Enhancing Gluten-Free Muffins with Cornelian Cherry (Cornus mas L.) and Carob–Taro–Rice Flour Blend: A Functional and Bioactive Approach

Featured Application

This study presents a complementary strategy for the gluten-free functional bakery sector. The approach successfully enables the development of nutritionally-enriched, sensorially-pleasing, and technologically-enhanced gluten-free muffins, effectively addressing common deficiencies in this product category.

Abstract

This study introduces a complementary ingredient strategy to improve gluten-free muffins by combining the bioactive properties of Cornelian cherry (Cornus mas L.) with the techno-functional advantages of a carob-taro-rice flour blend. A rice-only formulation served as the control, while other formulations included partial substitution with carob and/or taro flours, enriched with 0%, 4%, or 8% Cornelian cherry pulp, and were evaluated using a comprehensive set of physicochemical, textural, microstructural, and sensory analysis. The incorporation of pulp and flour substitution markedly increased total phenolic content, antioxidant capacity, and dietary fiber, reaching up to 7.9 times the levels observed in the rice-only control. Carob flour substitution reduced muffin hardness by 51–64%, indicating substantial textural improvement. Quantitative Descriptive Analysis (QDA), Principal Component Analysis (PCA), and heatmap clustering confirmed that carob and taro enhanced the sensory profile by increasing crumb porosity and reducing firmness. Image analysis supported these findings, showing that carob-containing blends exhibited a more desirable microstructure with larger air cell area (>48%) and greater circularity (>0.86), thus linking internal structure to improved texture. These results provide a practical approach for bakeries and food manufacturers to enhance the nutritional and sensory quality of gluten-free muffins.

1. Introduction

Gluten-free muffins are a dietary alternative for individuals with celiac disease, non-celiac gluten sensitivity, or wheat allergy [1]. However, the absence of gluten’s structural protein network often results in products with inferior texture and reduced consumer acceptability [2]. Quality limitations, particularly in crumb structure and mouthfeel, remain major barriers to consumer satisfaction in gluten-free bakery products, thereby reducing market appeal [3].

Plant-based flours are increasingly explored as alternatives to wheat in gluten-free formulations. Yet, replacing wheat flour often compromises product quality, yielding undesirable traits such as dryness and dense texture [1,3]. Rice (Oryza sativa), a staple for nearly half the world’s population, is the most widely used cereal in gluten-free products due to its neutral flavor, pale color, and functional versatility [4,5]. Nevertheless, rice flour is limited by its low fiber, mineral, and phytochemical content and poor functional properties such as weak textural stability and low water retention [6]. To overcome these shortcomings, complementary flours like taro (Colocasia esculenta L. Schott) and carob (Ceratonia siliqua L.) offer promising properties when blended with rice [7]. Taro flour, rich in amylopectin and mucilaginous compounds, improves moisture retention and mouthfeel [8,9], while carob flour contributes dietary fiber, natural sweetness, and antioxidant polyphenols [10,11], thereby supporting both texture and flavor diversity.

Cornelian cherry (Cornus mas L.) is a nutrient-rich fruit widely found in Türkiye, Romania, Bulgaria, Italy, and parts of Southern Europe and Southwest Asia [12]. It is valued for its high vitamin C content and bioactive compounds such as anthocyanins and phenolic acids [13]. Cornelian cherry is recognized for its notable bioactivity, encompassing antimicrobial, immunomodulatory, and potent antioxidant effects [14], which positions it as a promising functional food ingredient. Beyond its health potential, its vibrant color and distinctive tart-sweet flavor profile offer significant advantages for sensory enhancement. Cornelian cherry is already a well-established raw material in traditional food processing, particularly in its growing regions. As with other stone fruits, the fruit is routinely destoned, and the resulting pulp is the primary intermediate for a wide array of shelf-stable products such as jam, marmalade, pestil (a dried form of marmalade produced in the eastern part of Türkiye), paste, and sherbet [15,16]. It has also been successfully utilized in various food matrices such as meat, dairy, and confectionery products [13,17,18], but its application in gluten-free bakery products, especially muffins, remains largely unexplored. This gap presents a significant opportunity for novel ingredient strategies aimed at enhancing the quality and appeal of gluten-free products.

This study aimed to develop gluten-free muffins enriched with Cornelian cherry by employing a novel strategy based on a complementary flour blend. Rice flour served as a cost-effective base, blended with taro flour for moisture retention and carob flour for dietary fiber and natural sweetness. This approach targeted two key challenges in gluten-free baking: (i) improving technological functionality and nutritional quality through strategically formulated flour blends, and (ii) enhancing functionality and sensory appeal via bioactive enrichment from Cornelian cherry. By combining these ingredients in a functional bakery formulation, the study offers a practical strategy to improve both the quality and appeal of gluten-free bakery products.

2. Materials and Methods

2.1. Muffin Formulation

The base muffin formulation consisted of (per 100 g flour): white sugar (50%; Torku, Konya, Türkiye), whole eggs (50%; Evrenkaya, Afyonkarahisar, Türkiye), cow’s milk (30%; Torku, Konya, Türkiye), sunflower oil (30%; Orkide, İstanbul, Türkiye), and baking powder (4%; Dr. Oetker, İzmir, Türkiye). To evaluate the effects of alternative flours, rice flour was partially substituted with taro flour (Becos, Mersin, Türkiye) or carob flour (Tamtarım, Gümüşhane, Türkiye). Fresh Cornelian cherry (Cornus mas L.) fruits from the Cornaceae family were sourced from local markets (Afyonkarahisar, Türkiye). The mature fruits were carefully selected based on uniform size and appearance. Quality parameters of the selected fruits included pH of 3.21 ± 0.03, Brix value of 17.02° ± 0.02, and an average fruit weight of 3.29 ± 0.27 g. The fruits were blanched by immersing them in a stainless-steel basket in boiling distilled water for 2 min to soften them for destoning. They were then destoned using a 2 mm stainless steel sieve. The destoning process was considered complete upon visual inspection, with no visible stone fragments remaining. Finally, the destoned fruits were homogenized (Waring Blender, Model 8011, Torrington, CT, USA) to obtain fresh pulp for each trial.

The gluten-free muffin formulation used rice flour as the base material, chosen for its cost-effectiveness, availability, neutral flavor, and light color [4,7]. This control allowed for the evaluation of taro and carob flours as partial substitutes. Four flour blends (100% total) were prepared: (1) 100% rice (P), (2) 30% taro + 70% rice (PT), (3) 30% carob + 70% rice (PH), and (4) 30% taro + 30% carob + 40% rice (PHT). Each blend was enriched with 0%, 4%, or 8% Cornelian cherry pulp per 100 g flour, yielding twelve formulations. These were coded as P0, P4, P8 (rice-only controls); PT0, PT4, PT8 (rice-taro); PH0, PH4, PH8 (rice-carob); and PHT0, PHT4, PHT8 (rice-taro-carob).

Muffins were prepared following the established procedure [19], with slight modifications. Whole egg and sugar were mixed for 1 min at low speed using a hand mixer (Fakir Trinity, Fakir Werke GmbH, Bissendorf, Germany). Sunflower oil, milk, Cornelian cherry pulp, and the flour blend (with baking powder) were then added and mixed for 2 min. Portions of 30 ± 0.2 g batter were placed into paper cases in metallic trays and baked at 170 °C for 20 min in a preheated oven (Teka HAK 625 N, Teka Group, Mönchengladbach, Germany). This specific time-temperature combination was selected based on preliminary trials, as it ensured adequate internal structure setting without central collapse and produced a desirable golden-brown crust color (for rice-only controls) without excessive browning or drying, which was observed at higher temperatures. After cooling at room temperature for 30 min, muffins were packed in sealed plastic bags to minimize moisture loss and stored at 22 ± 1 °C. Each formulation was prepared in duplicate, and at least three muffins per batch were selected for evaluation. All analyses were performed on the muffins within 48 h of production, with texture and sensory evaluations completed within the first 24 h to ensure freshness.

2.2. Proximate Composition Analysis

The proximate composition of muffins was analyzed following AOAC official methods [20]: moisture (method 925.09), fat (method 920.39), total crude protein (method 954.01), ash content (method 923.03), and total dietary fiber (method 991.43). Available carbohydrates (CHO) were calculated by subtracting the percentages of all other measured components from 100%.

2.3. Baking Loss and Muffin Height

Baked muffins were cut vertically, and their maximum height was measured as the vertical distance from the base to the top at the center using a digital caliper. The weight of each batter portion in its muffin cup was recorded before baking. After baking and a 1 h cooling period at room temperature, the muffins in their cups were weighed again. Baking loss was calculated using Equation (1):

$$BakingLoss\left(\%\right)={\displaystyle \frac{\left(Batterweight-Muffinweight\right)\left(g\right)}{Batterweight\left(g\right)}}\times 100$$

(1)

2.4. Total Phenolic Content and Antioxidant Activity Analysis

Gluten-free muffin samples (3 g) were extracted with 22.5 mL methanol:acetone:water (1:1:1, v/v/v) following an established method [21], with minor modifications. The mixture was stirred (600 rpm, 30 min, IKA RT 15), centrifuged (6800× g, 4 °C, 30 min), and filtered. Analyses were performed on three muffins randomly selected from two independent batches. Total phenolic content (TPC) and DPPH radical scavenging activity were determined following a previously published method [22]. Results were expressed as mg gallic acid equivalents per g sample (mg GAE/g) for TPC and as percentage inhibition for DPPH.

2.5. Color Measurements

The color of the muffin crumb and crust was measured using a colorimeter (Minolta CR-400, Japan) with an 8 mm aperture size and D65 illuminant. Measurements were performed on two distinct regions: the crumb (analyzed on transversally sectioned surfaces) and the crust on three muffins randomly selected from the batches. Color values were expressed in the CIE L*a*b* system, and the total color difference (ΔE*) calculated between the batter and the final baked crumb and crust with Equation (2) [23]:

$${\mathsf{\Delta }\mathrm{E}}^{*}=\sqrt{({∆L^{*})}^2+({∆a^{*})}^2+({∆b^{*})}^2}$$

(2)

2.6. Texture Analysis

Texture Profile Analysis (TPA) was performed using a Texture Analyzer (TA-XT Plus, Stable Micro Systems, Godalming, UK) with a method adapted from [24]. For standardization, muffins were cut horizontally 2.5 cm from the base, and the upper half was discarded. Crumb samples were subjected to a double compression cycle (25% deformation of the initial height) using a 36 mm diameter flat-ended aluminum cylindrical probe at a test speed of 1 mm/s. A 5 s interval was maintained between cycles. Measurements were performed on three muffins randomly selected from each of the two independent production batches. Textural parameters including hardness, springiness, cohesiveness, chewiness, and resilience were calculated.

2.7. Image Processing

Image analysis was carried out to assess the crumb structure, following the established methodology described previously [25]. Muffins were horizontally cut at 2.5 cm from the base, and crumb images were captured using a Canon e510 scanner (Canon Inc., Tokyo, Japan) (600 dpi). Image processing was conducted with ImageJ v.154g software (National Institutes of Health, Bethesda, MD, USA) to measure total air cell area (mm²), average air cell size (mm), air cell (%), and circularity. Analyses were performed on three muffins randomly selected from each of the two independent production batches.

2.8. Sensory Analysis

The sensory characteristics were evaluated by ten trained adult panelists (6 females, 4 males, aged 25–40 years) from the Department of Food Engineering using quantitative descriptive analysis (QDA) following the previously published method [26], with modifications. During training, the panel developed consensus-based descriptors and definitions (Supplementary Table S1). Samples, coded with three-digit numbers, were presented randomly and evaluated in duplicate within 24 h of production. Panelists scored appearance, texture, mouthfeel, odor, and taste attributes on a 10 cm unstructured line scale, with palate cleansing between samples. Ratings were converted to numerical values for statistical analysis.

2.9. Statistical Analysis

All experimental data, including sensory scores, were analyzed using one-way analysis of variance (ANOVA) followed by Duncan’s multiple range test for post hoc comparisons (p < 0.05). Statistical analyses were performed using SPSS software (version 23.0, IBM, Armonk, NY, USA).

Multivariate analyses, including Principal Component Analysis (PCA) and heatmap generation, were conducted using ClustVis (custom BETA edition) [27]. Prior to these analyses, the sensory data were row-centered, scaled to unit variance, and constant variables were excluded. PCA was performed using singular value decomposition (SVD) with imputation. Heatmaps were generated using correlation distance metric and average linkage, and dendrograms were ordered using the “tightest cluster first” method without imposing a predefined number of clusters.

All analyses were conducted in at least triplicate, and data are presented as the mean ± standard deviation (SD).

3. Results and Discussion

3.1. Proximate Composition, Total Phenolic Content and Antioxidant Activity

Proximate composition of muffin samples is presented in

Table 1. Proximate composition (%), TPC (mg GAE/g), and AA (DPPH scavenging %) of muffin formulations prepared with rice (P), taro (T), and carob (H) flours containing 0%, 4%, or 8% Cornelian cherry pulp.

Sample	Moisture	Fat	Protein	Ash	Fiber	CHO	TPC	AA
P0	23.28 ± 0.63 bc	14.79 ± 0.27 abcde	5.77 ± 0.14 d	1.35 ± 0.04 a	1.19 ± 0.10 a	53.63 ± 1.1 ef	0.22 ± 0.01 a	47.71 ± 0.29 a
P4	23.30 ± 0.09 bc	14.53 ± 0.14 abc	5.51 ± 0.13 bc	1.36 ± 0.01 a	1.14 ± 0.14 a	54.16 ± 0.05 f	0.25 ± 0.01 a	54.25 ± 0.59 b
R8	23.56 ± 0.62 c	14.38 ± 0.2 a	5.59 ± 0.06 cd	1.39 ± 0.09 a	1.11 ± 0.15 a	53.98 ± 0.3 f	0.25 ± 0.02 a	59.38 ± 0.18 d
PT0	21.66 ± 0.41 a	15.04 ± 0.12 def	5.49 ± 0.03 bc	1.88 ± 0.03 b	3.23 ± 0.02 b	52.70 ± 0.61 ef	0.60 ± 0.01 b	56.58 ± 0.62 c
PT4	22.23 ± 0.26 ab	14.57 ± 0.1 2abcd	5.34 ± 0.07 b	1.83 ± 0.01 b	3.17 ± 0.03 b	52.86 ± 0.03 ef	0.61 ± 0.02 b	59.05 ± 0.23 d
PT8	23.13 ± 0.44 bc	14.49 ± 0.38 ab	5.31 ± 0.16 b	1.91 ± 0.01 b	3.10 ± 0.05 b	52.06 ± 0.07 e	0.76 ± 0.06 c	60.47 ± 0.21 e
PH0	21.75 ± 0.38 a	15.41 ± 0.13 f	5.65 ± 0.02 cd	2.06 ± 0.05 c	7.69 ± 0.05 c	47.43 ± 0.16 d	1.63 ± 0.04 d	66.57 ± 0.46 f
PH4	22.43 ± 0.86 abc	15.20 ± 0.25 ef	5.67 ± 0.11 cd	2.19 ± 0.07 cd	7.79 ± 0.02 c	46.72 ± 0.81 cd	1.66 ± 0.08 d	69.47 ± 0.75 g
PH8	23.20 ± 0.88 bc	15.15 ± 0.11 ef	5.59 ± 0.15 cd	2.24 ± 0.06 d	7.81 ± 0.03 c	46.01 ± 1.22 bcd	2.09 ± 0.03 f	71.14 ± 0.4 h
PHT0	22.21 ± 0.63 ab	15.24 ± 0.11 def	4.92 ± 0.09 a	2.62 ± 0.1 e	9.37 ± 0.12 d	45.65 ± 1.05 abc	2.04 ± 0.03 f	69.46 ± 0.38 g
PHT4	23.24 ± 0.26 bc	14.93 ± 0.13 bcdef	4.80 ± 0.08 a	2.76 ± 0.12 f	9.44 ± 0.1 d	44.82 ± 0.7 ab	1.88 ± 0.03 e	72.16 ± 0.56 ı
PHT8	23.64 ± 0.07 c	15.00 ± 0.33 cdef	4.82 ± 0.07 a	2.80 ± 0.07 f	9.38 ± 0.1 d	44.36 ± 0.64 a	2.24 ± 0.21 g	70.35 ± 0.15 gh

Different superscripts within a column indicate significant differences (Duncan’s test, p < 0.05).

. The moisture content of muffins slightly increased with fruit enrichment, likely due to the natural water content of the fruit pulp. The increase in fiber content upon partial substitution of rice flour with taro or carob flour (p < 0.05) aligns with previous reports of their high dietary fiber content [9,10,11]. This observed rise in dietary fiber with increasing levels of flour substitution is a well-established phenomenon in muffin fortification, consistently reported in studies utilizing diverse fiber-rich sources such as carob powder [28], grape pomace [29], and apple pomace [30]. The taro-carob blended formulations (PHT0–PHT8) exhibited the highest fiber content, representing a 7.9-fold increase over the rice-only control (P0: 1.19%). Carob-enriched samples showed a 6.5–6.6-fold increase, while taro blends demonstrated a 2.6–2.7-fold increase, all being statistically significant (p < 0.05). Dietary fiber in enriched formulations functionally modifies batter systems, demonstrating both rheological and metabolic effects. Specifically, fiber enhances batter viscosity and structural integrity while also helping to control blood sugar levels by slowing down the digestion and absorption of carbohydrates [9,11]. This strengthening effect on the product matrix is a common and desirable outcome of fiber addition. For instance, the incorporation of various dietary fibers, such as those from safflower or orange, has been consistently shown to enhance the viscoelastic properties of muffin batters, leading to final products with improved hardness and chewiness [31,32]. This supports the role of dietary fibers in promoting structural development in gluten-free baked goods. Protein content showed statistically significant variations (p < 0.05), while fat levels remained stable. Ash content, a proxy for minerals, was the lowest in the rice-only control (P0) but increased significantly (p< 0.05) with taro (PT0) and carob (PH0) substitution, suggesting higher mineral contributions (e.g., calcium, iron, potassium) [33]. These findings highlight the potential of taro and carob flours as valuable ingredients for significantly enhancing the dietary fiber and mineral profile of gluten-free muffins, thereby contributing to their nutritional value.

Partial substitution with taro (PT) and carob (PH) flours significantly improved the total phenolic content (TPC) and antioxidant activity. Carob blends showed the highest TPC due to their rich phenolic content, including gallic acid and tannins [10,33], while taro exhibited an intermediate increase (26.7%). The carob-taro combination (PHT) further enhanced bioactivity in a dose-dependent manner with Cornelian cherry pulp: 4% pulp yielded maximum antioxidant capacity, whereas 8% pulp resulted in the highest TPC in PHT8. The potential of Cornelian cherry enrichment for enhancing antioxidant activity in bakery products is further supported by a recent study on oily cakes made from peanut flour, where the addition of Cornelian cherry extract also led to a significant enhancement of the product’s bioactive profile [34]. The antioxidant capacity of gluten-free muffins is influenced by two key factors: (1) the composition of flour blends and (2) the presence of heat-stable bioactive compounds. Notably, Cornelian cherry anthocyanins (e.g., cyanidin-3-galactoside) contribute significantly due to their remarkable thermostability, maintaining redox activity even after high-temperature baking [35]. These pigments act as potent electron donors, neutralizing free radicals. Additionally, the Maillard reaction during baking generates melanoidins (brown polymeric compounds) with demonstrated radical-scavenging properties [36]. These melanoidins complement the fruit-derived antioxidants, resulting in a multi-mechanistic antioxidant system that enhances the muffins’ functional profile.

3.2. Color Analysis

As a primary visual cue for consumers, muffin color serves as a critical quality attribute. The specific color parameters (L *, a *, b *) for all formulations, which quantify these visual differences, are presented in detail in

Table 2. Color parameters of muffin formulations prepared with rice (P), taro (T), and carob (H) flours containing 0%, 4%, or 8% Cornelian cherry pulp.

	Crumb				Crust
Sample	L*	a*	b*	ΔE*	L*	A*	B*	ΔE*
P0	68.12 ± 0.26 h	2.60 ± 0.04 a	23.18 ± 0.37 g	4.34 ± 0.02 d	53.55 ± 0.28 f	21.57 ± 0.15 g	33.04 ± 0.17 i	5.75 ± 0.03 e
P4	66.41 ± 0.17 g	4.67 ± 0.13 b	17.37 ± 0.29 b	4.34 ± 0.05 d	53.29 ± 0.30 f	20.44 ± 0.36 f	31.51 ± 0.16 h	5.97 ± 0.03 f
P8	64.62 ± 0.40 f	5.83 ± 0.12 c	15.79 ± 0.18 a	4.59 ± 0.05 e	49.06 ± 0.72 e	19.89 ± 0.21 de	29.60 ± 0.26 g	5.77 ± 0.08 e
PT0	36.59 ± 0.17 c	11.75 ± 0.23 e	18.28 ± 0.39 c	3.06 ± 0.18 a	37.35 ± 0.69 b	15.57 ± 0.17 b	19.48 ± 0.32 a	3.90 ± 0.11 a
PT4	36.36 ± 0.40 c	11.79 ± 0.33 e	18.61 ± 0.30 cd	3.37 ± 0.19 b	37.25 ± 0.77 b	15.56 ± 0.24 b	20.06 ± 0.14 b	4.18 ± 0.09 b
PT8	36.04 ± 0.39 c	11.93 ± 0.33 e	18.49 ± 0.35 cd	3.31 ± 0.12 b	37.32 ± 0.60 b	14.98 ± 0.23 a	19.75 ± 0.40 ab	4.07 ± 0.10 ab
PH0	61.23 ± 0.38 e	8.25 ± 0.35 d	20.24 ± 0.31 f	4.33 ± 0.06 d	48.97 ± 0.55 e	20.22 ± 0.08 ef	28.37 ± 0.06 e	5.16 ± 0.05 d
PH4	60.76 ± 0.18 e	8.70 ± 0.31 d	20.06 ± 0.35 f	4.54 ± 0.02 e	47.05 ± 0.62 d	19.49 ± 0.16 d	27.27 ± 0.20 e	4.99 ± 0.10 cd
PH8	59.51 ± 0.47 d	8.23 ± 0.43 d	19.76 ± 0.32 f	4.48 ± 0.09 de	45.30 ± 0.67 c	18.70 ± 0.39 c	27.19 ± 0.01 f	4.87 ± 0.11 c
PHT0	34.72 ± 0.43 b	13.05 ± 0.11 f	18.96 ± 0.15 de	3.80 ± 0.09 c	33.11 ± 0.70 a	15.00 ± 0.29 a	22.06 ± 0.23 d	4.23 ± 0.18 b
PHT4	34.38 ± 0.26 ab	13.08 ± 0.32 f	18.99 ± 0.16 de	3.81 ± 0.11 c	33.44 ± 0.34 a	14.96 ± 0.26 a	21.32 ± 0.17 c	4.22 ± 0.08 b
PHT8	34.11 ± 0.37 a	13.08 ± 0.21 f	19.14 ± 0.06 e	3.81 ± 0.03 c	32.60 ± 0.60 a	15.20 ± 0.23 ab	21.15 ± 0.41 c	4.14 ± 0.16 b

Different superscripts within a column indicate significant differences (Duncan’s test, p < 0.05).

. Visual examples of the crust and crumb structure for representative formulations are provided in Supplementary Figure S1.

The crust exhibited significantly lower L* values (indicating a darker color) than the crumb, attributable to the Maillard reactions and caramelization [28]. Among crumb samples, the control (P0) sample displayed the highest lightness, while carob-enriched formulations (PH, PHT) showed the most pronounced darkening. Taro flour blends (PT) resulted in moderate darkening (L*: 59.51–61.23), whereas carob’s natural brown pigments caused more intense coloration. The incorporation of both carob and taro flours significantly enhanced the crumb’s chromaticity, leading to higher a* (redness) and b* (yellowness) values compared to the control samples made with rice flour. As expected, the crust was markedly darker than the crumb in all formulations, a phenomenon clearly demonstrated by the control’s crust color values. The addition of carob flour intensified this effect, producing an even darker crust (L ≈ 37) and shifting its color profile towards less pronounced red and yellow hues (i.e., reduced a* and b* values). Carob-substituted samples developed a characteristically darker and more uniform crumb color. This visually dense hue is not only a marker of potential nutrient density but also aligns with consumer perceptions of a richer, more wholesome product [37]. Fruit pulp enrichment darkened crumbs (decreasing L*), increased redness (a*), and reduced yellowness (b*), particularly in rice-only samples due to the anthocyanin-rich, reddish-pink pigmentation of Cornelian cherry (p < 0.05) [35]. However, in carob/taro blends, fruit pulp had only a secondary effect, as the flours’ dominant pigments masked its chromatic influence. The high ΔE* values between the batter and the final muffins confirmed significant color development during baking. This transformation was driven by the intrinsic pigments of the flours and their involvement in Maillard browning and caramelization reactions [28]. Furthermore, the thermal gradient within the muffin led to a clear divergence between the crust and crumb, which was quantitatively reflected in their ΔE* values. The crust exhibited higher ΔE* values, evidencing more intense browning reactions due to direct heat exposure. In contrast, the crumb showed lower ΔE* values, consistent with its shielded position and consequently milder color transformation.

3.3. Texture, Baking Loss and Height Results

In the absence of gluten, the starch gelatinization and gas retention dynamics that are critical to structure formation are fundamentally altered. Therefore, a rigorous assessment of physical parameters (such as baking loss, volume, and texture profile) is essential in the development of gluten-free muffins. Texture profile analysis (TPA) was performed to determine the muffins’ hardness, cohesiveness, chewiness, springiness, and resilience (

Table 3. Texture analysis, height and baking loss results of muffin formulations prepared with rice (P), taro (T), and carob (H) flours containing 0%, 4%, or 8% Cornelian cherry pulp.

Sample	Hardness	Springiness	Cohesiveness	Chewiness	Resilience	Height (mm)	Baking Lossc (%)
P0	1.83 ± 0.34 g	0.95 ± 0.06 ab	0.68 ± 0.02 a	1.19 ± 0.84 cd	0.44 ± 0.04 bc	40.4 ± 0.42 bc	18.04 ± 0.52 e
P4	1.95 ± 0.07 h	1.00 ± 0.00 ab	0.77 ± 0.02 bc	1.50 ± 0.52 ef	0.48 ± 0.01 bcd	39.6 ± 0.65 a	18.87 ± 0.45 f
P8	2.09 ± 0.36 i	1.07 ± 0.11 ab	0.83 ± 0.03 cd	1.87 ± 2.85 gh	0.52 ± 0.02 cd	39.9 ± 0.65 ab	18.64 ± 0.50 f
PT0	1.67 ± 0.07 f	0.99 ± 0.01 ab	0.80 ± 0.07 bcd	1.34 ± 1.12 de	0.34 ± 0.05 a	42.0 ± 0.61 de	15.33 ± 0.37 c
PT4	1.71 ± 0.27 f	1.09 ± 0.08 b	0.86 ± 0.04 cd	1.60 ± 2.23 efg	0.40 ± 0.03 ab	41.7 ± 0.27 d	16.84 ± 0.41 d
PT8	1.80 ± 0.23 g	1.17 ± 0.02 c	0.88 ± 0.01 d	2.48 ± 0.26 i	0.42 ± 0.02 ab	41.6 ± 0.55 d	16.48 ± 0.47 d
PH0	0.76 ± 0.11 a	1.66 ± 0.14 c	0.82 ± 0.04 cd	1.03 ± 1.60 bc	0.46 ± 0.04 bcd	43.1 ± 0.22 f	14.19 ± 0.31 a
PH4	0.83 ± 0.18 b	2.46 ± 0.08 d	0.83 ± 0.03 cd	1.70 ± 0.40 fg	0.48 ± 0.02 bcd	42.6 ± 0.42 ef	14.76 ± 0.44 b
PH8	0.90 ± 0.11 c	2.59 ± 0.11 d	0.87 ± 0.02 d	2.03 ± 1.54 C	0.53 ± 0.02 d	42.7 ± 0.45 f	15.05 ± 0.36 bc
PHT0	0.87 ± 0.25 bc	0.90 ± 0.01 a	0.73 ± 0.05 a	0.57 ± 0.11 a	0.43 ± 0.07 b	41.9 ± 0.65 d	16.83 ± 0.42 d
PHT4	1.01 ± 0.23 d	0.99 ± 0.00 ab	0.82 ± 0.04 cd	0.82 ± 0.60 ab	0.47 ± 0.02 bcd	40.7 ± 0.45 c	15.51 ± 0.41 c
PHT8	1.19 ± 0.14 e	1.01 ± 0.03 ab	0.85 ± 0.01 cd	1.02 ± 0.06 bc	0.47 ± 0.02 bcd	40.9 ± 0.42 c	15.48 ± 0.37 c

Different superscripts within a column indicate significant differences (Duncan’s test, p < 0.05).

).

The rice-only control samples (P) exhibited the highest hardness values, ranging from 1.83 to 2.09 N. In contrast, muffins containing carob flour (PH) showed markedly softer textures, with hardness values between 0.76 and 0.9 N. This represents a substantial 51–64% reduction in hardness compared to the rice-based controls. The softening effect of carob flour aligns with previous findings indicating that partial substitution (≤30%) with carob flour can yield texture profiles similar to those of wheat-based products [10]. Carob substitution also enhanced springiness and resilience (PH8: 0.53), whereas taro blends (PT) exhibited the highest chewiness (PT8: 2.48). Both carob and taro flours increased cohesiveness relative to rice-only control samples. The observed textural differences stem from the distinct functional properties of each flour: Rice flour (P) forms firm structures through starch retrogradation while lacking sufficient fiber, producing harder textures [38]. Taro flour (T) provided higher fiber content and amylopectin-rich starch that enhances water retention and produces softer products [39]. Carob flour (H) contributes optimal texture through its water-binding galactomannans [11], which promote elastic, aerated structures with improved moisture retention and desirable chewiness. Increasing pulp concentration (4–8%) led to increased hardness, particularly in rice-only (P0-P8) formulations, correlating with reduced muffin volume and height, consistent with previous reports [40]. Baking loss further influenced texture, as rice samples (P0-P8) demonstrated both the highest moisture loss and highest hardness values. In contrast, carob-enriched samples (PH0-PH8) showed lower baking loss and softer textures. Among all formulations, PH0 exhibited the most favorable textural characteristics, with maximum height (43.1 mm) and minimal hardness, indicating effective gas retention and a soft texture. Conversely, P8 showed the least desirable attributes in terms of both baking height and hardness. Fruit pulp incorporation significantly increased chewiness (p < 0.05), particularly in taro blends (an 84.4% increase from 1.34 to 2.48 N), likely through modified batter viscosity and stabilized air cell structure. Complementary image analysis of crumb microstructure further elucidates these textural formation mechanisms.

3.4. Image Processing Results

In gluten-free baking, the absence of gluten affects bubble stability and crumb structure, weakening the batter’s gas retention capacity. Image analysis provides a quantitative assessment of microstructural changes including pore formation and texture modifications induced by ingredient substitutions (e.g., taro or carob flour) and fruit pulp enrichment (Figure 1). Quantitative structural parameters from image analysis demonstrated that the air cell area (%) ranged from 16.4% (P8) to 50.9% (PH0), directly reflecting each formulation’s gas retention capacity (

Table 4. Microstructural parameters of muffin formulations prepared with rice (P), taro (T), and carob (H) flours containing 0%, 4%, or 8% Cornelian cherry pulp.

Sample	Total Air Cell Area	Average Air Cell Size	Air Cell Area (%)	Circularity
P0	38.98 ± 0.136 b	0.18 ± 0.001 b	24.17 ± 0.085 b	0.74 ± 0.002 a
P4	36.47 ± 0.107 b	0.23 ± 0.001 c	22.61 ± 0.066 b	0.72 ± 0.001 a
P8	26.50 ± 0.141 a	0.25 ± 0.003 c	16.43 ± 0.087 a	0.76 ± 0.001 b
PT0	48.39 ± 0.612 d	0.22 ± 0.002 bc	30.00 ± 0.380 d	0.81 ± 0.010 d
PT4	44.52 ± 1.240 c	0.18 ± 0.006 b	27.60 ± 0.769 c	0.79 ± 0.001 cd
PT8	43.61 ± 1.484 c	0.31 ± 0.045 d	27.04 ± 0.920 c	0.78 ± 0.011 bc
PH0	82.10 ± 0.375 g	0.12 ± 0.025 a	50.90 ± 0.232 g	0.87 ± 0.012 e
PH4	81.51 ± 1.790 g	0.18 ± 0.014 b	50.54 ± 1.110 g	0.87 ± 0.016 e
PH8	77.88 ± 0.675 f	0.18 ± 0.019 b	48.28 ± 0.418 f	0.86 ± 0.010 e
PHT0	82.56 ± 3.258 g	0.18 ± 0.033 b	50.78 ± 1.441 g	0.86 ± 0.013 e
PHT4	78.42 ± 1.139 f	0.11 ± 0.021 a	48.62 ± 0.706 f	0.87 ± 0.011 e
PHT8	72.87 ± 1.662 e	0.12 ± 0.001 a	45.18 ± 1.031 e	0.85 ± 0.001 e

^a–g Different superscripts within a column indicate significant differences (Duncan’s test, p < 0.05).

). Carob flour samples (PH0) exhibited superior total air cell area (cellularity), while rice flour with 8% pulp (P8) showed poor gas retention. These findings were consistent with texture and height measurements. PH0’s high cellularity corresponded to its minimal hardness (0.8 N) and maximum height (43.1 mm), whereas P8’s reduced gas retention (26.50 mm²) aligned with its elevated hardness (2.09 N) and limited rise (39.9 mm). Pore circularity (0–1 scale, where 1 = perfect circle), an indicator of structural integrity, was the highest in carob (H) formulations compared to the rice-only control (P0). Taro flour blends (PT8) demonstrated excessive pore expansion (0.31 mm), suggesting structural instability.

Cornelian cherry pulp generally reduced the air cell area in rice-based muffins due to the higher water content and acidity of the pulp [18], weakening gas retention. Increased pulp content also diluted structural components and impaired gas production, leading to uneven pore formation. However, carob-rich formulations showed less reduction in porosity because their high soluble fiber (pectin/and gums) improved viscosity and water retention, preserving structure [41]. Pulp incorporation affected pore structure differently depending on the formulation (p < 0.05). Rice flour (P) had the largest changes; porosity dropped from 24.2% to 16.4%, while pore size increased from 0.18 mm to 0.25 mm. In contrast, carob flour (H) maintained high porosity and stable circularity. These results confirm that carob blends (PH/and PHT) consistently exhibited superior microstructure (>48% air cell area, >0.86 circularity), whereas rice-taro combinations displayed pore instability.

3.5. Sensory Analysis Results

Sensory evaluation of gluten-free muffins provided comprehensive insights into product quality and consumer acceptance. Quantitative descriptive analysis revealed that carob-substituted formulations (PH and PHT) showed superior structural and sensory properties compared to rice-only samples (P). Carob-based muffins achieved the highest crumb porosity scores (Figure 2), consistent with their cellular structure observed using ImageJ. In contrast, porosity declined as pulp concentration increased in rice muffins. Carob formulations had the most intense crust and crumb color, a feature positively linked to perceived healthfulness [42]. Textural analysis indicated that rice muffins became progressively harder with pulp addition (p < 0.05), while carob-based samples were the softest, demonstrating carob flour’s softening effect in gluten-free matrices [43]. Springiness was markedly higher in PH samples, likely due to carob’s unique polysaccharide matrix, and moistness scores were also superior, reflecting the water-binding capacity of galactomannan-rich fibers [11]. Conversely, oiliness was more pronounced in rice-only samples, likely due to their lower fiber content and reduced water-holding capacity, which can increase the perception of surface oil. The addition of fruit pulp significantly enhanced the aromatic odor and taste intensity of the muffins (p < 0.05), an effect especially evident in rice-only and taro-substituted formulations. Among the samples, PH8 stood out with high scores for several attributes, particularly porosity, crust color, and springiness. In contrast, PHT8 achieved the highest score for aromatic odor. Both samples also demonstrated a desirable and long-lasting aftertaste, likely due to the slow release of carob-derived flavor compounds [44].

Multivariate analyses effectively supported the study’s findings. Principal Component Analysis (PCA) revealed two primary sensory dimensions, which collectively explained 83% of the total variance (PC1 = 69%, PC2 = 14%). The first principal component (PC1) primarily represented attributes associated with texture and appearance. PC1 showed a strong positive correlation with Hardness (0.31) and Oiliness (0.25), while it was negatively correlated with appearance-related attributes, including Crumb Porosity (−0.27), Crust Color (−0.33), and Crumb Color (−0.29). The second principal component (PC2) captured the mouthfeel and taste dimension. This component was strongly and positively correlated with mouthfeel attributes like chewiness (0.58) and oiliness (0.27). As seen in the PCA bi-plot (Figure 3), the samples clustered distinctly. This plot highlights the relationships between the samples and their corresponding sensory attributes, which are shown as points (samples) and vectors (attributes), respectively. The rice-only muffins (P0, P4, and P8) clustered with high scores for hardness and oiliness. In contrast, muffins with carob and taro substitutions separated from the control, aligning with desirable sensory vectors such as moistness, springiness, porosity, and aromatic odor. Notably, PH8 and PHT8 were positioned close to these desirable vectors, consistent with their high sensory scores for porosity, springiness, and aromatic odor. These results indicate that taro and carob substitution successfully enhanced the sensory profile of the gluten-free muffins.

Heatmap visualization further highlighted the relationships among the sensory attributes and samples (Figure 4). Red cells denoted scores above the mean and blue cells indicated those below, irrespective of whether the deviation represented a desirable attribute (e.g., high moistness) or an undesirable attribute (e.g., high hardness).

For instance, the P samples displayed red for hardness (+1.92) and blue for moistness (−1.57), reflecting an undesirable firm and dry profile. In contrast, PH4 and PH8 showed red shifts for moistness (+0.93 to +1.05) and aroma (+1.14 to +1.34), while maintaining lower hardness, resulting in a softer, moister, and more aromatic profile. Clustering was based on correlation distance and average linkage. The accompanying dendrogram grouped samples primarily by formulation, with rice-only (P), carob-based (PH), taro-based (PT), and combined (PHT) muffins forming distinct branches. At the attribute level, two main clusters were identified. Hardness and oiliness were grouped together on one branch, while moistness, aroma, aftertaste, color, porosity, and springiness clustered separately. This clustering indicates that firmness and oiliness were perceived together, while moistness, aroma, and appearance attributes defined a distinct profile across muffin formulations. Accordingly, rice-only samples tended to align with the hardness–oiliness cluster, reflecting their firmer and drier sensory profile, whereas carob- and taro-based formulations were more closely associated with the moistness–aroma cluster, consistent with their higher sensory acceptance. Overall, sensory analysis results demonstrated that carob and taro flours, particularly in combination with Cornelian cherry pulp at 4–8%, produced muffins with superior texture, enhanced flavor, and a more balanced sensory profile, while rice-only formulations were associated with less favorable sensory traits. Among the tested formulations, PH8 represented the most balanced profile across multiple attributes (porosity, moistness, springiness, and aroma), whereas PHT8 stood out specifically for its intense aromatic odor and taste. These insights support the development of nutritionally enriched gluten-free bakery products that better align with consumer expectations in terms of flavor, texture, and overall quality.

4. Conclusions

This study successfully achieved its aim of developing gluten-free muffins by employing a novel strategy based on a complementary flour blend and bioactive enrichment from Cornelian cherry. The approach effectively targeted the key challenges of improving technological functionality, nutritional quality, and sensory appeal. It demonstrated that combining a carob–taro–rice flour blend with bioactive-rich Cornelian cherry pulp can effectively overcome the major challenges in gluten-free baking: limited technological functionality and poor nutritional and quality. The formulations provided a comprehensive enhancement of the nutritional profile, as evidenced by markedly improved dietary fiber content (up to 7.9-fold), total phenolic content (2.24 mg GAE/g), and antioxidant activity (up to 72.16%). This nutritional enrichment was further supported by the significant increase in ash content, indicating a richer mineral composition. Carob flour substitution significantly reduced hardness by 51–64% and promoted a finer, more uniform microstructure with larger air cell area (>48%) and higher circularity (>0.86), contributing to greater springiness and better sensory perception. Quantitative Descriptive Analysis confirmed improvements in crumb porosity and softness, while specific formulations (PH8, PHT8) were characterized by balanced or more pronounced sensory attributes.

Overall, the findings present a practical formulation strategy that combines nutritional enrichment with desirable texture and sensory quality in gluten-free muffins, offering direct relevance to functional bakery product development. The formulation, based on commercially available flours and a shelf-stable fruit pulp, demonstrates strong potential for industrial scaling. While this study conclusively demonstrates the superiority of the proposed ingredient strategies compared with a standard rice-flour base, future work should include a consumer acceptance study to benchmark these formulations against leading commercial products to further validate their market potential. Further research should examine shelf-life performance, including staling kinetics and microbial stability, to directly validate and support commercial application.