Study On Early Warning System For Ketosis And Milk Fever Of Dairy Cows In Transition Period

Ketosis and milk fever have always been two main metabolic diseases during transition periodthat affect dairy cows production and may cause economical loss and low animal welfare.Therefore, for the purpose of economic and animal welfare, it is especially important to managedairy cows herd metabolic health using the notion of herd health and early warning method. Inthis study, we preliminarily researched early warn of ketosis and milk fever by the method ofcross sectional study, combined with statistically significant analysis, correlation analysis andregression analysis and receiver operating curve (ROC) analysis. It will provide a basis toeffectively monitor and prevent two metabolic diseases in the future.In this study, cross-sectional study was used to collect two intensive dairy farms that ownedmore than250lactating cows, fenced free-stall housing, fed a TMR and possessed a herdmanagement software in Heilongjiang Province. In farm A, all tested cows were randomlyselected from22cows at15d to2d prepartum,31cows at parturition, and40cows at2d to14dpostpartum. In farm B, all tested cows were randomly selected from32cows at15d to2dprepartum,30cows at parturition, and31cows at2d to14d postpartum. Blood samples from allcows were collected by tail vein early in the morning. In this study, plasma concentration ofglucose, BHBA and NEFA was detected to assess the energy balance status and prevalence ofketosis. Plasma Ca concentration was measured to assess the prevalence of hypocalcemia.Plasma concentration of TBIL, DBIL, IBIL, TP, ALB, GLO, ALT and AST was determined toreveal the relationship between ketosis and liver functions and the early warning effect on ketosis.Plasma concentration of Mg, P and K was analyzed to clarify the relationship betweenhypocalcemia and these element and the early warning effect on hypocalcemia.The results for ketosis shown that firs of all incidence of ketosis in herd A was40.00%and inherd B was45.16%. Secondly, Pearson correlation coefficient demonstrated that there was asignificantly negative correlation between ketosis and BCS (R=-0.19, P=0.04), plasma Gluconcentration(R=-0.21, P=0.03), and A/G value (R=-0.21, P=0.03), and a significantly positivecorrelation between ketosis and plasma concentration of NEFA(R=0.45, P=0.00), TBIL (R=0.40,P=0.00), DBIL (R=0.42, P=0.00), IBIL (R=0.37, P=0.00), TP (R=0.26, P=0.01) and GLO(R=0.35, P=0.00), AST activity (R=0.57, P=0.00), AST/ALT value (R=0.49, P=0.00). Finally,BCS, NEFA, Glu and AST were determined by binary regression analysis to early warn risk ofketosis. Risk of ketosis may be predicted by the thresholds that in two farms were>3.37for BCSwith25.0%sensitivity and92.3%specificity. In farm A were>3.37for BCS with25.0%and92.3%,>69.5U/L forAST with100%and46.2%,<4.16mmol/L for Glu with25.0%and23.1%, >0.25mmol/L for NEFA with50.0%and76.9%,>0.32mmol/L for BHBA with75.0%and76.9%, respectively. In farm B were>2.88for BCS with50.0%and70.6%,>68.0U/L for ASTwith57.1%and76.5%,<3.97mmol/L for Glu with64.3%and88.2%,>0.27mmol/L for NEFAwith64.3%and70.6%,>0.43mmol/L for BHBAwith78.6%and94.1%, respectively.The results for hypocalcemia shown that first of all incidence of hypocalcemia in herd A was90.32%and in herd B was86.67%, respectively. Secondly, Pearson correlation coefficientdemonstrated that there was a significant negative correlation between hypocalcemia and parity(R=-0.216, P=0.027), a significant positive correlation between hypocalcemia and plasmaconcentration of P (R=0.748, P=0.000), Mg (R=0.228, P=0.019), and no significant associationbetween hypocalcemia and plasma concentration of K (R=-0.084, P=0.391) concentrations, age(R=-0.112, P=0.252) and BCS (R=-0.056, P=0.570). Finally, plasma P were determined bybinary regression analysis to early warn risk of hypocalcemia. Risk of hypocalcemia may bepredicted by the thresholds that in farm A were <2.11mmol/L for Ca, with100%and88.2%,<1.83mmol/L for P, with75.0%and82.4%, respectively. In farm B were <2.12mmol/L for Ca,with98.7%and88.5%,<1.890mmol/L for P, with98.7%and65.4%, respectively.Conclusions obtained that BCS, AST, TBIL, DBIL, IBIL, TP, GLO, NEFA and Glu in plasmawere closely associated with ketosis, and BCS, AST, NEFA and Glu may be useful parametersfor risk prediction of ketosis. Parity, P and Mg in plasma were closely associated withhypocalcemia, and concentrations of Ca, P in plasma may be a useful parameter for riskprediction of hypocalcemia.