Associations of inflammation measured at 2 days in milk with metrics of milk production, health, and herd removal in multiparous Holstein cows

Most dairy cows experience some degree of inflammatory activation during the early postpartum period as a normal response to parturition and the initiation of lactation. Excessive inflammation during this time has been associated with various disease states, impaired reproductive success, and decreased productivity, but time points appropriate for measurement of inflammatory status are not well established. Our objective in this exploratory observational cohort study was to evaluate the relationships between inflammatory markers, specifically haptoglobin and serum amyloid A, measured at 2 DIM and important metrics of lactational success including milk yield through 10 wk in lactation, rumination and activity time during the periparturient period, risk of dyscalcemia, and time to herd removal. Blood was collected at 2 and 4 DIM from 266 multiparous Holstein cows on 4 commercial farms in New York. Serum samples collected at 2 DIM were analyzed for both haptoglobin and serum amyloid A concentrations, and samples from 4 DIM were analyzed for total serum calcium concentrations (tCa). Cows were identified as dyscalcemic if their tCa at 4 DIM was ≤2.2 mmol/L. Daily milk weights averaged by week through 10 wk in lactation and herd removal records were collected for each cow. A subset of cows (n = 101) had loggers that collected daily rumination times and activities from -14 d relative to parturition to 14 DIM. Linear mixed effects models were used to evaluate the relationships between inflammatory analytes and milk production, rumination time, and activity, and generalized linear models with a quasi-Poisson distribution were used to estimate risk for dyscalcemia. Time to event data for herd removal were modeled using Cox proportional hazards regressions. Confounders including parity, farm, disease status, and calving difficulty, as well as their interactions with analytes and time, were also included in our models where appropriate. Marginal means, risk ratios, and hazard ratios were calculated from their respective models at 2 concentrations of haptoglobin and serum amyloid A, independently, corresponding to low (haptoglobin: 0.4 g/L; serum amyloid A: 14 µg/mL), and high (haptoglobin: 1.1 g/L; serum amyloid A: 33 µg/mL) inflammatory status. Cows with higher inflammatory status produced less milk over the first 9 wk of lactation, ruminated less in the early postpartum period, had decreased activity in the early days following parturition, and had increased risk of dyscalcemia compared with less inflamed cows. Cows with high inflammation were also at a greater hazard of herd removal within 305 DIM. Our data support the hypothesis that more severe postpartum inflammation measured at 2 DIM is associated with reduced lactational success. This evidence suggests that these inflammatory markers might be useful in future diagnostic applications for identifying cows that are at risk of impaired performance in the early postpartum period.