Effect of extract or infusion of leaves of the Hibiscus sabdariffa L. in the production and storage of the beverage blends with cupuassu: physico-chemical and sensory acceptance

Abstract

This study aimed to produce beverages using extract or infusion of Hibiscus sabdariffa leaves with cupuassu pulp. The blended drinks with cold or hot extraction were formulated using the response surface methodology. Moreover, the physico-chemical and sensory stability of these beverages was performed for 180 days. The conditions production for beverage leaf extract were: 28% leaf extract, 72% cupuassu pulp, and 14°Brix. For beverage leaf infusion, were 37% leaf extract, 63% cupuassu pulp, and 13°Brix. Comparing the two beverages, the polyphenolic compounds and the antioxidant activity (ABTS and DPPH) were higher in the blends produced with leaf extract. Thus, the cold extract provided better extraction of these components in the leaves. During storage, pH, reducing and total sugars and soluble solids of blends increased linearly (p < 0.05). Nevertheless, vitamin C, polyphenolic compounds, and the antioxidant activity reduced linearly (p < 0.05). The sensory acceptance of blends containing leaf extract, in the hedonic scale, all sensory attributes were positively scored up to 135 days. At 180 days, the acceptance reduced for rejection. The blended beverage containing leaf infusion had all sensory attributes positively scored up to 90 days, with a rating between 6.27 and 7.42. At 135 and 180 days, the values were in the indifference region of the hedonic scale or acceptance region. Therefore, the blend with leaf infusion maintained better acceptance during storage when compared to leaf extract beverage. Thus, the blended beverages had good nutritional value and sensory acceptance and were acceptable for up to 135 days.

Introduction

The demand for functional drinks with health benefits has increased the beverage production containing Hibiscus sabdariffa. This plant belongs to the Malvaceae family and is cultivated in Asia, Africa, and Central and South America. It is commonly consumed as a vegetable, condiment, and tonic drink’s production (Pimentel-moral et al. 2018).

The calyx is the plant part more used. However, the leaves are also consumed as a leafy green vegetable in some countries. Compared with the calyx, the leaf of H. sabdariffa, the most abundant portion of foliage, is underutilized. The leaf is usually discarded, except in sub-Sahara Africa, where some consume it as a vegetable in soups and sauces and Brazil. In this country, the leaves are used in the production of typical food called cuxá rice that is consumed at the Maranhão state (Rezende et al. 2019; Zhen et al. 2016).

H. sabdariffa L. leaves have large amounts of organic acids (citric, ascorbic, oxalic, tartaric acids), flavonoids, anthocyanins, and other antioxidants. Researches show that the aqueous extract of leaves can increase the inhibitory activity of lipid peroxidation, decrease the levels of LDL, VLDL cholesterol, and increase HDL (Da-Costa-Rocha et al. 2014).

Fresh H. sabdariffa L. leaves are perishable, and its quality deteriorates due to microbial and physiological activities during the storage and transportation and hence requires immediate processing and preservation (Singh et al. 2014). Thus, beverage blends production can be very attractive in processing and storing. These drinks are easily consumable and use methods to preserve and maintain quality until consumed.

For the leaves use, can be produced extract or infusion. The aqueous extraction method can influence the nutritional composition and, consequently, its sensory characteristics. Brewing an infusion from the plant parts are still the most widely applied extraction procedures for preparing herbal teas. However, generally, the thermal treatment causes depletion of antioxidants. In contrast, plant sugars and amines can be involved in the Maillard reaction, yielding products with potential antioxidant properties (Bouabid et al. 2020; Zannou et al. 2020).

Most people dislike beverages made from H. sabdariffa extract, as it has an acidic and bitter taste. Thus, Mgaya-Kilima et al. (2014a) produced blends roselle extract with tropical juices to improve the aroma, flavor, nutritional, and antioxidant properties of the H. sabdariffa-fruit blends.

Therefore, the cupuassu (Theobroma grandiflorum), which is a tropical fruit of the Brazilian Amazon, can be an alternative to improve the sensory characteristics. The distinctive flavor of the cupuassu pulp provides its use as an ingredient in the manufacture of ice cream, juice, wines, jellies, and other products. Moreover, it has a yellowish-white pulp of high nutritional value and is a source of ascorbic acid (96–111 mg/100 g) and phenolic compounds (20.5 mg/100 g) (Pugliese et al. 2013).

During the storage of beverages, a wide range of biochemical changes occurs, which reduces the nutritional value. However, a proper understanding of these changes will help us to maintain the nutritional quality during processing and storage of juice (Singh and Sharma 2017).

Thus, no studies have been carried out to produce H. sabdariffa leaves with cupuassu, and on the effects of storage time. Furthermore, no investigation of the impact of cold or hot extraction on the beverage blends production. Therefore, this study aimed to produce beverage blends using extract or infusion of H. sabdariffa leaves with cupuassu pulp. Moreover, the physico-chemical and sensory stability of these beverages were performed.

Materials and methods

Raw material and extract and infusion production

Two types of aqueous extraction of H. sabdariffa L. leaves, which vary concerning the temperature and the time of extraction, were prepared: leaf extract (cold extraction) and infusion (hot extraction). For the leaf extract was production, the leaves were homogenized with water in the concentration of 15% (w/v) and maintained in cold temperature (7 °C) for 24 h. For the leaf infusion, it was used the same leaves concentration. However, the dissolution was performed in boiling water (100 °C) for 15 min. After, the leaf extract (1.0 L) and leaf infusion (1.5 L) were filtered through Whatman paper.

Extract and infusion of Hibiscus sabdariffa L. leaves, pulp cupuassu, and soluble solids content and beverage blends

The production conditions (vegetable ratio and solid soluble—SS) were evaluated by 2 central composites rotated experimental designs, changing the extract and infusion leaves (and consequently the cupuassu pulp) ratio and SS content. For the beverage containing leaf extract, the concentration changed from 13.79 to 56.21% (consequently the cupuassu pulp from 43.79 to 86.21%). In the drink containing leaf infusion, the concentration changed from 11.72 to 68.28% (consequently the cupuassu pulp from 31.72 to 88.28%). The SS in the two experimental designs changed from 10.17 to 15.83°Brix (Table 1). The experimental domain was chosen based on preliminary sensory tests for the concentration extract and infusion leaves and SS.

Table 1.

Experimental design and responses on overall liking in blend beverages containing extract or infusion of Hibiscus sabdariffa L. leaves and cupuassu with different ratio of vegetables in the mixed base

Assay	Leaf extract (%)	Cupuassu pulp (%)	Total soluble solids (°Brix)	Overall liking
1	50.00 (+1)	50.00	11.00 (−1)	6.85 ± 1.23
2	50.00 (+1)	50.00	15.00 (+1)	7.99 ± 1.14
3	20.00 (−1)	80.00	11.00 (−1)	6.90 ± 1.15
4	20.00 (−1)	80.00	15.00 (+1)	8.78 ± 1.88
5	56.21 (+α)	43.79	13.00 (0)	6.68 ± 1.72
6	13.79 (−α)	86.21	13.00 (0)	8.43 ± 1.23
7	35.00 (0)	65.00	10.17 (−α)	8.62 ± 1.58
8	35.00 (0)	65.00	15.83 (+α)	8.15 ± 1.27
9	35.00 (0)	65.00	13.00 (0)	8.67 ± 0.75
10	35.00 (0)	65.00	13.00 (0)	8.62 ± 1.27
11	35.00 (0)	65.00	13.00 (0)	8.65 ± 1.27

Assay	Leaf infusion (%)	Cupuassu pulp (%)	Total soluble solids (°Brix)	Overall liking
1	60.00 (+1)	40.00	11.00 (−1)	6.70 ± 0.95
2	60.00 (+1)	40.00	15.00 (+1)	7.30 ± 1.50
3	20.00 (−1)	80.00	11.00 (−1)	7.25 ± 1.27
4	20.00 (−1)	80.00	15.00 (+1)	7.80 ± 1.19
5	68.28 (+α)	31.72	13.00 (0)	6.50 ± 1.18
6	11.72 (−α)	88.28	13.00 (0)	7.00 ± 1.28
7	40.00 (0)	60.00	10.17 (−α)	6.30 ± 2.01
8	40.00 (0)	60.00	15.83 (+α)	7.00 ± 1.07
9	40.00 (0)	60.00	13.00 (0)	8.90 ± 1.07
10	40.00 (0)	60.00	13.00 (0)	8.70 ± 2.02
11	40.00 (0)	60.00	13.00 (0)	8.70 ± 1.43

The beverages were produced with 30% of the mixed base. The vegetable ratio was added according to the values of the experimental design. The SS were standardized by the addition of sucrose to reach the values of experimental design. After homogenization, the drinks were pasteurized using hot-fill processing at 90 °C for 60 s and cooled water. The overall acceptance was carried out. The results were performed by response surface methodology to obtain the optimum extract and infusion of H. sabdariffa L. leaves, cupuassu pulp, and SS content.

After finding the optimal conditions of production, two formulations of mixed beverages were produced, and measurements of pH, reducing and total sugars, vitamin C, polyphenolic compounds, and antioxidant activity were performed. The experiments were made in quintuplicate.

Storage of the blend beverages

For the storage, two formulations of mixed drinks were produced in quintuplicate and stored at 25 °C for 180 days. The storage temperature was chosen based on room temperature used for the storage of juices. In Brazil, the room temperature is around 25 °C. The storage time was chosen based on other researches with functional juices (Gonzalez-Larena et al. 2015; Tastan and Baysal 2015) or containing H. sabdariffa (Mgaya-Kilima et al. 2014b). In these formulations were added the food preservatives sodium metabisulfite (40 ppm) and potassium sorbate (500 ppm).

pH, reducing and total sugars, SS, vitamin C, polyphenolic compounds, antioxidant activity, and sensory acceptance were recorded before storage (day 0) and at intervals of 45 days during the 180 days.

Reducing and total sugars and soluble solids determinations

The content of reducing sugars was performed spectrophotometrically (Biospectro, SP-220, Curitiba, Brasil) using 3.5-dinitrosalicylic acid (DNS). For the total sugars, firstly, an acidic inversion was carried out, and then the total sugars were determined using the DNS (Miller 1959). The results were expressed in g glucose/100 mL. The SS analysis was performed directly using the digital refractometer (HI96801, Hanna, Woonsocket, USA).

Vitamin C determination

Vitamin C content was determined using 2.6-dichlorophenolindophenol reagent by titration (Topi 2020). Results were expressed as mg of ascorbic/100 mL.

Polyphenolic compounds and antioxidant activity analyses

Firstly, the phenolics were extracted from the drinks using a methanol/water solution of 50% (v/v) and an acetone/water solution of 70% (v/v). The extracts were used to determine the polyphenolic compounds and the antioxidant capacity. The polyphenolic compounds were quantified, according to Singleton and Rossi (1965). For the analysis, 0.5 mL of the sample (or water for blank) was mixed with 0.5 mL of the Folin–Ciocalteu solution (diluted in water 1:3) and 1 mL of 20% (w/v) sodium carbonate solution. The reading was done at 765 nm using a spectrophotometer (Biospectro, SP-220 model, Curitiba, Brasil). Gallic acid was the reference standard. Results were expressed as mg of gallic acid/100 mL.

The antioxidant activity was evaluated by ABTS and DPPH methods. For the ABTS method, aliquots of 30 μL of the extract were mixed to 3000 μL of ABTS⁺ radical. The absorbance was taken after 6 min. (Re et al. 1999). For the DPPH free radical method, aliquots of 30 μL of the extract were added to 1200 μL of DPPH solution (0.06 mM). The absorbance was measured at 515 nm after 30 min (Brand-Williams et al. 1995). Results were expressed as Trolox equivalent antioxidant activity/mL.

Sensory acceptance

One hundred ten untrained panelists performed sensory evaluation in each session. The sessions occurred in individual booths. Samples of 40 mL were served at 7 °C in glass cups. For the chosen vegetable ratio and SS contents using the central composite rotated experimental design, the samples were served in 6 sessions, 3 for each experimental design by the same panelists. The consumer acceptance measured for hedonic scale evaluated overall liking.

For the storage, the consumer acceptance measured for hedonic scale evaluated color, appearance, aroma, flavor, sweetness, acidity, body, and overall liking attributes. These attributes were chosen because they are the most important for juices based on literature studies (Bechoff et al. 2014; Ogundele et al. 2016; Monteiro et al. 2017). At the stage, two sessions were performed, one to each blend beverage. The hedonic scale used ranged from 1 (dislike extremely) to 9 (like extremely). The purchase intention was also evaluated by the use of a 5 point scale, ranging from 1 (I certainly would not buy) to (I would certainly buy) (Stone et al. 2004; Meilgaard et al. 1991).

The sensory analysis was performed after the results of the microbiological analyzes (American Public Health Association 2001). Thus, with the absence of total coliforms (< 3 MPN/mL) and mold and yeast counts (< 10 CFU/mL), the safety and suitability of the product for sensory tests were guaranteed.

Data analysis

Statistica software version 7.0 (StatSoft, USA) was used to build the experimental designs, the surface response overall liking graph, and to optimize the extract and infusion of H. sabdariffa L. leaves, cupuassu pulp, and SS content at 95% of confidence level.

One-way ANOVA analyzed the physico-chemical analysis with post hoc t test at the 95% confidence level. For the physico-chemical analysis data during storage, regression analysis was used.

For the hedonic scale data during storage, storage time was considered as a fixed source of variation and the panelist as a random effect. The non-parametric Friedman test analyzed the attributes at the 95% confidence level. For these data, the analyses were performed using XLSTAT software (Addinsoft, Paris, France). For the purchase intention, histograms were built and divided into three regions: percentages of 1 and 2 were named ‘‘I would buy it’’, the percentage at score 3 was called ‘‘Maybe I would buy it’’ and the percentages of 4 and 5 of ‘‘I would not buy it’’.

Results and discussion

Extract and infusion of H. sabdariffa L. leaves, pulp cupuassu, and soluble solids content

The overall liking of the mixed beverages was positively scored, with a rating between “like slightly” and “like extremely” (6–9) (Table 1). Therefore, it is possible to conclude that all assays constitute good market alternatives. This result is significant because 66.67% of consumers (Data not shown) said like of H. sabdariffa leaves. However, the consumption is by a typical salt food called cuxá rice. Therefore, the good acceptance of sweet food is a positive result.

Ogundele et al. (2016), investigated the effects of blends of H. sabdariffa extracts with vegetables (pineapple, orange, and carrot). These authors obtained an overall acceptability rating between 6 and 7 (like slightly and like moderately). Thus, the results reported herein indicate that cupuassu is a promising fruit for blends with H. sabdariffa leaves extract, because it provided higher acceptance scores.

The variable overall liking is often used to measure consumers’ overall hedonic responses to food. Therefore, in the present study, the results of this variable were used to obtain the optimums values of H. sabdariffa leaves extract and infusion, cupuassu pulp, and SS content.

In assay 5, with the highest leaf extract, the acceptance sensory was lower (6.68 ± 1.72) (Table 1). Bechoff et al. (2014), evaluating the infusions and syrups of H. sabdariffa calyces, reported a bitter taste, perceived negatively by the panelists. Moreover, these authors suggested that the addition of sugar would be masking the bitter tastes of H. sabdariffa. In the present study, the increased sugar improved acceptance. Comparing the assays 1 and 2, which have the same leaf extract concentration, it is observed an increase in the overall liking when the SS (sugars) increased.

For blend beverage produced with leaf extract, according to the estimated effects, the extract concentration, and SS affected the sensory acceptance, being significant at 95% confidence level. Data of Table 1 was fitted to the quadratic model for overall liking given in Eq. (1). The model was significant since the calculated F value (15.72) was higher than the listed F value (F5.5 = 5.05). Proper correlation coefficients were also obtained (R² = 0.94).

$$\text{Overall}\text{liking}=8.65+0.83\text{LE}-1.01{\text{LE}}^2+1.30\text{SS}-1.18{\text{SS}}^2+0.37\text{LE}.\text{SS}$$ 1

where LE leaf extract (%) and SS soluble solids (°Brix).

Figure 1a presents the surface graph built using Eq. (1). The increase in leaf extract and SS caused an increase in the overall liking, followed by a decrease at values above 28% and 14°Brix, respectively. The highest acceptance was obtained at leaf extract 28% (consequently 72% of cupuassu pulp) and SS of 14°Brix (critical point of Eq. (1), according to statistic program).

Fig. 1.

Surface graph of overall liking of blend beverages containing Hibiscus sabdariffa L. leaves extract (a) or infusion (b) and cupuassu pulp

Ogundele et al. (2016), using experimental design for the blend beverages containing H. sabdariffa extracts (HSE) with vegetables (pineapple, orange, and carrot), reported that formulations containing 26.30% of HSE were more acceptable by panelists. These authors concluded that HSE could be used as an ingredient for the production of a consumer acceptable functional beverage. This result is similar to the present study (28%), showing that it is possible the use of leaf extract for the blend functional.

For the leaf infusion, the lower overall acceptance was obtained in assay 7, which presented lower SS content (Table 1). Purbowati et al. (2020) reported that the production of H. sabdariffa functional drinks needs the addition of sweetening ingredients in a considerable amount to improve the flavor. Moreover, according to Monteiro et al. (2017), moderate to weak, bitter taste and astringency sensations often occur with perceptions of moderate to intense sourness in plant-based beverages, while moderate and high sweetness intensities are known to be generally suppressive of acid and bitter tastes. Thus, the sensory quality and acceptability of Hibiscus beverages are greatly affected by the sweetness. Therefore, this result emphasizes the importance of an adequate amount of SS.

For blend drink produced with leaf infusion, according to the estimated effects, the infusion concentration and SS affected the sensory acceptance, being significant at 95% of confidence level. Data of Table 1 was fitted to the quadratic model for overall liking given in Eq. (2). The model was significant since the calculated F value (11.73) was higher than the listed F value (F5.5 = 5.05). Proper correlation coefficients were also obtained (R² = 0.92).

$$\text{Overall}\text{liking}=8.77-0.44\text{LI}-1.74{\text{LI}}^2+0.53\text{SS}-1.83{\text{SS}}^2+0.02\text{LI}.\text{SS}$$ 2

where LI leaf infusion (%) and SS soluble solids (°Brix).

Figure 1b presents the surface graph built using Eq. (2). The increase in leaf infusion and SS increased the overall liking, followed by a reduction at values above 37% and 13°Brix, respectively. The highest acceptance was obtained at leaf infusion concentration 37% (consequently 63% of cupuassu pulp) and SS of 13°Brix.

When comparing the two blends, it is observed that it was possible to include a high percentage of infusion than extract in the beverage. This result is a consequence of the preparation form because, in the leaf extract, the green pigments provided a green appearance. This appearance was related to unfavorable by consumers. Moreover, it can be a consequence of polyphenolic compounds that is high in leaf extract (Table 2). Bechoff et al. (2014) reported that the concentration in polyphenols was significantly correlated to bitter taste in infusions and syrups of Hibiscus calyces.

Table 2.

Mean values and standard deviation of the physicochemical analyzes of extract and infusion leaves, cupuassu pulp, and blend beverages containing extract or infusion of Hibiscus sabdariffa L. leaves and pulp cupuassu (n = 5)

Parameters	Extract	Infusion	Cupuassu pulp	Formulations
				Extract/cupuassu1	Infusion/cupuassu2
pH	2.33 ± 0.00	2.82 ± 0.00	3.34 ± 0.00	3.16 ± 0.09a	3.30 ± 0.03a
RS3	0.01 ± 0.00	0.10 ± 0.12	3.30 ± 0.00	0.81 ± 0.02a	0.26 ± 0.03b
Total sugars4	0.21 ± 0.10	0.23 ± 0.01	9.01 ± 0.00	14.95 ± 0.23a	13.10 ± 0.02b
Vitamin C5	57.48 ± 0.00	57.50 ± 0.01	90.00 ± 0.00	38.26 ± 1.78a	36.01 ± 2.72a
PC6	171.96 ± 0.95	93.96 ± 0.37	74.90 ± 0.21	58.01 ± 0.05a	48.42 ± 0.94b
AA (ABTS)7	1.03 ± 0.05	0.56 ± 0.03	0.45 ± 0.04	0.35 ± 0.02a	0.29 ± 0.02b
AA (DPPH)7	0.11 ± 0.02	0.06 ± 0.01	0.05 ± 0.03	0.05 ± 0.01a	0.03 ± 0.00b

Means with different letters in the formulation’s columns differ according to the t test (p < 0.05)

¹Blend containing in mixed base 28% extract leaf, 72% cupuassu pulp, 14°Brix

²Blend containing in mixed base 37% infusion leaf, 63% cupuassu pulp, 13°Brix

³RS = reducing sugars expressed as g/100 mL of glucose

⁴Expressed as g/100 mL of glucose

⁵mg/100 mL

⁶Polyphenolic compounds expressed as mg of galic acid/100 mL

⁷AA = antioxidant activity expressed as µΜ Trolox/mL

Physico-chemical characteristics of beverage blends

The pH did not change (p > 0.05) between drinks (Table 2). H. sabdariffa leaves are known to be very acid (Meher et al. 2019), as can be seen in the extract and infusion pH values. However, herein, the high pH of the cupuassu pulp influenced the drinks mixed, which had pH values close to the pulp. This acidity reduction provided by cupuassu pulp in the blends is positive because the high acidity can compromise the beverage flavor.

The reducing and total sugars were higher in the blends produced with leaf extract (p < 0.05). Herein, the reducing sugars in beverages were provided by cupuassu pulp (3.30 g/100 mL) because the H. sabdariffa L. leaves did not have these compounds (Table 2). Moreover, the beverage with leaf extract has a higher pulp content (72%) when compared with those of leaf infusion. Mgaya-Kilima et al. (2014b), producing blends of H. sabdariffa with tropical fruits, also reported that the fruits were responsible for reducing sugars. These authors said that H. sabdariffa is known to have low sugar content. For total sugars, the similarity of values of each treatment was expected because the SS were standardized up to 14 and 13°Brix in the drink’s production. Thus, since these sugars are most likely the significant contributors to SS, these components are usually correlated to total sugars.

Vitamin C did not change (p > 0.05) between beverages (Table 2). The similarity in the vitamin C content in the two methods was not as expected because of the sensibility to heat this vitamin. Lima et al. (2010) reported high vitamin C stability with high temperatures. These authors said that this stability is explained by a protecting effect of the matrix used that was cashew apple. The authors also explained that degradation kinetics of food constituents might be related to molecular mobility, and the glass transition temperature has been used as the primary indicator of this mobility. Thus, the similarity of ascorbic acid obtained in the present study can be resultant of the used food matrices, H. sabdariffa leaves, and cupuassu fruit.

Meher et al. (2019) reported that the high acidity of H. sabdariffa leaves is a consequence of the highest amount of total organic acid (like ascorbic acid). These authors found values of 57.2 mg/100 g in puree concentrates. Thus, the values obtained herein for extract and infusion are according to observed by these authors. The vitamin C of the cupuassu pulp obtained in the present study was similar to the values reported by Pugliese et al. (2013) in fresh pulps (96–111 mg/100 g). Nowak et al. (2018) evaluating vitamin C—rich juices reported values of 32.55 and 25.00 mg/100 mL for sea buckthorn and cranberry. Therefore, the values obtained in the blends produced with leaf extract or infusion and cupuassu can be considered vitamin C-rich beverages because they had higher values (38.26 and 36.01 mg/100 mL, respectively).

The polyphenolic compounds and the antioxidant activity were higher in the beverages produced with leaf extract (p < 0.05) (Table 2).

According to Zhen et al. (2016), H. sabdariffa leaves contain high levels of polyphenol compounds, mainly chlorogenic acid and its isomers, quercetin, and kaempferol glycosides, which may contribute to its antioxidant capacity. Moreover, these authors concluded that leaf extract could reduce the free radicals in ABTS, proving its use as a functional beverage. Therefore, the results obtained herein are resultant by high polyphenolic compounds in leaf extract when compared to other raw materials. Moreover, the drinks of the present study provide functional effects by inserting of these nutritional components.

The high polyphenolic compounds and antioxidant activity in leaf extract when compared with leaf infusion, was due to aqueous extractions method. Bouabid et al. (2020), evaluating antioxidant properties of plant Atractylis gummifera L., reported that the highest concentration of polyphenols and antioxidant activity was found in the cold extract when compared to infusion. Moreover, these authors concluded that the active ingredients responsible for antioxidant better extracted cold than hot. Therefore, in the present study, the cold extract also provided a better extraction of these components in the leaves. This better extraction allowed that even using a smaller amount of the extract (28%), there was a higher concentration of these antioxidant compounds in the blend.

Storage of the beverage blends

pH, reducing, and total sugars and SS of blends increased linearly (p < 0.05) during storage (Table 3).

Table 3.

Mean values and standard deviation of the physicochemical analyzes of blend beverage containing extract or infusion of Hibiscus sabdariffa L. leaves and pulp cupuassu during storage for 180 days at 25 °C (n = 5)

Storage (days)	0	45	90	135	180
Extract Hibiscus sabdariffa L. leaves and pulp cupuassu beverage
pH1	3.16 ± 0.02	3.19 ± 0.01	3.19 ± 0.01	3.26 ± 0.02	3.31 ± 0.02
RS2	0.81 ± 0.06	3.58 ± 0.01	4.23 ± 0.02	5.77 ± 0.03	8.93 ± 1.00
Total sugars3	14.82 ± 1.32	14.98 ± 1.73	15.12 ± 0.79	15.81 ± 1.69	16.76 ± 0.87
TSS4	14.98 ± 0.16	15.04 ± 0.40	15.08 ± 0.19	15.08 ± 0.08	16.10 ± 0.27
Vitamin C5	38.26 ± 0.04	35.95 ± 0.05	33.63 ± 0.11	32.87 ± 0.06	27.55 ± 0.02
PC6	58.01 ± 1.04	39.30 ± 0.78	38.81 ± 0.80	37.51 ± 0.45	30.17 ± 0.18
AA (ABTS)7	0.35 ± 0.01	0.33 ± 0.01	0.31 ± 0.01	0.29 ± 0.00	0.25 ± 0.01
AA (DPPH)8	0.05 ± 0.01	0.04 ± 0.01	0.03 ± 0.00	0.03 ± 0.01	0.02 ± 0.01
Infusion Hibiscus sabdariffa L. leaves and pulp cupuassu beverage
pH9	3.30 ± 0.01	3.34 ± 0.02	3.42 ± 0.01	3.42 ± 0.02	3.57 ± 0.01
RS10	0.26 ± 0.05	0.65 ± 0.13	1.12 ± 0.38	1.29 ± 0.02	2.29 ± 0.39
Total sugars11	13.45 ± 0.24	13.53 ± 0.41	13.78 ± 0.27	14.07 ± 0.12	14.68 ± 0.19
TSS12	13.90 ± 0.32	14.20 ± 0.42	14.20 ± 0.66	14.34 ± 0.17	14.76 ± 0.25
Vitamin C13	36.01 ± 0.02	33.49 ± 0.02	32.76 ± 0.05	29.89 ± 0.01	28.81 ± 0.01
PC14	48.42 ± 0.60	43.51 ± 1.16	40.50 ± 0.86	29.46 ± 0.68	26.69 ± 0.03
AA (ABTS)15	0.29 ± 0.01	0.27 ± 0.03	0.26 ± 0.01	0.23 ± 0.01	0.22 ± 0.01
AA (DPPH)16	0.03 ± 0.00	0.03 ± 0.01	0.02 ± 0.00	0.02 ± 0.00	0.01 ± 0.01

¹Linear effect (y = 3.15 + 0.001x; R² = 0.88, p < 0.05), where y = pH, x = storage time, R² = determination coefficient

²RS = reducing sugars − g/100 mL of glucose, linear effect (y = 0.98 + 0.041x; R² = 0.94, p < 0.05), where y = RS

³Expressed as g/100 mL of glucose, linear effect (y = 14.56 + 0.01x; R² = 0.86, p < 0.05), where y = total sugars

⁴TSS = total soluble solids − °Brix, linear effect (y = 14.8 + 5.07x, R² = 0.58, p < 0.05), where y = TSS

⁵Expressed as mg/100 mL, linear effect (y = 38.55 − 0.054x, R² = 0.93, p < 0.05), where y = vitamin C

⁶PC = Polyphenolic compounds − mg of galic acid/100 mL, linear effect (y = 52.26 − 0.128x, R² = 0.77, p < 0.05), where y = PC

⁷AA = antioxidant activity − µΜ Trolox/mL, linear effect (y = 0.36 − 0.001x, R² = 0.93, p < 0.05), where y = AA (ABTS)

⁸AA = antioxidant activity − µΜ Trolox/mL, linear effect (y = 0.048 − 0.00001x, R² = 0.62, p < 0.05), where y = AA (DPPH)

⁹Linear effect (y = 3.24 + 0.001x; R² = 0.25, p < 0.05), where y = pH, x = storage time

¹⁰RS = reducing sugars − g/100 mL of glucose, linear effect (y = 0.18 + 0.01x; R² = 0.86, p < 0.05), where y = RS

¹¹Expressed as g/100 mL of glucose, linear effect (y = 0.91 + 0.05x, R² = 0.94, p < 0.05), where y = total sugars

¹²TSS = total soluble solids − °Brix, linear effect (y = 12.9 + 4.13x, R² = 0.88, p < 0.05), where y = TSS

¹³Expressed as mg/100 mL, linear effect (y = 35.79 − 0.04x, R² = 0.97, p < 0.05), where y = vitamin C

¹⁴PC = Polyphenolic compounds − mg of galic acid/100 mL, linear effect (y = 49.22 − 0.128x, R² = 0.95, p < 0.05), where y = PC

¹⁵AA = antioxidant activity − µΜ Trolox/mL, linear effect (y = 0.29 − 0.00001x, R² = 0.82, p < 0.05), where y = AA (ABTS)

¹⁶AA = antioxidant activity − µΜ Trolox/mL, linear effect (y = 0.031 − 0.00001x, R² = 0.63, p < 0.05), where y = AA (DPPH)

For the pH, according to Rehman et al. (2014), this increase with storage may be resultant of the acid hydrolysis of the polysaccharides into monosaccharides and disaccharides which are responsible for the increase in sweetness and decrease in sourness.

An increase in reducing sugars also was reported by Mgaya-Kilima et al. (2014a), evaluating the physico-chemical in the blend of H. sabdariffa calyces with fruit juice. According to these authors, this increase is due to the gradual conversion of nonreducing sugar (sucrose) and acids into reducing sugars (glucose and fructose).

For the total sugars and SS, the carbohydrates constitute the significant component of SS of fruit beverages and are mainly composed by similar proportions of three sugars, glucose and fructose, and sucrose (Gómez-López et al. 2018). Herein, more sucrose was added during beverage standardization, and it became the carbohydrate present in the highest concentration. Thus, the increase in °Brix is consistent with the rise of total sugars.

Olaniran et al. (2020), using preservatives for the to evaluate the storage stability of fruit juice blends, reported an increase in SS from 3.6 to 10.5% with 35 days. According to these authors, this increase was minimal and indicated a decrease in the rate of conversion of organic acid and hydrolysis of polysaccharides like starch, cellulose, pectin into simple sugar, thus increasing the shelf life of the juice. In the present study, the blends had a lower increase with storage of 180 days (0.40–7.48% and 2.32–6.67% for the drinks with leaf extract and infusion, respectively). Therefore, it can be concluded that these characteristics of beverages were preserved.

The vitamin C, polyphenolic compounds and the antioxidant activity reduced linearly (p < 0.05) during storage (Table 3).

For vitamin C, Singh and Sharma (2017) reported that degradation is common in all consumable items during storage and that storage condition has a more significant influence over ascorbic acid retention. Rabie et al. (2015), using the refrigerated temperature to minimize the vitamin C losses, reported a reduction of 22.37% in Physalis peruviana L. pasteurized juice with storage of 21 days. Herein, the blends had results similar to 180 days (28 and 20% for the drinks with leaf extract and infusion, respectively). Thus, it can be concluded that blends of H. sabdariffa leaves and cupuassu had the characteristic of beverages preserved.

Several researches with fruit juices indicated the phenolic compounds decrease with the storage (Wojdylo et al. 2014; Santos Filho et al. 2019). This reduction usually occurs as a result of the presence of dissolved oxygen in samples, which provided the oxidation of phenolic compounds (Nematollahi et al. 2016). However, the reduction was lower when compared to the other juices (Mgaya-Kilima et al. 2014a). Mgaya-Kilima et al. (2014a) also observed a reduction of polyphenolic compounds of the blend of H. sabdariffa calyces with fruit juice with 180 days. These authors reported a reduction of 51.01 and 66.16% in a mix of calyces with mango, in refrigerated (4 °C), and room (28 °C) temperature, respectively. The authors attributed this reduction to the transformation of some monomeric anthocyanins into polymeric compounds. Thus, the reduction obtained herein can be resultant of this change because the H. sabdariffa leaves are a source of anthocyanins (Wu et al. 2018). In the present study, the blend had a lower reduction with storage of 180 days at 25 °C (47.99 and 44.88% for the beverage with leaf extract and infusion, respectively). Therefore, just like with vitamin C, this bioactive compound was more preserved.

The antioxidant activity had a positive correlation with polyphenolic content. Nadeem et al. (2018) also observed this relation when stored a carrot-grape juice blend during 90 days.

For the sensory acceptance of beverages containing leaf extract, in the hedonic scale, all sensory attributes were positively scored up to 135 days, with a rating between ‘‘like slightly’’ and ‘‘like very much’’ (6.22 and 7.58). At 180 days, the acceptance reduced for rejection zone the scale for the most attributes (color, appearance, aroma, flavor, sweetness, and overall liking). According to Friedman’s test, for color and appearance, the higher values were with 0 and 45 days, followed by beverages with 90 and 135 days and lastly with 180 days. The attributes aroma, flavor, sweetness, acidity, body, and overall liking the lowest values were with 180 days (Table 4).

Table 4.

Sensory acceptance of blend beverage containing extract or infusion of Hibiscus sabdariffa L. leaves and pulp cupuassu during storage for 180 days at 25 °C

Storage (days)	0	45	90	135	180
Extract Hibiscus sabdariffa L. leaves and pulp cupuassu beverage
Color	7.58 ± 1.44A	7.48 ± 1.43A	6.68 ± 1.63B	6.33 ± 1.83B	4.38 ± 1.03C
Appearance	7.50 ± 1.55A	7.45 ± 1.42A	6.58 ± 1.65B	6.53 ± 1.74B	4.28 ± 1.01C
Aroma	6.37 ± 1.91A	6.85 ± 1.66A	6.85 ± 1.67A	6.98 ± 1.68A	4.62 ± 1.08B
Flavor	6.32 ± 1.98A	6.60 ± 1.90A	6.65 ± 1.78A	6.63 ± 1.07A	4.48 ± 1.03B
Sweetness	6.37 ± 1.78A	6.62 ± 1.83A	6.67 ± 1.71A	6.52 ± 1.06A	4.95 ± 1.33B
Acidity	6.22 ± 1.93A	6.43 ± 1.78A	6.38 ± 1.96A	6.48 ± 1.06A	5.00 ± 1.25B
Body	6.70 ± 1.57A	7.08 ± 1.29A	6.95 ± 1.42A	6.93 ± 1.69A	5.38 ± 1.98B
Overall liking	6.50 ± 1.68A	6.90 ± 1.62A	6.68 ± 1.71A	6.85 ± 1.80A	4.90 ± 1.97B
Infusion Hibiscus sabdariffa L. leaves and pulp cupuassu beverage
Color	7.42 ± 1.68A	7.38 ± 1.28A	6.93 ± 1.56A	5.47 ± 1.93B	5.07 ± 1.19B
Appearance	7.45 ± 1.51A	7.15 ± 1.41A	7.00 ± 1.48A	5.77 ± 2.13B	5.23 ± 1.95B
Aroma	6.77 ± 1.70A	6.77 ± 1.68A	6.48 ± 1.71B	5.68 ± 1.08C	5.40 ± 1.15C
Flavor	6.72 ± 1.91A	6.83 ± 1.88A	6.28 ± 1.91A	5.35 ± 1.34B	4.95 ± 1.56B
Sweetness	6.90 ± 1.80A	6.87 ± 1.81A	6.38 ± 1.00A	5.75 ± 1.26B	5.43 ± 1.27B
Acidity	6.63 ± 1.08A	6.50 ± 1.10A	6.27 ± 1.81A	5.97 ± 1.04B	5.45 ± 1.34B
Body	6.83 ± 1.76A	6.93 ± 1.65AB	6.67 ± 1.56AB	6.23 ± 1.66B	6.08 ± 1.72B
Overall liking	7.17 ± 1.47A	6.97 ± 1.75A	6.67 ± 1.69A	5.72 ± 1.96B	5.32 ± 1.17B

Means with different letters in the rows differ by the Friedman test (p < 0.05)

The drinks containing leaf infusion had all sensory attributes positively scored up to 90 days, with a rating between ‘‘like slightly’’ and ‘‘like very much’’ (6.27 and 7.42). At 135 and 180 days, except for flavor attribute with 180 days, the values were in the indifference region of the hedonic scale (neither like nor dislike) or acceptance region. Therefore, the blend with leaf infusion maintained better acceptance during storage when compared to leaf extract beverage. For Friedman test, color, appearance, flavor, sweetness, acidity, and overall liking had the lowest acceptance at the 135 and 180 days. The aroma had higher values with 0 and 45 days, followed by beverages with 90 and 135 days and lastly with 180 days. For the body, the drink stored with 135 and 180 days had less acceptance when compared with beginning storage (0 day) (Table 4).

According to Kathiravan et al. (2015), consumer acceptance is one of the most important criteria for any newly developed product. Therefore, the good acceptance of the products up to 135 days (leaf extract) or 180 days (leaf infusion) show that beverage blending is one of the best methods to improve the nutritional and sensory quality.

The reduction of color and appearance is resultant of a slight browning reported by consumers in the samples at the end storage. Pham et al. (2019) evaluated the browning during storage of orange juice. These authors reported that the formation of furfural and 5-hydroxymethylfurfural were highly correlated to the browning development of the juice during storage. In H. sabdariffa leaves, Zhen et al. (2016) reported the presence of 5-hydroxymethylfurfural during storage. Therefore, the browning reported by consumers can be due to 5-hydroxymethylfurfural presence.

For the aroma, the reduction in acceptance is a consequence of volatile compounds losses that occur with storage. Yang et al. (2015) reported that during the storage of the raspberry juice, the natural aroma compounds were gradually synthesized, transferred, and degraded, even derived from their original flavors. Therefore, these changes were perceived by consumers at the end of storage.

Jaworska et al. (2014) also observed a decrease in sensory acceptance of apple-whey beverages over 180 days. These authors reported that sweet and sour tastes had an essential impact on the flavor of drinks. Thus, in the present study, the reduction of flavor acceptance can be related to sweetness and acidity, which also reduced with storage. These results indicate that the increase of total sugars and pH (Table 3) affected the sensory acceptance of beverages at the end storage negatively.

For the body, according to Nawirska-Olszańsska et al. (2016), the viscosity of cloudy juices, regardless of the technology of their production, proved to be quite low. After storage, its values slightly decrease. Thus, the acceptance reduction of the body at the end storage is resultant of viscosity loss.

Overall liking is representing consumer acceptance (Kathiravan et al. 2015). Thus, in the present study, the overall liking also evidenced the other attributes.

Before releasing a beverage to the market, its essential to know the availability of purchase by the consumer is valued. The purchase intent of the blends had a reduction with the storage (Fig. 2a and b). This reduction was higher for drinks containing leaf extract. The highest percentage in this zone, “I would buy,” was higher up to 135 days in two drinks, reflecting the results of the hedonic scale.

Fig. 2.

Purchase intent of blend beverages containing Hibiscus sabdariffa L. leaves extract (a) or infusion (b) and cupuassu pulp during storage for 180 days at 25 °C

Conclusion

Although the use of calyx H. sabdariffa has been widely studied, few works have been done in this field considering the leaves. Herein, it was demonstrated that the blend with leaf extract or infusion and cupuassu pulp had good acceptance sensory and high nutritional value.

The conditions for beverage containing leaf extract production were 28% leaf extract, 72% cupuassu pulp, and 14°Brix. For beverage containing leaf infusion, were 37% leaf extract, 63% cupuassu pulp, and 13°Brix. The drink containing the cold extract leaf provided the extraction of polyphenolic compounds. The beverage blend with extract leaves provided the higher polyphenolic compounds and the antioxidant activity.

During storage, pH, reducing and total sugars and soluble solids of blends increased and vitamin C, polyphenolic compounds, and the antioxidant activity reduced. The blend with leaf infusion maintained better acceptance during storage. The sensory acceptance was maintained up to 135 days.