View article

Improvement of genetics of health is sought after by dairy producers, but it is at the same time difficult to achieve as heritability of health traits is low and amount of data collected is limited. Selection for health traits could be enhanced by using data on correlated traits. One such trait is body condition score (BCS), a measure of the fat deposition of a cow, that is used as an indicator of metabolic balance. Health records for 8 traits (mastitis, lameness, cystic ovarian diseases, left displaced abomasum, ketosis, metritis, milk fever and retained placenta) and multiple BCS records per lactation were available for Canadian Holsteins from the province of Quebec. Diseases were recorded by producers and checked for plausibility by veterinarians before being included in the national database. Six of the 8 health traits were grouped into: reproductive disease (cystic ovarian disease, metritis and retained placenta) and metabolic disease (left displaced abomasum, ketosis and milk fever). The relationship between BCS and disease resistance was analyzed by 2 different modelling approaches. The first was a multiple-trait model in which BCS recorded at 5 different time periods in the lactation were considered as different traits. Mastitis and reproductive disease before and after 50 days in milk were also considered as different traits. The second modelling approach was a bivariate (BCS and 1 health trait) longitudinal model.

Overall, there was a positive additive genetic correlation between BCS and resistance for mastitis, lameness and metabolic disease (rg< 0.59). Genetic correlations were negative between BCS and resistance to reproductive disease in the …