Metabolomic profile of barnyard millet (Echinochloa Esculenta) and its application in white carrot-based dairy product (kheer) preparation

Abstract

This study was conducted to develop a functional dairy product (Kheer) using barnyard millet (BM) (Echinochloa esculenta) and white carrot (Daucus carota) powder (WCP). The metabolic profile of BM and WCP was investigated for bioactive compounds. The results indicated that BM contained various biomolecules, including trans-3-indoleacrylic acid,5-aminovaleric acid, amino acids, etc. WCP contained acetylgenistein, glycitein, and genistein as the major compounds. A central composite rotatable design (CCRD) was applied to optimize the kheer formulation using BM (25–30%), brown sugar (20–25%), and WCP (5–10%) as evaluated by the sensorial, textural, and antioxidant properties of kheer. Based on the sensory reports (Overall acceptance, 8.8 ± 0.40), the optimal kheer ingredient composition was identified as 25 g of BM, 25 g of brown sugar, and 10 g of WCP. The optimized product achieved overall sensorial acceptance. The addition of WCP enhances the nutritional, textural, and antioxidant properties of kheer.

Highlights

• •LC-MS profiling identifies a broad spectrum of bioactive metabolites in millet.
• •The optimized kheer ingredients are 25 g BM, 25 g brown sugar, and 10 g WCP.
• •Developed kheer exhibits higher nutritional and antioxidant activity.
• •Sensory attributes remain acceptable despite ingredient substitution.
• •Barnyard millet and white carrot powder enrich the nutritional quality of kheer.

1. Introduction

The growing prevalence of lifestyle-related diseases such as obesity, diabetes, cardiovascular disorders, and metabolic syndromes has intensified global interest in nutritionally enhanced functional foods (Deokar et al., 2025; Sivaramakrishnan et al., 2025). These foods, in addition to providing basic nutrition, are formulated to offer additional health benefits that may reduce disease risk or promote optimal physiological functioning (Koumpouli et al., 2025). Functional foods enriched with bioactive compounds—such as dietary fiber, polyphenols, probiotics, flavonoids, and essential micronutrients—play a pivotal role in modulating metabolic processes, enhancing immune responses, and promoting gut health (Martirosyan, 2025). As scientific research continues to validate the therapeutic roles of specific nutrients and phytochemicals, functional foods are emerging as critical components of personalized nutrition and public health interventions aimed at curbing the global burden of chronic diseases (Ma et al., 2025). Functional foods and traditional dairy products intersect significantly in modern nutrition, offering both cultural value and health benefits. Dairy products like milk, yogurt, and kheer are deeply rooted in traditional diets and are naturally rich in nutrients such as calcium, protein, and B vitamins. When fortified or modified—by incorporating bioactive compounds, prebiotics, or low-glycemic ingredients—they can be transformed into functional foods that target specific health outcomes (Wazzan, 2024).

Kheer, an Indian dairy dessert, has primary ingredients that include milk, rice/wheat, sugar, dried fruits, and various flavourings like saffron and cardamom to enhance its taste (Bhosale et al., 2021). Kheer is a semi-solid dessert made from cereals, typically prepared by cooking rice in milk with sugar or jaggery until the rice starch fully gelatinizes. While milk and sugar are essential to kheer, the traditional recipe can be adapted by using substitutes like vermicelli or semolina instead of rice. However, this traditional dessert contains high levels of refined carbohydrates and sugars. Nevertheless, coarse grains like pearl millet and other millets are seldom incorporated into these value-added products (Jha et al., 2013). Adding a low GI ingredient such as barnyard millet (BM) and white carrot can find favor with consumers who are health-conscious, diabetic, or obese. White carrot is a dietetic, high-fiber, low-calorie vegetable with valuable compounds such as carotenoids and polyphenols (Da Silva Dias, 2014; Varshney & Mishra, 2022).

BM is recognized for its rich nutritional profile, including high levels of dietary fiber, proteins, essential amino acids, and micronutrients such as iron, zinc, calcium, and magnesium (Renganathan et al., 2020). Nutritionally, BM offers superior value compared to major cereals, with carbohydrate content typically ranging from 51.5 to 62.0 g per 100 g, which is lower than that of other millets, and its crude fiber content, ranging from 8.1 to 16.3%, surpasses that of any other cereal (Renganathan et al., 2020). This high fiber-to-carbohydrate ratio supports the gradual release of sugars into the bloodstream, aiding blood sugar regulation (Reynolds et al., 2024). Additionally, resistant starch in BM has been shown to reduce blood glucose, serum cholesterol, and triglyceride levels in animal studies (Kaur & Sharma, 2020; Varshney & Mishra, 2022). On the other hand, carrots were initially used for medicinal purposes before becoming popular as a food ingredient (Riaz et al., 2022). The chemical composition of carrots differs based on their color: carotenoids give orange and yellow shades, anthocyanins give the purple color, while white carrots lack any pigments (Yusuf, Tkacz, et al., 2021). Surles et al. (2004) investigated the consumer acceptance of five different colored carrots, revealing that orange and white varieties were preferred over the yellow, red, and purple varieties. White carrots contain procyanidins, a type of polyphenol recognized for its health benefits. Compounds such as 3-O-caffeoylquinic acid, 3-O-feruloylquinic acid, O-coumaroyl quinic acid, di-caffeoyl quinic acid derivatives, and di-ferulic acid derivatives have also been identified in white carrots (Yusuf, Wojdyło, et al., 2021). These compounds are known for their antioxidant and anti-inflammatory properties, etc., thereby contributing to improved metabolic and cardiovascular health.

Therefore, this research aims to replace traditional ingredients in kheer with low GI, high fiber, gluten-free, and phenolic compounds containing BM and WCP to develop a functional dairy product (Kheer). Before preparation, BM and WCP were investigated for the metabolomic profile. The metabolomics assessment of plant-based ingredients indicates the presence of bioactive compounds that could provide health benefits upon consuming functional foods (Chen et al., 2025; Wang et al., 2025). Additionally, the metabolic profile guides the functional product formulation to control or improve physical, sensory, and functional properties. Although there are few studies on the utilization of the BM for the development of kheer, these studies did not conduct optimization, nor did they investigate the combined effect of WCP to completely replace the traditional rice and sugar-based ingredients in the kheer. The novelty of the present study lies in the synergistic integration of BM and WCP within a thermal dairy matrix. Unlike traditional pigmented carrots, white carrots offer a unique profile of bioactive compounds (phenolic acids and dietary fibers) without altering the traditional visual appeal of dairy products. Optimization of dessert formulation was conducted using a CCRD methodology with BM ranging from 25 to 30%, brown sugar from 20 to 25%, and WCP varying from 5 to 10%. The developed formulations were evaluated for sensorial, textural, and functional properties of kheer. The research outcome of functional foods preparation using BM and WCP can encourage a shift away from refined grains and enhance the diversity of foods consumed. Also, a development of nutrient-dense, low-GI desserts, addressing the need for healthy, sustainable, and functional foods.

2. Materials and methods

2.1. Materials

Commercial-grade brown sugar (M-grade) (Natural Organic Khandsari Brand) procured from Jainava Foods, Indore, Madhya Pradesh, India. Pasteurized full cream milk of the Amul Gold brand, containing 6.0% fat and 9.0% solid-not-fat (SNF), was purchased from the local market in Varanasi, Uttar Pradesh, India. Dehusked BM was sourced from the local market in Mirzapur, while fresh, mature white carrots were obtained from local farmers in Mubarakpur, Azamgarh, Uttar Pradesh, India. The BM variety used (VL 172) was characterized by pale yellow-colored grains measuring approximately 1.9 mm in length. The white carrots ranged in length from 22.7 to 64.7 mm, with root diameters, considered a yield trait, between 17.2 and 41.3 mm. All chemicals used in this study were of analytical grade.

2.2. Preparation of barnyard millet powder and methanolic extract

The barnyard millet was thoroughly cleaned, dehulled, and ground into a fine powder using a Sieve Mesh No. 65. This powder was subsequently sealed in an airtight container for further experiments. Ten grams of barnyard millet powder were combined with 100 mL of 95% methanol and added with a zirconia bead to homogenize the mixture. The mixture was incubated at 30 °C for 48 h in a shaker incubator (100 rpm) (Kuhner, Germany). After incubation, the sample was centrifuged (REMI, India) for 30 min at 5000 rpm at a temperature of 4–6 °C. The resulting supernatants were collected and stored in glass bottles (Borosil, India) at refrigerated temperatures for further analysis.

2.3. Liquid chromatography– Mass spectrometry (LC-MS) analysis

The bioactive compounds in the barnyard millet methanolic extract were analyzed using LC-MS as described by Kumar et al. (2024) with slight modification. The extract was diluted with solvent to a final volume of 1500 μL. The sample was then vortexed at 2000 rpm for 2 min and centrifuged at 6000 rpm for 2 min. The supernatant was collected, filtered through a 0.22 μm syringe filter to remove hydrophobic compounds, and transferred into a vial. This vial was subsequently loaded into the auto-sampler for injection into the LC-MS system. The LC-MS system consisted of ultra-high-pressure liquid chromatography (Dionex UltiMate 3000) coupled with a tribrid high-resolution accurate mass spectrometer “Orbitrap Eclipse” (Thermo Fisher Scientific). The instrument setup included a Hypersil Gold C18 column (2.1 mm × 100 mm, 3.0 μm) at 40 °C, a 15 μL injection volume, a flow rate of 350 μL/min, and a run time of 55 min. The mobile phases used were solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile). A gradient mobile phase run from 2 to 98% of solvent B was used. The H-ESI at 3200 V was used for the ionization. The high-resolution MS was conducted in full scan, and MS2 scans at a scan range of 120–1000 m/z and a resolution of 60,000 m/z with HCD fragmentation. The data collection and peak annotation analysis were performed using Thermo Fisher Scientific's Compound Discoverer 3.3.2.3.1 software and ChemSpider database search.

2.4. Preparation of white carrot powder (WCP)

WCP was prepared following the methods outlined by Gupta and Rn (2017) and Kamel et al. (2023) with slight modifications. Fresh carrots were first thoroughly washed and thinly sliced. The slices were placed in muslin cloth and blanched in hot water at 95 °C for 2–3 min to inactivate enzymes. After blanching, they were cooled to room temperature, and surface moisture was removed by blotting. The blanched slices were then treated with a 0.25% potassium metabisulfite (KMS) solution for 20 min to extend shelf life. Subsequently, the slices were dehydrated in a tray dryer (Narang, India) at 55 °C for 12 h. The dried slices were ground into powder using a Bajaj mixer grinder (Mumbai, India), passed through a 0.5 mm sieve, and packed in low-density polyethylene (LDPE) pouches. The powder was stored at 15–20 °C under controlled conditions until further use (AOAC, 2000; Rai et al., 2011).

2.5. Electrospray ionization mass spectrometry (ESI-MS) analysis of WCP

A methanolic extract of white carrot extract was prepared according to a method adapted from Rai et al. (2011). One gram of WCP was weighed, and 10 mL of 70% methanol was added to the powder, and the mixture was stirred thoroughly using a magnetic stirrer for 5–10 min. Following extraction, the mixture was filtered through Whatman No. 1 filter paper to remove large particles, and the filtrate was centrifuged at 10,000 rpm for 15 min. The supernatant was then transferred to a rotary evaporator and concentrated at a temperature below 40 °C until the volume was reduced by half, with the final volume adjusted to 5 mL using the initial solvent. The concentrated extract was then passed through a 0.22 μm syringe filter to remove any remaining fine particles. Mass spectrometry (ESI-MS) analysis of the sample was conducted using a Micromass QTOF-2™ mass spectrometer (Waters, UK). The sample was ionized in positive electrospray ionization mode, with a needle voltage of 3 kV and a cone voltage of 30 V. The desolvation temperature was set at 350 °C with a nitrogen gas flow rate of 550 L/h, and a collision energy of 20 V was applied. Full scan mass spectrometry covered the m/z range of 50 to 1000, with a scan time of 0.5 s.

2.6. Preparation of BM and WCP-based kheer

Fresh full-cream milk (500 mL) was heated and concentrated to reduce its volume by 25%. BM (25–30 g) was prepared by removing extraneous materials and washing it 5–6 times with clean water. Once the milk was concentrated, the cleaned BM was added, and the mixture was heated at 65 ± 5 °C for 7 ± 1 min. WCP (5–10 g) was then added and further heated at 55 ± 5 °C for an additional 3 ± 1 min. Later on, brown sugar (20–25 g) was added, and cooking continued until the BM was tender. The kheer was then cooled to room temperature and stored under refrigeration for further analyses.

2.7. Formulation optimization using CCRD

CCRD was applied to optimize the process formulation variables using Stat-Ease Design Expert software (version 12). The independent variables considered in this experiment were the contents of brown sugar, BM, and WCP, with preliminary trials establishing baseline levels for each. The responses selected as output variables included color and appearance, body and texture, mouthfeel, flavor, and overall acceptability, allowing for the evaluation of the effects of the independent variables. The software generated a series of 20 experimental trials based on these responses (Table 1). Responses were randomly assessed after each experiment to determine the impact of the independent variables. A second-order polynomial equation was used to compare the predicted and observed values to the responses as follows:

$$Y={\beta }_0+\sum {\beta }_i{X}_i+\sum {\beta }_{\mathit{ij}}{X}_i{X}_j+\sum {\beta }_{\mathit{ii}}{X_i}^2+\text{€}$$

where, Y = Response; β₀ = Model constant; β_i = Linear effect of factor X_i; β_ij = Cross product of factor X_i and X_j; β_ii = Quadratic effect of factor X_i; € = Residual Error.

Table 1.

Response surface methodology of Kheer formulation using brown sugar, BM, and white carrot powder as independent variables in response to several sensory attributes.

Std	Run	Barnyard Millet (g)	Brown Sugar (g)	White Carrot (g)	Color And Appearance	Body And Texture	Flavor	Mouthfeel	Overall Acceptance
6	1	30	20	10	7.4	7.6	7	7.6	7.4
15	2	27.5	22.5	7.5	8.6	7.6	7.8	7.6	8
10	3	31.705	22.5	7.5	8.6	8.6	8	7.8	8
20	4	27.5	22.5	7.5	8.61	7.63	7.81	7.62	8.11
17	5	27.5	22.5	7.5	8.7	7.9	7.8	7.7	8.01
8	6	30	25	10	7.4	7.4	7.4	7.8	7.6
1	7	25	20	5	6.33	6.16	7.17	6.83	6.60
18	8	27.5	22.5	7.5	8.39	8	7.7	7.89	7.99
9	9	23.295	22.5	7.5	7	6.8	7.4	7	7.4
2	10	30	20	5	7	7.17	8	7.66	7.83
11	11	27.5	18.29	7.5	8.6	8	8	7.6	8
12	12	27.5	26.704	7.5	8.2	8	8.2	7.8	8.2
4	13	30	25	5	7.4	7.2	7.2	6.8	7
5	14	25	20	10	8.4	8.4	8.6	8.8	8.4
3	15	25	25	5	6.83	7.5	7	7.17	7.16
16	16	27.5	22.5	7.5	8.5	7.7	7.6	7.5	7.98
13	17	27.5	22.5	3.295	7.4	6.6	6.4	6.2	6
14	18	27.5	22.5	11.704	8.4	7.8	7.6	8.2	8.2
7	19	25	25	10	8.6	9	8.8	8.8	8.8
19	20	27.5	22.5	7.5	8.6	7.6	7.6	7.8	7.98

The adequacy of the developed model was evaluated using Analysis of Variance (ANOVA), in which P < 0.05 was considered significantly different. The high observed R² value in agreement with the predicted value confirms the model accuracy.

2.8. Physicochemical analysis

Total Soluble Solids (TSS) were measured using a refractometer (Bellingham and Stanley, UK) according to Association of official analytical chemists (1990). The pH was measured with a calibrated electronic pH meter (Thermo Scientific, Singapore). Moisture, fat, protein, crude fiber, carbohydrate, and ash content were determined using methods outlined in AOAC (2000). Antioxidant capacity was assessed using the DPPH inhibition assay as described by Baliyan et al. (2022), and Total Phenolic Content (TPC) was measured using the Folin-Ciocalteu method as per Blainski et al. (2013). Acidity was calculated according to the method in American Dairy Products Institute (2024), while reducing sugar content was analyzed using the Di-Nitro Salicylic Acid (DNS) method, and ascorbic acid content was determined by titration with DCPIP (Baliyan et al., 2022; Burhade et al., 2025).

2.9. Sensory characteristics of kheer

Before testing the samples, the panellists were instructed about the sensory characteristics to be assessed based on a 9-point hedonic rating scale. Sensory evaluation was conducted by 50 panellists (aged 20 to 35 years old) as active consumers of kheer. Since all panellists were regular consumers of kheer, no training was provided. Kheer samples were identified with random three-digit numbers and presented in a random sequence in order to eliminate any biases. Three-digit-number-marked paper cups were utilized to serve kheer samples. The panellists rinsed their palates with water before each sampling. Sensory evaluation included assessment of the taste, color, and appearance, body and texture, mouth feel, and overall acceptability. Duplicate assessments were performed, and mean values were used for the analysis.

Ethical permission was not needed to conduct the sensory evaluation of the kheer. However, panellists' consent has been obtained to participate in the sensory study and use the information they have provided on the sensory parameters of all kheer samples. Besides this, panellists' personal information, such as name, age, sex, and signature, was not disclosed at any part of the study or in the manuscript format to protect the participants' privacy and comply with legal standards.

2.10. Texture profile analysis (TPA) of kheer

TPA was performed on various textural parameters using a CT3 Texture Analyzer (AMETEK Brookfield, USA) with a TA36 probe (7 mm diameter stainless steel cylinder, 35 mm long) at a room temperature of approximately 25 ± 3 °C. The textural properties analyzed included adhesiveness, cohesiveness, chewiness, firmness, and gumminess.

2.11. Microbial analyses of the kheer

Microbial analyses comprising total plate count (TPC) and yeast and mold count were done following the method prescribed by AOAC (2000).

2.12. Statistical analysis

All analytical experiments were conducted in triplicate, with the results expressed as mean values ± standard deviation. A one-way ANOVA was performed on the data using Minitab 17 after the initial data processing in MS Excel. Statistical significance was set at P < 0.05. Model adequacy assessment involved regression coefficients (R²) and the lack-of-fit test.

3. Results and discussion

3.1. LC-MS profiling of barnyard millet

The LC-MS analysis of BM revealed a diverse concentration of metabolites. Untargeted metabolomics analysis using LC-MS identified 537 distinct compounds in BM. Among the annotated peaks in the mass spectrum, 44 compounds were categorized as major metabolites, including both primary and secondary types. Out of which, 36 putative bioactive metabolites are listed in Table 2. The base peak chromatogram of barnyard millet methanolic extract is presented in the supplementary material, Fig. 1S. The peak at m/z 391 was identified as bis-(2-ethylhexyl) phthalate, which could indicate environmental or laboratory contamination rather than a natural component. The peak at m/z 188 was identified as trans-3-indoleacrylic acid, a compound that has been reported to offer potential therapeutic benefits for conditions like inflammatory bowel disease (Wlodarska et al., 2017). A peak at m/z 134 was identified as L-Aspartic acid, which is essential for various physiological processes, including protein and nucleotide synthesis for cell growth, gluconeogenesis to maintain blood sugar levels, and ammonia detoxification through the urea cycle (Holeček, 2023). A peak at m/z 118 and m/z 104 was identified as 5-Aminovaleric acid and choline, respectively. A peak at m/z 123 was identified as nicotinamide, which is recognized as the amide form of niacin (vitamin B₃) (Fricker et al., 2018). Peak at m/z 268 was identified as adenosine, which has been reported to possess properties related to cholesterol homeostasis, cardio protection, anti-inflammatory, anticonvulsant, and neuroprotective effects, as well as anti-platelet and atheroprotective actions (Reiss et al., 2019; Tescarollo et al., 2020). Similarly, several other metabolites were detected in the methanolic extract of barnyard millet, as indicated in Table 2.

Table 2.

List of major chemical constituents identified from the methanolic extract of barnyard millet using LC-MS.

m/z	RT (minutes)	Phytochemical	M.W. (g/mol)	Structure	Molecular formula	Reference
391.28287	28.878	Bis-(2-ethylhexyl) phthalate	390.5561		C24 H38 O4	Javed et al. (2022); Satiyaningsih et al., (2024)
188.06993	2.001	Trans-3-Indoleacrylic acid	187.198		C11H9NO2	Wlodarska et al. (2017); Liu et al. (2023)
134.04439	0.684	L-Aspartic acid	133.0371		C4H7NO4	Holeček (2023)
118.08612	0.731	5-Aminovaleric acid	117.0788		C5H11NO2	Dhaher et al. (2014)
104.10704	0.698	Choline	104.173		C5H14NO	Goh et al. (2021)
123.05504	0.766	Nicotinamide	122.048		C6H6N2O	Fricker et al. (2018)
205.09647	2.001	D-(+)-Tryptophan	204.089		C11H12N2O2	Moghimani et al. (2024)
268.10297	0.761	Adenosine	267.0957		C10H13N5O4	Tescarollo et al. (2020); Reiss et al. (2019)
166.08571	1.288	L-Phenylalanine	165.0784		C9H11NO2	Akram et al. (2020); Pohle-Krauza et al., 2008); (Amin et al., 2021)
132.10149	0.821	L-Norleucine	131.0942		C6H13NO2	Ding et al., 2024
130.04951	0.764	L-Pyroglutamic acid	129.0422		C5H7NO3	Aiello et al., (2022)
116.07047	0.71	Proline	115.0632		C5H9NO2	Wu et al., (2011)
175.11839	0.668	L-(+)-Arginine	174.1111		C6H14N4O2	Wu et al., (2021)
148.05992	0.685	L-Glutamic acid	147.0526		C5H9NO4	Stamler et al. (2009)
282.27811	28.425	Oleamide	281.2708		C18H35NO	Reyes-Soto et al. (2022)
403.1373	22.888	Nobiletin	402.13		C21H22O8	Moazamiyanfar et al. (2023); Chen et al., (2023); Singh et al. (2021)
325.23401	27.73	2,3-dihydroxypropyl 12-methyltridecanoate	302.2448		C17H34O4	Yang et al. (2022)
373.12692	24.177	Tangeritin	372.1197		C20H20O7	Fatima and Siddique, (2019)
261.03601	0.764	Glucose 1-phosphate	260.0287		C6H13O9P	–
239.12697	24.616	1-(3-ethyl-2,4-dihydroxy-6-methoxyphenyl)butan-1-one	238.1197		C13H18O4	Aditya et al. (2022)
295.22577	23.178	(±)13-HpODE	312.229		C18H32O4	–
359.31454	28.987	1-Stearoylglycerol	358.3072		C21H42O4	Mondul et al. (2014)
152.05618	1.015	Guanine	151.0489		C5H5N5O	–
179.06966	27.84	4-Methoxycinnam-ic acid	178.0624		C10H10O3	Płowuszyńska and Gliszczyńska et al., (2021)
297.2027	26.795	Monolaurin	274.2135		C15H30O4	Nitbani et al. (2022); Subroto and Indiarto (2020)
)295.22574	26.837	9-Oxo-ODE	294.2184		C18H30O3	–
156.07628	0.677	L-Histidine	155.069		C6H9N3 O2	Holeček, (2020)
176.06999	7.803	Indole-3-acetic acid	175.0627		C10H9NO2	Shen et al. (2022)
192.06482	4.969	5-Hydroxyindole-3-acetic acid	191.0575		C10H9NO3	Li et al. (2020)
318.29935	24.706	2-Amino-1,3,4-octadecanetriol	317.2921		C18 H39NO3	–
326.30432	28.348	Oleoylethanolamide	325.297		C20H39NO2	De Filippo et al. (2023)
300.28873	26.843	Sphingosine (d18:1)	299.2815		C18H37NO2	–
282.11865	1.191	2’-O-Methyladenosine	281.1114		C11H15N5O4	–
328.31992	28.792	Stearoylethanolamide	327.3127		C20H41NO2	–
252.10822	0.981	2’-Deoxyadenosine	251.1009		C10H13N5O3	–
353.22885	24.877	(12Z)-9,10,11-trihydroxyoctadec-12-enoic acid	330.2396		C18H34O5	–

Fig. 1.

3D surface plot showing the interactive effect of Barnyard millet and sugar on color and appearance (A), body and texture (B), flavor (C), mouthfeel (D), and overall acceptance (E).

The metabolites mentioned above possess several bioactivities, including anti-depressant agent (Akram et al., 2020), insulin, glucagon, and gastric inhibitory polypeptide (Amin et al., 2020). Managing blood sugar levels and reducing lipid levels (Aiello et al., 2021) and improving insulin sensitivity (Ding et al., 2024). L-arginine strengthens the immune system, aids in fighting infections, and helps manage metabolic disorders such as diabetes, obesity, and hypertension (Wu et al., 2021). Compounds like 1-stearoyl glycerol and 2,3-dihydroxy propyl 12-methyl tri decanoate are involved in lipid metabolism by acting as substrates or intermediates in the synthesis and breakdown of lipids (Yang et al., 2022). Oleoylethanolamide is also crucial for regulating fat intake, energy balance, intestinal motility, and eating behaviour (De Filippo et al., 2023). The above-identified metabolites might play a significant role in determining the functional characteristics of the developed kheer. The presence of amino acids like L-arginine and L-glutamic acid is well understood for its physiological functions, including metabolic regulation and immune response. Moreover, indole derivatives and phenolic compounds have shown considerable antioxidant and anti-inflammatory effects in many studies. The inclusion of such components in the composition of the kheer can lead to an improvement in the quality and nutritional value of the kheer (Chandrasekara & Shahidi, 2011; Wu G, 2021).

3.2. ESI-MS/HRMS analysis of white carrot

The white carrot comprises acetylgenistin, glycitein, and genistein, members of the isoflavone class of phytoestrogens (Table 3). The ESI–MS/HRMS profile illustrating the bioactive compounds detected in white carrot is available in Supplementary Fig. 2S. WCP has minimal saturated fat and good content of fatty acids that help to improve endothelial activities and decrease the development of atherosclerosis (Ramachandran et al., 2020). Additionally, the organic compound dicyclohexylurea has demonstrated efficacy in reducing systemic blood pressure (Ghosh et al., 2008). Furthermore, eburicoic acid exhibits promising therapeutic potential for the management of type 2 diabetes and hyperlipidemia (Lin et al., 2017). Isolfavone-like metabolites such as genistein and glycitein are well known to exhibit antioxidant, anti-inflammatory, and cardioprotective functions. Incorporation of these in the dairy matrix could provide oxidative stability and improve the functional characteristics of the developed product (Shahidi & Ambigaipalan, 2015; Messina, 2016).

Table 3.

Compounds obtained from ESI-MS/HRMS analysis of White Carrot.

Name	Formula	Structure	Exact Mass
Acetylgenistin	C23H22O11		474.11621
Dicyclohexylurea	C13H24N2O		224.18886
Genistein	C15H10O5		270.05283
Glycitein	C16H12O5		284.06848
Eburicoic acid	C31H50O3		470.37601

3.3. Optimization of kheer ingredient parameters based on sensory response

The adequacy of developed models was checked. It can be concluded from the R² values that the predicted response was well-fitted with the experimental response. The adjusted and predicted R² values showed no major differences, which confirms the adequacy of the model. Lack-of-fit test results were non-significant (p > 0.05). RSM was employed as a numerical optimization technique to ensure all responses remained within acceptable ranges. Figs. 1 (A-E) illustrate the diverse response surface plots depicting the impact of varying amounts of barnyard millet and brown sugar on various sensory parameters. In the analysis focused on sensory attributes-based parameters, it was observed that trial 7, with a run order of 19, achieved the highest scores across various organoleptic attributes. These sensory scores were as follows: color and appearance, 8.6 ± 0.49; Body and Texture, 9 ± 0.00; Flavor, 8.8 ± 0.40; Mouthfeel, 8.8 ± 0.40; and Overall Acceptability, 8.8 ± 0.40. These findings suggest notable performance in sensory evaluation, highlighting the potential of trial 7 under run order 19 for further exploration and consideration.

3.4. Interactive effect of barnyard millet and brown sugar on the color and appearance of kheer

Color and appearance serve as a predominant factor in assessing the quality of food products. In this study, sensory scores for color and appearance varied from 6.33 to 8.7 (Fig. 1). Interestingly, run number 7 showed the minimum score, while run number 19 exhibited the maximum score for color and appearance. In run number 7, the levels of Barnyard millet, Brown sugar, and White carrot were 25.00, 20.00, and 5.00 g, respectively. Conversely, trial number 19 consisted of Barnyard millet, Brown sugar, and White carrot in quantities of 25.00, 25.00, and 10.00 g, respectively. These observations were utilized to develop a predictive model, as depicted in Eq. 1, which captures the interactive effects of barnyard millet, brown sugar, and white carrot on color and appearance:

Color and appearance = − 80.23 + 4.166*Barnyard millet +1.591*brown sugar +3.23*white carrot - 0.0060*Barnyard millet * brown sugar - 0.068*Barnyard millet * white carrot – 0.0140*brown sugar*white carrot - 0.0629*Barnyard millet²–0.0290*brown sugar²–0.0573*white carrot². – Eq. 1.

Fig. 1A represented a gradual increase in sensory scores for color and appearance with escalating levels of barnyard millet and brown sugar, reaching a peak around 28 g of barnyard millet and 22.5 g of brown sugar. Beyond this threshold, a decline in sensory scores for color and appearance was observed. This may be due to the prevalence of 2-methyl pyrazine and 2,5-dimethyl pyrazine formed by the Maillard reaction, causing product browning.

3.5. Interactive effect of barnyard millet and brown sugar on the body and texture of kheer

The sensory scores for body and texture displayed a range from 6.16 to 9, as outlined in Table 1. Detailed texture profile analysis for the control (A) and optimized (B) kheer samples is presented in Supplementary Fig. 3S.

Notably, the minimum score for body and texture was observed in trial run no. 7, while the maximum score was observed in trial run no. 19. These findings suggest a potential interactive effect among barnyard millet, brown sugar, and white carrot, as indicated by the following linear regression equation (Eq. 2).

Body and texture = − 37.058 + 1.704 Barnyard millet +0.796 brown sugar +2.820 white carrot - 0.0420 Barnyard millet * brown sugar - 0.0620 Barnyard millet * white carrot - 0.0192 brown sugar*white carrot −0.004612 Barnyard millet² + 0.01235 brown sugar²–0.03289 white carrot².

Fig. 1B shows the response 3D surface plot of body and texture influenced by the level of Barnyard millet and brown sugar concentration. From the figure, it can be observed that with an increase in the level of Barnyard millet and sugar, there is a gradual increase in the body and texture.

3.6. Interactive effect of barnyard millet and brown sugar on flavor

The sensory score of flavor in the different trials exhibited a range from 6.4 to 8.8, as presented in Table 1. The minimum flavor score was observed in trial run no. 17, while the maximum was noted in trial number no. 19. In trial no. 17, the levels of Barnyard millet, brown sugar, and white carrot were 27.50 g, 22.5 g, and 3.295 g, respectively. These observations suggest a potential linear relationship among these variables, as indicated by the following Eq. 3. This implies an interactive effect of Barnyard millet, brown sugar, and white carrot on the flavor profile, which warrants further investigation.

Flavor = 1.03801 + 0.7305*Barnyard millet – 1.0496*brown sugar +2.2263*white carrot - 0.008687*Barnyard millet * brown sugar - 0.080667*Barnyard millet * white carrot +0.0313*brown sugar * white carrot +0.000753*Barnyard millet² + 0.0233*brown sugar²–0.03884*white carrot².

Fig. 1C displays the 3D surface plot illustrating the influence of Barnyard millet level and brown sugar concentration on flavor perception. The plot reveals that an incremental rise in the Barnyard millet level is associated with a slight decrease in the sensory score for flavor. Conversely, an increase in sugar concentration initially reduces the flavor score, reaching a minimum at a specific concentration, before subsequently enhancing the flavor score with further increments in sugar content. The characteristic aroma of brown sugar is sweet, caramel, and has slight fruity notes. Certain flavoring compounds like furanones, pyrazines, aldehydes, alcohols, acids, and sulfur-containing compounds are present in brown sugar. Similarly, furfural, 2-methyl pyrazine, 2,5-dimethyl pyrazine, 2-furanmethanol, 2-methyl propanoic acid, and propanoic acid present in brown sugar enhance the flavor (Asikin et al., 2016; García et al., 2017).

3.7. Interactive effect of barnyard millet and brown sugar on mouthfeel

The sensory evaluation of mouthfeel ranges from 6.2 to 8.8. Notably, the lowest score for mouthfeel was recorded in trial run no. 17, while the highest score was achieved in trial runs no. 14 and 19. In trial no. 17, the respective quantities of Barnyard millet, brown sugar, and white carrot were 27.50, 22.5, and 3.295 g. Conversely, trials no. 14 and 19 were characterized by Barnyard millet, brown sugar, and white carrot quantities of 25.00, 20.00, 10.00 g and 25.00, 25.00, 10.00 g, respectively. These data can be effectively modelled using the following regression equation, which facilitates the interpretation of the interactive effects of Barnyard millet, brown sugar, and white carrot.

Mouthfeel = − 16.14256 + 1.1327*Barnyard millet - 0.001133*brown sugar +1.64515*white carrot - 0.020*Barnyard millet * brown sugar - 0.0533*Barnyard millet * white carrot +0.0146*brown sugar * white carrot - 0.00716*Barnyard millet² + 0.0098*brown sugar²–0.018479*white carrot².

Fig. 1D displays the 3D surface plot illustrating the influence of Barnyard millet levels and brown sugar concentrations on mouthfeel. The plot reveals that an increased presence of Barnyard millet correlates with heightened mouthfeel scores, whereas additional sugar demonstrates only a negligible effect on the sensory evaluation of kheer's mouthfeel. Furfural and 2-furanmethanol may be produced by dehydration of hexoses via caramelization (Yaylayan & Keyhani, 2000). 2-Methylpropanoic acid and propanoic acid are produced either from the cleavage of sugars or by microbial fermentation before heating sugarcane juice (Asikin et al., 2016).

3.8. Interactive effect of barnyard millet and brown sugar on overall acceptability

The sensory scores for overall acceptance exhibited a range of 6.00 to 8.8 (Table 1). The lowest score for overall acceptability was recorded in trial run no. 17, while the highest score for overall acceptability was observed in trial run no. 19. The data lend themselves to a linear regression analysis, as evidenced by Eq. 5.

Overall acceptability = −36.641 + 2.137*Barnyard millet +0.49925*brown sugar +2.326*white carrot – 0.03067*Barnyard millet * brown sugar - 0.0640*Barnyard millet * white carrot +0.0186*brown sugar * white carrot - 0.0177*Barnyard millet² + 0.004928*brown sugar²–0.05164*white carrot².

Fig. 1E displays the response 3D surface plot depicting the influence of Barnyard millet (BM) and brown sugar concentration on overall acceptability. The plot reveals a notable trend: an escalation in both BM and brown sugar levels correlates with an increase in overall acceptability.

3.9. Proximate and chemical composition of the control and developed kheer after optimization

Proximate and chemical composition analysis of the control sample and optimized product is presented in Table 4. The nutritional content and functional properties of the control sample and optimized product differed significantly, as shown by proximate and chemical composition analysis. Optimized Kheer has increased protein and phenolic content, indicating increased nutritional value and potential health benefits. The moisture content was relatively comparable between Control (58.93%) and optimized Kheer (59.96%), indicating almost equal hydration levels. Hydration levels are very crucial for maintaining texture and stability during shelf life (Pomeranz & Meloan, 1994). A slight increase in fat content, from 6.43% to 7.07%, may help in improving flavor and mouthfeel, as has also been reported in studies related to the enrichment of Kheer formulations (Gautam et al., 2014). The protein content increased significantly, from 5.30% to 13.66%, indicating an improvement in amino acid availability, as has been observed with plant–dairy blended products (Jha et al., 2011). The carbohydrate content reduced from 27.85% to 19.98%, reflecting a possibility of low glycemic index, which could be favorable for health-conscious consumers. The ash content indicated a slight increase from 1.49% to 1.55%, revealing somewhat higher mineral availability. This is consistent with findings from millet-based kheer mixes, where measurable ash reflects the contribution of cereal and dairy ingredients to the mineral fraction (Bunkar et al., 2012). The use of millets in dairy-based products increases their nutritional quality, dietary fiber, and mineral content compared to traditional cereal-based desserts. The nutritional improvements in the case of millet-based kheer formulations validate the suitability of using barnyard millet and white carrot powder as a value-added material. (Jha et al., 2011; Chandrasekara & Shahidi, 2011).

Table 4.

Proximate and Chemical Analysis of the Optimized and Control Kheer.

Parameter	Amount in mg/ 100 g or as % of the sample
	Control	Optimized Sample
Moisture (%)	58.93 ± 0.06	59.96 ± 1.96
Protein (%)	5.30 ± 1.70	13.66 ± 0.45
Fat (%)	6.43 ± 0.35	7.07 ± 0.13
Total solid (%)	41.06 ± 0.06	40.74 ± 1.27
Ash (%)	1.49 ± 0.34	1.55 ± 0.10
Carbohydrate (%)	27.85 ± 0.45	19.98 ± 1.71
Total phenolic content (mg of GAE/g)	16.35 ± 5.82	21.95 ± 18
% DPPH Inhibition	21.95 ± 18	24.99 ± 1.63
Reducing sugar (%)	8.45 ± 0.45	11.33 ± 0.48

Data represented as mean ± Standard Deviation, (n = 3).

Functional properties also showed evident enhancement. Total phenolic content increased from 16.35 to 21.95 mg GAE/g, which agrees with the work of Balasundram et al. (2006), who reported carrots as one of the natural sources of phenolic compounds with antioxidant activity. Similarly, DPPH inhibition increased from 21.95% to 24.99%, reflecting an increase in free-radical scavenging ability. Vijaya Kumar et al. (2015) reported a similar increase in antioxidant activity upon the addition of vegetables to dairy-based foods. This increase in antioxidant activity can be explained by the presence of bioactive components, mainly phenolic acids and isoflavones like genistein and glycitein, found in white carrot powder. These components have high radical-scavenging potential, thus enhancing the DPPH inhibition capability. Barnyard millet also has different amino acids and bioactive components, which may contribute to the nutritional quality of the product. Thermal treatment involved in kheer formulation may aid in the liberation of phenolics bound within the matrix, hence improving their extraction yield and antioxidant property. Besides, interactions between phytochemicals from plant sources and milk protein may promote the stability and functionality of the bioactive (Chandrasekara & Shahidi, 2011; Ozdal T et al., 2013). The reducing sugar content increased from 8.45% to 11.33%, presumably as a result of breaking down the complex carbohydrates during heating. Such observations were reported by Chandrasekara and Shahidi (2011), where the increase in reducing sugars was noticed after vegetables had undergone thermal processing, which contributed to sweetness enhancement and development of the Maillard reaction.

On the whole, the optimized Kheer reveals an improvement in nutritional composition with enhanced antioxidant properties, yet moisture and structural properties are comparable. Nevertheless, with these enhancements, consumer preference must be put into consideration since sweetness, texture, and flavor are crucial in product acceptability. Balancing nutritional improvements with desirable sensory qualities remains a critical consideration in food product development. The results indicated that the addition of BM and WCP could be a healthy substitute for dairy products, replacing traditional rice and white sugar ingredients. Further, this study suggests that a combination of BM and WCP could be practiced for other milk-based dessert products prepared around the globe. Detailed proximate analysis of white carrot and barnyard millet used in the formulation is included in Supplementary File Table S1.

3.10. Texture profile analysis control and optimized kheer samples

Texture profile analysis (TPA) of control and optimized kheer was performed for different parameters of texture using the probes TA36 at room temperature (Table 5). The texture profile analysis (TPA) graph illustrates the comparison between control and optimized Kheer samples presented in the supplementary material, Figure 3SA and 3SB. The compression tests were conducted using a cylindrical probe TA36, applying a trigger load of 5 g at a test speed of 1 mm/s. Various texture parameters of the developed functional Kheer were evaluated, including hardness, adhesion, cohesiveness, springiness, gumminess, and chewiness. The control sample exhibited the highest hardness, whereas the optimized sample demonstrated reduced hardness. Upon comprehensive analysis, the optimized sample weighing 100–150 g emerged as the most favorable among all variations. It was observed through the texture profile analysis that the optimized Kheer had lesser values for hardness, springiness, and chewiness than the control sample, suggesting that the optimized sample had a softer texture and lesser elasticity. It can be presumed that this decrease in hardness is probably due to the addition of barnyard millet, which might have interfered with the starch-protein complex, thus changing the properties of gelatinization, resulting in a softer consistency (Kavimani et al., 2016). The lower values obtained in springiness and chewiness indicate that the matrix has lesser strength and elasticity, thereby requiring less energy for chewing (Bourne, 2002). On the other hand, the increase in adhesiveness and gumminess is probably due to improved moisture content and higher amounts of soluble solids in the form of brown sugar and barnyard millet, making the optimized Kheer sticky (Arora & Patel, 2017).

Table 5.

Texture profile analysis (TPA) of control and optimized kheer.

Parameter	Control	Optimized sample
Hardness Cycle 1 (g)	35	25
Hardness Cycle 2 (g)	25	15
Adhesiveness (mJ)	0.1	0.8
Cohesiveness	0.71	1.5
Springiness (mm)	12.08	5.38
Gumminess (g)	25	37.5
Chewiness(mJ)	300.2	201.75

3.11. Effect of storage on the shelf life of functional kheer

The pH levels of kheer samples experienced a significant (p < 0.5) decrease, accompanied by a significant (p < 0.05) increase in the acidity during storage at refrigerated temperatures. The interaction effect between these two ingredients (BM and WCP) on the pH and acidity of the product was determined to be non-significant (p > 0.05), as detailed in Table 6. This phenomenon aligns with findings from a study by Aundhkar et al. (2024), who reported that the pH of little millet kheer underwent a significant decrease, while its acidity and viscosity increased significantly during storage at 6 + 1 °C. This may be due to the fiber content prevailing in little millet. The Maillard reaction occurring between sugar and amines leads to the formation of carbonyls, which may be responsible for pH decrease. Furthermore, the standard plate count (SPC) of the kheer exhibited a significant (p < 0.05) growth throughout the storage duration. Over the course of the 14-day storage period, there were no detectable levels of coliform, yeast, and mold (Table 6). However, it was noted that yeast and mold counts became apparent on the 17th day of storage. Importantly, coliform was not detected in the millet and carrot-based kheer samples in the present investigation, which may be due to antimicrobial and antioxidant compounds of added white carrot powder (Nitbani et al., 2022; Subroto & Indiarto, 2020).

Table 6.

Changes in pH, acidity, and microbial count of millet and white carrot base kheer during storage.

Storage duration (days)	pH		Acidity %		SPC (log CFU/g)
	Control	Optimize	Control	Optimize	Control	Optimize
0	6.29 ± 0.051	6.41 ± 0.01	0.282 ± 0.008	0.26 ± 0.001	Nil	Nil
2	6.29 ± 0.012	6.38 ± 0.005	0.29 ± 0.011	0.27 ± 0.001	Nil	Nil
4	6.31 ± 0.012	6.34 ± 0.01	0.32 ± 0.005	0.27 ± 0.005	Nil	Nil
6	6.28 ± 0.005	6.3 ± 0.049	0.34 ± 0.01	0.3 ± 0.006	1.46 ± 0.01	1.33 ± 0.01
8	6.26 ± 0.005	6.27 ± 0.005	0.37 ± 0.004	0.321 ± 0.005	2.83 ± 0.02	3.42 ± 0.33
10	6.21 ± 0.005	6.25 ± 0.006	0.41 ± 0.01	0.367 ± 0.003	3.65 ± 0.03	4.13 ± 0.17
12	6.17 ± 0.005	6.21 ± 0.0	0.45 ± 0.004	0.387 ± 0.0	4.57 ± 0.05	4.11 ± 0.14
14	6.11 ± 0.005	6.14 ± 0.007	0.53 ± 0.005	0.431 ± 0.006	6.49 ± 0.063	6.79 ± 0.11

Data represented as mean ± Standard Deviation (n = 3).

Microbial stability for 14 days at refrigerator temperature indicates that the product is fit for consumption as a fresh dairy-based dessert product. The product stability is comparable to that of other dairy-based products, which usually have a relatively short shelf life due to high moisture content and nutritional composition. In terms of commercial viability, the product can be considered viable if sold locally under cold storage conditions (Aneja et al., 2002). The absence of coliforms during the storage period indicates good hygienic quality and safety of the product, aligning with standard microbiological criteria for dairy products (Food Safety and Standards Authority of India, 2011).

4. Conclsuion

The present study shows that BM and WCP are promising functional ingredients for formulating a value-added dairy product (kheer). Metabolomic profiling has revealed that the presence of various bioactive compounds in BM and WCP could not only enhance the textural properties but also contribute to the functional properties of the product. Based on CCRD, optimization formulation content 25 g BM, 10 g WCP, and 25 g brown sugar to achieve the higher physical and sensory properties. The addition of BM and WCP substantially enhanced protein content, TPC, and DPPH radical inhibition ability of the optimized product compared to the control (P < 0.05). The developed optimized functional kheer shows overall acceptance in comparison with the control. Replacing rice and refined sugar with BM, WCP, and brown sugar further improved the product's health profile, without compromising consumer acceptability. Overall, this study highlights that traditional Indian sweets can be reformulated in innovative ways to create nutritionally enhanced functional food products. BM and WCP are valued ingredients that could provide a feasible approach to diversifying the dairy and functional foods sectors to meet consumers' emerging need for healthier, value-added alternatives. Nevertheless, there is a need for more research, including glycemic index determination and bioavailability evaluation, to determine the health value of the optimized product. Furthermore, pilot-scale production and economic feasibility of the industrial production of kheer need to be investigated.

CRediT authorship contribution statement

Vivek Gautam: Writing – original draft, Investigation, Formal analysis, Data curation. Anantita Sangsuriyawong: Writing – original draft, Investigation, Formal analysis, Data curation. Abhishek Dutt Tripathi: Writing – review & editing, Visualization, Supervision, Software, Resources, Project administration, Methodology, Investigation, Conceptualization. Saptaneel Ghosh: Writing – original draft, Software, Formal analysis. Preetam Banerjee: Writing – original draft, Software, Formal analysis. Muskan Kumari Gupta: Writing – original draft, Software, Formal analysis. Gunvantsinh Rathod: Writing – original draft, Software, Formal analysis. Aparna Agarwal: Writing – review & editing, Visualization, Supervision, Resources, Project administration, Methodology, Investigation, Conceptualization. Papungkorn Sangsawad: Writing – review & editing, Validation. Nilesh Nirmal: Writing – review & editing, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Conceptualization.

Consent for publication

All the authors have provided consent for the publication.

Ethical approval and consent to participate

Ethical approval not required.

Funding

No funding.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.