High‐Protein Hydrogels Composed of Pea–Collagen Protein and κ‐Carrageenan for the Dysphagia Diet

TL;DR

Researchers developed high-protein hydrogels using pea-collagen and κ-carrageenan to create nutritious, easy-to-swallow foods for people with dysphagia.

Contribution

The study introduces a novel high-protein hydrogel formulation suitable for dysphagia diets with good nutritional and textural properties.

Findings

Hydrogels with 20%–27.5% protein content were developed, suitable for mild dysphagia as per the IDDSI framework.

Higher collagen concentration increased water-holding capacity and did not negatively affect deformation or digestion.

The degree of proteolysis during in vitro digestion ranged from 55.08% to 65.67%.

Abstract

In this study, pea–collagen proteins and κ‐carrageenan were used to create hydrogels with high protein content (20%–27.5%) for dysphagia diets by varying the collagen concentration (7.5%–15%). The effects of collagen concentrations on physical, rheological, and textural properties; microstructure; suitability for a dysphagia diet; and the degree of proteolysis during in vitro digestion of hydrogels were explored. Temperature‐sweep analysis indicated that collagen did not participate in network formation and behaved more like an inactive filler. For this reason, the amount of nonnetwork proteins increased with increasing collagen concentration. This caused high water‐holding capacity (96.21%–99.75%), low hardness (1.58–1.82 N), and cohesiveness values. A higher collagen concentration in the hydrogel structure did not negatively affect the hydrogel's deformation, which was beneficial for the force required for food processing and swallowing, as well as for nutritional value. On the basis of the International Dysphagia Diet Standardization Initiative (IDDSI) framework, hydrogels were classified as Level 6, making them suitable for mild dysphagia. Collagen concentration only slightly altered the overall digestion pattern, whereas the degree of proteolysis at the end of the intestinal phase ranged from 55.08% to 65.67%. These findings provided new insights into the design and development of nutritious, palatable foods for people with swallowing difficulties.

Hydrogels composed of pea–collagen protein and κ‐carrageenan can be used in the diet of elderly people with dysphagia as a protein source, increasing food intake.

Figures

Frequency (A) and strain sweeps (B) in pea/collagen protein and κ‐carrageenan hydrogel samples with different collagen concentrations.

Amplitude sweeps of the pea/collagen protein and κ‐carrageenan hydrogels with different collagen concentrations. The filled dots represent elastic modulus (G′), and the empty dots represent loss modulus (G″).

Temperature sweep results of elastic modulus (G′) for hydrogel PP‐κcar‐C10, 12.5% w/w of pea protein solution (Pea P 12.5%), 0.6% w/w κ‐carrageenan solution (k‐carrag), and 10.0% collagen solution (Col P 12.5%). RS—calculated rheological synergism. The photos show different systems after the cooling and storage stages.

Network and nonnetwork proteins of pea–collagen protein and κ‐carrageenan hydrogels with different collagen concentrations (A) and schematic illustration of the proposed formation mechanism of pea–collagen protein and κ‐carrageenan hydrogels with the lowest (7.5%) and the highest (15%) collagen concentrations (B).

Visual appearance and SEM images of the pea/collagen protein and κ‐carrageenan hydrogels with different collagen concentrations; a1–a4 magnification 500×; b1–b4 magnification 1000×. The yellow arrows indicate fragments of the carrageenan network.

Fork pressure test, spoon pressure test, fork separation test of pea protein–collagen hydrogels.

Protein hydrolysis degree (%) in different hydrogel formulations during in vitro digestion. G0—gastric phase 0 min; G120—gastric phase 120 min; D120—intestinal phase 120 min; D240—intestinal phase 240 min. Data labeled with different lowercase letters (a–d) showed statistically significant differences (p < 0.05) between different hydrogel formulations under the same digestion phase. Data labeled with different uppercase letters (A and B) showed statistically significant differences (p < 0.05) between different digestion phases for individual hydrogel formulation. PP12.5—control pea protein solution (12.5% w/w) sample; C7.5—control collagen solution (7.5% w/w) sample; C15—control collagen solution (15% w/w) sample.

Tables

TABLE 1. Pea/Collagen protein and κ‐carrageenan hydrogel formulations with different collagen concentrations.

g/100g	PP‐κcar‐C7.5	PP‐κcar‐C10	PP‐κcar‐C12.5	PP‐κcar‐C15
Collagen	8.33	11.11	13.88	16.66
Pea protein	15.63
κ‐Carrageenan	0.6
Potassium chloride	0.1
Water	75.35	72.56	69.79	67.02
Total protein content, %	20	22.5	25	27.5

TABLE 2. Water‐holding capacity, swelling ratio, and color of the pea/collagen protein and κ‐carrageenan hydrogels with different collagen concentrations.

Samples	PP‐κcar‐C7.5	PP‐κcar‐C10	PP‐κcar‐C12.5	PP‐κcar‐C15
WHC, %	96.21 ± 0.96a	97.20 ± 2.46ab	99.75 ± 0.43bc	98.90 ± 1.06c
L*	68.26 ± 0.55c	66.10 ± 0.33b	64.73 ± 0.20a	64.68 ± 0.94a
a*	3.79 ± 0.06bc	3.62 ± 0.02abc	3.33 ± 0.11a	3.86 ± 0.30c
b*	19.22 ± 1.24a	20.82 ± 0.39bc	20.89 ± 0.27c	20.19 ± 0.57bc
Hardness, N	1.58 ± 0.16ab	1.77 ± 0.17ab	1.82 ± 0.04b	1.74 ± 0.12ab
Cohesiveness	−0.20 ± 0.01c	−0.13 ± 0.01a	−0.16 ± 0.03b	−0.17 ± 0.02b
SR, %	No visible phase separation

Full text

Introduction

1

Due to reduced appetite and a range of medical disorders (aggressive oncological treatment, shock, and stroke), hospital patients did not eat their full regular meals, resulting in a reduction in food intake far below their estimated energy requirements, which was especially evident in critically ill lying‐in patients (Patel and Martin 2008). Furthermore, patients with dysphagia often have difficulty swallowing safely. This leads to reduced oral intake of food and fluids, which can result in malnutrition, weight loss, and dehydration (Schwartz et al. 2018).

For a long period of time now, thickened liquids as well as texture‐modified foods (semisolid and solid) are a fundamental part of clinical practice for managing dysphagia in order to meet the basic nutritional needs of patients (Raheem et al. 2021). Foods must have certain textural and physical properties to facilitate swallowing, thereby improving safe swallowing in people with dysphagia and reducing the risk of aspiration pneumonia (Gallego et al. 2022). Besides this, dysphagia‐oriented formulations must be nutritionally balanced to improve caloric and nutritional status for individuals with dysphagia. In this context, an adequate daily protein intake (1.0–1.3 g/kg body weight) is highlighted as the most effective strategy for reducing muscle‐wasting disorders in patients (Nowson and O'Connell 2015). Protein‐enriched diets help to prevent loss of muscle strength (Volkert et al. 2024), and researchers are working hard to develop new formulations for dysphagia patients. First, thickened protein‐based beverages are a potential strategy to address protein deficiency for various patients (Streimikyte et al. 2020); however, the limitation of such formulations is that it is not convenient for people with severe dysphagia to consume large quantities of those products. Another direction that has recently attracted growing attention is developing dysphagia‐friendly food formulations suitable for 3D food printing, resulting in an appealing design that can also meet the sensorial pleasure of the patients. So far, one frequently unresolved issue is that they do not have high protein content, due to the need to balance printing performance (printability) with the nutritional composition of the food‐ink formulation (Hou et al. 2023; Zhang et al. 2024).

A third group of studies focuses on the development of complex emulsified and/or gelled food formulations that not only facilitate the swallowing process but also serve as a source of proteins. Rice starch–soybean protein composite gels were formulated by changing soybean protein ratios from 0% to 12% (Ren et al. 2024). Heated whey protein isolate‐pectin complexes were also used to prepare texture‐modified foods with high protein content (15%, w/v) (Fan et al. 2024). Authors explored the potential of mung bean starch–flaxseed protein composite gels for use in dysphagia diets (Min et al. 2023). Compared to single‐component gels, composite gels can modulate gel texture as well as digestive behavior and bioaccessibility of individual components (Yang et al. 2021).

We hypothesized that using protein mixture gels would not only achieve high protein content but also produce soft‐textured, safe‐to‐swallow gels with controlled protein digestion. In this context, pea protein caught our attention because, in addition to its high nutritional value, it has the ability to form soft, smooth gels whose texture and rheological properties can be modified with polysaccharides, such as carrageenan (Zhang et al. 2025; Ge et al. 2022). The second protein, collagen, which is not commonly used in such formulations, appeared to have considerable potential in both protein and peptide forms. Bioactive collagen peptides can regulate the synthesis of extracellular matrix proteins in various tissues and, by supplying the amino acid precursors required for endogenous collagen formation, can regulate this process. This is highly valuable for improving health, particularly joint and bone strength (Choi et al. 2021). Additionally, collagen has the advantage over other proteins of not forming rigid gels upon thermal denaturation, thereby allowing a significant increase in protein content. The selection of these proteins for the mixture was also based on their differing digestibility, which is known to be significantly higher in pea proteins (Zhu et al. 2022). The behavior of complex protein mixtures after ingestion into the gastrointestinal tract remains largely unexplored. Typically, most studies focus on the structural aspects of texture‐modified foods, whereas the digestive behavior of protein‐rich formulations needs to be assessed.

Herein, we propose to develop dysphagia‐oriented collagen–pea protein κ‐carrageenan hydrogels by changing collagen concentration. The resulting systems were characterized at structural and interaction levels by evaluating their physical, rheological, and textural properties. The potential of hydrogels to be implemented in the diet of dysphagia patients was evaluated by using International Dysphagia Diet Standardization Initiative (IDDSI) testing methods. Finally, in vitro simulated digestion analysis was performed in order to investigate how different hydrogel properties influence the digestibility of proteins to get a more detailed understanding of how this process could be controlled.

Materials and Methods

2

Materials

2.1

Collagen (bovine, I and III type mixture; 90% of protein, 20% hydrolyzed) and pea protein isolate (80% of protein, 5.5% of fat, 2.6% of carbohydrate, 4.1% of fiber, 1.9% of salt) were obtained from MyProtein (Manchester, UK). ĸ‐Carrageenan was obtained from Sigma Aldrich (Steinheim, Germany), and potassium chloride (KCl, ≥99% purity) was purchased from Eurochemicals (Vilnius, Lithuania). Sodium azide was used as a bacteriostatic agent to prevent microbiological deterioration (Sigma Aldrich, UK).

Preparation of Hydrogels

2.2

The calculated amount of κ‐carrageenan and distilled water was weighed into a beaker using an analytical balance, and the mixture was stirred on a magnetic stirrer for 30 min. After the carrageenan was completely dissolved in the water, the weighed amount of collagen (concentrations specified in Table 1) was added, and the beakers were left on the stirrer for about 15–20 min until complete mixing. Potassium chloride was added and stirred for a few more minutes. The purpose of adding potassium chloride was to induce gelation of carrageenan. κ‐Carrageenan forms strong and brittle gels in the presence of cations, such as K+ (Schefer et al. 2015). Finally, the required amount of pea protein was added and stirred. The beaker was covered with foil and left to stand for about 30 min to allow the proteins to swell. To obtain collagen–pea protein and κ‐carrageenan hydrogels, the beakers were heated in a water bath at 95°C for 30 min (ISOTEMP 205, Fisher Scientific, US), cooled at room temperature, and stored overnight at 4°C to allow complete gelation. A bacteriostatic compound (0.1% v/v) was incorporated into the hydrogels to prevent microbial degradation. Four different hydrogels have been produced and abbreviated as PP‐κcar‐C7.5, PP‐κcar‐C10, PP‐κcar‐C12.5, and PP‐κcar‐C15. Before the analysis, gels were kept at room temperature for 30 min.

Characterization of Bigels

2.3

Storage Stability

2.3.1

Hydrogel samples were stored at 4°C for 14 days. Stability was evaluated visually, and hydrogels were considered stable if they retained their shape and did not release moisture during storage.

The Syneresis Rate (SR)

2.3.2

The SR of the hydrogels after the freeze–thaw treatment was determined according to Basak and Singhal (2024), with some minor modifications. Samples were stored at −20°C for 24 h and then thawed at room temperature for 1 h, evaluating the amount of water released from the hydrogel structure. The SR was calculated from the difference between the weights of the hydrogels before and after freeze–thaw treatment using the following formula:

$$[eqn]$$

where SR is the syneresis rate, %; *W_i_

• is the hydrogel weight before freeze–thaw treatment, g; and *W_f_
• is the hydrogel weight after freeze–thaw treatment, g.

Water‐Holding Capacity (WHC)

2.3.3

For WHC determination, 5 g of the hydrogel was placed into 15 mL centrifuge tubes and centrifuged (Labofuge 200, Thermo Scientific, Bracknell, UK) at 4000 × g for 20 min at room temperature. The weight of the hydrogel samples before and after centrifugation (after removing separated water) has been used to estimate the WHC

$$[eqn]$$

where WHC is the water‐holding capacity, %; *W_i_

• is the hydrogel weight before centrifugation, g; *W_f_
• is the hydrogel weight after centrifugation, g.

Color

2.3.4

The tristimulus color coordinates (Lab**) of the hydrogels were measured using a Chroma meter CR‐410 colorimeter (Konica Minolta, Osaka, Japan), with D50 (50 mm) illumination area and 10° observer, within a ΔE**ab standard deviation of 0.07. Measurements were replicated five times.

Texture Analysis

2.3.5

Textural properties of the hydrogels (cylinder shape 3 × 3 cm^2^) were evaluated by using TA.XT plus texture analyzer (Stable Micro Systems, TA.XT Plus, Godalming, UK). The compression test was performed with a P/05S probe with the following parameters: trigger force—10 g, measurement speed—2 mm/s, and penetration distance—10 mm. Results were expressed as hardness (the maximum force required to deform the hydrogel) and cohesiveness (the strength of internal bonds within a hydrogel).

Rheological Properties

2.3.6

The rheological properties of the hydrogels were characterized using a rheometer MCR92 (Anton Paar, Graz, Austria) fitted with a 25 mm parallel plate test cell (gap of 1 mm) at 25°C.

Small Amplitude Oscillatory Shear (SAOS) Rheology

2.3.6.1

Frequency sweep: An oscillatory shear frequency sweep test was conducted from 0.1 to 100 rad/s within the linear viscoelastic region (LVR) (0.02% strain). The results were expressed as storage (G′) and loss (G″) modulus.

Temperature sweep: Temperature sweep analysis was conducted according to the methodology of Tanger et al. (2022), with minor modifications. Hydrogel samples were subjected to temperature increase from 25°C to 95°C, holding them at 95°C for 10 min, cooled down to 25°C, and then holding samples again at 25°C for 10 min. A temperature sweep was performed at a controlled temperature of −2°C/min (strain value = 1%, frequency = 1 Hz, within the LVR). In order to better understand the mechanism of gel formation, in addition to hydrogel samples, pure pea protein solution (12.5% w/w), κ‐carrageenan solution (0.6% w/w), and collagen solution (12.5% w/w) were also subjected to the same conditions. To protect the samples from dehydration, a solvent trap was used by initially covering the samples with silicone oil.

The rheological synergistic effect of each structuring ingredient was calculated as described by the following equation (Shahbazi et al. 2021) on the basis of the temperature sweep analysis results:

$$[eqn]$$

where RS is the rheological synergistic effect; G′mixture is the storage modulus of the mixed system, Pa; and G′ingredients is the storage modulus of each ingredient at 25°C after the end of the thermal cycle measurements, Pa.

Structural recovery rate: The structural recovery of the hydrogels was evaluated by strain sweep analysis. The sweep step strain tests were applied by changing the amplitude from the small strain (0.1%) to the large strain (50%) with an interval of 600 s. This amplitude loop has been repeated three times for a total of 1800 s. The recovery rate was calculated by using the following equation:

$$[eqn]$$

where RR is the structural recovery rate, %; G′at 0.1 is the initial storage modulus (G′) at 0.1% of the strain, Pa; and G′at 50 is the storage modulus at 50% of the strain, Pa.

Large Amplitude Oscillatory Shear (LAOS) Rheology

2.3.6.2

A strain sweep test was performed by logarithmically increasing strain from 0.1% to 1000% at a fixed frequency of 1 Hz. The critical strain of the LVR for the samples was calculated as the strain at which G′ deviated by more than 5% from its previous value (Shahbazi et al. 2021).

Scanning Electron Microscope (SEM) Analysis

2.3.7

The hydrogel samples were freeze‐dried with Alpha 1–4 LSC freeze‐drier (Martin Christ, Germany) for 18 h at 1 mbar, with the condenser temperature set to −55°C. A brand SEM was used to examine the microstructure of the hydrogels. Samples were placed on SEM Al stubs and then coated with 5 nm of gold in a Quorum Q150 model Au sputter coating unit (Quorum Tech, GB). The samples were acquired and imaged at a working distance of 10 mm using a Tescan Mira3 XMU (Czech Republic) at 10 kV, with a backscattered electron (BSE) detector, and representative images were recorded at magnifications of 500× and 1000×.

Network and Nonnetwork Protein

2.3.8

Network and nonnetwork proteins in hydrogels were determined according to the method described by Ge et al. (2022). After weighing 2 g of gel, it was poured with 20 mL of distilled water and left in a water bath at 25°C for 48 h (GFL 1092, Thermolab, Hanover, Germany). The Bradford assay was used to determine the amount of protein (soluble fraction) released from the hydrogel composition. Nonnetwork proteins were calculated as the percentage of proteins solubilized in the solvents relative to the initial protein concentration in the hydrogel. Proteins that did not release into the solvent were evaluated as network proteins and expressed as a percentage.

IDDSI Tests

2.3.9

The hydrogels were tested in accordance with Level 5 and 6 foods and classified according to the IDDSI framework. Spoon and fork pressure tests were used. The fork/spoon pressure testing employs the thumb finger to press the samples with a fork/spoon until the thumbnail blanches, which is equivalent to the tongue pressure during swallowing (∼17 kPa).

In Vitro Digestion

2.3.10

Static in vitro digestion of different hydrogel formulations was carried out in three stages, following the INFOGEST method proposed by Brodkorb et al. (2019) and adapted for older adults (Shani‐Levi et al. 2017). For the experiment, simulated salivary (without amylase; SSF), gastric (SGF), and intestinal fluids (SIF) and enzymes were used (pepsin 750 U/mL of digest, pancreatin 46 U/mL of digest, and bile salts 59.4 mg/mL of digest). For the gastric phase, pH was adjusted to 2 with 6 M HCl; for the intestinal phase, pH was adjusted to 7 using 6 M NaOH. The hydrogel‐digest mixture was stirred at 7 rpm in a water bath (Thermolab GFL 1092, Potsdam, Germany) at 37°C. Pepsin activity in the samples was inhibited by adding NaOH until the pH reached 7. At each time point (G0, G120, D120, and D240), the samples were cooled in ice water to 0°C–4°C and centrifuged (MPW‐260RH, MPW Med. Instruments, Warsaw, Poland) at 2683 × g for 20 min at 4°C. After centrifugation, the samples were filtered, and the soluble fraction was collected and stored at −18°C. The control pea protein solution (125% w/w) and the collagen (15% w/w; 30% w/w) solutions were prepared under the same conditions as for the hydrogel and used for analysis. To track the extent of digestion, the degree of proteolysis (DH) was measured (Larsen et al. 2004). DH was calculated using the following equation by evaluating the leucine equivalents using an external leucine standard curve:

$$[eqn]$$

where h is the amount of N‐terminal amines at each time point of in vitro digestion, and h tot is the total amount of N‐terminal amines determined after full hydrolysis with HCl.

Statistical Analysis

2.4

All experiments were conducted at least in triplicate and statistically analyzed using one‐way analysis of variance (ANOVA). The average values were compared using Duncan's multiple range test (p < 0.05). Statistical analyses were performed using SPSS 12.0 analysis software (StatSoft Inc., USA).

Results and Discussion

3

Characterization of Hydrogels

3.1

Physical Characterization

3.1.1

By considering the protein needs of dysphagia patients, four hydrogel formulations were developed by varying collagen concentrations (7.5%–15%) and total protein content (20%–27.5%). The aim of this section is to evaluate the effect of collagen concentration on the physical and structural properties of the resulting hydrogels, taking into account both nutritional needs and changes in food consumption abilities due to cognitive impairment in patients.

All developed hydrogels were stable and exhibited a solid‐like structure and self‐supporting ability after preparation (Table 2) and during subsequent storage. This feature makes it possible to use the hydrogel as a “finger food” by picking it up and bringing it to the mouth without crushing it between the fingers, providing an opportunity for self‐feeding, which is so needed for patients having an early apraxia (inability to handle a spoon and fork) (Delaide et al. 2020). Furthermore, it is also believed that self‐supporting hydrogels may be more appealing to patients with dysphagia than pureed or thickened liquid foods.

In terms of visual appeal, the specific yellowish‐brown color of the pea proteins and collagen led to the characteristic hydrogel color (Figure 5). In addition, increasing the collagen concentration resulted in a decrease in lightness (L* value from 68.26 to 64.68) and only a very slight increase in redness (a* value from 3.79 to 3.86) as well as yellowness (b* value from 19.22 to 20.19), so that the visual differences between the samples were negligible. It is believed that the formation of less‐dense structural hydrogel networks in samples with higher collagen concentrations altered light scattering and caused the hydrogel to darken (Lv et al. 2023). Highly similar color characteristics were observed in binary pea protein–psyllium hydrogels prepared with varying NaCl additions across different pH values (Hilal et al. 2024). The color and physical state of food are key factors in stimulating appetite and can positively affect patients’ food choices.

Low syneresis ratio and high WHC are desirable properties, indicating the ability of biopolymers to physically bind water molecules. Soft gelled foods also require a higher WHC to maintain a moist mouthfeel during consumption. WHC values were relatively high (>96%) across all hydrogel formulations, with slightly higher values at higher collagen concentrations (∼99%). This can be attributed to the greater amount of protein, leading to a higher number of molecular protein–water interactions (Min et al. 2023).

The freeze–thaw treatment resulted in no syneresis, demonstrating the strength of the protein network, which helped retain water within the hydrogel matrix under external stress. This property is useful in practical applications of hydrogels, for example, extending their shelf life through refrigeration while maintaining their structure.

Texture plays an important role in the palatability of the food, so TPA tests were performed in order to mimic oral mastication. The hardness of hydrogels was influenced by collagen addition, indicating slightly denser protein network formation. Consistent with our observations, previous studies reported a stronger correlation between hardness and total protein content (Kim et al. 2025). In general, hardness values (1.58–1.82 N) in all hydrogel formulations were similar or lower compared to other studies for dysphagia‐oriented foods (Kim et al. 2025; Dick et al. 2021), making them easy‐to‐bite and ensuring consumption with minimal chewing effort (disintegrate in the mouth by a tongue palate compression without mastication). In addition to high softness, appropriate cohesiveness plays an important role in the formation of a moist, homogeneous food bolus, ensuring a safe swallowing process (Hadde and Chen 2021). It was found that all cohesiveness values (0.13–0.20) were in the range of the recommended value for dysphagia patients (0.11–0.24) (Herrera‐Lavados et al. 2023), suitable to prevent decomposition of food bolus into small particles during safe swallowing. However, excessive cohesiveness can lead to chewing difficulties, making eating uncomfortable (Pure et al. 2021).

Dynamic Rheological Properties of the Hydrogels

3.1.2

Frequency Sweeps

3.1.2.1

The changes in storage modulus (G′) and loss modulus (G″) as a function of frequency for hydrogels with different collagen concentrations are shown in Figure 1A. All samples are confirmed to exhibit consistent elastic solid‐like behavior with G′ always exceeding G″, implying they are “solid‐like” materials, with only minimal dependence on the collagen protein concentration. However, a G′/G″ ratio <10 indicates that the structure of the resulting samples is typical of a weak gel (Xu et al. 2023), which correlates with the results of other authors who have developed different dysphagia‐oriented systems, that is, pea protein isolate‐based texture‐modified food by ultrasound treatment and salmon protein gels (Xu et al. 2023; Fei et al. 2024). Developing foods for the dysphagia diet, a weak gel structure is advantageous because it makes bites easier and may facilitate swallowing, which is particularly relevant for individuals with chewing disorders. It was also observed that tan δ values across the entire frequency range were less than 1, which may also be attributed to easier food bolus formation and to facilitate swallowing when the tan δ value is in the range of 0.1–1 (Talens et al. 2021).

Frequency (A) and strain sweeps (B) in pea/collagen protein and κ‐carrageenan hydrogel samples with different collagen concentrations.

G′ and G″ showed an increasing trend with increasing frequency, indicating a more flexible network structure and its sensitivity to applied deformation. Similar results have already been reported for protein and polysaccharide hydrogels (Basak and Singhal 2024; Odelli et al. 2024). The frequency dependence could also be explained by noncovalent physical cross‐linking, that is, the main factor in the composition of the gel network (Anvari and Joyner 2017), as well as at higher frequencies, the structure has less time to fully rearrange and is therefore more resistant to deformation (Lin et al. 2024).

Strain Sweep

3.1.2.2

LAOS measurements are advantageous over measurements in the SAOS range (tests are confined to the linear viscoelastic zone) due to the possibility of performing so‐called near real measurement in the complex food systems (Duvarci et al. 2017). This method may provide more detailed information on the behavior of the designed hydrogels in the non‐LVR corresponding to the large deformation occurring during oral processing, more specifically, crushing food with their tongue (because of reduced tooth or muscle strength) (Yu et al. 2024). In the range of relatively small strains (<1%), both G′ and G″ exhibit a constant plateau, whereas G′ was higher than G″, indicating a predominant elasticity (Figure 1B). As the strain increases, both moduli decrease until they intersect at the critical strain, which was 2.61% across all tested samples. This point represents the minimum tongue pressure required to squeeze the hydrogel in the mouth, as beyond this point, the structure disrupts, and the fluid‐like behavior starts to dominate (Ang et al. 2022). As shown, neither the changing collagen concentration nor the total protein content had any effect on the critical strain, allowing the total protein content to be increased without any drastic changes in the force required to break down the hydrogel. This can be explained by the fact that collagen does not form a gel network (Figure 3) and acts more as a filler than a structure‐forming element.

A step‐change strain‐sweep test was performed to evaluate the ability of hydrogels to recover their properties after deformation, an important factor in determining safe swallowing. This property is also closely related to the cohesiveness, indicating the integrity of the interior bonds within the food bolus. Both parameters are essential for meeting the basic requirements for dysphagia‐adapted food formulation. It can be seen that at low strain values (0.1%), hydrogels demonstrated elastic behavior with G′ > G″, corresponding to the results obtained in the amplitude sweep test (Figure 1B). When deformation increased to 50%, the hydrogel structure experienced significant structural damage (G″ > G′), and recovery was proportional to the content of collagen. As shown in Figure 2, increasing the collagen concentration from 7.5% to 15% reduced the rate of hydrogel structure recovery from 53.5% to 37.2%. The high resistance of the hydrogel structure to deformation, as well as high cohesiveness, may be undesirable from a safe swallowing perspective, as the force of the tongue may not be sufficient to deform it and move it into the throat (Xing et al. 2022). On the other hand, the structure's integrity and resistance to fragmentation during swallowing are also important factors in ensuring safe swallowing and reducing the risk of choking (Chen et al. 2024). As previously discussed, the determined cohesiveness values fall within the recommended range for patients with dysphagia. However, samples with higher collagen concentrations (12.5% and 15%) show relatively poor structural recovery, suggesting that some pharyngeal residues may form in the esophagus after rapid disintegration under masticatory forces (Wang et al. 2025). For the remaining samples, higher structural recovery and cohesiveness values could help to maintain both ease of swallowing (low resistance to the structural deformation) and reduced aspiration risk.

Amplitude sweeps of the pea/collagen protein and κ‐carrageenan hydrogels with different collagen concentrations. The filled dots represent elastic modulus (G′), and the empty dots represent loss modulus (G″).

Interactions Between Pea, Collagen Protein, and Carrageenan

3.1.3

When developing dysphagia diets, particularly those containing more than one macronutrient, it is necessary to understand the role of each component in gel formation and the interactions among components to achieve the most suitable system properties for safe consumption.

A dynamic temperature ramp sweep test was performed in order to evaluate the sol–gel transition properties, with additionally calculated rheological synergism. A hydrogel with a collagen concentration of 10% (PP‐κcar‐C1) was chosen to demonstrate the results, as hydrogels with other collagen concentrations showed similar trends.

The gradual increase in the G′ modulus in pure pea protein solution (Figure 3—PeaP 12.5%), caused by heat‐induced gelation, occurs due to the unfolding of protein molecules and the formation of aggregates, followed by gradual cross‐linking when the denaturation temperature is reached. Even though the main components of pea proteins, like 11S (legumin) and 7S (vicilin), denature at 77°C and 68°C, respectively, the gelation temperature is usually higher (Yang et al. 2021). It was previously reported that low solubility and low content of disulfide bonds in pea proteins result in poor gelling performance and formation of a visually soft, paste‐consistency gel (Ge et al. 2022; Zhang et al. 2022), as confirmed by the results of our study. Such a weak, brittle gel formulation could be characterized as having low self‐standing ability, poor WHC, and a lack of structural stability required for safe swallowing (Gu et al. 2025).

Temperature sweep results of elastic modulus (G′) for hydrogel PP‐κcar‐C10, 12.5% w/w of pea protein solution (Pea P 12.5%), 0.6% w/w κ‐carrageenan solution (k‐carrag), and 10.0% collagen solution (Col P 12.5%). RS—calculated rheological synergism. The photos show different systems after the cooling and storage stages.

During the cooling stage, κ‐carrageenan solution containing potassium ions starts to form a gel that involves coil‐helix transitions and double‐helix aggregation, as indicated by an increase in G′ at the step of the cooling process (Figure 3). κ‐Carrageenan has the ability to form a self‐standing gel and significantly enhances the gelling properties of pea proteins, as previously reported by various authors (Zhang et al. 2025; Bartkuviene et al. 2024). However, in our case, it was observed that the G′ value of the hydrogel composed of pea protein, κ‐carrageenan, and collagen was between the G′ values of pea protein and κ‐carrageenan (Figure 3), which contradicts the results of other authors. The main reason for this is the behavior of the protein collagen under a temperature sweep. No effect of the temperature was recorded, which was to be expected as it was a partially hydrolyzed protein and had no ability to gel. This was also confirmed by the calculated negative value of rheological synergy (−0.98), indicating that there is no synergistic effect between different components that strengthens the gel structure. On the contrary, it is believed that collagen acts as an inactive filler, weakening the interactions between pea proteins and carrageenan. Pea protein contains multiple amino acid side chains that can provide binding sites for κ‐carrageenan via their negatively charged groups (Geng et al. 2024). It is also believed that carrageenan did not form bonds with collagen molecules, as evidenced by the absence of network and nonnetwork proteins in the structure and microstructure.

Network and Nonnetwork Proteins

3.1.4

The increase in the collagen concentration significantly (p < 0.05) reduced the ratio of the network proteins, as indicated in Figure 4A. This phenomenon may be caused by two processes. First, this dependence indicates that collagen does not form a gel and remains trapped in the gel matrix as a soluble component. Second, collagen interferes with the formation of the pea protein network and reduces protein–protein and protein–polysaccharide interactions. It is believed that the presence of collagen in the system causes pea proteins and κ‐carrageenan to compete with collagen for water interactions, thereby reducing the mobility of pea proteins, a phenomenon that is particularly pronounced as collagen concentration increases (Basak and Singhal 2024).

Network and nonnetwork proteins of pea–collagen protein and κ‐carrageenan hydrogels with different collagen concentrations (A) and schematic illustration of the proposed formation mechanism of pea–collagen protein and κ‐carrageenan hydrogels with the lowest (7.5%) and the highest (15%) collagen concentrations (B).

Microstructure

3.1.5

To confirm our hypothesis, we observed the structure of heat‐induced hydrogels, as presented in Figure 5. Pea protein aggregates were clearly visible, varied in size, and described as having an amorphous shape, which was confirmed by other studies (Odelli et al. 2024). Thermal treatment at 95°C for 30 min unfolded the compact structure of pea proteins, exposing active sites and leading to the formation of amorphous aggregates from denatured proteins, primarily through interactions involving hydrophobic and, to a much lesser extent, disulfide bonds (Lin et al. 2023). During cool‐down, κ‐carrageenan transforms from a coil to a helix, causing gel formation through hydrogen bonds between the helical regions of the molecule (Kamlow et al. 2022), which are also visible in all images (Figure 5, indicated by the yellow arrows). Due to electrostatic interactions between pea proteins and κ‐carrageenan, the two gels interpenetrate, resulting in a stronger network. It was also found that increasing collagen concentration results in a coarser network with larger voids and cavities, in which individual fragments are not incorporated into the network structure. We assume that these individual fragments consist mainly of peas and collagen proteins. Collagen protein appears to act as an inactive filler, not binding to the network but merely absorbing water and becoming randomly embedded in the protein and polysaccharide network, thereby weakening it. It is believed to weaken both pea proteins and the carrageenan network, as well as their combined interactions, and to increase the formation of nonnetworked pea proteins within the structure. This phenomenon is more pronounced in the sample with the highest collagen concentration, corresponding to the lowest amount of network proteins (∼70%) (Figure 4). As discussed above, the potential mechanism of hydrogel structure formation is illustrated in Figure 4B in the samples with the lowest (7.5%) and the highest (15%) collagen concentrations.

Visual appearance and SEM images of the pea/collagen protein and κ‐carrageenan hydrogels with different collagen concentrations; a1–a4 magnification 500×; b1–b4 magnification 1000×. The yellow arrows indicate fragments of the carrageenan network.

IDDSI Tests

3.1.6

Following the methods provided by IDDSI, fork and spoon pressure tests were performed to assess the feasibility of the hydrogels as a dysphagia diet (presented in Figure 6). All samples were easily squashed with minimal force (indicated by slight whitening of the thumb tip ∼17 kPa) (Steele et al. 2014) and did not regain their original shape while removing the fork or spoon. These characteristics allow the produced hydrogels to be classified as Level 6 soft, bite‐sized. This shows that in order to form a tender food bolus ready for swallowing, biting, chewing, and oral processing are needed (Sugita et al. 2006), making it more suitable for patients with mild dysphagia. Obtained results correlated with the amplitude sweep test (structural deformation at the higher mechanical loads) as well as low hardness values (1.58–1.82 N). The same IDDSI level and similar hardness values were also reported in Marques et al. (2025) study with the mycoprotein‐based foods. Lower hardness values are beneficial for safe food consumption, as they require less chewing and tongue force (Sugita et al. 2006). Formulated hydrogels will also not be suited for patients with dysphagia who have loose teeth, extremely weakened chewing muscles, and reduced tongue mobility, as they require Level 5 food (Bitencourt et al. 2023). However, the nature of structure formation ensures that even higher collagen concentrations in the hydrogel do not adversely affect its deformation, which is beneficial for the force required for food processing and swallowing, as well as for nutritional value. Products developed for dysphagia diets should contain as high a protein content as possible to ensure that patients can obtain the necessary amount of these macronutrients even with reduced food intake. In this regard, the study aimed to achieve the highest possible collagen (15%) and total protein (27.5%) content without adversely affecting the physical properties necessary for safe swallowing.

Fork pressure test, spoon pressure test, fork separation test of pea protein–collagen hydrogels.

Additionally, we have assessed the behavior of the samples in terms of their ability to retain their shape and integrity when picked up with the fingers. Shape retention during finger pickup of the hydrogel allows bringing food to the mouth, providing self‐feeding ability, which is particularly important if the food is to be eaten with the hands.

In Vitro Digestion

3.1.7

Finally, the impact of different collagen concentrations on protein digestibility was comprehensively investigated, with results presented in Figure 7. As control samples, pure collagen and pea protein solutions were subjected to the same digestive conditions as hydrogels. When evaluating the hydrolysis of individual proteins, their digestibility varies markedly. First, at the beginning of the gastric stage, a higher degree of proteolysis (15.65%–18.14%) was observed in collagen protein solutions (C7.5 and C15), likely due to the use of partially hydrolyzed collagen in the experiment. Due to the relatively low gastric pepsin activity, only 11.82% of the pea proteins were hydrolyzed at the end of the gastric stage, which is consistent with the results of other studies reporting low digestibility of pea proteins under simulated gastric conditions (less than 14%) (Herrera et al. 2024; Jimenéz‐Munoz et al. 2021). After digestion in the gastrointestinal tract, the pea proteins were 100% hydrolyzed, whereas the collagen proteins in both collagen samples were hydrolyzed by 62%–65%, consistent with data reported in the literature (Zhu et al. 2022). The obtained results could be related to the complex structure of the collagen molecule, as it is a rigid, rod‐shaped molecule composed of three intertwined polypeptide chains with a high molecular weight (300 kDa) (McClements and Grossmann 2021). Such structural complexity causes difficulty for the enzymes to break down the collagen domain as peptide bonds are shielded from proteases (Haj et al. 2024).

Protein hydrolysis degree (%) in different hydrogel formulations during in vitro digestion. G0—gastric phase 0 min; G120—gastric phase 120 min; D120—intestinal phase 120 min; D240—intestinal phase 240 min. Data labeled with different lowercase letters (a–d) showed statistically significant differences (p < 0.05) between different hydrogel formulations under the same digestion phase. Data labeled with different uppercase letters (A and B) showed statistically significant differences (p < 0.05) between different digestion phases for individual hydrogel formulation. PP12.5—control pea protein solution (12.5% w/w) sample; C7.5—control collagen solution (7.5% w/w) sample; C15—control collagen solution (15% w/w) sample.

After both proteins were incorporated into the hydrogel formulation, the degree of protein hydrolysis was more similar to collagen than to pea protein digestibility, as shown in Figure 7. A higher concentration of collagen in the hydrogel increases the degree of proteolysis at the end of the gastric phase and only slightly alters the overall digestion pattern (degree of proteolysis at the end of the intestinal phase 55.08%–65.67%). The lower degree of proteolysis could be associated with the presence of pea proteins in an interconnected network of proteins and carrageenan, which made it more difficult for proteases to access and hydrolyze the proteins. The same trend was previously reported by Zhu et al. (2022) in cold‐set interpenetrating network hydrogel prepared with wheat bran arabinoxylans and pea protein isolates, where the degree of pea protein degradation reached <60%. Given that the increased collagen concentration caused an increase in the amount of nonnetwork proteins in the structure (Figure 4A), it was expected that this would accelerate the rate of protein hydrolysis. However, collagen, which is more difficult to hydrolyze, accounted for a much larger proportion of free proteins and additionally acted as a physical barrier around pea proteins, reducing enzymatic hydrolysis of these easily hydrolyzed proteins (Lin et al. 2025). In summary, a protein hydrolysis degree of more than 50% is considered satisfactory, particularly given that the total protein content across all hydrogels was relatively high, ranging from 20% to 27.5%.

Conclusion

4

In this study, a high amount of protein (20%–27.5%) in the hydrogel formulations was achieved by mixing pea protein with collagen and additionally using κ‐carrageenan. κ‐Carrageenan forms a self‐standing gel and significantly enhances the gelling properties of pea proteins, thereby enhancing the stability of hydrogel systems. Collagen appears to act as an inactive filler, not binding to the network but merely absorbing water and becoming randomly embedded in the protein and polysaccharide network, thereby weakening it. As the collagen concentration in the system increased, this phenomenon became more pronounced, as evidenced by an increase in the amount of nonnetwork proteins. This phenomenon helps to achieve the desired food properties that facilitate swallowing, such as a moist mouthfeel, softness, and the ability to deform easily but not break into many small pieces when swallowed. The hydrogels produced were classified as level 6 soft, bite‐sized, and therefore more suitable for patients with mild dysphagia, indicating that in order to swallow a food bolus safely, biting, chewing, and oral processing are needed. At the end of the intestinal stage, the degree of proteolysis was 55.08%–65.67% with a slight dependence on collagen concentration.

This study gives insight into how the structure formation mechanism is related to the swallow‐related properties of pea–collagen protein and κ‐carrageenan hydrogels, as well as the in vitro digestion process. On the basis of this, future studies will investigate the use of these hydrogels for the delivery of biologically active components. Furthermore, it will also be important to use sensory tests among dysphagia patients to establish whether these hydrogels are comfortably consumed, desirable, and acceptable.

Author Contributions

Deimantė Dagytė: investigation, formal analysis. Ieva Bartkuvienė: data curation, visualization. Milda Keršienė: resources, conceptualization. Sigita Jeznienė: methodology. Evren Gölge: methodology, investigation. Viktorija Eisinaitė: writing – original draft. Daiva Leskauskaitė: supervision.

Conflicts of Interest

The authors declare no conflicts of interest.