Effects of processing on onion skin powder added extrudates

Abstract

It is possible to enhance the functional properties of extruded products with the inclusion of fruit and vegetable by-products. Onion skin, a rich source of quercetin and fiber, is considered as waste in the industry and can be used as an alternative ingredient to improve the nutritional value of the extruded products. Three levels (3, 6, and 9%) of onion skin powder (OSP) were added to wheat flour and compared with control (0% OSP). The effect of the extrusion process on accessible quercetin, total phenolic content, and antioxidant activity of the samples were investigated. In addition, carbohydrate digestibility analyses were conducted for the products. Mass spectrometry (LC–MS/MS) results indicated that increasing the OSP level increased the quercetin content. The process caused the release of the entrapped quercetin from OSP, which was revealed by significantly higher quercetin levels for the extrudates. Some of the quercetin was lost during in vitro digestion process. Increasing the OSP level increased antioxidant activity and total phenolic contents of the samples. Total phenolic contents decreased significantly after the processing, yet antioxidant activities were not affected. The extruded products showed high amounts of rapidly available glucose (69.5 g/100 g). The OSP enhancement did not change the carbohydrate digestibility of products. The results indicated that the extrusion process could increase the level of accessible bioactive ingredients, and the level of functional compound addition can be optimized further.

Introduction

Development of ready to eat products, such as cereals and snacks, enhanced with functional compounds is a growing interest. Extrusion, which is a high pressure and temperature process, is often used for manufacturing such products (Leyva-Corral et al. 2016; Liu et al. 2019). This method offers modifications to the physicochemical and textural properties of the final product via chemical transformations and reactions during the process (Singh et al. 2019). Extrusion products are usually rich in starch content and have low nutritional values. Vegetables and fruits are rich sources of phytochemicals, which makes them a good substitute for incorporation into the extruded products to increase their nutritional value. Phytochemicals such as terpenoids (i.e., carotenoids) and phenolic compounds (i.e., flavonoids and isoflavonoids) are bioactive compounds (Alminger et al. 2014). Increasing interest in phytochemical consumption is mainly due to their association with health, such as cancer, diabetes, cardiovascular, and degenerative diseases prevention, as well as their antioxidant and anti-inflammatory properties (Onwulata 2012). Many studies investigated the effect of vegetables and fruit, such as tomato and grape pomace (Leyva-Corral et al. 2016), apple pomace (Liu et al. 2019), and artichoke leaf powder (Guven et al. 2018), enrichment of extruded products to increase their bioactive compounds contents and hence the antioxidant activity. Onion skin with a high dietary fiber content is a waste generated during industry peeling and is rich in phenolic acids (i.e., gallic acid and ferulic acid). Besides, it is one of the significant sources for quercetin (Lombard et al. 2005; Cheng et al. 2013). Quercetin, which is a flavonol, known to have antioxidant, anti-inflammatory, antihypertensive, anti-hypercholesterolemic, and anticancer properties (Świeca et al. 2013; Gawlik-Dziki et al. 2015; Ulusoy and Sanlier 2019).

Bioaccessibility is defined as the fraction of an ingested ingredient that is liberated from the food matrix during digestion (Aguilera 2019). Bioavailability, on the other hand, is defined as the fraction of the ingested ingredient or its metabolite that reaches the related site in circulation (Alminger et al. 2014). Therefore, the bioavailability of nutrients dependent on their release behavior from the ingested food matrix and their stability during digestion process. Specifically, bioaccessibility of phenolic compounds depends on many factors such as their solubility and extractability from the food matrix as well as the type of food processes used, such as thermal and mechanical processes (Alminger et al. 2014; Minekus et al. 2014). The molecular structure, which affects the solubility and extraction of phenolic compounds, can change due to pH and temperature variations throughout the food processes. The presence of dietary fiber and polysaccharides can decrease the extraction yield and prevent the enzyme-phenolic compound interactions as well as causing an increase in the medium viscosity, hence decrease the bioaccessibility (Alminger et al. 2014). In earlier studies, our group has shown that the extrusion increased in vitro bioaccessibility of lycopene (Tonyali et al. 2016), β-carotene and lutein (Ortak et al. 2017), and cynaroside and cynarin (Guven et al. 2018).

Food with a high glycemic index (GI), which is defined by the glucose concentration in the blood within 2 h of food consumption, is thought to have adverse health effects. The rate of starch digestion is one of the main factors determining the glycemic response. The rate at which glucose becomes available for absorption in the human small intestine can be reflected by glucose fractions determined by in vitro measurements (Englyst et al. 1999). Rapidly available glucose (RAG) and slowly available glucose (SAG) is the amount of glucose released after 20 and 120 min of in vitro digestion, respectively (Englyst and Hudson 1996). The remaining glucose is described as unavailable glucose (UG). The digestion of starch to its monomer glucose depends on many parameters such as food matrix (e.g., presence of proteins, lipids, and fibers), gelatinization degree of starch, and amylose to amylopectin ratio (Englyst and Hudson 1996). The extruded products that are rich-in-starch are examples of high GI products due to high temperature and shear conditions during processing causing complete starch gelatinization.

Although to our knowledge, there is not any study on the onion skin added extruded products, there are some recent studies focused on bread enriched with the onion skin (Gawlik-Dziki et al. 2013, 2015). This study aimed to investigate the effect of the extrusion process on accessible quercetin as well as the effect of OSP addition on the starch digestibility of extrudates. Changes in the total phenolic content and antioxidant activities of OSP added samples with extrusion were also investigated.

Materials and methods

Materials

Wheat flour and medium-size onions (Allium cepa L.) were purchased from Soke Un (Izmir, Turkey) and local markets (Ankara, Turkey), respectively. Chemicals used in LC–MS/MS were of liquid chromatography grade (Merck, Darmstadt, Germany), and the remaining chemicals were of analytical grade (Merck, Darmstadt, Germany, and Sigma-Aldrich, St. Louis, MO, USA). Pepsin (P7012), bile, bovine (B3883), pancreatin (7545), and Pefabloc were purchased from Sigma-Aldrich (Sigma-Aldrich, St. Louis, MO, USA). The enzymes for starch digestion analysis; pancreatin (7545), amyloglucosidase (A7095), invertase (I4504), and pepsin (P7125) were obtained from Sigma-Aldrich (Sigma-Aldrich, St. Louis, MO, USA). Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid), DPPH (2,2-diphenyl-1-picrylhydrazyl), and gallic acid (3,4,5-trihydroxybenzoicacid) were purchased from Sigma-Aldrich (Sigma-Aldrich, St. Louis, MO, USA), while Folin–Ciocalteau reagent was obtained from Merck (Merck, Germany).

Feed preparation

The outer dry skin layers of onions were removed, washed, and dried at room temperature for two weeks. The dried OS layers were ground in a grinder (KSW 445 CB, Bomann, Germany) and passed through a sieve (212 microns, 200 M.M B.S, Endecotts Ltd, London). The OSP was mixed with the wheat flour to have a final concentration of 3%, 6%, and 9% (w/w, db). The final moisture contents of mixtures were adjusted to 20 ± 2%. The control (no OSP added) and OSP added samples were kept in a refrigerator (4 °C) overnight. The feed samples were kept at room temperature for 2 h before the extrusion process to reach equilibrium. The schematic diagram of the experiments was presented in Fig. 1.

Fig. 1.

Schematic diagram of the experiments

Extrusion process

Laboratory scale twin-screw extruder (Feza Machine Co. Ltd., Istanbul, Turkey) was used for the extrusion process. The barrel length to diameter ratio (L/D) was 25:1, and the die diameter was 3 mm. The barrel temperature zones were set as 70 °C, 80 °C, 130 °C, and 150 °C. The feed flow rate and screw speed was 55 g/min and 250 rpm, respectively. The extruded samples were put into black bags and kept at − 20 °C until the analysis.

In vitro digestion simulation procedure for the accessible quercetin

In vitro digestion assays were performed at the Food Engineering Department of Ege University, Izmir, Turkey, according to the method described in Minekus et al. (2014). In brief, the digestion procedure was performed in three consecutive stages; oral, gastric, and duodenum. The compositions of the simulated salivary (SSF), gastric (SGF), and duodenum fluids (SDF) were given in Table 1.

Table 1.

The concentrations of SSF, SGF, and SDF

Stock concentration			SSF (pH 7)		SGF (pH 3)		SDF (pH 7)
			Stock volume	Concentration in SSF	Stock volume	Concentration in SGF	Stock volume	Concentration in SDF
	g L−1	mol L−1	mL	mmol L−1	mL	mmol L−1	mL	mmol L−1
KCl	37.3	0.5	15.1	15.1	6.9	6.9	6.8	6.8
KH2PO4	68	0.5	3.7	3.7	0.9	0.9	0.8	0.8
NaHCO3	84	1	6.8	13.6	12.5	25	42.5	85
NaCl	117	2	–	–	11.8	47.2	9.6	38.4
MgCl2 (H2O)6	30.5	0.15	0.5	0.15	0.4	0.1	1.1	0.33
(NH4)2CO3	48	0.5	0.06	0.06	0.5	0.5	–	–

SSF simulated salivary fluid, SGF simulated gastric fluid, SDF simulated duodenal fluid

Oral stage

1 g of sample was mixed with 5 mL of distilled water. 4 mL of SSF and 25 μL of CaCl₂ (0.3 M) were added to the mixture (Table 1). pH value was adjusted to 7 and the total volume was made up to 10 mL. Samples were incubated at 37 °C for 2 min to mimic the oral digestion.

Gastric stage

After incubation in the oral stage, 8 mL of SGF, and 5 μL of CaCl₂ (0.3 M) were added to the samples (Table 1). pH was adjusted to 3. Following 1 mL of pepsin solution (2000 U/mL) addition to the tubes, the total volume was completed to 20 mL. With incubation of samples at 37 °C for 2 h, the gastric digestion step was finalized.

Duodenal stage

After the completion of the incubation period in the gastric stage, 11 mL of SDF was added onto incubated samples (Table 1). 40 μL of CaCl₂ (0.3 M) and 2.5 mL of bile bovine solution (160 mM) were transferred into tubes. pH of samples was adjusted to 7. Later, 5 mL of pancreatic solution (100 U/mL) was added and the final volume was completed to 40 mL. The tubes were left in the incubator at 37 °C for 2 h to finalize the digestion.

After the incubation period, 45 μL of Pefabloc stock solution (500 mM) was added onto 5 mL of samples in order to prevent further digestion of samples. The mixture was centrifuged at 10000×g for 15 min. The supernatant was transferred to a new tube and kept at the freezer (− 20 °C) until further analysis.

Quantification method for quercetin (LC–MS/MS)

Extraction of quercetin from the samples before and after in vitro digestion

The quercetin was extracted from the samples following the method described in Takahashi et al. (2014). Briefly, the extrudate samples were ground in a grinder (KSW 445 CB, Bomann, Germany). 5 g of feed or ground extrudate sample was mixed with methanol (30 mL) and left in an ultrasonic bath (Jeiotech, Seoul, Korea) for 1 h. The mixture was centrifuged (2-16PK, Sigma Laborzentrifugen, Germany) at 15000×g for 10 min. The supernatant was separated, and the solid residue was mixed with 15 mL of methanol. The solid residue mixture was kept in the ultrasonic bath for one more hour. After one hour, the solid-methanol mixture was centrifuged again. The supernatant from the second centrifugation was added onto the first one and filtered through 0.45 μm polytetrafluoroethylene (PTFE) syringe-type filter and 0.2 μm PTFE syringe-type filter, respectively. The total volume of extract was completed to 50 mL with methanol.

For the samples after in vitro digestion, the procedure was slightly modified. The frozen supernatants from the digestion procedure were thawed at room temperature (~ 1 h) before the analyses. 0.5 mL of digested sample was transferred into every two Eppendorf tubes (Eppendorf, Hamburg, Germany). The tubes were centrifuged at 13,000 rpm (22673×g) for 90 s in a mini spin plus micro-centrifuge (Eppendorf, Hamburg, Germany). 0.4 mL of supernatant was mixed with 0.4 mL of mobile phase B (see LC–MS/MS analysis section) and tubes were centrifuged again. 0.2 mL of supernatant was diluted with 0.8 mL of mobile phase B and centrifuged again. After the second centrifugation, the supernatants from two tubes were combined and filtered through 0.45 μm and 0.2 μm syringe-type filters, respectively.

LC–MS/MS analysis

LC–MS/MS procedure was modified from Sánchez-Rabaneda et al. (2003). The analysis was carried out with a LC–MS/MS system equipped with ultra-high-performance liquid chromatography (UHPLC) pump (SPH1240™, Spark Holland) and electrospray ionization (ESI) interfaced with an auto-sampler (Alias™, Spark Holland). The oven temperature was set to 4 °C. The column was C₁₈ Synergi Fusion-RP 80 Å (Phenomenex, Aschaffenburg, Germany, 2.0 × 50 mm, 4 µm), and the column temperature was between 20 and 30 °C. The flow rate was 200 µL/min. The mobile phases A and B were 1% formic acid solution, and acetonitrile (CH₃CN) with 1% formic acid solution, respectively. The gradient profile (min/A%) was: (0.0/0), (1.0/15), (2.0/20), (2.2/10), (2.5/10), and (3.3/0).

The MS/MS detection was conducted by using a 3200 Q TRAP mass spectrometer (AB Sciex, Framingham, MA, USA) with the Turbo Ionspray source (negative ion mode). The recorded ion spray voltage was − 4500 V, the ion source temperature was 550 °C, and the injection volume was 10 µL. Nitrogen was used as the collision gas. Data processing and acquisition were performed by using Analyst 1.6 software.

In vitro bioaccessibility (%) of quercetin was calculated according to the equation below:

$$\text{Bioaccessibility}\left(\%\right)=100\times \left(\frac{\text{Amount}\text{after}\text{digestion}}{\text{Amount}\text{before}\text{digestion}}\right)$$ 1

Extraction procedure for antioxidant activity and total phenolic content

The extraction procedure was as described in Anton et al. (2009). 0.5 g of feed sample or ground extrudate sample was mixed with 12.5 mL acetone/water solution (80:20, v/v) and left on a magnetic stirrer (JeioTech-Multichannel Stirrer, MS-52 M) for 2 h. The mixture was centrifugated (Sigma, 2-16 PK, Germany) at 3000×g for 12 min. 0.45 μm syringe-type filter (Syringe Filter, PTFE) was used to filter the supernatants.

Total phenolic content analysis

Total phenolic content analysis was conducted according to the procedure described in Anton et al. (2009). In brief, 3 mL of Folin–Ciocalteau reagent was added to 0.4 mL of supernatants. After 5 min, 3 mL of sodium carbonate solution (60 g/L) was added. The mixture was left in the dark at room temperature for 90 min and the absorbance was read at 725 nm with a UV–Visible spectrophotometer. A calibration curve was prepared with gallic acid and the results were given as gallic acid equivalent (GAE) (mg)/dry weight (g).

Antioxidant activity analysis

The antioxidant activity of samples was measured by the method described in Anton et al. (2009). Briefly, 0.4 mL of filtered supernatants were reacted with 7.6 mL of DPPH solution (63 µM). The absorbance was read at 517 nm by using a UV–Visible spectrophotometer (Shimadzu, UV–Visible Spectrophotometer, UV-1700, Japan). A calibration curve of Trolox was prepared and the results were given as Trolox equivalent (TE) (µmol)/dry weight (g).

In vitro measurement of RAG, SAG and UG fractions

The in vitro starch digestions analyses were carried out by using the method described by Parada et al. (2011). The enzyme mixture used during the analysis was prepared fresh on the day of analysis with the following enzymes; pancreatin, amyloglucosidase, and invertase. 3 g of pancreatin was mixed with 20 mL of water. The mixture was mixed with vortex and stirred on the magnetic stirrer for 10 min. The mixed solution was centrifuged at 1500×g for 10 min. After centrifugation, the supernatant (15 mL) was transferred into another beaker and 4 mL of amyloglucosidase and 6 mL of invertase were added on it.

5 mL benzoic acid solution (50%) and 10 mL pepsin-guar gum solution (5 g pepsin/L and 5 g guar gam/L dissolved in 0.05 M HCI) were added on 0.7 grams of ground extrudate samples. The tubes were mixed with vortex briefly and placed into a water bath at 37 °C for 30 min. After 30 min, the samples were removed from the water bath. 5 mL enzyme mixture, 5 mL sodium acetate (0.5 M), and glass balls were added to each tube. The tubes were slowly shaken and put back to the water bath again (t = 0 min). 20 min later (t = 20 min), 0.2 mL of the mixture from each tube was transferred to another tube, which 4 mL of ethanol was added onto and labeled as G₂₀. The tubes containing the main mixture were placed into the water bath again immediately after the transfer. 100 min later (t = 120 min in total), the same procedure was applied and the tubes containing the ethanol were labeled as G₁₂₀. After the final transfer, the main tubes were put into the water bath at 100 °C for 30 min. The tubes were cooled for 15 min at room temperature before adding potassium hydroxide (0.7 M). Later, the tubes were further cooled in the water bath at 0 °C for 30 min. 1 mL acetic acid solution (1 M) and 320 μL amyloglucosidase solution were added onto 0.2 mL of mixture taken from the tubes. The new mixture was kept at the water bath at 70 °C for 30 min and 100 °C for 10 min, respectively. Finally, the tubes were left to cool to room temperature before adding 12 mL of ethanol, which was the total glucose (TG) sample.

The liquid samples from each sample set of G₂₀, G₁₂₀, and TG were centrifuged at 1500×g for 5 min, and the supernatants were separated. The supernatants were concentrated (Concentrator 5301, Eppendorf AG, Hamburg), then dissolved in 1 mL of distilled water, and filtered through 0.45 μm syringe-type filter. HPLC analyses of sugar fractions were conducted at METU Molecular Biology and Biotechnology R&D Center by adapting the official method of the European Community (NF EN 12630 1999). Measured values of G₂₀ (glucose released from the sample at 20 min), G₁₂₀ (glucose released from the sample at 120 min) and TG (total glucose in the sample) were used to calculate the glucose fractions. RAG (rapidly available glucose), SAG (slowly available glucose), and UG (unavailable glucose) were calculated according to the following equations (Parada and Aguilera 2011).

$$\text{RAG}\left(\text{g}/100\text{g}\right)={\text{G}}_{20}/\text{TG}\times 100$$ 2

$$\text{SAG}\left(\text{g}/100\text{g}\right)=\left({\text{G}}_{120}-{\text{G}}_{20}\right)/\text{TG}\times 100$$ 3

$$\text{UG}\left(\text{g}/100\text{g}\right)=\left(\text{TG}-{\text{G}}_{120}\right)/\text{TG}\times 100$$ 4

Statistical analysis

The results were analyzed by analysis of variance (ANOVA) using Minitab (Minitab Inc., State College, PA, USA, v17.2.1) to see any differences. When a significant difference (p ≤ 0.05) between samples was observed, Tukey’s Multiple Comparison Test was used.

Results and discussion

Contents and in vitro bioaccessibilities of quercetin

Increasing OSP content in the samples increased the quercetin content (Table 2). After the extrusion process an increase (p ≤ 0.05) was observed in quercetin contents of the 6% and 9% OSP added samples (Table 2), indicating that mechanical and thermal conditions of the extrusion process caused the release of quercetin from the cellular matrix and increased the extraction yield. A similar increase in quercetin contents was observed in onions after baking and sautéing (Lombard et al. 2005), in peanut kernel flours after roasting (Win et al. 2011), and in broccolis after boiling (dos Reis et al. 2015). While Lombard et al. (2005) attributed the increase of the quercetin amount to the concentration effect due to the loss of volatiles and water during cooking, dos Reis et al. (2015) and Win et al. (2011) suggested that the heat treatment disrupts the plant cells, releases the active ingredient, causes the breakdown of flavonol bonds, and exposes more quercetin in free form.

Table 2.

Quercetin contents of the feed and the extrudate samples before and after in vitro digestion with their calculated % in vitro bioaccessibility values

OSP level (%)	Quercetin content before in vitro digestion (μg quercetin/g dry weight)		Quercetin content after in vitro digestion (μg quercetin/g dry weight)		In vitro bioaccessibility (%)
	Feed	Extrudate	Feed	Extrudate	Feed	Extrudate
0	–	–	–	–	–	–
3	23.97 ± 2.87	24.58 ± 2.29	1.86 ± 0.41	10.90 ± 0.95	7.72 ± 0.79	44.34 ± 0.25
6	29.00 ± 0.76	60.49 ± 12.67	13.57 ± 0.49	23.88 ± 3.25	46.77 ± 0.45	39.79 ± 2.96
9	45.97 ± 6.03	100.50 ± 12.71	20.39 ± 0.06	55.30 ± 2.85	44.73 ± 5.72	55.29 ± 4.16

Results are mean ± SD (n = 2)

After the in vitro digestion simulations, measured quercetin contents of all OSP added feed and extruded samples were lower (Table 2), indicating that not all of the quercetin had survived the in vitro digestion conditions. The bioaccessibility of polyphenols changes with the particle size, the release from the matrix, the hydrophilic/lipophilic balance, the interaction with the food microenvironment, and the structure of the matrix (Alminger et al. 2014; Dueik and Bouchon 2016). Quercetin has poor solubility in water and is mainly carried by dietary fibers (Alminger et al. 2014). The extraction and solubility of quercetin are highly limited when it is entrapped within the dietary fiber matrix (Alminger et al. 2014). Moreover, the other components of food matrix such as polysaccharides interact with phenolic compounds via hydrogen bonding, hydrophobic interaction, covalent bonding, or physical entrapment (Salazar-López et al. 2018). Świeca et al. (2013) mentioned that the proteins in the wheat flour form complexes with phenols, and the bioaccessibility of these complexes depend on the pH, temperature, and protein and flavonoid concentrations. In their research, they noticed that onion skin supplementation amount (in the range of 1–4%) is correlated with the level of bound phenolics in wheat bread. Sivam et al. (2013) observed the phenolics form polyphenol—protein or—polysaccharide complexes during breadmaking steps (i.e., mixing) which affects the stability of phenols and reduces their extractability from the food matrix. Besides food components, the digestive enzymes (e.g., pancreatin and pepsin) might also interact with the phenol compounds and affect the bioaccessibility of quercetin (Salazar-López et al. 2018). Another factor that might interfere with the bioaccessibility is the harsh conditions of digestion.

The OSP added samples were exposed to acidic and slightly alkali conditions in the gastric and duodenal phase, respectively (Table 1). Although phenolic compounds survive under acidic conditions, they are not stable under weak alkali conditions (Liu et al. 2019). In Liu et al. (2019)’s study, the quercetin amount in apple pomace added extrudates decreased after in vitro intestinal phase due to degradation of quercetin under alkali conditions. In our study, it was clear that the extrusion process caused an increase in the available quercetin content, and with the decrease in quercetin contents after in vitro digestion calculated in vitro bioaccessibility (%) values were between 7.72 and 55.29 for OSP added samples (Table 2).

Total phenolic content of the feed and extrudate samples

The total phenolic content showed a positive correlation with increasing OSP amount for both feed and extruded samples (Fig. 2). Onion skin has high amounts of flavonoids and rich particularly in gallic acid, ferulic acid, quercetin, and quercetin derivatives (Cheng et al. 2013).

Fig. 2.

Total phenolic contents of OSP added feed and extrudate samples. Results are mean ± SD (n = 5). Significantly different values within feed or extrudate samples are marked by different lowercase letters (a, b, c, d) (p ≤ 0.05). Significantly different values between feed and extrudate samples are marked by different capital letters (A, B) (p ≤ 0.05)

The total phenolic contents of OSP added feed samples were significantly higher (p ≤ 0.05) than that of extrudate samples (Fig. 2). It ranged from 1.32 to 3.75 mg GAE/g dry sample, and 1.04–2.99 mg GAE/g dry sample for OSP added feed and extrudate samples, respectively. Previous studies have reported a decrease in phenolic content in extrudates of purple-flesh sweet potato flours (Soison et al. 2014), apple pomace incorporated oat flour (Leyva-Corral et al. 2016), buckwheat flour added cornflour (Singh et al. 2019), and artichoke leaf powder added wheat flour (Guven et al. 2018). This decrease in phenolic contents was attributed to the destruction of phenolic compounds as a result of the change in the molecular structure (Soison et al. 2014). Since the phenolic compounds are heat-labile, the free and conjugated phenolic compounds undergo thermal degradation during the extrusion process (Leyva-Corral et al. 2016). Moreover, the high shear stress of extrusion damages them, causing further loss (Leyva-Corral et al. 2016). Guven et al. (2018) suggested that decarboxylation of phenolic acids, such as gallic and ferulic acid in our study, could cause a decrease in total phenolic content. Korkerd et al. (2016) stated that the oxidation rate of phenolics increases under high temperatures of extrusion that causes low extraction yield.

Antioxidant activity of the feed and extrudate samples

The antioxidant activities of feed and extrudate samples significantly increased (p ≤ 0.05) with an increasing OSP amount (Fig. 3). Other researchers also found an increase in the antioxidant capacity of bread samples (Gawlik-Dziki et al. 2013) with the fortification of onion skin and, maize snacks (Beswa et al. 2016) with the enhancement of various types of vegetable and fruit parts.

Fig. 3.

Antioxidant activities of OSP added feed and extrudate samples. Results are mean ± SD (n = 5). Significantly different values within feed or extrudate samples are marked by different lowercase letters (a, b, c, d) (p ≤ 0.05). Significantly different values between feed and extrudate samples are marked by different capital letters (A, B) (p ≤ 0.05)

The antioxidant activity of feed and extrudate samples ranged from 15.3 to 44.42 µmol TE/g dry sample. Although antioxidants are heat-labile compounds, in our study, the extrusion treatment did not significantly (p > 0.05) change the antioxidant activity of samples (Fig. 3). Previous publications have reported a similar outcome (Soison et al. 2014; Leyva-Corral et al. 2016; Guven et al. 2018). Not having a significant change in the antioxidant activity after processing could be due to many reasons. Formation of Maillard reaction products or smaller phenolic compounds from polyphenols through degradation might mask the decrease in antioxidant compounds (Singh et al. 2019). Similarly, Sharma et al. (2015) did not observe a dramatic loss in antioxidant levels in onions after heat treatment (150 °C). Besides, the interaction of flavonoids with proteins might interfere with the assay (Altan et al. 2009). In Cheng et al. (2013), the antioxidant activity of onions had a higher correlation with the flavonoid content than the polyphenol content. This result indicates that the flavonoids, e.g., quercetin, are the main contributors to the antioxidant activity. The masking of these compounds might be the reason for not having a reduced antioxidant activity after processing and may need further study.

Starch digestibility

The starch digestibility analysis offers evaluation for the glycemic load of the extruded samples enhanced with OSP. RAG was significantly higher (p ≤ 0.05) than SAG and UG in the extruded samples (Fig. 4). RAG had 5.8- and 4.2-times higher values than SAG and UG, respectively. The starch digestibility is highly correlated with the gelatinization level. The gelatinization of starch is enhanced under high shear and high-temperature conditions of extrusion (Parada et al. 2011). The mechanical and thermal energy of extrusion transferred to the starch granules which causes the breakdown of chemical bonds in starch granules (Parada et al. 2011). The lower molecular weight compounds (i.e., decrease in size of amylose and amylopectin) are formed from disrupted starch granules and these newly-formed molecules are more accessible to digestive enzymes (Parada and Aguilera 2011). In our study, DSC thermograms showed that all the extrudate products had undergone full starch gelatinization (data not presented), which supports the results of the starch digestibility. Therefore, extruded products have higher RAG values compared to SAG values as a result of gelatinization. A similar inverse relationship between RAG and SAG was reported previously (Englyst et al. 1992; Rashmi and Urooj 2003).

Fig. 4.

RAG, SAG, and UG (%) values of OSP added extrudate samples. Results are mean ± SD (n = 3). Significantly different values between samples with different levels of OSP are marked by different lowercase letters (a, b) (p ≤ 0.05). Significantly different values between RAG, SAG, or UG of the same sample are marked by different capital letters (A, B) (p ≤ 0.05)

RAG, SAG, and UG of extruded samples did not significantly (p > 0.05) change with the incorporation of OSP at any level (3–9%) (Fig. 4). RAG was 69.5 ± 4.3 g/100 g, while SAG and UG were 12.5 ± 3.3 and 17.5 ± 4.4 g/100 g, respectively. Similarly, the addition of fiber to maize, potato, rice, and wheat extruded products (Parada et al. 2011) and inclusion of pea components (high fiber components) to barley extruded products (Brennan et al. 2016) up to 10% did not alter RAG, SAG, and UG in previous studies. Brennan et al. (2012) stated that fiber limits the starch gelatinization, which decreases the enzyme accessibility to starch granules and lowers the glycemic index. Capriles et al. (2008) attributed the reduction in starch digestibility to the entrapment of starch granules and degraded starch digestion products within dietary fiber cell structure and soluble dietary fiber gel matrix. Parada et al. (2011) mentioned that the proteins form a matrix around starch granules which hinders the accessibility of enzymes to the starch granules, while Al-Rabadi et al. (2011) suggested that proteins, as well as non-starch polysaccharides, compete with starch for water which limits the starch gelatinization. However, Brennan et al. (2016) discussed that the starch-protein matrix is destructed with extrusion, and Parada and Aguilera (2011) suggested that the digestibility of starch increases when the starch–protein matrix is destructed. Similar to the hypothesis of Parada and Aguilera (2011) on the formation and interference of lower molecular weight compounds, Neder-Suárez et al. (2016) stated these lower molecular weight compounds produce less-polymerized starch molecules which do not participate in the crystal structure and hence, decrease the resistant starch formation. Moreover, the authors added that enhanced starch gelatinization decreases resistant starch formation. Al-Rabadi et al. (2011) and Parada et al. (2011) mentioned the importance of particle size on the starch digestibility since the smaller the particle size, the faster the water penetration, and the starch gelatinization. Englyst and Englyst (2005) discussed that the degree of starch gelatinization is the dominant factor in the extent of starch digestibility.

The in vitro starch digestibility analysis showed that OSP added extrudates have high starch digestibility. After considering the hypothesis stated in the previous studies, we suggest that the high RAG is due to (1) small size of starch granules in wheat flour, (2) low gelatinization temperature and (3) low resistant starch formation. More than 90% of starch granules in wheat flour have a size of < 10 µm (Li et al. 2016). The extruded products had a gelatinization temperature around 63–65 °C and fully gelatinized during extrusion. The content of resistant starch was 1.6 ± 0.6 g/100 g and was not significantly different (p > 0.05) between control and OSP added extrudates.

Conclusion

The results indicated that the extrusion process disrupted the matrix and caused the release of the quercetin in the samples. Starch digestibility analysis indicated that the addition of OSP at this level (up to 9%) did not affect the measured glucose levels; rapidly available glucose, slowly available glucose, and unavailable glucose. This study provides a guideline for future studies in designing new generation food products with a higher level of bioactive compounds, which would serve as healthy alternatives to traditional ready-to-eat snacks.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.