Modified Okara improves Noodle Properties: Effects on digestibility, structure, and functionality

Abstract

This study developed a modified okara (the product of co-fermentation with Pleurotus ostreatus and Saccharomyces cerevisiae followed by cellulase treatment), evaluating its effects on noodle quality. Incorporation of 40 % modified okara significantly altered starch digestibility, reducing rapidly digestible starch (RDS) from 64 % to 50 %, while increasing slowly digestible starch (SDS) and resistant starch (RS) from 21 % and 6 % to 25 % and 14 %, respectively, which indicates a shift toward slower glucose release, beneficial for glycemic control. Structural characterization by FTIR, and SEM showed that okara promoted starch-protein-lipid complex formation and created porous networks that effectively encapsulated starch granules. These microstructural modifications were functionally significant, as dynamic rheological confirmed improved dough elasticity (G′ increased) at low okara concentrations. Additionally, soluble dietary fiber (SDF) content and ABTS radical scavenging ability increased to 9 % and 55 %, respectively, linked to liberated phenolics. These findings propose okara as a sustainable, functional ingredient for low-glycemic-index noodles, as it repurposes agro-industrial byproducts.

Graphical abstract

Highlights

• •Triple modification technology enhanced okara's functionality.
• •Reduced rapidly digestible starch via starch-protein-lipid complexes.
• •Modified okara improved the cooking characteristics of noodles.
• •Microporous okara structure entrapped starch granules, limiting enzyme access.

1. Introduction

Noodles, a staple food in Asia, exhibit a high glycemic index (GI) due to the predominance of rapidly digestible starch (RDS, 70–90 %) in refined wheat flour (Puligundla and Lim, 2021). This rapid digestibility leads to postprandial glycemic excursions, increasing the risk of metabolic syndrome. Functional ingredients such as soybean flour (Rani et al., 2019) and tartary buckwheat flour (Li et al., 2022) have been incorporated to modulate starch digestion. Yet, challenges remain in balancing processing adaptability and nutritional retention.

Okara, a byproduct of soybean processing, possesses unique nutritional-functional attributes owing to its high dietary fiber (38–60 % DW) and protein content (20–30 % DW). Recent studies have demonstrated its successful incorporation into various food matrices: development of gluten-free baked goods (Pesic et al., 2023), fiber-enrichment agent in yogurt (Tian et al., 2024), and substrate for fermentation-derived bioactive peptides (Shen et al., 2023). In addition, studies indicate that okara's dietary fiber physically encapsulates starch, reducing enzymatic accessibility (Kim et al., 2020), while protein-starch complexes promote resistant starch (RS) formation (Abdel-Mobdy et al., 2021). However, unmodified okara's high insoluble fiber, and free sulfhydryl groups impair gluten development by disrupting disulfide crosslinking, reducing its functional compatibility in wheat-based dough systems.

Conventional modification methods (chemical treatment, thermal processing, or single-strain fermentation) (Lin et al., 2020, Lin et al., 2020; Li et al., 2019) face critical limitations: chemical residues, heat-induced nutrient degradation, and inefficient lignin breakdown by yeast. Herein, we propose a triple-modification strategy: Edible mushroom fermentation: Pleurotus ostreatus secretes lignin peroxidase to degrade cellulose while generating β-glucans (Mutukwa et al., 2019), Yeast fermentation: Saccharomyces cerevisiae converts free sugars into flavor esters (Niçin et al., 2022), mitigating off-flavors, Enzymatic hydrolysis: Cellulase targets β-1,4-glycosidic bonds, thereby converting insoluble dietary fiber (IDF) to soluble dietary fiber (SDF), contributing to improved dough rheology.

By incorporating modified okara into noodles, this study systematically evaluates its impacts on nutritional functionality (dietary fiber, antioxidant capacity), processing properties (rheology, cooking loss), and structural characteristics (microstructure, protein secondary structure) using digestion assays, rotational rheometer, SEM, and FTIR. This approach not only advances low-GI noodle development but also promotes high-value utilization of okara, aligning with health and sustainability goals.

2. Materials and methods

2.1. Experimental materials

Soybeans were purchased from Heli Supermarket (China). Wheat flour, gluten enhancer (Jinliyuan), vital wheat gluten, and food-grade salt were obtained from Huimin Fresh Supermarket (Guiyang, Guizhou, China). Pleurotus ostreatus (Gaofeng 246) was provided by the Edible Fungi Research Institute in Xishui County (Guizhou, China), and Saccharomyces cerevisiae was sourced from Angel Yeast Co., Ltd. The solid medium (PDA) formulation (g/L) consisted of 25 g potato dextrose broth (PDB), 3 g KH₂PO₄, 1.5 g MgSO₄·7H₂O, 20 g agar, 0.08 g vitamin B1, and 2 g peptone, with unadjusted pH (natural pH). The liquid seed medium (g/L) comprised 25 g PDB, 3 g KH₂PO₄, 1.5 g MgSO₄·7H₂O, and 2 g peptone, with unadjusted pH (natural pH).

2.2. Fermentation and enzymolysis of okara

Fermentation and enzymolysis of okara according to the methods adapted from Xie et al. (2025) and Liu et al. (2024). Commercial soybeans were weighed, washed, and soaked in water at a 1:4 (w/v) ratio for 12 h at room temperature. The soaked soybeans were blended with water at 1:7 (w/v) and filtered to yield okara containing approximately 88 % moisture. The conditions for fermenting okara with Pleurotus ostreatus involved inoculating with 7.4 % Pleurotus ostreatus, adding 0.26 % sucrose, and fermenting at 27 °C for 7 days. The conditions for fermenting Okara with Saccharomyces cerevisiae included adjusting the moisture content to 95 % and adding food-grade potassium dihydrogen phosphate (0.12 %), magnesium sulfate (0.07 %), and sucrose (0.9 %), with inoculation of 2 % Saccharomyces cerevisiae (seed liquid concentration of 1 × 10⁸ CFU/mL), fermenting at 27 °C and 170 rpm for 24 h. The co-fermentation system was established by mixing Pleurotus ostreatus, and Saccharomyces cerevisiae fermented okara at 1:3.5 (w/v, 1 part P. ostreatus okara, and 3.5 parts S. cerevisiae okara), followed by static fermentation at 27 °C for 24 h (without agitation). Enzymolysis of Okara utilized food-grade cellulase on co-fermented okara, with enzymolysis performed for 2 h at 45 °C and pH 4.0, supplemented with 3.5 % cellulase.

2.3. Noodle making

Noodles were prepared using the Xie et al. (2023) method. Base formula: Wheat flour supplemented with 0.3 % gluten enhancer, 6 % vital wheat gluten, and 1 % food-grade salt. Experimental groups: Modified okara (The water content was adjusted to account for okara's moisture to maintain consistent dough hydration.) replaced 0 % (control), 30 %, 35 %, and 40 % of wheat flour (w/w, based on total dough mass including added water and ingredients). Wheat flour and okara were mixed with other ingredients to form a homogeneous dough. The dough was proofed at 35 °C and 75 % relative humidity for 30 min. The proofed dough was sheeted through a noodle machine to obtain uniform sheets. Sheets were cut into noodles (2.0 mm width, 25.0 cm long) and air-dried at 30 °C and 50 % RH for 2 h, producing the final noodle product.

2.4. Determination of antioxidant capacity

The ABTS radical scavenging activity of the noodles was assessed following a method adapted from Hu et al. (2022). A 0.50 g sample was mixed with 10 mL of 80 % methanol and subjected to room temperature at 150 rpm for 2 h, followed by 30 min of ultrasound-assisted extraction. The supernatant was collected after centrifuging the mixture at 6790×g for 15 min. A solution containing 7.00 mmol/L ABTS and 2.45 mmol/L potassium persulfate was prepared and left to react in the dark at room temperature for over 16 h. The absorbance was adjusted to 0.70 ± 0.02 at 734 nm using ethanol absolute. Subsequently, 2.00 mL of the supernatant was combined with 3.00 mL of ABTS solution incubated in darkness at room temperature for 30 min, and the absorbance was measured at 734 nm. The ABTS radical scavenging activity of the noodles was quantified using the following formula (1):

$$\begin{array}{c}ABTS(\%)=\frac{A1-A2}{A1}\times 100\%\end{array}$$ (1)

• A1: The absorbance without added ABTS solution,

• A2: The absorbance of sample and ABTS solution.

2.5. Determination of dietary fiber

The dietary fiber content was determined following the AOAC Official Method 991.43 (Aoac, 1998). Initially, samples were defatted with petroleum ether and desugared with 85 % ethanol. After these pretreatment steps, samples were oven-dried at 55 °C for 24 h to remove residual solvents before enzymatic digestion. Subsequently, the sample and enzyme were dissolved in 0.05 mol/L MES-TRIS buffer and sequential enzymatic treatment under the following conditions: high-temperature α-amylase at 95–100 °C for 35 min for starch hydrolysis, glucoamylase at 60 °C for 30 min for further starch degradation, and alkaline protease at 60 °C for 30 min for protein removal. The resulting mixture was centrifuged at 10610×g for 15 min, and the resulting precipitate was dried to a constant weight to obtain insoluble dietary fiber (IDF). The supernatant was combined with a four-fold volume of 95 % ethanol and left overnight. After centrifugating at 10610×g for 15 min, the precipitate was dried to a constant weight to obtain soluble dietary fiber (SDF).

2.6. Determination of in vitro starch digestion characteristics

The noodles' in vitro starch digestion characteristics were assessed using a modified Englyst method (Englyst and Cummings, 1985). One gram of porcine pancreatic amylase (14000 U/g, Yuanye Biotech, CAS#9000-90-2), which was dissolved in 100 mL of deionized water, stirred at 200 rpm, and centrifuged at 6790×g for 10 min to collect the supernatant. This enzyme solution was combined with 100 μL 100000 U/mL glucoamylase (100000 U/g, Ruiyang Biotech, CAS: 9032-08-0) to create the enzyme mixture, which was immediately used. A 0.500 g sample of ground noodles was mixed with 15 mL of sodium acetate buffer (0.2 mol/L, pH 5.2, pH not adjusted after sample preparation), vigorously shaken, and supplemented with 10 glass beads (4 mm diameter) to prevent aggregation. After a 30-min incubation at 37 °C with shaking at 150 rpm, 10 mL of the enzyme mixture was added and continuously stirred at 150 rpm. Hydrolysate samples of 1 mL each were withdrawn at 0 min, 20 min, 30 min, 60 min, 90 min, 120 min, 150 min, and 180 min and treated with 2 mL of ethanol absolute to halt enzymatic activity. This sampling scheme was designed to provide high temporal resolution during the initial rapid phase of hydrolysis (0 min–20 min), monitor the transition period (20 min–120 min) when reaction rates typically stabilize, and verify completion of the reaction during the final stage (120 min–180 min). To ensure accurate modeling of starch hydrolysis kinetics, eight strategically distributed time points were selected to capture dynamic changes in digestibility profiles while maintaining analytical precision. The reaction mixtures were centrifuged at 3000×g for 5 min at 4 °C, and the supernatants were analyzed directly for immediate measurements. Glucose levels at each time point were measured using the DNS method. Finally, the concentrations of rapidly digestible starch (RDS), slowly digestible starch (SDS), and resistant starch (RS) were determined using established calculations (2)–(4).

$$\begin{array}{c}RDS(\%)=\frac{G20\times 0.9}{TS}\times 100\end{array}$$ (2)

$$\begin{array}{c}SDS(\%)=\frac{(G120-G20)\times 0.9}{TS}\times 100\end{array}$$ (3)

$$\begin{array}{c}RS(\%)=\frac{TS-RDS-SDS}{TS}\end{array}$$ (4)

• G20: Glucose content at 20 min amylase hydrolysis time, mg,

• G120: Glucose content at 120 min amylase hydrolysis time, mg,

• TS: Mass of total starch in the sample, mg.

2.7. In vitro hydrolysis index and predicted glycemic index

The starch hydrolysis kinetics and predicted glycemic index (pGI) of noodles were analyzed according to the method of Goñi et al. (1997). Starch hydrolysis kinetics were calculated using the following formula (5).

$$\begin{array}{c}C=C^{\infty }\left(1-e^{-kt}\right)\end{array}$$ (5)

• C: Starch hydrolysis rate of samples at t min,

• C^∞: The maximum hydrolysis rate of samples at t min,

• k: The kinetic parameters of samples at t min.

The area under the hydrolysis curve of the noodle samples and the reference food white bread was fitted using Origin. The hydrolysis and predicted glycemic index (pGI) were calculated using the following formulas: (6) and (7).

$$\begin{array}{c}Hydrolysisindex\left(HI\right)=\frac{Areaunderhydrolysiscurveofsample}{\left(Referencefood\right)Areaunderhydrolysiscurveofwhitebread}\end{array}$$ (6)

$$\begin{array}{c}pGI=39.71+0.549HI\end{array}$$ (7)

2.8. Rheological properties

The rheological properties of okara dough were assessed based on the report from Guo et al. (2022). A 20 mm flat plate was employed as the fixture, with a gap of 1000 μm. The dough's storage modulus (G′), loss modulus (G″), and tanδ of the dough were measured within the linear viscoelastic region. Testing was conducted at a strain of 1 %, with the temperature maintained at 25 °C and the frequency ranging from 0.1 to 20 Hz.

2.9. Determination of cooking characteristics

Determination of noodle elongation: Fifteen noodle strands (20 cm long) were cooked in 500 mL boiling water until no white core remained. To standardize the measurement conditions while minimizing texture alteration, the cooked noodles were immediately transferred to cold water for 5 s to arrest further cooking before measuring their lengths. The elongation was calculated using the following formula (8).

$$\begin{array}{c}Elongation(\%)=\frac{cookednoodlelength-20cm}{20cm}\times 100\%\end{array}$$ (8)

Determination of noodle cooking loss: Noodles of equal mass were placed in 500 mL boiling water and cooked until no white core remained. Cooked noodles were removed, and the remaining cooking soup was further heated until reduced to 30 mL and dried in an air oven at 105 °C until constant weight was achieved (typically 24 ± 2 h). Constant weight was defined as less than 0.5 % mass variation between consecutive measurements. The calculation formula (9) for cooking loss was as follows.

$$\begin{array}{c}Cookingloss(\%)=\frac{Drymattermassofsoupafterdrying}{Rawnoodlemass}\times 100\%\end{array}$$ (9)

Determination of noodle water absorption: Noodles of equal mass were placed in 500 mL boiling water and cooked until no white core remained. Cooked noodles were immediately transferred to cold water for 5 s and weighed. The water absorption was calculated using the following formula (10).

$$\begin{array}{c}Waterabsorption(\%)=\frac{cookednoodleweight-rawnoodleweight}{rawnoodleweight}\times 100\%\end{array}$$ (10)

Determination of noodle breakage rate: Fifteen noodle strands (same length) were taken and cooked until no white core remained. Cooked noodles were removed, and the number of intact noodles was recorded. The breakage rate was calculated using the following formula (11).

$$\begin{array}{c}Breakingrate(\%)=\frac{15-numberofcompletenoodles}{15}\times 100\%\end{array}$$ (11)

2.10. Determination of noodle color

The color of noodles was measured using the method proposed by Xing et al. (2023). The L∗, a∗, and b∗ were measured using a calibrated colorimeter. Six parallel measurements were conducted, and the average value was determined.

2.11. Determination of noodle texture

The texture of the noodles was assessed with minor adjustments to the approach described by Ge et al. (2020). Three noodles, cooked until no white core remained, were placed on the sample platform of a texture analyzer. Texture Profile Analysis (TPA) was performed using a TA44 probe. Experimental parameters consisted of a pre-test speed of 10 mm/s, a test speed of 1 mm/s, a 5-s compression time for two cycles, a compression distance of 90 % of the noodle thickness, and a trigger force set at 10 g. Each measurement was replicated four times, and averages were calculated.

2.12. Determination of microstructure

The SEM was determined according to the method of Hong et al. (2024). Noodles were fixed with 2.5 % pentanediol for 2 h, washed with ethanol of varying concentrations, soaked in tert-butanol, and frozen in a refrigerator. After freezing, samples were dried using a freeze-dryer. The dried raw and cooked noodle samples were gold-coated on an aluminum plate with double-sided adhesive and observed under scanning electron microscopy (SEM) at 20.0 kV voltage. Representative images were captured at 1000 × and 5000× magnifications.

2.13. Determination of FTIR

The sieved samples were mixed thoroughly with a potassium bromide solution, and spectra were collected from 4000 to 400 cm⁻¹ at a resolution of 4 cm⁻¹, with each sample scanned 64 times (Shen et al., 2023). All spectra underwent baseline correction and vector normalization before spectral analysis to minimize scattering effects. Absorption peak positions in the amide I band (1600-1700 cm⁻¹) were determined using Fourier self-deconvolution and second derivative methods. Peak areas corresponding to different secondary structures were identified: 1610-1640 cm⁻¹ for intermolecular and intramolecular β-sheet, 1640-1650 cm⁻¹ for random coil, 1650-1660 cm⁻¹ for α-helix, and 1660-1700 cm⁻¹ for β-turn.

2.14. Determination of free sulfhydryl

The concentration of free sulfhydryl in the samples was determined with modifications to the method outlined by Gao et al. (2022). Initially, 1.25 g of sample was mixed thoroughly with 20 mL of Urea-Tris-HCl buffer solution (pH 8.0, comprising 6 mol/L urea, 2 mmol/L EDTA, and 1 % SDS), agitated at 25 °C for 1 h, and centrifuged at 10610×g for 10 min. Subsequently, 4 mL of the resulting supernatant was combined with 0.1 mL of buffer solution (10 mmol/L DTNB, 0.2 mol/L Tris-HCl, pH 8.0) and allowed to react in darkness at room temperature for 20 min. Absorbance at 412 nm was measured following calibration with the buffer solution, and a standard curve was established using L-cysteine as the reference standard.

2.15. Statistical analysis

The data were analyzed using SPSS 27.0 and Excel 2016, and results were reported as mean ± standard deviation. Statistical differences were assessed via one-way ANOVA (Tukey's multiple range test), with significance denoted by different letters (P < 0.05). PeakFit 4.12 was employed to model the secondary structure of proteins. Each indicator was measured in triplicate to ensure data accuracy.

3. Results and discussion

3.1. Functional enhancement

3.1.1. The antioxidant capacity and dietary fiber of noodles

Noodles' ABTS radical scavenging ability increased significantly with okara addition (Fig. 1A–B). At 40 % substitution, raw and cooked noodles achieved 44.0 % and 55.5 % scavenging rates, respectively, showing significant improvement versus control (P < 0.05). This antioxidant improvement correlated with marked changes in dietary fiber composition: soluble dietary fiber (SDF) increased from 2.4 % to 9.2 % in raw noodles and from 1.1 % to 9.4 % in cooked noodles (P < 0.05), while insoluble dietary fiber (IDF) rose from 9.7 % to 22.7 % and 7.1 %–17.3 % respectively (Fig. 1A–B). Thermal processing elevated soluble dietary fiber content (40 %), likely through the combined effects of cell wall degradation and phenolic release. The fermentation-enzymolysis pretreatment disrupted okara's cellular matrix, liberating bound polyphenols (particularly ferulic acid and sinapic acid derivatives) from fiber-polyphenol complexes (Zheng et al., 2024). During cooking, these released phenolics underwent thermal conversion (phenolic glycosides to free acids), while dietary fibers facilitated their homogeneous distribution, enhancing antioxidant bioavailability (Xu et al., 2020). This mechanistic framework aligns with established correlations between dietary fiber-bound phenolics and antioxidant capacity in plant matrices (Wang et al., 2021; Wu et al., 2023).

Fig. 1.

Changes in (A) raw noodles' ABTS radical scavenging activity, insoluble dietary fiber (IDF) and soluble dietary fiber (SDF), (B) cooked noodles' ABTS radical scavenging activity, insoluble dietary fiber (IDF) and soluble dietary fiber (SDF), (C) rapidly digestible starch (RDS), slowly digestible starch (SDS), and resistant starch (RS), and (D) starch hydrolysis curves and fit curves of noodles with 0 %, 30 %, 35 %, and 40 % okara. Means with different lowercase letters on the same type bars represented significant differences (P < 0.05).

3.1.2. The in vitro starch digestion characteristics of noodles

As the primary constituent of wheat noodles, starch digestibility significantly determines their nutritional quality. The incorporation of modified okara substantially altered starch digestibility profiles, reducing rapidly digestible starch (RDS) from 63.8 % to 49.8 % (P < 0.05) while increasing slowly digestible starch (SDS) and resistant starch (RS) from 21.0 % to 25.5 % and from 5.6 % to 13.6 %, respectively (Fig. 1C). The maximum SDS (25.5 %) and RS (13.6 %) contents were achieved at 40 % okara substitution, which nutritionally translates to delayed postprandial glucose response and lowered insulin levels (Zhou et al., 2019). This improvement stems from three synergistic mechanisms: enhanced dietary fiber content promoting phenolic acid formation that inhibits enzymatic hydrolysis (Ngo et al., 2024), increased system viscosity slowing gastric emptying and enzyme diffusion (Chen et al., 2020), and structural encapsulation of starch granules within the fiber-protein matrix. These findings position okara as an effective modifier for developing noodles with optimized starch digestibility. These changes yielded a reduced hydrolysis index (HI: 46.3–55.4) compared to control noodles (63.5) (Table 1), consistent with the observed decrease in RDS. Starch hydrolysis kinetics showed distinct digestion patterns - rapid hydrolysis within an initial 30 min followed by stabilization after 60 min (Fig. 1D), consistent with okara's effects in other starch systems (Kang et al., 2018). Notably, 40 % okara substitution achieved the most pronounced effect, lowering the hydrolysis rate to 38.8 % (versus 52.4 % in white bread reference), demonstrating dose-dependent inhibition of glucose release. Glycemic response analysis revealed a progressive decline in predicted glycemic index (pGI) with increasing okara content, with formulations exceeding 35 % substitution transitioning from high-GI (GI > 70) to medium-GI (55 ≤ GI ≤ 70) classification (Table 1). GI classification thresholds are low GI (≤ 55), medium GI (56–69), and high GI (≥ 70) (Wang et al., 2023, Wang et al., 2023). These results collectively establish triple-modified okara as an effective functional ingredient for developing reduced-glycemic-response noodle products.

Table 1.

Effect of okara content on the noodle maximum hydrolysis rate, calculated hydrolysis index (HI), and predicted glycemic index (pGI).

Okara content	0 %	30 %	35 %	40 %
C∞	52.4 ± 1.3a	46.1 ± 1.2b	45.4 ± 1.0b	38.8 ± 0.9c
HI	63.5	55.4	53.7	46.3
pGI	74.6	70.1	69.2	65.1

Notes: Means with different lowercase letters in the same row represented significant differences (P < 0.05), C^∞: The maximum hydrolysis rate, HI: Hydrolysis index, pGI: Predicted glycemic index.

3.2. Processing characteristics

3.2.1. Rheological properties

Rheological properties demonstrated that okara incorporation significantly modified dough viscosity, with storage modulus (G′) showing a biphasic response (initial increase followed by decrease) and loss modulus (G″) exhibiting consistent reduction (Fig. 2A–B), indicating concentration-dependent modulation of energy storage and dissipation capacities. Dynamic frequency sweeps were performed in triplicate for each formulation, with G′ consistently exceeding G″ across all formulations (tanδ <1, Fig. 2C), confirming maintained elastic dominance, the tanδ trajectory (initial decrease then increase with okara content) revealed optimal reinforcement at intermediate concentrations (0–30 %) where solid-like behavior peaked, beyond which (> 35 %) viscosity effects became predominant. This behavior stems from competing mechanisms: hydrophilic dietary fibers forming water-retentive gel structures that initially stabilize then, at higher concentrations, interfere with gluten-water interactions (Wang et al., 2021, Wang et al., 2021, Wang et al., 2021), and okara proteins that fortify the gluten network through complementary crosslinking until reaching a threshold where network disruption occurs. These rheological transitions directly correlate with the observed textural modifications in final noodle products (Li et al., 2024).

Fig. 2.

Changes in (A) storage modulus (G′), (B) loss modulus (G″), (C) ratio of G″ to G' (tanδ) of dough with 0 %, 30 %, 35 %, and 40 % okara. (D) and (E) cooking characteristics, and (F) redness (a∗), yellowness (b∗) and lightness (L∗) of noodles with 0 %, 30 %, 35 %, and 40 % okara. Means with different lowercase letters on the same type bars represented significant differences (P < 0.05).

3.2.2. The cooking characteristics of noodles

Cooking loss, a critical quality parameter reflecting nutrient retention during preparation, showed a biphasic response to okara incorporation: initially decreasing and then increasing, yet remaining significantly lower than control wheat noodles (P < 0.05) (Fig. 2D). The optimal 35 % okara substitution achieved minimal cooking loss (4.7 %), attributable to modified starch leaching behavior and gluten network reorganization (Jia et al., 2022), with fermentation-enzymolysis further enhancing cooking quality. Concurrently, okara addition elevated water absorption capacity to 251.9 % (Fig. 2E) through two complementary mechanisms: enhanced dietary fiber content providing additional water-binding sites and modified starch gelatinization dynamics. While increased hydration generally improved texture, excessive water absorption risked structural weakening, highlighting the importance of balanced formulation for optimal noodle quality.

The breakage rate evaluation revealed that all noodle formulations (including controls) maintained 0 % breakage (Fig. 2D), demonstrating that up to 40 % okara incorporation preserved structural integrity. Elongation characteristics followed a biphasic pattern, peaking at intermediate okara levels (30–35 %) before declining yet consistently exceeding wheat noodle benchmarks (Fig. 2E). Moderate okara addition (≤ 35 %) enhanced gluten network plasticity through optimized water distribution, excessive incorporation (> 35 %) led to water competition between dietary fibers and gluten proteins, ultimately constraining network expansion. Given the direct correlation between elongation and perceived quality, formulations should target less than 35 % okara to maintain optimal textural properties while maximizing nutritional benefits.

3.2.3. The color of noodles

The incorporation of modified okara significantly influenced noodle color parameters, with observed increases in redness (a∗) and yellowness (b∗) coupled with decreased lightness (L∗) (Fig. 2F). At 40 % substitution, noodles exhibited acceptable color values (a∗ = 2.6, b∗ = 14.0, L∗ = 63.3), despite the notable reduction in whiteness. Interestingly, a∗ remained stable across 30–40 % okara additions, suggesting a saturation effect. Three primary mechanisms contributed to these color changes: enzymatic browning and fermentation-induced darkening of okara imparting yellow pigments, caramelization of process-generated reducing sugars (Odey and Lee, 2020), and polyphenol oxidase (PPO) mediated reactions (Wen et al., 2023). While deviating from traditional bright white noodles, the resulting color profile aligns with consumer expectations for fiber-enriched noodle products.

3.2.4. The texture of noodles

Texture profile analysis revealed that 40 % okara incorporation significantly increased noodle viscosity (Fig. 3A–B), attributable to the enhanced water-binding capacity of dietary fibers and proteins in the fermented-enzymatic okara that promoted starch hydration (Korus et al., 2020). Hardness and gumminess showed a biphasic response, peaking at intermediate okara levels (30 % for gumminess at 103.5 g) before declining, yet consistently exceeding wheat noodle benchmarks, due to polysaccharide-protein interactions partially substituting gluten network functionality (Li et al., 2025; Kamble et al., 2019). Conversely, cohesiveness reached its minimum (0.4) at 30 % okara before recovering, while elasticity and chewiness progressively decreased from 0.6 mm to 0.6 mm and 0.8 mJ–0.5 mJ, respectively, reflecting water redistribution between okara's hydrophilic components and the gluten matrix (Fan et al., 2022). These texture modifications demonstrate the competing effects of okara's water absorption capacity and its structural role in the protein-carbohydrate network.

Fig. 3.

Changes in (A) and (B) texture of noodles with 0 %, 30 %, 35 %, and 40 % okara. Means with different lowercase letters on the same curve represented significant differences (P < 0.05).

3.3. Structural changes

3.3.1. SEM of noodles

Microstructural analysis revealed concentration-dependent effects of okara on noodle architecture (Fig. 4A–B), with optimal 30–35 % incorporation promoting uniform gluten networks that effectively encapsulated starch granules through three-dimensional protein-starch interactions (Liu et al., 2021), while higher concentrations (> 35 %) caused gluten dilution and starch granule exposure due to dominant dietary fiber interference (Wang et al., 2024). Thermal processing exacerbated these differences: cooked noodles showed progressive structural loosening with increasing okara content, as hydrophilic groups in modified dietary fibers (Zhao et al., 2023) enhanced water absorption during gelatinization, generating excessive hydration pressure weakened gluten matrix integrity (Van Ngo and Luangsakul, 2025). This disruption was compounded by phenolic-gluten interactions and reduced crosslinking efficiency, consistent with reported structural discontinuities at high okara levels (Xie et al., 2023). That ultimately explains the observed texture modifications across formulations.

Fig. 4.

Scanning electron micrographs of (A) raw noodles and (B) cooked noodles with 0 %, 30 %, 35 %, and 40 % okara. Noodles with 0 % (a, e), 30 % (b, f), 35 % (c, g), 40 % (d, h) okara.

3.3.2. FTIR of noodles

FTIR analysis revealed consistent spectral patterns across all okara formulations (Fig. 5A–B), confirming minimal alterations to fundamental starch molecular structures. Across all samples, a prominent broad peak at approximately 3390 cm⁻¹ signified typical O-H stretching vibrations in carbohydrates (Zhang et al., 2023). A distinct absorption band at 2930 cm⁻¹ corresponded to C-H stretching, characteristic of hydrogen atoms (Xu et al., 2022). Another notable peak near 2850 cm⁻¹ arose from the starch-lipid complex, and the peak appearing at approximately 1530 cm⁻¹ represented the starch-protein complex (Lin et al., 2020). That indicates the preservation of these molecular interactions during cooking. Notably, increasing okara content changed hydrogen bonding signals (1704-1610 cm⁻¹ region), particularly through intensified polymer hydroxyl interactions. The amide I (C=O stretching and N-H vibration) and amide II (N-H vibration and C-N stretching) bands between 1550 and 1500 cm⁻¹ (Meng and Li, 2021) provided critical insights into protein secondary structures, with subsequent peak deconvolution revealing specific conformational changes induced by okara incorporation.

Fig. 5.

The (A) raw noodles' FTIR, (B) cooked noodles' FTIR, (C) raw noodles' protein secondary structure, and (D) cooked noodles' protein secondary structure with 0 %, 30 %, 35 %, and 40 % okara. (E) Changes in free sulfhydryl of noodles with 0 %, 30 %, 35 %, and 40 % okara. Means with different lowercase letters on the same type bars represented significant differences (P < 0.05).

3.3.3. Protein secondary structure

Secondary structure analysis revealed concentration and thermal-dependent conformational changes, with β-sheet content increasing overall while β-turn decreased in raw noodles and showed a biphasic response (increase and then decrease) in cooked noodles as okara content rose (Fig. 5C–D). The 40 % okara formulation exhibited maximal random coil formation (raw noodles showing 16.4 %, cooked noodles showing 41.8 %), whereas α-helix content demonstrated opposing thermal behaviors: decreasing then increasing to 16.8 % in raw noodles while declining to 17.6 % in cooked noodles, though both exceeded raw noodle baselines. Thermal processing induced structural reorganization: heating promoted α-helix formation through partial chain unfolding and thermodynamic stabilization while cooling facilitated β-sheet development via reformed hydrogen bonds and hydrophobic interactions. However, excessive heating ultimately favored random coil formation through non-covalent bond disruption (Navneet et al., 2024). These modifications critically influenced gluten network quality, where increased β-sheet (Cui et al., 2022) and adjusted α-helix (Mu et al., 2023) contents respectively enhanced network complexity and elasticity, likely through dietary fiber-mediated protein restructuring that preferentially stabilized extended conformations while marginally compromising elastic recovery.

3.3.4. The free sulfhydryl in noodles

The stability and formation of gluten networks in noodles depend critically on intermolecular interactions, particularly hydrogen and disulfide bonds, with free sulfhydryl content serving as a key indicator of network integrity (Wang et al., 2023). Increasing okara incorporation significantly elevated free sulfhydryl levels (Fig. 5E), reflecting structural modifications through three mechanisms: dietary fiber disrupting disulfide bonds via active hydroxyl groups (Wang et al., 2021), reducible phenolic compounds cleaving existing bonds (Han et al., 2020), and water absorption during cooking exposing buried sulfhydryl that subsequently formed new crosslinks, explaining the lower free sulfhydryl content in cooked versus raw noodles. FTIR, protein secondary structure, and SEM revealed that cooked noodles exhibited greater structural disorder, confirming that okara primarily influences gluten networks through non-covalent interactions (hydrogen bonds, hydrophobic effects, and ionic bonds) rather than covalent crosslinking, consistent with observations in other fiber-enriched systems (Chen et al., 2021). While weakening the overall network strength (Zhan et al., 2019), these modifications contribute to the unique textural properties of okara-enriched noodles.

4. Conclusion

This study investigated that noodles made with modified okara have lower starch digestibility than pure wheat noodles. The co-fermentation with Pleurotus ostreatus and Saccharomyces cerevisiae and enzymolysis of okara enhanced the quality of the wheat noodles. The formation of starch-protein-lipid complexes and porous dietary fiber networks played a critical role in modulating starch digestibility, leading to a reduction in RDS and an increase in SDS and RS. The elevated dietary fiber promoted the retention of phenolic compounds, thereby enhancing the antioxidant capacity. Rheological analysis indicated that okara incorporation improved dough elasticity. However, excessive okara addition impaired gluten network formation, increasing viscosity and suboptimal dough performance. The hydrophilic properties of okara's dietary fibers enhanced water absorption and minimized cooking loss while preserving textural integrity. FTIR and SEM revealed that modified okara contributed to thermally stable composites, maintaining structural stability during cooking. The collective findings demonstrate that modified okara is a multifunctional ingredient capable of simultaneously enhancing nutritional quality and processing characteristics in noodle formulations, with optimal performance observed within a specific substitution range. However, sensory acceptance and shelf-life stability were not evaluated, which are critical for commercial viability. Future research should prioritize comprehensive sensory profiling and consumer studies to optimize acceptability, investigate shelf-life extension strategies, and explore synergistic effects with other fiber-rich byproducts to enhance nutritional profiles further. This work underscores the promise of bioprocessed okara in functional foods; however, its practical application is still lacking.

CRediT authorship contribution statement

Guohong Tian: Contributed data, Formal analysis, Performed the. Meidan Li: Software, Formal analysis. Shuqi Xing: Software. Laping He: Writing – review & editing, Funding acquisition, Supervision. Cuiqin Li: Advice on data treatment. Shunbin Qiao: Advice on data treatment. Ye Yuan: Software. Taixun Luo: Software.

Ethical statement

No humans or animals were subjected to the research.

Declaration of competing interest

We declare that we have no financial or personal relationships with others or organizations that can inappropriately influence our work.

Handling Editor: Dr. Yeonhwa Park

Data availability

Data will be made available on request.