Applications of metabolomics in cow health assessment.

Diseases and metabolic disorders in dairy cows have a negative impact on milk yield and quality, fertility, and overall health. Metabolomics has emerged as a powerful promising approach for investigating metabolic alterations associated with health status and for identifying potential biomarkers for disease diagnosis and monitoring of metabolic states. This review provides an overview of research conducted between 2010 and 2024, focusing on metabolic changes in dairy cows in relation to the most common and challenging diseases in dairy farming, as well as physiological states indicative of overall health. While several studies have reported promising metabolite candidates, many are limited by small sample size and lack of validation in larger, independent cohorts. As such, current findings should be regarded as exploratory rather than diagnostic tools. Future research should prioritize large-scale, well-controlled validation studies that account for confounding factors and aim to integrate metabolomics with other omics approaches to enhance biological interpretation and contribute to the development of more comprehensive health-monitoring strategies in the future.

Relevant publications were retrieved from the Web of Science database. The bibliographic search was conducted in December 2024 and included literature published since 2010. Using the keywords “metabolomics”, “disease”, and “bovine”, 124 studies investigating bovine diseases through metabolomics were identified. An additional 228 studies related to bovine health and metabolomics were retrieved using the search terms “metabolomics”, “bovine”, and “health”. Following a thorough screening process, publications were excluded if they were not original research articles, were unrelated to metabolomics, focused on animals other than cows, or addressed only management strategies in dairy production. Articles with substantially overlapping content were also excluded. Ultimately, 96 original research articles were selected and discussed in the present review.

Cows experience different metabolic stress under different physiological conditions. Metabolomics has also been applied to better understand metabolic changes associated with various physiological conditions such as pregnancy, transition and lactation and parity (see Table  3 ). Using 1 H NMR-based metabolomics, Sangwan et al. ( 2024 ) investigated the urinary metabolic profiles of early pregnant and non-pregnant Mithun (a cattle breed native to Northeast India). They identified six metabolites (i.e. kynurenine, kynurenate, 3-hydroxykynurenine, quinolinate, tyrosine, and leucine) as early pregnancy markers, all of which were higher in pregnant cows than in non-pregnant cows. 3-Hydroxykynurenine is a key intermediate in the kynurenine pathway and its elevation during pregnancy may be due to induction of the kynurenine pathway influenced by strong placental indoleamine 2,3-dioxygenase expression (Wang et al., 2020 ; Zong et al., 2016 ). Table 3 Summary of selected metabolomics studies investigating metabolic alterations associated with physiological status in dairy cows Physiological status Aim of study Metabolomics Conclusion References Pregnancy To identify potential pregnancy detection biomarkers Untargeted metabolomics data (270 metabolites) on urine from 9 non-pregnant and 8 pregnant cows explored by PCA, PLS-DA, and OPLS-DA Increased levels of kynurenine, kynurenate, 3-hydroxykynurenine, quinolinate, tyrosine, and leucine in pregnant cows were identified as potential early pregnancy markers Sangwan et al., ( 2024 ) To investigate the differences in urinary metabolome of pregnant and non-pregnant cows Untargeted 1 H NMR-based metabolomics data (24 metabolites) of urine collected on day 0, 10, 18, 35 and 42 of gestation (n = 6) and on day 0, 10 and 18 post-inseminations of non-pregnant (n = 6) cows and explored by ANOVA, PCA, PLS-DA, and OPLS-DA Anthranilate, 3-hydroxykynurenine, and tyrosine levels were higher in pregnant compared to non-pregnant cows Sarangi et al., ( 2022 ) Transition period To characterize changes of cows’ metabolic phenotype during the transition period Untargeted LC–MS, FIA-MS/MS-based metabolomics data (188 metabolites) on serum from cows at 42, 10 days before calving and 3, 21 and 100 days after calving, explored by PCA and PLS-DA Glycerophospholipids and sphingolipids decreased at 10 days pre-calving and 3 days post-calving, during the transition period Kenéz et al., ( 2016 ) Lactation To identify metabolic differences between lactating and non-lactating periods in cows Untargeted GC–MS-based metabolomics analysis identified 33 shared metabolites across rumen fluid, serum, milk, and urine, and 274 metabolites in mammary gland tissue, from 6 non-lactating and 6 lactating cows. Data were explored using PCA, PLS-DA, and OPLS-DA Lower levels of lactobionic acid, citric acid, orotic acid and oxamide in non-lactating compared to lactating periods were identified as potential biomarkers of the metabolic difference Sun et al., ( 2017 ) Heat stress To evaluate the thermotolerance differences between HH and HB under chronic HS and TN conditions Untargeted LC–MS-based metabolomics data (24 metabolites) on serum from 6 HH and 6 HB cows explored by t-test and OPLS-DA HH cows had lower betaine levels than HB cows; HB cows had lower betaine levels than healthy crossbred dairy buffaloes Gu et al., ( 2018 ) To analyze the differences in the metabolic profiles of serum and urine across three beef cattle breeds under heat stress Untargeted GC–MS-based metabolomics data (78 metabolites in serum and 81 in urine) from 8 XHC, 8 SXC and 8 JMY, explored by t-test PCA, and PLS-DA Under heat stress XHC showed the highest glucose, pyruvic acid, lactic acid, uric acid, α/γ-tocopherol and lowest malondialdehyde, and allantoin; SXC showed the highest glycerol, palmitic, oleic, linoleic, arachidonic acids, citric acid, aconitic acid, and fumaric acid. JMY breed had the highest methionine, glutamine, phenylalanine, tyrosine, asparagine, ornithine, urea, methylmalonic acid, methylcitric acid and the lowest putrescine Liao et al., ( 2018 ) Nitrogen utilization efficiency To identify metabolites associated with NUE in dairy cows Untargeted LC–MS/MS-based metabolomics data (439 metabolites) on plasma from 10 low NUE and 10 high NUE cows explored by One-way ANOVA and PLS-DA Sucrose, 1-( cis -13-docosenoic acid)-glycerol monoacylglycerol, 2-amino-6-hydroxyhexanoic acid, L-glutamine, and L-arginine were less abundant in high NUE cows Li et al., ( 2024a ) Milk yield To investigate potential biomarkers of milk production in HY and LY dairy cows Untargeted LC–MS-based metabolomics data (366 metabolites) on rumen fluid from 8 HY and 8 LY cows explored by PCA and OPLS-DA Rumen fluid of HY cows showed higher 3-hydroxyanthranilic acid, carboxylic acids, palmitic acid, and lower gentisic acid, caprylic acid, and myristic acid compared to LY cows Zhang et al., ( 2020d ) Residual feed intake To examine differences in milk metabolomes of high- and low RFI cows and identify biomarkers to predict RFI models in lactating Holstein cows Targeted DI/LC–MS/MS-based metabolomics data (118 metabolites) on milk from 20 high and 20 low RFI cows explored by PCA and OPLS-DA High RFI cows showed higher acylcarnitines (short-, medium-, and long-chain), lysophosphatidylcholines, valine, methionine, histidine, tryptophan, phenylalanine, arginine, lysine, proline, citrulline, ornithine, serine, alanine, glutamine, methylhistidine and pyruvic acid and lower citrate than low RFI cows. Strong RFI predictors in early lactation were decanoylcarnitine (C10), fumaric acid, citric acid, lysoPC a C18:2, in mid-lactation dodecanoylcarnitine (C12) and in late lactation phenylalanine and valine Hailemariam et al., ( 2023 ) Different forage treatment To characterize biomarkers and pathways for different forage treatment Targeted GC–MS-based metabolomics data (31 metabolites) on urine from 8 cows fed alfalfa hay and 8 cows fed corn stover explored by PCA and PLS-DA Increased hippuric acid and N-methyl-glutamic in cow urine are biomarkers for discriminating CS and AH diets Sun et al., ( 2016 ) To examine the impact of high, medium, and no pasture diets on milk metabolome throughout lactation 1 H NMR-based metabolomics data (43 metabolites) on milk from 27 cows with high-, 27 cows with medium- and 27 cows with no-pasture in their diet explored by PLS-DA Increased milk hippurate and dimethyl sulfone are potential biomarkers for pasture-based systems Rojas-Gómez et al., ( 2025 ) Fermented soybean meal To reveal the metabolic responses of dairy cows to the replacement of SBM with FSBM LC–MS-based metabolomics data (913 metabolites) on rumen fluid, blood, milk and urine from 6 SBM-fed and 6 FSBM-fed cows explored by PCA, OPLS-DA and ANOVA Replacing SBM with FSBM increased milk medium-chain fatty acids (i.e., C13:0, C14:0, C14:1, and C16:0) and decreased long-chain fatty acids (i.e., C17:0, C18:0, C18:1n9c, and C20:0); increased blood urea nitrogen, lactic acid, and uric acid, and decreased glucose Wang et al., ( 2022b ) Grain diet concentration To investigate the effect of the transition from a low to high grain diet on the rumen environment The combination of 1 H NMR (untargeted), GC–MS, Direct Flow Injection MS/MS (targeted)-based metabolomics data (93 metabolites) on ruminal fluid from 8 dairy cows fed varied amounts of barley grain (0, 15, 30, and 45%s of diet dry matter), explored by PCA and PLS-DA High-grain diets (> 30%) led to increased concentrations of various toxic, inflammatory, and exogeneous substances (e.g., putrescine, methylamines, ethanolamine, and short-chain fatty acids) in rumen fluid Saleem et al., ( 2012 ) ANOVA analysis of variance, CM clinical mastitis, DI/LC–MS/MS direct infusion chromatography-tandem mass spectrometry, FSBM fermented soybean meal, FIA-MS/MS flow injection analysis-tandem mass spectrometry, GC–MS gas chromatography-mass spectrometry, HB buffaloes in heat stress, HH Holstein cows in heat stress experiment, HS heat stress, HY high yield, JMY Jersey × Maiwa yak crossbred cattle, LC–MS liquid chromatography-mass spectrometry, LY low yield, lysoPC a lysophosphatidylcholine acyl, NMR nuclear magnetic resonance, NUE nitrogen utilization efficiency, OPLS-DA orthogonal partial least squares discriminant analysis, PCA principle component analysis, PLS-DA partial least squares discriminant analysis, RFI residual feed intake, SBM soybean meal, SXC Simmental × Xuanhan yellow cattle crossbred cattle, TN thermal-neutral, XHC Xuanhan yellow cattle Summary of selected metabolomics studies investigating metabolic alterations associated with physiological status in dairy cows ANOVA analysis of variance, CM clinical mastitis, DI/LC–MS/MS direct infusion chromatography-tandem mass spectrometry, FSBM fermented soybean meal, FIA-MS/MS flow injection analysis-tandem mass spectrometry, GC–MS gas chromatography-mass spectrometry, HB buffaloes in heat stress, HH Holstein cows in heat stress experiment, HS heat stress, HY high yield, JMY Jersey × Maiwa yak crossbred cattle, LC–MS liquid chromatography-mass spectrometry, LY low yield, lysoPC a lysophosphatidylcholine acyl, NMR nuclear magnetic resonance, NUE nitrogen utilization efficiency, OPLS-DA orthogonal partial least squares discriminant analysis, PCA principle component analysis, PLS-DA partial least squares discriminant analysis, RFI residual feed intake, SBM soybean meal, SXC Simmental × Xuanhan yellow cattle crossbred cattle, TN thermal-neutral, XHC Xuanhan yellow cattle As previously mentioned, the transition period in cows extends from late pregnancy to early lactation. During this time, dairy cows undergo major metabolic adaptations to meet the high energy demands of milk synthesis (Westhoff et al., 2024 ). These changes significantly affect carbohydrate, lipid and protein metabolism during the periparturient period (Yogeshpriya et al., 2024 ). Reflecting these shifts, a targeted LC–MS-based serum metabolomics study revealed notable changes in the metabolic profile across five time points (i.e., 42 and 10 days before calving and 3, 21 and 100 days after calving). Notably, PC aa C34:2, PC aa C36:2, and PV ae C34:3 were consistently increased throughout the studied period, indicating dynamic metabolic adaptations during the transition period (Kenéz et al., 2016 ). In particular, the period following parturition and onset of lactation is metabolically challenging for dairy cows. The metabolic status during this stage influences milk production and sustainability, as it is linked to mammary gland development and the coordinated regulation of systemic metabolism and physiology. To characterize the metabolic differences between lactating and non-lactating cows, Sun et al. ( 2017 ) conducted a GC-TOF–MS-based metabolomics study on four biofluids (rumen fluid, serum, milk and urine) and mammary gland tissue from 6 mid-lactation and 6 non-lactating Holstein cows. Creatine was identified as a key metabolite explaining biological variation across the biofluids, with potential associations to lactation performance and health in dairy cows. As an important component of the phosphocreatine energy shuttle and intermediate metabolite in the energy reactions, creatine supports fast adenosine triphosphate (ATP) regeneration in tissues with high energy demand (Wyss & Kaddurah-Daouk, 2000 ). In mammary gland tissue, four metabolites, including lactobionic acid, citric acid, orotic acid and oxamide, were increased in cows during lactation (Sun et al., 2017 ). However, since this comparison was cross-sectional, the findings do not provide temporal information on when these metabolic shifts occur or how metabolite levels change during the transition from lactation to non-lactation. A GC–MS-based serum and urine metabolomics study investigated three dairy cattle breeds (Simmental crossbred cattle-yaks and Xuanhan yellow cattle) in terms of their thermotolerance. The animals included in the study were in mid-lactation and under similar feeding and housing conditions during summer heat exposure. The results demonstrated that Xuanhan Yellow cattle exhibited elevated serum levels of pyruvic acid, lactic acid, and tocopherols, suggesting enhanced glycolytic activity and reduced oxidative stress. These metabolic traits suggest that Xuanhan Yellow breed has superior adaptability to hot and humid environments compared to the other two (Liao et al., 2018 ). In a similar study, Gu et al. ( 2018 ) used LC − MRM-MS to compare the serum metabolome of multiparous Holstein cows to Nili-Ravi × Murrah × Local crossbred buffaloes with similar lactation and under the same feeding. They found that dairy buffaloes exhibit greater thermotolerance, which was linked to increased levels of ketogenic and gluconeogenic amino acids (isoleucine, leucine, valine, lysine, and tryptophan) and linoleic acid and betaine. Across both studies, the metabolites associated with thermotolerant breeds were primarily linked to enhanced glycolytic activity and antioxidant capacity. These include organic acids like pyruvic acid, lactic acid, and the branched-chain amino acids and antioxidant compounds such as tocopherols, linoleic acid and betaine. Together, these metabolites demonstrate potential as biomarkers for assessing thermotolerance in cows. Beyond disease and individual variability assessment, metabolomics has also been used to explore key performance traits in dairy cows, including nitrogen utilization efficiency, milk yield and residual feed intake. These traits are related to the cows’ metabolic activity and overall productivity. For instance, an LC–MS/MS-based plasma metabolomics study identified five differential metabolites, i.e. sucrose, a monoacylglyceride, 2-amino-6-hydroxyhexanoic acid, L-glutamine and L-arginine, related to more efficient nitrogen utilization in dairy cows (Li et al., 2024a ). Similarly, LC–MS-based ruminal metabolite profiling differentiated high- and low-milk-yielding dairy cow groups based on 54 metabolites, with 24 significantly increased in the low-yield group (e.g. 3-hydroxyanthranilic acid, L-isoleucine, L-valine, L-tyrosine, uracil, thymine, cytosine and palmitic acid) and 30, including gentisic acid, caprylic acid and myristic acid, in the high-yield group (Zhang et al., 2020 c). In terms of feed efficiency, DI/LC–MS/MS-based milk metabolomics study identified stage-specific metabolic differences between high- and low-residual feed intake (RFI) dairy cows across early, mid and late lactation. High RFI cows had increased levels of medium- and long-chain acylcarnitines and lysophosphatidylcholines in early lactation and increased levels of short-chain acylcarnitines, phosphatidylcholines and amino acids, such as valine, methionine, tryptophan and histidine, in mid lactation. Late lactation was mainly characterized by increased levels of amino acids in high RFI cows. Specific compounds including decanoylcarnitine, dodecenoylcarnitine and phenylalanine were identified via ROC analysis as strong predictive biomarkers for feed efficiency. These findings support the potential of metabolites in guiding the selection of more productive and resource efficient dairy cows (Hailemariam et al., 2023 ). Metabolomics has also been applied to evaluate the effects of dietary intervention on cow performance. This was demonstrated in a study by Wang et al. ( 2022b ), who used LC–MS to analyze the metabolic profile of rumen fluid, blood, milk, and urine. They found that replacing soybean meal with fermented soybean meal significantly altered the metabolic profiles across all four biofluids. In milk, levels of medium chain fatty acids (C13:0, C14:1, and C16:0) increased, whereas levels of certain long-chain fatty acids (C17:0, C18:0, C18:1n9c, and C20:0) decreased. In addition, cows fed fermented soybean meal exhibited higher blood urea and lactic acid but lower glucose levels, compared to those fed conventional soybean meal Wang et al. ( 2022b ). In another study, Sun et al., ( 2016 ) used GC-TOF–MS to analyze urine samples from cows fed alfalfa hay versus corn stover diets and found significantly increased levels of hippuric acid in cows fed the corn stover and N-methyl-glutamic acid in cows fed the alfafa hay diet. In other feeding trial, 1 H NMR-based milk metabolomics was used to study the effect of varying proportions of pasture (high, medium, no pasture) in dairy cows. The results showed that hippuric acid was 1.30 and 0.89 fold higher in cows fed high and medium pasture, respectively, compared to non-pasture-derived milks (Rojas-Gómez et al., 2025 ). Saleem et al. ( 2012 ) showed the effect of high-grain diets on rumen health using 1 H NMR, GC–MS, and DFI-MS/MS, metabolomics. They found elevated levels of toxic, inflammatory, and non-physiological compounds in the rumen fluid of cows fed diets with over 30% grain, highlighting the value of multi-platform metabolomics in understanding the impact of diet on cow metabolism. Parity, defined as the number of times a cow has given birth, is closely related to cow performance, particularly in terms of production traits (Liu et al., 2024a ). As the performance of cows is affected by their metabolic state, it is essential to elucidate the metabolic alterations associated with parity. Through untargeted LC–MS/MS-based metabolomics, Liu et al. ( 2024b ) investigated metabolic profiles of serum and milk in Holstein cows from their first to fourth parity. The results showed that serum levels of histidine and glutamate were the lowest in first-parity cows and highest in third-parity cows, while glycine was significantly increased in the serum of fourth-parity cows. In addition, milk from first-parity cows showed higher levels of L-glutamate, citric acid, choline, urea and inositol, compared to higher-parity cows, while levels of citric acid and cholesterol decreased with increasing parity. These changes indicate a change in metabolic activity across parities, with younger cows exhibiting more biosynthetic and energy metabolic activity in milk. Many of the metabolite alterations observed in the studies investigating the effect of parity, feeding interventions, feeding efficiency, individual variation, pregnancy, transition and lactation overlap with those observed for diseases. For example, amino acids such as phenylalanine, glutamine, leucine and isoleucine shift as response to mastitis, ketosis, lameness, BRD, DA and residual feed intake. In addition, organic acids including hippurate, and lactate are biomarkers of pasture feeding and mastitis, acidosis, BRD and DA, respectively. Glucose has been found to shift in mastitis, ketosis, hypocalcemia, endometriosis and feeding. These findings show that it is challenging to isolate the disease-specific factors affecting the cow metabolome. Therefore, showing the need for experimental designs controlling the confounding factors.

Bovines, including domesticated cattle such as dairy cows, have been vital sources of high-quality animal protein for humans since ancient times. Dairy cows are valued for their milk, which is used to produce a wide variety of dairy products, from infant formula to cheese. Their role in human nutrition remains essential despite the recent green food transition initiatives, where plant-based foods are advocated as a more sustainable and growing part of human nutrition (Faber et al., 2020 ). Livestock is mentioned as the leading source of methane emissions in the global agriculture and waste sector (Stavert et al., 2022 ). However, there is a lack of comprehensive life cycle assessment (LCA) data quantifying the dairy industry’s contribution to global methane emissions, particularly in comparison to other industrial sources of greenhouse gases (Cruz-Rivero et al., 2025 ) Today, cows remain essential to feeding the global population, highlighting the need to optimize feeding practices (e.g. shifting from grain to grass and silage-based feeding), improve animal welfare, enhance milk and meat production while minimizing their environmental footprint and diseases that cause significant losses in the dairy farms. Several confounding factors influence milk yield, quality, processing performance, dairy cow health and reproductive performance, including their ability to successfully calve without developing health issues. Optimal nutrition and feeding management are crucial, as a diet balanced in carbohydrates, proteins, fats, minerals (notably calcium, phosphorus and magnesium), and vitamins (A, D and E), supports high milk production, improves milk composition and helps to prevent metabolic disorders like ketosis, acidosis, hypocalcemia, and cow health around calving ( Caixeta & Omontese, 2021 ; Tsermoula et al., 2025 ; Tufarelli et al., 2024 ). Consistent feeding schedules and high-quality forage improve rumen health, promoting efficient nutrient utilization and sustained milk yield and quality ( Eberly et al., 2023 ; Kmicikewycz et al., 2015 ). Over the past century, the average milk yield per dairy cow has substantially increased, primarily due to advances in breeding strategies and the adoption of efficient farming practices. While these improvements have enhanced reproductive performance, milk yield and quality, they have also been associated with a higher incidence of metabolic disorders, particularly around the calving period. Such conditions contribute to animal welfare concerns, reduced productive lifespan, and significant economic losses (Kappes et al., 2023 ). Metabolomics has emerged over the past two decades as a powerful tool for investigating the complex biochemical changes underlying disease pathophysiology and for assessing the overall health status of dairy cows through comprehensive metabolic profiling of cows’ biofluids (Overton et al., 2017 ). Metabolomics has emerged as an efficient approach for investigating dairy cow health, providing insights into disease etiology and progression, while facilitating early detection of metabolic disorders and diseases. Metabolomics research employs sensitive, selective and high-throughput analytical platforms such as nuclear magnetic resonance (NMR) spectroscopy, gas chromatography-mass spectrometry (GC–MS), and liquid chromatography-mass spectrometry (LC–MS). These analytical platforms generate rich yet complex metabolomics data that require dedicated pre-processing (Johnsen et al., 2017 ; Khakimov et al., 2020 ; Tsermoula et al., 2024 ) to transform raw instrumental outputs into informative metabolite tables suitable for downstream multivariate analysis. Through the integration of analytical chemistry and data analysis, metabolomics enables the profiling of metabolites in response to physiological stress, nutritional imbalances or pathological conditions (Li et al., 2024b ). In practice, metabolomics studies in dairy cattle have utilized plasma, serum, milk, urine, feces and ruminal fluid to investigate disease conditions and physiological states of dairy cows (Kim et al., 2021 ). These studies have contributed to the identification of biomarkers that indicate disease presence, progression or prognosis, as well as to assess the metabolic status of cows. Although metabolomics allows early detection of diseases in cows, it is not at the state to be implemented on farming sites as a diagnostic method due to the complexity, time and cost of the analytical techniques involved. However, the power of metabolomics to comprehensively map the biochemical changes in various biofluids makes it a promising approach for better understanding metabolic changes occurring in dairy cows due to a disease or other stresses (Goldansaz et al., 2017 ). Studies in cow metabolomics have generated data showing how changes in metabolites detected can be utilized to investigate metabolic disorders and diseases. These include papers demonstrating that milk from cows with subclinical mastitis, has altered levels of lactate, citrate, and short-chain fatty acids (Tong et al., 2019 ; Zhu et al., 2021 ). Similarly, cows with lameness have showed distinct blood serum metabolic profiles linked to inflammation and lipid metabolism, supporting the identification of affected animals before symptoms became severe (Dervishi et al., 2020 ). In ketosis, elevated levels of β -hydroxybutyrate and acetone in plasma and urine were detected days before clinical signs appeared, enabling timely nutritional interventions (Zhang et al., 2013 , 2021a , 2021b ). Metabolomic profiling has also revealed biomarkers for acidosis (e.g., increased lactate, histamine in ruminal liquid) (Mao et al., 2016 ; Saleem et al., 2012 ) and hypocalcemia (e.g., reduced serum calcium, altered vitamin D and amino acids), improving preventive strategies (Wilkinson et al., 2020 ). Overall, metabolomics enables the assessment of biomarkers that support early disease detection in cattle. This, in turn, can facilitate the development of practical on-farm tests for identifying affected animals before symptoms become severe and support the development of cow management strategies. This review synthesizes the analysis of nearly 100 publications that report metabolic alterations associated with common diseases and metabolic disorders in cows. It discusses metabolic alterations associated with mastitis, lameness and metabolic disorders such as acidosis, ketosis and hypercalcemia. We also highlight disease-specific metabolic shifts and attempt to correlate these metabolic variations with underlying etiology, pathological, and pathophysiological mechanisms. Additionally, the review summarizes metabolomics studies aimed at identifying potential biomarkers for early disease diagnosis. Lastly, we discuss metabolic fluctuations in dairy cows associated with pregnancy, the transition period, lactation, parity, and other factors that often confound the interpretation of diseases-related metabolic changes.