Upcycling Luffa cylindrica (Luffa Sponge) Seed Press Cake as a Functional Ingredient for Meat Substitute Formulations

Abstract

In the current context of environmental concerns and the search for sustainable food solutions, this study investigated the valorization of Luffa cylindrica seed press cake, a waste byproduct from oil extraction, as a functional ingredient for meat substitute formulations. The research systematically characterized the functional and bioactive properties of L. cylindrica seed press cake powder (LP) and its blends with tapioca flour (TF) at ratios of 30–70%. Techno-functional analyses included: hydration properties (water holding capacity, water absorption capacity, water absorption index, water solubility index, swelling power, oil absorption capacity); rheological characteristics; bioactive profiling through antioxidant assays (DPPH, ABTS, FRAP); and reducing sugar content determination. Meat substitute formulations were developed using an LP30/TF70 blend combined with coral lentils, red beet powder, and water, followed by a sensory evaluation and storage stability assessment. Pure L. cylindrica powder exhibited the highest water holding capacity (3.62 g H₂O/g) and reducing sugar content (8.05 mg GE/g), while tapioca flour showed superior swelling properties. The blends demonstrated complementary functional characteristics, with the LP30/TF70 formulation selected for meat substitute development based on optimal textural properties. The sensory evaluation revealed significant gender differences in acceptance, with women rating the product substantially higher than men across all attributes. The study successfully demonstrated the feasibility of transforming agricultural waste into a valuable functional ingredient, contributing to sustainable food production and representing the first comprehensive evaluation of L. cylindrica seed press cake for food applications. However, the study revealed limitations, including significant antioxidant loss during thermal processing (80–85% reduction); a preliminary sensory evaluation with limited participants showing gender-dependent acceptance; and a reliance on locally available tapioca flour, which may limit global applicability. Future research should focus on processing optimization to preserve bioactive compounds, comprehensive sensory studies with diverse populations, and an investigation of alternative starch sources to enhance the worldwide implementation of this valorization approach.

1. Introduction

In light of growing environmental concerns and the demand for sustainable food systems, the valorization of underutilized agri-food co-products—such as leaves, bark, and seed press cakes—has become a scientific and industrial priority. Luffa cylindrica (L.), also known as sponge gourd or loofa, offers notable potential in this context due to its nutritional and functional attributes [1]. A climbing annual plant from the Cucurbitaceae family, L. cylindrica is widely cultivated in subtropical regions, including South Asia, the Middle East, China, and the Americas [1,2]. Its young fruits, leaves, shoots, and flower buds are edible, while mature fruits develop a fibrous cellulose matrix that forms a natural sponge used in diverse applications such as exfoliation, cleaning, filtration, and packaging [2,3,4].

The fruit of Luffa cylindrica is a rich source of carbohydrates, vitamin C, and minerals, such as magnesium, calcium, sodium, potassium, iron, copper, zinc, and manganese [5], along with bioactive compounds like tannins, oxalates, phytic acid, saponins with immunomodulatory effects [6], glycosides, and flavonoids [7]. Its seeds are dull black, elliptical, and measure approximately 10–12 mm in length and 6–8 mm in width. They are edible and traditionally consumed roasted, similar to pumpkin seeds. Nutritionally, the seeds contain 33.55% protein, 29.51% carbohydrates, 22.17% fat, 6.47% fiber, and essential minerals (Ca, Zn, Mg, P) [8]. The seeds have demonstrated antiviral, anti-tumor, antioxidant, anti-inflammatory, and immunomodulatory activities [9]. Alkalase and tryptic protein hydrolysates support their potential in managing diabetes and hypertension [6]. The seed oil, dominated by linoleic acid (C18:2), exhibits antifungal, anti-inflammatory, and antitumor properties. It contains sterols (sitosterol, stigmasterol), tocopherols, phytol, hydrocarbons, glycolipids, and phospholipids, such as phosphatidylethanolamine [10], owing to its cytotoxicity toward skin cancer cells and its ability to inhibit protein synthesis in malignant tissues [11].

Luffa seed oil is currently produced only on a small, artisanal scale in regions, such as Africa and China, with limited use in cosmetics, pharmaceuticals, and to a lesser extent, food; its production is not recorded in FAO statistics. The seed press cake, a protein- and fiber-rich byproduct of oil extraction, has traditionally been used in agriculture, cosmetics, or biofuel applications; however, its valorization for human nutrition remains largely unexplored [12]. Introducing luffa seed press cake into the food sector presents an innovative opportunity to develop new, sustainable protein sources. This aligns with the increasing consumer demand for plant-based diets driven by health, environmental, and ethical concerns, particularly among younger populations [13]. Meat analogues made from plant proteins, including flours, isolates, and texturates, are being actively developed as alternatives to conventional meat, aiming to improve nutritional quality while reducing environmental and animal welfare impacts [14,15].

The aim of this study was to explore the potential of Luffa cylindrica seed press cake as a functional ingredient in meat substitute formulations by blending it with tapioca flour. A range of binary mixtures was evaluated to identify optimal proportions for application in alternative meat products. The focus was placed on the characterization of functional, rheological, and bioactive properties of the blends. This study is, to the best of our knowledge, the first to assess the combined techno-functional, viscometric, and antioxidant potential of L. cylindrica seed press cake and tapioca flour systems. The findings support the valorization of this agro-industrial byproduct as a sustainable food ingredient.

2. Materials and Methods

2.1. Raw Material

The raw material used in this study is a residue waste obtained after the transformation of the Luffa (L. cylindrica) seeds (LS—Figure 1D). This residue is often considered waste and is called Luffa seed press cake (Figure 1E). L. cylindrica seeds (LS) were collected by hand upon full maturity of the Luffa fruit (Figure 1A). The luffa fruits were grown to full maturity in the Gourbeyre commune, located in Guadeloupe (French West Indies) and harvested in February 2024. The LS was washed with distilled water, peeled to obtain the LS kernels and then dried at 50 °C for five hours in a food drier (Biosec, Tausereus, Camisano Vicentino, Italy). The dried seeds were ground to obtain oil, while the seed press cake was a byproduct. The seed press cake of L. cylindrica seeds was spread on a tray and dried for 24 h at room temperature in a closed, ventilated oven under controlled conditions. To ensure food safety, good processing practices were applied to promote effective dehydration. Microbiological analysis confirmed compliance with European safety standards, indicating that water activity and moisture levels were sufficiently low to prevent microbial growth and toxin formation. The dried seed press cake (LP) was vacuum-packed and stored for a maximum of 3 weeks.

Blends were prepared by combining LP and tapioca flour (TF) (Planteon, Borków Stary, Poland) at ratios of 30%/70%; 50%/50% and 70%/30% of LP to TF.

2.2. Characterization of Seed Press Cake of L. cylindrica Seeds Powder (LP), Tapioca Flour (TF) and Their Blends

2.2.1. Techno-Functional Characteristics

The water absorption capacity (WAC), oil absorption capacity (OAC), and hydrophilic/lipophilic index (HLI) were measured following the protocol of Nedviha and Harasym (2024), with slight modifications [16]. Briefly, the WAC was determined by mixing. For determination of the water absorption capacity (WAC), 3 g of a sample were mixed with 30 mL of distilled water in pre-weighed centrifuge tubes. To assess oil absorption capacity (OAC), an identical procedure was followed, replacing water with 30 mL of rapeseed oil. The mixtures were homogenized using a vortex mixer (Heidolph Reax, Schwabach, Germany) for 30 s, allowed to stand at room temperature for 10 min, and subsequently centrifuged at 3000× g (Thermo Fisher Scientific, Waltham, MA, USA). After centrifugation, the supernatants were carefully decanted. The remaining sediment, along with the tubes, was placed in a drying oven (Vindon Scientific, Rochdale, UK) at 50 °C for 25 min to remove residual moisture or oil. The WAC and OAC values were expressed as grams of liquid retained per gram of dry sample, based on Equation (1):

$$\mathbf{W}\mathbf{A}\mathbf{C}; \mathbf{O}\mathbf{A}\mathbf{C}={\displaystyle \frac{\mathbf{W}\mathbf{t}\mathbf{s}-\mathbf{W}\mathbf{t}}{\mathbf{W}\mathbf{s}}}$$

(1)

The hydrophilic/lipophilic index (HLI) was calculated as the ratio of water absorption capacity (WAC) to oil absorption capacity (OAC), serving as an indicator of the sample’s polarity-related binding preferences.

The water absorption index (WAI), water solubility index (WSI), and swelling power (SP) were assessed. Briefly, 3 g of the sample were suspended in 30 mL of distilled water in pre-weighed centrifuge tubes and subjected to thermal treatment at 90 °C for 10 min in a water bath (MLL147, AJL Electronics, Kraków, Poland). After allowing the suspensions to cool to ambient temperature, samples were centrifuged at 3000× g for 10 min. The sediment was weighed to determine water retention, while the supernatant was transferred to pre-weighed Petri dishes and dried in a laboratory oven (SML, Zalmed, Łomianki, Poland) at 110 °C for 24 h to quantify solubilized solids. WAI, WSI, and SP were then calculated and expressed as g/g or %, as appropriate.

Foaming properties were measured based on a modified procedure adapted from Kaushal et al. (2012) [17]. To determine foaming capacity (FC), 1 g of the test material was dispersed in 50 mL of distilled water in a graduated cylinder. The mixture was manually agitated in a vertical motion for 5 min and the initial and post-foaming volumes were recorded. FC was expressed as milliliters of foam per gram of dry matter. Foam stability (FS) was evaluated by monitoring the decrease in foam volume after 1 h at room temperature, with results expressed as a percentage of the initial foam volume. The calculation was based on Equation (2), as follows:

FS [%] = (V₁H/V₀) × 100

(2)

where V₁H represents the foam volume after 1 h and V₀ denotes the initial foam volume.

Emulsifying activity (EA) and emulsifying stability (ES) were assessed following the method of Kiiru et al. (2024), with slight modifications [18]. The procedure involved mixing the sample with predetermined amounts of water and rapeseed oil, followed by homogenization to form an emulsion. The emulsified phase was separated by centrifugation; the volume of the emulsion layer was measured to determine EA. To evaluate ES, the emulsions were subjected to mild heating and allowed to rest at room temperature, after which the retained emulsion volume was recorded.

The determination of dietary fiber content (both insoluble and soluble fractions) was conducted following a modified enzymatic-gravimetric protocol. The procedure was adapted from AOAC Method 991.43 and AACC Method 32-07.01, incorporating elements from AACC Methods 32-05.01 (total dietary fiber) and the dedicated method for soluble and insoluble fiber determination with additional adjustments [19].

The pasting properties of each sample were measured according to AACC 76-21.01 using a rapid viscoanalyzer (RVA 4500, Perten Instruments, Sydney, Australia). In particular, the parameters studied were as follows: maximum viscosity (PV); minimum viscosity (TV); reduction in viscosity (BD); final viscosity (FV); retrogradation viscosity (SB); and time to maximum viscosity (PT). The samples were subjected to controlled cooking and brewing conditions, beginning with 2 min at 50 °C, then an increase from 50 °C to 95 °C at a heating rate of 5 °C/min, held at 95 °C for 5 min. Then, the samples were cooled at 10 °C/min to 50 °C, with a final holding phase at 50 °C for 4 min.

2.2.2. Reducing Sugars and Antioxidant Characteristics

For the determination of bioactivity, an extract was obtained by combining 1 g of the sample with 10 mL of ethanol. The mixtures were subjected to continuous agitation at ambient temperature using a rotary shaker (MX-RD PRO, Chemland, Stargard, Poland) for 2 h. Following extraction, samples were centrifuged (MPW-350, MPW MED. INSTRUMENTS, Warsaw, Poland) at 3500× g for 15 min at 4 °C. The obtained supernatants were stored at 8 °C until further analysis.

The content of reducing sugars was determined by a modified dinitrosalicylic acid (DNS) assay following the method of Nedviha and Harasym (2024) [16]. A volume of 1 mL of DNS reagent was combined with 1 mL of each extract; the mixtures were heated in a boiling water bath for 5 min. After cooling to room temperature, absorbance was recorded at 535 nm. A glucose calibration curve (100–800 µg/mL) was used to quantify results, expressed as mg glucose equivalents per gram of dry sample (gDW).

The antiradical potential was assessed using the ABTS•+ method. Each sample (0.0204 mL) was mixed with 1.0 mL of a diluted ABTS•+ radical solution, prepared according to Sridhar and Charles (2019) [20]. Absorbance was measured at 734 nm, precisely 10 s after mixing. A Trolox standard curve (100–800 µmol/L) was used; results were presented as mg of Trolox equivalents (TE) per gram of dry sample.

To evaluate free radical scavenging capacity, the DPPH assay was used [20]. Briefly, 0.0345 mL of extract was mixed with 1 mL of a 0.1 mM methanolic DPPH solution. The mixtures were incubated for 20 min at room temperature; absorbance was measured at 517 nm. Trolox served as a reference standard (100–800 µmol/L); the antioxidant capacity was expressed as mg TE/gDW.

The ferric reducing antioxidant power (FRAP) assay was also employed to assess the electron-donating potential of the samples. Each extract (0.0035 mL) was added to 1 mL of freshly prepared FRAP reagent, as described by Bhajan et al. (2023) [21]. Following a 15 min incubation at 36 °C, absorbance was measured at 593 nm. Calibration was performed using ferrous sulfate (100–800 µmol/L); the results are expressed as mg of FeSO₄ equivalent per gram of dry matter.

Additionally, the molecular characteristics of the samples were investigated using Fourier transform infrared spectroscopy (FTIR). Spectra were acquired using a Nicolet 6700 FT-IR spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), equipped with an attenuated total reflectance (ATR) diamond crystal. Each measurement was based on 64 scans collected across the spectral range of 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹. Background subtraction was performed against air to ensure signal accuracy.

2.2.3. Microbial Characterization

The microbiological evaluation was conducted by suspending 10 g of the sample in 90 mL of buffered peptone water (BIOKAR Diagnostics, Allonne, France, 25.5 g/L). The mixture was homogenized using a Stomacher Mix-1 system (AES Laboratoire, Tours, France). Serial dilutions were prepared and plated on specific media for the enumeration of microbial groups.

Total mesophilic aerobic bacteria (TMC) were assessed using plate count agar (PCA), incubated at 30 °C for 72 h, in accordance with ISO 4833-1:2003b [22]. Coliform bacteria were enumerated on violet-red bile agar (VRBA), with incubation at 37 °C for 24 h, following ISO 4832:2006 [23]. The presence of Escherichia coli was verified using TBX (Tryptone Bile X-glucuronide) agar under aerobic conditions at 37 °C for 24 h, following ISO 16649-2 [24]. Yeasts and molds were quantified using Sabouraud dextrose agar supplemented with chloramphenicol (V08-059), with plates incubated at 25 °C for 5 days. All microbial counts were reported as colony-forming units per gram of dry Luffa seed press cake (CFU/g). Analyses were performed in triplicate.

2.3. Meat Substitute Formulations

2.3.1. Formulations

Five ingredients were necessary for the formulation: a combination of LP and TP, coral lentils used to add protein content in the final product, red beet powder used as a colorant to approximate the color of meat, and water, allowing for a homogenous texture in the final product. The LP was from Guadeloupe, whereas all other ingredients were from a local market in Poland.

First, a better proportion of LP and TP was selected according to the results obtained by the texture characterization. When the proportion of Luffa was higher, the dough was more friable and less consistent. LP30/TF70 had a similar texture to meat and was a better combination of LP/TF. Then, a pre-study was performed according to an experimental design to determine the amount of water to add. Five trials were tested. The 5 trials of paste, including for each LP30/TF70 and a precise amount of water, were placed in the freezer at −20 °C for 48 h and finally defrosted at room temperature (20 °C) for 1 h. After a visual observation, the proportion of 70% water added to the combination LP30/TP70 had a texture similar to that of cooked pork meat.

Other ingredients, such as coral lentil (Bionaturapol, Szeligi, Poland) and dehydrated beets (Planteon, Borków Stary, Poland), both reduced to powder through a millet grain, were added to the dough to finally obtain a pink meat color similar to water-cooked pork (control). The color between the meat substitute and pork was compared using the colorimeter.

All ingredients were well mixed in a blender, then heated in a bain-marie at 90 °C for 7 to 8 min to achieve a homogeneous and thickened paste. The final formulation included 50 g of LP30/TF70, 2.5 g of coral lentils, 2.5 g of red beets, and 150 g of water. Meatball substitutes were marinated in a sauce containing rapeseed oil, pepper, garlic, salt and paprika spices to improve the organoleptic characteristics for 30 min at room temperature. The meat substitutes were cooked in an airfryer (Ambiano, Hamburg, Germany) at 120 °C for 30 min.

2.3.2. Physicochemical and Sensory Analyses of Meat Substitute Formulation

The cutting force was measured using a texture analyzer (AXIS model FC200STAV50, AXIS, Gdańsk, Poland). Each specimen was supported at two ends, placed 8 cm apart, while a central force was applied via a wedge descending at a constant speed. The recorded hardness corresponded to the peak force (N) required to break the sample. Measurements were performed following storage periods of 2, 7, and 12 days at 4 °C.

Color attributes were determined using a tristimulus colorimeter operating under a D65 light source (CR-310 colorimeter, Konica Minolta, Ramsey, NJ, USA). The L* (lightness), a* (redness), and b* (yellowness) values were obtained as the mean of six individual readings collected from three distinct regions of the sample surface to ensure representativeness. These analyses were conducted after 2, 7, and 12 days of storage at −20 °C.

Twenty-four participants from the Wroclaw University (18–21 years old; 14 females and 10 male) were recruited. All participants provided informed consent before participation. A hedonic test was performed as a preliminary screening, in which each consumer evaluated the formulation retained and a control sample. The control sample was a commercial meat substitute from a supermarket in Wroclaw. Samples were distributed using a random presentation. Each meat was served on a white plate, labeled with randomly assigned three-digit codes. Consumers rated the overall acceptability color, aroma, taste, texture in mouth on a 5-point scale (0 = dislike very much, 5 = like very much). Between evaluating the samples, participants were required to rinse their mouths with mineral water and eat unsalted toast. Due to the limited number of participants (n = 24), no normality test was performed, as the sample size was insufficient for a reliable assessment of data distribution. The sensory data represent ordinal measurements on a 5-point hedonic scale, which typically does not follow normal distribution patterns, particularly with small sample sizes. Given these constraints, the sensory evaluation should be considered a preliminary screening rather than a definitive consumer acceptance study. The results are presented as descriptive statistics (means and frequency distributions) and should be interpreted with caution. Statistical comparisons were not performed due to the insufficient sample size for robust inferential analysis. Future research should incorporate larger sample sizes (n ≥ 50–100 participants per demographic group) to enable appropriate statistical testing and more generalizable conclusions about consumer acceptance patterns.

The sensory session took place in the sensory laboratory located in the Department of Biotechnology and Food Analysis of the Wroclaw University of Economics and Business.

2.4. Statistical Analysis

All analyses were performed in triplicate and results are reported as mean values. A one-way analysis of variance (ANOVA) and Tukey test were carried out to determine statistically significant differences (p < 0.05).

3. Results and Discussion

3.1. Characterization of Seed Press Cake of L. cylindrica Seeds Powder (LP), Tapioca Flour (TF) and Their Blends

3.1.1. Techno-Functional Characteristics

The absorption characteristics of L. cylindrica seed press cake powder and blends are shown in

Table 1. Absorption characteristics of L. cylindrica seed press cake powder and blends.

Sample	WHC [g H2O/g d.b.]	WAC [g H2O/g d.b.]	OAC [g oil/g d.b.]	HLI [g H2O/g oil]	WAI [g H2O/g d.b.]	WSI [g/100 g d.b.]	SP [g H2O/g d.b.]
LP 100%	3.62 ± 0.44 d	2.88 ± 0.02 e	1.69 ± 0.02 a	1.75 ± 0.03 e	3.59 ± 0.09 a	7.45 ± 0.64 d	3.25 ± 0.08 a
LP30/TF70	3.21 ± 0.31 d	2.10 ± 0.01 b	1.81 ± 0.08 b	1.22 ± 0.01 b	7.83 ± 0.23 d	3.37 ± 0.30 b	7.08 ± 0.28 d
LP50/TF50	2.59 ± 0.20 c	2.18 ± 0.02 c	1.82 ± 0.05 b	1.29 ± 0.04 c	6.33 ± 0.45 c	5.49 ± 0.44 c	6.24 ± 4.68 d
LP70/TF30	2.21 ± 0.11 b	2.47 ± 0.13 d	1.74 ± 0.03 b	1.51 ± 0.09 d	5.05 ± 0.10 b	6.07 ± 0.27 c	4.70 ± 0.35 b
TF 100%	1.76 ± 0.12 a	1.79 ± 0.03 a	2.08 ± 0.06 c	0.97 ± 0.01 a	10.90 ± 0.15 e	0.88 ± 0.38 a	9.74 ± 0.19 e

LP = luffa powder; TF = tapioca flour; WHC = water holding capacity; WAC = water absorption capacity; OAC = oil absorption capacity; HLI = hydrophilic/lipophilic index; WAI = water absorption index; WSI = water solubility index; SP = swelling power. Lowercase letters indicate significant differences in columns at p = 0.05.

.

The water holding capacity (WHC) values ranged from 1.76 ± 0.12 to 3.62 ± 0.44 g H₂O/g d.b., with pure L. cylindrica powder (LP 100%) exhibiting the highest value at 3.62 ± 0.44 g H₂O/g d.b., while pure tapioca flour (TF 100%) showed the lowest at 1.76 ± 0.12 g H₂O/g d.b. The blends demonstrated intermediate values, showing a clear decreasing trend with increases in luffa powder content. Water absorption capacity (WAC) exhibited values between 1.79 ± 0.03 and 2.88 ± 0.02 g H₂O/g d.b., with LP 100% the highest WAC at 2.88 ± 0.02 g H₂O/g d.b., while TF 100% showed the lowest value at 1.79 ± 0.03 g H₂O/g d.b. The blends showed a progressive decrease in WAC with increasing tapioca content. These results, showing that Luffa powder (LP) exhibits the highest WHC and WAC, can be attributed to its fibrous and highly porous structure, which provides a large surface area and abundant hydrophilic sites for water interaction and absorption. This finding aligns with studies on other high-fiber plant residues, where the presence of hydroxyl groups and open matrix architecture significantly enhances water retention capacity [25]. Similar water absorption behavior has been reported for various dietary fiber sources, such as okra peel, rice bran, and spent coffee grounds, all of which demonstrate strong WAC due to their microstructural features [26,27]. In the present study, Luffa cylindrica powder contained 59.72 g/100 g of insoluble dietary fiber (IDF) and 3.89 g/100 g of soluble dietary fiber (SDF), while the meat substitute formulation contained 16.29 g/100 g IDF and 2.92 g/100 g SDF. These values support the observed water absorption capacity of both materials.

However, the simultaneous decrease in WHC and increase in WAC with increasing LP concentrations can be explained by the distinction between water absorption and water retention. While LP is highly efficient at absorbing water into its structure, it lacks the ability to retain that water under external forces such as centrifugation or mechanical stress. WHC reflects not only water uptake, but also the strength and integrity of the surrounding matrix in holding the water in place.

As the proportion of LP increases and that of tapioca flour (TF) decreases, the composite matrix becomes more fragmented and fibrous, reducing its cohesiveness and structural integrity. Tapioca starch, in contrast, is well known for its gel-forming ability, contributing to a continuous and stable gel network that effectively traps and retains water [28]. When the TF content is lowered, this gel matrix is weakened or disrupted and the LP-dominated structure does not compensate sufficiently. As a result, even though more water is absorbed due to LP’s porosity (high WAC), less water is physically held within the matrix, leading to a lower WHC. These findings underscore the critical role of matrix cohesion and polymer interactions in determining the functional hydration properties of plant-based formulations.

The oil absorption capacity (OAC) remained relatively consistent across samples, ranging from 1.69 ± 0.02 to 2.08 ± 0.06 g oil/g d.b. Pure tapioca flour exhibited the highest OAC at 2.08 ± 0.06 g oil/g d.b., while pure L. cylindrica powder showed the lowest at 1.69 ± 0.02 g oil/g d.b. This trend suggests that TF’s starchy matrix, particularly upon gelatinization, may provide more nonpolar binding sites or physical spaces for oil entrapment. In contrast, although LP is fibrous and porous, its higher polarity and hydrophilicity limit its oil affinity. This is consistent with earlier reports, where protein-rich or starch-rich matrices showed higher OAC than fiber-dominant materials [29]. Additionally, the increased surface polarity of LP may hinder effective oil retention. The hydrophilic/lipophilic index (HLI) values ranged from 0.97 ± 0.01 to 1.75 ± 0.03 g H₂O/g oil with the highest HLI at 1.75 ± 0.03 for LP 100%, while pure tapioca flour exhibited the lowest, at 0.97 ± 0.01. The blends displayed decreasing HLI values with increases in tapioca content, indicating a shift toward a greater lipophilic character. LP-rich samples exhibited higher HLI values (up to 1.75), consistent with the hydrophilic nature of cell wall polysaccharides in Luffa (e.g., cellulose, hemicellulose). The observed changes support the view that LP contributes to hydrophilic functionality, while TF introduces more hydrophobic or amphiphilic behavior, which is relevant for emulsion-type meat analogues [30].

The water absorption index (WAI) showed values ranging from 3.59 ± 0.09 to 10.90 ± 0.15 g H₂O/g d.b.; TF 100% again had the highest WAI at 10.90 ± 0.15, indicating the strong water uptake and swelling behavior typical of gelatinized starches. Pure L. cylindrica powder showed the lowest WAI at 3.59 ± 0.09. The blends exhibited increasing WAI with higher tapioca contents. The LP-rich samples showed lower WAI values, consistent with their limited swelling due to the rigid, non-gelatinizing nature of dietary fiber matrices. The water solubility index (WSI) varied considerably, from 0.88 ± 0.38 to 7.45 ± 0.64 g/100 g d.b. For soluble residuals, pure tapioca flour showed the lowest WSI at 0.88 ± 0.38, while LP 100% exhibited the highest at 7.45 ± 0.64, suggesting partial solubilization of some fiber components or fiber–starch interactions that promote breakdown into smaller soluble fractions during processing. Comparable patterns have been reported in blends of soluble and insoluble fibers with starches [31]. The swelling power (SP) displayed substantial variations, ranging from 3.25 ± 0.08 to 9.74 ± 0.19 g H₂O/g d.b.; TF 100% exhibited the highest swelling power at 9.74 ± 0.19, while pure L. cylindrica powder showed the lowest at 3.25 ± 0.08. The blends demonstrated intermediate values that increased with the tapioca contents. SP was significantly higher in TF-dominated samples, confirming the gel-forming and swelling properties of tapioca starch, which exhibits extensive granule swelling upon heating [28]. In contrast, LP, being composed mostly of non-gelatinizing, lignocellulosic material, contributes little to swelling, resulting in lower SP values in LP-rich formulations.

These results indicate that the addition of tapioca flour significantly influenced the functional properties of L. cylindrica seed press cake powder, generally resulting in decreased water holding and absorption capacities but also in an enhanced swelling power and water absorption index, suggesting complementary functional characteristics that could be optimized for specific food applications.

Figure 2 presents the foaming and emulsifying characteristics of L. cylindrica seed press cake powder and its blends with tapioca flour.

The foaming capacity (FC) values demonstrated a gradual decrease with increases in tapioca flour content. Pure L. cylindrica powder (LP 100%) exhibited the lowest foaming capacity, at approximately 53%, while pure tapioca flour (TF 100%) showed the highest at around 58%, consistent with the known foaming properties of starches, particularly those rich in amylopectin such as tapioca [32]. The blends displayed intermediate values, indicating a linear relationship between L. cylindrica content and foaming capacity. Tapioca’s gelatinization and viscosity-enhancing properties can effectively trap air, leading to increased foam formation. The foam stability (FS) values remained relatively consistent across all samples, ranging narrowly from 102% to 112%. The highest FS was observed for the LP30/TF70 blend (112%), followed by LP50/TF50 (109%), and pure Luffa cylindrica powder (105%). The lowest FS was recorded for LP70/TF30 (102%). Pure tapioca flour exhibited a value of approximately 110%, indicating good foam-stabilizing capacity on its own. This suggests that the partial inclusion of fibrous LP (30%) might reinforce the matrix, helping to stabilize air bubbles within the gel network formed by TF. Similar effects have been observed when combining starches with moderate levels of insoluble dietary fiber, which can enhance the mechanical strength of the foam structure without excessively disrupting air entrapment [33].

Emulsifying activity (EA) and emulsion stability (ES) are essential parameters in food systems, particularly for plant-based meat analogues where oil-in-water emulsions contribute to texture and mouthfeel. The emulsifying activity (EA) showed more pronounced variations among the samples. The lowest emulsifying activity was seen in TF 100%, at about 5.4%, while pure L. cylindrica powder showed a slightly higher EA value of about 8.7%. The blends demonstrated higher EA values, ranging between 10.9% and 15.8%, where the highest EA was observed for the LP70/TF30 blend, indicating that a higher proportion of Luffa cylindrica powder supports the formation of emulsions. This can be attributed to its fibrous, porous structure and potential protein residues, which increase surface activity and oil droplet dispersion [34]. The emulsion stability (ES) values ranged from approximately 6% to 12%. Interestingly, the highest ES was found in the LP 100% sample, despite its moderate EA. This suggests that LP possesses inherent properties, such as a highly porous and fibrous matrix, that allow it to stabilize emulsions over time, likely by forming a strong interfacial barrier that resists coalescence. The emulsifying stability (ES) of the blends containing 30% and 50% luffa powder was not significantly compromised, maintaining relatively high values (respectively 10.2% and 10.45%). In contrast, the blend with 70% LP exhibited the lowest ES (5.95%), indicating that a higher proportion of fibrous material may disrupt the emulsion structure more substantially. The uniformity in emulsion stability suggests that both components and their combinations maintain stable emulsions once formed, regardless of the initial emulsifying activity differences. The complementary nature of these ingredients suggests the potential for optimizing blend ratios based on specific application requirements in food formulation.

Figure 3 illustrates the pasting profiles of L. cylindrica seed press cake powder blends with tapioca flour over a controlled temperature cycle, revealing distinct viscosity development patterns during the heating, holding, and cooling phases.

Pure L. cylindrica powder (LP 100%) demonstrated no viscosity development throughout the entire temperature profile, maintaining a baseline near zero mPa·s across all phases of the analysis. This confirms the absence of starch or other paste-forming components in the seed press cake powder. TF 100% exhibited the most pronounced pasting behavior, with the initial viscosity remaining low until approximately 65 s, when rapid viscosity development began. The viscosity increased dramatically to reach a peak of approximately 3300 mPa·s at around 81 s, corresponding to a temperature of approximately 85 °C. During the holding phase at 95 °C, the viscosity decreased to approximately 2400 mPa·s, indicating substantial breakdown under continuous shear. Upon cooling, the viscosity increased moderately to reach a final value of approximately 3000 mPa·s at 193 s, demonstrating significant setback.

The LP30/TF70 blend showed intermediate pasting behavior, with viscosity development beginning around 57 s. The LP50/TF50 blend demonstrated further reduced pasting characteristics, with gradual viscosity development starting around 65 s. The LP70/TF30 blend exhibited minimal pasting properties, with very gradual viscosity development beginning at around 73 s. The profile showed no distinct peak, with viscosity gradually increasing to approximately 400 mPa·s during the heating phase. The viscosity remained stable during holding and showed only a slight increase during cooling to reach a final value of approximately 500 mPa·s, demonstrating a minimal setback tendency.

These pasting profiles clearly demonstrate the dilution effect of L. cylindrica seed press cake on tapioca starch functionality, with increasing L. cylindrica contents resulting in progressively reduced peak viscosity, breakdown, and setback values [16,33]. The results suggest that L. cylindrica seed press cake acts as a non-gelling diluent that interferes with starch granule swelling and paste formation, potentially through physical hindrance or competition for available water.

The dilution of pasting properties observed with increases in the Luffa cylindrica contents can be attributed to its intrinsic composition, particularly to the absence of starch and the presence of fibrous cell wall materials that interfere with starch gelatinization. Studies have confirmed that L. cylindrica is rich in cellulose and hemicellulose, which can restrict water availability and physically obstruct starch granule swelling [35]. This aligns with our finding that pure L. cylindrica (LP 100%) showed no measurable viscosity under thermal treatment, indicating the lack of gelatinizable constituents. Moreover, L. cylindrica has demonstrated hydrophilic behavior due to its porous structure, which can increase water holding but not contribute to pasting viscosity, further supporting its role as a non-gelling diluent [36]. The observed increase in pasting temperature and peak time in blends with higher L. cylindrica contents may reflect delayed starch gelatinization due to water competition and the formation of fiber–starch interactions, a phenomenon also observed in bio-composite studies using Luffa cellulose [37].

Table 2. Pasting properties of L. cylindrica seed press cake powder and blends.

Sample	Peak Viscosity	Trough Viscosity	Breakdown Viscosity	Final Viscosity	Setback Viscosity	Peak Time	Pasting Temperature
	[mPa·s]					[min]	[°C]
LP 100%	— a	— a	— a	— a	— a	— a	— a
LP30/TF70	3073 ± 98 d	1540 ± 65 d	1534 ± 34 d	2594 ± 85 d	1055 ± 150 d	4.2 ± 0.00 b	74.33 ± 0.11 c
LP50/TF50	1415 ± 294 c	987 ± 142 c	428 ± 152 c	1516 ± 200 c	528 ± 60 c	4.7 ± 0.17 c	73.92 ± 0.60 c
LP70/TF30	475 ± 20 b	447 ± 12 b	28 ± 8 b	555 ± 19 b	108 ± 7 b	6.6 ± 0.13 d	75.90 ± 0.07 d
TF 100%	6644 ± 45 e	2431 ± 28 e	4213 ± 72 e	3870 ± 234 e	1439 ± 206 e	3.7 ± 0.00 b	72.65 ± 0.07 b

LP = luffa powder; TF = tapioca flour; PV = peak viscosity; TV = trough viscosity; BD = breakdown viscosity; FV = final viscosity; SB = setback viscosity; PT = peak time. Lowercase letters indicate significant differences in columns at p = 0.05—indicates no measurable viscosity parameters.

presents the pasting properties of L. cylindrica seed press cake powder and its blends with tapioca flour, revealing significant variations in viscometric characteristics across different formulations.

Pure L. cylindrica powder LP 100% exhibited no measurable viscosity parameters, indicating an absence of pasting properties under the test conditions. This suggests that L. cylindrica seed press cake alone lacks the starch content necessary for gelatinization and paste formation.

The peak viscosity (PV) values varied considerably among samples containing tapioca flour, ranging from 475 ± 20 to 6644 ± 45 mPa·s. Pure tapioca flour (TF 100%) demonstrated the highest peak viscosity, at 6644 ± 45 mPa·s, while LP70/TF30 showed the lowest, at 475 ± 20 mPa·s. The blends exhibited a clear inverse relationship between L. cylindrica content and peak viscosity. Trough viscosity (TV) followed a similar pattern to peak viscosity, ranging from 447 ± 12 to 2431 ± 28 mPa·s. Pure tapioca flour showed the highest trough viscosity, while LP70/TF30 exhibited the lowest at 447 ± 12 mPa·s.

The breakdown viscosity (BD), representing paste stability during heating, showed dramatic variations from 28 ± 8 to 4213 ± 72 mPa·s. Pure tapioca flour exhibited the highest breakdown at 4213 ± 72 mPa·s, indicating lower paste stability, while LP70/TF30 showed minimal breakdown at 28 ± 8 mPa·s, suggesting excellent paste stability. The final viscosity (FV) ranged from 555 ± 19 to 3870 ± 234 mPa·s. Pure tapioca flour displayed the highest final viscosity, at 3870 ± 234 mPa·s, while LP70/TF30 showed the lowest, at 555 ± 19 mPa·s. The blends exhibited decreasing final viscosity with increasing L. cylindrica content.

Setback viscosity (SB), indicating retrogradation tendency, varied from 108 ± 7 to 1439 ± 206 mPa·s. Pure tapioca flour showed the highest setback, at 1439 ± 206 mPa·s, suggesting strong retrogradation potential, while LP70/TF30 exhibited minimal setback, at 108 ± 7 mPa·s.

The peak time ranged from 3.7 ± 0.00 to 6.6 ± 0.13 min. Pure tapioca flour reached peak viscosity fastest, at 3.7 ± 0.00 min, while LP70/TF30 required the longest time, at 6.6 ± 0.13 min. The pasting temperature exhibited variations from 72.65 ± 0.07 to 75.90 ± 0.07 °C. Pure tapioca flour showed the lowest pasting temperature, at 72.65 ± 0.07 °C, while LP70/TF30 demonstrated the highest, at 75.90 ± 0.07 °C. The absence of measurable viscosity in pure Luffa cylindrica powder confirms its lack of gelatinizable starch and aligns with previous studies demonstrating that fibrous, low-starch materials suppress paste formation under thermal shear conditions [38]. The progressive decline in peak, breakdown, and final viscosities with increasing L. cylindrica content reflects the well-documented water competition and physical interference effects of insoluble dietary fibers in starch matrices [39]. In these blended systems, fiber particles hinder the swelling of starch granules, leading to delayed gelatinization and reduced peak viscosity—a phenomenon also observed with oat and rice flour mixtures.

The sharp decrease in breakdown viscosity at high L. cylindrica ratios suggests improved paste thermal stability, as the starch network is reinforced or constrained by fiber presence, resisting disintegration under heating [40]. Similarly, lower setback viscosities indicate suppressed retrogradation, which has been linked to reduced amylose mobility in the presence of fibrous components [41]. The increased pasting temperature and peak time in the fiber-rich blends point to restricted water diffusion, requiring more energy for starch gelatinization [42].

These results clearly demonstrate that the incorporation of L. cylindrica seed press cake powder significantly modifies the pasting properties of tapioca flour, generally resulting in reduced viscosity parameters, enhanced paste stability, and increased pasting temperatures, suggesting its potential as a viscosity modifier in food formulation applications. L. cylindrica seed press cake acts as a rheological modifier by diluting starch and impeding its gelatinization and retrogradation, confirming its value as a functional, non-gelling ingredient in starch-based systems [43].

3.1.2. Reducing Sugars and Antioxidant Characteristics

Table 3. Reducing sugars content and antioxidant capacity of L. cylindrica seed press cake powder and blends.

Sample	Reducing Sugars [mg GLE/g DM]	DPPH [mg TE/g DM]		ABTS [mg TE/g DM]		FRAP [mg FeSO4/g DM]
		EtOH	H2O	EtOH	H2O	EtOH	H2O
LP 100%	8.05 ± 0.68 d	4.58 ± 0.18 a	4.25 ± 0.25 a	4.56 ± 0.04 a	3.67 ± 0.70 a	0.101 ± 0.007 b	0.126 ± 0.016 c
LP30/TF70	5.32 ± 1.31 b	5.68 ± 0.07 c	4.55 ± 0.51 a	5.26 ± 0.02 d	5.72 ± 0.33 d	0.046 ± 0.006 a	0.090 ± 0.023 b
LP50/TF50	4.13 ± 0.37 a	5.15 ± 0.05 b	4.91 ± 0.02 a	5.19 ± 0.07 c	5.11 ± 0.01 c	0.053 ± 0.015 a	0.050 ± 0.000 a
LP70/TF30	5.10 ± 0.92 b	5.04 ± 0.13 b	4.41 ± 0.15 a	5.06 ± 0.07 b	4.66 ± 0.12 b	0.070 ± 0.001 ab	0.109 ± 0.010 bc
TF 100%	6.90 ± 0.20 c	5.66 ± 0.13 c	6.27 ± 0.06 b	6.11 ± 0.18 e	6.33 ± 0.06 e	0.072 ± 0.038 ab	0.041 ± 0.005 a
sample		***		***		*
solvent		**		ns		ns
sample * solvent		**		*		ns

LP = luffa powder; TF = tapioca flour. Mean values with different lowercase letters imply significant differences between means in rows at p < 0.05. Second-order interaction analysis *—p < 0.05; **—p < 0.01; ***—p < 0.001, ns—nonsignificant (p > 0.05).

presents the reducing sugars content and antioxidant capacity of L. cylindrica seed press cake powder and its blends with tapioca flour, evaluated through multiple assays using both ethanolic and aqueous extraction methods.

The reducing sugar content showed significant variations among samples, ranging from 4.13 ± 0.37 to 8.05 ± 0.68 mg glucose equivalent/g dry matter. Pure L. cylindrica powder (LP 100%) exhibited the highest reducing sugar content at 8.05 ± 0.68 mg GE/g DM, significantly higher than all other samples.

It has been confirmed that simple sugars can also lead to the formation of Maillard reaction products in the presence of amino acids at lower temperatures [44,45]. However, since the mixtures of luffa powder and tapioca flour were not subjected to heat treatment at this stage, the formation of Maillard reaction products was not expected.

It has been shown that the total amino acid content in Luffa cilindrica can be as high as 32.1 g/100 g. It is indicated that the amino acid found in the largest amount in this raw material is glutamic acid [46].

The DPPH radical scavenging activity in ethanolic extracts ranged from 4.58 ± 0.18 to 5.68 ± 0.07 mg Trolox equivalent/g DM. LP 100% showed the lowest activity at 4.58 ± 0.18 mg TE/g DM; the following blends exhibited higher values: LP30/TF70 and TF 100%, both at 5.68 ± 0.07 and 5.66 ± 0.13 mg TE/g DM, respectively.

Tapioca is a product obtained from cassava root (Manihot esculenta), which, in terms of composition, is almost pure starch; therefore, the content of antioxidant components in this raw material is relatively low [47,48].

However, it has been confirmed that cassava root, which is the source of tapioca, contains antioxidants such as ascorbic acid, flavonoids, tannins, and saponins [49]. Perhaps during the process of processing cassava into tapioca, a certain amount of antioxidant compounds was retained in the raw material used; thus, adding this raw material to Luffa Powder may have resulted in increased antioxidant activity [50].

In aqueous extracts, the DPPH values ranged from 4.25 ± 0.25 to 6.27 ± 0.06 mg TE/g DM, with pure tapioca flour showing the highest activity at 6.27 ± 0.06 mg TE/g DM and pure L. cylindrica powder the lowest, at 4.25 ± 0.25 mg TE/g DM.

Previous studies have shown that potentially present antioxidants in tapioca, such as tannins (especially those with a low degree of polymerization) and saponins, are also highly soluble in ethanol and water–ethanol solutions [51,52]. Hence, the values for antioxidant activity measured by the DPPH and ABTS methods in the ethanol extracts of luffa powder and tapioca flour mixtures are probably higher than in the aqueous extracts.

Higher values of antioxidant activity measured by the DPPH and ABTS methods in the aqueous extract than in the ethanol extract for tapioca flour may be due to the several times higher solubility in water of L-ascorbic acid, which is also a potential antioxidant present in tapioca [53].

The ABTS radical scavenging capacity in ethanolic extracts demonstrated values from 4.56 ± 0.04 to 6.11 ± 0.18 mg TE/g DM. Pure tapioca flour exhibited the highest activity, at 6.11 ± 0.18 mg TE/g DM, while pure L. cylindrica powder showed the lowest, at 4.56 ± 0.04 mg TE/g DM. In the aqueous extracts, ABTS values ranged from 3.67 ± 0.70 to 6.33 ± 0.06 mg TE/g DM, with pure tapioca flour again showing the highest activity, at 6.33 ± 0.06 mg TE/g DM and pure L. cylindrica powder the lowest, at 3.67 ± 0.70 mg TE/g DM.

The reducing activity values (measured by FRAP) in the ethanol extracts ranged from 0.046 to 0.101 mg FeSO₄/g DM. Pure L. cylindrica powder showed the highest reducing power, at 0.126 mg FeSO₄/g DM, while the LP30/TF70 blend showed the lowest reducing activity, at 0.046 mg FeSO₄/g DM. The remaining samples showed intermediate values: TF 100% at 0.072; LP70/TF30 at 0.070; and LP50/TF50 at 0.053 mg FeSO₄/g DM. In aqueous extracts, FRAP values ranged from 0.041 to 0.126 mg FeSO₄/g DM, with pure L. cylindrica powder showing the highest value, at 0.126 mg FeSO₄/g DM and pure tapioca flour showing the lowest reducing activity, at 0.041 mg FeSO₄/g DM.

The lower values of reducing substances in the mixtures of L. cylindrica powder and tapioca flour were probably due to the overall low content of reducing substances in pure tapioca flour compared to pure L. cylindrica powder.

Studies confirm that the bioactive compounds contained in tapioca, which have the ability to donate electrons (such as carotenoids or polyphenols) and thus reduce iron (III) ions to iron (II) ions, occur in much smaller amounts than in L. cylindrica powder [54,55].

The statistical analysis revealed significant interaction effects between the sample type and extraction solvent for the DPPH (p < 0.01) and FRAP (p < 0.05) assays, while ABTS showed significant solvent effects (p < 0.01) but no significant sample–solvent interaction. The sample type alone showed no significant effect on antioxidant activities across all assays.

These results indicate that L. cylindrica seed press cake powder possesses notable antioxidant properties, particularly in terms of ferric reducing power, while the addition of tapioca flour generally enhanced radical scavenging activities in both the DPPH and ABTS assays. The extraction solvent significantly influenced the recovery of antioxidant compounds, with varying effects, depending on the specific assay method employed.

It can be assumed that the tapioca flour used to enrich luffa powder largely retained its initial antioxidant content during the main cassava root processing processes such as cleaning, dehydration and drying. As a result, the prepared and tested mixtures of these two raw materials, such as luffa powder and tapioca flour, present significant potential as a raw material for the preparation of functional products with significant antioxidant properties.

The strong broad band, at ca. 3280 cm⁻¹ in the Luffa cylindrica seed press cake and its blends with cassava flours (Figure 4), correspond to the free hydroxyl groups’ stretching υ(OH) mode and those involved in the intra- and inter-molecular hydrogen bonds (HB). The spectra show several distinct absorption regions characterizing the samples’ molecular composition. The FT-IR/ATR analysis suggests a medium lipid content in the seed press cake and flour blends. This is evidenced by sharp bands of medium intensity at the following wavenumbers: 3008 ν(=C-H)/cis, 1744 (ν(C=O)), 1456 (δ(CH)/CH2), 1414 (ρ(=CH)/cis), 1149 (ν(C-O), δ(-CH2-)), 707 (ρ(-(CH2)n-), and ρ(C=C)/-, cis) cm⁻¹ [56]. The intense vibration at 1638 and 1536 cm⁻¹ can be attributed to the amide I (ν(C=O), δ(N-H)) and II (δ(N-H), ν(C-N)) bonds from the protein compounds [57]. This indicates the cake’s high protein content and the blends’ beneficial fortification. The characteristic symmetrical and asymmetrical -C-O-C- vibrations for carbohydrates in the 1100–900 cm⁻¹ range are observed [56]. Adding more of the cakes to the flour has a positive effect on the nutritional properties of the mixture, i.e., it increases the concentration of proteins and unsaturated fats at the expense of starch.

The FTIR spectra present the molecular fingerprints of L. cylindrica seed press cake powder and its blends with cassava flour at different ratios: E1 (50/50%), E2 (30/70%), and E3 (70/30%). This reveals the characteristic functional groups and their interactions across the 4000–600 cm⁻¹ spectral range.

In the high-frequency region (3500–3000 cm⁻¹), all samples exhibit a broad absorption band centered at approximately 3280 cm⁻¹, corresponding to O-H stretching vibrations from hydroxyl groups involved in intra- and intermolecular hydrogen bonding. The pure L. cylindrica seed press cake (L) shows a relatively sharp and intense peak, while the blends (E1, E2, E3) display broader bands with varying intensities, indicating different degrees of hydrogen bonding networks between the seed press cake and cassava flour components.

The region between 2800–3000 cm⁻¹ shows distinct peaks at 2925 and 2854 cm⁻¹, attributed to the asymmetric and symmetric C-H stretching vibrations of methylene (CH₂) groups. These peaks appear most prominently in pure L. cylindrica seed press cake, suggesting a higher lipid content, with the peak intensity decreasing in the blends proportionally to the seed press cake content.

A significant peak at 1744 cm⁻¹ is observed across all samples, corresponding to the C=O stretching vibrations of ester groups, likely from residual oil or lipid components in the seed press cake. The intensity of this peak is highest in pure seed press cake and decreases systematically in the blends (E3 > E1 > E2), correlating with the seed press cake content.

The prominent bands at 1638 and 1536 cm⁻¹ can be attributed to amide I and amide II vibrations, respectively, indicating substantial protein content in the seed press cake. These peaks show C=O stretching and N-H bending vibrations characteristic of protein secondary structures. The intensity of these protein-related peaks decreases with the reduced seed press cake content in the blends.

The fingerprint region (1500–900 cm⁻¹) reveals complex overlapping peaks characteristic of carbohydrate structures. Notable peaks at 1456, 1414, 1375, 1341, 1227, 1193, 1149, 1103, 1078, and 1016 cm⁻¹ represent various C-O stretching, C-H bending, and C-C skeletal vibrations typical of polysaccharides. The peak at 1149 cm⁻¹ particularly indicates C-O stretching coupled with CH₂ bending vibrations.

In the 1000–900 cm⁻¹ region, the characteristic carbohydrate fingerprint shows symmetrical and asymmetrical C-O-C vibrations of glycosidic linkages. The blends exhibit enhanced intensity in this region compared to pure seed press cake, reflecting the higher starch content from cassava flour addition. The E3 blend (70% cassava flour) shows the most pronounced peaks in this region, particularly at 1016 cm⁻¹.

The low-frequency region (900–600 cm⁻¹) displays several peaks, including those at 857, 797, 744, 707, and 575 cm⁻¹. The peak at 707 cm⁻¹ is particularly significant, attributed to ρ(-(CH₂)_n^-) rocking vibrations and C=C bending of cis double bonds, confirming the presence of unsaturated fatty acids in the seed press cake.

The spectral patterns clearly demonstrate the complementary nature of the blend components. Pure L. cylindrica seed press cake shows strong lipid and protein signatures, while the addition of cassava flour enhances the carbohydrate-related absorption bands. The systematic changes in peak intensities across the blends (E1, E2, E3) correlate with their composition ratios, confirming successful blending and indicating molecular-level interactions between the seed press cake proteins/lipids and cassava starch components. This FTIR analysis validates the potential of these blends as functional ingredients with balanced nutritional profiles, combining proteins, lipids, and carbohydrates.

The viable cell counts of the microbial groups investigated in the Luffa seed press cake powder indicated that all microorganisms tested (TMC, coliforms, E. coli, yeast and mold) were below, and in accordance with, European commission regulations concerning the microbiological criteria for foodstuffs, both for food safety and process hygiene criteria. The mean of the triplicate showed the following results: 2 CFU/g for TMC and 0 CFU/g for the other microorganisms.

3.2. Meat Substitute Formulations

Figure 5 presents the cross-sectional structures of frozen–thawed doughs, revealing distinct morphological characteristics between pure tapioca flour systems and L. cylindrica blend formulations at varying water concentrations.

The upper panel displays tapioca flour:water doughs at concentrations ranging from 20% to 50%. At lower concentrations (20–30%), the cross-sections exhibit relatively uniform, dense structures with minimal void formation. As the flour concentration increases to 40%, the structure becomes more heterogeneous, with the appearance of larger air pockets and irregular void spaces. At the highest concentration (50%), the dough shows a more compact structure with reduced porosity, likely due to limited water availability for ice crystal formation during freezing.

The lower panel illustrates L. cylindrica blend doughs at concentrations from 25% to 35%. These samples demonstrate markedly different structural characteristics compared to pure tapioca flour systems. At 25% concentration, the cross-section reveals a highly porous, sponge-like structure with interconnected void spaces, suggesting extensive ice crystal formation and subsequent sublimation during the freeze–thaw cycle. The 30% concentration sample shows a more organized pore structure with distinct, uniformly distributed cavities throughout the matrix. At 35% concentration, the structure becomes denser with smaller, more regularly distributed pores, indicating reduced ice crystal formation due to lower water content.

The structural differences between the two dough systems are particularly evident in their response to freeze–thaw treatment. Pure tapioca flour doughs maintain relatively compact structures with limited void formation, suggesting a strong gel network formation that resists ice crystal damage. In contrast, L. cylindrica blend doughs exhibit more extensive structural modification, with pronounced honeycomb-like patterns, indicating weaker gel structures that are more susceptible to ice crystal disruption.

The color variations observed across samples also provide insights into component distribution and water migration patterns. Darker regions in the L. cylindrica blends likely indicate areas of concentrated seed press cake particles, while lighter areas suggest zones of higher starch concentration or ice crystal locations. The more uniform coloration in tapioca flour doughs indicates better component integration and more homogeneous water distribution.

These freeze–thaw structural analyses demonstrate that the incorporation of L. cylindrica seed press cake significantly alters dough architecture and water-binding properties, creating more open, porous structures that could influence final product texture and quality. The concentration-dependent structural changes suggest optimal formulation windows for specific textural outcomes in frozen dough applications.

3.3. Meat Substitute Quality

Table 4. The physicochemical characteristics of L. cylindrica meat substitute.

2		7			12
Sample	L*		a*	b*		Cutting Force [N]
Meat Substitute
2 days storage	40.36 ± 3.22 b		11.53 ± 1.19 b	10.55 ± 1.49 a		5.62 ± 1.05 a
7 days storage	32.20 ± 1.67 a		10.70 ± 1.00 b	10.59 ± 1.15 a		11.82 ± 4.41 b
12 days storage	39.88 ± 4.45 b		8.79 ± 0.83 a	12.87 ± 0.78 b		20.03 ± 2.40 c
Controls
LP30/TF70	51.23 ± 0.21 b		−1.47 ± 0.08 a	12.17 ± 0.00 a		---
Pork	46.73 ± 0.17 a		10.79 ± 0.21 b	19.81 ± 0.00 b		----
	Reducing Sugars [mg GLE/g DM]		DPPH [mg TE/g DM]	ABTS [mg TE/g DM]		FRAP [mg FeSO4/g DM]
LP30/TF70	5.32 ± 1.31 b		5.68 ± 0.07 b(E) 4.55 ± 0.51 a (W)	5.26 ± 0.02 a (E) 5.72 ± 0.33 a (W)		0.05 ± 0.02 a (E) 0.10 ± 0.02 b (W)
Meat Substitute	1.34 ± 0.09 a		0.99 ± 0.03 b (E) 0.63 ± 0.02 a (W)	0.82 ± 0.01 b (E) 0.54 ± 0.06 a (W)		0.58 ± 0.02 b (E) 0.14 ± 0.00 a (W)

Different superscript letters within a column indicate significant differences according to Dunkan’s test.

presents the physicochemical characteristics of L. cylindrica meat substitute during storage, including color parameters, textural properties, and bioactive profiles compared to those of the control samples.

The visual appearance of meat substitutes showed progressive changes during storage. At 2 days, the samples exhibited a uniform brown coloration with smooth surface characteristics. After 7 days storage, slight darkening was observed with minimal surface changes. By 12 days, the samples displayed more pronounced color changes with some surface irregularities, although the overall structural integrity was maintained.

The color parameters of the L. cylindrica meat substitute demonstrated significant changes during the 12-day storage period. The L* values (lightness) showed a notable decrease from 40.36 ± 3.22 to 32.20 ± 1.67 after 7 days, indicating substantial darkening, before partially recovering to 39.88 ± 4.45 by day 12. This darkening phenomenon is consistent with non-enzymatic browning reactions, particularly Maillard reactions between reducing sugars and amino acids, which continue during refrigerated storage [58]. The presence of amino acids from L. cylindrica seeds, particularly glutamic acid, provides reactive substrates for these browning reactions [59].

The progressive decrease in a* values (redness) from 11.53 ± 1.19 to 8.79 ± 0.83 over 12 days suggests oxidative degradation of red pigments, likely the betalains from the red beet component. This degradation pattern aligns with findings by Herbach et al. (2006) [60], who demonstrated that betalains are susceptible to oxidation even under refrigerated conditions. The relatively stable b* values (yellowness) ranging from 10.55 to 12.87 indicate that yellow pigments, possibly carotenoids from coral lentils, showed better stability during storage [61].

The cutting force measurements revealed significant textural changes during storage. Initial hardness at 2 days was 5.62 ± 1.05 N, increasing substantially to 11.82 ± 4.41 N after 7 days storage, indicating firmness development. The dramatic increase in cutting force from 5.62 ± 1.05 N to 20.03 ± 2.40 N over 12 days represents a 3.6-fold increase in product hardness. This progressive hardening can be attributed to multiple mechanisms: moisture migration, protein aggregation, and starch retrogradation from the tapioca component [62]. The observed hardness values after 12 days (20.03 N) approach those reported for certain meat products, suggesting that the texture evolution may actually improve meat-like characteristics during initial storage [63].

Bioactive properties of the formulated products showed distinct profiles. The reducing sugar content (DNS assay) was 1.34 ± 0.09 mg glucose equivalent/g DM for the meat substitute, compared to 5.32 ± 1.31 for the LP30/TF70 flour control. This reduction indicates sugar consumption or modification during processing. The substantial reduction in reducing sugar content from 5.32 ± 1.31 mg GLE/g DM in the LP30/TF70 control to 1.34 ± 0.09 mg GLE/g DM in the final product indicates significant sugar consumption during thermal processing. This 75% reduction can be attributed to Maillard reactions occurring during the 120 °C cooking process, where reducing sugars react with amino acids to form complex flavor and color compounds [64].

Luffa cylindrica seeds have been confirmed to contain amino acids, mainly glutamic acid, aspartic acid and arginine [59]. The DPPH antioxidant activity in ethanolic extracts was 0.99 ± 0.03 mg TE/g DM for the meat substitute versus 5.68 ± 0.07 for the LP30/TF70 control. The aqueous extracts were 0.63 ± 0.02 mg TE/g DM for the meat substitute compared to 4.55 ± 0.51 for the control, indicating significant antioxidant loss during thermal processing.

The ABTS radical scavenging capacity demonstrated similar patterns, with the meat substitute containing 0.82 ± 0.01 mg TE/g DM in ethanolic extracts versus 5.26 ± 0.02 for the LP30/TF70 control. The aqueous extracts were 0.54 ± 0.06 mg TE/g DM for the meat substitute compared to 5.72 ± 0.33 for the control.

The DPPH radical scavenging activity showed marked decreases in both ethanolic (5.68 to 0.99 mg TE/g DM) and aqueous extracts (4.55 to 0.63 mg TE/g DM), representing 82.6% and 86.2% losses, respectively. Similarly, the ABTS values decreased from 5.26 to 0.82 mg TE/g DM (84.4% loss) in the ethanolic extracts. These substantial losses align with the thermal degradation of phenolic compounds and other antioxidants during high-temperature processing, as reported by Çubukçu et al. (2019) [65].

The degradation of betalains from the red beet component likely contributed significantly to antioxidant loss. Muramatsu et al. (2023) [66] demonstrated that betanins are almost completely degraded when heated at 100 °C for 60 min, explaining the substantial antioxidant activity reduction in our 120 °C, 30 min cooking process.

Interestingly, the FRAP values showed an opposite trend, increasing from 0.05 to 0.58 mg FeSO₄/g DM in the ethanolic extracts and from 0.10 to 0.14 mg FeSO₄/g DM in the aqueous extracts. This unexpected increase in reducing power could be attributed to the formation of Maillard reaction products (MRPs) with reducing properties, as demonstrated by Morales and Jiménez-Pérez (2001) [67]. The higher value of reducing activity of the meat substitute could be due to the presence of reducing substances from coral lentils and red beets. It was confirmed that coral lentils contain substances with strong reducing properties, such as phenolic acids, gallic acid, ferulic acid, caffeic acid, and p-coumaric acid, and carotenoids such as beta-carotene [68,69].

In turn, in beetroot, the main carotenoids are beta-carotene and lutein, which have strong reducing properties [70]; this could have influenced the increase in the reducing activity of the obtained meat substitute.

The comprehensive analysis reveals that while the L. cylindrica meat substitute undergoes significant changes during refrigerated storage, particularly in texture and color, it maintains structural integrity over 12 days. The initial bioactive losses during processing are substantial, but no further significant degradation was observed during storage, suggesting the good stability of the remaining compounds. The texture evolution toward increased firmness may actually be beneficial for consumer acceptance, as it approaches the textural properties of conventional meat products.

Figure 6 presents a comparative sensory profile illustrating the preliminary acceptance patterns of meat substitutes between young men and women for two different formulations (Substitute A and Substitute B), evaluated across five sensory attributes on a 5-point hedonic scale. These results should be interpreted as preliminary due to the limited sample size. It clearly reveals what may be pronounced gender differences in the sensory evaluation of meat substitutes, with women demonstrating potentially higher hedonic ratings for Substitute B across all sensory attributes compared to men. Women are reported to have a better predisposition for meat alternatives than men, mainly due to their concerns about animal welfare and the environment [71] which could translate into more favorable sensory evaluations. The particularly high ratings women gave for taste (≈4.0) and overall acceptability (≈3.8) for Substitute B align with broader research indicating that women, compared to men, have been shown to respond differently to foods [72], particularly in their hedonic responses to novel food products.

The most striking gender difference was in the taste evaluation, where women rated Substitute B at approximately 4.0 points while men’s ratings remained notably lower at approximately 2.2 points. This substantial gap might be explained by the greater preference for sweet and sour flavors among females, compared to the preference for bitter and salty flavors among males [73]. This suggests that the flavor profile of Substitute B may have aligned more closely with female taste preferences. Additionally, non-vegetarians seem to be reluctant to try meat analogs due to the belief that consuming healthy products might compromise taste [74]; since men typically consume more meat than women, this reluctance could manifest as lower taste ratings.

Interestingly, men showed a slight preference for Substitute A in the texture evaluation (≈2.6 vs. ≈2.2 for Substitute B), contrasting with the women’s strong preference for Substitute B’s texture (≈3.8 vs. ≈3.2). This divergence may relate to different expectations for meat-like textures between genders. Compared to the meat burger, plant-based burgers had an off-flavor; they were perceived as less juicy, dry, and granular under blind conditions [75]; men’s higher consumption of traditional meat might make them more sensitive to textural differences that deviate from conventional meat products.

Women’s higher ratings for color (≈3.5 for Substitute B) compared to men (≈2.5 for both substitutes) align with research showing that before consumption, shape, color, and appearance have a greater influence on consumer acceptance compared to flavor and texture [74]. The gender difference in the visual evaluation may reflect women’s generally higher health consciousness and females’ propensity for breakfast regularity. Higher fruit and vegetable intake can be tied to greater health consciousness, often socially encouraged among women [73].

The overall acceptability scores suggest clear gender-based preferences, with women showing apparent discrimination between formulations (≈3.8 for Substitute B vs. ≈3.0 for Substitute A), while men exhibited more uniform and lower ratings (≈2.4 and ≈2.3). In one study, maleness is positively correlated with the consumption of mammal muscle meat; however, females are more accepting of vegetarian and vegan diets than males [76]. This fundamental difference in dietary orientation may influence the hedonic evaluation of meat substitutes.

These gender-specific sensory responses could have important implications for meat substitute formulation and marketing. The variety of novel protein alternatives on the market is increasing; there are many new product innovations potentially prompting consumers to change their nutrition habits [77]. The results tentatively indicate that Substitute B’s formulation successfully appeals to female consumers but might require modifications to enhance acceptance among male consumers, who represent a significant portion of the meat-eating population. The relatively uniform and moderate responses from male participants across both substitutes suggest that further testing with a larger sample is needed and indicate that sensory data would be key to understanding the physicochemical characteristics of novel plant proteins to support sensory developments [76]. Product developers should consider gender-specific formulations or marketing strategies that address the distinct sensory preferences and expectations of different consumer segments. The use of a 5-point hedonic scale for evaluation follows established sensory science practices, although it is worth noting that open-ended questions could further help to analyze sensory drivers and barriers to liking. A sensory study on consumer valuation for plant-based meat alternatives [78] could provide additional insights into the specific attributes driving these gender differences. Future research should investigate whether these gender differences persist across different cultural contexts and age groups, as vegetarians and vegans gave a significantly higher rating to the products than omnivores and, to some extent, flexitarians [78], suggesting that dietary habits may interact with gender in complex ways.

3.4. Limitations and Future Prospects

Despite the promising results, several limitations must be acknowledged to provide a balanced assessment of this valorization approach. The substantial reduction in antioxidant activity during thermal processing, with 80–85% losses in DPPH and ABTS values, indicates that current processing conditions require optimization to preserve bioactive compounds. The preliminary sensory evaluation, conducted with only 24 participants, revealed pronounced gender differences in acceptance, which indicates that the current formulation may not appeal universally, particularly to male consumers who demonstrated lower acceptance scores across all attributes. Additionally, the focus on tapioca flour, while strategically chosen for local resource utilization, may limit global applicability in regions where alternative starches are more accessible. The sensory evaluation was conducted with only 24 participants, which limits the statistical power and generalizability of the findings. This sample size is insufficient for robust statistical testing of consumer preferences and precludes definitive conclusions about demographic differences in product acceptance.

Future research directions should prioritize processing optimization studies to minimize bioactive compound degradation through modified thermal treatments or protective ingredient incorporation. Comprehensive sensory studies with larger, demographically diverse populations are essential to identify formulation modifications that enhance universal consumer acceptance. An investigation of alternative starch sources would significantly enhance global applicability and allow for adaptation to different regional agricultural systems. Extended nutritional characterization beyond dietary fiber content, including protein quality assessment and mineral bioavailability studies, would strengthen the nutritional positioning of this ingredient. Furthermore, scale-up investigations addressing processing challenges, economic viability analysis, and extended shelf-life studies under various storage conditions are crucial for successful commercialization. The development of different product formats beyond meat substitutes could expand the application opportunities and maximize the valorization potential of L. cylindrica seed press cake across diverse food systems.

4. Conclusions

This study successfully demonstrated the potential of Luffa cylindrica seed press cake as a valuable functional ingredient for sustainable meat substitute formulations, effectively transforming an agricultural waste byproduct into a food-grade material. The comprehensive characterization revealed that L. cylindrica seed press cake possesses unique functional properties, including a high water holding capacity, significant antioxidant activity, and substantial protein content, which complement the gel-forming and swelling properties of tapioca flour.

The LP30/TF70 blend emerged as the optimal formulation, providing balanced techno-functional characteristics suitable for meat analog development. The resulting meat substitute exhibited acceptable textural properties that improved during storage, with cutting force values approaching those of conventional meat products after 12 days. While thermal processing reduced antioxidant activities by 80–85%, the products maintained structural integrity and demonstrated good microbiological safety throughout the storage period.

This research represents the first comprehensive scientific evaluation of Luffa cylindrica seed press cake as a functional food ingredient, establishing a novel pathway for agricultural waste valorization that significantly advances the field of sustainable food ingredient development. The work makes several unique contributions to food science and technology. Methodologically, we developed the first systematic characterization protocol for L. cylindrica seed press cake’s techno-functional properties, providing benchmark data for water holding capacity, pasting behavior, and antioxidant profiles that were previously unavailable in the scientific literature. This establishes essential foundational knowledge for future food applications of this underutilized byproduct.

The study uniquely demonstrates successful transformation of an agricultural waste stream into a value-added functional ingredient through systematic blend optimization with complementary materials. Unlike previous valorization studies that focus primarily on single-component characterization, our research provides a complete development pathway from waste characterization through to the final product formulation and sensory evaluation. This holistic approach contributes significantly to circular economy principles by demonstrating the practical implementation of waste-to-food conversion.

From a technological perspective, the work advances meat substitute formulation science by introducing a novel fiber-rich ingredient that enhances both nutritional and functional properties while addressing sustainability concerns. The research contributes to global food security discussions by demonstrating how locally available agricultural wastes can be transformed into protein-alternative products using regionally accessible processing techniques. This represents a significant advancement in sustainable protein development that is particularly relevant for the tropical and subtropical regions where L. cylindrica cultivation is established but seed press cake utilization remains unexplored.

The identification of significant gender differences in meat substitute acceptance represents an important contribution to consumer behavior research in the plant-based food sector, providing insights that can guide targeted product development and marketing strategies. This socio-demographic analysis adds a crucial dimension to technical food development that is often overlooked in ingredient-focused research. Women showed a generally higher acceptance across all sensory attributes, suggesting that formulation optimization might benefit from considering demographic preferences to maximize market penetration.

This research contributes to the circular economy by providing a viable pathway for agricultural waste valorization while addressing the growing demand for sustainable protein alternatives. The findings establish L. cylindrica seed press cake as a promising ingredient for the plant-based food industry, offering both functional benefits and environmental sustainability. Future research should focus on optimizing processing conditions to minimize bioactive compound losses and developing formulations that enhance acceptance among male consumers.