Modelling the effects of psyllium and water on dough parameters using Mixolab® and their relationship with physical properties and acceptability of gluten-free bread

This study aimed to investigate the effects of psyllium (P) and water (W) on dough Mixolab® parameters, and their relationship with gluten-free bread (GFB) physical properties and acceptability. A 22 factorial design with three center points was used, in which P levels ranged from 2.86 to 17.14% and W levels from 82.14 to 117.86% on a flour basis. Samples were 1 Trabalho apresentado no CBCP 2020 Congresso on-line Brasileiro de Tecnologia de Cereais e Panificação, selecionado para publicação na forma de artigo completo. Research, Society and Development, v. 9, n. 11, e77591110589, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i11.10589 2 compared to a control GFB (0P:100W), and data were evaluated using regression models and multiple factor analysis (MFA). The predicted model equations were significant (Radj= 8299%, p<0.05) and showed that P increased dough consistency (C1), protein weakening (C2), gelatinization (C3), stability (C4) and retrogradation (C5) of starch, whereas W or its interaction with P decreased these parameters. MFA’s three dimensions explain 94.86% of the total variation. Factor 1 (57.02%) positively discriminates the loaf-specific volume and all acceptability attributes, but negatively discriminates crumb firmness and C1, C2, C3, C4, and C3-C4 Mixolab parameters, especially in the 2.86P:82.14W sample. Factor 2 (26.30%) positively discriminates the C5, C1-C2, and C5-C4 Mixolab parameters and central points of the study, but negatively discriminates the control GFB. Factor 3 (11.54%) positively discriminates crumb moisture and 2.86P:117.86W and 17.14P:117.86W samples, unlike 2.86P:82.14W, which is negatively discriminated. We found results regarding dough Mixolab parameters to explain P and W influence and its capability of predicting GFB physical properties and acceptability.


Introduction
The increasing demand for high-quality gluten-free bread (GFB) poses a major challenge for food scientist, chefs, bakers, and the food industry due to the growing number of individuals with or without gluten intolerance following a gluten-free diet worldwide .
Given that GFB quality (especially its texture) is compromised by the lack of viscoelastic network in its dough, it is necessary to design matrices to meet breadmaking requirements. In this sense, and in view of the growing demand for functional foods beneficial to health, the psyllium (P)a soluble fiber obtained from the seed husks of Plantago ovata Forsk -has aroused food scientist's interest due to its role in helping intestinal transit, cholesterol, blood glucose, and satiety control (Belorio & Gómez, 2020;Franco et al., 2020).
When hydrated, P has important technological properties in food application, particularly its solubility and high-water binding and retention capacity, resulting in thickening and gel formation (Pejcz et al., 2018;Yu et al., 2017).
Recent studies showed that incorporating P into the dough may increase its viscosity and gas holding capacity, and improve GFB volume, structure, texture, appearance, acceptability, and shelf-life (Cappa et al., 2013;Fratelli et al., 2018;Mancebo et al., 2015;Santos et al., 2020;Ziemichód et al., 2019), besides increasing its fiber content and reducing its glycemic response (Fratelli et al., 2018). However, the impact of P on gluten-free dough and bread properties depends on its added levels, water content, and other ingredients present in the formulation (Cappa et al. 2013;Mancebo et al. 2015;Fratelli et al. 2018). Studies addressing how gluten-free dough affects GFB properties, especially with the incorporation of P, are still scarce.
Mixolab® is a device that has been successfully used to assess wheat dough systems under similar breadmaking conditions (Rosell et al., 2007), providing a thoroughly rheological analysis; it is also applicable to gluten-free systems (Matos & Rosell, 2013).
However, the relationship between gluten-free dough rheology and GFB properties using P is still little explored, requiring further research.
Our previous work reported P as a promising ingredient to improve GFB physical, sensory, and nutritional properties concomitantly (Fratelli et al, 2018). Considering that, we further investigated the effects of P and water on dough Mixolab® parameters and their relationship with GFB physical properties and acceptability. Research, Society and Development, v. 9, n. 11, e77591110589, 2020 (CC BY 4.

Methodology
This research is a lab study of quantitative nature, of which was relationed the dough and bread properties.
Rice flour (Urbano Agroindustrial Ltda.) and cassava starch (General Mill Brasil Alimentos Ltda.), purchased from local Brazilian stores, psyllium ( P) (VITACEL® Psyllium P95), donated by JRS Latinoamericana Ltda, and water were the materials employed in this study.
A full 2² factorial design was adopted in four trials, in which P levels ranged from 2.86 to 17.14% and W levels from 82.14 to 117.86%, and the three-center points repetition contained 10% P and 100% W on a flour weight basis (75% rice flour and 25% cassava starch). These trials were compared to a control (0P:100W), totalling eight dough samples.
The dough Mixolab® thermomechanical properties were related to the physical properties (loaf-specific volume, crumb moisture, and firmness) and acceptability (appearance, color, aroma, texture, flavor, and overall liking evaluated by 53 consumers) of the GFB developed and studied by Fratelli et al. (2018).
Data were compared using one-way ANOVA at p <0.05 and Tukey's test, and evaluated using regression models considering adjusted R² ≥ 70% and p <0.05. These analyses were performed using Statistica 13.5 software (Tibco Inc., USA, 2018), and the relationships between doughs and bread properties were verified considering the multiple factor analysis (MFA) using XLSTAT 2020.2 software (Addinsoft, USA, 2020). Figure 1 shows the dough thermomechanical curves, and in the sequence, Table 1 presents these parameters values.  Values are mean ± standard deviation. Different letters in the same column are different significantly (Tukey test p <0.05). * water-level adjusted to the initial design to enable the analysis. Source: Authors.

Results and Discussion
The combination of high P and low W levels found in trial 2 impair the analysis given the high force exerted on the device. The W level was thus increased to the minimum required value to enable the analysis. The highest peaks in the main Mixolab parameters were observed in trial 2, whereas the control trial (0P:100W) showed the lowest peaks ( Figure 1 and Table 1).
C1 dough consistency increased the most with the growing P addition, whereas W affected this parameter the least. In this study, the increasing P contribution reduced C2 points ( Figure 1) and is in agreement with the observations by Pejcz et al. (2018) on the wheat dough with added P (4 and 8%). In the same way, the C1-C2 rate increased due to P addition but only differs between the samples with low P concentration (trial 3-2.86P:117.86W) and control (0P:100W) (Table 1). However, Santos et al. (2020) observed torque reduction in C2 at the highest P levels (12.5%) combined with the highest levels of chickpea flour (100%).
According to these authors, high P levels are highly capable of forming complexes with system proteins through both ionic and nonionic interactions, thus affecting dough strength .
The dough properties with a potential tendency of relations with bread characteristics were characterized by points C3 to C5 (Figure1 , Table 1). Different P and W levels resulted in different doughs consistency in these stages. Compared to the control, incorporating P into the dough significantly increased consistency at C3 (from 14 to 50 times), which is more evident in trial 3 for presenting the highest concentration. However, we observed no difference between control and trial 1, with the lowest P and W levels, regarding C4 and C5 ( Figure 1, Table 1). According to Pejcz et al. (2018), P is capable of absorbing forty times its weight in water, which could strongly affect dough functional and technological properties, including rheology, as it would limit the W content available to starch hydration. Incorporating P into the dough increased starch gelatinization susceptibility (C3-C2), amylase activity (C4-C3), and retrogradation tendency (C5-C4). These parameters may influence bread quality, especially staling kinetics during storage (Rosell et al., 2007). In general, the observed changes in parameters may be explained by the interaction between starch and different incorporated P levels competing for the available W content on doughs systems.
Based on these results and statistical calculation, we adjusted the trial 2 to real levels of factorial design. Table 2 shows prediction equations for the experimental model. Research, Society andDevelopment, v. 9, n. 11, e77591110589, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i11.10589 Table 2. Predicted model equations indicating psyllium and water effects a and their interaction on the Mixolab® properties of gluten-free dough.
Based on Table 2 results, we may verify that P increased all primary parameters (from C1 to C5) evaluated by Mixolab, while W or its interaction with P decreased them. In turn, P, W, or their interaction, caused the opposite effect on secondary parameters. Most parameters adjusted to models (R 2 adj= 82-99%, p=0.000), but not C3-C4 and C5-C4 (R 2 adj < 70%). Our findings corroborate those reported by Aprodu & Banu (2015), who verified that dough parameters varied with different hydration levels due to starch and fiber competition for water. Incorporating P into a dough of low water level increases C4 and C5; in turn, incorporating it into a dough of high W level decreases C3 and C5. Figure 2 shows the relationships between dough and bread properties, whereby MFA three dimensions explain 94.86% of the total variation.
Factor 3 (11.54%) positively discriminates crumb moisture due to higher W levels, explaining properties. This finding indicates a favorable dough consistency for GFB with greater loafspecific volume, as well as greater sensory acceptability of all evaluated attributes, observed by the proximity between these axes and lower crumb firmness (Figures 2a, 2b).
Our finding corroborates those reported by our previous studies (Fratelli et al., 2018), as the presence of P may positively influence crumb softness if the dough is properly hydrated. Conversely, deficient hydration decreases loaf volume, crumb softness, and sensory acceptability.

Conclusion
We found balanced P and W levels to strongly affect dough parameters, resulting in GFB with greater loaf-specific volume, crumb softness, and sensory acceptability for the evaluated attributes.
Mixolab dough parameters explain P and W influence and indicate its potential for predicting GFB physical properties and acceptability, and that it could be helpful in guiding further studies in both gluten-containing and gluten-free breadmaking.