Calving body condition score combined with milk test data and rectal tempreture improved the prognostic value of non-invasive markers for infectious diseases in Holestein cows

Abstract

Early lactating dairy cows, are predisposed for calving-related diseases. Early prediction of “at risk” animals increases the likelihood of a successful recovery. In this regard, test-days milk data and body condition score (BCS) at calving have been evaluated as potential indicators of cows at-risk, but results were inconclusive. We hypothesised that a combined use of easily accessible data (BCS at calving, net BCS change over the first 2 weeks of lactation, rectal temperature (RT) and first 2 weeks test-days milk data) could improve the prognostic value. A total of 117 multiparous Holstein cows were selected according to their BCS at calving and divided into two experimental groups: a high BCS (HBCS; BCS ≥ 4.0) and a normal BCS group (NBCS from 3.25 to 3.5). The following data were collected and evaluated for each cow: milk yield and composition at the first two milk test-days (week 1 and 2 postpartum), BCS at calving and at the first two milk test-days and RT measured daily from day 1 (parturition) to 14 postpartum. Cows were furthermore monitored for subclinical mastitis (SCM), metritis and endometritis in the postpartum period, and diseases diagnosed using standardized definitions. The predicting value of variables were evaluated by Receiver Operating Characteristic (ROC) analysis. None of the determined variables provided accuracy individually in prediction of post-partum diseases. Only the mean value of rectal temperature at week 2 (RT2) showed an area under the curve (AUC) higher than 0.60 (AUC = 0.67; P < 0.05); however, with a sensitivity of 0.59 and specificity of 0.81, it cannot be considered a strong predicting marker for endometritis. A combination of markers, which included BCS at calving, net BCS change during first 2 weeks of lactation, milk fat, milk protein, milk fat to protein ratio (FPR) and RT, provided the best accuracy in prediction of SCM and uterine infections. The prediction strength of the combined mode was substantially higher compared to using each of the parameters alone. In conclusion, the combination of easily accessible observational data collected over the first 14 d of lactation, improves the prediction of SCM, metritis and endometritis compared to the prognostic value of the individual markers.