Effects of Pitaya Peel Supplementation on Nutritional Quality, Overall Sensory Acceptance, In Vitro Glycemic Index, and Antioxidant Release from Fiber-Enriched Cookies

,


Introduction
Cookie is a well-known bakery product due to its low manufacturing cost, convenience of use, and long shelf-life. However, conventional cookies are high in starch, sugar, and fat but low in dietary fber and antioxidants which make them unhealthy for daily use [1]. Te addition of antioxidant dietary fber (ADF) to cookie formulation has attracted great attention. Diferent ADF sources originated from plant-byproducts have been added to cookie recipes as functional ingredients [2]. However, the bioavailability of bioactive compounds in ADFs which were added to the cookie recipe depends on many factors such as their chemical structure, molecular interactions with other compounds in the food matrix, as well as their release during gastrointestinal digestion [3].
Pitaya peel powder (PPP) has a high fber content, approximately 79% dw, and the ratio of insoluble fber to soluble fber is about 1.32; it also contains phenolics, betacyanins, and ascorbic acid with high antioxidant activities [4]. In 2016, wheat four was replaced by PPP at ratios of 5%, 10%, and 15% in the cookie recipe, and the obtained product had improved crude fber and ash content [5]. However, the efects of the PPP ratio in the cookie formulation on the dietary fber composition as well as the antioxidant activity of the product have not been reported. cookie samples incorporated with 10% PPP was examined. Te results showed that reduction in particle size of PPP decreased the dietary fber content of the cookie but enhanced its antioxidant content and activity [6]. Nevertheless, the glycemic index as well as the release of antioxidants from the PPP-incorporated cookies during in vitro digestion remained unknown.
In this study, diferent ratios of PPP were used in the cookie formulation. Te objective of this work was to investigate the efects of PPP supplementation on the nutritional quality of cookie samples, especially on their fber composition and antioxidant activity. In addition, the glycemic index and antioxidant release from the PPPadded cookies were in-vitro evaluated to clarify the potential use of this ADF source in the making of healthy food products.

Materials.
Pitaya fruits with purple-red peel and whitefesh (Hylocereus undatus) were harvested from a local farm in Dak Lak province (Vietnam). At the laboratory, the fruits were washed with tap water and drained; the pulp was manually separated, and the peel was then cut into 5 × 2 cm pieces. Each drying batch was conducted with 1.5 kg of fresh peel pieces at 60°C in a dryer (SF30, Memmert, Büchenbach, Germany) for 8 h, and the moisture content was achieved at 10 ± 1 (%). Te dried peel was crushed in a cutter mill and passed through a 70-mesh sieve. From 50 kg of pitaya fruits, about 12 kg of fresh peel and 1.4 kg of pitaya peel powder were collected. Te obtained pitaya peel powder was preserved in polyethylene bags at 4°C for experimentation.

Cookie Preparation.
Te cookie recipe included 150 g of wheat four and PPP, 46.61 g of isomalt, 0.18 g of acesulfame potassium, 0.66 g of table salt, 46.61 g of fresh chicken eggs (yolk and white), 74.51 g of butter, 1.61 g of baking soda powder, 0.6 g of vanilla, and 13.01 g of water. In the cookie preparation, the weight of PPP to the total weight of PPP and wheat four was changed: 0 (control), 10, 15, 20, and 25% (w/ w). Te cookie-making procedure was performed as described elsewhere [6].  [7]. Total dietary fber (TDF) content was computed as the sum of insoluble and soluble fber content. Total starch was analyzed using the AOAC 996.11 method [7]. Ash was evaluated by the AOAC 930.30 method [7]. Total carbohydrate was determined by subtracting the sum of lipid, protein, and ash percentages from 100%.

Betacyanin Content.
In order to extract betacyanin, 1 g of cookie samples were put into a 50 mL beaker containing 20 mL of distilled water. Te extraction was performed with an ultrasonic probe (VC750, Sonics, Newtown, CT, USA) at 150 W for 15 min. Te extraction temperature was kept at roughly 30°C by placing the beaker in a bath (WCB-22, Daihan, Seoul, Korea) with cold water. Te slurry was then centrifuged at 25°C and 3500 × g and for 5 min (ROTOFIX 32A, Hettich, Tuttlingen, Germany). Te supernatant was collected, and the procedure was carried out three times. Te resulting extracts were well mixed. Betacyanin content (Bc) was measured with a spectrophotometric method described by Ee et al. [8] and calculated as follows: where A is the absorbance (538 nm); M is the molecular weight of betacyanins (M = 550 g mol −1 ); f is the dilution factor and V is the pigment solution volume (mL); ε is molar extinction coefcients of betacyanins (65,000 L mol −1 cm −1 ); L is the path length of the cuvette (1 cm); W is the dry weight of the cookie sample used for extraction (g).

Total Phenolic Content. About 1 g of cookie samples
were added to a 100 mL beaker with 40 mL of a 60% (v/v) methanol solution for phenolic extraction. Te process of extraction was carried out using an ultrasonic probe (VC750, Sonics, Newtown, CT, USA) at 150 W for 15 min. Te extraction temperature was also kept at roughly 30°C as previously mentioned. Te slurry was separated by centrifugation (ROTOFIX 32A, Hettich, Tuttlingen, Germany) at 25°C and 3500 × g for 10 min, and the leftover material was extracted once more under the same circumstances. Te pooled supernatants were utilized to calculate the total phenolic content. By using the Folin−Ciocalteau reagent and spectrophotometric method, the total phenolic content was determined [9].

Determination of Antioxidant Activity.
Te phenolic and betacyanin extracts were mixed and utilized to assess the antioxidant capacity, which was measured by ferric reducing antioxidant power (FRAP) [10] and DPPH radical scavenging activity assays [11]; the results were presented as μmol Trolox equivalent per 100 g dry weight of the sample.

Sensory Evaluation.
After 24 h of baking, the cookie samples were evaluated by sixty untrained panelists who were casually recruited from the staf and students of Tay Nguyen University. A nine-point hedonic scale was used to evaluate the overall acceptability of Toledo et al. [12]. Te panelists were asked to give score from 1 point (extremely dislike) to 9 points (extremely like).

Estimation of the In Vitro Glycemic
Index. Te in vitro glycemic index of cookie samples was estimated following the method of Englyst et al. [13] with some adjustments. Approximately 2 g samples of cookies incorporated with 0, 10, 15, 20, and 25% PPP and white bread as a reference food item were crushed by a crusher (Model A11, Ika, Staufen, Germany) and mixed with 30 mL distilled water; the pH value was changed to 2.5 by 0.8 mL of 1 M HCl solution and maintained for 10 min. Ten, 1 mL pepsin solution (10% w/v prepared in 0.05 M HCl) was supplemented and kept in the dark at 120 rpm and 37°C for 30 min using an incubator shaker (Innova 40, New Brunswick Scientifc, NJ, USA). After that, the pH value was modifed to 6.9 by 1 mL 1 M NaHCO 3 solution before adding 5 mL pancreatin and 0.1 mL amyloglucosidase solution (2.5% w/v prepared in 0.01 M phosphate saline bufer with pH of 6.9). Te mixture was flled up to 50 mL with 0.01 M PBS pH 6.9 and further incubated at 37°C and 120 rpm. Sampling was performed at 0, 15, 30, 60, 120, and 180 min of the incubation; about 1 mL of the aliquot was separated and mixed with 4 mL of 99.5% v/v ethanol for enzyme inhibition. Te centrifugation was performed at 3000 × g for 15 min. Te supernatants were then collected to measure reducing sugar content by the spectrophotometric method using 3,5-dinitrosalicylic acid reagent and glucose as a standard [14]. Te hydrolyzed starch content was calculated by multiplying the reducing sugar content by 0.9. Te starch hydrolysis kinetics were described by a nonlinear model suggested by Goñi et al. [15] as follows: where C is the hydrolyzed starch content (g/L) at time t, C ∞ is the hydrolyzed starch content (g/L) after 180 min, k is the kinetic constant, and t is the reaction time (min). Te digestion curves were simulated and regressed using OriginLab OriginPro 8.5 software (OriginLab Corporation, Northampton, MA, USA) to obtain the value of k.
Te area under the curve of starch hydrolysis S was computed as follows: where t 0 = 0 and t f = 180 min. Te hydrolysis index (HI) was computed while white bread was used as reference food: Te in vitro glycemic index (GI) based on the bread reference was calculated as follows: When glucose is used as a reference food item (GI � 100), the glycemic index of white bread is 71. Te obtained results in this study were therefore multiplied by 0.71 to convert them into glycemic indexes based on glucose [16].

Evaluation of Antioxidant Release Using
In Vitro Sequential Digestion. Cookie samples with PPP ratios of 0 (control), 10, 15, 20, and 25% were subjected to in vitro sequential digestion following the method described by Colantuono et al. [17] with minor modifcations. Briefy, four steps of digestion including salivary, gastric, intestinal, and colon steps were carried out.
In the salivary digestion step, 2.5 g of cookie samples were crushed by a crusher and added to 15 mL of distilled water, the pH value of which was adjusted to 6.5; the mixture was incubated for 2 min and centrifuged. Te obtained pellet was suspended in 15 mL of distilled water for the next gastric digestion step; the suspension pH was modifed to 2.2 using 6 M HCl; about 2.5 mL of pepsin solution (12.5 mg/mL dissolved in 0.1 M HCl) was added; the resulting mixture was further incubated for 1 h and centrifuged. In the intestine digestion step, 15 mL of distilled water was supplemented to the resulting pellet; the pH value was altered to 7.5 using 1 M NaOH; about 2.5 mL of pancreatin solution (10 mg/mL dissolved in 1 M NaOH) was added; the obtained mixture was then incubated for 4 h and centrifuged. In the colon digestion step, the pellet was resuspended in 15 mL of distilled water and the pH value was changed to 8.0 before adding 2.5 mL of pronase E (1 mg/mL); the mixture was incubated for 1 h and centrifuged. Finally, the pellet was added to 17.5 mL of distilled water; the pH was then adjusted to 4.0, followed by the addition of 75 μL of Viscozyme L; the mixture was incubated for 16 h and centrifuged. For the salivary, gastric, intestinal, and colon steps of the in vitro digestion, the incubation was carried out at 37°C and 120 rpm in the incubator shaker. At each step, the supernatants (soluble fractions) were collected by centrifugation at 4°C and 4924 × g for 15 min (Allegra X-15R, Beckman Coulter, IN, USA) for the measurement of the release of betacyanin and total phenolic contents as well as antioxidant activities by FRAP and DPPH assays, whereas the pellets were used for the next steps. Blanks of reagents were also made in each step of the in vitro digestion. Te in vitro sequential digestion test was carried out in triplicate.
Te release percentage (RP) of antioxidant activity and content from the cookie samples was calculated as follows: where N 1 , N 2 , N 3, and N 4 are the antioxidant activity or the content of betacyanins or phenolics released from the investigated cookie sample at the salivary, gastric, intestinal, and colon steps, respectively; N is the initial antioxidant activity or the content of betacyanins or phenolics of the investigated cookie sample.

Statistical Analysis.
Each cookie sample was conducted in triplicate; the results were shown as means ± standard deviation (n � 3). One-way analysis of variance and multiple range tests with a signifcant level of p < 0.05 were performed by Statgraphics Centurion XV.I software (Manugistics Inc., Rockville, MN, USA).

Efects of Pitaya Peel Supplementation on Nutritional
Quality and Overall Acceptability of Cookies. When the ratio of PPP in the formulation increased from 0 to 25%, the total lipid content of the PPP-supplemented cookies was similar to that of the control cookies, while the protein, starch, and total carbohydrate content of the PPP-added cookies decreased by 7%, 21%, and 2%, respectively (Table 1). In contrast, the ash content of PPP-fortifed cookies was 83-220% higher than that of the control cookies. Some important minerals for human nutrition such as calcium, potassium, magnesium, zinc, and iron are identifed in pitaya peel [6]. Enhancement in the PPP ratio from 0% to 25% increased the total insoluble and soluble fber contents of the cookies by 7.1, 7.4, and 6.7 times, respectively. Cookie samples supplemented with 15% PPP or higher ratio were considered as foods high in fber since they had a total dietary fber content of over 6% dw [18]. Furthermore, when the level of PPP increased from 10 to 25%, the IDF:SDF ratio slightly changed, ranging from 1.85 to 1.93. Te IDF:SDF ratio of food is an important index for improvement in health benefts associated with dietary fber consumption; this ratio should be varied from 1.0 to 2.3 to achieve the healthy physiological efects of both soluble and insoluble fractions [19]. Diferent fber materials have various IDF:SDF ratios. When bulgur bran was added to the cookie formulation at 5-20%, the IDF and SDF of the obtained product ranged from 2.80% to 9.61% and 0.43% to 0.63%, respectively, and the IDF:SDF ratio of the cookies was therefore increased from 6.5 to 15.3 [20]. Dietary fber materials with high insoluble fber content are reported to be pretreated for the conversion of insoluble into soluble fber before being supplemented to cookie formulation to improve the balance of insoluble and soluble fractions [21]. It can be concluded that PPP is a potential dietary fber source for cookie fortifcation since the obtained product had a proper IDF: SDF ratio.
Te control cookies did not contain betacyanins since they were not identifed in wheat four. Te addition of PPP to the cookie recipe highly improved the antioxidant content and activity of the product. When the level of PPP increased from 0 to 25%, the phenolic content of the product enhanced by 204%. Te DPPH radical scavenging activity and FRAP of the cookies supplemented with 25% PPP were 7.0 and 5.4 times, respectively, greater than those of the control without PPP addition. Table 1 also shows that all PPP-added cookie samples were deemed to be acceptable due to their overall acceptance scores being greater than 5. When the ratio of PPP increased from 0 to 25%, the sensory acceptability of the cookies declined by 20.4%, which was probably due to the enhanced hardness (data not shown) and the obvious change in color ( Figure 1). It can be noted that the change in the PPP ratio from 0 to 15% did not afect the sensory acceptability of the product.

Efects of Pitaya Peel Supplementation on the In Vitro
Glycemic Index of Cookies. In this work, the selected nonlinear model was well ftted to the experimental data with R 2 > 0.94 (the data are not shown); this model was therefore used to calculate the GI of cookies samples. Te GI of all cookie samples varied from 79.9 to 64.4 in relation to white bread as a reference food or from 56.7 to 45.7 in relation to glucose reference. Te efects of PPP supplementation to the cookie recipe on the anticipated glycemic index of the product based on glucose reference are depicted in Figure 2.
It should be noted that the control cookie sample (GI = 56.7) was categorized as having a medium GI (GI = 55-69), while all PPP-supplemented cookie samples (GI = 45.7-51.9) were classifed as having a low GI (GI < 55). Te control cookies with a medium GI in this study could be attributed to the use of isomalt instead of sucrose in the product recipe. An increase in the PPP ratio in the cookie formulation from 0 to 25% reduced the GI of the product by about 19%. Te reduction in GI of PPP-supplemented cookies was due to their enhanced levels of SDF which enhances the viscosity of the digesta, thus impeding amylase access to their substrates [22]. In addition, there was evidence that amylase-catalysed starch hydrolysis may be affected due to the interaction between pectin and amylase [23]. Polygalacturonic acid, a pectin component of pitaya peel [24], might interact electrostatically with α-amylase and act as a noncompetitive inhibitor of this enzyme [25]. Moreover, cookies incorporated with a high PPP ratio contained more phenolic compounds which might reduce the glycemic index by inhibiting α-amylase and/or α-glucosidase as well as the interactions between polyphenol and starch molecules [26].

Release of Betacyanins, Phenolics, and Antioxidant Activities from Cookie Samples Incorporated with Diferent Ratios of Pitaya Peel Powder during the In Vitro Sequential
Digestion. Table 2 shows that the betacyanins of all PPPincorporated cookies were released during the four steps of the digestive tract. Te content of betacyanins liberated at the salivary step was much higher than that at the three remaining steps, and the value accounted for 60.0-60.5% of the initial content. It can be explained by their high water solubility [27]. Te higher the PPP ratio in the cookie recipe, the greater the content of betacyanin released at the salivary, gastric, and intestinal steps. It is probably due to the high initial betacyanin levels of cookies with the increased PPP ratio. Previously, betanin was not detected in colon fermentation fuid when the dosage of 23 mg/mL was used in human simulated gastrointestinal digestion [28]. In this study, the betacyanin content liberated at the colon step was statistically similar for all fber-enriched cookies and the value was 9.3% of the original level. Te presence of betacyanins in the colon can protect humans from diet-induced obesity and its related metabolic disorders thanks to the improvement of infammatory status and modulation of the gut microbiota [29]. Te release percentage of betacyanins was nearly 99%, and this value was statistically similar for all cookies fortifed with various PPP ratios ( Figure 3). According to Tesoriere et al. [30], when the purifed betalains were subjected to in vitro digestion, their loss was observed in the intestinal step, while the loss did not occur when raw red beet or cactus red fruit with betalains were used in the in vitro model. It can be confrmed that the bioavailability of betalains depends not only on their chemical stability in the digestive tract but also on the proximate composition of the food matrix. For the frst time, fber-enriched cookies were proven to have great potential for betacyanin protection in the in vitro digestion model. Similarly, phenolics were also released during the four steps of the digestive tract (Table 2). At each digestion step, the content of released phenolics tended to increase with the increased PPP ratio in the cookie recipe. Te highest release of phenolic compounds was recorded at the gastric step, and the obtained value varied from 103.3 to 188.6 mg GAE/100 g dw sample. It was anticipated that the low pH and the presence of pepsin in the gastric step would create favorable conditions for the release of some phenolics that were covalently or noncovalently bound to the protein in the food matrix, making these bioactives more accessible [31]. In contrast, some polyphenols in pomegranate peels and pomegranate peel-enriched cookies were released most at the simulated salivary step [17]. Various results are probably due to diferent impacts of the food matrix. Phenolic compounds were still liberated from all PPP-incorporated cookie samples throughout the large intestine step. It is possible that Viscozyme L and Pronase E were responsible for the release of nonextractable polyphenols associated with dietary fbers [31] and/or some low-molecular-weight phenolics linked to the melanoidins formed during the baking [32]. From the health point of view, the potential release of dietary phenolic compounds in the colon is useful. It is reported that phenolic compounds afect gut health since these bioactives impact the balance between major groups of gut bacteria including Bifdobacterium spp., Bacteroidetes, and Firmicutes, in comparison to the controls [33]. Furthermore, in the colon, dietary phenolic compounds appear to improve the formation of some useful metabolites by gut bacteria and/or restrict their production of detrimental molecules. Following the in vitro incubation of rutin, quercetin, chlorogenic acid, and cafeic acid, an increase in the overall synthesis of short-chain fatty acids was recorded; these metabolites provide energy for the growth of colonocytes and play a major role in preventing colorectal cancer [33]. Conversely, consumption of proanthocyanidin-rich extracts from grape seeds has the potential to reduce toxic products from the protein fermentation in the large intestine such as ammonia, phenol, pcresol, and skatole [34].
Te total content of phenolics released from all cookie samples ranged from 173.2 to 372.9 mg GAE/100 g dw in the in vitro digestive tract. Te released percentage of phenolic compounds ranged from 108 to 153% of the initial content of the undigested cookie samples and tended to decrease as the PPP ratio in the cookie formulation increased (Figure 3). It is hypothesized that increased soluble fber content in cookie samples might reduce the release of phenolic compounds. Food matrix with highly soluble fber content increases the viscosity of gastrointestinal fuids, restricting the peristaltic mixing process that promotes the transport of digestive enzymes to their substrates [35]. As a result, the release of phenolics which are bound with fber and other compounds such as carbohydrates, proteins, or lipids [31] in the food matrix is limited. In the same column and in the same step of the in vitro digestion, the values with diferent lowercase letters (a-e) are signifcantly diferent (p < 0.05); in the same column and with the same ratio of pitaya peel powder, the values with diferent capital letters (A-C) are statistically diferent (p < 0.05); dw: dry weight; GAE: gallic acid equivalent; DPPH: 2,2-diphenyl-1-picrylhydrazyl; TE: trolox equivalent; FRAP: ferric reducing antioxidant power. Table 2 also reveals that the antioxidant activities of the PPP-incorporated cookies were all signifcantly greater than those of the control cookies at all digestion steps. In the salivary step, the accessible antioxidant activities of the cookie incorporated with 25% PPP evaluated by FRAP and DPPH assays were 7.3 and 6.4 times, respectively, greater than those of the control cookies. At each digestion step, the increase in the PPP ratio in the cookie recipe signifcantly enhanced the antioxidant activity of the digested aqueous fractions. Te highest antioxidant activities were recorded at the gastric digestion step for all cookie samples; the value varied from 71.7 to 267.8 μmol TE/100 g dw for FRAP and from 52.6 to 318.1 μmol TE/100 g dw for DPPH radical scavenging capacity. Tis is probably related to the highest content of phenolic compounds liberated from the cookie samples at this step; the content of released phenolics positively correlated with the FRAP and DPPH radical scavenging capacity; their correlation coefcients (r) were 0.99 (p < 0.01) and 0.92 (p < 0.05), respectively.
Te total FRAP and DPPH radical scavenging capacities of all digested cookie samples were greater than those of the predigested samples. Te released percentage of antioxidant activities evaluated by FRAP and DPPH assays varied from 123 to 195% and 193 to 271%, respectively. Tis can be explained by the release of bound phenolics from the cookie matrix during the in vitro sequential digestion. In addition, the Maillard reaction products which are produced during the cookie baking [36] as well as the cookie proteins of which might be cleaved to release peptides [37] may contribute to the in vitro antioxidant activity. Nevertheless, when the ratio of PPP increased from 0 to 25%, the released percentage of antioxidant activities in the cookie samples decreased. Tis result was in accordance with the recent study, in which the increase in decafeinated cofee silverskin from 0 to 6% in the biscuit recipe resulted in a decreased release percentage of antioxidant activity from 745 to 394% [38].

Conclusions
Increased levels of PPP in the cookie formulation signifcantly enhanced betacyanin, phenolic, and dietary fber contents as well as the antioxidant capacities of the product. Te sensory acceptability of cookies added with 15% PPP and that of the control cookies were statistically similar. Te supplementation of PPP reduced the predicted GI of the product. In the in vitro sequential digestion, betacyanins, phenolics, and antioxidant activities of cookie samples were released at all salivary, gastric, intestinal, and colon steps, and the highest release of antioxidants was recorded at the gastric step. Te enhanced level of PPP in the cookie recipe signifcantly improved the antioxidant activities of the aqueous fractions at all digestive steps. PPP would be a prospective antioxidant and dietary fber ingredient to be added to cookie products.

Data Availability
Data are available on request.

Ethical Approval
Te study was covered by a general approval from the Ethics Review Board at the University of Social Sciences and Humanities, Vietnam National University -Ho Chi Minh City (VNU-HCM). Participants were instructed to read an information sheet and sign a consent form. Participants gave voluntary consent and were assured that their responses would remain confdential. Tey were informed that they could withdraw at any point without any consequences. Participants received monetary compensation for their participation.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Authors' Contributions
Ti Hai Anh Mai was responsible for investigation, methodology, formal analysis, and writing of the original draft. Ti Tu Tra Tran was responsible for visualization, data curation, and supervision (equal). Van Viet Man Le was responsible for conceptualization, project administration, supervision (equal) and writing, reviewing, and editing.