Reference Values

    Based on methodology used for the metabolic profiling reference values can either represent expected values for the equivalent population of healthy animals, or specific threshold (cut point) analyte concentrations that have been statistically associated with a specific disease or collective disease risk. Analyte reference values should represent the population mean (or median if not normally distributed) and variation from a defined population of animals clinically evaluated to be free of disease and other health problems and fed an appropriate diet. Reference values for each metabolite need to be refined to minimize inherent variability due to effects of age, physiologic state, production level, and other cow-specific factors on analyte concentration and improve sensitivity of analyte to environmental (i.e., nutritional) influences. At present, few laboratories have specialized blood analyte reference criteria that are adjusted for age, physiologic state, and time relative to calving effects. Research is currently underway to develop appropriate metabolic profiling reference criteria.

     Threshold or cut point criteria are derived from statistical modeling using logistic regression and calculating odds ratios or relative risk. In this process, the prevalence of a specific disease (i.e., retained placenta, ketosis, metritis, etc.) or any disease event is related to various concentrations of a specific analyte to determine if significant predictive relationships exist. For example, fresh cows with serum BHB concentrations ≥ 12.5 mg/dL (1200 µmol/L) were 8 times more likely to experience a left displaced abomasum. A number of studies have defined disease risk relationships to various blood analyte concentrations.

     Expected analyte concentrations for healthy periparturient mature dairy cows are presented in Tables 3 and 4.  Standards for defining appropriate reference values for metabolic profiling have been suggested. Greater clinical adoption of metabolic profiling testing is predicated upon development of robust reference criteria to improve diagnostic interpretation.

     Serum or plasma concentrations of NEFA and BHB have been the most studied in the periparturient dairy cow. Higher NEFA concentrations in either the close-up dry (≥ 0.4 mEq/L) or fresh (≥ 0.6 mEq/L) period are associated with increased risk for many periparturient diseases.  Prepartum BHB concentrations are not predictive of disease, but postpartum concentrations are very sensitive indicators of disease risk. Subclinical ketosis diagnosis has been defined by BHB concentrations of 12.5 or 14.5 mg/dL (1200 or 1400 μmol/L). However, BHB concentrations of 10 mg/dL (0.96 mmol/L) and greater are associated with increased risk of a cow experiencing some postpartum disease.

     Other blood analytes have also been shown to be predictive of disease risk. Albumin concentrations in the close-up (≤ 3.25 g/dL) and fresh (≤ 3.4 g/dL) periods have been associated with increased disease risk. Low total protein (≤ 6 g/dL) in fresh cows has also been suggested to indicate disease risk concerns, though hypergammaglobulinemia resulting from inflammatory response can confound total protein interpretations. Calculated NEFA-to-Cholesterol ratio was suggested as an index of liver function and indicator of hepatic lipidosis. The ratio is calculated with both measures expressed on a mmol/L basis (mg/dL cholesterol x 0.02586 = mmol/L). Cows with increased NEFA:Cholesterol ratio in the close-up dry (≥ 0.2) and fresh (≥ 0.3) periods were are greater risk for postpartum disease. Other analytes did not show significant disease risk associations, but cows collectively experiencing postpartum disease had lower albumin, UN, glucose, and cholesterol and higher NEFA, BHB, NEFA:Cholesterol ratio, and AST concentrations compared to healthy cows.

Table 1. - Expected range (95% confidence interval) in variouis blood analyte concentrations over the periparturient period for healthy, mature dairy cows. (Close-up dry is defined as -3 to -21 days prior to calving; Fresh cows defined as 3 to 30 days in milk.)

 * Analyte has been shown in one or more studies to be predictive for disease risk.

 Table 2. - Fresh cow mineral concentrations in healthy population and concentrations that are of concern for potential disease risk.

Analyte	Adequate Range	Concern Levels
Calcium*	8.7 - 11.0 mg/dL (2.17 - 2.74 mmol/L)	< 8 mg/dL (<2.0 mmol/L)
Phosphorus	4.5 - 8.0 mg/dL (1.45 - 2.58 mmol/L)	< 3.5 mg/dL (<1.13 mmol/L)
Magnesium*	2.0 - 3.5 mg/dL (0.82 - 1.43 mmol/L)	<1.5 mg/dL (<0.62 mmol/L)
Sodium*	137 - 148 mEq/L	< 137 mEq/L
Potassium*	3.8 - 5.2 mEq/L	< 3.0 or > 5.5 mEq/L
Copper	0.6 - 1.5 mg/mL (9.4 - 23.6 mmol/L)	<0.45 or > 4 mg/ml (<7.1 or >63 mmol/L)
Iron	130 - 250 mg/dL (23.3 - 44.9 mmol/L)	< 130 or > 1800 mg/dL (<23 or > 322 mmol/L)
Zinc	0.8 - 1.4 mg/ml (0.89 - 1.3 mmol/L)	< 0.5 or > 3 mg/ml (<7.6 or > 45.9 mmol/L)
Selenium, serum	70 - 100 ng/ml (0. 89 - 1.3 mmol/L)	< 35 or > 800 ng/ml (<0.44 or >10.1 mmol/L)
Selenium, whole blood	120 - 250 ng/ml (1.5 - 3.2 mmol/L)	<50 or > 1900 ng/ml (<0.63 or > 24 mmol/L)
Serum Vitamin A*	225 - 500 ng/ml	< 150 ng/ml
Serum Vitamin E*	3 - 10 ug/mL	< 3.0 ug/ml
Vitamin E:Cholesterol Ratio	2.5:6.0	<1.5

*Analyte has been shown in one ore more studies to be predictive for disease risk.

Reference:  Anderson, D. E. and Rings M. (2009) Current Veterinary Therapy: Food Animal Practice St. Louis, MO: Saunders Elsevier.

Sometimes labs report values using units that may be different than the ones used in the reference ranges.  Try this unit conversion calculator to get your data into the desired units:

SI Unit Conversion Calculator

 This PDF contains a table of all the conversion factors for transforming conventional US units into SI units and vice versa:

SI/ US Unit Conversion Table

Pertinent References from Bibliography

Metabolic profile testing for Jersey cows in Louisiana: reference values. Roussel, J. D., S. H. Seybt. et al. 1982. Am J Vet Res 43(6): 1075-7.

Metabolic Profile Testing in Dairy Herds: Wrong Answer or Wrong Question? Eicher, R.  Acta Veterinaria Scandinavica 2003, 44(Suppl 1) p28.