Application of a D‐Optimal Mixture Design for the Development and Optimization of a Pasta Enriched With Pumpkin Products

ABSTRACT

Fortification of wheat‐based products with high‐nutritional‐value ingredients is becoming of great interest as a strategy for producing high‐quality sustainable food products. Therefore, pumpkin products such as pulp (PPU), peel (PPE), and seed (PS) were blended into wheat pasta at different proportions and combinations using D‐optimal mixture design (DMD), resulting in 12 mixtures between PPU, PPE, and PS flours. Fifteen responses were analyzed and predicted models with high accuracy (p < 0.02) and a satisfactory coefficient of determination (R ² > 0.92). The optimum composition was 85% semolina, 11.85% PPU, 0% PPE, and 3.15% PS, achieving a desirability of 86.04%. It corresponded with the following properties: moisture of 9.66%, ash of 1.49%, protein of 11.89%, fat of 1.63%, carbohydrates of 75.06%, total phenolics of 107.22 mg/100 g dry weight (DW), total flavonoids of 19.27 mg/100 g DW, carotenoids of 10.17 mg/100 g dry powder (DP), and antioxidant activity of 15.34%. The cooking loss, optimal cooking time, and water absorption capacity reached values of 5.04%, 8.5 min, and 211.86%, respectively. The color parameters L*, a*, and b* values were 84.38, 2.76, and 27.3, respectively. The sensory analysis scores of the DMD mixtures disclosed that the mixture with 12% PPU and 3% PS was the most appreciated, and these proportions were quite similar to the optimized formula, which could be a major asset in combining nutritional and sensory profiles. This new formulation advocated that pasta fortified with pumpkin products could confer improved nutritional, cooking, and sensorial profiles.

Practical Applications

For the production of products with appealing consumer attributes, adding additives to pasta formulas is a suitable replacement. The amount of ingredients added can affect the pasta's quality and produce properties that could impact the final result. It is possible to identify the ingredients that support pasta quality using the D‐optimal combination design, which provides pertinent information for the food sectors. The findings defined the proportions of pumpkin products, including pulp, peel, and seed, in pasta. They also disclosed their remarkable potential in enhancing the quality parameters, nutritional value, and cooking profile of pasta while maintaining a satisfactory sensorial profile, thereby promoting the industrial development of high‐value alimentary paste.

1. Introduction

“Pasta” is the Italian term for “dough” and refers to a traditional, ancient, and highly versatile food with a long shelf life currently consumed worldwide (Nilusha et al. 2019). The International Pasta Organization (IPO) reported that global pasta production will reach 16.9 million tons in 2022. Major market shares come from regions including North America, Central and South America, Africa, the Middle East, Asia, Australia, and Europe (both United and other European countries), with key contributing countries such as Italy, the United States, Brazil, Russia, and Turkey. Among consumers worldwide, Italy, Tunisia, Venezuela, Greece, and Peru rank as the top five countries in pasta consumption (Roshini et al. 2025).

From a culinary perspective, pasta is essentially made by combining water and semolina, occasionally with other ingredients (Bresciani et al. 2022). Semolina is valued for its high yellow pigment content, low lipoxygenase activity, and gluten‐forming proteins, which determine its high cooking quality, ensuring minimal cooking losses (CLs), preventing overcooking, and providing a firm and elastic texture (Sobota et al. 2020). Pasta components are usually subject to a continuous process that includes three major phases: dosing and mixing, kneading and shaping (by extrusion or lamination), and drying (Bresciani et al. 2022). The extrusion technique, which is a method of pressing combined components out through a small hole known as a die to form and shape the materials, is one of the most widely used technologies in pasta manufacturing. Significant shear stresses are applied to the dough during the extrusion process, which may weaken the protein structure. However, concurrent pressure application gives the pasta a high compactness that enables it to endure cooking (Jalgaonkar et al. 2019).

Although pasta is a significant source of complex carbohydrates and contains low levels of fats and sodium, it is deficient in several nutrients, such as dietary fibers, proteins (essential amino acids), minerals, and vitamins (Dello Russo et al. 2021; Vital et al. 2020). It is worth noting that pasta is regarded by the Food and Drug Administration (FDA) and the World Health Organization (WHO) as a suitable vehicle for nutritional fortification (Bianchi et al. 2021; Nilusha et al. 2019). Consequently, many studies have focused on the fortification of pasta with different products, such as cereals, vegetables, fruits, legumes, and seeds, which in many cases came as by‐products of the food chain to enhance its nutritional and antioxidant value while maintaining desirable sensory and functional properties (Bas‐Bellver et al. 2024; Castillejo et al. 2025). Examples include fortification with parsley leaf powder (Sęczyk et al. 2016), asparagus flour (Vital et al. 2020), carob powder (Ribes et al. 2025), mushroom powder (Lu et al. 2018), persimmon flour (Lucas‐González et al. 2020), tomato and linseed by‐products (Estivi et al. 2024), and vegetable powders such as beet, carrot, and kale (Sobota et al. 2020).

The Cucurbitaceae family includes the commonly grown pumpkin (Cucurbita spp.), a highly nutritious and valuable crop. The most widely utilized portion of the pumpkin fruit is the pulp, which generates by‐products such as peels, seeds, and fibrous components (Arshad et al. 2025). Generally, industry processing and household disposal generate 18%–21% of pumpkin by‐products (Fatima et al. 2025). An estimated 1.35–5.4 and 0.92–1.38 million tons of pumpkin peel and seed, respectively, are produced each year globally from the approximately 27 million tons of pumpkins produced (Tran and Nguyen 2025; Mehdizadehtapeh et al. 2026).

The increasing demand for pumpkins is largely driven by their low caloric content and high concentrations of beneficial compounds, including dietary fibers, polysaccharides, pectin, and phenolic compounds (mainly flavonoid pigments) (Ihedinachi et al. 2025). The attractive orange color of pumpkins is attributed to carotenoids (Leichtweis et al. 2025), which are increasingly valued for their antioxidant activity. Pumpkins are also rich in vitamins A, C, B1, B2, and B9, as well as minerals such as potassium, calcium, magnesium, sodium, and iron (Ihedinachi et al. 2025).

The incorporation of pumpkin into bakery and cereal‐based products has been widely investigated. Pumpkin product flour was added to wheat flour for cupcake (Batista et al. 2018), cookie (Kumari et al. 2021; Anitha et al. 2020), sponge cake (Ghaboos et al. 2018), noodle (Farzana et al. 2023), and bread (Aljobair 2024) formulations.

To address growing production issues, the food industry requires a strict statistical, experimental design (ED), and operational excellence (OPEX) strategy. The amount of work and experimentation involved in product development increases as more factors are examined. The design of experiments (DoE) provides an effective way to reduce process variability by assessing multiple parameters at once (Antony et al. 2024). Mixture designs (MDs) are especially notable among the different EDs. It is important to emphasize that all MDs are followed by a mathematical regression model that can be linear if the blending effect is purely additive or quadratic, special cubic or full cubic if the response exhibits curved, nonlinear behavior as a result of component interactions (Galvan et al. 2021).

D‐optimal mixture design (DMD) is a statistical approach widely applied in food formulation studies to optimize multicomponent systems efficiently (Fidaleo et al. 2021; Kamali Rousta et al. 2021). Unlike traditional factorial designs, DMD allows for the evaluation of the effects and interactions of mixture components while minimizing the number of experimental runs needed. This design is particularly suitable for complex formulations where ingredients are proportions of a whole, as it enables the identification of optimal combinations that maximize multiple responses simultaneously (e.g., nutritional, functional, and sensory attributes). Although DMD has been successfully applied in the optimization of beverages, bakery, and dairy products, its application in pasta formulation with pumpkin‐derived ingredients has not been previously reported.

Therefore, the novelty of this study resides in using DMD for the formulation of pasta in which semolina is partially replaced with three pumpkin‐derived flours: pulp (PPU), peel (PPE), and seeds (PS). MDs provide significant advantages by allowing efficient experimentation to elucidate the effects and interactions of ingredients with fewer trials, thus conserving resources. They enable staged optimization processes, such as first optimizing physicochemical or textural properties and subsequently assessing sensory qualities. The DMD method enhances this efficiency by automatically generating formulations within predefined limits, simplifying the development of formulations. Unlike traditional EDs, MDs define factors as proportions of a mixture, facilitating the creation of novel products with potential commercial value or process improvements. This approach encourages innovation and expands the possibilities in scientific research and product development (Luthfiyanti et al. 2024; Squeo et al. 2021). The optimization was conducted to obtain a product with increased nutritional (ash, protein, fat, carbohydrates), antioxidant (total phenolic, flavonoid, and carotenoid contents, total antioxidant activity), functional (CL, optimal cooking time [OCT], water absorption capacity [WAC]), and color (L*, a*, and b*) properties. In addition, sensory evaluation was performed for all DMD mixtures, and the optimal formulation and control pasta were further characterized in terms of mineral, fiber content, and carotenoid composition.

2. Materials and Methods

2.1. Pumpkin Supplement Preparation and Ingredients

Pumpkin (variety: Cucurbita pepo L.) was harvested in Bejaia Department, Algeria (latitude 36°33′00.00″ N and longitude 4°42′00.00″ E) during September 2023. It was washed, rinsed, and peeled to separate the flesh from the rind and seeds. Each part was dried at 40°C in a ventilated oven (Model 55 UN, Memmert, Germany) until constant weight. After that, they were milled and sifted to produce a powder particle size ≤ 250 µm. Semolina was collected from a local market.

2.2. Pasta Processing

Tagliatelle pasta flour was prepared with semolina and different concentrations of the pumpkin mixture PPU–PPE–PS in 15% of the total weight of wheat flour. For this formulation, the appropriate range of each flour ingredient, such as PPU (X ₁: 3%–12%), PPE (X ₂: 0%–6%) and PS (X₃: 0%–6%), was selected through a preliminary trial where each ingredient was enriched individually from 0% to 15% in pasta and then submitted to a sensory evaluation (Table 1).

TABLE 1.

Different levels of pasta enrichment with pumpkin pulp (PPU), peel (PPE), or seed (PS) flours individually.

	Ingredients	Amount (g/100 g)
		Control	3%	6%	9%	12%	15%
Fortification	PPU flour	0	3	6	9	12	15
	PPE flour	0	3	6	9	12	15
	PS flour	0	3	6	9	12	15
Basic pasta formulation	Semolina	100	97	94	91	88	85

Durum wheat flour (300 g), pumpkin mixture, and the necessary amount of water were combined in the pasta extruder's mixing chamber for 20 min. The aggregation of moist flour was placed in the extrusion chamber including a single screw (length, 30 cm; diameter, 5.5 cm) that ended with a bronze die (hole diameter, 1.70 mm) (Laboratory scale pasta press MAC 30S‐Lab extruder, ItalPast, Parma, Italy) was used to extrude the dough into a tagliatelle shape with screw speed of 50 rpm, pressure was approximately 3.4 bar, and the temperature of the pasta after the extrusion was between 27°C and 28°C. Extruded pasta was dry in a drying chamber (8 h at 40°C) to a final moisture content of approximately 10%. The dry pasta (DP) was stored in refrigerator‐safe low‐density polyethylene (LDPE) bags for further analysis. Before the analysis, pasta was homogenized to a particle size of 0.5 mm.

2.3. Experimental Design and Optimization Procedure

In this study, DMD was implemented to design experiments with three ingredients, PPU, PPE, and PS, to establish the best proportion of the fortified pasta formulation. For the repeatability of the procedure, a total of 12 experiments were created for the selected ingredient levels (Table 2). Equation (1) outlines the optimization procedure for predicting responses relying on a polynomial model for q components (Presenza et al. 2022):

$$Y_i=\sum _{i=1}^q{\beta }_i{x}_i+\sum \sum _{i<j}^q{\beta }_{ij}{x}_i{x}_j+\sum \sum \sum _{i<j<k}^q{\beta }_{ijk}{x}_i{x}_j{x}_k+\epsilon$$ (1)

where Y_i is the predictive response variable including nutritional profile (moisture, protein, fat, carbohydrates, total polyphenolic content [TPC], total flavonoid content [TFC], total carotenoid content, antioxidant activity by 2,2‐diphenyl‐1‐picrylhydrazyl [DPPH]), cooking parameters (CL, OCT, WAC), and color analysis (L*, a*, b*). While β_i , β_ij , and β_ijk are the linear and nonlinear coefficients, x_i and x_j represent the proportions of independent ingredients.

TABLE 2.

D‐optimal mixture design applied to pasta enrichment optimization.

Mixture	Component proportions			PPU flour (g/100 g)	PPE flour (g/100 g)	PS flour (g/100 g)
	X1	X2	X3
1	0.80	0.00	0.20	12.00	0.00	3.00
2	0.42	0.18	0.40	6.30	2.70	6.00
3	0.68	0.00	0.32	10.20	0.00	4.80
4	0.56	0.22	0.22	8.40	3.30	3.30
5	0.20	0.40	0.40	3.00	6.00	6.00
6	0.20	0.40	0.40	3.00	6.00	6.00
7	0.80	0.20	0.00	12.00	3.00	0.00
8	0.56	0.22	0.22	8.40	3.30	3.30
9	0.60	0.00	0.40	9.00	0.00	6.00
10	0.41	0.40	0.19	6.15	6.00	2.85
11	0.60	0.40	0.00	9.00	6.00	0.00
12	0.56	0.22	0.22	8.40	3.30	3.30

Abbreviations: PPE, pumpkin peel; PPU, pumpkin pulp; PS, pumpkin seed.

To minimize the influence of unexplained variability in observed responses resulting from external factors, the tests were conducted in a random order. The three ingredients, PPU flour (X ₁), PPE (X ₂), and PS (X ₃), were taken as independent variables, while the response variables were moisture content (Y ₁), ash (Y ₂), protein content (Y ₃), fat content (Y ₄), carbohydrate content (Y ₅), TPC (Y ₆), TFC (Y ₇), total carotenoid content (Y ₈), antioxidant activity by DPPH (Y ₉), CL (Y ₁₀), OCT (Y ₁₁), WAC (Y ₁₂), lightness (L*) (Y ₁₃), redness (a*) (Y ₁₄), and yellowness (b*) (Y ₁₅).

During optimization, the ingredients were kept in range, while ash, protein, TPC, TFC, total carotenoid, and antioxidant activity were maximized. Moisture, fat, carbohydrates, and cooking quality parameters of the product were minimized. The optimal mixture was selected using the desirability function, a widely used technique that identifies the relevant operating conditions to obtain the best response values. It serves as an objective function that quantifies the preferred value. This method transforms each response into a unique desirability function that is scaled from 0 to 1. The following equation illustrates how the design factors are chosen to optimize the overall desirability, where $d_i(y_i)$ is the desirability of the individual responses and n represents the number of responses (Talens et al. 2022):

$$D={\left(\prod _{i=1}^n{d}_i\left(y_i\right)\right)}^{1/n}$$

2.4. Quality Parameter Determination

2.4.1. Proximate Composition

The analysis of moisture content, ash, crude protein, and fat of pasta products was carried out using the AACC methods (American Association of Cereal Chemists 2000). The carbohydrate amount was determined by deducting the total moisture, protein, fat, and ash from 100. The results were reported as percentages (%).

2.4.2. Quantification of Bioactive Compounds

Phenolics from pasta samples were extracted following the method of Yilmaz et al. (2024). Briefly, 1 g of the pasta sample was combined with 10 mL of 70% methanol for 2 min. After sonication in a water bath (VWR, Leuven, Belgium) for 10 min, the mixture was centrifuged for 20 min at 4°C at 12,100 × g. The supernatant was collected for TPC and TFC assessments.

TPC was evaluated by the Folin–Ciocalteu method. The results were expressed as mg gallic acid equivalent (GAE) per 100 g DP. The TFC was estimated by the aluminum chloride method. The results were represented in mg of quercetin equivalent (QE) per 100 g DP (Oufighou et al. 2024).

Carotenoids were assessed using the procedure described by Brahmi et al. (2018), and the results were expressed as mg β‐carotene equivalent (β‐CE) per 100 g DP.

The antioxidant potential of pasta samples was determined using the DPPH radical scavenging activity following the method of Armellini et al. (2018). The results were expressed as percentages as follows: DPPH (%) = [(Absorbance_control − Absorbance_sample)/Absorbance_control] × 100.

2.4.3. Cooking Quality Evaluation

CL was measured after homogenization of the cooking water, and 25 mL was removed and dried in an oven (Model 55UN, Memmert, Germany) at 105°C for 24 h. After weighing the residue, the results are expressed as g solids per 100 g of DP (American Association of Cereal Chemists 2000).

The OCT evaluation was estimated after cooking 50 g of pasta with 500 mL of boiling tap water in a Durex beaker. After being cooked for different times, the pasta was immediately drained. The elimination of the white, uncooked core in a strand of pasta pressed between two glass plates served as a marker for OCT.

The WAC of pasta reflects the cooked pasta's ability to retain water to a greater or lesser extent and was determined following the formula (Torres et al. 2021):

$$\mathrm{WAC}=\left[\left(W_1-W_2\right)/W_2\right]\times 100$$

where W ₁ (g) is the weight of cooked pasta and W ₂ (g) is the weight of raw pasta.

2.4.4. Color Measurement

The color of the different formulations was determined using a Minolta Chroma Meter CR‐200 (Minolta Camera Co. Ltd., Osaka, Japan), where the three color‐reflectance values, CIELAB L*, a*, b*, were obtained for each combination (Kamali Rousta et al. 2020).

2.5. Sensory Analysis

2.5.1. Panel Characteristics

Fifteen trained panelists (6 men and 9 women, aged between 28 and 55 years) consisting of members of the University of Bejaia's (Algeria) sensory analysis laboratory, carried out the sensory evaluation. All panelists received training 2 h a week continuously until they were accustomed to evaluating various attributes of different types of food products before performing the sensory evaluation to precisely identify and describe the sensory characteristics. To guarantee ethical compliance and voluntary involvement, panelists taking part in the sensory evaluation first completed an informed consent form prior to data collection (Thomas et al. 2022; Carpentieri et al. 2024).

2.5.2. Design of Sensory Experiment and Protocol Evaluation

In the sensory analysis laboratory, each panelist was placed in an individual booth, well‐ventilated and equipped with fluorescent lights at ambient room temperature (25°C), where each participant was requested to rinse his/her mouth before starting and between tasting. Pasta samples were cooked at their OCT, drained, and marked with unique, randomly generated three‐digit codes and presented in white plastic plates. Panelists were given detailed instructions and guided through the evaluation procedure using a 5‐point hedonic scale considering the intensity (from 0 to 5) of the following descriptors: color (light/dark), smell (light/strong), texture (smooth/rough), stickiness (not‐sticky/sticky), aspect (soft/hard), and taste (bad/good). In addition, the level of sensory acceptance was quantitatively assessed using a scale ranging from 0 to 20 (Iacovino et al. 2025).

2.5.3. Mineral and Fiber Content of the Optimized Formula

The mineral content of the optimized pasta was analyzed by x‐ray fluorescence (XRF) according to (Nielson et al. 1991) by a spectrometer S4 Pioneer (Bruker Corporation, Billerica, MA, USA) equipped with an Rh anticathode x‐ray tube (20–60 kV, 5–150 mA, and 4 kW maximum), five analyzer crystals (LiF200, LiF220, Ge, PET, and XS‐55), a sealed proportional counter for light element detection and a scintillation counter for heavy elements with slight modifications. The recorded spectrum was evaluated by the fundamental parameter method using the Spectraplus software EVA 1.7. Mineral content was expressed as g/kg DW and mg/kg DW for major minerals and trace elements, respectively.

The fiber content of the optimized pasta was determined by the enzymatic‐gravimetric method according to McCleary et al. (2010).

2.5.4. UHPLC–DAD Analysis of Optimized Pasta Carotenoids

Carotenoids from pasta optimized powder were extracted following the procedure described by Pinna et al. (2022), with minor modifications. Briefly, 1 g of dried pasta powder was homogenized with 20 mL of a hexane/isopropanol mixture (60:40, v/v) using an ultrasonic bath (Sonorex Digiplus, Bandelin electronic GmbH & Co. KG, Berlin, Germany) for 30 min at 45°C. The resulting extract was first filtered through filter paper (Grade 589/2, Whatman, Maidstone, UK), followed by a 0.2‐µm PTFE membrane filter. The filtered extracts were collected in amber glass vials to protect them from light degradation.

The carotenoids were determined from the pasta optimized mixture using a UHPLC system (Shimadzu, Kyoto, Japan) equipped with a DGU‐20A degasser, LC‐30AD quaternary pump, SIL‐30AC autosampler, CTO‐10AS column heater, and SPDM‐20A photodiode array detector. Briefly, the separation phase was performed on a reverse‐phase C30 Develosil column (250 × 4.6 mm i.d., 5‐µm particle size, Nomura Co., Kyoto, Japan). The mobile phase consisted of a binary solvent system: A (methanol:water, 97/3 v/v) and B (MTBE) with a flow rate maintained at 1.0 mL/min and an injection volume of 20 µL. The gradient elution profile lasted 60 min as follows: 0 min 84% A, 11.25 min up to 78% A, 30 min up to 55% A, 60 min up to 65% A, and the peaks were detected at 450 nm. The identification was performed by comparing the retention time and elution profile of pure standards of β‐carotene and lutein (Figures S1 and S2). Three separate peaks appeared corresponding to three major components present in pumpkin products (Kulczyński and Gramza‐Michałowska 2019): lutein (Peak a) with a retention time of 10.64 min, Peak b, which corresponded to α‐carotene with a retention time of 25.03 min, and β‐carotene (Peak c) with a retention time of 27.89 min. For quantification, calibration curves of lutein and β‐carotene at different concentrations ranging from 0.01 to 20 ppm were prepared.

2.6. Statistical Analysis

JMP software (Version 14.0, SAS Institute, Cary, NC, USA, 1989–2019) was used for MD model building. ANOVA followed by Fisher's LSD test was performed to compare mean values. If the p < 0.05, the findings were considered statistically significant using STATISTICA software (Version 8.0, StatSoft Inc., Tulsa, OK, USA, 2007). For the sensory data, the treatment was assessed by XLSTAT (Version 9.0, Addinsoft, Paris, France). All data are provided as the mean ± SD of three repetitions.

3. Results and Discussion

3.1. Preliminary Study

The DMD plan was preceded by a preliminary study involving a sensory analysis to define the appropriate ranges of pumpkin products used for enrichment. Each pumpkin powder was tested individually at levels ranging from 3% to 15%, resulting in a total of 15 pasta formulations. For PPU flour, a very high consumer preference was found between 100% and 90% till 12% of enrichment, whereas at 15%, the appreciation decreased significantly (10%). A high preference for PPE flour was also observed up to an incorporation level of 6%, but above this level, the appreciation decreased considerably (from 30% for 9% of enrichment to 0% for 12% and 15% of enrichment). The same trend was observed for PS flour, where the appreciation was between 100% and 80% for 3% and 6% incorporation, respectively. However, over this level, the acceptance was very low (Table 3).

TABLE 3.

Consumer preference of pasta individually enriched with different levels of pumpkin‐derived flours.

Pumpkin‐derived flours	Enrichment level (%)	Percentage of preference
Control	0	100%
PPU	3	100%
	6	100%
	9	100%
	12	90%
	15	10%
PPE	3	100%
	6	100%
	9	30%
	12	0%
	15	0%
PS	3	100%
	6	80%
	9	0%
	12	0%
	15	0%

Note: Enrichment level of 3%–15%, means 3–15 g of pumpkin‐derived product per 100 g of pasta.

Abbreviations: PPE, pumpkin peel; PPU, pumpkin pulp; PS, pumpkin seed.

Based on these findings, maximum enrichment concentrations for pumpkin products in pasta were established at the point where the sensory quality of the pasta reached the acceptable threshold (above 50%). The ranges used for the DMD building were PPU (X ₁: 3%–12%), PPE (X ₂: 0%–6%), and PS (X₃: 0%–6%).

3.2. Optimization of Pumpkin Pasta Formulations

3.2.1. Proximate Composition

The moisture content of the different formulations varied between 9.48% and 9.93%, which was lower than that of the control pasta (11.5%). For ash content, the values of the enriched pasta varied between 1.07% and 1.53%, which were higher than those of the control pasta (0.67%). The increase in ash content can be attributed to the inclusion of pumpkin flours, which contain higher ash content than wheat flour. Since ash content reflects the total mineral content, samples with higher ash content are assumed to have a higher mineral content (Hussain et al. 2022). The protein content significantly varied (p < 0.05), ranging from 10.9% in the pasta mixture enriched with PPU and PPE (without PS) to 16.15% in the mixture containing PPU and PS (without PPE). Indeed, PS possesses higher amounts of proteins than PPU and PPE. These results were higher than those of the control (9.08%) and consistent with findings by Farzana et al. (2023), who reported a protein content of 14.75 ± 0.13% in noodles enriched with 20% pumpkin flour. The fat content varied from 0.77% to 2.6%, and formulations containing 6 g of PS flour/100 g of DP showed the highest fat content values (Table 4). PS are considered a rich source of essential fatty acids such as palmitic, stearic, oleic, and linoleic acids, which enhance their nutritional profile. Beyond their nutritional benefits, dietary fats also play a key role in improving the sensory qualities of food by enhancing flavor retention. In fact, they are essential for the effective absorption and transport of fat‐soluble vitamins, which are critical for various physiological functions (Öztürk and Turhan 2020). Similarly, Das et al. (2021) reported a fat content of 5.67 ± 0.10 g/100 g for biscuits enriched with 10% of whole PS flour. Carbohydrates represented the major fraction in the different pasta formulations, ranging between 70.44% and 77.32%, and this amount decreased compared with the control pasta (78.1%). A similar trend was observed in crackers enriched with pumpkin flour, where carbohydrate values decreased from 73.53 g/100 g in the control sample to 69.87 g/100 g at 20% of incorporation (Gallardo et al. 2025).

TABLE 4.

Nutritional composition, cooking quality and color of pasta mixtures obtained from the D‐optimal mixture design using pumpkin‐derived flours.

		D‐optimal mixture design
	CTRL	1	2	3	4	5	6	7	8	9	10	11	12
PPU	0	0.80	0.42	0.68	0.56	0.20	0.20	0.80	0.56	0.60	0.41	0.60	0.56
PPE	0	0.00	0.18	0.00	0.22	0.40	0.40	0.20	0.22	0.00	0.40	0.40	0.22
PS	0	0.20	0.40	0.32	0.22	0.40	0.40	0.00	0.22	0.40	0.19	0.00	0.22
Moisture (%)	11.5 ± 0.026a	9.93 ± 0.16b	9.62 ± 0.01e	9.88 ± 0.08bc	9.64 ± 0.08de	9.48 ± 0.04e	9.48 ± 0.04e	9.91 ± 0.05b	9.62 ± 0.11e	9.87 ± 0.12bcd	9.53 ± 0.04e	9.87 ± 0.07bcd	9.65 ± 0.13cde
Ash (%)	0.67 ± 0.05b	1.49 ± 0.15a	1.35 ± 0.01a	1.40 ± 0.49a	1.50 ± 0.03a	1.43 ± 0.28a	1.40 ± 0.05a	1.39 ± 0.04a	1.53 ± 0.04a	1.27 ± 0.02a	1.29 ± 0.16a	1.07 ± 0.02ab	1.51 ± 0.14a
Protein (%)	9.08 ± 0.144e	12.19 ± 0.67c	15.01 ± 0.02ab	12.8 ± 0.05c	11.72 ± 0.67cd	14.62 ± 0.16b	14.20 ± 0.40b	10.90 ± 0.29d	11.60 ± 0.24cd	16.15 ± 0.66a	14.36 ± 0.22b	14.83 ± 0.66b	11.68 ± 0.44cd
Fat (%)	0.65 ± 0.03f	1.60 ± 0.01d	2.28 ± 0.01b	1.82 ± 0.04cd	1.25 ± 0.08e	2.40 ± 0.06ab	2.60 ± 0.17a	0.77 ± 0.01f	1.21 ± 0.08e	2.22 ± 0.14b	1.85 ± 0.00c	0.83 ± 0.00f	1.18 ± 0.09e
Carbohydrates (%)	78.1 ± 0.06a	74.78 ± 0.90cd	71.74 ± 0.02gh	74.09 ± 0.50de	75.89 ± 0.66bc	72.07 ± 0.32fg	72.31 ± 0.31fg	77.03 ± 0.32ab	76.03 ± 0.19bc	70.49 ± 0.62h	72.96 ± 0.35efg	73.39 ± 0.61ef	75.98 ± 0.30bc
TPC (mg GAE/100 g DW)	52.82 ± 0.248d	106.53 ± 1.06b	114.08 ± 1.44a	111.63 ± 2.99a	106.13 ± 0.10b	99.87 ± 1.97c	99.33 ± 0.60c	102.91 ± 0.40bc	105.66 ± 1.18b	112.59 ± 2.85a	103.42 ± 0.61bc	99.64 ± 1.14c	106.20 ± 1.71b
TFC (mg QE/100 g DW)	0.78 ± 0.036h	19.07 ± 0.98ef	18.65 ± 1.26f	18.83 ± 0.49f	22.42 ± 1.55cd	19.48 ± 0.03cdef	19.32 ± 0.33def	26.19 ± 1.75b	22.09 ± 0.62cde	14.63 ± 0.87g	26.06 ± 0.87b	29.94 ± 1.94a	22.61 ± 0.47c
Carotenoids (mg β‐CE/100 g DW)	1.77 ± 0.101h	10.16 ± 0.12b	7.17 ± 0.15efg	8.88 ± 0.15c	7.51 ± 0.02e	6.88 ± 0.18g	6.86 ± 0.14g	11.27 ± 0.18a	7.39 ± 0.15ef	7.92 ± 0.11d	7.03 ± 0.09fg	8.28 ± 0.14d	7.48 ± 0.19e
DPPH (%)	2.45 ± 0.066f	15.39 ± 0.42ab	15.45 ± 0.74ab	15.33 ± 0.36abc	15.93 ± 0.18a	12.82 ± 0.41de	12.55 ± 0.53e	15.50 ± 0.56ab	15.80 ± 0.10ab	14.54 ± 0.84bc	14.01 ± 0.60cd	13.99 ± 0.21cd	16.03 ± 0.08a
CL (%)	4 ± 0.505c	5.04 ± 1.23ab	5.67 ± 0.16a	5.05 ± 0.04ab	5.50 ± 0.13ab	5.14 ± 0.12ab	5.08 ± 0.11ab	5.47 ± 0.10ab	5.55 ± 0.06ab	4.95 ± 0.24b	5.23 ± 0.21ab	5.02 ± 0.25ab	5.53 ± 0.29ab
OCT (min)	10 ± 0.2a	8.48 ± 0.04bc	8.11 ± 0.01cd	8.55 ± 0.26b	8.01 ± 0.06de	9.66 ± 0.09a	9.60 ± 0.13a	7.85 ± 0.03de	8.00 ± 0.02de	8.26 ± 0.33bcd	7.66 ± 0.07ef	7.26 ± 0.08f	8.03 ± 0.05de
WAC (%)	178.33 ± 7.637e	208.60 ± 3.84d	230.00 ± 5.57ab	222.83 ± 2.57abc	221.00 ± 2.84bc	215.20 ± 2.31cd	215.00 ± 4.58cd	230.00 ± 5.50ab	221.34 ± 0.97bc	225.80 ± 1.91abc	224.20 ± 4.78abc	233.67 ± 2.52a	221.91 ± 2.43abc
L*	89.75 ± 0.238a	84.32 ± 0.13cd	82.79 ± 0.05f	84.07 ± 0.16cde	84.03 ± 0.37cde	83.40 ± 0.10ef	83.61 ± 0.05def	84.23 ± 0.04cde	84.20 ± 0.26cde	84.57 ± 0.20c	85.66 ± 0.41b	86.28 ± 0.23b	84.02 ± 0.79cde
a*	0.36 ± 0e	2.74 ± 0.01a	2.22 ± 0.06c	2.58 ± 0.04b	2.33 ± 0.04c	1.73 ± 0.06d	1.75 ± 0.04d	2.62 ± 0.06ab	2.34 ± 0.06c	2.51 ± 0.06b	2.25 ± 0.03c	2.21 ± 0.02c	2.27 ± 0.03c
b*	18.57 ± 0.127g	27.08 ± 0.14cd	27.37 ± 1.40cd	26.57 ± 0.31d	30.29 ± 0.05b	23.31 ± 0.10e	23.32 ± 0.03e	31.82 ± 0.28a	30.29 ± 0.16b	27.88 ± 0.32c	20.50 ± 0.05f	22.64 ± 0.03e	30.32 ± 0.25b

Note: Data are presented as the mean standard deviation; data values of each parameter with distinct letters in rows are significantly different (p < 0.05).

Abbreviations: CL, cooking loss; DPPH, antioxidants by DPPH method; OCT, optimal cooking time; PPE, pumpkin peel; PPU, pumpkin pulp; PS, pumpkin seed; TFC, total flavonoid content; TPC, total phenolic compounds; WAC, water absorption capacity.

3.2.2. Quantification of Bioactive Compounds and Antioxidant Activity

The addition of pumpkin by‐products in pasta formulations significantly (p < 0.05) affected their bioactive compound composition. The highest TPC was observed in Mixture 2 (114.07 mg GAE/100 g DP), and the lowest TPC was observed in Mixture 6 (99.33 mg GAE/100 g DP). These values are similar to those reported by Farzana et al. (2023), who found TPC values ranging from 52.88 ± 2.11 mg GAE/100 g DP to 162.53 ± 2.07 mg GAE/100 g DP in pumpkin flour‐enriched noodles. The results are lower than those of Aljobair (2024), who enriched bread with pumpkin, cucumber, and watermelon peel flour and reported a TPC value of 6.56 ± 0.27 mg GAE/g DP for 10% substitution. Algarni (2020) assessed the TPC of a sponge cake enriched with between 5% and 20% of pumpkin flour and found high values ranging from 7.50 ± 1.05 to 16.18 ± 2.04 mg GAE/g.

For TFC, values ranged from 14.63 to 29.94 mg QE/100 g DP, and the enrichment markedly increased their concentration compared with the control (0.78 mg QE/100 g DP). Hussain et al. (2022) developed biscuits using different substitution levels of PPU, PPE, and PS and noticed that biscuits containing 15% PS powder presented the highest TFC (60.74 ± 0.10 mg CE/100 g DP).

The carotenoid content of the pasta mixtures ranged from 6.879 to 11.265 mg β‐CE/100 g DP, revealing that, as expected, mixtures with higher pulp concentrations contained the greatest carotenoid levels. Indeed, Ninčević Grassino et al. (2023) demonstrated that the total carotenoid content and profile depend on the species, cultivar, and parts of the pumpkin (pulp, peel, or seed), with the pulp generally containing the highest amount of carotenoids such as β‐carotene, α‐carotene, lutein, and zeaxanthin, while the seeds contain significantly less carotenoids. These findings are in line with those of Hussain et al. (2022), who established that biscuits enriched with PPU powder had higher carotenoid yields than those enriched with PPE or PS. Similar trends were reported by Algarni (2020) for sponge cake enriched with 20% pumpkin flour (8.34 ± 1.58 mg/100 g). The outcomes are higher than those reported by Kampuse et al. (2015) for wheat bread enriched with 20% of pumpkin residue powder, with a mean value of 2.21 mg/100 g DW.

DPPH free radicals were retained to assess the antioxidant potential of the pasta formulations. The control pasta exhibited an inhibition potential (DPPH) of only 2.45%, which increased considerably with the incorporation of pumpkin products in pasta, fluctuating from 12.55% to 16.03%. Cupcakes fortified with 10% pumpkin powder exhibited a comparable inhibition potential of 11.69% (Sello and Mostafa 2017). In addition, the supplementation of bread with 10% vegetable peel powder (green pumpkin, watermelon, and cucumber) exhibited a comparable DPPH inhibition capacity of 10.08 ± 0.01% (Aljobair 2024).

Pumpkin is an abundant source of diverse antioxidants, including polyphenols and carotenoids, which collectively contribute to its strong health‐promoting properties. Significant variation in antioxidant composition and concentration has been observed among different parts of the field pumpkin (Olowolajua et al. 2024). Previous studies have shown that PPU, PPE, and PS present very high TPC, TFC, and carotenoid contents and possess strong antioxidant potential (Asif et al. 2017; Bahramsoltani et al. 2017). Therefore, incorporating pumpkin into pasta not only increases its antioxidant capacity but also enriches it with bioactive compounds that promote health, making pumpkin an excellent functional ingredient for improving dietary quality.

3.2.3. Cooking Properties

The gluten network and starch gelatinization, along with their ability to retain water, are closely linked to the cooking quality of pasta. Supplementary ingredients introduced into wheat semolina can interact with the gluten–starch network and compete for available water, which may hinder optimal matrix development and influence cooking behavior (Lucas‐González et al. 2020; La Gatta et al. 2017).

CL was the first parameter assessed to evaluate pasta cooking quality; it reflects the quantity of dry matter released into the cooking water. For optimal quality, CL should be as low as possible and should not exceed 8% (Dick and Youngs 1988). The extruded pasta mixtures were characterized by a low CL between 4.95% and 5.67%. These values are elevated compared to the control (4%). The increase in CL may be a consequence of the breakdown in the interaction between proteins and starch, combined with an uneven distribution of water within the pasta matrix. The competitive water absorption of dietary fibers limits the amount of water available, which in turn inhibits the swelling of starch granules (Purnima et al. 2012). In addition, amylose is released from the inside of the starch granule due to high CL, and a significant loss of amylose could increase its stickiness, leading to potential customer rejection (Vimercati et al. 2020). Similar trends have been observed in extruded pasta fortified with 5%–25% detoxified Matri flour (4.8%–5.8%) (N. Ahmad et al. 2018), pasta fortified with tomato peel (Kadri et al. 2024), and pasta enriched with onion skin (Michalak‐Majewska et al. 2020).

The addition of pumpkin products in pasta had a significant effect (p < 0.05) on the OCT, which is an important parameter for the product's purchase. This parameter decreased significantly (p < 0.05) after the addition of the mixtures from 10 min (control) to 7.26–9.66 min. The lowest OCT (7.26 min) was recorded in Mixture 11, and the highest was recorded in Mixture 5. Some researchers have demonstrated a strong correlation between the reduction in OCT and the increase in dietary fiber content, which facilitates the hydration and gelatinization of starch granules (Bianchi et al. 2021). Pumpkin products are known to be rich in dietary fibers, particularly pectin, present in the peel (Villamil et al. 2023; Nyam et al. 2013). The findings are consistent with those of Oyeyinka et al. (2021) for pasta enriched with Bambara groundnut, Teterycz et al. (2020) for legume‐enriched pasta, and Reddy Surasani et al. (2019) for pasta supplemented with protein isolate from pangas processing waste. These results are also much lower than those reported by Filipčev et al. (2023) for wild garlic pasta and Kamali Rousta et al. (2021) for multigrain pasta.

WAC is the ability of a sample to store water inside its structure. The WAC results ranged from 208.6% for Mixture 1 to 233.6% for Mixture 11, which aligns with the results of Farzana et al. (2023). It is important to highlight that WAC is strongly associated with the protein network, as proteins can form structures similar to gluten (Carpentieri et al. 2024; Kushwaha et al. 2026). In addition, flour is more suitable for baking applications if it has a higher capacity to absorb water (Falade and Amadi 2024). Generally, a high‐quality pasta without the inclusion of additional ingredients has WAC in the range of 150–200 g of water/100 g pasta (Bianchi et al. 2021).

3.2.4. Evaluation of Color Parameters

Pasta color is a key parameter in the appreciation of the consumer from the sensory perspective. Consumer preferences have shifted toward natural food colorants due to safety concerns and perceptions of chemical‐free products. This trend extends to pasta enrichment, where plant‐based pigments provide vibrant colors while enhancing nutritional profiles with bioactive compounds (Axentii et al. 2023; Özyurt et al. 2015).

The addition of pumpkin products to pasta formulations affected their color parameters, where a significant increase (p < 0.05) in b* and a decrease in lightness L* compared with the control (18.57 and 89.75, respectively) were observed, likely attributed to the presence of pumpkin carotenoids, which impart an appealing orange‒yellow color to the pasta. The pasta formulations exhibited positive a* values, indicating a shift toward red tones rather than the greenish hues observed in the control pasta (a* = 0.36). b*, a*, and L* values ranged between 20.5 and 31.82, between 1.73 and 2.74, and from 82.79 to 86.28, respectively. Similar results were reported by Nanthachai et al. (2020) and Kampuse et al. (2015), who observed increases in b* and L* values following the addition of pumpkin flour to instant noodles and bread, respectively. In a study reported by Mirhosseini et al. (2015), on gluten‐free pasta enriched with pumpkin flour, a* values increased from 0.94 ± 0.28 for the control to 1.35 ± 0.41 for 25% pumpkin pulp flour (PPF) incorporation. Cookies enriched with PPE and PS from 5% to 20% revealed comparable outcomes in terms of b* to current results (Asaduzzaman et al. 2025).

3.3. ANOVA Analysis

Table 5 shows the model performances for each studied parameter. These performances include statistical parameters such as R ², p value, lack of fit, and RMSE. Very high coefficients of determination (R ²) were recorded, ranging between 0.92 and 1. These values were very close to the $R_{\mathrm{adj}}^2$ values ranging between 0.82 and 0.99, with all constructed models being significant (p < 0.05). Henika (1982) reported that the coefficients of determination for constructed models of more than 0.75 are relatively adequate for the prediction of parameters. The results also suggested that the special cubic model was the best fit for ash, protein, fat, carbohydrates, TPC, DPPH, OCT, a*, and b*, whereas the linear models were best fit for moisture, TFC, CL, WAC, and L*, and the quadratic model was best fit for carotenoids. Table 6 shows the different polynomial equations developed and used to illustrate the relationship between formulation components and response variables.

TABLE 5.

ANOVA of the experimental model.

	Linear terms			Nonlinear terms				Model
Responses	X 1	X 2	X 3	X 1 X 2	X 1 X 3	X 2 X 3	X 1 X 2 X 3	R 2	${\mathit{R}}_{\mathbf{adj}}^2$	p	Lack of fit (Prob. > F)	RMSE	CV
Moisture	10.029***	9.802***	10.102***	−0.293418	−0.725	−1.890	−2.555	0.99	0.98	< 0.0001	0.194	0.02	0.206
Ash	1.260***	−0.852**	−0.657*	3.483***	3.798***	8.659***	−5.863**	0.99	0.98	< 0.0001	0.732	0.014	1.033
Protein	15.070***	45.114***	46.220***	−62.149**	−59.554**	−125.167**	120.717*	0.98	0.95	< 0.001	0.127	0.39	2.924
Fat	3.014***	8.382***	9.207***	−19.337***	−15.534**	−25.167**	36.723**	0.99	0.98	< 0.0001	0.115	0.0815	4.981
Carbohydrates	70.401***	37.575***	35.217***	78.141**	71.790**	143.293***	−148.5625**	0.99	0.98	< 0.0001	0.26	0.347	0.404
TPC	113.852***	118.958***	167.841***	−66.691	−107.476*	−174.810*	322.09*	0.97	0.94	< 0.001	0.292	1.233	1.167
TFC	22.577**	38.174*	7.005	−1.979	4.091	−12.468	20.783	0.98	0.95	< 0.001	0.05	0.96	4.438
Carotenoids	16.113***	10.033*	13.377**	−18.629	−26.898*	−192.359	263.563	0.99	0.97	0.0001	0.058	0.2292	2.84
DPPH	14.772***	2.921	10.215**	20.118*	9.13	24.479*	6.273	0.99	0.97	< 0.001	0.179	0.207	1.401
CL	5.453***	2.620*	5.966***	3.853	−2.815	3.262	11.128	0.97	0.94	< 0.001	0.576	0.0628	1.19
OCT	7.165***	2.128**	3.379***	10.413***	12.028***	27.506***	−46.471***	1	0.99	< 0.0001	0.205	0.025	0.301
WAC	216.038***	236.92**	319.147***	35.465	−161.476	−250.848	−192.48	0.92	0.82	0.012	0.144	2.954	1.327
L*	83.686***	96.619***	81.307***	−14.926	6.8153	−21.888	2.954	0.96	0.91	0.0024	0.082	0.278	0.33
a*	3.616***	3.756**	3.471**	−5.736*	−4.212*	−7.488*	14.562*	0.99	0.98	< 0.001	0.258	0.0516	2.244
b*	18.686**	−57.947**	−1.803	170.708**	72.262	212.526**	−236.761*	0.97	0.93	0.0015	0.2	0.974	3.636

TABLE 6.

Mathematical model of the different responses.

Response	Mathematical model
Moisture	10.029X 1 + 9.802X 2 + 10.102X 3
Ash	1.260X 1 − 0.852X 2 − 0.657X 3 + 3.483X 1 X 2 + 3.7984X 1 X 3 + 8.659X 2 X 3 − 5.863X 1 X 2 X 3
Protein	15.307X 1 + 45.114X 2 + 46.22X 3 − 62.149X 1 X 2 − 59.554X 1 X 3 − 125.167X 2 X 3 + 120.717X 1 X 2 X 3
Fat	3.014X 1 + 8.382X 2 + 9.207X 3 − 19.337X 1 X 2 − 15.534X 1 X 3 − 25.167X 2 X 3 + 36.723X 1 X 2 X 3
Carbohydrates	70.401X 1 + 37.575X 2 + 35.217X 3 + 78.141X 1 X 2 + 71.79X 1 X 3 + 143.293X 2 X 3 − 148.5625X 1 X 2 X 3
TPC	113.852X 1 + 118.958X 2 + 167.841 × 3 − 107.476X 1 X 3 − 174.81X 2 X 3 + 322.09X 1 X 2 X 3
TFC	22.577X 1 + 38.174X 2
Carotenoids	161.13X 1 + 10.0329X 2 + 13.3777X 3 − 26.8988X 1 X 3
DPPH	14.772X 1 + 10.215X 3 + 20.118X 1 X 2 + 24.479X 2 X 3
CL	5.453X 1 + 2.62X 2 + 5.966X 3
OCT	7.165X 1 + 2.128X 2 + 3.379X 3 + 10.13X 1 X 2 + 12.028X 1 X 3 + 27.506X 2 X 3 − 46.471X 1 X 2 X 3
WAC	216.038X 1 + 236.92X 2 + 319.147X 3
L*	83.686X 1 + 96.619X 2 + 81.307X 3
a*	3.616X 1 + 3.756X 2 + 3.471X 3 − 5.736X 1 X 2 − 4.212X 1 X 3 − 7.488X 2 X 3 + 14.562X 1 X 2 X 3
b*	18.686X 1 − 57.947X 2 + 170.708X 1 X 2 + 212.526X 2 X 3 − 236.761X 1 X 2 X 3

3.4. The Influence of Pasta Enrichment on Response Characteristics

3.4.1. Physicochemical and Antioxidant Properties

Linear terms were mostly found to have a statistically significant (p < 0.05) positive effect. On the other hand, interaction terms between pulp and peel (X ₁ X ₂) were also significant (p < 0.05) for all responses except for moisture, TPC, and TFC. Interaction terms between pulp and seed (X ₁ X ₃) were not significant for moisture, TFC, and DPPH. The X ₂ X ₃ interaction was not significant only for moisture and TFC. A three‐way interaction was also observed to be significant (p < 0.05) in almost all responses except for moisture, TFC, and DPPH.

3.4.2. Cooking Properties

Linear terms were found to be positively significant in all the responses related to cooking quality. Interaction terms were only significant (p < 0.05) for OCT, with a positive effect between the factors, but the three‐way interaction (X ₁ X ₂ X ₃) was negative.

3.4.3. Color Property

Linear terms were all significant for the color (positive effect for L* and a* and negative for b*). The interaction between X ₁ X ₂ and X ₂ X ₃ was significant for a* and b*, and the interaction X ₁ X ₃ was significant (p < 0.05) in a*. The three‐way interaction (X ₁ X ₂ X ₃) was positively significant (p < 0.05) for a* and negatively significant for b*.

3.5. Contour Plot Interpretation

Figure 1 displays the ternary contour plots created using model equations within the restricted area. The contour plots (Figure 1a) clearly showed that the moisture values progressively increased with increasing PPU values. This result was expected due to the high moisture content of PPU. The ash values increased when the contour plot increased toward the PS flour vertex and declined in the PPU (Figure 1b). The same trend was observed regarding protein and fat (Figure 1c,d), considering that PS flour has a greater concentration of these nutrients than PPU flour. On the other hand, the carbohydrate and carotenoid contents (Figure 1e,h) increased with increasing amounts of PPU flour.

FIGURE 1.

Ternary contour plots of the effect of pumpkin product enrichment on moisture (a), ash (b), protein (c), fat (d), carbohydrates (e), TPC (f), TFC (g), carotenoids (h), DPPH (i), CL (j), OCT (k), WAC (l), L* (m), a* (n), and b* (o).

Figure 1f,g,i indicate that increasing the proportion of PPU flour improved TPC, TFC, and DPPH in pasta formulation. This trend can be attributed to the high concentrations of phenolic compounds naturally present in PPU, which are well recognized for their potent antioxidant activity (Oufighou et al. 2024).

The incorporation of increased proportions of PPU flour considerably enhanced the nutritional profile of the pasta and improved its functional attributes by augmenting its antioxidant capacity, thereby contributing to the mitigation of oxidative stress and supporting overall health benefits (Farzana et al. 2023).

CL and WAC (Figure 1j,l) were positively affected by the addition of a higher quantity of PPU flour. Indeed, Kamali Rousta et al. (2020) reported that the augmentation in CL is directly related to the reduction in the gluten network. The high fiber content and water‐binding capacity of pumpkin also allow for meaningful water absorption through hydrogen bonding (Khan et al. 2019).

The contour plots of the color parameters L*, a*, and b* (Figure 1m–o, respectively) showed that PPU flour addition raised the pasta lightness and yellowness. This is due to the presence of the β‐carotene pigment responsible for the pumpkin color (Farzana et al. 2023). The same results were reported by Othman et al. (2023), who stated that the color intensity of the doughnut increased as more pumpkin powder was incorporated, largely attributable to the high levels of carotenoids, which are natural pigments present in pumpkin.

3.6. Validation of Pumpkin Pasta Formulation

Enhancing nutritional, cooking, and sensory qualities was the main goal of the explicit optimization of PPU, PPE, and PS powders mixed with wheat flours.

To achieve desirable product quality features, multi‐response optimization was used in conjunction with the desirability function approach. The individual desirability results (d_i ) presented in Figure 2 ranged between 0.75 (DPPH) and 0.95 (carbohydrates). These results are considered satisfactory according to the desirability scale since d_i  < 0.3 denotes unsatisfactory performance (Boateng 2023). When the overall desirability function is maximized to 86.04%, the predicted model constructed using coded values from the optimized formulation is 0.79 PPU, 0 PPE, and 0.21 PS, corresponding to the real mass fraction in grams: 11.85 g of PPU powder, 0 g of PPE powder, and 3.15 g of PS powder. This maximized formulation achieved a moisture content of 9.66%, ash content of 1.49%, protein content of 11.89%, fat content of 1.63%, and carbohydrate content of 75.06%. The values achieved for TPC, TFC, carotenoids, and DPPH were 107.22 mg GAE/100 g DW, 19.27 mg QE/100 g DW, 10.17 mg β‐CE/100 g DP, and 15.34%, respectively. In addition, CL, OCT, and WAC reached values of 5.04%, 8.5 min, and 211.86%, respectively. For color parameters, the L*, a*, and b* values were 84.38, 2.76, and 27.3, respectively.

FIGURE 2.

Desirability function plot. CL, cooking loss; OCT, optimal cooking time; TFC, total flavonoid content; TPC, total polyphenolic content; WAC, water absorption capacity.

These findings demonstrate an effectively optimized mixture that maximizes nutritional, cooking, and sensorial components while satisfying the defined desirability requirements. The optimization process is shown to be reliable, as evidenced by the predicted values falling within tight confidence intervals. To validate the model, pasta was prepared in triplicate using the optimized formulation. No significant differences (p > 0.05) were perceived between the different experimental and predicted responses (Table 7).

TABLE 7.

Predicted and experimental results.

Variables	Moisture	Ash	Protein	Fat	Carbs	TPC	TFC	Carotenoids	DPPH	CL	OCT	WAC	L*	a*	b*
Selected model	Linear	Special cubic	Special cubic	Special cubic	Special cubic	Special cubic	Linear	Quadratic	Special cubic	Linear	Special cubic	Linear	Linear	Special cubic	Special cubic
Predicted values	9.91	1.49	11.89	1.64	75.065	107.21	19.31	10.22	15.34	5.04	8.50	211.79	84.38	2.77	27.27
Experimental value	9.89 ± 0.03	1.48 ± 0.03	11.673 ± 0.371	1.62 ± 0.15	75.34 ± 0.42	106.71 ± 1.24	18.93 ± 0.41	10.27 ± 0.69	15.07 ± 0.16	5.08 ± 0.02	8.60 ± 0.17	210.59 ± 1.30	84.28 ± 0.46	2.72 ± 0.04	27.24 ± 1.62
Desirability function	0.78	0.85	0.79	0.92	0.98	0.88	0.8	0.79	0.75	0.87	0.95	0.82	0.94	0.85	0.94

Abbreviations: Carbs, carbohydrates; CL, cooking loss; OCT, optimal cooking time; TFC, total flavonoid content; TPC, total polyphenolic content; WAC, water absorption capacity.

3.7. Sensory Analysis

3.7.1. Preference Mapping

A trained panel conducted a sensory analysis of the different mixtures of the DMD. From a sensory standpoint, this research enabled us to identify which of these pairings was the most favored. Preference mapping (Figure 3) depicts the relation between consumer preference and descriptive sensory data. This approach contributes to product marketing, research, and development by producing foods that satisfy consumer needs (Olawoye and Gbadamosi 2020).

FIGURE 3.

Preference mapping of the different pasta mixtures and their attributes. Mixture 1: (12 g PPU, 3 g PS), Mixture 2: (6.3 g PPU, 2.7 g PPE, 6 g PS), Mixture 3 (10.2 g PPU, 4.8 g PS), Mixture 4, 8, 12 (8.4 g PPU, 3.3 g PPE, 3.3 g PS), Mixture 5, 6 (3 g PPU, 6 g PPE, 6 g PS), Mixture 7 (12 g PPU, 3 g PPE), Mixture 9 (9 g PPU, 6 g PS), Mixture 10 (6.15 g PPU, 6 g PPE, 2.85 g PS), Mixture 11 (9 g PPU, 6 g PPE).

The control wheat pasta achieved the highest scores in appearance, taste, and overall acceptability, with a preference level of 100%. Among all the formulations, two mixtures were placed in the red zone within an 80%–100% preference percentage. These mixtures corresponded to the fractions: 12 g PPU, 0 g PPE, and 3 g PS (Mixture 1) and 12 g PPU, 3 g PPE, and 0 g PS (Mixture 7). Both mixtures were highly preferred regarding their very smooth appearance and their very pleasant taste. It should be highlighted that a substantial amount of PPU powder with a small amount of PPU and PS are included in these mixes. According to Ersedo (2019), bread enriched with 10% PPU flour was the most preferred for its good appearance, but an increase in the supplementation level led to a reduction in the appreciation. In another study, the addition of up to 15% PPU flour to noodles improved the flavor and taste but diminished the textural quality (Farzana et al. 2023). The other combinations were the least preferred, all placed in the blue zone, with only 18% of panelists expressing satisfaction for Mixtures 3 and 11, and 9% for the other mixtures. These mixtures were characterized by a strong smell, dark color, and a rough texture originating from PPE powder due to a significant portion present in each of these mixtures. The same pattern was noted in the addition of vegetable peel powders (green pumpkin, watermelon, and cucumber) into bread, where color, taste, and flavor were negatively impacted (Aljobair 2024).

3.7.2. Radar Plots

Figure 4 depicts the radar chart of the sensory analysis for pasta attributes. The results demonstrate that control pasta and Mixtures 1 and 7 gave the best taste with the highest score compared to other samples (4.53 in “control,” 4.4 in “Mixture 1,” and 3.76 in “Mixture 7”). Regarding their aspect, they were rated as “soft” compared to other mixtures considered to have a “rough” aspect. Considerable differences were observed for color intensity between pasta formulations; Mixtures 4, 8, 10, 11, and 12 (scores of 3.3, 3.4, 3.46, 3.4, and 3.26, respectively) were rated as “dark” due to the presence of pumpkin peel powder. The color intensity for other mixtures was rated medium to light for Mixture 1 (score 1.9). The same tendency was observed for the smell attribute, where Mixtures 4, 8, and 12 received the highest intensity scores (3.06, 3.2, and 3.02) compared to the control, Mixture 1, and Mixture 7, which were considered to have a light smell. Considering stickiness and texture attributes, low intensities were observed, indicating pasta formulations with low stickiness and a soft texture. To determine the significance of differences observed in sensory properties from descriptive sensory analysis, the results were subjected to ANOVA (Table 8). Color, aspect, and taste presented significant differences (p > 0.05), but no significant differences were observed for smell, texture, and stickiness (p > 0.05).

FIGURE 4.

Radar plot of the sensory profile of the 12 pasta mixtures with the control. Mixture 1: (12 g PPU, 3 g PS), Mixture 2: (6.3 g PPU, 2.7 g PPE, 6 g PS), Mixture 3 (10.2 g PPU, 4.8 g PS), Mixture 4, 8, 12 (8.4 g PPU, 3.3 g PPE, 3.3 g PS), Mixture 5, 6 (3 g PPU, 6 g PPE, 6 g PS), Mixture 7 (12 g PPU, 3 g PPE), Mixture 9 (9 g PPU, 6 g PS), Mixture 10 (6.15 g PPU, 6 g PPE, 2.85 g PS), Mixture 11 (9 g PPU, 6 g PPE).

TABLE 8.

Analysis of sensory evaluation significance through ANOVA on a 5‐point scale.

Sample	Color	Smell	Texture	Stickiness	Aspect	Taste
Control	1.70 ± 0.70d	1.33 ± 0.48b	2.01 ± 0.53b	1.86 ± 0.63b	4.33 ± 0.79a	4.53 ± 0.51a
Mixture 1	1.90 ± 0.91c	1.90 ± 0.79ab	2.21 ± 0.79ab	2.72 ± 0.79ab	3.80 ± 0,53ab	4.40 ± 0.73a
Mixture 2	2.33 ± 0.89abcd	2.73 ± 0.88a	2.50 ± 0.74ab	2.50 ± 1.08ab	2.60 ± 0.98c	2.33 ± 0.72de
Mixture 3	2.35 ± 0.59bcd	2.50 ± 0.99a	2.31 ± 0.72ab	2.90 ± 1.23ab	3.20 ± 0.56bc	1.93 ± 0.79e
Mixture 4	3.30 ± 0.99ab	3.06 ± 0.88a	3.00 ± 0.84a	2.81 ± 0.94ab	2.68 ± 1.06c	3.23 ± 0.86bcd
Mixture 5	2.60 ± 0.82abcd	2.53 ± 0.74a	2.60 ± 0.91ab	2.53 ± 0.96ab	2.63 ± 1.01c	2.66 ± 0.97cde
Mixture 6	3.23 ± 0.96ab	2.80 ± 0.86a	2.85 ± 0.63ab	2.50 ± 0.89ab	2.87 ± 0.70c	3.40 ± 1.05bc
Mixture 7	2.53 ± 0.74abcd	2.13 ± 0.63ab	2.28 ± 0.61ab	2.20 ± 0.70ab	3.50 ± 0.86bc	3.76 ± 0.41ab
Mixture 8	3.40 ± 0.58ab	3.20 ± 0.88a	3.10 ± 0.84a	3.00 ± 0.94ab	2.60 ± 1.06c	3.00 ± 0.86bcd
Mixture 9	2.71 ± 1.01abc	2.57 ± 1.05a	2.57 ± 0.89ab	2.78 ± 0.88ab	2.93 ± 0.91bc	3.13 ± 0.74bcd
Mixture 10	3.46 ± 0.63a	2.33 ± 0.61ab	2.60 ± 0.73ab	2.66 ± 0.81ab	2.56 ± 0.70c	3.20 ± 1.01bcd
Mixture 11	3.40 ± 0.79ab	2.25 ± 0.86ab	2.66 ± 0.84ab	2.28 ± 1.05ab	2.68 ± 0.83c	2.13 ± 0.63e
Mixture 12	3.26 ± 0.59ab	3.02 ± 0.68a	3.00 ± 0.63a	2.81 ± 0.94ab	2.69 ± 0.90c	3.00 ± 0.44bcd

Note: Within the same column, distinct letters denote significant differences (p < 0.05) based on Tukey's test. Scores: color (1 = very light/5 = very dark), smell (1 = very light/5 = very strong), texture (1 = very smooth/5 = very rough), stickiness (1 = nonsticky/5 = very sticky), aspect (1 = very soft/5 = very hard), taste (1 = very bad/5 = very good).

The results of the sensory analysis demonstrated that one of the most preferred mixtures by the panelists matches the optimal formulation predicted by the MD. The ability to develop a food product that blends sensory and nutritional aspects is an essential advantage.

3.8. Fiber and Mineral Content of the Optimized Formula

The incorporation of pumpkin products into pasta powder led to a significant increase in the dietary fiber content, rising from 0.2 g/100 g in the control to 5.6 g/100 g in the optimized formula. It is well known that pumpkin products are a great source of dietary fiber (G. Ahmad and Khan 2019). The incorporation of PS and rinds at a level of 5% into bread increased the total dietary fiber content compared with the control from 2.3% to 4.3% and to 3% for PS and rinds, respectively (Nyam et al. 2013). A similar trend was observed in pumpkin blended cake, where the fiber amount increased from 0.83% in T1 (100/00: refined wheat flour/pumpkin) to 1.85% in T8 (70/30: refined wheat flour/pumpkin) (Bhat and Bhat 2013).

The enrichment of wheat pasta with pumpkin products at 15% resulted in considerable increases in Mg, Al, Si, P, Cl, Ca, and above all K (Table 9). Indeed, the inclusion of mineral‐rich components in processed food products serves as a strategic approach to combat mineral deficiencies (Oliveira et al. 2023). Numerous studies have demonstrated that pumpkin products are an excellent source of potassium, predominantly found in high amounts in the flesh, and calcium, which is mainly concentrated in the seeds (Adubofuor et al. 2016; Amin et al. 2019). Potassium plays a direct role in establishing the membrane potential, thereby influencing the excitability of nerve and muscle cells (Stryjecka 2025). These high levels of calcium and potassium content in pumpkin make it an ideal food for middle‐aged and elderly individuals by helping to prevent osteoporosis and manage hypertension (Hussain et al. 2022).

TABLE 9.

Mineral content of the optimized pasta formulation.

Elements	Control (%)	Pasta (%)
C6H10O5	99.1600	98.540
Na	0.0220	0.0280
Mg	0.0743	0.1310
Al	0.0140	0.0200
Si	0.0458	0.0786
P	0.1850	0.2610
S	0.1230	0.1440
Cl	0.0645	0.1350
K	0.2650	0.5636
Ca	0.0412	0.0892
Cr	0.0005	0.0004
Mn	0.0010	0.0009
Fe	0.0010	0.0021
Cu	0.0001	0.0002
Zn	0.0015	0.0020
Br	0.0004	n.d.
Rb	n.d.	n.d.
Sr	n.d.	0.0008
Zr	n.d.	n.d.
Ba	n.d.	n.d.

Abbreviation: n.d., not determined.

3.9. UHPLC–DAD Analysis of Carotenoid Content of the Optimized Pasta

Carotenoids from the optimized pasta formulation were characterized using UHPLC–DAD analysis (Figure S1). The carotenoid contents in pasta were quantified by referencing calibration curves, and the findings were 4.34 ppm of lutein and 1.17 ppm of β‐carotene. Our findings are consistent with other research studies (Ninčević Grassino et al. 2023) reporting that lutein was the most abundant carotenoid in PS and that α‐carotene and β‐carotene were mostly abundant in PPU and PPE. Similar profiles were also observed by (Gavril et al. 2024). Many studies have reported the carotenoid content in pumpkin extracts, such as Pinna et al. (2022), who found an amount of α‐carotene and β‐carotene in the extract of Cucurbita moschata varieties of 12.60 and 19.45 µg/g, respectively. In fresh Hokkaido pumpkins, (Atencio et al. 2022) found β‐carotene to be more abundant in PPU than in PPE and PS, with a mean value of 15.04 ± 0.38 µg/g pumpkin FW, while lutein was more abundant in PPE (14.10 ± 2.37 µg/g pumpkin FW). Matić et al. (2024) optimized the extraction of carotenoids from C. moschata L. using an eco‐friendly method, and the contents of β‐carotene and α‐carotene were 3.53 ± 0.24 and 3.88 ± 0.19 mg/100 g DW, respectively. However, data on the carotenoid levels in pasta enriched with pumpkin are currently lacking.

4. Conclusion

The current investigation aimed mainly to enhance the physicochemical, nutritional, and sensory qualities of pasta, which is extensively utilized and consumed across the world due to its fundamental role in diets worldwide, contributing significantly to daily caloric intake and nutritional sustenance. A DMD was successfully employed to develop a new pasta formulation incorporating three pumpkin by‐products. This statistical approach enabled the achievement of an optimized pasta formulation exhibiting optimal physicochemical and nutritional properties by the incorporation of 15% pumpkin by‐product flours (11.85% PPU flour and 3.15% PS flour). Modeling of the experimental data facilitated the development of predictive equations capable of estimating the system behavior under various factor combinations. Interestingly, the significant increase in mineral, fiber, and carotenoid content in the pumpkin‐enriched pasta demonstrated both qualitative and quantitative improvements compared with conventional wheat pasta. The sensory analysis demonstrated the high appreciation of the optimal formula, with preference levels between 80% and 100%. Overall, this study confirmed that pasta can be effectively enriched with pumpkin by‐products to enhance its nutritional value and to create functional food products. Nevertheless, improving pasta quality through extrusion parameter optimization remains important for facilitating industrial scale‐up. In addition, future studies focusing on the nutritional bioavailability and bioaccessibility of these functional ingredients, along with product shelf life, are needed to clarify the health benefits of fortified pasta.

Author Contributions

Amira Oufighou: writing – original draft, methodology, conceptualization, formal analysis, data curation, investigation, validation. Fatiha Brahmi: conceptualization, formal analysis, methodology, supervision, writing – review and editing. Sabiha Achat: supervision, validation, formal analysis. Leila Smail‐benazzouz: formal analysis, methodology, software. Younes Arroul: writing – review and editing. Sarah Slimani: writing – review and editing. Sidahmed Saadi: writing – review and editing. Lila Boulekbache‐Makhlouf: project administration. Francisco Artés‐Hernández: resources, writing – review and editing. Noelia Castillejo: writing – review and editing.

Funding

This work is a result of the AGROALNEXT program and was supported by MICIU with funding from EU NextGeneration (PRTR‐C17.I1) and by the Seneca Foundation with funding from the Autonomous Community of the Region of Murcia (CARM).

Ethics Statement

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Figure S1‐S2: jfds71172‐sup‐0001‐FigureS1‐S2.docx

Data Availability Statement

The original contributions presented in the study are included in the article and Supporting Information, and further inquiries can be directed to the corresponding author.