Food Research International 130 (2020) 108993

Contents lists available at ScienceDirect

Food Research International

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

In vitro digestion effect on mineral bioaccessibility and antioxidant bioactive T

compounds of plant-based beverages
Joyce Grazielle Siqueira Silvaa, Ana Paula Rebellatoa, Elem Tamirys dos Santos Caramêsa,
⁎
Ralf Greinerb, Juliana Azevedo Lima Pallonea,
a
Department of Food Science, School of Food Engineering, University of Campinas, Campinas, São Paulo, Brazil
b
Department of Food Technology and Bioprocess Engineering, Max Rubner-Institut, Federal Research Institute of Nutrition and Food, Karlsruhe, Germany

A R T I C LE I N FO A B S T R A C T

Keywords: Consumption of plant-based beverages (PBB) is a growing trend; and have been used as viable substitutes for
Plant-based beverages dairy based products. To date, no study has comparatively analyzed mineral composition and effect of in vitro
Minerals digestion on the bioaccessibility of different PBB. The aim of this research was to investigate the content of
Bioaccessibility essential minerals (calcium (Ca), magnesium (Mg), iron (Fe), zinc (Zn)) and to estimate the effect of in vitro
Phenolic compounds
digestion in plant-based beverages, and their antioxidant bioactive compounds (phenolic compounds and an-
Antioxidant capacity
tioxidant capacity). Moreover, the presence of antinutritional factors, such as myo-inositol phosphates fractions,
Myo-inositol phosphates
Chemometrics were evaluated. Samples of PBB (rice, cashew nut, almond, peanut, coconut, oat, soy, blended or not with
another ingredients, fortified with minerals or naturally present) and milk for comparison were evaluated. TPC
ranged from 0.2 mg GAEq/L for coconut to 12.4 mg GAEq/L for rice and, the antioxidant capacity (DPPH)
ranged from 3.1 to 306.5 µmol TE/L for samples containing peanut and oat, respectively. Only a few samples
presented myo-inositol phosphates fractions in their composition, mostly IP5 and IP6, especially cashew nut
beverages. Mineral content showed a wide range for Ca, ranging from 10 to 1697.33 mg/L for rice and coconut,
respectively. The Mg content ranged from 6.29 to 251.23–268.43 mg/L for rice and cashew nut beverages,
respectively. Fe content ranged from 0.76 mg/L to 12.89 mg/L for the samples of rice. Zinc content ranged from
0.57 mg/L to 8.13 mg/L for samples of oat and soy, respectively. Significant variation was observed for Ca
(8.2–306.6 mg/L) and Mg (1.9–107.4 mg/L) dialyzed between the beverages, with lower concentrations of Fe
(1.0 mg/L) and Zn (0.5 mg/L) in dialyzed fractions. This study provides at least 975 analytically determined
laboratory results, providing important information for characterization and comparison of different plant-based
beverages.

1. Introduction 2015)
In this context, the nutritional properties of the PBB may vary ac-
Plant-based beverages (PBB) have stood out and are a growing cording to the raw material used, type of processing, and compounds
worldwide trend in the food sector, they are used as a substitute for used in fortification, among others. These factors may influence aspects
cow's milk or as an alternative to juices and other beverages (Sethi, such as; particle size, rheology, stability, color and composition of
Tyagi, & Anurag, 2016). In addition, factors such as lactose intolerance, macro and micronutrients (Jeske & Arendt, 2018; Jeske, Zannini, &
cow's milk allergy, heart disease (caused by high cholesterol levels), as Arendt, 2017). Thus, depending on the composition of the raw material
well as vegetarian, vegan, and flexitarian diets have contributed to the used in its preparation, bioactive substances such as phenolic com-
increased consumption (Chalupa-Krebzdak, Long, & Bohrer, 2018; pounds, fibers, flavonoids, vitamins and minerals may be present in the
Diarra, Nong, & Jie, 2005; Mäkinen, Uniacke-Lowe, O’Mahony, & final product, and consequently, provide health benefits (Briviba, Gräf,
Arendt, 2015; Sethi et al., 2016). Due to the impact of this demand, Walz, Guamis, & Butz, 2016).
food industries are investing in the development of new non-dairy Studies have shown that oats, rice, cashews, Brazil nuts, and al-
products based on nuts, cereals and seeds, as well as almond, oat, rice, monds are rich in phenolic compounds (45–456 mg gallic acid
soy and coconut beverages, among others (MINTEL, 2016; Tecnavio, equivalent/100 g) and have proven antioxidant capacity (Chang,

⁎
Corresponding author.
E-mail address: jpallone@unicamp.br (J.A.L. Pallone).

https://doi.org/10.1016/j.foodres.2020.108993
Received 4 October 2019; Received in revised form 17 December 2019; Accepted 7 January 2020
Available online 10 January 2020
0963-9969/ © 2020 Elsevier Ltd. All rights reserved.
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

Alasalvar, Bolling, & Shahidi, 2016; Gong et al., 2017; John & Shahidi, capacity and phenolic compound content and effects on mineral bioa-
2010; Kaur, Whitson, Ashton, Katopo, & Kasapis, 2018; Kornsteiner, cessibility in PBB. Moreover, the obtained analytical data was evaluated
Wagner, & Elmadfa, 2006). In addition to phenolic compounds and by multivariate analysis.
antioxidant capacity, PBB can naturally have in their composition mi-
nerals such as; phosphorus (P), potassium (K), zinc (Zn), manganese 2. Material and methods
(Mn), copper (Cu), sodium (Na), selenium (Se), calcium (Ca), magne-
sium (Mg) and iron (Fe), which are characterized as nutrients (Felberg, 2.1. Samples
Antoniassi, Deliza, Freitas, & Modesta, 2009; dos Santos, 2015). Thus,
the evaluation of nutritional and functional properties (such as essential 24 commercial samples of liquid and powder PPB were evaluated
mineral content and bioaccessibility) and content of phenolic com- (Supplementary material 1). Samples included, 7 rice samples, 5
pounds are required for proper characterization of these beverages cashew nut samples, 5 coconut samples, 3 oat samples, 2 almond
(Briviba et al., 2016; Codina-Torrella, Guamis, Ferragut, & Trujillo, samples, one sample of peanut and one soy. Samples were either
2017; Faccin, Miotto, Vieira, Barreto, & Amante, 2009; dos Santos, blended or not with other ingredients, or fortified with added minerals
2015). or not. One sample of sterilized coẃs milk was also evaluated for
Essential minerals, including Ca, Fe, Mg and Zn, have an important comparison with the PBB, totaling 25 samples. The samples obtained
role in the proper functioning of the body (Quintaes & Diez-Garcia, were available commercially in Campinas, São Paulo, Brazil. The study
2015). Therefore, the ingestion of foods that contain bioaccessible mi- evaluated 12 brands of PBB and one brand of milk. The powder samples
nerals is essential for the body to perform vital functions. Bioaccessi- were prepared with ultra-pure water according to label instructions and
bility studies are evaluated by in vitro digestion assays, such as the used in all the analysis. For the myo-inositol phosphates analysis, the
dialysis method; which involves nutrient transport through a semi- samples were freeze-dried for further analysis.
permeable membrane of defined pore size, similar to the intestinal pore,
and simulates the transition time of the food during digestion, with 2.2. Evaluation of total phenolic compounds and antioxidant capacity
consideration to changes in pH, agitation and temperature similar to
what occurs in the body (Alegría-Torán, Barberá-Sáez, & Cilla-Tatay, 2.2.1. Sample preparation
2015; Miller, Schricker, Rasmussen, & Van Campen, 1981; Sahuquillo, Sample preparation for total phenolic compounds (TPC) and anti-
Barberá, & Farré, 2003). However, the bioaccessibility may be affected oxidant capacity determination was performed according to the method
by different chemical forms of minerals and their interaction with ve- proposed by Varga, Jójárt, Fónad, Mihály, and Palágyi (2018), with
getable components used in beverage production (Chen et al., 2005; some modifications. For the extraction step, a total 4 mL of 80% me-
Mäkinen et al., 2015; Zhao, Martin, & Weaver, 2005). thanol solution was added at 400 μL increments to each sample, fol-
Some components that may be present in beverages, in addition to lowed by manual shaking in a 15 mL falcon tube. The tubes were then
nutritional compounds are myo-inositol phosphates (IP3, IP4, IP5, IP6), placed for 10 min in an ultrasound bath (Model: Unique, Brand: Catel)
considered antinutritional substances that may interfere with essential and centrifuged for 5 min at 6000 rpm (Refrigerated Centrifuge SL -
mineral absorption (Zhang, Önning, Öste, Gramatkovski, & Hulthén, 706, SOLAB). After centrifugation, the supernatant (extract) was col-
2007). The negative effect attributed to myo-inositol phosphates in the lected for analysis.
absorption of some nutrients is due to six reactive phosphate groups in
their chemical structure, which may form complexes with divalent ca- 2.2.2. Phenolic compounds and antioxidant capacity determination
tions such as Mg2+, Mn2+, Fe2+, Cu2+, Zn2+ and Ca2+ (Greiner, The TPC content was determined according to the method described
Konietzny, & Jany, 2006; Kumar, Sinha, Makkar, & Becker, 2010). by Singleton and Rossi (1965). In a dark room, the reaction was per-
Some compounds like oxalates, tannins, fibers, bivalent cations (in formed using 85 µL sample extract, 43 µL Folin-Ciocalteu reagent (1N)
excess) and fatty acids (in excess) could also interfere in the availability and 212 µL Na2CO3 (75 g/L). After 25 min of reaction, the absorbance
of minerals. However, myo-inositol phosphates are one of the most in- was read on 96-well microplate in a multi-mode microplate (BMG
trusive compounds related to mineral absorption reduction (Amalraj & Labtech, Germany, model Fluostar Omega) at 725 nm. Analytical
Pius, 2015; Bosscher, Van Caillie-Bertrand, Van Cauwenbergh, & curves with a gallic acid standard (standard antioxidant) at con-
Deelstra, 2003). centrations from 1 to 50 mg/L were performed. TPC contents were
Studies have already detected myo-inositol phosphates in almonds, expressed in mg of GAEq/L sample. Analyses were performed in tri-
cashew nuts, Brazil nuts, rice and oats (Duong, Clark, Lapsley, & Pegg, plicate, and the limit of quantification was 1 mg/L.
2017a, 2017b; Zhang et al., 2007). However, information about the Antioxidant capacity was performed by the DPPH (2,2-Diphenyl-1-
myo-inositol phosphate concentrations of plant-based beverages is picrylhydrazyl, Sigma Aldrich, USA) Scavenging Assay as proposed by
scarce. Bernat, Chafer, Chiralt, and Gonzalez-Martinez (2014) and Sánchez-Moreno, Larrauri, and Saura-Calixto (1998), with modifica-
Singhal, Baker, and Baker (2017) defined minerals in PBB and com- tions. A calibration curve was determined with Trolox standard solution
pared the different beverages according to their mineral content and (6-hydroxy-2,5,7-tetramethylchroman-2-carboxylic acid, Sigma Al-
vitamins, based on available label information and in databases. Other drich, USA) diluted in methanol at concentrations from 0.6 mM to
than these, comparative analytical studies of different PBB considering 20 mM. The extracts of each sample or trolox solution (85 µL) were
mineral content and in vitro digestion effects on bioaccessibility, total added of 312 µM of DPPH reagent in methanol (25 mg/L) for the assay.
phenolic compounds, antioxidant activity and myo-inositol phosphates Absorbance values during the kinect were measured at 515 nm in the
are scarce. multi-mode microplate (BMG Labtech, Germany, model Fluostar
Few studies have evaluated the mineral bioaccessibility in samples Omega). All samples were analyzed in triplicate and DPPH results were
such as soy beverages (Theodoropoulos, Turatti, Greiner, Macedo, & expressed as µmol Trolox Equivalent per liter of sample (µmol TE/L).
Pallone, 2018), but they did not compare different types of beverages The limit of quantification was considered as 0.6 mM.
composed of different raw materials, making the current study novel
research. 2.3. Myo-inositol phosphates quantification
The aims of this study are to investigate the content of essential
minerals (calcium, iron, magnesium and zinc); to estimate the effect of Samples were freeze-dried (for 24 h up to a pressure of < 200 μHg)
in vitro digestion on mineral bioaccessibility in PPB; determine the and maintained at −40 °C until myo-inositol phosphate (InsP6, InsP5,
presence of antinutritional factors in PBB (myo-inositol phosphates InsP4, InsP3) quantification. 1 g of the freeze-dried samples were ex-
fractions) and possible effects on mineral bioacessibility; antioxidant tracted with 2.4% (w/w) HCl for 3 h at 22 °C, according to Feitosa et al.

2
 J.G.S. Silva, et al.

Table 1
Total phenolic compounds, antioxidant capacity myo-inositol fractions and molar ratio.
Type Sample Total phenolic compounds Antioxidant capacity Myo-inositol fractions Myo-inositol molar ratios

TPC (mg GAE/L) DPPH (µmol TE/L) IP4 (µmol/L) IP5 (µmol/L) IP6 (µmol/L) (IP5 + IP6):Ca (IP5 + IP6):Fe (IP5 + IP6):Zn (IP5 + IP6):Mg

Rice 1 1.10 ± 0.08 jklm 7.82 ± 1.82 e < LOQ < LOQ < LOQ ND ND ND ND
2 0.91 ± 0.20 klm 283.55 ± 18.81 a < LOQ < LOQ < LOQ ND ND ND ND
3 1.88 ± 1.10 hijk 301.93 ± 1.94 a < LOQ < LOQ < LOQ ND ND ND ND
4 1.48 ± 0.03 ijkl ND < LOQ < LOQ < LOQ ND ND ND ND
5 1.07 ± 0.16 jklm 28.58 ± 2.53 de < LOQ < LOQ < LOQ ND ND ND ND
6 1.55 ± 0.14 ijkl 150.21 ± 2.82 b < LOQ < LOQ < LOQ ND ND ND ND
7 12.39 ± 0.82 a 112.25 ± 16.36 c < LOQ < LOQ < LOQ ND ND ND ND
Cashew nut 8 3.21 ± 0.10 f 146.77 ± 4.78 b < LOQ 0.4 ± 0.02 b 0.67 ± 0.02 c 0.80 15.28 24.72 0.13
9 4.76 ± 0.05 e 297.19 ± 3.25 a < LOQ 0.59 ± 0.02 ab 1.07 ± 0.08 b 0.99 20.33 27.76 0.16
10 4.24 ± 0.14 e 300.21 ± 0.37 a < LOQ 0.54 ± 0.05 b 0.76 ± 0.02 c 1.15 18.29 18.36 0.16
11 8.27 ± 0.24 c 46.83 ± 3.58 d < LOQ 0.67 ± 0.02 a 0.82 ± 0.08 c 0.61 8.06 26.05 0.13
12 2.98 ± 0.11 fg 151.75 ± 0.73 b 2.05 ± 0.04 0.56 ± 0.01 ab 0.69 ± 0.03 c 0.03 18.42 37.84 0.18

3
Almond 13 1.96 ± 0.00 ghij 304.83 ± 0.97 a < LOQ < LOQ < LOQ ND ND ND ND
14 0.53 ± 0.11 lm 151.64 ± 4.32 b < LOQ < LOQ 0.17 ± 0.02 d 0.03 ND ND 0.13
Peanut 15 6.74 ± 0.03 d ND < LOQ 0.32 ± 0.04 c 1.46 ± 0.03 a 0.05 12.76 28.73 0.26
Coconut 16 1.90 ± 0.03 hijk 306.23 ± 0.37 a < LOQ < LOQ < LOQ ND ND ND ND
17 1.47 ± 0.11 ijkl 3.07 ± 0.56 e < LOQ < LOQ < LOQ ND ND ND ND
18 0.21 ± 0.05 m 290.61 ± 30.95 a < LOQ < LOQ < LOQ ND ND ND ND
19 7.39 ± 0.05 cd 299.99 ± 1.29 a < LOQ < LOQ < LOQ ND ND ND ND
20 7.77 ± 0.36 c 12.59 ± 1.71 e < LOQ < LOQ < LOQ ND ND ND ND
Oat 21 4.77 ± 0.16 e 151.96 ± 0.37 b < LOQ < LOQ < LOQ ND ND ND ND
22 2.36 ± 0.20 fghi ND < LOQ < LOQ < LOQ ND ND ND ND
23 2.65 ± 0.30 fgh 306.46 ± 2.40 a < LOQ < LOQ < LOQ ND ND ND ND
Soy 24 10.99 ± 0.30 b 150.32 ± 3.08 b < LOQ < LOQ 0.77 ± 0.02 c ND 12.83 6.05 0.10
Milk 25 4.79 ± 0.37 e 298.05 ± 1.29 a < LOQ < LOQ < LOQ ND ND ND ND

Results are expressed as mean ± standard deviation, n = 3. Mean with different letters in the same column indicate significant difference (p < 0.05) as determinate using one way ANOVA and Tukey test at 95% of
confidence.
TPC: Total phenolic compounds, DPPH: antioxidant capacity and myo-inositol fractions (IP4; IP5; IP6). TPC (mg GAE/L) expressed as mg gallic acid equivalent per liter of sample. DPPH (µmol TE/L) expressed as
micromoles of Trolox equivalent per liter of sample. IP3 content was under the LOQ.
Food Research International 130 (2020) 108993
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

(2018). Tulsa, USA). The multivariate analysis was conducted by Principal

The extracts were centrifuged at 30.000g for 30 min to remove any Component Analysis (PCA) and performed using the MatLab R2019a
solid particles. Thereafter, the clear supernatants were diluted with (MathWorks, USA) with PLS-toolbox version 8.6 (Eigenvector Research
water (1:25) and applied to a column containing AG1-X8, 100–200 Inc, 2010) to evaluate all the results performed.
mesh resin. The column was washed with 10 volumes (10 times the
column volume) of water and 10 volumes of 25 mM HCl. The myo- 3. Results and discussion
inositol phosphates were eluted with 5 volumes of 2 M HCl. The eluate
obtained was concentrated in a vacuum evaporator till dry and the 3.1. Evaluation of total phenolic compounds, antioxidant capacity and
residue was dissolved in 10 mL water. 20 µL of the samples were myo-inositol phosphates
chromatographed on Ultrasep ES 100 RP18 (2 × 250 mm) to quantify
the myo-inositol phosphates. The column was run at 45° C and Total phenolic compounds, antioxidant capacity and myo-inositol
0.2 mL min−1 with an eluant consisting of formic acid:methanol:water : phosphates were evaluated in the 24 samples of plant-based beverages
tetrabutylammonium hydroxide (44:56:5:1.5 v/v), pH 4.25, as de- and also in the cow’s milk sample and the results obtained were pre-
scribed by Sandberg and Ahderinne (1986). A standard mixture of the sented in Table 1.
individual myo-inositol phosphate esters (InsP3–InsP6) was used. The
limits of detection and quantification for the method were of 3.1.1. Total phenolic compounds
8.0 × 10−9 and 12 × 10−9 mol/L, respectively. The total phenolic compounds (TPC) were significantly different
between evaluated samples. The TPC ranged from 0.20 mg GAEq/L for
2.4. Method validation for essential mineral evaluation by FAAS sample 18 (coconut) to 12.39 mg GAEq/L for sample 7 (rice) (Table 1).
A study about the phenolic compounds in different types of rice, was
The method was validated using samples of rice and cashew nut performed by Setyaningsih, Saputro, Carrera, Palma, and García-
beverages and according to the parameters of sensitivity, linearity, limit Barroso (2019), they found a total of 13.49 mg/kg of TPC in white rice,
of detection and quantification, accuracy, and recovery as suggested by similar to the value determined in sample 7 of rice beverage (12.39 mg
the guidelines on validation (AOAC, 2016; Brasil, 2016). GAEq/L), but higher than the values determined for the others rice
Prior to analysis approximately 500 mL of each sample was homo- beverage samples analyzed in this study. However, sample 7 has in its
genized in a blender (Philips Wallita). Powder samples were prepared composition quinoa and cocoa, which are known to be plants with high
with ultra-pure water according to label and used during the analysis. amounts of total phenolic compounds (Abderrahim et al., 2015). Cacao
The mineralization step was performed according to Silva, Orlando, has a relevant amount of flavonols, such as epicatechin and catechin
Rebellato, and Pallone (2017), with modifications. Sample volumes and methylxanthines (such as theobromine); while quinoa contains a
were adjusted to 4 mL and 6 mL of nitric acid added, then left to stand good profile of phenolic acids such as vanilic, chlorogenic and caffeic
for one night and then 2 mL of hydrogen peroxide (30%) added. The acids as well as some flavonoids (Zięba, Makarewicz-Wujec, &
mixture was mineralized for 4 h at 130 °C on a digester block and Ca, Kozłowska-Wojciechowska, 2019).
Mg, Fe and Zn were evaluated by flame atomic absorption spectrometer Rodríguez-Roque, Rojas-Graü, Elez-Martínez, and Martín-Belloso
(FAAS), model AAnalyst 200. Mineralized samples were placed into a (2013) determined TPC in soy milk and found 61.4 mg/100 mL, which
nebulizer and mixed with air-acetylene flame (2.5/10 L h−1) at ap- is higher than the found in this research (10.99 mg/L). The TPC in
proximately 2000 °C. A deuterium lamp was used for correction of soymilk was already determined in other studies and ranged from 96 to
background radiation and hollow cathode lamps for determination of 320 mg of gallic acid equivalent/100 g (Tyug, Prasad, & Ismail, 2010).
iron (248.3 nm), calcium (422.67 nm), magnesium (285.21 nm), and These results have a wide range once it depends on the soybean variety,
zinc (213.86 nm) (PerkinElmer). climatic and soils conditions, besides that the amount of grain used to
produce the beverage can also be a relevant variable once it is not
2.5. In vitro assay for bioaccessibility estimation of essential minerals standardized.
The raw materials used in the production of PBB such as almonds,
The in vitro assay to estimate bioaccessibility was performed using cashew nuts, rice and soybeans could present antioxidant activity and
the dialysis method according to Jovani, Barbera, Farre, and Aguilera different profiles of phenolic compounds, such as flavonoids, iso-
(2001) and Luten et al. (1996), with modifications. Approximately flavones, phenolic acids, and others (Herbello-Hermelo et al., 2018).
20 mL of each beverage was added to a 250 mL erlenmeyer flask and TPC values found in the phenol-explorer database show that some of
the pH adjusted to 2.0 by adding 6 mol L−1 HCl. For the gastric di- these raw materials could present a high variation in the TPC according
gestion step, 0.6 g of freshly prepared pepsin solution (1.6 g in 10 mL to the different raw materials, origin and processes applied. For al-
0.1 mol L−1 HCl) was added, the flasks were mixed and incubated at monds the TPC ranged from 126,8 to 418,00 mg/100 g; oats from 1,46
37 °C for 2 h. The digests were cooled, in an ice bath, for the titratable to 389,70 mg/100 g; peanuts from 396 to 420 mg/100 g and cashew
acidity determination. After that, a dialysis bag (Sigma-Aldrich, nuts from 137 to 274 mg/100 g (Phenol-Explorer, 2019).
35x21x30 mm, porosity of 25 Å, pore size of 14,000 Da) containing Therefore, the different contents of TPC found in this research for
water (25 mL) and NaHCO3 equivalent to the titratable acidity was the PBB could be associated to several factors: the type, origin or
added to the flasks and incubated for 37 °C for 30 min. Then a mixture amount of the raw material used during the process (since many
of 5 g of a bile-pancreatin solution (0.4% of pancreatin (w/v) and 2.5% companies do not declare the amount of raw material used in produc-
of bile extract (w/v) in 0.1 mol L−1 NaHCO3) were added and the in- tion); and the proportion of water added and presence of possible ad-
cubation proceeded for 2 h at 37 °C. ditives, which could cause an alteration of the contents of phenolic
The dialyzed fractions content were transferred into a falcon tube, compounds and other compounds (Herbello-Hermelo et al., 2018;
acidified with 1 mL HNO3 65% v/v and the mineral content determined Mendes et al., 2019; Rodríguez-Roque et al., 2013).
by FAAS.
3.1.2. Antioxidant capacity (DPPH)
2.6. Statistical analysis The DPPH method is based on a measure of the consumption of
DPPH radical by an antioxidant compound, in this case by a bioactive
The data was performed in triplicate and the results were evaluated compound present in the samples (Sousa, Oliveira, Oliveira, Moraes, &
with an analysis of variance (one-way ANOVA) and Tukey test, with Abarza, 2015).
95% of confidence using the software Statistica 8.0 (StatSoft, Inc., In this context, the antioxidant capacity (DPPH) ranged from 3.07 to

4
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

306.46 µmol TE/L for samples 17 (peanut) and 23 (oat), respectively. Udomkun et al., 2019).
Exception was observed for samples 4 (rice), 15 (peanut) and 22 (oat) In literature, few studies were found about PBB and the impact of
where antioxidant capacity was not detected (Table 1). Rice samples, 2 phytates on mineral bioaccessibility. In this context, we evaluated the
(283.55 µmol TE/L) and 3 (301.93 µmol TE/L); cashew nut, 9 effect of molar proportions of phytates (IP5 + IP6) on mineral avail-
(297.19 µmol TE/L) and 10 (300.21 µmol TE/L); almond samples, 13 ability (Fe, Ca, Mg and Zn). In our study, only cashew nuts (8 to 12),
(304.83 µmol TE/L); coconut samples 16 (306.23 µmol TE/L), 18 almond (14), peanut (15) and soy (24) samples presented phytates
(290.61 µmol TE/L) and 19 (299.99 µmol TE/L) and oat sample 23 (IP5 + IP6)/mineral molar ratio values. For the molar ratio [phytates]:
(306.46 µmol TE/L) presented the highest DPPH values and showed no [iron] the values ranged from 8.06 to 20.33; for zinc, the [phytate]:
significant difference (p < 0.05) in relation to cow's milk (sample 25). [zinc] molar ratio ranged from 6.05 to 37.84, while the molar ratio
In addition, it is relevant that at least one sample of each PBB type [phytate]:[calcium] ranged from 0.03 to 1.15 among the samples of
presented in our study about 300 µmol TE/L, proving that, as expected, cashew nuts, almond, peanut and soy.
these beverages present some antioxidant capacity. To predict mineral availability, it is ideal that the [phytate]:[iron]
Sudjaroen, Thongkao, and Suwannahong (2018) determined the molar ratio is less than 0.4 (Hurrell, 2004). Although Hurrell et al.
antioxidant capacity in cashew nut shell waste through DPPH assay and (1992) indicated that there is a strong phytate inhibiting effect on iron
found an antioxidant capacity of 57.1 µMET/mL, higher than the even when the molar ratio is 0.2. Therefore, in the current study the Fe
“DPPH values” determined in the cashew nut beverages. bioaccessibility of all the cashew nut, peanut and soy samples could be
It is important to highlight that antioxidant assays can be classified negatively affected by phytates.
according to the base mechanism; Hydrogen Atom Transfer (HAT) or The suggested critical value of the [phytate]:[calcium] molar ratio
Single Electron Transfer (SET). The DPPH assay was used in this study is less than 0.17 (Sánchez-Moya et al., 2019; Udomkun et al., 2019;
combines these two mechanisms; the DPPH radical reacts fast through Umeta, West, & Fufa, 2005). However, Morris and Ellis (1985) men-
electron transfer and slowly with hydrogen atom transfer. In this con- tioned that a [phytate]:[calcium] > 0.24 M ratio is already responsible
text, each type of compound reacts in its own way, phenols (without an for decreasing calcium absorption. Then, the Ca bioaccessibility of the
aromatic ring) present fast reactions, while phenols with aromatic rings cashew nut samples (8–11) could be negatively affected by phytates.
and acid groups present slow reactions. Once the DPPH radical site The molar ratio [phytate]:[magnesium] ranged from 0.10 to 0.26,
hinders the access by phenolic compounds, all the reactions are ex- and these values were considered too low to interfere with Mg bioac-
pected to slow, especially HAT (Schaich, Tian, & Xie, 2015). cessibility (Israr, Frazier, & Gordon, 2013). Zinc availability is corre-
The composition of the PBB evaluated in this research is distinct lated when the [phytate]:[zinc] molar ratio is less than 18 (Sánchez-
from each other, not only the raw matrices, but also by the ingredients Moya et al., 2019). Although, Ellis et al. (1987) recommended a molar
used by the producers. The samples with the highest value of TPC are ratio ≤ 10 [phytate]:[zinc] for adequate bioavailability of zinc. In this
not the same that showed the highest values of antioxidant capacity by context, only soy (sample 24) was not influenced by the antinutritional
DPPH assay. This fact could be related to the matrix effects, DPPH assay effect of phytic acid.
limitations and the unknown bioactive compound composition of each Theodoropoulos et al. (2018) studied soy beverages and observed
sample. The identification of the phenolic compounds profile was not that myo-inositol fractions (IP3, IP4, IP5, IP6) were present in almost all
performed in this study, we conducted a preliminary study about the the soy beverages studied. It ranged from not detected (ND) to
total phenolic compounds of the samples. 0.2 μmol/g for IP3, 0.2 to 0.9 μmol/g to IP4, 0.2 to 5.2 μmol/g to IP5
In this context, we observed that there is a large variation on the and 0.49 to 19.7 μmol/g for IP6. These results are congruent with the
antioxidant capacity and composition of evaluated beverages. The an- results found in our research, where IP6 was also the most prevalent
tioxidant capacity could be low (ND) or high according to the PBB component in the samples with a mean concentration of
chosen for consumption. 11.88 ± 0.28 μmol/g.
The myo-inositol fractions were evaluated in almond meal and al-
3.1.3. Quantification of myo-inositol phosphates mond brown skins and six forms of myo-inositol phosphates were de-
The myo-inositol phosphates (IP3, IP4, IP5, IP6) were evaluated in tected, they ranged from 8 to 12 µmol/g in the meal and 5 to 14 µmol/g
samples of PBB (Table 1). The myo-inositol trisphosphate (IP3) could in the brown skins and IP6 was the dominant form (Duong, Clark,
not be detected in the beverages. The myo-inositol tetrakisphosphate Lapsley, & Pegg, 2017b). Other research detected myo-inositol phos-
(IP4) was only observed in the non-blended cashew nut beverage phates in cashew nuts (5.02 µmol/g), Brazil nuts (20.08 µmol/g), rice
(2.05 ± 0.04 µmol/L) of the brand D (sample 12). The concentrations (97.36 µmol/g) and oats (54.47 mg/240 g) (Zhang et al., 2007).
of myo-inositol pentakisphosphate (IP5) ranged from 0.32 to However, in this research we did not observed myo-inositol phosphates
0.66 µmol/L, for the samples 15 (peanut) and 11 (cashew nut), re- in rice and oats beverages; probably due the amount of raw material,
spectively. Except for the peanuts, all the samples that presented IP5 and variations of, used in production process and dilution with water
fractions were from the cashew nut class, blended or not with other (moisture variation).
components.
A myo-inositol hexakisphosphate (IP6, phytate) was observed in all 3.2. Essential mineral content and in vitro digestion for bioaccessibility
the cashew nut beverages, one almond, peanut and soy. The IP6 con- estimation of Ca, Fe, Mg and Zn
centration ranged from 0.16 to 1.46 µmol/L. The lowest concentration
was found in the almond beverage and the highest in the peanut bev- 3.2.1. Method validation for calcium, iron, magnesium and zinc in PBB
erage. All the cashew nut beverages and the soy beverage were con- The parameter evaluation for linearity and sensitivity were made
sidered equal. using analytical curves (five to seven points) for each element, Ca
The antinutritional effect of myo-inositol hexaphosphate (IP6), or (0.5–5.0 mg/L), Fe (0.1–2 mg/L), Mg (0.025–0.5 mg/L), and Zn
phytic acid, and also IP5, in reducing or inhibiting the absorption of (0.05–0.35 mg/L). The curves were linear with suitable percentages of
minerals are known. These compounds actively bind to divalent cations explained variance (R2 > 0.99). The regressions were considered
such as Ca, Fe and Zn, forming insoluble complexes and as a con- significant through ANOVA evaluation and no lack of fit for the linear
sequence decrease the absorption of these minerals in the body models were observed.
(Hurrell, 2004; Sandberg et al., 1999). Thus, mineral availability can The limits of detection (LOD) and quantification (LOQ) values for
also be estimated by evaluating the phytate (IP5 + IP6)/mineral molar the minerals were Fe, Ca, Mg and Zn were 0.04; 0.45; 0.01 and
ratio and can be used as a tool to predict the inhibitory effect of this 0.05 mg/L, respectively. The LOQ were 0.11; 0.81; 0.02 and 0.07 mg/L
antinutrient on mineral bioavailability (Sánchez-Moya et al., 2019; for Fe, Ca, Mg and Zn, respectively.

5
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

Precision and accuracy were evaluated in rice and cashew nut samples 21 and 23 (oat) had a calcium content equal to milk
beverages. Precision was evaluated using parameters of repeatability (1531.0 mg/L), but all of were fortified with calcium. They also pre-
and intermediate precision. The repeatability was considered appro- sented more calcium than soy, a more common PBB in the markets.
priate for the analysis, due to the relative standard deviation (RSD) When evaluating Ca fortified samples, we observed that despite the
obtained for 7 repetitions of each mineral (RSD rice values: Ca 3.46, Fe different types of calcium, there was significant variation between the
1.95, Mg 4.76 and Zn 4.76; RSD cashew nut values: Ca 4.57, Fe 3.84, fortified samples.
Mg 4.76 and Zn 4.76) being lower than the maximum RSD established A similar comparison was observed by Chalupa-Krebzdak et al.
by the guidelines of validation (AOAC, 2016; Brasil, 2016). According (2018). They evaluated the USDA Food Composition Database of
to the guidelines, samples with concentrations of up to 1, 10 and common PBB products (soy, coconut, almond, and cashew nut) and
100 ppm and 0.1% of analyte, the RSD should be 11, 7.3, 5.3 and 3.7, observed that the Ca content was highly variable. The fortified bev-
respectively. Intermediate precision, was assessed using ANOVA on erages contained higher total Ca when compared to milk, but non-for-
three different days and was not significantly different (P > 0.05), tified PBB had drastically lower levels of calcium, ranging from 0 to
with 95% of confidence. 188 mg/100 g for non-fortified and fortified, respectively.
The accuracy assay was performed according to INMETRO (Brasil, The dialyzed calcium ranged from 8.2 to 306.6 mg/L, for the sam-
2016) and was evaluated through recovery tests for all minerals. The ples 18 (coconut) and 24 (soy), respectively, corresponding to 1.6 and
rice and cashew nut beverage samples were fortified with 50, 100% or a 45.9% of calcium dialyzed. For other beverages the Ca dialyzed was
known concentration of each mineral, regarding the initial value pre- under the LOQ. Some beverages such as samples 2, 4, 5 (rice), 12
sent in the samples studied. If residual carbon was less than 0.1% in the (cashew nut) and 24 (soy) had higher amounts and percentages of
beverages after mineralization, according Gouveia, Silva, Costa, dialyzed calcium than milk; with 13.1, 15.4, 21.2, 17.2 and 45.9% of
Nogueira, and Nóbrega (2001), then the digestion procedure was con- dialyzed calcium and only 7.10% in milk.
sidered efficient for analysis of plant-based beverages by FAAS. Theodoropoulos et al. (2018) demonstrated that soy beverages have
a percentage of Ca dialysis ranging from 3.3 to 5.4%, values lower than
3.2.2. Mineral content and effect of in vitro digestion on bioaccessibility what was found in the current study (45.9%), likely due the type of
Several minerals in PBB were evaluated, including calcium, mag- calcium used for fortification. Besides that, they also observed that
nesium, iron and zinc and bioaccessibility was assessed through in vitro samples treated with phytase, had a decrease in IP fractions, which
digestion assay. The analyses were also carried out for a milk sample for increased the dialyzed Ca percentage.
comparison. The differences between the dialyzed calcium might be related to
The total calcium content and dialyzed fractions were evaluated in the interactions of Ca with the other compounds of the food matrix, and
the samples and compared (Table 2). may also be affected by the calcium compounds used during fortifica-
Results show a high variation in the calcium content of the bev- tion (Chalupa-Krebzdak et al., 2018). The PBB studied were fortified
erages, ranging from 10.00 mg/L to 1697.33 mg/L, for samples 1 (rice) with calcium carbonate, tricalcium phosphate, calcium lactate and
and 20 (coconut), respectively. The highest amounts were observed in calcium from algae. Studies observed that different compounds (che-
samples that were fortified with some calcium compound (some brands mical forms) of calcium had different behaviors in Ca bioaccessibility,
used calcium lactate, calcium phosphate, calcium carbonate and cal- are affected by pH, molecular interactions with other compounds of the
cium from algae). Some of them, sample 6 (rice), sample 12 (cashew digestive tract, and other factors (Lorieau et al., 2018). Other studies
nut), sample 13 (almond), sample 15 (peanut), sample 20 (coconut), show that calcium carbonate in puffed rice extrudates and rice noodles

Table 2
Content of total, dialyzed and percentage of dialysis for calcium and magnesium.
Type Sample Calcium Magnesium

Total Ca (mg/L) Dialyzed Ca (mg/L) Dialysis (%) Total Mg (mg/L) Dialyzed Mg (mg/L) Dialysis (%)

Rice 1 10.00 ± 0.85 i < LOQ ND 12.63 ± 0.24 h 2.46 ± 0.21 j 9.5
2 1149.33 ± 48.01 d 151.15 ± 7.94 cd 13.2 27.69 ± 2.30 ijh 8.91 ± 0.62 ij 34.8
3 170.80 ± 12.12 hi 15.81 ± 1.68 f 9.3 122.80 ± 9.44 fgh 11.24 ± 1.79 ij 9.2
4 1091.50 ± 97.10 d 168.30 ± 17.50 bcd 15.4 6.29 ± 0.47 h 1.73 ± 0.21 j 30.2
5 858.20 ± 44.15 ef 182.02 ± 7.49 bc 21.2 24.82 ± 2.34 ijh 12.32 ± 0.19 ij 53.2
6 1619.00 ± 68.61 ab 89.52 ± 9.82 e 5.5 48.99 ± 3.25 ij 19.84 ± 0.89 ghi 35.7
7 142.43 ± 0.67 hi < LOQ ND 31.37 ± 1.14 ijh 15.73 ± 1.91 hi 56.6
Cashew nut 8 53.70 ± 3.25 i < LOQ ND 196.33 ± 13.76 bc 35.13 ± 8.09 def 17.7
9 67.10 ± 2.00 hi < LOQ ND 251.23 ± 14.20 a 40.03 ± 1.54 de 15.4
10 45.34 ± 0.11 i < LOQ ND 195.30 ± 13.95 bc 31.21 ± 0.23 efg 13.3
11 97.21 ± 2.14 hi < LOQ ND 268.43 ± 2.83 a 45.88 ± 4.59 cd 17.1
12 1637.33 ± 107.87 ab 281.31 ± 25.36 a 17.2 168.33 ± 9.40 cd 60.93 ± 8.16 b 31.3
Almond 13 1482.67 ± 152.18 bc 89.82 ± 9.66 e 6.1 153.67 ± 15.77 def 42.51 ± 4.31 de 24.1
14 266.47 ± 21.37 h 9.44 ± 0.38 f 3.5 32.98 ± 2.66 ijh 8.73 ± 1.12 ij 22.5
Peanut 15 1641.70 ± 135.36 ab 143.97 ± 12.72 d 8.8 169.60 ± 17.44 cd 44.74 ± 1.01 cd 23.6
Coconut 16 35.46 ± 2.86 i 8.25 ± 0.60 f 23.3 157.60 ± 21.66 de 56.20 ± 1.25 bc 30.4
17 876.70 ± 10.48 e 34.94 ± 4.22 f 3.9 26.69 ± 1.40 ijh 8.07 ± 1.40 ij 27.5
18 511.80 ± 24.79 g 8.21 ± 0.51 f 1.6 131.63 ± 5.62 efg 107.42 ± 10.65 a 65.3
19 527.90 ± 30.22 g 10.91 ± 0.59 f 2.1 206.83 ± 20.91 b 62.43 ± 0.76 b 25.7
20 1697.33 ± 92.59 a 197.11 ± 4.72 b 11.6 121.16 ± 8.59 gh 42.98 ± 3.37 de 32.9
Oat 21 1477.33 ± 121.96 bc 78.40 ± 6.31 e 5.3 56.00 ± 2.34 i 19.09 ± 1.33 ghi 29.6
22 139.43 ± 1.19 hi < LOQ ND 22.01 ± 2.25 jh 3.30 ± 0.55 j 14.2
23 1359.67 ± 72.86 c 32.69 ± 0.74 f 2.4 92.60 ± 5.96 h 26.39 ± 4.06 fgh 27.1
Soy 24 668.00 ± 18.19 fg 306.65 ± 13.57 a 45.9 179.37 ± 6.29 bcd 60.02 ± 6.30 b 28.1
Milk 25 1531.00 ± 33.06 abc 108.64 ± 11.22 e 7.1 94.08 ± 3.52 h 34.60 ± 2.07 def 35.6

Results are expressed as mean ± standard deviation, n = 3. Mean with different letters in the same column indicate significant difference (p < 0.05) as
determinate using one way ANOVA and Tukey test at 95% of confidence.

6
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

are more bioaccessible than tricalcium phosphate, but not in com- Table 3
parative studies of bioaccessibility of all different calcium compounds Content of total iron and zinc.
(Janve & Singhal, 2018). Iron Zinc
However, it is not possible to determine the exact influence of each Type Sample Total Fe (mg/L) Total Zn (mg/L)
calcium compound on bioaccessibility in PBB, because some brands do
Rice 1 < LOQ < LOQ
not inform which type of calcium is used. This information could ensure
2 < LOQ 0.86 ± 0.05 k
that future studies can be performed in order to estimate the best kind 3 0.92 ± 0.04 hi < LOQ
of calcium source for plant-based beverages. 4 0.77 ± 0.00 i < LOQ
Despite fortification and different calcium compounds, phenolic 5 0.87 ± 0.09 i 1.47 ± 0.07 hij
compounds and phytates could affect Ca bioaccessibility (Vitali, 6 12.89 ± 0.92 a 1.10 ± 0.08 ijk
7 1.59 ± 0.15 ghi 1.78 ± 0.05 gh
Vedrina Dragojević, & Šebečić, 2008). In our research we observed that
Cashew nut 8 3.91 ± 0.20 de 2.83 ± 0.32 d
TPC could influence Ca bioaccessibility and that most of the samples 9 4.56 ± 0.57 d 3.91 ± 0.27 c
with phytates are related to low or undetected Ca bioaccessibility, as 10 3.97 ± 0.17 d 4.63 ± 0.017 b
observed for most of the cashew nut samples. 11 10.32 ± 0.26 b 3.74 ± 0.02 c
12 3.79 ± 0.20 de 2.16 ± 0.10 efg
The total magnesium content ranged from 6.29 to
Almond 13 2.13 ± 0.23 fgh 1.68 ± 0.27 ghi
251.23–268.43 mg/L for samples 4 (rice), 9 and 11 (cashew nut with 14 < LOQ < LOQ
Brazil nut and cashew nut with cacao), respectively (Table 2). The Peanut 15 7.79 ± 0.47 c 4.05 ± 0.21 bc
samples with the highest amounts of Mg were in the cashew nut class Coconut 16 3.79 ± 0.24 de 1.59 ± 0.05 ghi
and with no added Mg. However, they contained Brazil nut and cacao 17 < LOQ 0.65 ± 0.04 k
18 < LOQ 2.16 ± 0.06 efg
with Mg also in their composition, likely leading to the differences in
19 2.68 ± 0.14 efg 2.43 ± 0.14 def
Mg from the other beverages due to the presence of these raw materials 20 10.51 ± 1.04 b 1.97 ± 0.12 fgh
(Cardoso, Duarte, Reis, & Cozzolino, 2017). Oat 21 2.13 ± 0.16 fgh 0.58 ± 0.06 k
Coconut samples 18 and 19 are from the same brand and Mg was 22 < LOQ 0.65 ± 0.03 k
added, but despite this, it did not have the highest Mg content when 23 1.91 ± 0.06 ghi < LOQ
Soy 24 3.35 ± 0.27 def 8.32 ± 0.63 a
compared with the beverages of cashew nuts, which naturally have a Milk 25 < LOQ 2.65 ± 0.19 de
high Mg content. Some beverages from almond, peanut, coconut, soy
and all the cashew nut class had higher amounts of Mg when compared Results are expressed as mean ± standard deviation, n = 3. Mean with dif-
with milk. ferent letters in the same column indicate significant difference (p < 0.05) as
It was possible to quantify dialyzed Mg in all the samples, and it determinate using one way ANOVA and Tukey test at 95% of confidence.
ranged from 1.90 to 107.41 mg/L for the samples 4 (rice) and 18 (co- Values for dialyzed (mg/L) and percentage of dialysis (%) for iron and zinc were
conut), respectively. This corresponds to 30.2 to 65.3% of dialysis. This under the LOQ.
means that the sample with the lowest Mg content also had the lowest
amount of dialyzed Mg, but the beverage with the highest content of amounts under the LOQ. Only sample 16 (coconut) had 1.02 mg/
total Mg was not the same with the highest Mg dialyzed. L ± 0.15 of dialyzed Fe, which represented 31.76% of dialyzed iron.
Sample 18 had the highest dialyzed Mg (65.3%) and is fortified with According to the PCA results (Fig. 1), the low dialysis percentage of iron
Mg, despite this, sample 19, also from the same brand and fortified with could be related to samples with high amounts of total and dialyzed Ca,
Mg only had 25.7% of dialyzed Mg. Sample 19 has in its composition Zn, and with levels of IP4, IP5 and IP6, suggesting competition between
cacao, which could be related to the low Mg dialyzed when compared the minerals and a negative influence of the myo-inositol phosphates.
to the sample without cacao (sample 18), probably because cacao is Strain and Cashman (2009) related that a competition between divalent
known as source of phenolic compounds, that are related to the de- minerals could occur, as Fe, Ca and Zn might decrease Fe bioaccessi-
crease of mineral bioaccessibility (Gibson, Newsham, Gibson, & bility.
Newsham, 2018). Zinc content ranged from 0.57 mg/L to 8.13 mg/L for samples 21
The comparison between samples fortified with Mg compounds and (oat) and 24 (soy), respectively. Other beverages with high zinc content
ones that have naturally occurring Mg, show that both samples have were cashew nuts (8, 9, 10, 11) and coconut beverages (18 and 19)
bioaccessible Mg considered are equal, as observed for samples 19 fortified with zinc. However, it was not possible to quantify Zn for some
(fortified) and 12 (naturally occurring), therefore naturally occurring samples of rice, almond and oat, as the values were under the LOQ for
Mg samples could have Mg dialyzed amounts similar to samples for- the method.
tified with Mg salts. This might be related to the other compounds The soy beverage was the only that had an amount of Zn dialyzed,
present in the samples which could interact with the minerals. Sample with 0.51 mg/L ± 0.03, representing 6.15% of zinc dialyzed. This
12 is not fortified with Mg and has less total Mg content (compared to could be related to the initial amount of zinc in the samples and also
sample 19), however it has a lower TPC (2.98 mg GAE/L) when com- due to possible interactions of zinc with the matrices in the sample.
pared with sample 19 (7.77 mg GAE/L); possibly due the added cacao Other studies analyzed Fe and Zn in soy beverages, naturally occurring
which is a source of phenolic compounds (5624.23 mg/100 g) (Phenol- or added, and observed an increase in the bioaccessibility of Fe and Zn
Explorer, , 2019). Therefore, reinforcing the hypothesis that TPC may in the beverages when smaller amounts of myo-inositol phosphates
decrease the absorption of Mg. Furthermore, sample 19 has cacao in its were present (Theodoropoulos et al., 2018).
composition, which is a source of phenolic compounds (5624.23 mg/ From the results of essential minerals content analysis, we estimated
100 g). the contribution of minerals Ca, Mg, Fe and Zn towards the
The total content of iron and zinc are presented in the Table 3. Recommended Daily Intake (RDI) was calculated from 200 mL of PBB
Total iron content ranged from 0.76 mg/L to 12.89 mg/L for the and compared to equal amounts of traditionally consumed or know
samples 4 (rice) and 6 (rice), respectively. The difference in iron levels beverages of cow's milk and soybean, as they are traditionally con-
between the two samples is that sample 4 is not fortified while the sumed and known beverages. According to the combined Food
sample 6 has iron added. The following samples contained high iron Agriculture Organization and World Health report (FAO/OMS, 2001)
content; 11 (cashew nut) and 20 (coconut) with 10.32 and 10.50 mg/L, the approximate RDI for Ca, Mg, Fe and Zn is 1000 mg, 400 mg, 18 mg
with no significant difference. These have not been fortified with iron, and 8 mg, respectively.
however, they have cacao as a common additive. Thus, considering the samples with the highest contents of each
Dialysis assays showed that the samples had of dialyzed iron mineral, the ingestion of one portion of coconut (sample 20) will

7
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

Fig. 1. Principal Component Analysis for evaluation of total and bioaccessible minerals, phenolic compounds, antioxidant capacity, and myo-inositol phosphates.
Graphs of scores (A) and loadings (B).

contribute 33.9% of Ca RDI, while cow and soybean milks will con- that were added or absent in the formulation; mineral fortification (Ca,
tribute 30.6 and 13.4%, respectively. For Mg, intake of a portion of Fe, Mg, Zn, Se), different chemical types of fortificant compounds
cashew (sample 11) will contribute 13.4%, while intake of a portion of (calcium carbonate, calcium lactate, calcium phosphate, calcium from
cow's milk and soy will contribute 4.7 and 9.0% of the RDI. Regarding algae), addition of vitamins and other compounds such as sweeteners,
iron, intake of a single portion of cow's milk will not contribute to the maltodextrin, polydextrose, starch, gums, thickeners, and stabilizers,
RDI due to the low content of this mineral in its composition, while soy might also have influence in the sample discrimination.
and rice milk (sample 6) may contribute to 3.3 and 13.0% of the RDI, According to the properties evaluated in this study, the composition
respectively. Considering zinc, intake of a portion of cow's milk and of the PBB could vary notably according to different sample composi-
soybeans will contribute 2.7 and 8.3% of RDI, respectively, while tion and production process, the ingredients added, raw material used
cashew (sample 10) will contribute 4.63% of RDI for this mineral. and percentage of raw material added. This is in accordance with the
The Principal Component Analysis (PCA) was performed as an ex- results of Chalupa-Krebzdak et al. (2018) and (Jeske et al., 2017) that
ploratory analysis to evaluate all data obtained and determine possible evaluated some properties of PBB.
correlations between the samples and the variables studied (Fig. 1).
The PCA was built using the 25 samples in triplicate and with 13 4. Conclusions
variables (TPC, antioxidant capacity, IP4, IP5, IP6, total calcium, iron,
magnesium, zinc and dialyzed calcium, iron, magnesium and zinc), The study demonstrated that some samples of PBB have bioactive
resulting in a 75 × 13 matrix, representing 975 assays. The data was potential, according the total phenolic compound content and anti-
auto-scaled, to give equal importance to all variables with differing oxidant capacity in their composition. The sample with the highest TPC
magnitudes. was a rice beverage with 12.39 mg GAEq/L, considered higher than
The first and second PC’s together represent 48.44% of the total milk; and the highest antioxidant capacity was an oat beverage, with
explained variance. Other PC’s were tested and they did not contribute 306.46 µmol TE/L, considered equal to milk. They also have low or no
to the increase of total explained variance, which indicates that other detectable myo-inositol phosphates content; however, samples of
components could have more impact in sample differentiation. The cashew nuts had IP fractions that could bind with minerals.
graph of scores display how the samples are distributed. The graph It was possible to establish a method for mineral evaluation (Ca, Fe,
indicates the formation of distinct groups (coconut, cashew nut, peanut, Mg and Zn) in different PBB. The method presented suitable parameters
soy) and one large group composed of mixed sample types. The load- for sensitivity, linearity, limit of detection and quantification, accuracy,
ings graph indicates the variables and influences leading to the se- and recovery for the mineral quantification in plant-based beverages.
paration of groups. PC1 separated rice, almond, oats, and milk samples In addition, the content and bioaccessibility of essential minerals
and one sample of coconut (17) from the coconut, cashew nuts, peanut evaluated varied widely; samples of PBB not fortified with Ca presented
and soy beverages groups. less amount of Ca content than cow's milk, however samples that were
The main group of cashew nut is related to high amounts of total fortified presented comparable values for total Ca and more Ca bioac-
and dialyzed Mg and also the presence of IP5 and IP6. On the other cessibility than cow's milk. The beverages presented high amounts of
hand, the samples of rice, almond, and oats were separated, probably Mg content and dialyzed Mg, with also Fe and Zn in their composition;
because they contained, in general, low amounts of all variables. however, low amounts of dialyzed Fe and Zn were verified. Regarding
Therefore, it is possible differentiate samples of coconut, cashew nut, the type of raw material used in the production process and beverage
peanut and soy from the other samples. According to the PCA, milk did composition, it was possible to demonstrate that PBB are able to pro-
not have considerable differences and was unable to be distinguished vide a higher percentage of RDI than cow’s milk, with either fortified or
from the PBB. In addition, cashew nut samples (not fortified with Ca) non-fortified samples. Indicating that PBB are a viable alternative, and
have in their composition IP5 and IP6, and that the Ca dialyzed was not are able to deliver the equal amount or better than the RDI of regular
detected in all of them, indicating a possible negative effect on bioac- cow’s milk.
cessibility of this mineral influenced by this antinutritional factor. Therefore, the substitution of milk for PBB should be carefully
Differentiation between beverages was achieved possibly not only evaluated, according to the RDI of each mineral. In addition, consumers
by raw material used, but also by the presence of different ingredients are able to determine nutritional value due to the available information

8
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

of beverage composition (ingredients used), by evaluating how much of Chen, J., Zhao, X., Zhang, X., Yin, S., Piao, J., Huo, J., ... Chen, C. (2005). Studies on the
the raw material was added and presence of other ingredients that may effectiveness of NaFeEDTA-fortified soy sauce in controlling iron deficiency: A po-
pulation based intervention trial. Food and Nutrition Bulletin, 26(2), 177–186. http://
or may not improve the nutritional value of this kind of product. www.unu.edu/unupress/food/Unupress.htm.
Codina-Torrella, I., Guamis, B., Ferragut, V., & Trujillo, A. J. (2017). Potential application
CRediT authorship contribution statement of ultra-high pressure homogenization in the physico-chemical stabilization of tiger
nuts’ milk beverage. Innovative Food Science and Emerging Technologies, 40, 42–51.
https://doi.org/10.1016/j.ifset.2016.06.023.
Joyce Grazielle Siqueira Silva: Conceptualization, Data curation, Diarra, K., Nong, Z. G., & Jie, C. (2005). Peanut milk and peanut milk based products
Formal analysis, Investigation, Methodology, Project administration, production: A review. Critical Reviews in Food Science and Nutrition, 45(5), 405–423.
https://doi.org/10.1080/10408390590967685.
Validation, Visualization, Writing - original draft, Writing - review & Duong, Q. H., Clark, K. D., Lapsley, K. G., & Pegg, R. B. (2017a). Determination of myo-
editing. Ana Paula Rebellato: Data curation, Formal analysis, inositol phosphates in tree nuts and grain fractions by HPLC–ESI–MS. Journal of Food
Visualization, Writing - original draft, Writing - review & editing. Elem Composition and Analysis, 59, 74–82. https://doi.org/10.1016/j.jfca.2017.02.010.
Duong, Q. H., Clark, K. D., Lapsley, K. G., & Pegg, R. B. (2017b). Quantification of inositol
Tamirys dos Santos Caramês: Formal analysis, Methodology, Writing
phosphates in almond meal and almond brown skins by HPLC/ESI/MS. Food
- original draft. Ralf Greiner: Funding acquisition, Resources, Chemistry, 229, 84–92. https://doi.org/10.1016/j.foodchem.2017.02.031.
Supervision, Writing - review & editing. Juliana Azevedo Lima Ellis, R., Kelsay, J. L., Reynolds, R. D., Morris, E. R., Moser, P. B., & Frazier, C. W. (1987).
Pallone: Conceptualization, Funding acquisition, Resources, Phytate: Zinc and phytate x calcium: Zinc millimolar ratios in self-selected diets of
American, Asian Indians, and Nepalese. Journal of the American Dietetic Association,
Supervision, Visualization, Writing - review & editing. 87, 1043–1047.
Faccin, G. L., Miotto, L. A., Vieira, L. D. N., Barreto, P. L. M., & Amante, E. R. (2009).
Acknowledgements Chemical, sensorial and rheological properties of a new organic rice bran beverage.
Rice Science, 16(3), 226–234. https://doi.org/10.1016/S1672-6308(08)60083-9.
Feitosa, S., Greiner, R., Meinhardt, A., Müller, A., Almeida, D. T., & Posten, C. (2018).
The authors acknowledge the Coordination of Improvement of Effect of traditional household processes on iron, zinc and copper bioaccessibility in
Higher Level Personnel (CAPES) - Finance Code 001 and process black bean (Phaseolus vulgaris L.). Foods, 7(8), 123. https://doi.org/10.3390/
foods7080123.
88881.188691/2018-01, the National Council for Scientific Felberg, I., Antoniassi, R., Deliza, R., Freitas, S. C. D., & Modesta, R. C. D. (2009). Soy and
Development (CNPq) – Process Numbers 142415/2016-2 and 142414/ Brazil nut beverage: Processing, composition, sensory, and color evaluation. Ciência e
2016-6 and the Sao Paulo Research Foundation (FAPESP) – Process Tecnologia de Alimentos, 29(3), 609–617. https://doi.org/10.1590/S0101-
20612009000300024.
Number 2018/09759-3. Gibson, M., Newsham, P., Gibson, M., & Newsham, P. (2018). Chocolate/Cacao. Food
Science and the Culinary Arts, 341–352. https://doi.org/10.1016/B978-0-12-811816-
Declaration of Competing Interest 0.00017-8.
Gong, E. S., Luo, S. J., Li, T., Liu, C. M., Zhang, G. W., Chen, J., ... Liu, R. H. (2017).
Phytochemical profiles and antioxidant activity of brown rice varieties. Food
The authors declare none conflict of interest. Chemistry, 227, 432–443. https://doi.org/10.1016/j.foodchem.2017.01.093.
Gouveia, S. T., Silva, F. V., Costa, L. M., Nogueira, A. R. A., & Nóbrega, J. A. (2001).
Appendix A. Supplementary data Determination of residual carbon by inductively-coupled plasma optical emission
spectrometry with axial and radial view configurations. Analytica Chimica Acta,
445(2), 269–275. https://doi.org/10.1016/S0003-2670(01)01255-7.
Supplementary data to this article can be found online at https:// Greiner, R., Konietzny, U., & Jany, K. D. (2006). Phytate - an undesirable constituent of
doi.org/10.1016/j.foodres.2020.108993. plant - based foods? Journal Fur Ernahrungsmedizin, 8(3), 18–28.
Herbello-Hermelo, P., Lamas, J. P., Lores, M., Domínguez-González, R., Bermejo-Barrera,
P., & Moreda-Piñeiro, A. (2018). Polyphenol bioavailability in nuts and seeds by an in
References vitro dialyzability approach. Food Chemistry, 254, 20–25. https://doi.org/10.1016/J.
FOODCHEM.2018.01.183.
Hurrell, R. F. (2004). Phytic acid degradation as a means of improving iron absorption.
Abderrahim, F., Huanatico, E., Segura, R., Arribas, S., Gonzalez, M. C., & Condezo-Hoyos,
International Journal for Vitamin and Nutrition Research, 74(6), 445–452.
L. (2015). Physical features, phenolic compounds, betalains and total antioxidant
Hurrell, R. F., Juillerat, M. A., Reddy, M. B., Lynch, S. R., Dassenko, S. A., & Cook, J. D.
capacity of coloured quinoa seeds (Chenopodium quinoa Willd.) from Peruvian
(1992). Soy protein, phytate, and iron absorption in humans. The American Journal of
Altiplano. Food Chemistry, 183, 83–90. https://doi.org/10.1016/J.FOODCHEM.2015.
Clinical Nutrition, 56(3), 573–578. https://doi.org/10.1093/ajcn/56.3.573.
03.029.
Israr, B., Frazier, R. A., & Gordon, M. H. (2013). Effects of phytate and minerals on the
Alegría-Torán, A., Barberá-Sáez, R., & Cilla-Tatay, A. (2015). Bioavailability of minerals
bioavailability of oxalate from food. Food Chemistry, 141(3), 1690–1693. https://doi.
in foods. In M. G. Guardia, & S. Garrigues (Eds.). Handbook of mineral elements in food
org/10.1016/j.foodchem.2013.04.130.
(pp. 41–67). Wiley.
Janve, M., & Singhal, R. S. (2018). Fortification of puffed rice extrudates and rice noodles
Amalraj, A., & Pius, A. (2015). Bioavailability of calcium and its absorption inhibitors in
with different calcium salts: Physicochemical properties and calcium bioaccessibility.
raw and cooked green leafy vegetables commonly consumed in India - An in vitro
LWT, 97, 67–75. https://doi.org/10.1016/J.LWT.2018.06.030.
study. Food Chemistry, 170, 430–436. https://doi.org/10.1016/j.foodchem.2014.08.
Jeske, S., & Arendt, E. K. (2018). Past, present and future: The strength of plant-based
031.
dairy substitutes based on gluten-free raw materials. Food Research International, 110,
AOAC (2016). Association of official analytical chemists international. Appendix F:
42–51. https://doi.org/10.1016/J.FOODRES.2017.03.045.
Guidelines for standard method performance requirements.
Jeske, S., Zannini, E., & Arendt, E. K. (2017). Evaluation of physicochemical and gly-
Bernat, N., Chafer, M., Chiralt, A., & Gonzalez-Martinez, C. (2014). Vegetable milks and
caemic properties of commercial plant-based milk substitutes. Plant Foods for Human
their fermented derivative products. International Journal of Food Studies, 3(April),
Nutrition, 72(1), 26–33. https://doi.org/10.1007/s11130-016-0583-0.
93–124.
John, J. A., & Shahidi, F. (2010). Phenolic compounds and antioxidant activity of Brazil
Bosscher, D., Van Caillie-Bertrand, M., Van Cauwenbergh, R., & Deelstra, H. (2003).
nut (Bertholletia excelsa). Journal of Functional Foods, 2(3), 196–209. https://doi.org/
Availabilities of calcium, iron, and zinc from dairy infant formulas is affected by
10.1016/j.jff.2010.04.008.
soluble dietary fibers and modified starch fractions. Nutrition, 19(7–8), 641–645.
Jovani, M., Barbera, R., Farre, R., & Aguilera, E. M. (2001). Calcium, iron, and zinc uptake
https://doi.org/10.1016/S0899-9007(03)00063-7.
from digests of infant formulas by Caco-2 cells. Journal of Agricultural and Food
Brasil (2016). INMETRO. Instituto Nacional de Metrologia, Normalização e Qualidade
Chemistry, 49(7), 3480–3485.
Industrial.Orientações sobre Validação de Métodos de Ensaios Químicos, DOQ-
Kaur, J., Whitson, A., Ashton, J., Katopo, L., & Kasapis, S. (2016). Effect of ultra high
CGCRE-008.
temperature processing and storage conditions on phenolic acid, avenanthramide,
Briviba, K., Gräf, V., Walz, E., Guamis, B., & Butz, P. (2016). Ultra high pressure homo-
free fatty acid and volatile profiles from Australian oat grains. Bioactive Carbohydrates
genization of almond milk: Physico-chemical and physiological effects. Food
and Dietary Fibre(September), https://doi.org/10.1016/j.bcdf.2016.09.002.
Chemistry, 192, 82–89. https://doi.org/10.1016/J.FOODCHEM.2015.06.063.
Kornsteiner, M., Wagner, K. H., & Elmadfa, I. (2006). Tocopherols and total phenolics in
Cardoso, B. R., Duarte, G. B. S., Reis, B. Z., & Cozzolino, S. M. F. (2017). Brazil nuts:
10 different nut types. Food Chemistry, 98(2), 381–387. https://doi.org/10.1016/j.
Nutritional composition, health benefits and safety aspects. Food Research
foodchem.2005.07.033.
International, 100, 9–18. https://doi.org/10.1016/J.FOODRES.2017.08.036.
Kumar, V., Sinha, A. K., Makkar, H. P. S., & Becker, K. (2010). Dietary roles of phytate and
Chalupa-Krebzdak, S., Long, C. J., & Bohrer, B. M. (2018). Nutrient density and nutri-
phytase in human nutrition: A review. Food Chemistry, 120(4), 945–959. https://doi.
tional value of milk and plant-based milk alternatives. International Dairy Journal, 87,
org/10.1016/j.foodchem.2009.11.052.
84–92. https://doi.org/10.1016/J.IDAIRYJ.2018.07.018.
Lorieau, L., Le Roux, L., Gaucheron, F., Ligneul, A., Hazart, E., Dupont, D., & Floury, J.
Chang, S. K., Alasalvar, C., Bolling, B. W., & Shahidi, F. (2016). Nuts and their co-pro-
(2018). Bioaccessibility of four calcium sources in different whey-based dairy ma-
ducts: The impact of processing (roasting) on phenolics, bioavailability, and health
trices assessed by in vitro digestion. Food Chemistry, 245, 454–462. https://doi.org/
benefits - A comprehensive review. Journal of Functional Foods, 26, 88–122. https://
10.1016/J.FOODCHEM.2017.10.108.
doi.org/10.1016/j.jff.2016.06.029.
Luten, J., Crews, H., Flynn, A., Van Dael, P., Kastenmayer, P., Hurrel, R., ... Frolich, W.

9
J.G.S. Silva, et al. Food Research International 130 (2020) 108993

(1996). Interlaboratory trial on the determination of the in vitro iron dialysability Fast determination of phenolic compounds in rice grains by ultraperformance liquid
from food. Journal of the Science of Food and Agriculture, 72(4), 415–424. chromatography coupled to photodiode array detection: Method development and
Mäkinen, O. E., Uniacke-Lowe, T., O’Mahony, J. A., & Arendt, E. K. (2015). validation. Journal of Agricultural and Food Chemistry, 67(10), 3018–3027.
Physicochemical and acid gelation properties of commercial UHT-treated plant-based Silva, J. G. S., Orlando, E. A., Rebellato, A. P., & Pallone, J. A. L. (2017). Optimization and
milk substitutes and lactose free bovine milk. Food Chemistry, 168, 630–638. https:// validation of a simple method for mineral potential evaluation in citrus residue. Food
doi.org/10.1016/J.FOODCHEM.2014.07.036. Analytical Methods, 10(6), 1899–1908. https://doi.org/10.1007/s12161-016-0748-3.
Mendes, M. K. D. A., Oliveira, C. B. D. S., Veras, M. D. A., Araújo, B. Q., Dantas, C., Singhal, S., Baker, R. D., & Baker, S. S. (2017). A comparison of the nutritional value of
Chaves, M. H., & Vieira, E. C. (2019). Application of multivariate optimization for the cow’s milk and nondairy beverages. Journal of Pediatric Gastroenterology and Nutrition,
selective extraction of phenolic compounds in cashew nuts (Anacardium occidentale 64(5), 799–805. https://doi.org/10.1097/MPG.0000000000001380.
L.). Talanta, 205, 120100. https://doi.org/10.1016/J.TALANTA.2019.06.100. Singleton, V. L., & Rossi, J. A. (1965). Colorimetry of total phenolics with phosphomo-
Miller, D. D., Schricker, B. R., Rasmussen, R. R., & Van Campen, D. (1981). An in vitro lybdic-phosphotungstic acid reagents. American Journal of Enology and Viticulture,
availability method for estimation iron availability from meals. The American Journal 16(3), 144–158.
of Clinical Nutrition, 43, 2248–2256. Sousa, R. M. F., Oliveira, G. K. F., Oliveira, A. D., Moraes, S. A. L. D. M., & Abarza, R. A.
MINTEL (2016). Food & Drink Trends 2017. From http://www.mintel.com/global-food- (2015). 2,2-diphenyl-1- picrylhydrazyl methods. In C. Ruiz-Capillas, & L. M. L. Nollet
and-drink-trends (retrieved November 30, 2018). (Eds.). Flow injection analysis of food additives (pp. 547–577). (1st ed.). .
Morris, E. R., & Ellis, R. (1985). Content of zinc, iron, calcium and their absorption in- Strain, J. J. S., & Cashman, K. D. (2009). Minerals and trace elements. In M. J. Gibney, S.
hibitors in food commonly consumed in Ethiopia. In C. Kies (Ed.). ACS symposium A. Lanham-New, A. Cassidy, & H. H. Vorster (Eds.). Introduction to human nutrition
series 275: Nutritional bioavailability of calcium (pp. 63). Washington, DC, USA: (pp. 188–237). (2nd ed.). .
American Chemical Society. Sudjaroen, Y., Thongkao, K., & Suwannahong, K. (2018). Antioxidant, antibacterial, and
Phenol-Explorer (2019). Database on polyphenol content in foods. From http://phenol- cytotoxicity activities of cashew (Anacardium occidentale) nut shell waste.
explorer.eu/foods/classification (retrieved December 5, 2019). International Journal of Green Pharmacy, 2(1), 229–234.
Quintaes, K. D., & Diez-Garcia, R. W. (2015). The importance of minerals in the human Tecnavio (2015). Nutrient density and nutritional value of milk and plant-based milk
diet. In M. De la Guardia, & S. Garrigues (Eds.). Handbook of mineral elements in food alternatives. From http://www.technavio.com/report/global-non-alcoholic-
(pp. 1–21). . beverages-dairy-alternative-drinks-market (retrieved November 30, 2016).
Rodríguez-Roque, M. J., Rojas-Graü, M. A., Elez-Martínez, P., & Martín-Belloso, O. Theodoropoulos, V. C. T., Turatti, M. A., Greiner, R., Macedo, G. A., & Pallone, J. A. L.
(2013). Soymilk phenolic compounds, isoflavones and antioxidant activity as affected (2018). Effect of enzymatic treatment on phytate content and mineral bioacessability
by in vitro gastrointestinal digestion. Food Chemistry, 136(1), 206–212. https://doi. in soy drink. Food Research International, 108, 68–73. https://doi.org/10.1016/J.
org/10.1016/J.FOODCHEM.2012.07.115. FOODRES.2018.03.018.
Sahuquillo, A., Barberá, R., & Farré, R. (2003). Bioaccessibility of calcium, iron and zinc Tyug, T. S., Prasad, K. N., & Ismail, A. (2010). Antioxidant capacity, phenolics and iso-
from three legume samples. Nahrung/Food, 47(6), 438–441. flavones in soybean by-products. Food Chemistry, 123(3), 583–589. https://doi.org/
Sánchez-Moya, T., Martínez-Hernández, V. M., López-Nicolás, R., González-Bermúdez, C. 10.1016/J.FOODCHEM.2010.04.074.
A., Ros-Berruezo, G., & Frontela-Saseta, C. (2019). Effect of soaking and inositol Udomkun, P., Tirawattanawanich, C., Ilukor, J., Sridonpai, P., Njukwe, E., Nimbona, P., &
phosphate content on in vitro mineral availability in edible seaweeds. Journal of Vanlauwe, B. (2019). Promoting the use of locally produced crops in making cereal-
Applied Phycology, 31(3), 1981–1989. https://doi.org/10.1007/s10811-018-1677-x. legume-based composite flours: An assessment of nutrient, antinutrient, mineral
Sánchez-Moreno, C., Larrauri, J. A., & Saura-Calixto, F. (1998). A procedure to measure molar ratios, and aflatoxin content. Food Chemistry, 286, 651–658. https://doi.org/
the antiradical efficiency of polyphenols. Journal of the Science of Food and Agriculture, 10.1016/J.FOODCHEM.2019.02.055.
76, 270–276. Umeta, M., West, C. E., & Fufa, H. (2005). Content of zinc, iron, calcium and their ab-
Sandberg, A. S., & Ahderinne, R. (1986). HPLC method for determination of inositol sorption inhibitors in foods commonly consumed in Ethiopia. Journal of Food
triphosphates, tetraphosphates, pentaphosphates and hexaphosphates in foods and Composition and Analysis, 18(8), 803–817. https://doi.org/10.1016/J.JFCA.2004.09.
intestinal contents. Journal of Food Science, 51(3), 547–550. https://doi.org/10. 008.
1111/j.1365-2621.1986.tb13875.x. Varga, M., Jójárt, R., Fónad, P., Mihály, R., & Palágyi, A. (2018). Phenolic composition
Sandberg, A. S., Brune, M., Carlsson, N. G., Hallberg, L., Skoglund, E., & Rossander- and antioxidant activity of colored oats. Food Chemistry, 268, 153–161.
Hulthe, L. (1999). Inositol phosphates with different number of phosphate groups Vitali, D., Vedrina Dragojević, I., & Šebečić, B. (2008). Bioaccessibility of Ca, Mg, Mn and
influence iron absorption in humans. American Journal of Clinical Nutrition, 70, Cu from whole grain tea-biscuits: Impact of proteins, phytic acid and polyphenols.
240–246. Food Chemistry, 110(1), 62–68. https://doi.org/10.1016/J.FOODCHEM.2008.01.056.
dos Santos, M. G. (2015). Avaliação de estabilidade do extrato hidrossolúvel de castanha-do- Zhang, H., Önning, G., Öste, R., Gramatkovski, E., & Hulthén, L. (2007). Improved iron
brasil (Bertholletia excelsa). Goiânia: Dissertação de mestrado. Universidade Federal bioavailability in an oat-based beverage: The combined effect of citric acid addition,
de Goiás. dephytinization and iron supplementation. European Journal of Nutrition, 46(2),
Schaich, K. M., Tian, X., & Xie, J. (2015). Hurdles and pitfalls in measuring antioxidant 95–102. https://doi.org/10.1007/s00394-006-0637-4.
efficacy: A critical evaluation of ABTS, DPPH, and ORAC assays. Journal of Functional Zhao, Y., Martin, B. R., & Weaver, C. M. (2005). Calcium bioavailability of calcium
Foods, 14, 111–125. https://doi.org/10.1016/J.JFF.2015.01.043. carbonate fortified soymilk is equivalent to cow’s milk in young women. The
Sethi, S., Tyagi, S. K., & Anurag, R. K. (2016). Plant-based milk alternatives an emerging Journal of Nutrition, 135(10), 2379–2382. https://doi.org/10.1093/jn/135.10.2379.
segment of functional beverages: A review. Journal of Food Science and Technology, Zięba, K., Makarewicz-Wujec, M., & Kozłowska-Wojciechowska, M. (2019).
53(9), 3408–3423. https://doi.org/10.1007/s13197-016-2328-3. Cardioprotective mechanisms of cocoa. The Journal of the American College of
Setyaningsih, W., Saputro, I. E., Carrera, C. A., Palma, M., & García-Barroso, C. (2019). Nutrition, 38(6), 564–575.

10