Functional beverage design based on fresh milk , tarwi ( Lupinus mutabilis ) beverage and oatmeal ( Avena sativa )

The tarwi is an Andean legume with a high nutritional value from which a vegetable beverage can be obtained. Cereals, like oat, have good characteristics as a prebiotic for the production of functional drinks, whose consumption is currently increasing. The objective of the research was to design a probiotic fermented beverage based on fresh milk, tarwi beverage, and oatmeal. An optimal treatment consisted of 1.9% oatmeal, 39.9% tarwi beverage, 46.2% fresh milk, 10.0% honey, and 2.0% probiotic culture; determined by applying a rotatable central composite design of surface response methodology. It had a probiotic count of 3.47x108 cfu/mL, a protein content of 3.75%, and overall acceptability of 7 points, which corresponds to “I like it very much”. The result was experimentally validated. Likewise, the shelf life of the optimal beverage was 20 days at 5 °C with appropriate functional, nutritional, and sensory characteristics.


INTRODUCTION
The synergistic or symbiotic combination of prebiotics and probiotics that found in products such as food, medicine, and supplements are defined as a combination that beneficially affects the host by improving the survival and implementation of live microbial dietary supplements in the gastrointestinal tract. Symbiotics are more effective than probiotics or prebiotics used in isolation 16 .
Oatmeal, mainly as flakes, is included in the human diet for its health benefits 18 . It is rich in β-glucans, proteins, starch, and phenolic compounds, and it is known to have anti-cancer and hypocholesterolemic properties. The availability of soluble fibers, both oligosaccharides, and polysaccharides, have a prebiotic effect. Despite this, there is little information available on the use of oats as a possible carrier of probiotics 19 .
The tarwi (Lupinus mutabilis) is a non-widely used legume found in several regions of Peru 6 that should be part of our diet because of its high nutrient content. Its protein content is even higher than that of soybeans, and its fat content is similar 20 . After dehydration, it is used in different culinary dishes (mote, salads, soups, stews, desserts, and ceviche) or to obtain flour used in baking. Beverage (tarwi juice) is also produced from tarwi and consumed at breakfast 10 . Vegetable beverage is an aqueous extract of legumes, cereals, or seeds. A trend of these new products is the application of the fermentation process 21 .
The objective of this research was to design and optimize a fermented functional beverage based on fresh milk, tarwi beverage, and oatmeal with appropriate probiotic viability, protein content, and overall acceptability during refrigerated storage.

Tarwi beverage extraction
Debittered grains were combined with water in a 40:60 ratio and homogenized in an industrial blender (Metalmecánica Agroindustrias) of 16 L capacity, at 3600 rpm for 5 min. Grains were then filtered using a stainless-steel mesh (0.173 mm opening) and sterile gauze cloth, separating the retained solids from supernatant. This operation was carried out twice to obtain a homogeneous solution. Finally, the solution was pasteurized at 85 °C for 15 min, cooled, and stored in refrigeration at 4 °C 21 (with author's modifications).

Preparation of the functional beverage
The fresh milk and tarwi beverage were mixed in a stainless-steel kettle with the oatmeal in the amounts indicated in the experimental design (Table  1). Afterward heated up to 70 °C, to add the honey and obtain its total solution, pasteurized at 85 °C for 15 min, and cooled to 37 °C to inoculate 2% (w/v) of the probiotic culture. Then, it was incubated at 37 °C for 5 h and finally stored in refrigeration at 4 °C for 24 h for analytical evaluations. Table 1 shows the formulations of the functional beverage with fresh milk, tarwi beverage, and oatmeal.

Physicochemical characteristics
For soluble solids 22 , the method of the Association of Official Analytical Chemists-AOAC was used; for titratable acidity 23 and proteins content 24 , the AOAC method, and dietary fiber 25 according to AOAC, using the Sigma TDF-100 Enzymatic Kit.

Probiotic bacteria count
De-Man Rogosa Sharpe (MRS) media used (Merck, USA) was enriched with 10% of a 0.050% L-Cysteine solution to improve anaerobic conditions in the agar. The inoculum and the functional beverage were homogenized properly and 1 mL was taken and serially diluted until reaching 10 -7 . After that, the pour plate method was used for the probiotic bacteria count in duplicate and incubated in anaerobiosis at 37 °C for 48 h. Results were expressed in cfu/mL 11 .
Overall acceptability A 9-point hedonic scale was used, anchored between 1 that is equivalent to "I dislike it very much" and 9 to "I like it very much", with intermediate point of 5 "I neither like nor dislike it". The sensory tests were performed in a laboratory room with bright natural daylight. All samples were marked with codes and were randomly provided to evaluators on white plates. The sample consisted of 20 mL of room temperature beverage in randomly numbered containers with three digits. The test was performed in two sessions with 30 untrained panelists, recruited among students and staff of Antenor Orrego University, mainly aged 20-30. In the first session, five samples were delivered, and in the second, 4. All participants involved in the evaluation were thoroughly informed about the manufacturing process and the sensory attributes of the product 21,26 .

Statistical design
The design was based on a base formulation of acidified milk beverage consisting of 88% fresh milk, 10% honey, and 2% probiotic culture of the final volume. From this, a rotatable central composite design (DCCR) of the response surface methodology (MSR) was used to evaluate the effect of replacing fresh milk with tarwi beverage and oatmeal, and their combinations on soluble solids, titratable acidity, dietary fiber, protein content, probiotic bacteria count and overall acceptability. This resulted in a total of 9 treatments and 4 repetitions of the central point, as seen in table 1. Analysis of variance was applied with a 95% confidence level; to validate the models obtained, R 2 > 0.800 was considered, p<0.050, and the lack of adjustment greater than 0.05. The optimization was developed using the superposition of contours of the most relevant variables in the functional beverage, obtaining the maximum compound desirability as a statistical indicator that validates the product design. Minitab Statistical Software, version 17 (Minitab Inc., 2014) was used. Optimal beverage shelf life Shelf life was determined using the methodology of evaluation in real time of the quality attribute lactic acidity until reaching 0.55%, correlated with the overall acceptability (data not shown). Table 2 shows the results of the physicochemical charac te ris t ic s; pr o bio t ic bac te ria co unt and overall acceptability in the different treatments of the functional beverage. Soluble solids fluctuated between 15.4-17.0%. The highest values were in those formulations due to high content of oatmeal and fresh milk. The acidity varied between 0.198-0.250%, showing higher results in drinks with a higher content of oatmeal and tarwi beverage. Protein content results were between 3.62-3.86%, showing higher values in formulations with higher oatmeal and intermediate tarwi beverage. Dietary fiber showed results between 0.14-0.36%, denoting the highest values in the formulations with the highest amount of oatmeal and tarwi beverage. The probiotic count fluctuated between 3.77x10 7 -3.20x10 9 cfu/mL, with the formulations with highest counts being the ones with the highest amount of oatmeal and intermediate tarwi beverage. Acceptance by the panelists varied between 5.6-8.5 points, with the highest acceptance in beverages with lower content of oatmeal and higher amount of tarwi beverage. Table 3 shows the regression models and fit of the rotatable central composite design for the variables evaluated in the functional beverage based on fresh milk, tarwi beverage and oatmeal. Table 4 shows the optimization for probiotic counts, protein content and overall acceptability. Figure 1 shows contour surface of the rotatable central composite design for probiotic counting (A), protein content (B) and overall acceptability (C). Figure 2 shows the optimization of probiotic counts, protein content and overall acceptability.   Shelf life optimal beverage 20 days at 5 °C

Physicochemical characteristics
The treatments with the highest content of soluble solids were F3 constituted by (oatmeal 3%, tarwi beverage 30% and fresh milk 55%) and F4 (oatmeal 2.25%, tarwi beverage 27.9% and fresh milk 57.8), as seen in table 2. This is because these ingredients have more dissolved solids in their composition. Salamanca et al. 27 reported values of 17-24% in a fermented beverage based on borojó juice, yogurt, and honey; Randazzo et al. 28 with 5.90 -15.7% in a fermented non-dairy beverage based on Mediterranean fruit juices with kefir microorganisms in water.
The treatments with the highest titratable acidity were F2 (oatmeal 3%, tarwi 40% and fresh milk 45%) and F7 (oatmeal 2.25%, tarwi 42.1% and fresh milk 43.7); as seen in table 2. Lactic cultures have a higher rate of acidification in legume kinds of milk, such as soy, in comparison to dairy substrates because they have lower buffer capacity, which allows S. thermophilus, as a homofermentative species, to produce more lactic acid 29 . Flores et al. 10 report 0.07-0.09% in a fermented beverage based on tarwi juice and kiwicha; Battistini et al. 29 with 0.178-0.213% in a fermented beverage based on soy milk with the addition of inulin and fructooligosaccharides; Sánchez et al. 30 with 0.45% in a fermented beverage based on milk, whey, sucrose, and curuba extract.
The highest protein content was found in treatments F5 (3.31% oatmeal, 35% tarwi beverage and 49.7% fresh milk) and F2 (3% oatmeal, 40% tarwi beverage and 45% fresh milk); as seen in table 2. Products derived from oatmeal such as flakes and flour denote high quantity and quality of protein that can be used in functional beverages 18 . Castañeda et al. 20 mentioned that vegetable beverage such as tarwi and soy have slightly higher amounts of protein than cow's milk, and, when combined in different proportions for the manufacture of yogurt, the final product obtains values between 3.86 and 3.93%. Kumar et al. 31  The highest content of dietary fiber was found in beverages made up of F5 (3.31% oatmeal, 35% tarwi and 49.7% fresh milk) and F2 (3% oatmeal, 40% tarwi and 45% fresh milk); as seen in table 2. Nionelli et al. 18 mentioned that oatmeal shows impressive characteristics for food processing, in particular, due to its high fiber content and bioactive compounds beneficial to health. Buriti et al. 32 indicated values between 0.620-2.25% in a fermented beverage based on goat's milk, fruit pulp, and galactomannans of Caesalpinia pulcherrima seeds; Din et al. 33 with 0.200-0.400% in a functional beverage with Barley β-glucan. Hernández et al. 34 , with 0.600% in a fermented beverage based on rice flour and sesame seeds.

Probiotic bacteria count
The highest count of probiotic bacteria, which is beneficial for consumers, was obtained in beverages F4 (oatmeal 2.25%, tarwi 27.9% and fresh milk 57.8%) and F3 (oatmeal 3%, tarwi 30% and fresh milk 55%); as seen in table 2. The viability of probiotics in fermented beverages depends on factors such as their nature, the substrate, processing, and storage conditions 10 . Gupta et al. 19 explained that the presence of prebiotic compounds and non-digestible carbohydrates present in cereals such as barley and oat, as well as, malt substrates stimulate the growth of probiotics in fermented beverages and other beneficial bacteria of the intestinal tract. The hydrolyzate kiwicha and tarwi beverage have a synergistic effect with the Lactobacillus and Bifidobacterium cultures on the counts of these probiotics in the preparation of a fermented beverage 10 . Battistini et al. 29 reported 7.76x10 7 -1.18x10 8 cfu/ mL in a fermented beverage based on drinkable soy with the addition of inulin and fructooligosaccharides; Randazzo et al. 28 with 3.9x10 7 -1x10 8 cfu/mL in a fermented non-dairy beverage based on Mediterranean fruit juices with kefir microorganisms in water; Dos Santos et al. 35 found values 3.10x 7 -2.20x10 8 cfu/mL of Lactobacillus lactis in drinkable soy fermentation with kefir and addition of inulin and values of 6.64x10 7 ; 3.25x10 7 and 3.07x10 7 cfu/mL were reported by Łopusiewicz et al. 36 in the development of non-dairy kefir-like fermented beverage based on flaxseed oil cake.

Overall acceptability
The major perception of sensory characteristics expressed as general acceptability was found in F9 (oats 1.50%, tarwi 30% and fresh milk 56.5%) and F7 (oats 2.25%, tarwi 35% and fresh milk 50.8%); as seen in table 2. Sensory results are essential and related to product quality, substrate composition and the concentration of its components 10 . Nionelli et al. 18 mentioned that the acceptability of this type of beverage depends on the fermentation since it generates changes in the profile of its sensory characteristics such as acidity, taste, smell, and viscosity. The best formulation of a fermented beverage based on soybeans and chickpea reported by Wang et al. 21

Experimental design and response surface models
The results analyzed with the rotatable central composite design (Table 3), determined that the quadratic model denoted the most considerable fitted of the data for soluble solids, acidity, probiotic count, protein and overall acceptability, and the linear model for dietary fiber. Likewise, the significance of the models was evaluated through analysis of variance that presented a significant effect (p<0.05). Besides regression coefficients, the equations for each response were determined. Figure 1A shows the contours generated for the DCCR, which represents the analysis area at the limits of the independent variables, denoting that the region of interest for the probiotic count was between 1.5-3% and 30-40% of oatmeal and tarwi beverage, respectively. In figure 1B, analysis indicated that between 2.5-3% and 32.5-40% of oatmeal and tarwi beverage, respectively, was the region of interest for protein content. In figure 1C for overall acceptability, the area of interest was between 1.5-2.5% and 32.5-40% of oatmeal and tarwi beverage, respectively. Figure 2 shows the optimization of the independent variables on the probiotic count, protein content, and overall acceptability responses, considered as more important for the health benefit and nutritional contribution they provide to consumers. By superposition the regions of interest of the contour surfaces, it was determined that the optimal area corresponds to 1.85% oatmeal and 39.9% tarwi beverage, achieving a probiotic count of 3.47x10 8 cfu/mL, 3.75% of protein content and 7 points of overall acceptability ("I like it very much"). Table 4 presents the validation of the optimization for the probiotic count, protein content, and overall acceptability. The shelf life of the optimal formulation stored at 5 °C was tested presenting, on the 20th day of storage, a probiotic count of 3.89x10 7 cfu/mL, a protein content of 3.54%, and overall acceptability of 7.60 points, which corresponds to "I like it very much". In addition, soluble solids of 13.8%, acidity 0.550%, and dietary fiber 0.210%. This beverage, according to Peruvian labeling regulations, would present the octagon high in sugar (using a natural sweetener, honey), but nutritionally provides the benefits of high protein content, dietary fiber, probiotic bacteria and does not contain artificial preservatives foods, that is recommended by the world health institutions, as well as, a good overall acceptability that demonstrates the intention to purchase the product for possible commercialization as a functional beverage.

CONCLUSION
The design of a functional fermented beverage containing 1.85% oatmeal, 39.9% tarwi beverage, 46.2% fresh milk, 10% honey, and 2% probiotic culture was optimized using a rotatable central composite design and experimentally validated. Likewise, the shelf life of the functional beverage was determined to be 20 days of storage at 5 °C with appropriate functional, nutritional, and sensory characteristics.
Founding Source. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.