Profiling of different bioactive compounds in functional drinks by high-performance liquid chromatography.

Abstract

In the present work, an HPLC method is proposed to simultaneously detect and quantify water- and fat-soluble vitamins, phenolic compounds, carotenoids and chlorophylls in a single run, by using an ultradeactivated C18 column and gradient separation using trifluoroacetic acid, water and methanol. It is shown that the HPLC method provides baseline separation of all these compounds with good resolution values in 40 min. Moreover, other figures of merit of the method show a good linear response and low detection limits for all the compounds considered in the present study. Furthermore, the usefulness of this method is demonstrated via its successful application to the analysis of different beverages from different natural origin (orange, strawberry, apple, peach pineapple, plum and blackcurrant juices, soybean milk, beers) without the need of any previous sample preparation. A good correlation is also found by comparing the total phenol content (measured by Folin-Ciocalteu method) with the sum of total phenolic compounds obtained using the proposed HPLC method. By using statistical tools, the main compounds associated with antioxidant activity of the extracts (measured by 1,1-diphenyl-2-picrylhydrazyl radical scavenging) were assessed.

Full text

Page 1

PROFILING OF DIFFERENT BIOACTIVE COMPOUNDS IN
1
FUNCTIONAL DRINKS BY HPLC 
2
José A. Mendiola1, Francisco R. Marin2, F. Javier Señoráns2, Guillermo Reglero2, Pedro
J. Martín1, Alejandro Cifuentes1, Elena Ibáñez1
3
4
5
6
7
8
9
1Instituto de Fermentaciones Industriales (CSIC). Juan de la Cierva, 3. 28006 Madrid,
Spain
2Sección Departamental Ciencias de la Alimentación (Unidad Asociada al CSIC)
Universidad Autónoma de Madrid, Campus de Cantoblanco, 28049 Madrid, Spain.
10
11
12
Abstract
In the present work, an HPLC method is proposed to simultaneously detect and quantify 13
water and fat soluble vitamins, phenolic compounds, carotenoids and chlorophylls in a 14
single run, by using an ultradeactivated C18 column and gradient separation using 15
trifluoroacetic acid, water and methanol. It is shown that the HPLC method provides 16
baseline separation of all these compounds with good resolution values in 40 min. 17
Moreover, other figures of merit of the method show a good linear response and low 18
detection limits for all the compounds considered in the present study. Furthermore, the 19
usefulness of this method is demonstrated via its successful application to the analysis 20
of different beverages from different natural origin (orange, strawberry, apple, peach 21
pineapple, plum and blackcurrant juices, soybean milk, beers) without the need of any 22
previous sample preparation. A good correlation is also found by comparing the total 23
phenol content (measured by Folin-Ciocalteau method) with the sum of total phenolic 24
compounds obtained using the proposed HPLC method. By using statistical tools, the 25
main compounds associated with antioxidant activity of the extracts (measured by 26
DPPH radical scavenging) were assessed.27
28
Keywords: functional foods, water soluble vitamins, fat soluble vitamins, phenolic 29
compounds, HPLC-DAD, Folin-Ciocalteau, Antioxidant activity, PLS.30

 Presented at the 7th Meeting of the Spanish Society of Chromatography and Related Techniques,
Granada, Spain, 17-19 October 2007

Page 2

31
32
33
1.- Introduction
Nowadays, the consumers’ demand of new functional foods including those enriched 34
with vitamins is clearly rising. Usually mixtures of different vitamins are added to food 35
products, whose complex composition make the analysis and quantification of vitamins 36
and other biologically active compounds difficult. The analysis of such a mixture by 37
conventional methods usually involves tedious and time-consuming procedures, since it 38
requires the use of numerous analytical and/or sample preparation protocols. 39
Vitamins are a structurally heterogeneous group of essential food constituents that are 40
required in small amounts for normal growth and maintenance of human health. They 41
are generally divided in two main groups: lipid- and water-soluble vitamins. Until 42
recently [1], this classification has been used to determine the type of analytical method 43
applied, being the most common approach the analysis of each vitamin individually 44
(mostly by using HPLC) or even a multiple water- or fat- soluble vitamins 45
determination [2-8].Nevertheless, some methods have been developed that combine the 46
analysis of water and fat soluble vitamins in one step such as using TLC [9] or HPLC 47
[1]. Thus, in the work developed by Kledjus et al. [1], the simultaneous determination 48
of water- and fat-soluble vitamins in pharmaceutical preparations and food samples was 49
achieved after a careful optimisation of all the factors influencing the separation such as 50
type of column, pH, temperature, gradient elution and so on. 51
Even considering the enormous interest of vitamins as micronutrients in food 52
samples, there exist other compounds with known biological activity that can, at the 53
end, have a similar or even greater contribution in the functional activity of a food 54
and/or beverage. These compounds (e.g., polyphenols, carotenoids, chlorophylls, etc) 55
can play an important role in the associated bioactivity of a food product and, therefore, 56
their determination can also be of great importance for food and nutrition studies. These 57
food components have the common property of being labile and so degraded by light, 58
heat or oxygen, also they are extremely complex, with dozens of very similar 59
components or isomers which makes its separation extremely difficult as next 60
discussed. 61
The basic feature of all polyphenols is the presence of one or more hydroxylated 62
aromatic rings, which seemed to be, in fact, responsible for their properties as radical 63

Page 3

scavengers [10-16]. Some analytical strategies have been reported [17] for the 64
determination of polyphenols. Their presence may interfere in the analysis of vitamins, 65
so usually a clean up step is needed, namely solid-phase extraction, liquid extraction, 66
etc. In spite of this difficulty, some analytical methods based on HPLC have already 67
been developed to quantify some phenolic compounds along with some vitamins [18, 68
19]. Carotenoids can have many physiological functions. Given their structure, 69
carotenoids are efficient free-radical scavengers, and they enhance the immune system 70
[20-26]. On the other hand, some biological effects have also been associated to 71
chlorophylls [20]. The analysis of carotenoids and chlorophylls is commonly done by 72
HPLC and several methods have been developed to simultaneously determine these 73
type of compounds [27-29]. 74
The goal of this work was to develop an HPLC method (using diode array detection) 75
for the simultaneous separation, detection and quantification of water- and fat-soluble 76
vitamins, phenolic compounds, chlorophylls and carotenes in a single run. The 77
developed method should allow obtaining a “bioactive profile” of different food 78
matrices. The usefulness of the HPLC method was corroborated through its application 79
to the analysis of several commercial drinks, some of them fortified with vitamins. 80
Moreover, based on statistical analysis, it was possible to associate the antioxidant 81
activity of the functional beverages to their content on some target compounds. 82
83
2.- Experimental
84
2.1.- Chemicals and reagents
85
Trifluoracetic acid (TFA) was purchased from Scharlau Chemie (Barcelona, Spain). 86
All vitamins, fat- (-tocopherol, menadione, cholecalciferol, and -tocopherol 87
succinate) and water soluble vitamins (thiamine, niacine, folic acid, pantothenic acid, 88
ascorbic acid, riboflavin, cobalamine) were from Accustandard (New Haven, CT, 89
USA). All standard phenolic compounds (caffeic acid, coumaric acid, catechin, 90
catechol, cyanidine chloride, gallic acid, genistein, hesperidin and quercetin) were 91
purchased from Extrasynthese (Genay, France) and chlorophylls and zeaxanthin from 92
DHI (Hørsholm, Denmark). Milli-Q water was obtained using a Millipore purification 93
system (Millipore), methanol of HPLC grade was purchased from LabScan (Dublin, 94
Ireland). 95

Page 4

Standards solutions (3 mg/ml) were prepared using solutions of 0.01% TFA in water 96
(polar vitamins, caffeic acid and gallic acid), methanol:wather (the rest of phenolic 97
compounds) or methanol (pigments and fat soluble vitamins) and stored in darkness 98
under freezing conditions (<-20ºC). 99
100
2.2.- HPLC conditions
101
The method developed in the present work was modified from Klejdus et al [1]. 102
Analytes were separated on a ACE-100Å C18 (150 mm × 4.6 mm, 3 μm particle size) 103
(Advanced Chromatographic Technologies, Aberdeen, UK) in a single run using 104
combined isocratic and linear gradient elution with a mobile phase consisting of 0.010% 105
trifluoroacetic acid (solvent A, pH=3.9) and methanol (solvent B) at the flow rate of 0.7 106
ml/min using an Agilent 1100 HPLC apparatus. The gradient profile (A%:B%) started 107
at 96:4 and increased up to 95:5 in the first 4 min, then linearly changed up to 2:98 108
during the next 7 min, then it was constant in the next 6 min, increased up to 0:100 in 2 109
min, then constant until a total analysis time of 38 min and finally linearly increased up 110
to 96:4 to reach initial conditions. Different wavelengths were selected for monitoring 111
the HPLC separation: 280, 350, 450 and 660 nm. 112
Direct injection of pure vitamin standards was done by dissolving them in TFA 0.01% 113
or 100% methanol depending on the polarity of the vitamin. For polyphenols, all 114
samples were injected dissolved in hydromethanolic solutions. 115
116
2.3.- Food samples
117
Different commercial drinks enriched in vitamins were analyzed along with two beer 118
samples, namely: 119
a) Orange Juice: sterilized canned orange juice. 120
b) Soy-orange drink: mixture of orange juice with “soybean milk” with vitamin E. 121
c) Strawberry Yogurt: drinkable yogurt with strawberry juice. 122
d) Multifruit-Milk Juice: partially fermented milk with Bifidus, mixed with peach, 123
apple and orange juices enriched in vitamins A, C and E. 124

Page 5

e) Antioxidant Juices: mixture of juices rich in antioxidants, two flavours were 125
analyzed: orange+raspberry+acerola (Antiox 1) and pineapple+plum+blackcurrant 126
(Antiox 2) 127
f) Beer: beer pilsner type with 5.4% alcohol, as declared in label 128
g) Light Beer: light beer pilsner type with 0.0% alcohol, as declared in label.129
h) Fruit purée: mixture of different fruit purees and juices, include apple, carrot, 130
pineapple, passion fruit and sweet corn. 131
132
The only clean up used was filtering through a 0.45 μm nylon syringe filter (Symta, 133
Madrid, Spain). Fruity samples were diluted 1:1 with water. Beer was degassed using an 134
ultrasonic bath to avoid CO2 to enter into the chromatographic system. 135
136
2.4.- Antioxidant activity
137
The antioxidant activity of beverages was measured using the DPPH method. All the 138
measurements were carried out according to a procedure previously used in juices by 139
Klimczak et al. [30]. Briefly, 0.1 mL of juice sample (filtered using syringe nylon filter 140
of 0.45 μm and diluted with distilled water 10:1) was added to 2.46 mL of 1,1-diphenyl-
2-picrylhydrazyl radical (DPPH; 0.025 g·L−1 in 50% ethanol) and mixed by vortex for
141
142
1 min. The absorbance of the samples was measured at 515 nm every 5 min for 30 min 143
using the spectrophotometer Genesys (Thermo Finnigan, Madrid, Spain). For each 144
sample, three separate determinations were carried out. The percentage of DPPH, which 145
was scavenged (%DPPHsc) was calculated as in [30]. 146
147
2.5.- Total polyphenol content
148
Total polyphenols were determined using the Folin–Ciocalteu method [31]. 149
Previously, a precipitation of proteins was carried out using HCl 1,48% and 150
centrifugation. The method consists of mixing 2 ml of Na2CO3 (2% w/v in water) with 151
100 l of the supernatant; after 3 minutes of reaction, 50 l of Folin reagent (Scharlau 152
Chimie, Barcelona, Spain) was added. After 30 min of incubation the absorbance was 153
measured at 750 nm. The standard calibration curves were prepared using pure caffeic 154

Page 6

acid (Sigma, Madrid, Spain) and caffeic acid + BSA (Bovine serum albumin, Sigma, 155
Madrid, Spain). The results were expressed in mg of caffeic acid equivalents per liter of 156
beverage. 157
158
2.6.- Statistical analysis
159
In order to correlate the antioxidant activity found in the beverages with the different 160
compounds detected using the HPLC method, two statistical strategies were used, 161
namely: Partial Least Squares (PLS) regression and Forward Stepwise Multiple Linear 162
Regression (FSMLR), were tested to predict the antioxidant activity of the samples 163
using the complete profile of components of HPLC analysis or considering only a few 164
selected components, respectively. The STATISTICA program for Windows, version 165
7.1, was used for data processing (StatSoft, Tulsa, OK, USA (2006), 166
www.statsoft.com). 167
168
3.- Results
169
As mentioned, the method developed in the present work has been optimized based 170
on the work done by Kledjus et al [1]; the most important modifications of the method 171
were in the phase gradient and in the selected analytical column. The main challenge 172
encountered for the simultaneous determination of water- and fat-soluble vitamins, 173
phenolic compounds, carotenoids and chlorophylls was the different physico-chemical 174
properties of the compounds tested. As mentioned under Experimental section, 5 water 175
soluble vitamins and 2 forms of fat-soluble vitamins were used as standards along with 176
8 phenolic compounds (corresponding to 7 families of polyphenols), zeaxanthin and 177
chlorophyll-a. To be able to separate such a complex mixture composed of polar and 178
non polar analytes in a single run, the selection of the chromatographic column, among 179
other parameters, is crucial. Therefore, an ultra inert HPLC column whose main features 180
are high efficiency and low silanol activity was selected in the present work. Besides, 181
this column allows working under highly acidic conditions without deterioration of the 182
separation achieved.183
Another important challenge is the selection of the most suitable wavelength for 184
simultaneous detection and quantification of all the compounds. In this method, the use 185

Page 7

of a diode array detector allowed us to select 4 different wavelengths according to the 186
spectral characteristics of all compounds tested. Thus, 280 nm was chosen as the most 187
suitable detection wavelength for simultaneous vitamin-polyphenol-pigment 188
determination [1], while 350 nm was selected for flavones, flavonols, chalcones and 189
aurones, 450 nm for carotenoids, chlorophylls and anthocyanins, and 660 nm for 190
chlorophylls. 191
By optimizing the separation conditions it was possible to achieve the elution of all 192
the standards in a single run with a good resolution. As can be seen in Figure 1, water- 193
soluble vitamins elute from minutes 2 to 10 while phenolics elute between min 6 and 194
14. Fat-soluble vitamins as well as photosynthetic pigments are eluted in the last part of 195
the chromatogram, that is, after 20 minutes. Interestingly, with the information provided 196
by both, retention time and spectra of each compound, a profile of the bioactive 197
components of a food sample can be obtained using this method.198
199
3.1.- Figures of merit
200
The HPLC method was evaluated in terms of linear response range, limit of detection 201
(LOD), limit of quantification (LOQ) and precision, obtaining the results shown in 202
Table 1. External calibration was used to perform the quantification of the compounds.203
Moreover, RSDn=3 values obtained for intra-day repeatability were better than 3.4% 204
and 1.9% for peak areas and migration time, respectively. On the other hand, RSD 205
values obtained for three different days were better than 6.7% and 2.4% for peak areas 206
and migration time, respectively, assuring an adequate quantitation of the analytes.207
As can be seen in Table 1, all studied standards provide a good linear response. Thus, 208
the widest linear response was provided by pyridoxine, while the anthocyanin 209
(cyanidine chloride) provide the narrowest linear response and the worse linear
adjustment (R2). The limit of detection (LOD) and limit of quantification (LOQ) were
210
211
calculated according to IUPAC [32]: 212
213
214
215

Page 8

bl blL
Skxx

Equation 2 216
where: xL is the area limit,
bl
x is the mean of the blank areas, k is the numerical factor 217
used (k=3 for limit of detection LOD, k=10 for limit of quantification LOQ) and Sbl is 218
the standard deviation of blanks.219
All calibration curves were obtained using 280 nm despite it is not the maximum 220
wavelength of absorbance for all compounds, i.e., zeaxanthin and chlorophyll. 221
However, the selection of a single wavelength to quantify at the same time all the 222
compounds present in functional drinks can be useful for both, routine analysis and 223
profiling of bioactive compounds using a method as simple as possible that can be run 224
in a HPLC with single wavelength detector. 225
226
3.2.- Total phenolic content and antioxidant activity
227
In our study, antioxidant activity of beverages was determined using DPPH free 228
radical scavenging and concentration of total polyphenols was determined using the 229
Folin-Ciocalteau method. Both assays have proven their efficacy in juices [30]. Table 2 230
shows the mean of the results obtained in each assay (analysis were done by triplicate).231
In terms of antioxidant activity, the best results were obtained with both, Antiox 1 and 232
Antiox 2 beverages while the worse results were obtained with beer and the multifruit-233
milk beverage. This behavior can somehow be correlated to their total phenolic content 234
since Antiox 1 was the drink with higher phenolic content while the multifruit-milk 235
beverage was the lowest. Despite these results, there is not a clear relationship among 236
the total phenolic content and the radical scavenging activity, as can be observed in 237
Figure 2A. This lack of relationship could be associated with the presence of vitamins 238
(with antioxidant properties) in the drinks (i.e. vitamins C and E), which do not 239
contribute to the total phenolic measurement but can have synergistic effects with the 240
antioxidant activity of phenolic compounds [33].241
242
3.3.- Food samples analysis
243
The chromatograms obtained for all beverages (after dilution 1:1 with water) can be 244
seen in Figure 3. In general, good resolution was achieved. As can be seen, the amount 245

Page 9

of fat soluble vitamins and pigments was not very significant, because the non-polar 246
zone of the chromatograms is almost empty. The quantification of vitamins and 247
polyphenols detected is shown in Table 3. In general terms, there exist a good 248
correlation between the quantification of total phenolic compounds using Folin-249
Ciocalteau method and the total phenolic compounds quantified using the present 250
HPLC-DAD method, as can be seen in Figure 2B. 251
The main phenolic compounds present in orange juice are flavanones, as expected 252
[34-36], but the vitamin C content of this juice was very low, probably due to the 253
sterilization step during its manufacture. In fact, concentration of vitamin C is a 254
significant indicator of orange juice quality and, therefore, how it has been processed 255
[37, 38]. 256
The analysis of the soy-orange drink reveals a high similarity with the pure orange 257
juice but with two main differences: vitamin E and isoflavones. Vitamin E has been 258
detected in the soy-orange drink, probably being incorporated as additive E-307 (α- 259
tocopherol) to avoid lipid peroxidation of the fat content of “soy-milk” [39]. As for the 260
isoflavones, their origin come clearly from the soy-milk [40-42]. 261
Both Antiox drinks present a high content in ascorbic acid and the highest 262
concentration of anthocyanins. In spite of it, these beverages showed quite different 263
profiles. Antiox 1 posseses the highest content in thiamine and flavanones that can be 264
related to both, its antioxidant activity and its high total phenolic content. The phenolic 265
compounds observed in both beverages were closely related with those phenolic 266
compound that are typically detected in fruits, i.e., flavanones and flavonols in oranges 267
[30], anthocyanins in berries [43], acerolas [44] and plums [45] or cinnamic acids and 268
flavan-3-ols in pineapples [46, 47]. 269
Qualitative profiles of beers is quite similar being the main differences quantitative. 270
Both present an important content of vitamins B, which is higher in the “traditional 271
beer”; in fact, “light beer” shows lower content of most of the compounds analyzed. The 272
content of vitamin C could be related with the addition of E-300, ascorbic acid, as 273
additive (as declared in label). In terms of phenolic compounds, the main families are 274
the phenolic acids and flavanones, but in light beer the presence of benzenediols could 275
be detected. The composition of beers obtained using this method is in accordance with 276
previously published works [48].277

Page 10

It is important to remark the presence of tocopherol in milky drinks (multifruit and 278
strawberry). Vitamin E is stabilized by the non polar fraction of the milk. The 279
strawberry yogurt is the only beverage lacking ascorbic acid. These beverages posses an 280
important content of thiamine, and their phenolic profile correspond with previously 281
published profiles of fruits that take part in the mixtures [10, 36, 43, 49, 50]. 282
Regarding photosynthetic pigments (carotenoids and chlorophylls), none of the 283
previously discussed samples contained such compounds above their detection limit. In 284
order to prove the efficacy of the HPLC method in real samples, a fruit purée was 285
analyzed. In this case the compounds detected were similar to those detected in the 286
multifruit juice, but in this case a certain amount of zeaxanthin was found. Zeaxanthin is 287
a typical compound found in corn, Zea mays [51].288
289
3.4.- Statistical analysis
290
In order to correlate the composition of beverages with its antioxidant activity, two 291
statistical strategies were used: partial least squares (PLS) regression and forward 292
stepwise multiple linear regression (FSMLR). 293
By using the PLS analysis four principal components could be assessed, but only 294
using the first one the model ( ˆ
p
0ii
i 1

yb b X

) could explain 91.1% of variability 295
(R2=0.911). Figure 4A shows the predicted values for antioxidant activity (ActAntOx), 296
obtained from PLS regression with all the components of the chromatographic profile 297
and 1 component, versus the observed values. As can be seen, the fit for the predictions 298
of ActAntOx can be considered adequate. From the standardised regression coefficients, 299
the most important variables were the anthocyanins eluted at 11.7 and 11.8 min (with 300
coefficients of correlation of 0.93 and 0.91, respectively), the benzenediol eluted at 9.3 301
min (0.74), flavanones eluted between 13.2 and 14.2 min (0.74) and sum of flavanones 302
(0.73). The root-mean-square error of this calibration was RMSEC=6.32, this value is 303
defined by the equation:
ˆ
y
()2
ii
y
RMSEC
n


where n is the number of samples (in 304
this study n=8), yi is the true ActAntOx. 305

Page 11

A relative error,
iy
RMSEC
RE 
, of 11.5% was obtained in this case. Using this model 306
the difference among the predicted value of antioxidant activity and the observed is 307
below 5 for all beverages, except for the multifruit-milk drink, whose difference is 15. 308
The reason of this difference can be due to the low amount of phenolic compounds 309
present in this sample. 310
By using the FSMLR, an equation ˆ
p
0ii
i 1

yb b X

is assumed, where b0 is the intercept 311
of the model, bi is the regression coefficient for the compound (Xi), p is the number of 312
compounds in the model, and ˆiy is the radical scavenging activity calculated with the 313
model. Values of 4.0 and 3.9 were used for F-to-enter and F-to-Remove, respectively, 314
and a limit of 10 steps was fixed. The Multiple Regression module of STATISTICA 315
software was used for calculation. The selected variables were the compounds that 316
eluted between 11.6 and 11.8 min. The value of determination coefficient was
R2=0.978, and the value of the root-mean-square error of calibration was RMSEC=3.09.
317
318
A relative error of RE=5.6% was obtained in this case. Fig. 4B shows the predicted 319
values obtained using the fitted regression equation versus the observed values for 320
antioxidant activity. As can be seen, the fit for the predictions of DPPH radical 321
scavenging activity can be considered appropriate. With this statistical procedure the 322
differences of all the predicted values with the observed antioxidant activities are below 323
7.324
325
4.- Conclusions
326
The results obtained for the simultaneous determination of polyphenols, carotenoids, 327
chlorophylls and water and fat soluble vitamins using HPLC-DAD confirm that the 328
proposed method can be successfully applied to the routine analysis of these kind of 329
compounds and to obtain a clear profiling of the bioactive compounds in a functional 330
food sample. All vitamins, whether fat or water soluble, phenolic compounds and 331
photosynthesis-related pigments were baseline separated with good resolution values 332
and good linear response. The analytical method was applied to the determination of 333
vitamin and polyphenols of functional drinks by direct injection with the only cleanup 334
of a filtration step. A good correlation could be found by comparing the total phenol 335
content measured by Folin-Ciocalteau method with the sum of phenolic compounds 336

Page 12

detected by HPLC-DAD. By using statistical tools the main compounds associated with 337
antioxidant activity of the functional drinks (measured by DPPH radical scavenging) 338
belonged to the anthocyanin and flavanone families. Easy prediction of the antioxidant 339
activity of a functional beverage can be obtained by knowing the composition in terms 340
of these families of compounds. 341
342
343
5.- Acknowledgements
344
This work has been funded by a CICYT project (AGL2005-06726-C04-02) and the 345
NOCHEMFOOD project (STREP FP6-23060-2006, European Commission, Sixth 346
Framework Programme of Research and Development, Food Quality and Safety 347
Thematic Priority).348
349

Page 13

REFERENCES350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
[1]B. Klejdus, R. Mikelova, V. Adam, J. Zehnálek, J. Vacek, R. Kizek, V. Kubán,
Anal. Chim. Acta 517 (2004) 1.
J.A. Paixao, J.M. Campos, J. Liq. Chromatogr. Rel. Technol. 26 (2003) 641.
I.N. Papadoyannis, G.K. Tsioni, V.F. Samanidou, J. Liq. Chromatogr. Rel.
Technol. 20 (1997) 3203.
C.M. Cho, J.H. Ko, W.J. Cheong, Talanta 51 (2000) 799.
X. Huang, D. Lin, Y. Chen, H. Zhang, Fenxi Huaxue 27 (1999) 812.
P. Chen, W.R. Wolf, Anal. Bioanal. Chem. 387 (2007) 2441.
A. Jedlicka, J. Klimes, Chem. Papers 59 (2005) 202.
D.E. Breithaupt, S. Kraut, Eur. Food Res. Technol. 222 (2006) 643.
L.A. Kartsova, O.A. Koroleva, J. Anal. Chem. 62 (2007) 255.
S.U. Lule, W. Xia, Food Rev. Int. 21 (2005) 367.
L.R. Fukumoto, G. Mazza, J. Agric. Food Chem. 48 (2000) 3597.
L. Almela, B. Sánchez-Muñoz, J.A. Fernández-López, M.J. Roca, V. Rabe, J.
Chromatogr. A 1120 (2006) 221.
D. Villaño, M.S. Fernandez-Pachón, M.L. Moyá, A.M. Troncoso, M.C. Garcia-
Parrilla, Talanta 71 (2007) 230.
O. Benavente-García, J. Castillo, J. Lorente, A. Ortuño, J.A. Del Rio, Food
Chem. 68 (2000) 457.
M.E. Cuvelier, H. Richard, C. Berset, J. Am. Oil Chem. Soc. 73 (1996) 645.
L. Jaime, J.A. Mendiola, M. Herrero, C. Soler Rivas, S. Santoyo, F.J. Señorans,
A. Cifuentes, E. Ibañez, J Sep. Sci. 28 (2005) 2111.
A. Escarpa, M.C. González, J. Chromatogr. A 897 (2000) 161.
M. Tasioula-Margari, O. Okogeri, Food Chem. 74 (2001) 377.
A.J. Blasco, I. Barrigas, M.C. González, A. Escarpa, Electrophoresis 26 (2005)
4664.
B. Schoefs, in Adv. Food Nutr. Res., 2005, p. 41.
H. Tapiero, D.M. Townsend, K.D. Tew, Biomed. Pharmacother. 58 (2004) 100.
X.D. Wang, R.M. Russell, Nutr. Rev. 57 (1999) 263.
S. Voutilainen, T. Nurmi, J. Mursu, T.H. Rissanen, Am. J. Clin. Nutr. 83 (2006)
1265.
N.I. Krinsky, Nutrition 17 (2001) 815.
V.P. Palace, N. Khaper, Q. Qin, P.K. Singal, Free Radic. Biol. Med. 26 (1999)
746.
T.W. Goodwin, The biochemistry of the carotenoids, Chapman and Hall,
London, 1984.
J.A. Mendiola, F.R. Marin, S.F. Hernandez, B.O. Arredondo, F.J. Senorans, E.
Ibanez, G. Reglero, J Sep. Sci. 28 (2005) 1031.
O. Montero, M.D. Macías-Sánchez, C.M. Lama, L.M. Lubián, C. Mantell, M.
Rodríguez, E.M. De La Ossa, J. Agric. Food Chem. 53 (2005) 9701.
M.M. Mendes-Pinto, A.C. Silva Ferreira, C. Caris-Veyrat, P.G. De Pinho, J.
Agric. Food Chem. 53 (2005) 10034.
I. Klimczak, M. Malecka, M. Szlachta, A. Gliszczynska-Swiglo, J. Food Comp.
Anal. 20 (2007) 313.
O. Folin, V. Ciocalteu, J. Biol. Chem. 73 (1927) 627.
IUPAC, Compendium of Chemical Terminology - the Gold Book (electronic
version) (2006), http://goldbook.iupac.org/L03540.html
J.D. Box, Water Res. 17 (1983) 511.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]

Page 14

[34]
[35]
A. Gil-Izquierdo, M.I. Gil, F. Ferreres, J. Agric. Food Chem. 50 (2002) 5107.
P. Rapisarda, A. Tomaino, R. Lo Cascio, F. Bonina, A. De Pasquale, A. Saija, J.
Agric. Food Chem. 47 (1999) 4718.
A. Bocco, M.E. Cuvelier, H. Richard, C. Berset, J. Agric. Food Chem. 46 (1998)
2123.
P. Elez-Martínez, R.C. Soliva-Fortuny, O. Martín-Belloso, Eur. Food Res.
Technol. 222 (2006) 321.
M.C. Corrêa De Souza, M. De Toledo Benassi, R.F. De Almeida Meneghel,
R.S. Dos Santos Ferreira Da Silva, Brazilian Arch. Biol. Tecnhol. 47 (2004)
391.
D.J.M. Gómez-Coronado, E. Ibañez, F.J. Rupérez, C. Barbas, J. Chromatogr. A
1054 (2004) 227.
M.D. Chiarello, J.L. Le Guerroue, C.M.S. Chagas, O.L. Franco, E. Bianchini,
M.J. Joao, J. Food Biochem. 30 (2006) 234.
A.K. Tripathi, A.K. Misra, J. Food Sci. Technol. 42 (2005) 111.
USDA-ARS, United States Dept. of Agriculture (Agricultural Research Service)
- Iowa State University Database on the Isoflavone Content of Foods Release
1.4 (2007), Beltsville, MD, USA:
http://www.ars.usda.gov/Services/docs.htm?docid=6382
S. Häkkinen, M. Heinonen, S. Kärenlampi, H. Mykkänen, J. Ruuskanen, R.
Törrönen, Food Res. Int. 32 (1999) 345.
V.L. Lima, E.A. Melo, M.I.S. Maciel, F.G. Prazeres, R.S. Musser, D.E.S. Lima,
Food Chem. 90 (2005) 565.
M. Vizzotto, L. Cisneros-Zevallos, D.H. Byrne, D.W. Ramming, W.R. Okie, in
Acta Horticulturae, 2006, p. 453.
B.F. de Simón, J. Pérez-Ilzarbe, T. Hernández, C. Gómez-Cordovés, I. Estrella,
J. Agric. Food Chem. 40 (1992) 1531.
E.R. Elkins, R. Lyon, C.J. Huang, A. Matthys, J. Food Comp. Anal. 10 (1997)
285.
J.M. Sendra, J.V. Carbonell, Evaluación de las propiedades nutritivas,
funcionales y sanitarias de la cerveza, en comparación con otras bebidas, Ed.
Fundación Cerveza y Salud, Madrid, 1999.
L. Wada, B. OU, J. Agric. Food Chem. 50 (2002) 3495.
USDA-ARS, United States Dept. of Agriculture (Agricultural Research Service)
- Database for the Flavonoid Content of Selected Foods, Release 2.1 (2007),
Beltsville, MD, USA: http://www.ars.usda.gov/Services/docs.htm?docid=6231
G. Beecher, I.M. Buzzard, J. Holden, D. Trainer, C. Spangler, C. Davis, A.L.
Eldridge, S. Sehakel, D. Haytowitz, S. Gebhardt, S. Bhagwat, FASEB J. 11
(1997).
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]

Page 15

439
440