Milk fever (MF), also known as clinical hypocalcemia, is a common calcium metabolism disorder in perinatal cows that is characterized by hypocalcemia, muscle weakness, paralysis of the limbs, and depression, with coma occurring in severe cases.Currently, information regarding the overall effects of such treatments on metabolism in cows with MF is scant.The purpose of the research was to study the metabolic profiling of serum samples from cows with MF, and explore other underlying pathological mechanisms of this disease.In our current study,Thirty-two Holstein cows (day 0 postpartum) with parities of three to six, body condition scores of approximately 3.25, and similar milk yields were selected for our study.The cows were maintained in free-stall housing in Heilongjiang, China.Production data were generated using the Dairy Comp 305 dairy management software system (Valley Agricultural Software, Tulare, CA, USA).Based on their clinical presentation and serum calcium levels, 24 healthy cows were assigned to the control group (No MF symptoms and calcium concentration > 2.5 mmol/L), and 8 cows were assigned to the MF group (MF symptoms and calcium concentration < 1.4 mmol/L).All serum samples were placed under the signal acquisition probe of an Avance Ⅲ 500-MHz Digital NMR Spectrometer (Bruker, Ettlingen,Germany), and the NMR spectra were recorded using a Call-Purcell-Meiboom-Gill (CPMG) sequence (90 [τ-180-τ] n-acquisition) with 64 timed scans and a 10-ms total relaxation period (transverse relaxation and longitudinal relaxation) to inhibit the protein signal.The free induction decay (FID) signal was measured following transverse relaxation, and the FID signals were Fourier transformed to obtain the1H NMR spectra.After pre-processing of data,the multivariate pattern recognition, including the principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA) and orthogonal signal correction-partial least square-discriminate analysis (OSC-PLS-DA) were carried out.Then a one-way analysis of variance was performed using the SPSS statistical software (IBM, Armonk, NY, USA).The level of statistical significance was set at P < 0.05 for the identification of differentially expressed metabolites.In the OSCs plot, data points to the left (Group MF) of the vertical axis were negative, and those to the right side (Control group) were positive.In the loading plot, the data points above the horizontal axis were positive, and those under the horizontal axis is negative.Therefore,if a positive or negative shift occurred in the loading plot for a metabolite in Group MF, the metabolite was considered to be decreased or increased, respectively, compared to the level of the metabolite in control group.Nine metabolites in Group MF were present at significantly different levels, compared with those of control group.Glucose, alanine, glycerol,phosphocreatine, and gamma-aminobutyrate (GABA) were decreased, and β-hydroxybutyrate (BHBA), acetone,pyruvate, and lysine were increased in cows with MF.Most of these were sugars and amino acids involved in various energy metabolism pathways.The different metabolites in cows with MF reflected the pathological features of negative energy balance and fat mobilization, suggesting that MF is associated with altered energy metabolism.The1 H-NMR spectroscopy can be used to understand the pathogenesis of MF and identify biomarkers of the disease.