Effect of pre-treated homogenized egg liquid addition on yogurt quality and metabolic profiles Authors: Yuru Shui, Gan Hu, Xinyu Jiang, Di Zhu, Zhifeng Zhao, Yizi Yuan, Yunxiao Xie, Jiafu Lin, Guofeng Jin, Fang Geng Published: 2026 ABSTRACT AbstractThis study prepared yogurt by adding pre-treated egg liquid to milk to modify the texture and nutritional composition, and systematically analyzed the effects of fermentation and egg liquid addition on yogurt metabolites. Results showed that the addition of pre-treated egg liquid significantly improved the nutritional value of both milk and yogurt, enhanced the sense of thickness in taste, and resulted in a more compact microstructure. A total of 935 metabolites were identified through metabolomics analysis. Compared with ordinary yogurt, the abundance of “amino acids and its metabolites”, “organic acid and its derivatives”, “glycerophospholipids”, and “fatty acyls” in egg-milk yogurt was significantly increased. Differentially abundant metabolites were mainly enriched in pathways such as “glycerophospholipid metabolism” and “ABC transporters”. Specifically, the addition of egg liquid increased small peptides (e.g., γ-Glu-Phe), glycerophospholipids (e.g., LPC(0:0/22:6), LPC(0:0/20:4)), and long-chain polyunsaturated fatty acids (e.g., FFA (22:4), FFA (22:5)), which maintained a significant advantage after fermentation. These findings provide a theoretical basis and new insights for the processing of egg–dairy composite products and the development of functional yogurts.Keywords: Egg-enriched yogurt, Metabolomics, Pre-treated homogenization technology, Quality characteristics Introduction Yogurt is a semi-solid fermented dairy product produced from heat-treated milk inoculated with Lactobacillus bulgaricus and Streptococcus thermophilus for fermentation. Due to its richness in protein, calcium, and beneficial microorganisms, yogurt has become an important component of the daily diet (Vaseghi Bakhshayesh et al., 2024; Wu et al., 2023). The formation of traditional yogurt gel depends mainly on the aggregation of precipitated milk proteins (Asaduzzaman et al., 2021). During the fermentation process, lactic acid bacteria degrade lactose to produce lactic acid, gradually lowering the pH to the isoelectric point of casein at approximately pH 4.6 (Kumari et al., 2025). At this point, casein micelles carry less net negative charge on their surfaces, reducing electrostatic repulsion. Meanwhile, enhanced hydrophobic interactions and molecular cross-linking promote casein assembly into a three-dimensional network gel structure (Aryana and Olson, 2017). Eggs contain abundant high-quality protein (approximately four times that of milk), fat, vitamins, and minerals, as well as other nutrients. In particular, egg white proteins possess excellent functional properties (Hu et al., 2025a; Zhai et al., 2026), including gelation, emulsification, and foaming capacities, enabling their extensive applications in food products such as preserved eggs and egg tofu (Jia et al., 2025; Lv et al., 2022). Recent studies have demonstrated that the incorporation of egg-derived ingredients into yogurt systems may provide a potential approach for modulating yogurt gel networks and improving the functional properties of yogurt. Bouhadi et al. (2021) found that adding 2%–3% egg white protein effectively enhanced the dry matter and protein content of steamed yogurt. Yang et al. (2024) introduced liquid whole egg at concentrations ranging from 0% to 30% into set-type yogurt and found that a 5% addition increased yogurt hardness by 6.28 times and viscosity by 6.1 times. Ridawati and Alsuhendra (2024) compared the sensory and physicochemical properties of “Eggurt” (an egg white-based yogurt) made from different sources of egg whites with traditional milk yogurt, finding that free-range egg white products were most similar to conventional yogurt in texture and flavor. Zang et al. (2023) systematically investigated the gel formation mechanism during the fermentation of egg-based yogurt, identifying hydrophobic interactions and disulfide bonds as the primary driving forces maintaining the gel network structure, which contributed up to 93% in the final gel state. However, most of these studies focus on the direct addition of single egg white or whole egg components, with analytical methods largely limited to apparent indicators such as physicochemical properties and sensory evaluation. Meanwhile, direct incorporation of raw egg liquid often faces challenges including poor thermal stability, tendency to induce gel structure deterioration, and flavor incompatibility. Metabolomics, as a high-throughput and systematic analytical approach, can comprehensively reveal the composition of metabolites in complex food systems and their dynamic changes during processing and fermentation (Xie et al., 2025). Based on this, our research group previously developed a technique for preparing egg white fluid gels by subjecting egg whites to ultrasonic or homogenization pretreatment, followed by heating at 72 °C to form a pre-treated, solution-like egg liquid (Hu et al., 2025b, 2025c; Zhu et al., 2024). In this study, the pre-treated egg liquid was added to milk to modify yogurt texture and nutritional composition, particularly to enhance the protein content of the fermented dairy product. The effects of pre-treated egg liquid on yogurt quality characteristics, microstructure, and rheological properties were systematically evaluated using texture analysis, electronic tongue sensory evaluation, and scanning electron microscopy (SEM). Additionally, widely targeted metabolomics was employed to analyze the differential abundance changes of amino acids, organic acids, and lipid metabolites before and after fermentation, revealing at the molecular level how the pre-treated egg liquid contributes to flavor formation and nutritional enhancement in yogurt. These findings not only provide a theoretical basis for the processing of egg–milk composite products but also offer new insights for the development of functional yogurts. Material and methods Material and chemicals reagents Fresh eggs were purchased from Fengji Food Group Co., Ltd. (Mianyang, China). Pure milk was purchased from Inner Mongolia Mengniu Dairy Co., Ltd. (Hohhot, China). White granulated sugar was purchased from Guangzhou Fuzheng Donghai Food Co., Ltd. (Guangzhou, China), and the starter culture was purchased from Angel Yeast Co., Ltd. (Yichang, China). Preparation of pre-treated homogenized egg white and whole egg liquid The preparation process was performed following our previous methods (Zhu et al., 2024; Hu et al., 2025b, 2025c), as illustrated in Fig. 1A. Preparation process flowchart of pre-treated egg liquid and yogurt samples. A: Preparation of homogenized liquid egg (HLE*) and homogenized egg white (HEW*). B: Preparation of four experimental samples (M, YM, EM, YEM). * indicates samples treated by homogenization followed by pre-coagulation. Preparation of pre-treated whole egg liquid (HLE*): Fresh eggs were manually shelled, and 1000 mL of whole egg liquid was subjected to three cycles of high-pressure homogenization at 20 MPa using a high-pressure homogenizer (SCIENTZ-207A, Ningbo Xinzhi Biotechnology Co., Ltd., Ningbo, China). The homogenized sample was then heated at 72 °C for 10 min to obtain HLE*. Preparation of pre-treated egg white liquid (HEW*): Fresh eggs were manually shelled, and the yolks were separated from the egg whites. A total of 1000 mL of egg white was subjected to three cycles of high-pressure homogenization at 20 MPa. The homogenized sample was then heated at 72 °C for 10 min to obtain HEW*. Sample preparation The preparation procedures for the four groups of samples were as follows (Fig. 1B). Preparation of pure milk (M): Pure milk (920 g) and sucrose (80 g) were mixed. The milk was heated to 75 °C, and sucrose was added with stirring for 10 min to ensure uniform mixing. The mixture was then cooled to 65 °C and homogenized once at 20 MPa using a homogenizer. Preparation of pure milk yogurt (YM, yogurt by milk): Sample M was heated to 80 °C and held for 10 min for sterilization. After sterilization, the sample was immediately covered and cooled in an ice-water bath to 40 °C. Then, 1 g of commercial yogurt starter culture (Baizuan, Angel Yeast Co., Ltd., Yichang, China), mainly containing Lactobacillus delbrueckii subsp. Bulgaricus and Streptococcus salivarius subsp. Thermophilus according to the product instructions, was added. The mixture was stirred for 1 min using a sterilized container until no obvious particles remained, and was immediately poured into sterilized glass yogurt bottles. Fermentation was conducted at 42 ± 0.5 °C for 12 h, and the resulting product was designated as pure milk yogurt. Preparation of egg-milk mixture (EM, egg-milk): Milk (770 g), pre-treated homogenized whole egg liquid (HLE*, 50 g), pre-treated homogenized egg white liquid (HEW*, 100 g), and sucrose (80 g) were combined. The milk was heated to 75 °C, and sucrose was added with stirring for 10 min. The mixture was then cooled to 65 °C, and HLE* and HEW* were added with further stirring for 5 min. The mixture was then homogenized once at 20 MPa to obtain the egg-milk mixture. Preparation of egg-milk yogurt (YEM, yogurt by egg-milk): Sample EM was heated to 80 °C and held for 10 min for sterilization. After sterilization, the sample was immediately covered and cooled in an ice-water bath to 40 °C. Then, 1 g of commercial yogurt starter culture (Baizuan, containing two bacterial strains) was added. The mixture was stirred for 1 min using a sterilized container until no obvious particles remained, and was immediately poured into sterilized glass yogurt bottles. Fermentation was conducted at 42 ± 0.5 °C for 12 h, and the resulting product was designated as egg-milk yogurt. Physicochemical parameter analysis Color measurement The L*, a*, and b* values of YM and YEM were measured using a colorimeter (CR9, Shenzhen Threenh Technology Co., Ltd., Shenzhen, China). Each sample was tested five times. Texture analysis Texture properties of YM and YEM were evaluated using a texture analyzer (TA. TOUCH, Shanghai Baosheng Industrial Development Co., Ltd., Shanghai, China) in TPA mode with a P/0.5 cylindrical probe. Testing parameters: 1 mm/s speed (pre-, test, and post-compression), 8 gf trigger force, 8 mm compression, and 5 s interval. Each sample was tested five times. Taste evaluation Taste characteristics of YM and YEM were analyzed using an electronic tongue (SA402B, INSENT Inc., Japan). After removing the surface layer and the portion in contact with the container, 20 g of each of the three samples per group were combined and blended into a homogeneous paste using a blender. Then, a 40 g aliquot of the paste was diluted with 40 mL of ultrapure water, filtered, and injected into the instrument. The system was equipped with six pre-calibrated sensors (C00, AE1, CA0, CT0, AAE, GL1) and three standard electrodes. Each sample was analyzed in triplicate. Determination of nutritional components The determination of basic nutritional components in the samples was conducted in accordance with Chinese national standard methods. Specifically, moisture, crude protein, fat, ash, and mineral contents were measured by direct drying, Kjeldahl, Soxhlet extraction, high-temperature ashing, and inductively coupled plasma mass spectrometry methods, respectively. Prior to the determination of total fat and total protein contents, the samples were subjected to vacuum freeze-drying (ZL-20D, Guangdong Zhongleng Refrigeration Technology Co., Ltd., Guangdong, China) under the following conditions: cold trap temperature −80 °C, vacuum pressure 10–20 Pa, and drying time 36 h. Microstructure observation The microstructures of YM and YEM were observed using a scanning electron microscope (SEM 300, Carl Zeiss AG, Germany). After freeze-drying, the surface layer and the portion in contact with the container were removed, and the fractured surface was sputter-coated with gold. Observation was performed at an accelerating voltage of 10 kV with a scale bar of 10 μm. Rheological property measurement The rheological properties of the samples were measured using a rheometer (MCR702e, Anton Paar Trading Co., Ltd., Shanghai, China). Temperature ramp test: The shear rate was set at 1 s−1, the gap was 1 mm, and the temperature was ramped from 25 °C to 100 °C to monitor viscosity changes with temperature. Shear viscosity sweep: At 25 °C, the shear rate was varied logarithmically from 0.1 s−1 to 100 s−1 to determine the relationship between viscosity and shear rate. Dynamic rheological measurement: At 25 °C with 0.1% strain and 1–100 rad/s frequency sweep, storage modulus (G′) and loss modulus (G’’) were determined. Metabolomics analysis Sample extraction Thawed samples (50 mg) were mixed with 300 μL of 20% acetonitrile–methanol internal standard solution, vortexed for 3 min, and centrifuged (12,000 rpm, 10 min, 4 °C). The supernatant (200 μL) was incubated at −20 °C for 30 min and re-centrifuged (12,000 rpm, 3 min, 4 °C). An aliquot of 180 μL was then transferred to an injection vial insert. UPLC-MS/MS conditions Chromatographic separation was performed on a Waters ACQUITY UPLC HSS T3 C18 column (1.8 μm, 2.1 mm × 100 mm) at 40 °C with a flow rate of 0.4 mL/min and an injection volume of 2 μL. Mobile phase A was ultrapure water with 0.1% formic acid, and mobile phase B was acetonitrile with 0.1% formic acid. Gradient elution program: 0 min, 95% A; 11.0 min, 10% A; 12.0 min, 10% A; 12.1 min, 95% A; and 14.0 min, 95% A. Mass spectrometry was performed on a triple quadrupole instrument in MRM mode using electrospray ionization (ESI) with the following source conditions: source temperature, 500 °C; spray voltage, +5500 V or −4500 V; gas I and II pressures, 55 and 60 psi; curtain gas pressure, 25 psi. Metabolite identification and quantification Metabolites were identified by comparing retention times, fragment ion pairs, and tandem mass spectrometry (MS/MS) data with those in an in-house MWDB database (MetWare database). According to the identification criteria, metabolites were classified into different confidence levels. Level 1 identification was achieved based on the matching of authentic standards, retention time (RT), precursor ion (Q1), product ion (Q3), and MS/MS spectral information. Level 2 identification was based on the matching of authentic standards, RT, Q1, and Q3 information. Quantitative analysis was conducted using MRM mode, with relative quantification based on peak area calculations. Quality control (QC) samples were periodically injected throughout the analytical sequence to monitor analytical stability. During data preprocessing, metabolites with QC sample coefficient of variation (CV, equivalent to RSD) values < 0.3 were retained for subsequent statistical analysis. Moreover, the stability and reproducibility of the analytical platform were evaluated using QC samples. Statistical analysis Data are presented as mean ± standard deviation. Group comparisons were analyzed using independent sample t-tests, with p < 0.05 * and p < 0.01 ** considered significant. Origin software was used for data visualization. Results and analysis Effect of pre-treated egg liquid on the sensory and nutritional quality of yogurt To verify the consistency of fermentation, the final pH values of both yogurt groups were determined. The pH values of YM and YEM were 4.37 ± 0.16 and 4.45 ± 0.10, respectively, with no significant difference observed between the two groups (p > 0.05). These results suggest that the addition of pre-treated homogenized egg liquid did not adversely affect the fermentation process of yogurt (Fig. S1). In terms of appearance, compared with M and YM, EM and YEM exhibited a distinct yellow color (Fig. 2A). Further analysis of color differences among the yogurt samples using a colorimeter revealed that the L* values for both groups were greater than 90.0, indicating high brightness. However, the L* value of YEM was slightly lower than that of YM, while its a * and b *values were significantly higher than those of YM (p < 0.01). Both groups exhibited negative a * values, indicating a slight greenish shift in YM and YEM, but without noticeable red or green tones (Fig. 2B). These results indicated that the addition of egg liquid slightly reduced the brightness of yogurt while significantly increasing its yellowness, resulting in a more pronounced yellow color. Similarly, Yang et al. (2024) also found that the addition of egg liquid increased the yellowness (b* value) of yogurt. Regarding texture, the hardness of YEM was significantly greater than that of YM (p < 0.01), while its springiness was slightly lower than that of YM, suggesting that the addition of pre-treated egg liquid enhanced the gel structure of yogurt (Fig. 2C). Chewiness was significantly greater in YEM than in YM (p < 0.01), demonstrating that adding egg liquid enhanced yogurt chewiness. Cohesiveness, which reflects the strength of internal structural forces within the yogurt gel, was significantly lower in YEM than in YM (p < 0.01). Effect of pre-treated egg liquid on the quality characteristics of yogurt. A: Appearance comparison. B: Colorimetric analysis. C: Texture and electronic tongue analysis. D: Nutritional composition analysis. The taste profiles of the two yogurts were analyzed using six sensors of an electronic tongue, enabling the determination of five basic tastes (sour, sweet, bitter, salty, and umami) and astringency (Fig. 2C). The results showed that the sourness value of YM was −13.02, slightly below the tasteless threshold (−13), whereas the sourness value of YEM increased to −12.16, indicating a weaker electronic response to sour taste. Furthermore, the bitterness sensor detected richness at low concentrations (Wei et al., 2016), and differences in bitterness sensor responses among samples may reflect variations in bitter taste-related compounds. These results revealed that the YEM group exhibited different electronic tongue response patterns compared with YM, characterized by variations in sourness-, umami-, and bitterness-related sensor responses. Overall, the addition of pre-treated egg liquid modified the electronic tongue response patterns of yogurt without markedly altering the overall taste profile, consistent with the conclusion of Yang et al. (2024) that egg white imparts only a subtle, delicate flavor to yogurt without masking the typical yogurt attributes. Nutritional composition analysis showed that the moisture content of YEM was significantly lower than that of YM (p < 0.01), which can be attributed to the higher proportion of pure milk in YM, as pure milk has a higher water content than the other ingredients in the YEM formulation (Fig. 2D). The protein, fat, and ash contents of YEM were all significantly higher than those of YM (p < 0.01). Mineral element analysis revealed that both yogurts contained abundant minerals, with YEM showing higher levels of K, Mg, Na, P, and S compared to YM (Fig. 2D). These results indicated that the addition of HEW* and HLE * significantly increased the protein, fat, and multiple mineral contents of yogurt, thereby enhancing its nutritional value. Effect of pre-treated egg liquid on the microstructure and rheological properties of yogurt The microstructures of the two yogurts were observed by SEM to better understand their textural properties. After freeze-drying, both yogurt samples lost moisture, and their microstructures consisted of accumulated protein aggregates with numerous pores that could accommodate whey. The pores in YM were irregular in size and relatively large in diameter. After the addition of pre-treated egg liquid, YEM exhibited a denser texture, with more uniformly distributed pores of smaller diameter (Fig. 3A). This could restrict the rearrangement of the protein network to some extent, thereby reducing whey exudation in the yogurt (Qing et al., 2023). Furthermore, the denser microstructure of YEM also helped explain why its hardness was significantly greater than that of YM (p < 0.01). These results indicated that the addition of HEW * and HLE * led to a denser and more stable gel structure in yogurt. Effect of pre-treated egg liquid on the microstructure and rheological properties of yogurt. A: Scanning electron microscopy (SEM) images. B: Rheological properties. Temperature ramp tests were performed on the two types of milk before fermentation (Fig. 3B). From 0 to 85 °C, the trends of M and EM were similar. The viscosity of the M group remained at a low level throughout the heating process, increasing only slightly after 90 °C. In contrast, the viscosity of EM also remained at a low level (<10 mPa s) before 90 °C but began to increase sharply from 90 °C and reached 116 times that of M at 100 °C. This was attributed to the gradual heat-induced coagulation of the added HEW * and HLE * at this temperature, during which they transformed from a viscoelastic liquid into a gel-like structure. Conventional thermal sterilization processes for yogurt are typically carried out at 85–90 °C for a certain period of time. These results indicated that after the addition of HEW* and HLE * to yogurt, the product could still withstand the sterilization temperature and duration used in conventional yogurt production. Shear rheology results showed that the viscosity of all groups gradually decreased with increasing shear rate (Fig. 3B), indicating that they were pseudoplastic fluids with shear-thinning behavior (Akshit et al., 2025; Guerra et al., 2022; He et al., 2026), consistent with the flow properties observed in other yogurts (Nemati and Mozafarpour, 2024). Moreover, compared with YM, the YEM group maintained a relatively higher viscosity at high shear rates, suggesting that the addition of pre-treated egg liquid enhanced the mechanical resistance of the gel network. This may be attributed to the increased fat content in yogurt resulting from the addition of egg liquid, which hindered fluid flow. On the other hand, the incorporation of egg liquid altered the charge state of the particles, increasing electrostatic repulsion between them, thereby enhancing flow resistance and contributing to the formation of a more stable system. Dynamic rheological results showed that the elastic modulus G′ was significantly lower than the viscous modulus G’’ for both the M and EM groups, indicating that the samples before fermentation were more inclined toward viscoelastic liquids. The fermented samples exhibited significantly higher elastic modulus (G′) than viscous modulus (G’’), indicating the formation of a predominantly elastic gel network after fermentation (Fig. 3B). Metabolite profiling The total ion current (TIC) chromatograms of different quality control (QC) samples showed highly overlapping retention times and peak intensities (Fig. S2), indicating stable MS signals and good analytical reproducibility. A total of 935 metabolites were detected across the four groups (M, EM, YM, and YEM), comprising 455 in positive ion mode and 480 in negative ion mode. These were classified into 16 categories, including 234 “amino acid and its metabolites”, 136 “organic acid and its derivatives”, 90 “benzene and substituted derivatives”, 85 “glycerophospholipids”, 79 “nucleotide and its metabolites”, 76 “heterocyclic compounds”, 70 “fatty acyls”, 51 “alcohol and amines”, 38 “carbohydrates and its metabolites”, 20 “coenzyme and vitamins”, 17 “hormones and hormone related compounds”, 15 “aldehyde, ketones, esters”, 7 “bile acids”, 3 “tryptamines, cholines, pigments”, 2 “glycerolipids”, and 12 “others” (Fig. 4A). Metabolomic profiles of milk (M), egg-milk (EM), yogurt by milk (YM), and yogurt by egg-milk (YEM). A: Classification and quantity distribution of 935 metabolites identified in the four samples. B: Comparison of total metabolite abundance among the four sample groups. Based on mass spectrometry signal intensity, the abundances of metabolites in the four samples were quantitatively compared (Fig. 4B). The results showed that, prior to fermentation, the total abundance of the EM group was significantly higher than that of the M group. Similarly, after fermentation, the total abundance of the egg-milk yogurt group (YEM) was also higher than that of the pure milk yogurt group (YM), indicating that the addition of pre-treated egg liquid significantly increased the overall metabolite abundance in the system. Regarding compound classification, the top five abundant categories varied among the four groups, but were primarily concentrated in “fatty acyls”, “organic acid and its derivatives”, “carbohydrates and its metabolites”, “amino acid and its metabolites”, and “benzene and substituted derivatives”. This suggested that while maintaining the homeostasis of the core metabolic profile, the egg-milk composite system was associated with increased levels of several metabolites related to flavor and nutritional properties through the addition of egg liquid. Specifically, compared with ordinary milk (M), the abundances of “organic acid and its derivatives”, “amino acid and its metabolites”, “glycerophospholipids”, and “fatty acyls” in the egg-milk mixture (EM) were significantly increased. Notably, the abundance of “glycerophospholipids” in EM was approximately three times that of M. However, during the fermentation process from EM to YEM, the abundance of “glycerophospholipids” decreased significantly. This may be attributed to the degradation of glycerophospholipids by lactic acid bacteria during fermentation, converting them into lysophospholipids, free fatty acids, and other small-molecule flavor precursors. Moreover, compared with YM, the abundances of “amino acid and its metabolites”, “organic acid and its derivatives”, and “fatty acyls” in YEM were significantly increased, which may contribute to the formation of taste-related metabolites and flavor characteristics of yogurt. Therefore, to further elucidate the impact of pre-treated egg liquid on yogurt quality, this study focused on an in-depth analysis of the aforementioned highly contributory categories, namely “amino acid and its metabolites”, “organic acid and its derivatives”, as well as lipid metabolites (lysophospholipids and free fatty acids), which were involved in structural support and aroma precursor formation. Key differentially abundant metabolites (DAMs) Principal component analysis of the four milk groups (M, EM, YM, YEM) revealed that PC1 and PC2 accounted for 64.77% of the total variance. The four replicates within each group formed distinct clusters with clear separation between groups (Fig. 5A), indicating good reproducibility and reliability. Differentially abundant metabolites (DAMs) were identified using criteria including variable importance in projection (VIP ≥1) and absolute log2 fold change (|log2FC|) > 1, and were visualized via volcano plots (Fig. 5B–D). In the comparison between EM and M, a total of 317 DAMs were identified (299 higher in the EM group and 18 higher in the M group; Fig. 5B), primarily comprising 104 “amino acid and its metabolites”, 64 “glycerophospholipids”, 25 “nucleotide and its metabolites”, and 33 “organic acid and its derivatives”. The top five DAMs ranked by VIP value were 2-hydroxy-3-methyl butanoic acid, DL-3-phenyllactic acid, LPE (O-18:1/0:0), LPC(0:0/22:6), and LPC(0:0/20:4), all of which were higher in the EM group (Fig. 5E). Compared to YM, YEM identified 236 DAMs (216 higher in the YEM group and 20 higher in the YM group; Fig. 5C), mainly including 82 “amino acid and its metabolites”, 64 “glycerophospholipids”, 18 “nucleotide and its metabolites”, and 17 “organic acid and its derivatives”. The top five DAMs by VIP value were LPC(0:0/20:4), LPC(0:0/22:6), 2-hydroxy-3-methyl butanoic acid, Ile-Ile, and Leu-Leu, all of which were elevated in the YEM group (Fig. 5E). In the comparison between YEM and EM, 375 DAMs were identified (205 higher in YEM and 170 higher in EM; Fig. 5D), predominantly consisting of 115 “amino acid and its metabolites”, 57 “organic acid and its derivatives”, 57 “glycerophospholipids”, and 45 “nucleotide and its metabolites”. The top five DAMs by VIP value were malic acid, inosine, LPE (16:0/0:0), LPE (14:0/0:0), and benzoic acid. Among these, benzoic acid was higher in YEM, while the other four were higher in EM (Fig. 5E). These core DAMs were mainly involved in lipid metabolism, amino acid metabolism, and organic acid metabolism, providing key clues for further in-depth investigation of the regulatory mechanism of egg liquid addition on yogurt quality. Metabolomic analysis of milk (M), egg-milk (EM), yogurt by milk (YM), and yogurt by egg-milk (YEM). A: Principal component analysis (PCA) score plot. B-D: Volcano plots of pairwise comparisons (B: EM vs. M; C: YEM vs. YM; D: YEM vs. EM). E: Top 5 differentially abundant metabolites ranked by VIP value. Subsequently, KEGG enrichment analysis was performed on DAMs from pairwise comparisons of EM vs. M, YEM vs. YM, EM vs. YEM, and M vs. YM (Table S1). Compared with pure milk, DAMs in the milk supplemented with pre-treated egg liquid (EM vs. M) were significantly enriched in the pathways of “glycerophospholipid metabolism”, “nucleotide metabolism”, “linoleic acid metabolism”, “aminoacyl-tRNA biosynthesis”, and “valine, leucine and isoleucine biosynthesis” (p < 0.05). Between ordinary yogurt and egg-based yogurt (YEM vs. YM), the DAMs were significantly enriched in the pathways of “glycerophospholipid metabolism”, “fatty acid biosynthesis”, “linoleic acid metabolism”, “ABC transporters”, and “biosynthesis of unsaturated fatty acids” (p < 0.05). Compared with egg-based milk, the DAMs in fermented egg-based yogurt (EM vs. YEM) were significantly enriched in the pathways of “glycerophospholipid metabolism”, “2-oxocarboxylic acid metabolism”, “biosynthesis of amino acids”, “glyoxylate and dicarboxylate metabolism”, and “citrate cycle (TCA cycle)” (p < 0.05). Compared with unfermented pure milk, the DAMs in fermented pure milk yogurt (M vs. YM) were significantly enriched in the pathways of “biosynthesis of amino acids”, “ABC transporters”, “2-oxocarboxylic acid metabolism”, “valine, leucine, and isoleucine degradation”, “arginine biosynthesis”, “alanine, aspartate, and glutamate metabolism”, and “glycerophospholipid metabolism” (p < 0.05). Contribution of pre-treated egg liquid addition to the metabolic profiles of milk and yogurt Eggs are widely recognized as a nutrient-dense food source. Egg white is rich in high-quality proteins (including ovalbumin, ovotransferrin, lysozyme, etc.), while egg yolk contains abundant glycerophospholipids, long-chain polyunsaturated fatty acids, fat-soluble vitamins, and various mineral elements (Liu et al., 2026; Zhang et al., 2025). To further analyze the impact of egg-derived ingredients on dairy product quality, this study employed metabolomics technology to systematically analyze the effects of pre-treated egg liquid on the metabolite composition of milk, compare the metabolic profiles between YEM and YM, and examine the changes in metabolites of M and EM before and after fermentation. We focused on analyzing the top 10 DAMs ranked by VIP value within the categories of “organic acid and its derivatives”, “amino acid and its metabolites”, “lysophospholipids”, and “free fatty acids”. Compared with M, the total abundance of these four categories of metabolites in EM was significantly higher (Fig. 6). Similarly, the total abundance of these four categories of metabolites in YEM was also significantly higher than that in YM (Fig. 7), indicating that the addition of pre-treated egg liquid comprehensively increased the metabolite abundance in both milk and fermented milk. In terms of organic acids, the abundances of 2-hydroxyisocaproic acid, DL-3-phenyllactic acid, 2-hydroxy-2-methyl butyric acid, 2-hydroxy-3-methyl butanoic acid, and L-3-phenyllactic acid in EM were more than 30 times higher than those in M (Fig. 6E). These hydroxy acids are metabolic intermediates of amino acids (e.g., valine and leucine), suggesting that egg-derived ingredients contributed additional amino acid-related metabolites and organic acids associated with flavor formation. Notably, DL-3-phenyllactic acid and L-3-phenyllactic acid, which are produced from phenylalanine metabolism, suggested that the addition of egg liquid may impart a unique flavor to yogurt (Yu et al., 2015). Meanwhile, these organic acids introduced by egg liquid maintained a significant advantage after fermentation, with their abundances in YEM being 1.4- to 2.8-fold higher than those in YM. These organic acids not only contribute a mild sour taste but also serve as flavor modifiers to increase the richness of the taste profile. Comparison of key differentially abundant metabolites between milk (M) and egg-milk (EM). A: Total abundance of organic acids and their derivatives. B: Total abundance of amino acids and small peptides. C: Total abundance of lysophospholipids. D: Total abundance of free fatty acids. E: Heatmap of the top 10 differentially abundant metabolites ranked by VIP value in each subclass. Comparison of key differentially abundant metabolites between yogurt by milk (YM) and yogurt by egg-milk (YEM). A: Total abundance of organic acids and their derivatives. B: Total abundance of amino acids and small peptides. C: Total abundance of lysophospholipids. D: Total abundance of free fatty acids. E: Heatmap of the top 10 key differentially abundant metabolites ranked by VIP value in each subclass. Amino acids and their metabolites are key substances affecting the nutritional value and taste characteristics of dairy products (Yan et al., 2025). In this study, a total of 234 “amino acids and its metabolites” were identified, representing the most abundant metabolite category. The total abundance of “amino acids and small peptides” and the levels of individual metabolites in EM were significantly higher than those in M (p < 0.05). Notably, small peptides such as γ-Glu-Phe, Ile-Val, Glu-Leu, and Leu-Val exhibited fold changes greater than 20, and significant increases were observed in essential amino acids including L-tryptophan, L-valine, L-isoleucine, DL-leucine, and lysine (Fig. 6E), indicating that pre-treated egg liquid is a rich source of amino acids and small peptides. Among these, branched-chain amino acids (valine, leucine, and isoleucine) can serve as precursors for α-keto acids and subsequently contribute to the formation of volatile flavor compounds during fermentation. Meanwhile, significant increases in taste-active amino acids such as L-tyrosine and L-phenylalanine also provided sufficient nitrogen sources and taste precursors for subsequent fermentation. On the other hand, these egg liquid-derived amino acids and small peptides maintained significant advantages after fermentation, with 7 out of the top 10 free amino acids ranked by VIP value being consistent with the DAMs identified in EM vs. M. The significantly increased small peptides in YEM mainly included Leu-Glu, γ-Glu-Phe, Val-Leu, Leu-Ala, Glu-Tyr, and Leu-Leu (with fold changes greater than 10.0), further suggesting that egg-derived proteins contributed to the enrichment of amino acids and peptides in yogurt. Lipid metabolites, particularly glycerophospholipids and fatty acyls, not only serve as key structural components supporting the gel network but also act as crucial precursors for various aroma compounds. In this study, the total abundance of “lysophospholipids” in EM was approximately threefold higher than that in M (Fig. 6C), with specific lysophosphatidylcholines (LPC) and lysophosphatidylethanolamines (LPE), including LPC(0:0/22:6), LPE (22:6/0:0), LPE (0:0/22:6), and LPC(20:4/0:0), showing 44.5- to 190.9-fold higher abundance in EM compared with M (Fig. 6E). Glycerophospholipids in egg yolk are rich in long-chain polyunsaturated fatty acids (PUFAs), especially docosahexaenoic acid (DHA, 22:6) and arachidonic acid (AA, 20:4), which play vital physiological roles in infant brain and retinal development. Incorporation of these lipid components may improve the nutritional profile of dairy products. Regarding free fatty acids (FFA), the abundance of FFA (22:5) in the EM group was 12.2-fold higher than that in the M group, while FFA (20:2), FFA (18:3), FFA (22:6), FFA (22:4), FFA (16:1), and FFA (18:2) were also significantly elevated to varying degrees (all greater than 2.0-fold; Fig. 6E). These PUFAs, particularly those from the ω-3 and ω-6 series, play essential roles in maintaining cardiovascular health, modulating immune function, and promoting nervous system development. Meanwhile, these egg liquid-derived lysophospholipids and free fatty acids remained significantly higher after fermentation. In the YEM group, the abundances of LPC(0:0/22:6), LPC(0:0/20:4), LPC(O-16:0), LPE (O-18:1/0:0), and LPC(20:3/0:0) were approximately 22- to 244-fold higher than those in YM (Fig. 7E), and the aforementioned six PUFAs were also significantly increased by 1.6- to 2.8-fold (among which FFA (22:5) increased by 13.5-fold). Interestingly, although the abundance of lysophospholipids decreased significantly and that of free fatty acids increased significantly during fermentation from EM to YEM, suggesting possible phospholipid hydrolysis or lipid transformation during fermentation, the final product (YEM) still contained significantly higher levels of lysophospholipids and free fatty acids compared with conventional yogurt without egg liquid addition (YM). This may be attributed to the higher initial abundance of glycerophospholipids introduced by egg liquid. Furthermore, we observed similar trends in DAMs between the M vs. EM and YM vs. YEM comparisons, suggesting that the metabolic characteristics of egg-derived ingredients contributed substantially to the observed differences between conventional and egg-enriched yogurt. Metabolic changes during the fermentation of egg-based yogurt After fermentation with lactic acid bacteria, the metabolite profiles changed significantly. Compared with the EM group, the total abundance of “organic acid and its derivatives” in YEM was significantly increased (p < 0.01), mainly including DL-3-phenyllactic acid, L-3-phenyllactic acid, hydroxyphenyllactic acid, (R)-2-hydroxybutyric acid, benzoic acid, 2-hydroxy-2-methyl butyric acid, and 2-hydroxyisocaproic acid, with fold changes ranging from 12.1 to 80.4 (Fig. 8). Studies have shown that lactic acid bacteria degrade proteins into small peptides and free amino acids via extracellular proteases and peptidases. After entering the cells, amino acids can be converted into corresponding α-keto acids by aminotransferases (Smit et al., 2005). As central metabolic intermediates, α-keto acids can be further converted into hydroxy acids. However, the abundances of malic acid (0.05-fold) and hippuric acid (0.02-fold) decreased significantly after fermentation (Fig. 8E). The decrease in malic acid may be associated with the metabolic conversion of malic acid by lactic acid bacteria during fermentation (Gerardi et al., 2019), a process that reduces the sharp sourness of yogurt, resulting in a milder and smoother mouthfeel (Li et al., 2024). Previous studies have suggested that malolactic enzyme and malic enzyme may participate in malic acid conversion (Landete et al., 2013; Ozcan et al., 2021). We speculate that similar enzymatic activities may also exist in the yogurt system of this study. Zheng et al. (2022) also found that malic acid content showed a significant decreasing trend during yogurt fermentation and storage. Hippuric acid is a naturally occurring metabolite in dairy products. It has been reported that hippuric acid is negatively correlated with benzoic acid in dairy products, and hippuric acid may be harmful to health (Del Olmo et al., 2017; Sun et al., 2021). Similarly, during milk fermentation with Lactobacillus plantarum CCFM 8610, the strain degraded hippuric acid to produce benzoic acid, thereby reducing potential health risks and enhancing product storage stability (Huang et al., 2022). In this study, we also observed a substantial decrease in hippuric acid abundance after fermentation, suggesting that the lactic acid bacteria used possess similar hippuric acid-degrading and -converting abilities, which may influence the metabolic characteristics of fermented milk. These results indicated that the lactic acid bacteria fermentation process generated abundant hydroxy acids and aromatic lactic acid derivatives through amino acid metabolic pathways, which imparted a mild sourness and rich flavor to yogurt, contributing significantly to yogurt quality formation. Similar conclusions were obtained for ordinary milk (M) before and after fermentation. Compared with M, the total abundance of “organic acid and its derivatives” in YM was significantly increased (p < 0.01), among which the abundance of lactic acid increased by 80.6-fold. DL-3-phenyllactic acid, L-3-phenyllactic acid, and 2-hydroxyhexanoic acid were also observed to increase by 1217.1- to 2814.9-fold after fermentation, and 2-hydroxy-2-methyl butyric acid, 2-hydroxy-3-methyl butanoic acid, and hydroxyphenyllactic acid increased by 126.5- to 186.5-fold after fermentation (Fig. S3). However, the abundances of hippuric acid (0.05-fold) and malic acid (0.1-fold) decreased significantly after fermentation. Comparison of key differentially abundant metabolites between egg-milk (EM) and yogurt by egg-milk (YEM). A: Total abundance of organic acids and their derivatives. B: Total abundance of amino acids and small peptides. C: Total abundance of lysophospholipids. D: Total abundance of free fatty acids. E: Heatmap of the top 10 differentially abundant metabolites ranked by VIP value in each subclass. Free amino acids are key precursors of various aroma compounds in yogurt, such as aldehydes, alcohols, and amines (Warakaulle et al., 2025). Compared with EM, the total abundance of “amino acids and small peptides” in YEM did not change significantly (Fig. 8B). However, the abundances of small peptides such as Phe-Met, Tyr-Pro, and Val-Leu increased substantially, with fold changes ranging from 26.5 to 105.4. Meanwhile, free amino acids such as L-valine and L-tryptophan also showed an increasing trend. However, unlike pure milk fermentation, where most amino acids significantly increased, some amino acid abundances in the YEM group decreased significantly, including lysine, L-tyrosine, aspartic acid, L-glutamine, and L-cysteine, with fold changes ranging from 0.11 to 0.33 (Fig. 8E). This difference may be attributed to the more complex amino acid composition in the fermentation system following egg liquid addition, with certain amino acids being preferentially consumed during fermentation for growth metabolism and flavor precursor synthesis. For comparison, after fermentation of ordinary milk (M vs. YM), the total abundance of “amino acids and small peptides” increased significantly, with substantial accumulation of small peptides such as Val-Phe-Ala, Tyr-Pro, Pro-Met, and Ala-Ala (20.0- to 85.4-fold), and significant increases in free amino acids including L-tryptophan, L-valine, L-isoleucine, and DL-leucine, providing a material basis for the formation of flavor compounds (Fig. S3). This indicated that during fermentation, lactic acid bacteria efficiently hydrolyzed milk proteins (mainly casein and whey protein), releasing large amounts of free amino acids and small peptide fragments. These results suggested that pre-treated egg liquid itself might provide a sufficient nitrogen source (the EM group was significantly higher than the M group), resulting in a balanced release and consumption of amino acids during fermentation, thus resulting in no significant change in total abundance. In contrast, fermentation of pure milk was dominated by milk protein hydrolysis, manifesting as a significant increase in total abundance. In terms of lipid metabolites, compared with EM, the abundance of “lysophospholipids” in YEM generally decreased significantly (Fig. 8C). In particular, the abundances of LPE (0:0/18:2), LPE (0:0/14:0), LPE (14:0/0:0), and LPE (18:3/0:0) decreased by 0.01- to 0.02-fold (Fig. 8E), indicating that lactic acid bacteria may have extensively utilized or transformed these lysophospholipids as sources of phosphorus, nitrogen, and energy for growth. García-Cano et al. (2020) confirmed that phosphatases in lactic acid bacteria can efficiently hydrolyze various glycerophospholipids. The significant enrichment of the “glycerophospholipid metabolism” pathway in KEGG also demonstrated the presence of active phospholipid metabolism during fermentation (Table S1). In contrast, the abundances of various free fatty acids in YEM increased significantly (Fig. 8D). The abundances of long-chain polyunsaturated fatty acids, such as FFA (22:6), FFA (20:2), FFA (22:4), and FFA (18:3), in YEM were 1.3- to 1.9-fold higher than those in the EM group. Saturated and monounsaturated fatty acids, such as FFA (14:0), FFA (12:0), and FFA (18:1), also accumulated significantly (Fig. 8E). Studies have shown that lactic acid bacteria-associated lipolytic activity can contribute to lipid transformation during fermentation, decomposing it into glycerol and fatty acids. Meanwhile, lipids synthesized by the growth of lactic acid bacteria themselves are secreted into the milk system, resulting in significant synthesis and accumulation of glycerides, most polar lipids, and functional long-chain unsaturated fatty acids and their metabolites, such as arachidonic acid, oleic acid, and linoleic acid (Gao et al., 2022; Lordan et al., 2020). These free fatty acids contribute to the characteristic taste of yogurt, whereas the accumulation of long-chain polyunsaturated fatty acids substantially enhances its nutritional profile. A similar trend of a general decrease in lysophospholipids and a significant increase in free fatty acids was also observed in ordinary milk fermentation (M vs. YM; Fig. S3), confirming that the utilization of phospholipids by lactic acid bacteria is a common phenomenon during fermentation. It should be noted that in this study, the YEM group was prepared by partially replacing milk with pre-treated homogenized egg liquid; therefore, differences in protein, lipid, mineral, and metabolite compositions between the two groups were inevitable. The specific contributions of altered raw material composition to the changes in yogurt quality characteristics and metabolic profiles require further investigation. Conclusion This study demonstrated that pre-coagulation homogenization technology could improve the overall quality of egg-based yogurt. The addition of pre-treated egg liquid significantly enhanced the nutritional composition of yogurt by increasing protein, fat, and mineral contents, while also contributing to a denser microstructure with more uniformly distributed pores compared with conventional yogurt. Metabolomics analysis revealed that the addition of egg liquid increased the total abundance of key metabolites, including hydroxy acids, small peptides, glycerophospholipids, and free fatty acids, which remained significantly higher after fermentation. Further analysis indicated that differential metabolites between the two yogurts were significantly enriched in pathways such as “glycerophospholipid metabolism”, “fatty acid biosynthesis”, “linoleic acid metabolism”, “ABC transporters”, and “biosynthesis of unsaturated fatty acids”. By directly introducing abundant amino acids and glycerophospholipids, pre-coagulation homogenization technology effectively improved the processing compatibility of egg liquid, thereby significantly enhancing the nutritional quality and flavor precursor foundation of dairy products. These findings provide new insights and a theoretical basis for the development of functional egg-enriched fermented dairy products. CRediT authorship contribution statement Yuru Shui: Writing – original draft, Writing – review & editing, Validation, Methodology, Data curation; Gan Hu: Writing – original draft, Methodology, Funding; Xinyu Jiang: Investigation, Validation; Di Zhu: Methodology; Zhifeng Zhao: Writing – original draft, Writing – review & editing; Yizi Yuan: Investigation; Yunxiao Xie: Writing – review & editing, Conceptualization, Data curation; Jiafu Lin: Writing – review & editing; Guofeng Jin: Conceptualization; Supervision; Fang Geng: Validation; Writing – review & editing, Funding. Declaration of competing interest The authors declare no competing financial interest. Acknowledgments This research was supported by the Cross-Innovation Open Project of Food Flavor and Health, Beijing Technology & Business University (FFHCI-2026065), Sichuan Science and Technology Program (2026NSFSC1028), Chengdu Science and Technology Program (2024-YF05-01096-SN), and Sichuan Innovation Team Project of National Modern Agricultural Industry Technology System (SCCXTD-2026-24). Handling Editor: Dr. Yeonhwa Park Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.crfs.2026.101507. Contributor Information Yunxiao Xie, Email: xieyunxiao@cdu.edu.cn. Jiafu Lin, Email: linjiafu@cdu.edu.cn. Supplementary data The following is the Supplementary data to this article. Multimedia component 1 References Associated Data Supplementary Materials Multimedia component 1