Partial substitution of wheat flour with kiwi starch: Rheology, microstructure changes in dough and the quality properties of bread

Abstract

Kiwi starch (KS) is a new fruit-derived starch-based food material. In this study, wheat flour was partially replaced with 10–20% KS to make bread, and the influence of this substitution on mixed flour, dough processing performance, bread quality, and shelf life was investigated. KS substitution improved the water-binding ability of mixed flour, making it easier to gelatinize while improving viscoelasticity but reducing the integrity of the dough's gluten network structure. As the substitution rate increases, the hardness, air-cell ratio, and width-to-height ratio of bread significantly increased, while the springiness, resilience, baking loss, and specific volume reduced significantly (p < 0.05). KS enriched the bread's color and flavor by promoting the Maillard reaction during baking. Overall acceptability of 10% KS group was highest in sensory evaluation. KS substitution significantly reduced starch digestibility and expected glycemic index (GI), inhibited mold growth and reproduction during storage and prolonged the shelf life of the bread at 25 °C.

Graphical abstract

Highlights

• •Kiwi starch (KS) diluted the gluten network, improving the dough's viscoelasticity.
• •KS enriched the bread's color and flavor, 10% KS group had the highest sensory score.
• •Adding KS can inhibit bread's digestibility, help stabilize postprandial blood sugar.
• •KS substitution extended the shelf life of bread, but made it more prone to staling.
• •KS is a potential low-GI food ingredient and a clean label ingredient.

1. Introduction

Wheat bread is a staple diet in many nations and a major source of carbohydrates in European diets. The 2018 Whole Grain and Refined Grains-Family Grocery Purchases survey in the United States found that >90% of bread products contain refined wheat ingredients (Dunford et al., 2022). Type 2 diabetes (T₂D) is highly correlated with long-term consumption of refined carbohydrates with a high glycemic index (GI) (Livesey et al., 2019). Prevention of T₂D is a major goal of the World Health Organization and the International Diabetes Federation, and consuming foods that elicit lower-GI and insulin responses may help reduce the incidence of these chronic metabolic disorders (Demirkesen-Bicak et al., 2021). Therefore, developing low-GI breads is important for preventing chronic metabolic diseases such as T₂D.

In recent years, researchers have focused on modifying or partially substituting ingredients (Amini Khoozani et al., 2020), adding dietary fiber or antioxidant components (Chen et al., 2019; Gkountenoudi-Eskitzi et al., 2023), developing wheat flour with high amylose starch to modulate the postprandial glycemic response to bread. Gkountenoudi-Eskitzi et al. (2023) added 30% acorn flour and chickpea flour in a bread formulation, finding that the antioxidant properties of bread were enhanced and the GI was significantly reduced. Compared to traditional wheat bread, wheat bread supplemented with 50% high amylose (AM) had a sixfold increase in resistant starch (RS) content, while the digestibility was reduced by 80% (Li et al., 2022). However, these breads are not well known to the public due to low consumer acceptance and insufficient commercialization; thus, new solutions must be explored to obtain commercially viable, good-quality, lower-GI bread.

Kiwi starch (KS) is a high-value byproduct of solid waste in the kiwifruit industry. This high-quality, fruit-derived starch is available in great quantity. It is rich in polyphenols and dietary fibers, with low pH, high antioxidant activity, and high AM and RS content (Wang et al., 2022; Wang, Lan, et al., 2021), making it a promising development resource. Adding KS to starch-based food materials is potentially useful for developing low-GI foods. The high AM of KS confers higher thermal stability, making it resistant to digestion and absorption (Li, Dhital, & Gidley, 2023). AM has a more compact structure that is not easily hydrolyzed. Adding KS may also increase the RS content of starch-based foods and inhibit starch hydrolysis (Li et al., 2022). However, studies on KS have primarily been theoretical, and no studies have been published on its application in food processing. Previous studies have shown that bread additives can significantly improve glycemic response and nutritional properties but negatively impact bread quality, which largely depends on the level of substitution. For instance, Wang, Lao, et al. (2021) replaced wheat flour with whole quinoa flour and found that <20% substitution had no significant effect on bread's hardness and chewiness, both of which are determinants of consumer acceptance. However, substitution >20% caused bread's hardness to significantly increased and the specific volume (SV) decreased by 10%. At 40% substitution, the starch digestibility of the bread was reduced by 17%, but SV was significantly reduced by 57%, leading to a significant decline in consumer acceptance. Thus, partially replacing wheat flour with KS can regulate the postprandial glycemic response to bread, but the substitution level may also significantly affect bread quality, affecting consumer acceptance. Therefore, it is vital to investigate the effects of various KS substitution ratios on bread digestibility and quality to support the development of low-GI bread.

This article used different proportions of KS (10%, 15%, and 20%) to replace wheat flour in bread, and evaluated the impacts on the hydration and gelatinization characteristics of composite flour, as well as the rheological properties, mixing properties, and dough microstructure. This article also explored the impact of KS substitution on bread quality and shelf life so as to develop low-GI bread with excellent sensory quality and consumer acceptance.

2. Materials and methods

2.1. Materials and chemicals

2.1.1. Materials

Starch-rich Ruiyu kiwifruit was picked from the kiwifruit experimental station in Wugong County, Shaanxi Province. The dry matter content of kiwifruit was about 18.36%, the soluble solids content (SSC) was about 7.0 °Brix, the hardness was about 100 N, and the starch content was about 9.18% by fresh weight (FW). KS was extracted from the kiwifruit in the laboratory. High-gluten wheat flour, high-active dry yeast, soft sugar, and butter were purchased from nearby supermarkets.

2.1.2. Chemicals

Porcine pancreatic α-amylase (23 U/mg), pancreatic lipase (57 U/mg), invertase (≥300 U/mg), and glucosidase (8 × USP) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Pectinase (500 U/mg) and cellulase (50 U/mg) were purchased from Yuan Ye Biotechnology (Shanghai, China). Starch (A148–1-1) and glucose (A154–1-1) kits were purchased from Nanjing Jiancheng Bioengineering Institute (Jiangsu, China). All other reagents were analytical grade and purchased from Xinfang Chemical Reagent (Yangling, China).

2.2. Preparation of kiwi starch

The kiwifruit skin and seeds were removed, the flesh cut into pieces, and then suspended in ice water at 40%. The flesh mass was beaten to obtain the pulp, and then adjust the pH to 4.00 using food-grade sodium bicarbonate. Then, following the addition of 0.2% mass fractions of cellulase and pectinase, the pulp was enzymatically hydrolyzed for 2 h at 50°C in a water bath. Centrifuge the hydrolysate for 5 min at 11, 500 ×g, remove the supernatant, and wash the sediment with distilled water. After repeating the above centrifugation and washing three times, the collected sediments were placed in a 40 °C oven and dried to a constant weight, then ground the dried extract and pass through a 100-mesh sieve to obtain KS (Wang, Lan, et al., 2021).

2.3. Product processing

Wheat flour (135 g) was combined with 1.4% high-activity dry yeast, 12% soft white sugar, 1% salt, 6% butter, 55% water, and 4.8% milk powder to yield ∼250 g of dough by a dough mixer. The dough was placed at 28°C to rise, then shaped and placed in a square mold (10 × 10 × 10 cm) for fermentation at 38°C. After fermentation, the bread was baked for 30 min at 180°C for both upper and lower fires in a preheated oven; then, the upper fire was adjusted to 200°C for 8 min. All tests were conducted after the bread had cooled to room temperature. The groups of pure wheat flour were designated as WF, and the groups of 10%, 15%, and 20% KS replacement wheat flour were designated as KF10, KF15, and KF20, respectively.

2.4. Flour characteristics

2.4.1. Basic indicators

The moisture content of wheat flour and KS was measured using a moisture meter (MB27ZH, OHAUS instrument, New Jersey, USA). The fat content, protein content, and crude fiber content of wheat flour and KS were determined according to the methods described in Chinese national standards The National Standard of China, 2016a, The National Standard of China, 2016b, and The National Standard of China. GB/T 5009.10–2003, 2003, respectively. The starch purity of wheat flour and KS was determined using a starch reagent kit based on acid hydrolysis combined with anthrone colourimetry. The pH and amylose content were measured using the method of Lan et al. (2024). The results are listed in Table S1.

2.4.2. Hydration properties

According to the method of Cornejo & Rosell (2015), the flour's water absorbance index (WAI), water solubility index (WSI), and swelling power (SP) were calculated. Uniformly disperse the sample (1.000 ± 0.001 g) in 10 mL of distilled water, gelatinize at 95°C for 15 min, then cool to room temperature, and centrifuged at 5, 400 ×g for 10 min. The weight of the sample was recorded as m₀, the weight of the wet sediment was recorded as m₁, and the supernatant was dried to a consistent weight and recorded as m₂.

$$\mathrm{WAI}\left(\mathrm{g}/\mathrm{g}\right)=\frac{m_1}{m_0}$$ (1)

$$\mathrm{WSI}\left(\%\right)=\frac{m_2}{m_0}\times 100\%$$ (2)

$$\mathrm{SP}\left(\mathrm{g}/\mathrm{g}\right)=\frac{m_1}{m_0-m_2}$$ (3)

2.4.3. Pasting properties

According to the descriptions of Lan et al. (2024), the pasting properties of different flour samples were determined by a rapid viscosity analyzer (Tec Master, Perten Instruments, Stockholm, Sweden). The peak time (Ptime), pasting temperature (PT), peak viscosity (PV), trough viscosity (TV), breakdown value (BD), setback value (SB), and final viscosity (FV) were automatically analyzed by the instrument supporting software.

2.5. Dough characteristics

2.5.1. Mixolab

The dough mixing characteristics were analyzed using a Mixolab mixing tester (Mixolab2, Chopin Technologies, Paris, France). Chopin + standard was selected for testing. The results were expressed as water absorption (WA, %), C1 (N·m), C2 (N·m), C3 (N·m), C4 (N·m), development time (TF, min), and stability time (TS, min) (Sun, Bu, et al., 2023).

2.5.2. Rheological properties

The dynamic rheological properties of the dough were measured by Rheometers as described elsewhere (DHR-1, Waters, Milford, USA) (Atudorei et al., 2022). Two grams of dough without yeast was placed on the platform, and frequency scanning was performed using 40-mm plates with 1-mm plate spacing, stress 1.0%, temperature 25°C, and scanning frequency 0.1–20 Hz.

2.5.3. Textural properties

The textural properties of dough were evaluated using a TA-XT Plus Texture Analyzer (TA. XT PLUS/50, Stable Micro Systems, Godalming, UK) in texture profile analysis mode (Qin et al., 2021). Samples (30 g) of dough made without yeast were formed into uniform pellets and compressed twice to 40% strain with a 36-mm aluminum probe (P/36R) at 1 mm/s and 5 g trigger force.

2.5.4. Scanning electron microscopy

The dough samples were freeze-dried and sprayed with gold, then observed by scanning electron microscopy (SEM, Nano SEM-450, FEI, Czech Hillsboro, USA) with a magnification of 500. Fix a small amount of KS and wheat flour onto a metal sample table using conductive double-sided tape and observe at the magnification of 3000.

2.6. Product quality

2.6.1. Baking properties

The calculated baking characteristics included specific volume (SV), width-to-height ratio, baking loss, and air-cell ratio. The volume determination was based on the rapeseed replacement method (Wang et al., 2023) but with millet instead of rapeseed. SV and baking loss were calculated as described by Amini Khoozani et al. (2020) and Guadalupe-Moyano et al. (2022). The air-cell ratio of the bread was calculated using Image J software.

$$\text{Baking loss}\left(\%\right)=\frac{\text{Dough weight}-\text{Bread weight}}{\text{Dough weight}}$$ (4)

$$\mathrm{SV}\left({\mathrm{cm}}^3/\mathrm{g}\right)=\frac{\text{Apparent volume of bread}}{\text{Bread weight}}$$ (5)

2.6.2. Crumb texture

Samples (15 × 15 × 15 mm) were obtained from the center of each freshly baked bread, and the texture analysis was performed as described for dough in Section 2.5.3. Textural parameters included hardness, springiness, cohesiveness, gumminess, chewiness, and resilience.

2.6.3. Crust and crumb color

Bread crust and crumb color characteristics were analyzed using a Colorimeter (CS-820, Caipu Technology, Hangzhou, China) in reflection mode. The crumbs and crusts were measured at nine different points, and lightness (L*), red-greenness (a*), yellow-blueness (b*), chroma (C*), hue (h), and ΔE were used to express color profiles. ΔE was calculated using formula (6)

$$\mathrm{\Delta E}=\sqrt{{\left(L_0^{\ast }-L^{\ast }\right)}^2+{\left(a_0^{\ast }-a^{\ast }\right)}^2+{\left(b_0^{\ast }-b^{\ast }\right)}^2}$$ (6)

where L₀*, a₀*, and b₀* are the L*, a*, and b* of WF, respectively.

2.6.4. Sensory evaluation

Thirty evaluators assessed the bread samples based on their appearance, color, flavor, texture, taste, and overall acceptability using a nine-point hedonic scale. The scale ranged from 9 to 1, with 1 meaning dislike very much, 5 meaning neither like nor dislike, and 9 meaning like very much (Atudorei et al., 2022).

2.7. In vitro starch digestion

In vitro digestion was evaluated as described by Wang et al. (2022). The digestive data of the bread samples were analyzed by fitting the first-order kinetic formula (7) and calculating the expected glycemic index (eGI) of bread samples according to the description formula (10) of Guadalupe-Moyano et al. (2022).

$${\mathrm{C}}_{\mathrm{t}}={\mathrm{C}}_{\infty }\left(1-{\mathrm{e}}^{-\mathrm{kt}}\right)$$ (7)

$$\mathrm{AUC}={\mathrm{C}}_{\infty }\left({\mathrm{t}}_{\mathrm{f}}-{\mathrm{t}}_0\right)-\left(\frac{{\mathrm{C}}_{\infty }}{\mathrm{k}}\right)\left(1-{\mathrm{e}}^{-\mathrm{k}\left({\mathrm{t}}_{\mathrm{f}}-{\mathrm{t}}_0\right)}\right)$$ (8)

$$\mathrm{HI}\left(\%\right)=\frac{{\mathrm{AUC}}_{\text{Sample}}}{{\mathrm{AUC}}_{\text{White bread}}}\times 100$$ (9)

$$\mathrm{eGI}=8.198+0.862\times \mathrm{HI}$$ (10)

C_t represents the glucose concentration of starch hydrolysis at the reaction time t, C_∞ is the estimated value of starch hydrolysis at the end of the reaction, and k is the first-order kinetic coefficient. The area under the curve (AUC) of starch hydrolysis was calculated using formula (8). The hydrolysis index (HI) was calculated using formula (9), using commercial white bread as a reference.

2.8. Storage characteristics

The mold growth of bread samples was monitored at 25°C for 0–6 days; Using an MB25ZH moisture meter (MB27ZH, Ohaus, New Jersey, USA) to detect the moisture content of bread during storage. The XRD diffraction patterns of bread samples during storage were obtained using the X-ray diffraction analyzer (D8 Advance A25, Bruker Corporation, Billerica, Germany). Operating parameters: pipe pressure = 40 kV, pipe flow = 30 mA, and scanning area 2θ = 5–40°. Calculate the relative crystallinities of samples using the Jade 6.5 software (Li et al., 2022).

2.9. Statistical analysis

One-way analysis of variance, Duncan's test, and correlation analysis were performed using SPSS 22 (IBM Corp., Armonk, USA). Other images were drawn using Excel 2016 (Microsoft, Redmond, USA) and Origin 9.1 (OriginLab, Southampton, USA). All experiments were repeated at least three times. The results were shown in the form of mean ± standard deviation (SD).

3. Results and discussion

3.1. Effects of KS substitution on mixed flour

3.1.1. Hydration properties

The hydration properties of starch can reflect the interaction between starch and water molecules, which can be influenced by the types of starch, the ratio of AM to amylopectin, other components in starch such as protein and fat, and the particle size of starch (Wang et al., 2022). Thus, it is critical to investigate the hydration characteristics of starch in water systems. As shown in Table 1, 10–20% KS substitution considerably improved the WA and WAI of the mixed flour (p < 0.05), indicating that KS had stronger water absorption than wheat flour, and the increased WA means that bread has a higher production rate after KS substitution. This was mainly because KS substitution increased the total starch and AM contents (Table S1). Additionally, Mohebbi et al. (2018) reported that wheat flour with high amylose corn starch exhibited stronger WA, consistent with our results. Compared with the WF group, the WSI of the KF group was significantly reduced (p < 0.05). Table S1 shows that the AM contents in KS and WF were 35.70% and 27.84%, respectively, which resulted in an increase in the AM content of the mixed flour after KS substitution. The formation of partial helical structures between amylose and amylopectin, as well as the entanglement between amylose chains, may enhance the rigidity of starch granules, thereby inhibiting the expansion of starch granules, resulting in a decrease in AM leaching and WSI (Jia et al., 2023). This may also be because KS is a typically small granular starch with a large specific surface area, which is more likely to form amylose-fat complexes in the presence of fat (Li, Gong, et al., 2020), thereby hindering the leaching of AM and thus decreasing the WSI of the mixed flour. In addition, within the KS substitution range in this experiment, KS substitution below 20% did not significantly affect the SP (p > 0.05) of the mixed flour.

Table 1.

Effects of KS substitution on the hydration properties, pasting properties of flours.

		WF	KF10	KF15	KF20
Hydration properties	WA (%)	58.70 ± 0.00c	60.00 ± 0.00b	60.90 ± 0.00b	62.00 ± 0.00a
	WAI (g/g)	6.47 ± 0.06c	6.62 ± 0.08ab	6.72 ± 0.02ab	7.15 ± 0.55a
	WSI (%)	4.45 ± 0.32a	3.55 ± 0.36b	2.93 ± 0.57c	2.80 ± 0.09c
	SP (g/g)	6.77 ± 0.08a	6.86 ± 0.08a	7.35 ± 0.52a	7.35 ± 0.56a
Pasting properties	Ptime (min)	5.89 ± 0.03a	5.82 ± 0.04ab	5.82 ± 0.04ab	5.78 ± 0.04c
	PT (°C)	89.93 ± 0.49a	88.78 ± 0.80ab	88.03 ± 0.08bc	87.18 ± 0.80c
	PV (cP)	1373.00 ± 38.97c	1417.67 ± 20.60b	1411.33 ± 4.04bc	1564.33 ± 8.50a
	TV (cP)	850.67 ± 34.12c	901.00 ± 9.17bc	944.33 ± 13.43b	1008.00 ± 50.27a
	FV (cP)	1618.00 ± 48.66c	1724.33 ± 20.50b	1812.67 ± 1.53ab	1911.67 ± 93.19a
	BD (cP)	517.00 ± 11.31a	508.00 ± 4.24ab	549.00 ± 5.66bc	578.50 ± 23.33c
	SB (cP)	767.33 ± 15.63c	823.33 ± 11.68b	868.33 ± 12.10ab	903.67 ± 43.94a

Note: The results are expressed as mean ± standard deviation (n = 3), and there is a significant difference in the values of different letters in the same line (p < 0.05).

3.1.2. Pasting properties

The gelatinization of starch is an irreversible thermally induced phase transition process, during which the starch structure gradually changes from ordered to disordered (Li, Dhital, Gilbert and Gidley, 2020, Li et al., 2023). The pasting properties of different starch samples are shown in Table 1, which demonstrates that 10% KS substitution had no significant effect on the Ptime and PT of the mixed flour (p > 0.05). However, when the KS substitution amount was 15% and 20%, PT and Ptime were significantly reduced, respectively (p < 0.05), which was mainly related to the high protein content of WF (Table S1). On the one hand, proteins can be wrapped on the surface of starch granules to inhibit starch swelling (Sun, Qian, et al., 2023). At the same time, KS reduces the protein content of the mixed flour and weakens the inhibitory effect, thus decreasing PT and Ptime (Li, 2022; Wang et al., 2020). On the other hand, protein may compete with starch for free water during gelatinization, resulting in higher-protein WF being more difficult to gelatinize, so WF has the longest Ptime and the highest PT (Ding et al., 2021; Li, 2022). The above results show that KS is more prone to gelatinization, and the gelatinization process is rapid, indicating that adding a certain proportion of KS in the industrial production of bread can shorten the baking time of the product, which is conducive to saving energy consumption.

In addition, compared with WF, the PV, TV, and FV of the mixed flour after KS substitution increased significantly overall. This could be because of the high viscosity of KS itself (Wang, Lan, et al., 2021) or KS substitution increasing the total starch content of the mixed flour (Table S1). Thakaeng et al. (2021) also reached a similar conclusion by using unripe green banana flour to partially replace wheat flour in making bread. BD reflects the hot paste stability of the mixed flour under high temperature and shear force and the damage degree of starch granules during gelatinization (Wang et al., 2020). The BD values of the mixed flours significantly increased from 517 cP (0%) to 549–578 cP (p < 0.05) when KS substitution ≥15%, suggesting that ≥15% KS substitution caused a decrease in the hot paste stability of the mixed flour and aggravated the damage degree of the starch granules. SB reflects the degree of starch aging. Starch with high AM content helps to form an ordered rigid structure and is often more prone to aging (Biduski et al., 2018). Therefore, the SB of the mixed flour showed a significant, dose-dependent increase (p < 0.05). It also shows that KS has potential application value in producing foods that need moderate aging, such as vermicelli and bean jelly.

3.2. Effects of KS substitution on dough properties

3.2.1. Texture properties

The texture characteristics of dough are important indicators for evaluating the processing characteristics of flour and are of great significance for the processing adaptability of dough. Table 2 shows that compared with WF, the hardness of KF15 and KF20 significantly increased (p < 0.05), mainly because KS increased the RS content in the mixed flour (Wang et al., 2022). Previous study has shown that higher RS content produces harder dough (Guadalupe-Moyano et al., 2022). In addition, KS is a B-type starch with small particles that has stronger WA and stacking ability compared to A-type wheat starch with larger particles, resulting in insufficient hydration of dough gluten and easier formation of tight structures after the addition of KS (Li et al., 2021), thereby increasing the stiffness of the dough. Cohesiveness reflects the resistance of its internal structure, and resistance refers to the ability of the dough to return to its original state after deformation. With the increase of KS substitution, the cohesiveness and resilience of dough first increased and then decreased. This is mainly because when a small amount of KS substitution is used, KS particles fill the pores of the gluten network and enhance the physical interaction between starch and the gluten network, increasing the rigidity and deformation resistance of the dough. However, when the KS substitution amount is too large, the protein content in the dough significantly decreases (Table S1), resulting in a considerable decrease in the integrity of the gluten network structure and the dough's capacity to resist deformation being weakened. After KS substitution, the adhesiveness of the dough considerably decreased, however, there was no significant difference in the various substitution levels (p > 0.05), indicating that KS addition was beneficial for dough shaping.

Table 2.

Effects of KS substitution on the textural properties and mixolab parameters of flours and the baking properties and textural properties of breads.

			WF	KF10	KF15	KF20
Dough	Textural properties	Hardness (g)	178.30 ± 13.02c	199.12 ± 15.19bc	212.05 ± 7.15b	238.99 ± 7.03a
		Adhesiveness	−96.36 ± 67.95a	−296.00 ± 28.66b	−288.97 ± 32.58b	−315.06 ± 14.82b
		Cohesiveness	0.57 ± 0.03b	0.69 ± 0.02a	0.66 ± 0.03ab	0.63 ± 0.04b
		Resilience	0.59 ± 0.04b	0.64 ± 0.02a	0.62 ± 0.02ab	0.57 ± 0.03b
	Mixolab parameters	C1 (N.m)	1.10 ± 0.03a	1.10 ± 0.02a	1.09 ± 0.01a	1.10 ± 0.02a
		C2 (N.m)	0.48 ± 0.00a	0.28 ± 0.01b	0.24 ± 0.01c	0.22 ± 0.01d
		C3 (N.m)	1.52 ± 0.00bc	1.49 ± 0.02c	1.57 ± 0.04ab	1.62 ± 0.01a
		C4 (N.m)	1.37 ± 0.02b	1.43 ± 0.01ab	1.5 ± 0.06a	1.49 ± 0.00a
		TF (min)	5.03 ± 0.28a	2.07 ± 0.69b	1.23 ± 0.57b	0.93 ± 0.13b
		TS (min)	10.60 ± 0.14a	8.45 ± 0.21b	4.90 ± 0.85c	3.65 ± 0.21c
		C1-C2 (N.m)	0.61 ± 0.03c	0.82 ± 0.02b	0.86 ± 0.01ab	0.88 ± 0.01a
Bread	Baking properties	Width/height ratio	0.99 ± 0.01a	1.03 ± 0.02b	1.13 ± 0.03c	1.30 ± 0.04c
		SV (mL/g)	3.71 ± 0.08a	3.59 ± 0.05a	3.29 ± 0.15b	2.67 ± 0.13c
		Dough mass (g)	243.33 ± 0.58a	244.50 ± 0.50a	243.67 ± 1.04a	244.67 ± 1.04a
		Bread mass (g)	219.83 ± 1.04c	221.33 ± 1.44bc	222.50 ± 0.50b	225.00 ± 0.50a
		Baking loss (%)	9.66 ± 0.32a	9.27 ± 0.46ab	8.89 ± 0.28b	8.37 ± 0.68c
		Air-cell ratio (%)	23.97 ± 0.16d	32.21 ± 0.10c	36.26 ± 0.23b	38.74 ± 0.34a
	Textural properties	Hardness (g)	57.83 ± 8.35c	87.51 ± 0.98b	91.80 ± 7.74b	149.58 ± 3.88a
		Springiness	1.78 ± 0.09a	1.37 ± 0.22b	1.36 ± 0.26b	0.98 ± 0.01c
		Cohesiveness	0.81 ± 0.02a	0.73 ± 0.00ab	0.74 ± 0.04ab	0.66 ± 0.08c
		Gumminess	46.80 ± 5.49c	63.85 ± 0.29b	68.27 ± 6.54b	98.03 ± 9.33a
		Chewiness	75.56 ± 2.96c	79.08 ± 1.57c	97.45 ± 0.61b	125.45 ± 3.19a
		Resilience	0.39 ± 0.02a	0.34 ± 0.00ab	0.37 ± 0.03ab	0.32 ± 0.05c

Note: The results are expressed as mean ± standard deviation (n ≥ 3), and there is a significant difference in the values of different letters in the same line (p < 0.05).

3.2.2. Mixing properties

Mixolab can reflect the rheological properties of dough under the dual constraints of mixing and temperature. The specific parameters are shown in Fig. 1C and Table 2. Among them, C1 (N·m) is the target torque during mixing, through which we can obtain the WA of the mixed flour, and C2 (N·m) represents the degree of protein weakening, with smaller values indicating more severe weakening. C3 (N·m) characterizes the PV; C4 (N·m) characterizes the hot paste stability and resistance of starch to enzymatic hydrolysis (Pérez-Rodríguez et al., 2023), with larger values indicating poorer hot paste stability and thermal stability and stronger resistance to enzymatic hydrolysis. C1-C2 (N·m) represents the total protein weakening value of the mixed flour during the mixing process; the larger this value, the smaller the gluten strength of the mixed flour. TF (min) is the time of gluten formation in the dough, and the larger the value, the stronger the gluten strength of the flour. TS (min) is the stability time, and a larger value indicates that the dough is more resistant to stirring and the gluten network is less susceptible to damage.

Fig. 1.

Rheological properties of doughs. (A) Storage (G′) and loss (G′′) moduli as versus frequency (ω); (B) Dynamic loss tangent tanδ versus frequency; (C) Dough mixing curve of different KS substitution amount.

Compared with WF, the C2 value of the mixed flour after KS substitution showed a dose-dependent significant decrease, while TF and TS were significantly reduced (p < 0.05). This was mainly attributed to KS substitution diluting the total protein content of the mixed flour, increasing protein weakening in the system, and decreasing the overall gluten strength of the mixed flour, thereby decreasing the gluten network stability of the dough. The significant increase of the C1-C2 value of the mixed flour after KS substitution also confirmed this. In addition, correlation analysis was conducted on some parameters of the dough's texture properties and mixing characteristics, as well as the bread's baking and texture characteristics, as shown in Fig. 6. The results showed dough hardness was positively related to C1-C2 (r = 0.81, p < 0.01), indicating that the increase in hardness was closely related to the destruction of protein network structure. Increasing KS substitution was associated with an increasing trend in C3, which agreed with the PV results mentioned in Section 3.1.2 above. When the KS substitution ratio increased to over 15%, C4 significantly increased (p < 0.05), which coincided with the significant increase in BD mentioned above, indicating that KS substitution reduced the thermal paste stability of the mixed flour and helped to improve its enzymatic resistance.

Fig. 6.

Correlation analysis of selected parameters for dough and bread. (D) represents the parameters of the doughs (in order to distinguish the texture parameters of dough and bread).

3.2.3. Rheological properties

Characterizing the rheological behavior of dough is crucial for optimizing formulas and improving bread-making (Sun et al., 2020), as shown in Fig. 1A. With the increase of KS substitution, in all samples, the storage modulus G′ and loss modulus G′′ increased with the angular frequency ω. The increase indicates that KS substitution improves viscoelasticity. Moreover, G′ is always greater than G′′, indicating that the elastic characteristics of the dough after KS substitution dominate. This is mainly because KS is a small particle starch that tends to form a denser stacking structure, enhancing the physical filling effect of starch particles in the gluten network, enhancing uniformity and density, and enhancing the overall elasticity of the dough (Guo et al., 2022). The increase in total starch content increases dough elasticity (Amini Khoozani et al., 2020).

The comprehensive viscoelasticity of a dough system can be expressed as tan δ (G′′/G′). According to Fig. 1B, the tan(δ) of all dough samples was <1; all dough samples exhibited greater elasticity than viscosity, exhibiting typical solid-like characteristics. Thus, KS substitution somewhat improved dough viscoelasticity, but the original solid-like state of the dough was unchanged.

3.2.4. Scanning electron microscopy

The morphological structure of wheat flour and KS were observed to assess changes in the microstructure of the dough after replacing wheat flour with KS. Fig. 2A is the microstructure of wheat flour, which is mainly composed of large A-type wheat starch particles (green circles), small B-type wheat starch particles (purple circles), and gluten (yellow circles) (Li, Gong, et al., 2020). From Fig. 2B, it can be observed that KS exhibited irregular polygons with smooth surfaces, and partial particles were damaged, similar observations were made by Wang, Lan, et al. (2021) in the SEM of KS.

Fig. 2.

SEM micrograph (×3000) of (A) wheat flour and (B) kiwi starch; (C), (D), (E) and (F) are the SEM micrograph (×1000) of WF, KF10, KF15 and KF20 doughs, respectively; (G), (H), and (I) indicate the appearance and cross-sectional images of WF, KF10, KF15 and KF20, and the bread's cross-sectional images processed by Image J (used to calculate the air cell ratio).

Table S1 shows that KS has lower protein and higher starch content. Therefore, the gluten network and starch concentration of the dough replaced by KS will change accordingly, causing the concentration of gluten to drop and influencing the dough's mechanical strength (Grassi de Alcântara et al., 2020). Figs. 2C–F show the SEM microstructure of different dough samples. As shown in Fig. 2C, the gluten network structure of the WF group was continuous and dense. The starch particles were well encapsulated, with few pore structures and varying sizes of pores (red circles), consistent with previous reports on the microstructure of pure wheat dough (Sun, Bu, et al., 2023; Wang, Lao, et al., 2021). However, the gluten protein content in the dough was reduced after the 10% KS substitution, and the gluten network showed partial breakage (Fig. 2D, yellow circle). In addition, due to the filling of small particles of KS in the gluten network, it could be observed that the pores between the gluten networks decreased and the number increased (red circle). As the KS substitution amount increased to 15%, the gluten network further fractured (Fig. 2E, yellow circle), and the pores between the networks were further reduced; with some pore depths decreasing, KS starch particles began to be exposed in the gluten network. When the substitution amount increased to 20%, the relatively loose gluten network structure could no longer accommodate the increasing number of KS particles, and the gluten network almost collapsed on a large area. Most starch particles were exposed on the surface of the network structure, and the pores became smaller and more widely distributed, and the depth of the pores (Fig. 2F, red circle) significantly decreased. Previous study has indicated that the addition of B-type wheat starch to wheat dough resulted in the B-type particles filling the pores in the gluten network and reducing their quantity (Guo et al., 2022), consistent with our findings.

3.3. Product quality

3.3.1. The baking properties of bread

Table 2 lists the bread's baking characteristics. Compared with WF, KS bread width/height decreased significantly, and the air-cell ratio increased significantly from 23.97% to 38.74% (p < 0.05). This trend is also clear from the appearance and cross-sectional view of the bread in Figs. 2G–I. The WF bread had a loose texture and uneven distribution of air cells, while the KS bread was dense and uniformly porous. Increasing KS substitution significantly reduced bread height and volume, making the air cells denser and more uniform.

The SV of bread decreased significantly from 3.71 to 2.67 cm³/g (p < 0.05) when at ≥15% KS substitution. Moreno-Araiza et al. (2023) also observed a reduction in SV when substituting 20% green pea flour for wheat flour. KS has a lower pH (Table S1) and may thus inhibit yeast fermentation at higher substitution levels. The low protein content of KS may also dilute the gluten protein content, thereby reducing SV. Therefore, yeast amount and fermentation time should be optimized to achieve the ideal SV. Similarly, the baking loss after KS substitution showed a similar trend. Analysis of correlations revealed a positive correlation between baking loss and SV (Fig. 6, r = 0.93, p < 0.01). The surface area accessible for evaporating water increased with increasing SV, leading to greater water loss during baking (Gidari- Gounaridou et al., 2023). Within the substitution range of 10–20%, there was no significant variation in dough mass (p > 0.05), while baking loss decreased significantly (p < 0.05). Therefore, the bread mass generally increased after KS substitution, verifying that KS substitution improves bread yield and production efficiency.

3.3.2. The textural properties of bread crumbs

The texture of the bread is an important quality indicator and factor in consumer acceptance. The textural characteristics of different breads are shown in Table 2. Compared with WF, KS substitution significantly increased hardness, gumminess, and chewiness and reduced springiness (p < 0.05). The cohesiveness and resilience of the KF group showed a gradually decreasing trend; at 20% KS substitution, the cohesiveness and resilience of the bread were significantly reduced (p < 0.05). Chen et al. (2019) obtained similar results by adding mango peel powder. The filling effect of small starch granules and the dilution of the gluten network caused the increase in bread hardness, as described in Section 3.2.4. The higher chewiness and gumminess scores of KF15 and KF20 breads indicated that chewing and swallowing consume more energy. The significantly reduced cohesiveness and resilience of KF20 implied that the bread was more prone to fracture and crumble. It was difficult to restore its original shape when subjected to external pressure, which may reduce the sensory experience when consumed.

Hardness was significantly positively correlated with the air-cell ratio (r = 0.88, p < 0.01) and negatively correlated with SV (r = −0.94, p < 0.01), verifying that the change in hardness was closely related to the air-cell ratio and SV of bread (Fig. 6). Notably, there was a positive correlation (r = 0.88, p < 0.01) between the hardness of the bread and the dough. Thus, the texture of the bread is greatly influenced by the properties of the dough throughout processing.

3.3.3. The color characteristics of crust and crumb

Bread color has the most direct and intuitive effect on consumers. KS substitution affected color, and the top, crust, and bottom significantly differed from WF (Figs. 2G, H, and 3B). Compared with WF bread, the ΔE of the crumb in the KF groups were > 3, indicating a visible change in crumb color associated with KS substitution (Fig. 3C). KS substitution caused the top color of the bread to be lighter, while the bottom color deepened (Fig. 3B). At 10% KS substitution, the SV of the bread did not change significantly and had little effect on the heating and coloring of the top. Thus, the top of the KF10 bread was evenly colored. At 15% substitution, the bread color became lighter and was uneven due to the significant decrease in SV and the uneven contact between the top of the bread and the top mold. The Maillard and caramelization reactions mediate bread coloration during baking. The deepening color at the bottom of the bread is mainly due to the yellow-green color of KS and the increased content of damaged starch, which yields more sugar that confers brown pigmentation via the Maillard and caramelization reactions (Li, Tilley, et al., 2023). Due to the yellow-green color of KS, increasing substitution levels lead to increasing b*, C*, and h° and decreasing L* in the breadcrumb (Fig. 3B and C), indicating that KS substitution shifts the crumb color to yellow-green, this color shift in the crumb after KS substitution may make it attractive to consumers (Kelanne et al., 2024).

Fig. 3.

Color and sensory characteristics of breads. (A) Overall chromaticity distribution diagram; (B) Local amplification of chrominance distribution, significant differences in a* are marked by various lowercase letters, whereas significant differences in b* are marked by distinct uppercase letters (p < 0.05), and the more scattered the dots, the more uneven the color; conversely, the more concentrated the dots, the more uniform the color; (C) L*, C*, h°, ΔE of crumbs; (D) Rose diagram of artificial sensory evaluation, in which there is a statistically significant difference in the numerical values of different letters on the column under the same indicator (p < 0.05). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3.3.4. Sensory evaluation

An artificial sensory evaluation experiment was carried out to investigate the effects of replacing wheat flour with KS on the sensory properties of bread (Fig. 3D). The results revealed no significant difference in color, appearance, texture, and taste scores between KF10 and WF bread (p > 0.05). Flavor analysis suggested that 10–15% KS substitution gave the bread a unique kiwi fruit flavor, enriching the flavor of the bread and yielding a flavor score significantly higher than that of WF (p < 0.05). However, at 20% KS substitution, the fruit aroma was too strong, overpowering the original wheat bread flavor and reducing its flavor score. Compared to the other breads, KF20 bread had a substantially lower texture score (p < 0.05), mainly because the bread became hard and crumbled when eaten at this substitution level. This is consistent with the significantly increased hardness and decreased cohesiveness associated with KF20 in Section 3.3.2. Overall, at ≥15% KS substitution, the color, appearance, taste, and overall acceptability of bread are significantly reduced (p < 0.05), primarily due to the reduction of SV, uneven color, and tendency to crumble, all of which affect the evaluator's eating experience. In summary, overall acceptability was highest for KF10 (8.50 points), which maintained WF's original appearance, texture, and taste while endowing the bread with a unique fruity aroma and attractive yellow-green tones, earning unanimous praise from the evaluators.

3.4. In vitro starch digestion

Starch is the primary ingredient in bread and is generally classified as rapidly digested (RDS), slowly digestible (SDS), or resistant (RS) by its rate of enzymatic hydrolytic digestion. RDS is quickly digested and absorbed by the human body, rapidly increasing blood sugar. In contrast, in the small intestine, SDS is slowly digested, which helps prevent cardiovascular disease (Cui et al., 2022). RS can enhance insulin sensitivity, reduce blood sugar levels, and improve intestinal health. The in vitro digestive effect of KS substitution in bread significantly impacted the rate of starch hydrolysis, eGI, and starch composition.

The hydrolysis curves of all bread samples showed typical exponential growth, with rapid enzymatic hydrolysis within the first 40 min and slow enzymatic hydrolysis after 120 min reaching a maximum at 180 min, indicating that digestion followed first-order kinetics (Fig. 4A). Compared with WF, 10–20% KS substitution reduced the rate of starch hydrolysis, which was confirmed by the significant reduction of C_∞ and HI (p < 0.05) after KS substitution (Fig. 4C). This may be because KS is rich in polyphenols and dietary fiber (Wang, Lan, et al., 2021), which inhibit the activity of starch-digesting enzymes. This is consistent with the conclusion reached in Mixolab that the KS substitution improves the enzymatic resistance of the system. Wang, Lao, et al. (2021) added whole quinoa flour to bread also slowed down the rate of starch hydrolysis. As shown in Fig. 4B, compared with WF, the RDS associated with various levels of KS substitution significantly decreased from 24.99% to 19.63–21.32%, while RS significantly increased by 4.62-15.32% in a dose-dependent trend (p < 0.05), indicating that KS-substituted bread was not easily digestible and can prevent rapid increases in blood glucose levels. RS is highly resistant to enzymatic hydrolysis and is not digestible or absorbed in the small intestine (Jiali et al., 2023), which is also a critical factor in decreasing starch hydrolysis rate of bread. According to Fig. 4B, the eGI of the WF group was 78.18, classifying it as a high-GI food, while the eGI of the KF15 and KF20 groups was 68.19 and 63.45, respectively, classifying them as medium GI (Zhao et al., 2023). Therefore, supplementation with KS can improve bread by inhibiting starch digestion, thereby maintaining stable blood sugar levels. These results indicate that KS is a promising raw material for low-GI food development.

Fig. 4.

In vitro digestibility of bread samples. (A) Starch hydrolysis rate (CK: white bread); (B) Starch composition (RDS, SDS, RS); (C) HI, eGI, C∞ of bread samples.

3.5. Storage characteristics

3.5.1. Shelf-life evaluation

Bread provides an ideal environment for mold because of its comparatively high moisture content. At room temperature, fresh bread without preservatives has a short shelf life of a few days; therefore, we explored the growth of mold in bread stored at 25°C for 6 days. At 4 days, visible colonies appeared in the WF bread, but none were apparent in the KS bread (Fig. 5A). By day 6, colonies of varying sizes were apparent on the KS breads, but still significantly less than on the WF breads. The slightly acidic pH (5.0–6.0) and high water activity (0.95–0.97) make bread susceptible to mold contamination (El Houssni et al., 2023). Previous studies have shown that KS has the characteristic of low pH (3.43–4. 28) (Lan et al., 2024); therefore, adding KS with low pH (Table S1) reduced the bread's pH, inhibiting mold growth and reproduction. Similarly, Dopazo et al. (2023) also found that adding fermented whey reduced the pH value of the bread and extended its shelf life. Galanakis et al. (2018) found that olive polyphenols and other natural antioxidants (such as α-tocopherol and ascorbic acid) as antibacterial agents could prolong the shelf life of the bread. Therefore, KS rich in polyphenols and antioxidants (Wang et al., 2022) may help prolong the shelf life of bread at 25°C. In summary, KS substitution inhibits mold growth and reproduction and extends the shelf life of bread at 25°C, suggesting KS is a potential clean-label ingredient with antimicrobial properties.

Fig. 5.

Changes in (A) microbiological shelf-life of bread samples, (B) x-ray diffractograms, (C) moisture content, and (D) starch crystallinity. Significant differences between different breads at the same storage duration are shown by values with different lowercase letters (p < 0.05). Significant changes between the same bread at different storage durations are shown by values with different uppercase letters (p < 0.05).

3.5.2. Moisture

Given that mold colonies were present on all bread samples by day 6, subsequent experiments only monitored moisture changes over 4 days of storage at 25°C (Fig. 5B). Over time, the moisture of bread in all groups significantly decreased (p < 0.05). On days 0, 2, and 4, the KS bread's moisture content was higher than that of the WF bread, suggesting that the addition of KS improved water retention.

3.5.3. X-ray diffraction

Bread is characterized by aging during cooling and storage. In addition to water migration, the main cause of aging in bread is starch recrystallization. Fig. 5D shows the changes in the XRD spectrum of bread stored at 25°C for 0, 2, and 4 days. B-type crystals (with characteristic peaks at 2θ = 17°) are usually generated by the formation of a double helix between AM and amylopectin chains during starch storage (Kang et al., 2018), while the V-shaped crystal of the amylose-lipid complex (with characteristic peaks at 2θ = 19.5°) is developed during baking and cooling (Wang et al., 2023). On day 0, there was no obvious characteristic peak at 2θ = 17° in WF and KS breads, but there were obvious characteristic peaks at 2θ = 19.5°, indicating that the aging of fresh bread mainly occurs during baking and cooling. With increasing storage time, the peak intensity at 2θ = 17° gradually increased for all bread types but was greater for KF than WF. Peak intensity at 2θ = 19.5° had no significant changes, revealing that the development of B-type crystals was the primary cause of aging during storage, which was aggravated by the addition of KS. Similar to our findings, Kang et al. (2018) also found that bread containing different hydrocolloids and modified starches produced a distinct characteristic peak near 2θ = 19.5° on day 0 that did not change significantly during storage.

The crystallinity of all bread types increased significantly with storage duration at 25°C, but the increase was greater at higher substitution levels, consistent with the SB and XRD findings (p < 0.05) (Fig. 5D). This difference is attributable to the KS-mediated increase in AM content. AM molecules are a straight chain structure with small branches, which reduce their spatial distribution and facilitate orientation such that bread with high AM content is more prone to aging (Corrado et al., 2023). Conversely, due to the high WA of KS, its water retention supports starch degradation (Arp et al., 2020), thus exacerbating aging.

4. Conclusion

This study systematically investigated the effects of 10–20% substitution on the quality and storage characteristics of the bread. The results showed that KS substitution significantly enhanced the water binding capacity of the mixed dry ingredients and significantly reduced the WSI. At ≥15% KS substitution, the heat stability of the dough was reduced, and damage to the starch granules was aggravated, making them more prone to gelatinization and reducing processing energy consumption. KS substitution weakened the gluten network, interrupting the contiguous network structure of the dough, thereby affecting its processing performance and improving the quality of the finished bread. Increasing amounts of KS substitution caused the air cells of the bread to become denser and more uniform. A 15% substitution significantly reduced the SV and baking loss while increasing mass, indicating that adding KS improved productivity. The KS substitution improved the Maillard reaction during baking, which gave the bread a unique fruity flavor and gradually shifted the crumb to a yellowish green, which was more attractive to consumers. The KF10 group had the highest sensory scores. In addition, 10–20% KS substitution inhibited starch digestibility, changing high-GI bread into medium-GI bread and better supporting postprandial blood glucose stability. Moreover, KS substitution improved the water-holding capacity of the bread, inhibited mold growth and reproduction, and extended the shelf life at 25°C. However, the bread was more prone to aging. KS is a potentially useful material for developing low-GI, clean-label food. Future research should focus on process optimization and formulation improvements to obtain better breads. The results presented here provide a theoretical basis for developing low-GI bread and applying KS in other products.

CRediT authorship contribution statement

Zhenyun She: Writing – review & editing, Writing – original draft, Resources, Methodology, Investigation, Conceptualization. Qinyu Zhao: Writing – original draft, Software, Resources, Methodology, Investigation. Danting Hou: Project administration, Methodology, Investigation. Jiaqi Wang: Writing – review & editing, Software, Resources, Methodology. Tian Lan: Writing – review & editing, Methodology, Investigation. Xiangyu Sun: Writing – review & editing, Investigation, Funding acquisition, Conceptualization. Tingting Ma: Writing – review & editing, Resources, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material: Table S1. Chemical composition of kiwi starch (KS) and wheat flour (WF).

Data availability

Data will be made available on request.