear whether storing
gels for 7 days at 4 °C had an increasing or decreasing effect on gel firmness compared to
freshly prepared gels (Figure 28bc (M OBC) and Figure 30).
Figure 30. Firmness of acid milk gels with a non-heat treated oat bran concentrate (M OBC, 23.2±0.3%β-glucan and 7.8±1.0% fatty acids). M OBC was stored at 4, 22, and 40 °C for up to 12 weeks, before used inacid milk gels. A control sample without added fibre was also prepared (on the left). N = 2x4.
L OBC gels were generally the firmest of all the OBC gels, having the largest areas under
the force vs. distance curve. However, behaviour of gel firmness over the weeks varied
depending on the storage temperatures of L OBC: Gels with L OBC stored at 4 °C were the
firmest at week 8, decreasing towards week 12. Gels with L OBC stored at 40 °C grew firmer
week by week, and gels with L OBC stored at 22 °C showed no differences in firmness
56
(Figure 31). The effect of the 7 day’s storage of gels at 4 °C was most pronounced in L OBC
gels, especially at weeks 8 and/or 12 at 4 and 22 °C (Figure 28bc (L OBC) and Figure 31).
Although L OBC gels showed increasing likeness to the firmness of the control gel, L OBC
gels continued to be short, smooth and pleasant.
Figure 31. Firmness of acid milk gels with a non-heat treated oat bran concentrate (L OBC, 19.9±0.2%β-glucan and 1.3±0.2% fatty acids). L OBC was stored at 4, 22, and 40 °C for up to 12 weeks, before used inacid milk gels. A control sample without added fibre was also prepared (on the left). N = 2x4.
57
3.6 Discussion
3.6.1 Lipid degradation behaviour of OBC samples
To fully understand the effect degraded lipids have on β-glucan functionality, the different
pre-treatments (no heat treatment (H OBC), heat treatment (M OBC), no heat treatment and
defatting (L OBC)) as well as storage temperatures (4, 22, 40 °C) were chosen to create a
range of oat ingredients with different stages of lipid hydrolysis and oxidation. Regarding
different OBCs, it was hypothesised that H OBC would represent a rancid material with both
hydrolysis and oxidation products, M OBC would be a commercial industry-type oat bran
concentrate with indications of lipid oxidation in long-term storage, and due to defatting,
L OBC would have very little, if any, indications of lipid degradation. Lipid extraction
performed on L OBC was efficient and the ingredient behaved according to the hypothesis:
the changes in the lipid phase were not detectable due to low lipid concentrations. This is
why this section will concentrate on the results of H and M OBCs.
Based on previous studies, storage temperature was known to affect lipid degradation
kinetics, and it was hypothesised that 4 °C would slow down and 40 °C would accelerate the
degradation reactions (Kaur et al. 2018). The effect of storage temperature and its impact on
reaction kinetics (Labuza 1971) could be seen in the results of both H and M OBC. The
accelerating effect of higher storage temperatures on the extent of lipid degradation was
emphasized in, for example, the formation of FFA in H OBC (Figure 14) and hexanal in
M OBC (Figure 20), as well as in the degradation of tocols in both samples (Figures 17 and
18). The temperature most probably also provided favourable conditions for enzyme activity,
as samples stored at 40 °C quickly formed FFAs. However, this formation rate did not stay
as high as in the beginning, which could also be attributed to higher temperatures. FFAs
might have reacted further at a faster rate than they were formed, appearing as a slowed-
down formation rate.
H OBC had a higher total FA content than M OBC (Table 2), which probably resulted from
M OBC having a slightly higher β-glucan content. During storage, total neutral lipids
decreased in H OBC (Figure 15). TAGs were degraded to DAGs and FFAs, but as DAGs
and FFAs would react further at a faster rate than TAGs were degraded, the total amount of
neutral lipids would slowly decrease. A higher storage temperature increased the relative
58
portion of DAGs and FFAs in the samples, which is probably related to increased enzyme
activity at 40 °C.
The presence or absence of a heat treatment expectedly affected lipase activity, and thus, the
presence of hydrolytic products. H OBC was not heat treated and was the only OBC to have
its lipase still active (Table 2), which explains why it was the only OBC ingredient to form
FFAs. However, only little enzyme activity related volatiles, such as 2-pentylfuran, were
found (Figure 19), even in the stored raw material (Figure 23). The heat treatment also
affected the amounts of tocols. α-tocotrienol was the most abundant tocol in both samples,
which is typical for an oat bran ingredient (Bryngelsson et al. 2002). However, H OBC had
significantly higher concentrations of α-tocotrienol compared to M OBC both at the
beginning and at the end of the storage time (Figures 17 and 18). M OBC was heat treated,
which could explain the concentration difference, as tocols are reported to be sensitive to
heat during processing (Bryngelsson et al. 2002).
Quite surprisingly, H OBC did not show indications of lipid oxidation. Even after the
extended storage period, oxidation products, such as hexanal, in H OBC were scarce. The
extraction fibre in use could have affected the extracted volatile compounds, as previous
studies with volatiles related to oat lipid oxidation used DVB/CAR/PDMS-fibres (e.g.
Damerau et al. 2014). However, using a CAR/PDMS-fibre should not have adversely
affected volatile profiles, but either had a potentially neutral or improving effect on the
volatile extraction process (Poinot et al. 2007, Raffo et al. 2015). Furthermore, the same
extraction fibre was used in the analysis of other OBC samples and volatiles were detected
in M OBC, which supports the use of CAR/PDMS-fibre. To be sure, more studies on the
impact of DVB/CAR/PDMS- and CAR/PDMS- fibres on the extraction of oat volatile
compounds should be conducted.
The overall lipid stability of H OBC in storage was still a puzzle, as usually non-heat treated
oats are thought to become rancid quicker because of active enzymes, giving reason to
inactivate the enzymes by heat treating oats (Molteberg et al. 1996). The stability of H OBC
lipids could lie in oxidation kinetics and the presence of antioxidants. The higher
concentrations of α-tocotrienol and slower α-tocotrienol degradation rates, even during 12
weeks’ storage at 40 °C, could tell of high antioxidant activity that could protect lipids from
degradation. On the other hand, tocols also exhibit a proxidative nature, as side reactions in
tocol degradation can be prooxidative (Kamal-Eldin and Appelqvist 1996). There are also
59
other antioxidants in the bran that could possibly affect oxidation kinetics. Purified oat
β-glucan with residual phytic acid is shown to quench and bind iron and scavenge hydroxyl
radicals (Wang et al. 2017), which could participate in protecting the lipids from oxidation
and the formation of both primary and secondary oxidation products. During storage,
β-glucan molecules did not degrade and the molecular weight remained high, over 1000 kDa.
But then again, Wang et al. (2017) explain that in an aqueous food system, β-glucans with
low molecular weight were more efficient in retarding oxidation than with high molecular
weight.
The presence of antioxidants is not a fully satisfactory explanation to the stability of H OBC
lipids. When considering the location of lipids in oat bran, the lipids are mainly located inside
thick cell walls of the aleurone layer (Miller and Fulcher 2011). Compared to other studies
that experimented with non-heat treated oats, the oat samples were wholegrain flours,
meaning the lipids in their oat samples were mainly from the starchy endosperm, where the
cell walls are lighter (Grundy et al. 2018) and the lipids thus more easily accessed and
susceptible to oxidation (Molteberg et al. 1996; Heiniö et al. 2002; Lehtinen et al. 2003).
The lipid degrading enzymes from nearby tissues should be able to mobilize lipid reserves
even through thick cell walls (Lenova et al. 2010), and lipid hydrolysis, although minimal,
should promote lipid oxidation by providing substrates for LOX and autoxidation by the
formation of FFAs (Ekstrand et al. 1992). However, it might just be that in the bran
ingredient, the cell structures were still intact and lipids protected and immobile. Larger
particle sizes have been shown to protect β-glucan from degradation in sourdough bread
(Johansson et al. 2018). The authors visualized, how in coarse oat brans the cell walls were
more intact, which limited the accessibility of β-glucan to degrading enzymes. It might be
possible that in H OBC, the lipid reservois were not accessible to lipid degrading enzymes,
or to free radicals, for that matter, resulting in a more stabile OBC product. The puzzling
fact was that M OBCs were produced using the same method as for H OBC, the only
difference being the heat-treatment, which should not affect oat cell structures. However, for
M OBC, part of the processing was done in the industry, where mere processing scale aspects
might influence the end product.
M OBC showed signs of lipid oxidation as typical lipid autoxidation related volatiles (Zhou
et al. 1999a) were produced (Figures 20 and 21) and tocopherols and tocotrienols were
degraded in the course of time and progress of lipid oxidation (Figure 18). Hexanal is known
to first fill the airspace fairly quickly and then to dissipate and/or be absorbed into the oat
60
matrix (Zhou and Decker 1999; Lehto et al. 2003), which is the most probable explanation,
why hexanal amounts decreased in M OBC after 8 weeks of storage at all storage
temperatures (Figure 20). Tocols are known to quench free radicals by their own expense
(Kamal-Eldin and Appelqvist 1996), which is why with time, when the oxidation reaction
chains advanced and the amount of formed free radicals increased, tocols were consumed
and their amounts deceased. The same degrading effect on antioxidants could be seen in the
amount of total phenolic compounds in the study of Rakic et al. (2014). Even though lipid
degradation was observed in M OBC, it did not cause changes in β-glucan molecular weight
in the dry stored ingredients. It seems that β-glucan needs an aqueous food matrix to undergo
degradation (e.g. Kivelä et al. 2009b).
To verify the compatibility and thus comparability of H and M OBC, it would be beneficial
to analyse the particle sizes of both OBCs. It would also be interesting to measure LOX
activity in H and M OBCs to better understand enzyme activity or inactivity in the OBC
ingredients.
3.6.2 Effects of OBC addition on acid milk gel functionality
OBCs were included in acid milk preparations at dosages aiming to achieve the same amount
of β-glucan in each of the final gel to deliver the 1 g of β-glucan per portion, as specified in
the health claims. Adding OBC to yogurts significantly increased acidification times in a
similar way as in the study of Rosa-Sibakov et al. (2016) using xylans as dietary fibre
enrichment. This effect might be contributed to the amount of soluble components in the
OBC, as they might affect the buffering properties of the solution. In previous studies, adding
oat fibre to yogurt has had different effects on the acidification times of yogurts. In the study
done by Fernández-García et al. (1998), oat fibre did not affect the acidification time, but in
the study done by Raikos et al. (2018), β-glucan from spent brewer’s yeast reduced the
acidification times. However, it must be acknowledged that in both studies the yogurts were
acidified with LAB and the former had also sweeteners affecting culture activity in their
yogurts, which means their yogurts were even more complex systems than GDL acidified
acid milk gels, making it challenging to compare results.
The acid milk gel textures varied a lot with both added OBC type and OBC storage time.
Both acid milk gels and OBCs are complex, heterogeneous systems. Mixing the two together
makes it increasingly difficult to isolate the effect of oxidized lipids on β-glucan
61
functionality, especially as it seems that there were no changes in β-glucan MW. There might
be other interactions affecting the final result. For example, because lipids have been found
to interfere separating oat components as fractions (Sibakov et al. 2011), OBCs might still
have some starch in them. The presence of milk ingredients, such as skim milk powder, have
been found to affect the behaviour of oat starch, including pasting, textural, rheological,
freeze-thaw, and swelling properties, both during and after cooking (Kumar et al. 2018).
Therefore, to get an idea on the effect that β-glucan might have on the gel, it is fitting to
observe the possible changes the gel matrix undergoes with the OBC addition. Some changes
to be followed are the changes in water retaining properties, as they can tell of β-glucan
functionality in the acid milk gel matrix. In this study, adding OBCs to the acid milk gels
improved the water retaining properties significantly. This can be explained with previous
knowledge on the water-binding characteristics of β-glucan molecules, as they are known to
bind water and make viscous solutions (Lazaridou and Biliaderis 2007), even in acid milk
gels (Lazaridou et al. 2008). The effect of the additional fibre network in the gel and
increased viscosity resulted in no spontaneous syneresis and less or no forced syneresis
(WHC) in the gels of the experiment (Figures 26 and 27, Table 3). The effect of storing gels
could be seen as increased firmness. When the protein network rearranges and concentrates,
serum will be released, which increases the gel protein content with an outcome of a firmer
gel (Ercili-Cura et al. 2013). This could explain why control gels were more brittle than gels
with OBC, and why gels, both control gels and gels with OBCs, became denser and firmer
after 7 days’ storage in the fridge (Figures 29, 30, and 31).
The molecular weight of β-glucan also indicates β-glucan functionality. The β-glucan
molecular weight was measured in fresh and stored acid milk gels with M OBC stored 0 and
12 weeks at 40 °C. The molecular weights remained unchanged, over 1000 kDa. However,
the method used supported only molecular weights up to 1000 kDa, meaning it was not sure
if the large β-glucans had actually degraded, but were still over the detection limit. In the
study of Wang et al. (2016), nearly pure β-glucans were added to an emulsion with highly
oxidised lipids, resulting in β-glucan degradation and lowered molecular weight and
viscosity. This is not the same when adding OBC to acid milk gels with mildly degraded
lipids. OBCs are complex systems, with parts of plant tissue possibly intact, and act as the
carriers of β-glucans and lipids. This means that in the OBCs, the β-glucans are not as
accessible as in the emulsion study with nearly pure β-glucan, which already hinders the
possibility of the degradation of β-glucans. Additionally, as it is probable that the β-glucans
in the OBCs create a viscous secondary gel network within the acid milk gel (Lazaridou et
62
al. 2008), both the protein network and the secondary β-glucan network act as steric effects,
hindering the β-glucan degradation even further.
Because the solubility of β-glucans affect their concentration in the food matrix and thus the
technological and functional properties (Lazaridou and Biliaderis 2007), it would have been
of great benefit to measure β-glucan concentrations in the gels after acidification to see how
much β-glucan has been solubilized. The amount of soluble β-glucan in the product could
also affect the product’s nutritive value as a carrier of β-glucan.
3.6.3 Limitations of study
This study had some challenges with repeatability when preparing acid milk gels. The
challenges most probably arose from the properties of the raw material. The OBCs were not
fully soluble, which made it difficult to disperse the material in a repeatable manner. The
agitation was done with magnetic stirring, and speed and length of agitation seemed to have
an impact on the solubility of the material. It was only after the heat treatment, which was
designed for protein denaturation, that OBC sedimentation was somewhat restrained.
Solubility of β-glucan is a general technological challenge in high moisture food systems
(Aktas-Akyildiz et al 2018).
When performing gel penetration tests, the gel temperatures were not constant as laboratory
and fridge temperatures varied despite attempts to control them. Having some deviation in
gel firmness results is considered acceptable, however, the fluctuating gel temperatures
added to repeatability challenges when measuring gel firmness. Low repeatability makes it
difficult to draw conclusions on the effect of storage and lipid degradation on β-glucan
functionality in acid milk gels. This why the gel firmness results should be regarded as
suggestive, and used to support other results describing β-glucan functionality, for example
viscosity.
In order to complete the storage trial and all the experiments in a limited time, not all things
could be taken into account, including slight variations in OBC compositions. Although the
aim of preparing OBC with differences in their lipid degradation states was obtained, the
OBCs most probably had slight compositional variations firstly due to small differences in
β-glucan contents, but also in L OBC due to its defatting procedures. These slight
compositional variations might have affected gel properties. To maintain a constant β-glucan
63
content in the gels, the amount of OBCs in the recipe slightly varied, affecting the dry matter
content of the final product. The milk protein amounts were fixed, so the possible variations
in dry matter contents of acid milk gels were based on added OBCs.
H OBC lipids were expected to degrade rapidly, both with hydrolysis and oxidation.
However, as H OBC did not perform as expected, additional experiments were needed,
rendering the prepared H OBC batch scarce. This lead to making only one batch of acid milk
gels per time point instead of two batches, which could add to the errors both in results and
in interpretation of the results. However, the effect is assumed to be minimal, as the standard
deviations within the batches were small, and the trend of changes in gel firmness over time
were clear.
Concerning β-glucan molecular weight analysis, the reliable detection range of the method
was only up to 1000 kDa. This study would have benefited of a range at least up to 2000 kDa,
as there were indications of high molecular weight β-glucans, which could have halved in
sized without detecting it. Additionally, for volatile analyses, the reference material revealed
to have excessive lipid degradation, as the followed volatile compounds did not stay stable.
Having a clogged CAR/PDMS-fibre was ruled out by changing the fibre in the middle of the
analyses. Not being sure if the instability originates form the instrument or the reference
material might add to errors in the results.
64
4 CONCLUSIONS
A storage trial was designed to better understand the impact of lipid hydrolysis and oxidation
in oat bran concentrates on the functionality of β-glucan in acid milk gels, a yogurt type food
model. In order to design functional foods, it has been of great interest to investigate, how
oat bran concentrates with high β-glucan concentrations (~20%) function in regard to
progressive lipid degradation both in dry oat ingredients and/or in yogurt models. It was also
investigated whether different oat bran concentrates caused differences in yogurt texture and
stability.
Lipid hydrolysis was observed in non-heat treated oat bran concentrates (H OBC, 9% fatty
acids, 18% β-glucan) and lipid oxidation was observed in heat treated oat bran concentrate
(M OBC, 8% fatty acids, 23% β-glucan). However, neither caused changes in β-glucan
molecular weight in the dry stored ingredients. Defatting oat bran concentrates was efficient
and changes in the lipid phase could not be followed (L OBC, 1% fatty acids, 20% β-glucan).
Acid milk gels with added OBCs had excellent water retaining properties and were softer
and smoother than without added OBCs. Changes in gel texture were observed when using
different OBCs and when using them after storage. However, the changes might be attributed
to other factors than molecular weight of β-glucan or products from lipid degradation in the
dry stored ingredient. Assessing OBC particle sizes and the β-glucan molecular weight from
more of the acid gels would be interesting to better understand the changes in β-glucan
molecular weight and functionality in regard to lipid degradation.
There have not been similar storage studies on hybrid ingredients researching the impact of
oxidized lipids on β-glucan. Thus this study provides novel information on the interactions
of lipid degradation (hydrolysis and oxidation) and β-glucan functionality. Even when lipids
were oxidized within the dry oat ingredient, β-glucan molecules did not degrade, which
might be due to having protective cell wall structures intact in the oat bran concentrates.
65
REFERENCES
Aktas-Akyildiz E, Sibakov J, Nappa M, Hytönen E, Koksel H, Poutanen K. 2018. Extraction of soluble β-glucan from oat and barley fractions: Process efficiency and dispersion stability. J Cereal Sci 81:60–68.
[AACC] American Association for Clinical Chemistry. 1989. AACC committee adopts oat bran definition.Cereal Foods World 34:1033–1034.
Åman P, Rimsten L, Andersson R. 2004. Molecular weight distribution of β-glucan in oat-based foods. CerealChem 81:356–360. doi: 10.1094/CCHEM.2004.81.3.356.
Andersson AAM, Börjesdotter D. 2011. Effects of environment and variety on content and molecular weightof β-glucan in oats. J Cereal Sci 54:122–128. doi: 10.1016/j.jcs.2011.03.003.
Aryana KJ, Olson DW. 2017. A 100-Year Review: Yogurt and other cultured dairy products. J Dairy Sci100:9987–10013. doi: 10.3168/jds.2017-12981.
Auty MAE. 2011. Microscopy. Microstructure of milk constituents and products. In: Fuquay JW, Chief Editor.Encyclopedia of Dairy Sciences. 2nd ed. Cambridge, MA, USA: Elsevier Ltd. p 226–234.
Biermann U, Wittmann A, Groch W. 1980. Occurrence of bitter hydroxy fatty acids in oat and wheat. FettWiss Technol 82:236–24. Referenced from: Doehlert DC, Angelikousis S, Vick B. 2010. Accumulation ofoxygenated fatty acids in oat lipids during storage. Cereal Chem 87(6):532-537.
Bingham SA, Day NE, Luben R, Ferrari R, Slimani N, Norat T, Clavel-Chapelon F, Kesse E, Nieters A, BoeingH, Tjonnelans A, Overvad K, Martinez C, Dorronsoro M, Gonzalez CA, Keu TJ, Teichopoulou A, Naska A,Vineis P, Tumino R, Krogh V, Buoeno-de-Mesquita HB, Peeters PHM, Berglund G, Hallmans G, Lund E,Kaaks R, Riboli E. 2003. Dietary fibre in food and protection against colorectal cancer in the EuropeanProspective Investigation into Cancer and Nutrition. EPIC: an observational study. Lancet 361:1496–1501.doi: 10.1016/S0140-6736(03)13174-1.
Błaszczyk K, Wilczak J, Harasym J, Gudej S, Suchecka D, Królikowski T, Lange E, Gromadzka-Ostroeska J.2015. Impact of low and high molecular weight oat beta-glucan on oxidative stress and antioxidant defense inspleen of rats with LPS induced enteritis. Food Hydrocoll 51:272–280. doi: 10.1016/j.foodhyd.2015.05.025.
Brennan CS, Tudorica CM. 2008. Carbohydrate-based fat replacers in the modification of the rheological,textural and sensory quality of yoghurt: Comparative study of the utilisation of barley beta-glucan, guar gumand inulin. Int J Food Sci Technol 43:824–833. doi: 10.1111/j.1365-2621.2007.01522.x.
Brownlow S, Morais Cabral JH, Cooper R, et al. 1997. Bovine β-lactoglobulin at 1.8 Å resolution - Still anenigmatic lipocalin. Structure 5:481–495. doi: 10.1016/S0969-2126(97)00205-0.
Bryngelsson S, Dimberg LH, Kamal-Eldin A. 2002. Effects of commercial processing on antioxidants in Oats(Avena sativa L.). Cereal Chem 50:1890–1896. doi: Doi 10.1094/Cchem.2000.77.5.551.
Carter DC, Ho JX. 1994. Structure of Serum Albumin. Adv Protein Chem 45:153–203. doi: 10.1016/S0065-3233(08)60640-3. Referenced from: Considine T, Patel HA, Anema SG, Singh H, Creamer LK. 2007.Interactions of milk proteins during heat and high hydrostatic pressure treatments - A Review. Innov Food SciEmerg Technol 8:1–23. doi: 10.1016/j.ifset.2006.08.003.
Chen YC, Chen CC, Chen ST, Hsieh JF. 2016. Proteomic profiling of the coagulation of milk proteins inducedby glucono-delta-lactone. Food Hydrocoll 52:137–143. doi: 10.1016/j.foodhyd.2015.06.005.
Choe E. 2008. Effects and mechanisms of minor compounds in oil on lipid oxidation. In: Min DB, Editor. FoodLipids: Chemistry, Nutrition, and Biotechnology. 3rd ed. Boca Raton, FL, USA: CRC Press. p 449–474.
Considine T, Patel HA, Anema SG, Singh H, Creamer LK. 2007. Interactions of milk proteins during heat andhigh hydrostatic pressure treatments - A Review. Innov Food Sci Emerg Technol 8:1–23. doi:10.1016/j.ifset.2006.08.003.
66
Cosgrove JP, Church DF, Pryor WA. 1987. The kinetics of the autoxidotion of polyunsaturated fatty acids.Lipids 22:299–304.
Cui W, Wood PJ. 2000. Relationships between structural features, molecular weight and rheological propertiesof cereal β-D-glucans. In: Nishinari K, Tipton CM, Editors. Hydrocolloids: Fundamentals and Applications inFood, Biology and Medicine. Amsterdam: Elsevier Science & Technology. p 159–168.
Dalgleish DG. 2011. On the structural models of bovine casein micelles - Review and possible improvements.Soft Matter 7:2265–2272. doi: 10.1039/c0sm00806k.
Damerau A, Moisio T, Partanen R, Forssell P, Lampi AM, Piironen V. 2014. Interfacial protein engineeringfor spray-dried emulsions - Part II: Oxidative stability. Food Chem 144:57–64. doi:10.1016/j.foodchem.2013.04.006.
De Kruif CG. 1997. Skim milk acidification. J Colloid Interface Sci 185:19–25. doi: 10.1006/jcis.1996.4548.
Dello Staffolo M, Bertola N, Martino M, Bevilacqua A. 2004. Influence of dietary fiber addition on sensoryand rheological properties of yogurt. Int Dairy J 14:263–268. doi: 10.1016/j.idairyj.2003.08.004.
DeVries JW, Camire ME, Cho S, Craig s, Gordon D, Jones JM, Li B, Lineback D, Prosky L, Tungland BC.2001. The definition of dietary fiber. Cereal Foods World 46:112–129.
Dhingra D, Michael M, Rajput H, Patil RT. 2012. Dietary fibre in foods: A review. J Food Sci Technol 49:255–266. doi: 10.1007/s13197-011-0365-5.
Doehlert DC, McMullen MS. 2001. Optimizing conditions for experimental oat dehulling. Cereal Chem78:675–679. doi: 10.1094/CCHEM.2001.78.6.675.
Ekstrand B, Gangby I, Akersson G. 1992. Lipase activity in oats - distribution, pH dependance and heatinactivation. Cereal Chem 69:379–381.
Elizaquível P, Sánchez G, Salvador A, Fiszman S, Dueñas MT, López P, Fernández de Palecia P, Aznar R.2011. Evaluation of yogurt and various beverages as carriers of lactic acid bacteria producing 2-branched (1,3)-β-d-glucan. J Dairy Sci 94:3271–3278. doi: 10.3168/jds.2010-4026.
Ercili-Cura D, Lantto R, Lille M, Andberg M, Kruus K, Bucher J. 2009. Laccase-aided protein modification:Effects on the structural properties of acidified sodium caseinate gels. Int Dairy J 19:737–745. doi:10.1016/j.idairyj.2009.06.007.
Ercili-Cura D, Lille M, Legland D, Gaucel S, Poutanen K, Partanen R, Lantto R. 2013. Structural mechanismsleading to improved water retention in acid milk gels by use of transglutaminase. Food Hydrocoll 30:419–427.doi: 10.1016/j.foodhyd.2012.07.008.
[EFSA] European Food Safety Authority. 2010. Scientific Opinion on the substantiation of a health claimrelated to oat beta - glucan and lower ing blood cholesterol and reduced risk of (coronary) heart disease pursuantto Article 14 of Regulation (EC) No 1924/2006. EFSA J 8(12):1885. doi: 10.2903/j.efsa.2010.1885.
[EFSA] European Food Safety Authority. 2011. Scientific Opinion on the substantiation of health claimsrelated to beta-glucans from oats and barley and maintenance of normal blood LDL-cholesterol concentrations(ID 1236, 1299), increase in satiety leading to a reduction in energy intake (ID 851, 852), reduction of post-prandial glycaemic responses (ID 821, 824), and “digestive function” (ID 850) pursuant to Article 13(1) ofRegulation (EC) No 1924/2006. EFSA J 9(6):2207. doi: 10.2903/j.efsa.2011.2207.
Fernández-García E, McGregor JU, Traylor S. 1998. The addition of oat fiber and natural alternativesweeteners in the manufacture of plain yogurt. J Dairy Sci 81:655–663. doi: 10.3168/jds.S0022-0302(98)75620-6.
Fu R, Li J, Zhang T, Zhu T, Cheng R, Wang S, Zhang J. 2018. Salecan stabilizes the microstructure andimproves the rheological performance of yogurt. Food Hydrocoll 81:474–480. doi:10.1016/j.foodhyd.2018.03.034.
67
Gahruie HH, Eskandari MH, Mesbahi G, Hanifpour MA. 2015. Scientific and technical aspects of yogurtfortification: A review. Food Sci Hum Wellness 4:1–8. doi: 10.1016/j.fshw.2015.03.002.
Givens DI, Davies TW, Laverick RM. 2004. Effect of variety, nitrogen fertiliser and various agronomic factorson the nutritive value of husked and naked oats grain. Anim Feed Sci Technol 113:169–181. doi:10.1016/j.anifeedsci.2003.11.009.
Grundy MML, Fardet A, Tosh SM, Rich GT, Wilde PJ. 2018. Processing of oat: The impact on oat’s cholesterollowering effect. Food Funct 9:1328–1343. doi: 10.1039/c7fo02006f.
Haginaka J. 2000. Enantiomer separation of drugs by capillary electrophoresis using proteins as chiralselectors. J Chromatogr A 875:235–254. doi: 10.1016/S0021-9673(99)01168-1.
Head D, Cenkowski S, Arntfield S, Henderson K. 2011. Storage stability of oat groats processed commerciallyand with superheated steam. LWT - Food Sci Technol 44:261–268. doi: 10.1016/j.lwt.2010.05.022.
Heneen WK, Banas A, Leonova S, Carlsson AS, Marttila S, Debski H, Stymne S. 2009. Plant Signal Behav4(1):55–56.
Heneen WK, Karlsson G, Brismar K, Gummeson PO, Marttila S, Leonova S, Carlsson AS, Bafor M, Banas A,Mattsson B, Debski H, Stymne S. 2008. Fusion of oil bodies in endosperm of oat grains. Planta 228:589–99.
Heiniö RL, Lehtinen P, Oksman-Caldentey KM, Poutanen K. 2002. Differences between sensory profiles anddevelopment of rancidity during long-term storage of native and processed oat. Cereal Chem J 79:367–375.doi: 10.1094/CCHEM.2002.79.3.367.
Hitayezu R, Baakdah MM, Kinnin J, Henderson K, Tsopmo A. 2015. Antioxidant activity, avenanthramidesand phenolic acid contents of oat milling fractions. J Cereal Sci 63:35–40.
Holt C, Sawyer L. 1993. Caseins as rheomorphic proteins: Interpretation of primary and secondary structuresof the αS1-, β-, and κ-caseins. Chem Soc Faraday Trans 89:2683–2692. doi: 10.1039/ft9938902683.
Hyde S, Verdin D. 1968. Oxidation of methyl oleate induced by 60Co γ-radiation. Part 2. - Emulsions of methyloleate in water. Trans Faraday Soc 155–162.
Ingold KU. 1969. Peroxy radicals. Acc Chem Res 2:1–9. doi: 10.1021/ar50079a001.
Johansson DP, Andersson R, Alminger M, Landber R, Langton M. 2018. Larger particle size of oat braninhibits degradation and lowers extractability of β-glucan in sourdough bread – Potential implications forcholesterol-lowering properties in vivo. Food Hydrocoll 77:49–56. doi: 10.1016/j.foodhyd.2017.09.016.
Juhola S, Klein N, Käyhkö J, Schmid Neset TS. 2017. Climate change transformations in Nordic agriculture?J Rural Stud 51:28–36. doi: 10.1016/j.jrurstud.2017.01.013.
Kaláb M, Allan-Wojtas P, Phipps-Todd BE. 1983. Development of microstructure in set-style nonfat yogurt -A review. J Food Struct 2:51–66.
Kamal-Eldin A, Appelqvist L-Å. 1996. The chemistry and antioxidant properties of tocopherols andtocotrienols. Lipids 31:671–701.
Kamal-Eldin A, Görgen S, Pettersson J, Lampi A-M. 2000. Normal-phase high-performance liquidchromatography of tocopherols and tocotrienols: Comparison of different chromatographic columns. JChromatogr A 881:217–227. doi: 10.1016/S0021-9673(99)01346-1.
Kaukonen O, Sontag-Strohm T, Salovaara H, Lampi AM, Sibakov J, Loponen J. 2011. Foaming of differentlyprocessed oats: Role of nonpolar lipids and tryptophanin proteins. Cereal Chem 88:239–244. doi:10.1094/CCHEM-11-10-0154.
Kaur J, Whitson A, Ashton J, Katopo L, Kasapis S. 2018. Effect of ultra high temperature processing andstorage conditions on phenolic acid, avenanthramide, free fatty acid and volatile profiles from Australian oatgrains. Bioact Carbohydrates Diet Fibre 15:21–29. doi: 10.1016/j.bcdf.2016.09.002.
68
Keogh GF, Cooper GJS, Mulvey TB, McArdle BH, Coles GD, Monro JA, Poppitt SD. 2003. Randomizedcontrolled crossover study of the effect of a highly β-glucan-enriched barley on cardiovascular disease riskfactors in mildly hypercholesterolemic men. Am J Clin Nutr 78:711–718.
Kerr JA. 1966. Bond dissociation energies by kinetic methods. Chem Rev 66:465–500. doi:10.1021/cr60243a001.
Kim HJ, Min DB. 2008. Tocopherol stability and prooxidant mechanisms of oxidized tocopherols in lipids.435–447. doi: doi:10.1201/9781420046649.ch16.
Kivelä R, Gates F, Sontag-Strohm T. 2009a. Degradation of cereal beta-glucan by ascorbic acid inducedoxygen radicals. J Cereal Sci 49:1–3. doi: 10.1016/j.jcs.2008.09.003.
Kivelä R, Henniges U, Sontag-Strohm T, Potthast A. 2012. Oxidation of oat β-glucan in aqueous solutionsduring processing. Carbohydr Polym 87:589–597. doi: 10.1016/j.carbpol.2011.08.028.
Kivelä R, Nyström L, Salovaara H, Sontag-Strohm T. 2009b. Role of oxidative cleavage and acid hydrolysisof oat beta-glucan in modelled beverage conditions. J Cereal Sci 50:190–197. doi: 10.1016/j.jcs.2009.04.012.
Klensporf D, Jeleń HH. 2008. Effect of heat treatment on the flavor of oat flakes. J Cereal Sci 48:656–661.doi: 10.1016/j.jcs.2008.02.005.
Knudsen JC, Skibsted LH. 2010. High pressure effects on the structure of casein micelles in milk as studied bycryo-transmission electron microscopy. Food Chem 119:202–208. doi: 10.1016/j.foodchem.2009.06.017.
Konak ÜI, Ercili-Cura D, Sibakov J, Sontag-Strohm T, Certel M, Loponen J. 2014. CO2-defatted oats:Solubility, emulsification and foaming properties. J Cereal Sci 60:37–41. doi: 10.1016/j.jcs.2014.01.013.
Kumar L, Brennan M, Zheng H, Brennan C. 2018. The effects of dairy ingredients on the pasting, textural,rheological, freeze-thaw properties and swelling behaviour of oat starch. Food Chem 245:518–524. doi:10.1016/j.foodchem.2017.10.125.
Laguerre M, López Giraldo LJ, Lecomte J, Figueroa-Espinoza MC, Baréa B, Weiss J, Decker EA, VilleneuveP. 2009. Chain length affects antioxidant properties of chlorogenate esters in emulsion: the cutoff theory behindthe polar paradox. J Agric Food Chem 57:11335–11342. doi: 10.1021/jf9026266.
Lampi A-M, Damerau A, Li J, Moisio T, Partanen R, Forssel P, Piironen V. 2015. Changes in lipids and volatilecompounds of oat flours and extrudates during processing and storage. J Cereal Sci 62:102–109. doi:10.1016/j.jcs.2014.12.011.
Lascorz M, Drapron R. 1987. Lipoxygenase-lipase system in oat hulls in relation to dormancy. Phytochemistry26:14–16.
Law AJR, Leaver J. 1998. Effects of acidification and storage of milk on dissociation of bovine casein micelles.J Agric Food Chem 46:5008–5016. doi: 10.1021/jf980748p.
Lazaridou A, Biliaderis CG. 2007. Molecular aspects of cereal β-glucan functionality: Physical properties,technological applications and physiological effects. J Cereal Sci 46:101–118. doi:10.1016/J.JCS.2007.05.003.
Lazaridou A, Serafeimidou A, Biliaderis CG, Moschakis T, Tzanetakis Nl. 2014. Structure development andacidification kinetics in fermented milk containing oat β-glucan, a yogurt culture and a probiotic strain. FoodHydrocoll 39:204–214. doi: 10.1016/j.foodhyd.2014.01.015.
Lazaridou A, Vaikousi H, Biliaderis CG. 2008. Impact of mixed-linkage. 1→3, 1→4. β-glucans on physicalproperties of acid-set skim milk gels. Int Dairy J 18:312–322. doi: 10.1016/j.idairyj.2007.08.005.
Lee WJ, Lucey JA. 2004. Structure and physical properties of yogurt gels: Effect of inoculation rate andincubation temperature. J Dairy Sci 87:3153–3164. doi: 10.3168/jds.S0022-0302(04)73450-5.
69
Lehtinen P, Kiiliäinen K, Lehtomäki I, Laakso S. 2003. Effect of heat treatment on lipid stability in processedoats. J Cereal Sci 37:215–221. doi: 10.1006/jcrs.2002.0496.
Lehtinen P, Laakso S. 2004. Role of lipid reactions in quality of oat products. Agric Food Sci 13:88–99. doi:10.2137/1239099041838085.
Lehto S, Laakso S, Lehtinen P. 2003. Enzymatic oxidation of hexanal by oat. J Cereal Sci 38:199–203. doi:10.1016/S0733-5210(03)00028-6.
Leonova S, Grimberg A, Marttila S, Stymne S, Carlsson AS. 2010. J Exp Bot 61(11):3089–3099.
Liu D, Yu Y, Zhang J, Liu X, Wang M, Hemar Y, Regenstein JM, Zhou P. 2017. Biochemical and physico-chemical changes of skim milk during acidification with glucono-δ-lactone and hydrogen chloride. FoodHydrocoll 66:99–109. doi: 10.1016/j.foodhyd.2016.12.030.
Liukkonen K, Kaukovirta-Norja A, Laakso S. 1993. Elimination of lipid hydrolysis in aqueous suspensions ofoat flour. J Cereal Sci 17:225–265.
Lucey JA, Singh H. 1997. Formation and physical properties of acid milk gels: A review. Food Res Int 30:529–542. doi: 10.1016/S0963-9969(98)00015-5.
Lucey JA, Tamehana M, Singh H, Munro PA. 1998. A comparison of the formation, rheological propertiesand microstructure of acid skim milk gels made with a bacterial culture or glucono-δ-lactone. Food Res Int31:147–155. doi: 10.1016/S0963-9969(98)00075-1.
Luhaloo M, Ma A, Andersson R, Óman P. 1998. Compositional analysis and viscosity measurements ofcommercial oat brans. 142:142–148.
Mäkelä N, Sontag-Strohm T, Maina NH. 2015. The oxidative degradation of barley β-glucan in the presenceof ascorbic acid or hydrogen peroxide. Carbohydr Polym 123:. doi: 10.1016/j.carbpol.2015.01.037.
Makki K, Deehan EC, Walter J, Bäckhed F. 2018. The impact of dietary fiber on gut microbiota in host healthand disease. Cell Host Microbe 23:705–715. doi: 10.1016/j.chom.2018.05.012.
Mälkki Y, Myllymäki O, Teinilä K, Koponen S, inventors. Avena Oy, Helsinki, applicant. 19.04.2001. Methodfor preparing an oat product and a foodstuff enriched in the content of β-glucan. WO 01/26479 A1.
Marchin S, Putaux JL, Pignon F, Léonil J. 2007. Effects of the environmental factors on the casein micellestructure studied by cryo transmission electron microscopy and small-angle x-ray scattering/ultrasmall-anglex-ray scattering. J Chem Phys 126:. doi: 10.1063/1.2409933.
Mårtensson O, Andersson C, Andersson K, Öste R, Holst O. 2001. Formulation of an oat-based fermentedproduct and its comparison with yoghurt. J Sci Food Agric 81:1314–1321. doi: 10.1002/jsfa.947.
Martin F, Cayot N, Marin A, Journaux L, Cayot P, Gervais P, Cachon R. 2009. Effect of oxidoreductionpotential and of gas bubbling on rheological properties and microstructure of acid skim milk gels acidified withglucono-δ-lactone. J Dairy Sci 92:5898–5906. doi: 10.3168/jds.2009-2491.
McCleary BV, Codd R. 1991. Measurement of (1-3)(1-4)-β-D-glucan in barley and oats: A streamlinedenzymic procedure. J Sci Food Agric 55:303-312.
Meydani M. 2009. Potential health benefits of avenanthramides of oats. Nutrition Reviews 67(12)731-735.
Miller SS, Fulcher RG. 2011. Microstructure and chemistry of the oat kernel. In: Webster FH, Wood PJ,Editors. OATS: Chemistry and Technology. 2nd ed. St. Paul, Minnesota: American Association of CerealChemists, Inc. p 77–94. Referenced from: Grundy MML, Fardet A, Tosh SM, Rich GT, Wilde PJ. 2018.Processing of oat: The impact on oat’s cholesterol lowering effect. Food Funct 9:1328–1343. doi:10.1039/c7fo02006f.
Molteberg EL, Magnus EM, Bjørge JM, Nilsson A. 1996. Sensory and chemical studies of lipid oxidation inraw and heat-treated oat flours. Cereal Chem 73:579–587.
70
[LUKE] Natural Resources Institute Finland. 2017. Cereals balance sheet, 2016 – 2017 (provisional) and 2015- 2016 final figures. Available at: http://stat.luke.fi/en/cereals-balance-sheet-2016-2017-provisional-and-2015-2016-final-figures_en. Accessed 30 Apr 2018.
[NNR] Nordic Council of Ministers. 2012. Nordic Nutrition Recommendations 2012. Integrating nutrition andphysical activity, 5th edn. doi: 10.6027/Nord2014-002.
Parker JK, Hassell GME, Mottram DS, Guy RCE. 2000. Sensory and instrumental analyses of volatilesgenerated during the extrusion cooking of oat flours. J Agric Food Chem 48:3497–3506. doi:10.1021/jf991302r.
Peltonen-Sainio P, Lehtinen P, Kontturi M, Rajala A, Kirkkari AM. 2004. Impact dehulling oat grain toimprove quality of on-farm produced feed 2. Groat breakage and storability. Agric Food Sci 13:29–38. doi:10.2137/1239099041837950.
Peterson DM. 2001. Oat antioxidants. J Cereal Sci 33:115–129. doi: 10.1006/jcrs.2000.0349.
Piazza GJ, Nuñez A, Foglia TA. 2003. Epoxidation of fatty acids, fatty methyl esters, and alkenes byimmobilized oat seed peroxygenase. J Mol Catal B Enzym 21:143–151. doi: 10.1016/S1381-1177(02)00122-4.
Poinot P, Grua-Priol J, Arvisenet G, Rannous C, Semenou M, Le Bail A, Prost C. 2007. Food Res Int 40:1170–1184. doi:10.1016/j.foodres.2007.06.011.
Raffo A, Carcea M, Castagna C, Magri A. 2015. J Chromatogr A 1406:266–278.
Raikos V, Grant SB, Hayes H, Ranawana V. 2018. Use of β-glucan from spent brewer’s yeast as a thickenerin skimmed yogurt: Physicochemical, textural, and structural properties related to sensory perception. J DairySci 101:1–11. doi: 10.3168/jds.2017-14261.
Rakić S, Janković S, Marčetić M, Živković D, Kuzevski J. 2014. The impact of storage on the primary andsecondary metabolites, antioxidant activity and digestibility of oat grains. Avena sativa). J Funct Foods 7:373–380. doi: 10.1016/j.jff.2014.01.022.
Regand A, Chowdhury Z, Tosh SM, Wolever TMS, Wood P. 2011. The molecular weight, solubility andviscosity of oat beta-glucan affect human glycemic response by modifying starch digestibility. Food Chem129:297–304. doi: 10.1016/j.foodchem.2011.04.053.
Reische DW, Lillard DA, Eitenmiller RR. 2008. 15 Antioxidants. In: Akoh CC, Min DB, editorss. Food Lipids:Chemistry, Nutrition, and Biotechnology. 3rd ed. Boca Raton, FL, USA: CRC Press. p 409–433.
Rinaldi L, Rioux LE, Britten M, Turgeon SL. 2015. Invitro bioaccessibility of peptides and amino acids fromyogurt made with starch, pectin, or β-glucan. Int Dairy J 46:39–45. doi: 10.1016/j.idairyj.2014.09.005.
Robinson DS, Wu Z, Domoney C, Casey R. 1995. Lipoxygenases and the quality of foods. Food Chem 54:33–43. doi: 10.1016/0308-8146(95)92659-8.
Roefs SPFM, Van Vliet T. 1990. Structure of acid casein gels 2. Dynamic measurements and type of interactionforces. Colloids and Surfaces 50:161–175. doi: 10.1016/0166-6622(90)80260-B.
Rosa-Sibakov N, Hakala TK, Sözer N, Nordlund E, Poutanen K, Aura AM. 2016. Birch pulp xylan works asa food hydrocolloid in acid milk gels and is fermented slowly in vitro. Carbohydr Polym 154:305–312. doi:10.1016/j.carbpol.2016.06.028.
Rousseau D. 2000. Fat crystals and emulsion stability - A review. Food Res Int 33:3–14. doi: 10.1016/S0963-9969(00)00017-X.
Ryynänen M, Lampi A-M, Salo-Väänänen P, Ollilainen V, Piironen V. 2004. A small-scale sample preparationmethod with HPLC analysis for determination of tocopherols and tocotrienols in cereals. J Food Compos Anal17:749–765. doi: 10.1016/j.jfca.2003.09.014.
71
Sah BNP, Vasiljevic T, McKechnie S, Donkor ON. 2016. Physicochemical, textural and rheological propertiesof probiotic yogurt fortified with fibre-rich pineapple peel powder during refrigerated storage. LWT - Food SciTechnol 65:978–986. doi: 10.1016/j.lwt.2015.09.027.
Sahan N, Yasar K, Hayaloglu AA. 2008. Physical, chemical and flavour quality of non-fat yogurt as affectedby a β-glucan hydrocolloidal composite during storage. Food Hydrocoll 22:1291–1297. doi:10.1016/j.foodhyd.2007.06.010.
Schaich KM, Shahidi F, Zhong Y, Eskin NAM. 2013. Lipid Oxidation. In: Eskin MNA, Shahidi F, Editors.Biochemistry of Foods. 3rd ed. Amsterdam: Elsevier. p 419–478.
Schwartz H, Ollilainen V, Piironen V, Lampi A-M. 2008. Tocopherol, tocotrienol and plant sterol contents ofvegetable oils and industrial fats. J Food Compos Anal 21:152–161. doi: 10.1016/j.jfca.2007.07.012.
Shahidi F, Wanasundara PKJPD. 2008. Extraction and Analysis of Lipids. In: Akoh CC, Min DB, Editors.Food Lipids. 3rd edn. Boca Raton, FL, USA: CRC Press. p 125–156.
Sibakov J. 2014. Processing of oat dietary fibre for improved functionality as a food ingredient [PhD Thesis].VTT Science 67. Espoo: Aalto University School of Chemical Technology. 116 p.
Sibakov J, Myllymäki O, Holopainen U, Kaukovirta-Norja A, Hietanimi V. Pihlava JM, Poutanen K, LehtinenP. 2011. Lipid removal enhances separation of oat grain cell wall material from starch and protein. J Cereal Sci54:104–109. doi: 10.1016/J.JCS.2011.04.003.
Sibakov J, Myllymäki O, Suortti T, Kaukovirta-Norja A, Lehitnen P, Poutanen K. 2013. Comparison of acidand enzymatic hydrolyses of oat bran β-glucan at low water content. Food Res Int 52:99–108. doi:10.1016/j.foodres.2013.02.037.
Sonnenburg ED, Sonnenburg JL. 2014. Starving our microbial self: The deleterious consequences of a dietdeficient in microbiota-accessible carbohydrates. Cell Metab 20:779–786. doi: 10.1016/j.cmet.2014.07.003.
Soupas L, Huikko L, Lampi A-M, Piironen V. 2005. Esterification affects phytosterol oxidation. Eur J LipidSci Technol 107:107–118. doi: 10.1002/ejlt.200401096.
Srigley CT, Mossoba MM. 2017. Current analytical techniques for food lipids. Food Saf Innov Anal Tools SafAssess 33–64. doi: 10.1002/9781119160588.ch3.
Suortti T. 1993. Size Exclusion chromatographic determination of b-glucan with post column reactiondetection, J. Chromatogr 633:105-110.
Tiwari U, Cummins E. 2009. Factors influencing β-glucan levels and molecular weight in cereal basedproducts. Cereal Chem 86:290–301. doi: 10.1094/CCHEM-86-3-0290.
Tomic N, Dojnov B, Miocinovic J, Tomasevic I, Smigic N, Djekic I, Vujcic Z. 2017. Enrichment of yoghurtwith insoluble dietary fiber from triticale – A sensory perspective. LWT - Food Sci Technol 80:59–66. doi:10.1016/j.lwt.2017.02.008.
[VRN] Valtion Ravitsemusneuvottelukunta. 2014. Terveytta ruoasta - Suomalaiset ravitsemussuositukset2014, 4th edn. Valtion ravitsemusneuvottelukunta, Tampere.
Vanegas-Azuero A-M, Gutiérrez L-F. 2018. Physicochemical and sensory properties of yogurts containingsacha inchi (Plukenetia volubilis L.) seeds and β-glucans from Ganoderma lucidum. J Dairy Sci 101:1020–1033. doi: 10.3168/jds.2017-13235.
Walstra P, Wouters JTM, Geurts TJ. 2006. Dairy Science and Technology. 2nd ed. Boca Raton, FL, USA: CRCPress. 782 p.
Wang YJ, Maina NH, Ekholm P, Lampi AM, Sontag-Strohm T. 2017. Retardation of oxidation by residualphytate in purified cereal β-glucans. Food Hydrocoll 66:161–167. doi: 10.1016/j.foodhyd.2016.11.019.
72
Wang YJ, Mäkelä N, Maina NH, Lampi AM, Sontag-Strohm T. 2016. Lipid oxidation induced oxidativedegradation of cereal beta-glucan. Food Chem 197:1324–1330. doi: 10.1016/j.foodchem.2015.11.018.
Wang Y. 2018. Role of lipids and phytate in oxidative stability of cereal beta-glucan [PhD] Thesis. EKT Series.Helsinki: University of Helsinki. 68 p.
Whitehead A, Beck EJ, Tosh S, Wolever TMS. 2014. Cholesterol-lowering effects of oat beta-glucan: A meta-analysis of randomized controlled trials. Am J Clin Nutr 100:1413–21. doi: 10.3945/ajcn.114.086108.
Wolever T, Susan TM, Alison GL, Brand-Miller J, Duncan AM, Hart V, Lamarche B, Thomson BA, Duss R,Wood P. 2010. Physicochemical properties of oat beta-glucan influence its ability to reduce serum LDLcholesterol in humans: a randomised clinical control trial. Am J Clin Nutr 92:723–732. doi: 10.3945.
Yang Z, Piironen V, Lampi A-M. 2017. Lipid-modifying enzymes in oat and faba bean. Food Res Int 100:335–343. doi: 10.1016/j.foodres.2017.07.005.
Zhou M, Robards K, Glennie-Holmes M, Helliwell S. 1999a. Oat Lipids. JAOCS, J Am Oil Chem Soc 76:159–169. doi: 10.1007/s11746-999-0213-1.
Zhou M, Robards K, Glennie-Holmes M, Helliwell S. 1999b. Analysis of volatile compounds and theircontribution to flavor in cereals. J Agric Food Chem 47:3941–3953. doi: 10.1021/jf990428l.
Zhou S, Decker EA. 1999. Ability of amino acids, dipeptides, polyamines, and sulfhydryls to quench hexanal,a saturated aldehydic lipid oxidation product. J Agric Food Chem 47:1932–1936. doi: 10.1021/jf980939s.
Zhu F, Du B, Xu B. 2016. A critical review on production and industrial applications of beta-glucans. FoodHydrocoll 52:275–288. doi: 10.1016/j.foodhyd.2015.07.003.
73
APPENDIX
Appendix 1. An exemplary graph for the analysis of gel firmness with TA
A deformation test was performed on 40 g aliquots of gel, acidified in plastic cups (ø 3.5 cm), pH 4.6±1.0, by penetrating the gel with a ø 25 mm cylinder probe at a constant speed of1.0 mm/s to a distance of 75% of the sample height. The area under the force vs. distance curve (up to 15 mm) was determined and taken as a measure of gel firmness