Artigoconfitados
Artigoconfitados
Artigoconfitados
net/publication/296058973
CITATIONS READS
23 5,318
5 authors, including:
All content following this page was uploaded by Alice Vilela on 06 July 2016.
KEYWORDS ABSTRACT
Candying, osmotic dehydration,
fructooligosaccharides, galactooligosacchar- Candied fruits produced with healthier ingredients for obtaining low glycemic index, low
ides, fructose, sorbitol and maltitol, sensorial energy and/or dietary fiber enriched products were develop by substituting the
evaluation, textural characteristics traditionally used osmotic dehydration sugars sucrose and glucose by fructose, sorbitol,
4
maltitol, fructooligosaccharides and galactooligosaccharides. Quality of candied fruits was
Corresponding author.
evaluated chemically, rheologically and sensorially. For all fruits the final composition was
TEL: 1351-25-035-0592;
FAX: 1351-35-935-0480;
dependent on the osmotic dehydration agent used and, in all cases, the final composition
EMAIL: [email protected] show that fruits can be impregnated successfully with the healthier ingredients.
Each of the different osmotic dehydration agents significantly influenced the
Received for Publication September 9, 2015 texture and sensorial profile of the candied fruits, but the changes were dependent
Accepted for Publication January 24, 2016 on the fruits, with the texture and sensorial quality of some products being better
Published online Article Accepted on January and other poorer than the commercial candied fruits. For all fruits tested, the
27, 2015 osmotic dehydration agent sorbitol imprinted peculiar characteristics. The
chemical composition of the final products show that candied fruits obtained with
doi:10.1111/jtxs.12177
these alternative osmotic dehydration agents are suitable healthier alternatives of
the traditional candied fruits. Still, only a consumer’s trial would assess the
palatability of the different fruits/osmotic dehydration agents tested.
PRACTICAL APPLICATIONS
One of the main objectives of food processing is the preservation of perishable foods.
Candying of fruits results in a decrease of water activity by osmotic dehydration,
prolonging fruits shelf-life. This process changes the textural and sensorial properties but
also has an impact on its dietary value, due to sucrose and glucose used in candying,
accused for being responsible for various illness. Therefore, the candying industry is
looking for substitutes, those so called “healthy ingredients” that can be used for
osmotic dehydration and yield products with similar or even better quality properties
(structural, textural and organoleptic). Fructose, sorbitol, maltitol, and dietary fiber
oligosaccharides, are possible candidates for substitution of sucrose in the candying
industry. With this work we aim to increase the knowledge about modifications at the
textural and sensorial level occurring on these fruits after the candying process to
adequately control candying and improve the quality of candied fruits.
INTRODUCTION
fruits results in a decrease of water activity by osmotic dehy-
One of the main objectives of food processing is the preser- dration, which slows down deteriorative reactions and
vation of perishable foods in a stable form that can be stored increases fruits microbial stability, hence prolonging fruits
for consumption after the production period. Candying of shelf-life. Candying changes the textural and sensorial
properties of the final products, as fruits and vegetables are power compared with sucrose is reduced to 1/3 and the
enriched with the osmotic dehydration agent, traditionally a energy is reduced to 8.37 kJ g21 (Guggisberg et al. 2011).
mixture of sucrose and glucose, this will impact not only its GOS as functional food constituents play a special role as
sensory properties, but also has an impact on its dietary prebiotics in gastrointestinal tract. A rapid uptake and fer-
value (Torreggiani and Bertolo 2001). mentation of available carbohydrates contribute to main-
Candied fruits and vegetables can be added to foods such taining the micro biota in the human small intestine
as desserts, confectionery, and bakery products. After addi- (Zoetendal et al. 2012).
tional drying, they can also be used as components of cereals With this work we aim of increasing the knowledge about
or snacks for direct consumption (Torreggiani and Bertolo modifications at the texture and sensory level occurring on
2001). Nevertheless sucrose, the main sugar used in candying fruits after the candying process (osmotic dehydration with
industry, due to its nutritional properties has been blamed for fructose, sorbitol, maltitol, FOS and GOS). This knowledge is
various illness, including hyperactivity in children, type 2 dia- needed to adequately control the candying process and improve
betes, dyslipemia and cardiovascular diseases (Barnard et al. the quality of candied fruits, namely with desirable texture and
1998). In view of this healthy issues, the candying industry is sensory characteristics, which is decisive for consumer accep-
looking for substitutes of sucrose, those so called “healthy tance on the market. Candied fruits with healthy ingredients
ingredients” that can be used for osmotic dehydration and can be used as energetic delicacy and healthy goody.
yield a product with similar or even better quality properties,
namely structural, textural and organoleptic properties.
Fructose, sorbitol, maltitol, and dietary fiber oligosaccha- MATERIAL AND METHODS
rides, namely fructooligosaccharides (FOS) and galactooligo-
saccharides (GOS) are possible candidates for substitution of Fruit Material and Sample Preparation
sucrose in the candying industry. Fructose is a simple sugar
Pineapple, orange-peel and citron samples for osmotic dehy-
found in many foods and is one of the three most important
dration were kindly supplied by the company Douromel
blood sugars along with glucose and galactose. Fructose is the
(Tabuaço, Portugal). Fruit samples were supplied stabilized
sweetest naturally occurring sugar and is often recommended
in a sodium chloride solution (5%) containing sodium
for people with diabetes mellitus or hyperglycemia, because it
metabisulfite (75 mg/L). Before candying they were thor-
has a very low Glycemic Index (GI) relative to sucrose.
oughly desalted by immersion in tap water for 12 h with five
Polyols like sorbitol and maltitol, in contrast to glucose
which is absorbed in the small intestine by active transport, water renewals (total time 60 h). Osmotic dehydration
are absorbed relatively slowly by “passive transport” or diffu- agents (fructose, sorbitol, maltitol, and actilight [FOS]) used
sion. As a result only a limited fraction of them pass throw were supplied by Brenntag (Oporto, Portugal) and Vivinal
the intestinal membrane. Maltitol is absorbed in amounts of GOS was a kind gift of Friesland Foods Domo (Holland).
50–75% and his energy is about 30–50% lower than that of For the osmotic dehydration, fructose, sorbitol, maltitol,
sucrose (Dobreva et al. 2013). Control of blood glucose, lip- FOS and GOS solutions with 65 Brix were used at 60C,
ids and weight are the three major goals of diabetes manage- according to the methodology described by Abra~ao et al.
ment today. Sorbitol is slowly absorbed. Therefore, when (2013). Pineapple samples were cut into ring slices; orange-
sorbitol is used, the rise in blood glucose and the insulin peel into small pieces of about 2.0 cm in length and 4.0 cm in
response associated with the ingestion of glucose is signifi- width and citron samples into pieces of about 6.0 3 6.0 cm.
cantly reduced. The reduced caloric value (2.6 calories per
gram versus 4.0 for sugar) of sorbitol is consistent with the Water Activity (aw) and Weight Gain (WG)
objective of weight control. Products sweetened with sorbitol
in place of sugar may be useful in providing a wider variety The weight gain (WG) was calculated according to the fol-
of reduced calorie and sugar free choices to people with dia- lowing equations:
betes (Calorie Control Council 2015).
FOS has interesting technological properties both as a wsample
t 2 W0sample
WG5 (1)
low-caloric sweetener and bulking agent, as well as a fat sub- W0sample
stitute. In addition, it presents beneficial effects on human
physiology such as low cariogenic sugar substitute, nondi- Moisture was determined gravimetrically after freeze-
gestibility and modification of the colonic microflora. FOS drying samples until constant weight.
presents a sweet taste and prebiotic properties and is pro- Water activity (aw) was measured using a LabSwift-aw
duced by an enzymatic transformation of sucrose to short- (Novasina, Switzerland) hygrometer. Measurements were
chain fructooligosaccharides (scFOS), similar to short-chain made at room temperature (20C). The effect of the different
inulin, with no aftertaste or artificial. The relative sweetening ingredients used on water activity was evaluated by
determining the Ks constant, which is supposed to be some- PA100 pre-column (Dionex) using an isocratic elution
what related with the chemical structure of nonelectrolyte with a 60 mM NaOH solution containing 15 mM sodium
solute, by applying Eq. (2) developed by Norrish (1966) on acetate and 2 mM Ba(OH)2. The injection volume was 25
the basis of a very simple Eq. (3) for calculation of activity 3L, the flow rate was 0.7 mL/min and the column tem-
coefficients proposed by Hildebrand and Scott (1962) which perature was maintained at 358C during the run. All elu-
states that for an aqueous solution, ents were kept under nitrogen to reduce carbonate
buildup and biological contamination. The electrochemi-
2
aw 5Xw e2Ks Xs (2) cal detector consisted of an Au working electrode, Ag/
AgCl reference electrode, and Ti counter electrode. The
ln c 5Ks Xs2 (3)
ED cell waveform was 10.1 V from 0.00 to 0.40 s, then
where c is the activity coefficient of water, aw is the measured 22.0 V from0.41 to 0.42 s, and a ramp 22.0 to 10.6 V
water activity, Xw is the mole fraction of water, Xs is the mole from 0.42 to 0.43 s, followed by 20.1 V from 0.44 to
fraction of solute (ingredient). For the Actilight ingredient, 0.50 s (end of cycle). The integration region was from
composed by a mixture of different fructooligosaccharides 0.2 s to 0.4 s. For quantification of sucrose, fructose, glu-
(FOS), the apparent molecular weight was used (617 g/mol) cose, sorbitol and maltitol, galactose and lactose, a stand-
taking into account the relative amounts of the different oli- ard curve was used with pure standards. For actilight, and
gosaccharides present (glucose 1 fructose 1 sucrose <4%, vivinal GOS, as these ingredients were composed by a
GF2 (oligosaccharide containing one glucose and 2 fructose mixture of short chain fructoligosaccharides plus small
units) 30.38%, GF3 50.18% and GF4 15.69%) (Kr ol and amounts of glucose, fructose and sucrose for actilight and
Grzelak 2006), for GOS an apparent molecular weight of galactose, glucose, lactose and short chain galactooligosac-
417.8 g/mol was used taking into account the Vivinal GOS charides, and the lack of pure standards for all oligossa-
composition (Galactose, glucose, lactose, Degree of Polymer- charides, the ingredients were used for calibration,
ization (DP) varying from 2 to >4: (DP2) GOS, DP3 GOS, corrected for the amount of glucose, fructose and sucrose
DP4 GOS and DP >4 GOS) (Coulier et al. 2009) and for present in actilight and for galactose, glucose and lactose
polydextrose a mean molecular weight of 2000 g/mol was on Vivinal GOS.
used (Craig 2001). Standard ingredient solutions of concen- The caloric value was calculated by using the sugar com-
tration between 40-70% were prepared unless the ingredient position determined by sugar analysis and the energy con-
solubility did not allow it, and the water activity was meas- version values of the 2008/100/CE Directive. For FOS, due to
ured as described above. Three measurements were made for the limitations in quantification described previously, the
each sample. caloric value was calculated by estimating the amount of
FOS by subtracting the solid content of fruits, water content
and fructose present to 100%.
Sugar Composition and Energy
For sugar composition and energy determinations, three
For the quantification of sugars (maltitol, sucrose, fruc- measurements were made for each sample.
tose, glucose, fructooligosaccharides and galactoglucooli-
gosaccharides) on the final osmodehydrated fruits, 1.00 g
of dried fruit was homogenized (Ultra-turrax, IKA T25 Color Measurements
digital) with 50 mL of water. After diluting the homoge- Osmodehydrated fruit pieces were tested in a Hunterlab col-
nate with water to 200 mL, the content was further diluted orimeter to obtain the a*, b* and L* parameters as an aver-
and analyzed by high performance anion exchange chro- age of three measurements. The equipment was calibrated
matography with pulsed amperometric detection with a white standard (L* 5 97.71; a* 5 20.59 and
(HPAEC-PAD, ICS-3000, Dionex). For the analysis of b* 5 2.31) and each sample was scanned at three different
sucrose, glucose, fructose, sorbitol and maltitol, the sepa-
locations. From these values, chroma (C*) was calculated
ration was performed with a CarboPac PA-20 column
according to the Eq. (4) (Hunt 1991):
(150 mm 3 3 mm) with a CarboPac PA20 precolumn
(Dionex) using an isocratic elution with a 10 mM NaOH qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 ffi
solution containing 2 mM Ba(OH)2. The injection volume C 5 a Þ 1 ð b Þ 2 (4)
was 5 lL, the flow rate was 0.3 mL/min and the column
temperature was maintained at 35C during the run. For Difference between two colors or DE* was given by Eq.
the analysis of fructooligosaccharides and galactoglucooli- (5) (Hunt 1991):
gosaccharides the separation was performed with a Carbo- 2 2 2
Pac PA-100 column (250 mm 3 3 mm) with a CarboPac DE 5 冑 L1 2L2 1 a1 2a2 1 b1 2b2 (5)
Rheological Properties and Texture Profile tests, with individual booths for each panelist. A sensory
Analysis vocabulary was developed over two sessions lasting 1.5 h.
The final vocabulary consisted of 23 attributes for pineaplle
Stress and strain at fracture and texture profile analysis
fruits (Supporting Information Table 2S), 32 for orange peel
(hardness, fracturability, elasticity, adhesiveness, cohesiveness
and citron fruits (Supporting Information Tables 3S and 4S,
and chewiness) were performed using a texturometer
respectively).
equipped with a 30 kg compression cell (TA-XT2i Texture
The panel was calibrated in a pretest where two samples
Analyzer, Stable Micro Systems, Godalming, Surrey, UK).
were evaluated. One with extremely low attributes and
Measurements were performed with a lubricated 75-mm
diameter cylindrical probe. Stress and strain at fracture were another with high attributes level. Attribute intensities were
performed on the fruits after osmotic dehydration (five rep- scored with a 5-point intensity scale (0 – lower intensity to 5
licates). Test samples cut in cylinders (30.0 mm in length – higher intensity). The samples were left at room tempera-
and 10.6 mm in diameter) were compressed between two ture for 2 h to obtain a temperature of about 18C and were
lubricated flat Teflon surfaces. The fruits were deformed at a presented in a white plate. Panelists were required to cleanse
constant crosshead speed of 2 mm/s up to 80% deformation. their palates with a bite of low-salt cracker, a sip of room
Assays were performed at room temperature. The force (F) temperature water and a small time lag before every sample.
and height at fracture (h) obtained from these tests were
converted to true normal stresses (rtrue) and strains (etrue), Statistical Analysis
assuming constancy of sample volume during compression,
using the following equations (Hamann 1983): Principal Component Analysis (PCA) and cluster analysis
(the average Euclidean distance, method UPGMA—
Fðh0 2 DhÞ unweighted pair-group average) were performed using the
rtrue 5 (6) software Statistica (Stat Soft Inc) 2010. The analysis of the
pR2 h0
fruit sensory profile was performed using the spider graphic.
Dh
etrue 52ln 12 (7)
h0
RESULTS AND DISCUSSION
For texture profile analysis a two-cycle compression was
performed on nonlubricated samples. All experiments were Selection of the Healthy Ingredients for
conducted in a controlled temperature room at 20C. The Candying
crosshead was allowed to descend at the rate of 2 mm/s to a
total deformation 80% compression. When the compression One of the main objectives of candying of fruits and vegeta-
stroke was completed, plunger abruptly reversed direction bles is lowering the water activity by impregnation of the
and started its upward stroke at 2 mm/s. Then a second fruits and vegetables with solids (Torreggiani 1993) that can
down and up cycle was run on the same sample. All opera- bind water and increasing the shelf life of fruits and vegeta-
tions were automatically controlled by the Texture Analyzer. bles by inhibition of bacterial and fungal growth (Marın
The instrument automatically recorded the force–displace- et al. 1998). The water activity of the traditional candied
ment or force–time curve. Ten replicates were conducted. fruits is in average 0.831 (oscillating from 0.80 to 0.86, based
in our previously results from measurement of a range of
commercial candied products), so of the various ingredients
Candied Fruits Descriptive Sensorial available today, besides their functional properties, they
Analysis must have the appropriate technological properties to be
Candied fruits sensory evaluation was performed by a panel used for production of candied fruits. For evaluation and
of 12 participants aged from 35–50 years old. All participants prediction of water activity in nonelectrolyte solutions the
were experienced in food sensory evaluation. The panelists Norrish equation was used (Norrish 1966). In Supporting
were nonsmokers and refrained from wearing perfume and Information Table 1S, there are presented the Norrish con-
drinking or consuming foods that could affect performance stant Ks and the estimated amount of ingredient needed to
in the hour before tasting. obtain a water activity of 0.83.
All the samples were evaluated in triplicate in three tasting The values obtained for sucrose, glucose, fructose, sorbi-
sessions, one session per week, from 10.00 h a.m. to 12:00 h. tol, xylitol, glycerol are in accordance with literature values
Sucrose candied fruits were presented as a comparative sam- (Chirife et al. 1980; Chirife et al., 1982; Sereno et al. 2001;
ple. Sessions were carried out under controlled temperature Baeza et al., 2010) but the value obtain for eritritol is not in
conditions (20 6 2C) and relative humidity (60 6 20%) in a accordance with that of Chirife et al. (1980) that obtained a
laboratory specifically constructed for performing sensory value of 1.34. As far as we know there is no data on the
literature available for maltitol, lactitol, GOS, FOS and poly- months after storage at room temperature. Two important
dextrose, although the values obtained for these last three environmental parameters that determine the ability of molds
ingredients were calculated based on the apparent molecular to grow on food are water activity (aw) and temperature (T)
weight dependent on their real composition that in the case (Marın et al. 1998), the effect of aw on mold growth being
of FOS and GOS contain other sugar besides the oligosac- more important than temperature (Holmquist et al. 1983).
charides. Although there is observed a trend in the number For the fruits OD with fructose and sorbitol no mold growth
of hydroxyl groups in the molecule and the value of Ks, this was observed after one year of storage at room temperature.
relation is not perfect and this may be attributed to different As expected the final sugar composition of the fruits was
spatial orientation of the hydroxyl groups, presence of the dependent of the OD agent used (Table 1, Fig. 1). Sucrose
hemiacetal ring in sugars and anomeric configuration of the was absent in all the fruits analyzed except those candied
reducing end (Norrish, 1966; Baeza et al. 2010). with sucrose and glucose, however, fructose and glucose were
For eritritol the amount of solids needed to obtain a water present in very small values, in all the fruits, independently of
activity of 0.83 exceeds the solubility of the ingredient exper- the OD agent used (sorbitol, maltitol, FOS or GOS). The low
imentally determined, the same being observed for maltitol. values of fructose and glucose present are derived from the
However, eritritol was not used because the maximum natural sugars present in the fruits and vegetables before can-
amount that could be used was 40 g/100 g and maltitol dying and also from the presence of these sugars in the acti-
could be used in a much higher concentration 60 g/100 g, light and Vivinal GOS ingredients. The exceptions are the
although lower than that needed for having the desired water fruits candied with fructose were this sugar occurs in higher
activity, was still high enough to perform the experiments. concentrations and is the sugar responsible for the amount of
For the dietary fiber ingredients the amount needed are total sugars found in these fruits (Table 1, Fig. 1). Also the
approximately the same (69–74), and taking into account fruits candied with Vivinal GOS presented significant
that polydextrose does not present a sweet flavor it was amounts of glucose (and galactose) and lactose, a higher
excluded. For the alditols, xylitol is the ingredient needing amount than that present in the initial ingredient used (Fig.
lower mass to decrease the water activity, nevertheless it pre- 1). This is due to the lower molecular weight of these sugars
liminary sensorial analysis revealed an unpleasant taste at when compared to the galactooligosaccharides that increase
this level (results not show). The same was found for glyc- their diffusion into the fruit during the osmotic dehydration
erol. Also for lactitol the products obtained after candying process (Panagiotoua et al. 1999). The same is observed for
did not show a sweet taste and were unpleasant (results not the FOS present in the actilight, as there is an enrichment of
shown) as its sweetness is much lower than that of maltitol the lower molecular weight FOS when compared to the origi-
and sorbitol (Mortensen 2006). So taking these factors into nal formulation (Fig. 1). Due to this different diffusion rates
account the ingredients selected for producing healthier can- of the oligosaccharides the amount of FOS and GOS on the
died fruits were fructose, sorbitol, maltitol, GOS and FOS. final product is difficult to measured, as different molecular
weight oligosaccharides present a different response on the
pulsed amperometric detector (Koch et al. 1998).
Chemical Composition, Shelf-Life and
The energy density values of the commercial and alter-
Energy Value of Candied Fruits
native candied fruits are listed in Table 1. As can be
Experimental values of fruits weight gain, % of residual observed the commercial fruits and fruits candied with
water, water activity at the end of the osmotic dehydration fructose, as expected contained the higher energy values,
(OD) process for the different fruits and OD agents are pre- in average 251 kcal/100 g for commercial fruits and 221
sented in Table 1, along with the observation of microbial kcal/100 g for the fruits candied with fructose. The fruits
growth during storage at room temperature. Orange peel candied with sorbitol and maltitol ranged second in
samples presented the highest value of weight gain for all the energy values, with fruits candied with sorbitol and mal-
OD agents whereas, pineapple samples presented the lowest titol presenting in average 140 kcal/100 g, showing that
values in the OD agents maltitol, GOS and fructose, whereas, the substitution of the candying agent by these two aldi-
citron samples presented the lowest values in the OD agents tols result in a 44% decrease in the energy value in rela-
sorbitol and FOS. Fruits OD with FOS or GOS present, in tion to the commercial candied fruits. For the fruits
general, a higher percentage of residual water and a higher candied with FOS and GOS the energy value is even
value of water activity (aw), followed by the fruits osmodehy- lower, 122 kcal/100g, representing a 51% decrease in the
drated with maltitol. These results are in line with the effect energy value. For the FOS and GOS candied fruits the
of each ingredient on water activity discussed previously. It amount of dietary fiber present in 100g of candied prod-
was only in the fruits osmodehydrated with this three agents uct represent 132% and 79%, respectively, of the recom-
(maltitol, FOS and GOS) that microbial growth, namely mul- mended daily adequate intake of dietary fiber (Institute
ticellular filamentous fungi, occurred at the end of 3 to 4 of Medicine 2014).
P‡ OP§ C¶ P OP C P OP C P OP C P OP C P OP C
WG (%) 107 111 103 108 125 118 108 123 100 101 125 116 110 128 120 2 2 2
Water (%) 35.5 31.9 38.3 31.9 35.2 34.9 33.3 33.3 35.2 35.6 33.3 37.0 32.4 33.5 35.3 23.9 31.3 27.0
Glucose† 0.94 0.10 0.21 1.95 0.11 0.81 1.02 0 0 15.4 2.74 9.34 0.19 0 0 29.8 22.9 23.9
Fructose† 0.93 0.05 0.15 0.64 0 0.27 11.0 0 0.11 58.0 50.0 57.2 0.06 0.08 0.26
Sucrose† 0.06 0.08 0.01 0 0 0 0 0 0 0 0 0 44.5 38.7 27.9
Lactose† 8.56 8.20 6.86
Sorbitol† 59.2 58.7 55.8 2 2 2 2 2 2 2 2 2 2 2 2
Maltitol† 2 2 2 57.4 54.6 54.2 2 2 2 2 2 2 2 2 2
FOS† 2 2 2 2 2 2 94.6 170 105 2 2 2 2 2 2
GOS† 30.7 31.1 28.0
Total† 61.1 58.9 56.1 60.0 54.7 55.3 107 170 106 54.7 42.0 44.2 58.2 50.0 57.2 74.4 61.7 52.1
aw 0.820 0.804 0.824 0.881 0.873 0.889 0.888 0.888 0.905 0.890 0.884 0.931 0.819 0.799 0.818 0.792 0.840 0.802
Energyk 150 142 135 148 131 134 140# 106 103 157 106 121 233 200 229 298 247 208
MG** 2 2 2 4 4 3 4 3 3 4 3 3 2 2 2 2 2 2
FIG. 1. SUGAR COMPOSITION OF CANDIED ORANGE PEELS WITH A. COMMERCIAL PRODUCT, B. FRUCTOSE, C. SORBITOL, D. MALTITOL, E.
ACTILIGH (FOS), F. VIVINAL GOS (GOS), AND COMPOSITION OF THE G. ACTILIGHT, H. VIVINAL GOS INGREDIENTS
GLC, Glucose; FRU, Fructose; GAL, Galactose; SUC, Sucrose; LAC, Lactose: GOS, Galactooligosaccharides; GF2, kestose; GF3, Nystose; GF4,
Fructosylnystose.
DC* (chromaticity or saturation) and DE* 5 total color dif- 2003). Changes on cell walls microstructure are the main
ference value, present higher values for orange peel samples, contributing factor to the modification of the textural prop-
independently of the OD agent used, however, the highest erties, including cellular turgor and cell wall integrity. This
value is found when the OD agent used is FOS. The OD pro- change in behavior is probably related with the different
cess, independently of the OD agent used, changes the com- solid content of the different fruits. Krokida et al. (1999)
position of the internal fruits solution, making it more observed lowering the water content of plant tissues, and
homogenous throughout the osmodehydrated sample, hence an increase in solid content, the larger will be the
which leads to more homogenous refractive indexes, pro- needed stress to break the samples. The results show that as
moting light absorption against scattering, resulting in can- a result of the OD agent, the osmodehydrated fruits became
died products with increased transparency (Chiralt and more brittle and less tough as the water content was reduced
Talens 2005). and the solid content was increased, and this may be related
In order to understand the influence of the OD agent in to the reinforcement of the cell walls due to the increase in
the fracture behavior of the osmodehydrated fruits, uniaxial total sugars concentration in the osmodehydrated fruits tis-
compression tests were performed. Osmotic dehydration sues (Torreggiani 1993). Thus, true normal stresses (rtrue) is
with the five different osmotic dehydration agents changed higher on citron fruits candied with sucrose and maltitol
their fracture behavior (Table 2). The classification of tex- and true normal strains (etrue) are higher on orange peel
tural terms for solids and semi-solids gave rise to a profiling fruits candied with FOS (Table 2).
method of texture description (TPA) applicable to both sen- To develop a deeper insight of the internal data structure
sory (Brandt et al. 1963) and instrumental measurements and evaluate the effect of the different osmotic dehydration
(Szczesniak et al. 1963). Some sensory parameters, particu- agents used on color and rheological properties of the final
larly the mechanical ones, seem to be fairly straight forward. product, a principal component analysis was performed on
For example, hardness, firmness and softness are on a scale the data set. PCA was performed for each fruit separately as
of resistance of the food to the applied compressive forces. they have a very different cellular and chemical structure.
However, as state by Szczesniak (2002) it is still not known For all fruits PCA yielded three principal components
exactly whether the human being reacts to the physical stress explaining >89% of the total variance in the original data
or to the strain, and how to simulate with instruments the set.
high strain rates experienced in the mouth. As can be observed in Fig. 2, from the score plot obtained
Hardness and fracturability increased, for all the fruits, by representing PC1 vs. PC2 for pineapple and citron and
with the exception of pineapple, when the OD agent used PC2 vs PC3 for orange peel, no alternative osmotic dehydra-
was FOS (actlight) compared with the fruits candied with tion agent yielded products with similar attributes to the
sorbitol, maltitol, GOS and fructose, nevertheless, these val- commercial candied fruits. PC2 and PC3 for orange peel
ues are slightly higher in commercially candied fruits (fruits were used, instead of PC1 and PC2, because were the factors
candied with sucrose). Elasticity is higher in orange peel, that showed the correlations and accounted for the variabili-
independently of the OD agent used, with the exception of ty in the desired variables that were discussed in the other
FOS. The same occur for cohesiveness and chewiness; two fruits. Also depending on the fruits the most variable
Orange peel presents the higher values, independently of the characteristics were different as can be observed from the
OD agent used. loadings plot obtained for the different fruits. The loadings
Also the texture parameters stress (rtrue) and strain (etrue) express how well the new PCs correlate with the original var-
of the osmodehydrated fruits changed for the different OD iables and their proximity in the scatter allow to observe
agents (Table 2). The edible part of the fruits and is mainly which variables are correlated. In all fruits the rTrue was cor-
composed of thin-walled parenchymatous cells composed by related with fruit hardness and fracturability measured by
pectic polysaccharides, cellulose, hemicelluloses, and minor TPA. The percentage of water in the final products was cor-
components such as structural proteins and phenolic com- related with the lightness for all fruits. etrue was negatively
pounds (Vincken et al. 2003). Many aspects of fruits cell correlated with the percentage of water for pineapple and
structure and mechanical properties are affected during citron, not being observed the same correlation for the
osmotic dehydration with sugar solutions, such as alteration orange peel. Also L* and b* were correlated for orange peel
of cell walls, splitting of the middle lamella, lysis of mem- and citron and not for pineapple.
branes, and tissue shrinkage (Mafra et al. 2007; Waldron Nevertheless for the different fruits the different OD agents
et al. 1997). did not result in the same set of changes. For example for the
The complex microstructure of the fruits influences the pineapple sorbitol resulted in products with a higher hardness,
candying process, since water and solutes leave the interior fracturability, rtrue, chewiness and elasticity with the other OD
of the cells and sugars diffuses to and penetrates into the free agents rendering products with higher adhesiveness, and the
spaces of the cell structure (Peir o et al. 2006; Quiles et al. commercial product having a higher etrue. On the other hand
TABLE 2. COLOR, TEXTURE PROFILE, STRESS AND DEFORMATION AT FRACTURE BY UNIAXIAL COMPRESSION OF PINEAPPLE, ORANGE PEEL AND CITRON AFTER OSMOTIC DEHYDRATION WITH
SIX DIFFERENT OSMOTIC DEHYDRATING AGENTS*
P† OP‡ C§ P OP C P OP C P OP C P OP C P OP C
L*¶ 38.5 40.6 30.7 47.3 44.1 29.8 44.9 45.7 34.0 45.8 39.2 33.3 45.8 41.3 29.6 38.4 33.2 23.3
a* 20.14 0.00 22.50 21.22 4.65 20.96 21.48 6.85 20.29 20.77 1.20 21.87 21.15 3.16 20.69 20.63 3.02 20.26
b* e 13.8 21.5 8.64 12.0 33.2 7.37 17.9 34.6 10.9 22.0 22.5 7.09 13.6 26.7 5.72 12.5 14.0 3.78
DC* (chroma) 1.42 8.14 5.35 0.75 19.32 3.66 5.53 20.94 7.07 9.48 8.67 3.68 1.22 12.67 1.99 12.48 14.30 3.79
9
HEALTHY CANDIED FRUITS QUALITY A. VILELA ET AL.
FIG. 2. PRINCIPAL COMPONENT ANALYSIS (PCA) PERFORMED TO EVALUATE THE EFFECT OF THE DIFFERENT OSMOTIC DEHYDRATION AGENTS
USED ON COLOR AND RHEOLOGICAL PROPERTIES OF THE FINAL PRODUCT. PCA WAS PERFORMED FOR EACH FRUIT SEPARATELY AS THEY
HAVE A VERY DIFFERENT CELLULAR AND CHEMICAL STRUCTURE
for orange peel the commercial product presented a higher citrus and fresh aroma); Flavor (stewed fruit, characteristic,
hardness and rtrue, with FOS and GOS enriched products con- citrus, sweet, tangerine, caramel, bitter, fresh and rotten); Tex-
taining a higher adhesiveness and etrue. For citron both the ture (stiff, gummy, fibrous, firm, fine, consistent and rough);
commercial and FOS enriched product presented a higher Taste (sweet, acid and bitter); Aftertaste (bitter, acid, sweet,
hardness, fracturability and rtrue and FOS presenting addition- astringent, spicy and herbaceous). The sum of average values
ally a higher adhesiveness and sorbitol presenting the lower assigned by the panelists for each group of attributes is
hardness, fracturability and rtrue. These results show that the marked on the corresponding axis. The sensory profile reveals
effect of the OD agent on the color and rheological properties itself when the dots are connected. The analysis of the Fig. 3A
of the final product are dependent on the product itself. shows that pineapple fruits candied with sorbitol (symbol -o-)
presents a sensory profile quite distinct from the other samples
in what concerns the attributes Flavor (pineapple, candy,
Candied Fruits Sensory Profile
sweet, fresh) and the taste/textural/mouth-feel - crumbly tex-
The sensory profile of each fruit/OD agent is shown graphi- ture, fibrous texture, hard texture, uneven texture, rough
cally in Fig. 3A, B and C. We did not made the sensorial aspect, pineapple flavor, sweet taste, sweet flavor and sweet
evaluation of fruits candied with GOS because as the glucose aftertaste (not shown in the graphics due to attributes group-
levels were very high in the products candied with GOS, due ing). Maltitol decreased the panelist perception of almost all
to the composition of the ingredient, if we did it, the flavor the attributes and, pineapple samples candied with fructose
would be certainly the result of the presence of glucose and and FOS present a very similar profile.
not from the GOS itself. In what concerns orange peel sensory characteristics, the
Due to a large number of attributes, graphs of sensory pro- analysis of Fig. 3B shows, once again, that fruits candied
file were very confused and difficult to understand, therefore, with sorbitol are very distinguished from the others in
we grouped the attributes of the same category. For pineapple almost all the parameters evaluated, namely: Aroma (charac-
samples: Color; Aspect (juicy, glossy, rough and oxidized); teristic, citrus and fresh); Flavor (stewed fruit, tangerine);
Aroma (pineapple, stewed fruit and, sweet aroma); Flavor Taste (acid), Texture (firm and fine) and Sweet aftertaste.
(pineapple, candy, sweet, fresh, cooked fruit and liquor); Tex- However, fructose is the sugar that gives to orange peel more
ture (crumbly, soft, fibrous, hard, uneven, crunchy and suitable characteristics (not shown in the graphics due to
rough); Taste (sweet and acid); Aftertaste (sweet, spoiled pine- attributes grouping), namely: orange peel characteristic
apple and persistent). For orange peel and citron samples: color, characteristic flavor, citrus flavor and fresh flavor. As it
Color; Aspect (juicy, shiny and rough); Aroma (characteristic, happened with pineapple, maltitol also decreased the
FIG. 3. SENSORY PROFILE OF PINEAPPLE (A), ORANGE PEEL (B) AND CITRON (C) CANDIED SAMPLES WITH FIVE DIFFERENT OD AGENTS.
COMMERCIAL (-D-); SORBITOL (-O-); FRUCTOSE (w); FOS (-•-); MALTITOL (-䉬-)
panelist perception, but, in this case, for all the sensory PC2) represent a percentage higher than 70% of the original
parameters evaluated. However, orange peel candied with variability, so according to Mardia et al. (1979) they satisfac-
maltitol and FOS presents a very similar profile. In citron torily explain the variability among the samples tested. In
sensory evaluation (Fig. 3C) all the sugars present a very Supporting Information Table 2S, we find the principal com-
similar sensory profile, however, FOS seems do diminish the ponent loadings for pineapple candied samples with maltitol,
caramel flavor when compared with the rest of the OD sorbitol, fructose, FOS and commercial samples.
agents used and sorbitol enhanced the tangerine flavor (not From PCA analysis it can be inferred that pineapple candied
shown in the graphics due to attributes grouping). with FOS share a similarity with the commercial candied pine-
Some authors have reported that for osmodehydrated apple and fructose candied pineapple, these last two samples
products the most important characteristics are flavor and showing the higher similarity, and maltitol and sorbitol candied
texture that depend on the solute used in preparing the fruits being the most different (Fig. 4A,D). In factor 1 positive
osmotic solution and can generate residual flavor (Brande- coordinate, with a loading value higher than 0.7 we have the
lero et al. 2005). In all three fruits evaluated, compared to attribute adhesiveness. In factor 2 positive coordinates the attrib-
the commercial candied fruits, the osmodehydrated fruits utes yellow color, rough aspect, oxidized aspect, stewed fruit
had differences in flavor acceptance and some flavors are aroma, cooked fruit and spoiled pineapple aftertaste (Support-
more or less evident in the commercial candied fruits than ing Information Table 2S). This attributes, or parameters, are
in the FOS, maltitol, sorbitol or fructose candied fruits. As the ones responsible for positioning fructose, FOS and commer-
also reported by Konopacka et al. (2009), the majority of the cial pineapple samples in the upper- right PCA quadrant (Fig.
sensory attributes investigated turned out to be significantly 4A). However, maltitol candied pineapple samples are on the
influenced by the osmotic agent used. lower- right PCA quadrant, which means that the attributes
To develop a deeper insight of the internal data structure responsible for his position are in factor 1 positive coordinate
and evaluate the effect of the different osmotic dehydration the attribute adhesiveness and in factor 2 negative coordinates
agents used on sensorial properties of the final product and the attribute soft texture (Supporting Information Table 2S). On
possible correlations between the sensorial texture and instru- the other hand, sorbitol pineapple candied samples are posi-
mental texture and distinguishing the weight of each descrip- tioned in the lower-left PCA quadrant (Fig. 4A), which means
tor to the similarities of the studied osmodehydrated fruits, a that the attributes responsible for his position, in factor 1 nega-
principal component analysis was performed on the data set. tive coordinates are: juicy aspect, glossy aspect; pineapple aroma;
Again, PCA was performed for each fruit separately as they sweet aroma, crumbly texture, fibrous texture, hard texture,
have a very different cellular and chemical structure. In all uneven texture, rough texture, pineapple flavor, sweet taste, acid
the PCAs performed, the first two factorial axes (PC1 and taste, candy flavor, sweet flavor, fresh flavor, liquor flavor, sweet
FIG. 4. PCA SAMPLES PROJECTION AND PHENOGRAMS OBTAINED BY CLUSTER ANALYSIS, OF SENSORY, CHEMICAL COMPOSITION AND
TEXTURE PROFILE DATA OF PINEAPPLE (A AND D), ORANGE PEEL (B AND E) AND CITRON (C AND F) SAMPLES CANDIED WITH THE SUGARS
MALTITOL, SORBITOL, FRUCTOSE, FOS AND SUCROSE (COMMERCIAL CANDIED FRUITS)
aftertaste, persistent aftertaste, hardness, fracturability, elasticity, and rtrue (Fig. 4B). Fructose and maltitol candied orange peel
chewiness and rtrue (Supporting Information Table 2S). For the samples are on the lower- right PCA quadrant, which means
pineapple samples the sensorial parameters crumbly texture, that the attributes responsible for his position are lower values
fibrous texture, hard texture, uneven texture and rough texture of shiny appearance, rough aspect, characteristic aroma, citrus
were positively correlated with the instrumental texture parame- aroma, fresh aroma, stewed fruit flavor, stiff texture, gummy
ters hardness, fracturability, elasticity, chewiness and stress at texture, fibrous texture, firm texture, fine texture, consistent
fracture and negatively correlated with adhesiveness. texture, rough texture, characteristic flavor of orange peel; sweet
A similar analysis can be made with the data on Supporting taste, acid taste, bitter taste, sweet flavor, tangerine flavor, flavor
Information Table 3S (principal component loadings for orange of caramel, bitter flavor, fresh flavor, rotten flavor, sweet after-
peel candied samples with the same osmodehydrated agents). taste, astringent aftertaste and elasticity when compared to sor-
From the PCA it can be inferred orange peels candied with mal- bitol candied orange peels and also lower values citrus flavor,
titol and fructose share similar characteristics, and in another bitter aftertaste, acid aftertaste, spicy aftertaste, hardness, frac-
group we can find orange peels candied with FOS and com- turability, adhesiveness and rtrue when compared to commercial
mercial orange peels (Fig. 4B, E). Once again, sorbitol samples and FOS candied orange peels. For candied orange peels all sen-
are distinguished from the other four candied products. Orange sorial textural parameters, stiff texture, gummy texture, fibrous
peel candied with sorbitol are positioned in the lower-left PCA texture, firm texture, fine texture, consistent texture, rough tex-
quadrant (Fig. 4B), which means that the attributes responsible ture, are correlated with the instrumental elasticity.
for his position in factor 1 negative coordinates are: shiny Although being a citrus fruit (with similar characteristics with
appearance, rough aspect, characteristic aroma, citrus aroma, orange), the OD agents FOS, fructose, sorbitol and maltitol
fresh aroma, stewed fruit flavor, stiff texture, gummy texture, imprinted peculiar characteristics to the citron fruits. In fact, in
fibrous texture, firm texture, fine texture, consistent texture, Fig. 4F we can see that fruits candied with maltitol and fructose
rough texture, characteristic flavor of orange peel; sweet taste, share some similar characteristics, followed by fruits candied with
acid taste, bitter taste, sweet flavor, tangerine flavor, flavor of FOS and commercial candied citron and then, again, citron can-
caramel, bitter flavor, fresh flavor, rotten flavor, sweet aftertaste, died with sorbitol are very distant from the other ones. On Sup-
astringent aftertaste and elasticity (Supporting Information porting Information Table 4S are presented the PCA component
Table 3S). Commercial orange peel and FOS candied orange loadings for citron candied samples with FOS, sorbitol, fructose,
peels are better defined by the positive coordinates of factor 2 maltitol and commercial candied citron. The commercial sam-
(PC2), attributes such us citrus flavor, bitter aftertaste, acid ples are positioned in the upper-left PCA quadrant (Fig. 4C).
aftertaste, spicy aftertaste, hardness, fracturability, adhesiveness The negative factor 1 attributes that better characterize those
DOBREVA, V., HADJIKINOVA, M., SLAVOV, A., HADJIKINOV, MORTENSEN, A. 2006. Sweeteners permitted in the European
D., DOBREV, G., and ZHEKOVA, B. 2013. Functional Union: safety aspects. Scand. J. Food Nutr. 50(3), 104–116.
properties of maltitol. J. Agric. Sci. Tech. 5(2), 168–172. NORRISH, R.S. 1966. An equation for the activity coefficients
GUGGISBERG, D., PICCINALI, P., and SCHREIER, K. 2011. and equilibrium relative humidities of water in confectionery
Effects of sugar substitution with Stevia, ActilightTM and Stevia syrups. J. Food Technol. 1, 25–39.
combinations or PalatinoseTM on rheological and sensory PANAGIOTOUA, N. M., KARATHANOSA, V. T., and
characteristics of low-fat and whole milk set yoghurt. Int. Dairy MAROULISA, Z.B. 1999. Effect of osmotic dehydration agent on
J. 21(9), 636–644. osmotic dehydration of fruits. Dry Technol. 17 (1–2), 175–189.
HAMANN, D.D. (1983). Structural failure in solid foods. In R., DIAS, V.M.C., CAMACHO, M.M., and MARTINEZ-
PEIRO,
Physical Properties of Food, (E.B. Bagley and M. Peleg, eds.) pp. NAVARRETE, N. 2006. Micronutrient flow to the osmotic
351–383, AVI Publishing Co., Westport. solution during grapefruit osmotic dehydration. J. Food Eng.
HILDEBRAND, J.H., and SCOTT, R.L. 1962. Regular Solutions, 74(3), 299–307.
Englewood Cliffs, N.J., Prentice-Hall. QUILES, A., PEREZ-MUNUERA, I., HERNANDO, I., and
HOLMQUIST, G.U., WALKER, H.W., and STAHR, H.M. 1983. LLUCH, M.A. 2003. Impact of mass transport on
Influence of temperature, pH, water activity and antifungal microstructure of Granny Smith apple parenchyma during
agents on growth of Aspergillus flavus and A. parasiticus. J. osmotic dehydration. J. Sci. Food Agric. 83(5), 425–429.
Food Sci. 48, 778–782. SERENO, A.M., HUBINGER, M.D., COMESANA ~ J.F., and
HUNT, R.W.G. 1991. Measuring color, 2nd Ed., Ellis Horwood, CORREA, A. 2001. Prediction of water activity of osmotic
New York. solutions. J. Food Eng. 49, 103–114.
INSTITUTE OF MEDICINE. 2014. Dietary Reference Intakes STAT SOFT INC 2010. Electronic Statistics. Tulsa, Oklahoma.
(DRIs): Recommended Dietary Allowances and Adequate SZCZESNIAK, A.S. 2002. Texture is a sensory property. Food
Intakes, Total Water and Macronutrients. Food and Nutrition Qual. Prefer. 13, 215–225.
Board, Institute of Medicine, National Academies. https://fnic. SZCZESNIAK, A.S., BRANDT, M.A., and FRIEDMAN, H. 1963.
nal.usda.gov/sites/fnic.nal.usda.gov/files/uploads/DRI_RDAs_ Development of standard rating scales for mechanical
Adequate_Intakes_Total_Water_Macronutrients.pdf (accessed parameters of texture and correlation between the objective and
10 January 2015). sensory methods of texture evaluation. J. Food Sci. 28, 397–403.
KOCH, K., ANDERSSON, R., and AMAN, P. 1998. Quantitative TORREGGIANI, D. 1993. Osmotic dehydration in fruit and
analysis of amylopectin unit chains by means of high- vegetable processing. Food Res. Int. 26(1), 59–68.
performance anion-exchange chromatography with pulsed TORREGGIANI, D., and BERTOLO, G. 2001. Osmotic pre-
amperometric detection. J. Chromatogr. A, 800, 199–206. treatments in fruit processing: chemical, physical and structural
KONOPACKA, D., JESIONKOWSKA, K., KLEWICKI, R., and effects. J. Food Eng. 49(2), 247–253.
BONAZZI, C. 2009. The effect of different osmotic agents on VINCKEN, J.P., SCHOLS, H.A., OOMEN, R., MCCANN, M.C.,
the sensory perception of osmo-treated dried fruit. J. Hortic. ULVSKOV, P., VORAGEN A.G.J and VISSER R.G.F. 2003. If
Sci. Biotech. 84, 80–84. homogalacturonan were a side chain of rhamnogalacturonan I.
KROKIDA, M.K., KIRANOUDIS, C.T., and MAROULIS, Z.B. Implications for cell wall architecture. Plant Physiol. 132(4),
1999. Viscoelastic behavior of dehydrated products during 1781–1789.
rehydration. J. Food Eng., 40(4), 269–277. WALDRON, K.W., SMITH, A.C., PARR, A.J., NG A., and
B., and GRZELAK, K. 2006. Qualitative and quantitative
KROL, PARKER, M.L. 1997. New approaches to understanding and
composition of fructooligosaccharides in bread. Eur. Food Res. controlling cell separation in relation to fruit and vegetable
Technol. 223(6), 755–758. texture. Trends Food Sci. Tech. 8(7), 213–221.
MAFRA, I., BARROS, A.S., and COIMBRA, M.A. 2007. The ZOETENDAL, E.G., RAES, J., VAN DEN BOGERT, B.,
combined effects of black oxidizing Fig olive process and ARUMUGAM, M., BOOIJINK, C.C., TROOST, F.J., BORK, P.,
ripening on the cell wall polysaccharides of olive pulp. WELS M., DE VOS, W.M., and KLEEREBEZEM, M. 2012. The
Carbohydr. Polym. 68(4), 647–657. human small intestinal microbiota is driven by rapid uptake
MARDIA, L.V., KENI, J.T., and BIBBY, J.M. 1979. Multivariate and conversion of simple carbohydrates. ISME j. 6, 1415–1426.
analysis, 1st Ed., Academic Press, London.
MARIN, S., SANCHIS, V., TEIXIDO, A., RAMOS, A.J., VINAS, I.,
and MAGAN, N. 1998. Environmental factors, in vitro
interactions, and niche overlap between Fusarium monoliforme,
SUPPORTING INFORMATION
F. proliferatum, and F. graminearum, Aspergillus and Penicillium Additional Supporting Information may be found in the
species from maize grain. Mycol. Res. 102, 831–837. online version of this article.