Short communication: Reference limits for blood analytes in Holstein late-pregnant heifers and dry cows: Effects of parity, days relative to calving, and season.

Reference limits for metabolic profiles in Holstein late-pregnant heifers and dry cows were determined considering the effects of parity, days relative to calving, and season. Blood samples were collected from 104 pregnant heifers and 186 dry cows (68 primiparous and 118 pluriparous) from 60 to 10 d before the expected calving date in 31 dairy farms in northeastern Italy. Sampling was performed during summer (182 samples) and the following winter (108 samples). All the animals were judged as clinically healthy at a veterinary visit before sampling. Outliers were removed from data of each blood analyte, and variables that were not normally distributed were log transformed. A mixed model was used to test the fixed effects of parity (late-pregnant heifers, primiparous or pluriparous dry cows), class of days relative to calving (60-41 d, 40-21 d, 20-10 d), season (summer or winter), and the interactions between parity and class of days relative to calving and between parity and season, with farm as random effect. Single general reference limits and 95% confidence intervals were generated for analytes that did not vary according to fixed effects. Whenever a fixed effect included in the model significantly affected a given analyte, specific reference limits and 95% confidence intervals were generated for each of its levels. Albumin, urea, triglycerides, alanine aminotransferase, aspartate aminotransferase, creatinine kinase, conjugated bilirubin, calcium, phosphorus, magnesium, potassium, chloride, zinc, copper, and iron concentrations were not influenced by any of the fixed effects. Total protein, globulins, creatinine, glucose, alkaline phosphatase, gamma glutamyltransferase, lactate dehydrogenase, and sodium plasma concentrations were affected by parity. The class of days relative to calving had a significant effect on the concentrations of total protein, globulins, fatty acids, cholesterol, total bilirubin, and sodium. Season affected plasma concentrations of creatinine, glucose, fatty acids, lactate dehydrogenase, and sodium. Interactions between parity and class of days relative to calving and between parity and season did not significantly affect any of the blood analytes tested. The reference limits and the 95% confidence intervals for blood analytes determined in the study could help dairy practitioners to improve the accuracy of metabolic profile interpretation in Holstein late-pregnant cattle.