The objectives of this study were to characterize the epidemiology of subclinical hypocalcemia (SCH) in Holstein dairy cows by assessing the temporal associations of plasma Ca concentrations in the first 4 d in milk (DIM) with the risk of cows being diagnosed with metritis or displaced abomasum (or both), and milk production across the first 15 wk of lactation. A prospective cohort study was conducted in 2 dairy herds in New York State, in which cows had a blood sample collected daily for the first 4 DIM. A total of 396 Holstein cows (137 primiparous and 259 multiparous) were enrolled. Multivariable Poisson regression models were built to evaluate the associations of plasma Ca concentration at each of the 4 d following parturition with the risk of primiparous cows being diagnosed with metritis and multiparous cows being diagnosed with metritis, displaced abomasum, or both. Similarly, generalized linear mixed models were built to evaluate the associations of plasma Ca concentration with milk production across the first 15 wk of lactation. Plasma Ca concentration was assessed on a continuous scale in all models; dichotomization and SCH classification only occurred in the final models if the Ca concentration variable was meaningful by creating an optimized threshold based on receiver operating characteristic curve analysis. Plasma Ca concentration assessed at 1 DIM was not associated with the risk of metritis in primiparous cows, but an association was observed at 2, 3, and 4 DIM (critical thresholds were plasma Ca concentration ≤2.15, 2.10, and 2.15 mmol/L, respectively). Plasma Ca concentration was associated with the risk of metritis or displaced abomasum diagnosis (or both) for 2nd parity animals at 2 DIM (threshold ≤1.97 mmol/L), and at 4 DIM for 3rd and greater lactations (threshold ≤2.20 mmol/L). Reduced plasma Ca concentration was associated with higher milk production when assessed at 1 DIM in primiparous and multiparous cows, and lower milk production when assessed at 4 DIM in multiparous cows only. For primiparous cows, plasma Ca concentration was not associated with lower milk production at any of the DIM assessed. In conclusion, assessments of SCH at the individual cow level must take into account the DIM of Ca concentration measurement and parity of the cow, as the epidemiology of the disorder was demonstrated to be highly dependent on these variables. This study advances the knowledge of the epidemiology of SCH and better establishes thresholds for optimizing SCH diagnosis.

The objectives of this study were to determine the effect of decreasing dietary cation-anion difference [DCAD; (Na + K) - (Cl + S)] of the prepartum diet on aspects of mineral metabolism, energy metabolism, and performance of peripartum dairy cows. Multiparous Holstein cows (n = 89) were enrolled between 38 and 31 d before expected parturition and randomized to treatments in a completely randomized design (restricted to balance for previous 305-d mature equivalent milk production, parity, and body condition score) at 24 d before expected parturition. Treatments consisted of a low-K ration without anion supplementation [CON; n = 30, DCAD = +18.3 mEq/100 g of dry matter (DM)]; partial anion supplementation to a low-K ration (MED; n = 30, DCAD = +5.9 mEq/100 g of DM); and anion supplementation to a low-K ration to reach a targeted average urine pH between 5.5 and 6.0 (LOW; n = 29, DCAD = -7.4 mEq/100 g of DM). Cows were fed a common postpartum diet and data collected through 63 d in milk. Urine pH (CON = 8.22, MED = 7.89, and LOW = 5.96) was affected quadratically by decreasing prepartum DCAD. A linear relationship between urine pH and urine Ca:creatinine ratio was observed (r = -0.81). Plasma Ca concentrations in the postpartum period (d 0 to 14; CON = 2.16, MED = 2.19, and LOW = 2.27 mmol/L) were increased linearly with decreasing prepartum DCAD. A treatment by parity (second vs. third and greater) interaction for postpartum plasma Ca concentration suggested that older cows had the greatest response to the low DCAD diet and older cows fed LOW had decreased prevalence of hypocalcemia after calving. A quadratic effect of decreasing DCAD on prepartum DMI was observed (CON = 13.6, MED = 14.0, and LOW = 13.2 kg/d). Milk production in the first 3 wk postpartum was increased linearly with decreasing DCAD (CON = 40.8, MED = 42.4, and LOW = 43.9 kg/d) and DMI in this period also tended to linearly increase (CON = 20.2, MED = 20.9, and LOW = 21.3 kg/d). Overall, effects on intake and milk yield analyzed over wk 1 to 9 postpartum were not significant. This study demonstrates that feeding lower DCAD diets prepartum improves plasma Ca status in the immediate postpartum period and results in increased DMI and milk production in the 3 wk after parturition. Compared with no anion supplementation or lower levels of anion supplementation, greater improvements were observed with the lower DCAD feeding strategy, in which an average urine pH of 5.5 to 6.0 was targeted.

The objective of this study was to evaluate the association of postpartum plasma Ca concentration with early-lactation disease outcomes, culling within 60 d in milk, pregnancy to first service, and milk production. A total of 1,453 cows from 5 commercial dairy farms in New York State were enrolled in a prospective cohort study from February to November 2015. Blood samples were collected within 12 h of parturition, and plasma was submitted to a diagnostic laboratory for total Ca measurement. Early-lactation disease, reproductive performance, and milk production from Dairy Herd Improvement Association (DHIA) test-day data were compiled from each farm's management software. Multivariable Poisson regression models were built to evaluate the association of plasma Ca with the risks of retained placenta (RP), metritis, displaced abomasum (DA), clinical mastitis, culling within 60 d in milk, and pregnancy to first service. Repeated-measures ANOVA were used to evaluate the association of Ca at parturition with milk production across the first 9 DHIA tests. Herd was considered a random effect in all models. Primiparous cows were modeled separately from multiparous cows if differential responses were observed. Calcium was not associated with the risk of RP, metritis, clinical mastitis, or pregnancy to first service in primiparous or multiparous cows. For multiparous cows only, higher Ca concentration tended to be associated with increased culling within the first 60 d in milk. Multiparous cows with Ca ≤1.85 mmol/L had an increased risk of being diagnosed with a DA compared with cows with Ca >1.85 mmol/L. For the milk production models, Ca was not associated with the amount of milk produced within the first 9 DHIA tests in primiparous cows; however, multiparous cows with Ca ≤1.95 mmol/L produced, on average, 1.1 kg more milk per day across the 9 DHIA tests than their multiparous counterparts with Ca >1.95 mmol/L. Our results indicate that plasma Ca concentration measured within 12 h of parturition is a poor predictor of early-lactation health outcomes. Reduced Ca concentration in the immediate postpartum period was associated with higher milk production in multiparous cows. From these results, we caution that studies attempting to categorize subclinical hypocalcemia based on a single sample in the immediate postpartum period could misclassify the disorder.

The objectives of this study were to (1) compare a test for serum measurement of total Ca (tCa), Mg, and P (VetTest Chemistry Analyzer, IDEXX Laboratories Inc., Westbrook, ME) to reference methods (spectrophotometric assays on a Beckman Coulter 640e automated clinical chemistry analyzer; Beckman Coulter, Brea, CA), (2) determine the relationship between ionized Ca (iCa) and reference method tCa in the immediate postpartum period, and (3) assess the relative value of these blood Ca indices as predictors of neutrophil oxidative burst activity. Samples were collected from multiparous Holstein cows (n = 33) over the first 5 d in milk. A total of 183 samples for objective 1 and 181 samples for objective 2 were available. Neutrophil oxidative burst activity was assessed once between 2 and 5 d in milk (n = 29). Linear regression demonstrated strong relationships between serum tCa, Mg, and P concentrations measured by the VetTest compared with the reference method. Bland Altman analysis indicated that the VetTest values were higher than the reference method by 0.22 mmol/L for tCa, 0.12 mmol/L for Mg, and 0.16 mmol/L for P. Compared with hypocalcemia categorized at ≤2.0 or ≤2.125 mmol/L with the reference method tCa, thresholds for the VetTest measured tCa of ≤2.23 mmol/L (sensitivity = 87%, specificity = 89%) or ≤2.30 mmol/L (sensitivity = 86%, specificity = 96%) could be used. The relationship between whole-blood iCa and reference method serum tCa differed by sampling time point after calving. Compared with identification of hypocalcemia with serum tCa measurements from the reference method (thresholds of ≤2.0 and 2.125 mmol/L), a whole-blood iCa threshold of ≤1.17 mmol/L resulted in the highest combined sensitivities (94 and 82%) and specificities (80 and 94%) at either threshold. Ionized Ca measurements were more consistently related to outcomes of neutrophil oxidative burst activity measured in vitro. The VetTest measurements of serum tCa reliably identified hypocalcemia when thresholds were adjusted to account for the bias of the test. The variation in the relationship between iCa and reference method tCa in the days following parturition suggest that these measures cannot be used interchangeably as indicators of Ca status. The more consistent associations between iCa and in vitro measures of neutrophil function, compared with tCa, indicated that this may be a more sensitive predictor of functional outcomes associated with postpartum Ca status.