Serum metabolomic profiling for early detection of ketosis in high-yielding dairy cattle using nuclear magnetic resonance spectroscopy

Introduction

The consensus among dairy metabolic disease specialists is that ketosis represents a spectrum rather than a binary state, with subclinical forms detectable at the metabolic level days to weeks before clinical signs emerge [1,2] . This view is supported by the observation that cows with serum BHB concentrations between 1.0 and 1.4 mmol/L, conventionally classified as subclinical ketosis, already show depressed dry matter intake, reduced milk yield, and impaired reproductive performance compared with herd-mates below this threshold [3] . Current diagnostic practice, however, relies almost exclusively on BHB measurement, whether by laboratory enzymatic assay or handheld cow-side meters. While BHB is a sensitive marker for the ketogenic state, it captures only one arm of the metabolic disruption. Ketosis involves simultaneous changes in lipid mobilisation (elevated non-esterified fatty acids and ketone bodies), gluconeogenesis (depressed glucose and glucogenic amino acids), tricarboxylic acid cycle flux (reduced citrate and succinate), and amino acid catabolism (altered glutamine, alanine, and branched-chain amino acid ratios) [4,5] . A single analyte cannot reflect this multi-pathway perturbation. Metabolomics, the comprehensive measurement of small molecules in a biological sample, offers a solution. Nuclear Magnetic Resonance (NMR) spectroscopy is particularly suited to this application because it is quantitative, highly reproducible (inter-assay CV typically below 3%), requires minimal sample preparation, and can simultaneously detect dozens of metabolites in a single acquisition lasting less than 10 minutes [6] . Earlier metabolomic studies of bovine ketosis have used mass spectrometry-based platforms, which provide greater sensitivity but at the cost of lower reproducibility and more complex sample handling [7,8] .

Objectives 1) To characterise the serum metabolomic profile of highyielding dairy cows with clinical ketosis, subclinical ketosis, and healthy metabolic status using 1H-NMR spectroscopy; 2) To develop a PLS-DA classification model for ketosis status based on the multi-metabolite signature; and 3) To identify the metabolites with the greatest discriminatory power for incorporation into a reduced diagnostic panel.

Material and Methods

Material

Results

1H-NMR spectroscopy identified and quantified 32 metabolites across the 192 serum samples. Of these, 18 showed statistically significant differences (FDR-adjusted p<0.05) between healthy and clinically ketotic animals. The most prominently altered metabolites fell into three categories: ketone bodies (BHB, acetone, acetoacetate) and NEFA, which were elevated in ketotic animals; glucose and glucogenic precursors (alanine, glutamine), which were depressed; and TCA cycle intermediates (citrate, succinate), which were also reduced.

Discussion

The PLS-DA model achieved excellent separation between healthy and clinically ketotic cows, with cross-validated predictive metrics (Q-squared = 0.81) indicating robust generalisation. This performance exceeds that of BHB alone (Q-squared = 0.62) and BHB combined with NEFA (Q-squared = 0.68), supporting the hypothesis that multi-metabolite profiling captures disease-relevant information that single analytes miss. The metabolite panel aligns with established biochemistry. The three ketone bodies reflect hepatic ketogenesis from mobilised fatty acids, while NEFA elevation marks the upstream lipolytic drive. The reduction in glucose, alanine, and glutamine indicates impaired gluconeogenesis and increased amino acid catabolism to compensate for the energy deficit. The TCA cycle intermediates citrate and succinate are reduced because oxaloacetate is diverted to gluconeogenesis, slowing the cycle and channelling more acetyl-CoA toward ketogenesis [4,5] . The most clinically relevant finding is the improved classification of subclinical cases. BHB alone correctly identified only 58.3% of subclinical cows, while the sevenmetabolite panel achieved 79.2%. This improvement arises because the multi-analyte signature detects the coordinated metabolic shift simultaneous ketone elevation, glucose depression, and amino acid depletion that characterises the transition from negative energy balance to ketosis, even when BHB alone has not crossed the 1.4 mmol/L clinical threshold [7, 8] .

From a practical standpoint, the translation of NMR-based metabolomics to routine dairy diagnostics faces cost and throughput barriers. A 600 MHz NMR spectrometer is a major capital investment, and sample processing, while rapid per sample, requires trained operators. However, the metabolite panel identified here could be implemented on cheaper, faster analytical platforms such as flow-injection mass spectrometry or multiplexed immunoassays, which are already used in human clinical metabolomics [9] . The seven-metabolite panel, rather than the full 32-metabolite profile, would be a realistic target for such translational development. The recommendation arising from these findings is that the development of multi-analyte ketosis screening panels should be pursued as a priority for dairy herd health. The seven metabolites identified here are biologically coherent, analytically accessible, and collectively provide diagnostic information that no single analyte can match. Looking ahead, three steps would advance this work toward practical implementation. The first is validation in a larger, geographically diverse cohort spanning multiple breeds, management systems, and lactation stages. The second is the development of a cost-effective analytical platform, whether multiplexed immunoassay, portable infrared spectroscopy, or microfluidic chip-based detection, capable of measuring the seven-metabolite panel at the farm level within the economic constraints of commercial dairy operations. The third is a prospective intervention trial testing whether early treatment of cows identified as subclinical by the metabolomic panel (but not by BHB alone) improves milk yield, reproductive performance, and health outcomes sufficiently to justify the cost of expanded screening.

Conclusion

References

1. Herdt TH. Ruminant adaptation to negative energy balance: influences on the etiology of ketosis and fatty liver. Veterinary Clinics of North America: Food Animal Practice. 2000;16(2):215-230.
2. McArt JAA, Nydam DV, Oetzel GR. Epidemiology of subclinical ketosis in early lactation dairy cattle. Journal of Dairy Science. 2012;95(9):5056-5066.
3. Ospina PA, Nydam DV, Stokol T, Overton TR. Associations of elevated nonesterified fatty acids and betahydroxybutyrate concentrations with early lactation reproductive performance and milk production in transition dairy cattle in the northeastern United States. Journal of Dairy Science. 2010;93(4):1596-1603.
4. Drackley JK. Biology of dairy cows during the transition period: the final frontier? Journal of Dairy Science. 1999;82(11):2259-2273.
5. White HM. The role of TCA cycle anaplerosis in ketosis and fatty liver in periparturient dairy cows. Animals. 2015;5(3):793-802.
6. Emwas AH, Roy R, McKay RT, Tenori L, Saccenti E, Gowda GAN, et al. NMR spectroscopy for metabolomics research. Metabolites. 2019;9(7):123.
7. Sun LW, Zhang HY, Wu L, Shu S, Xia C, Xu C, et al. 1Hnuclear magnetic resonance-based plasma metabolic profiling of dairy cows with type I and type II ketosis. Journal of Dairy Science. 2014;97(3):1552-1562.
8. Xu C, Sun LW, Xia C, Zhang HY, Zheng JS, Wang JS. 1H-nuclear magnetic resonance-based plasma metabolic profiling of dairy cows with fatty liver. Asian-Australasian Journal of Animal Sciences. 2016;29(2):219-229.
9. Soininen P, Kangas AJ, Wurtz P, Suna T, Ala-Korpela M. Quantitative serum nuclear magnetic resonance metabolomics in cardiovascular epidemiology and genetics. Circulation: Cardiovascular Genetics. 2015;8(1):192-206.
10. Klein MS, Almstetter MF, Schlamberger G, Nurnberger N, Dettmer K, Oefner PJ, et al. Nuclear magnetic resonance and mass spectrometry-based milk metabolomics in dairy cows during early and late lactation. Journal of Dairy Science. 2010;93(4):1539-1550.
11. Dervishi E, Zhang G, Hailemariam D, Goldansaz SA, Deng Q, Dunn SM, et al. GC-MS metabolomics identifies metabolite alterations that precede subclinical mastitis in the blood of transition dairy cows. Journal of Proteome Research. 2017;16(2):433-446.
12. Zhang H, Wu L, Xu C, Xia C, Sun L, Shu S. Plasma metabolomic profiling of dairy cows affected with ketosis using gas chromatography/mass spectrometry. BMC Veterinary Research. 2013;9:186.
13. Goldansaz SA, Guo AC, Sajed T, Steber MA, Wishart DS. Livestock metabolomics and the livestock metabolome: a systematic review. PLoS ONE. 2017;12(5):e0177675.
14. Jeong JK, Choi IS, Moon SH, Lee SC, Kang HG, Jung YH, et al. Effect of two treatment protocols for ketosis on the resolution, postpartum health, milk yield, and reproductive outcomes of dairy cows. Theriogenology. 2018;106:5359.