Blood parameters predicted from milk spectra are candidate indicator traits of hyperketonemia—A retrospective study in the Italian Holstein population

At the onset of lactation, high-producing dairy cows commonly face a negative energy balance and metabolic disorders, such as hyperketonemia. Blood concentrations of nonesterified fatty acids (NEFA), BHB, cholesterol, glucose, and urea provide valuable information about the metabolic, health, and nutritional status of lactating cows. Milk mid-infrared (MIR) spectroscopy has been successfully used for the prediction of several health traits in the last years, including blood metabolite concentrations, even though the model accuracy is moderate. In fact, MIR-predicted blood parameters are useful for population screening and may be used for selective breeding if they are heritable and genetically variable within a population. In the present study, we estimated h² and genetic correlations of BHB, NEFA, glucose, cholesterol, and urea and assessed their genetic correlations with milk yield and composition traits in the Italian Holstein population using phenotypes of 9,943 cows in 460 herds. Two datasets were considered: early lactation (8,277 test-day records between 5 and 35 DIM) and whole lactation (105,293 test-day records, between 5 and 305 DIM). The h² and genetic variability of blood traits were greater in early than whole lactation, suggesting that there is room to manipulate metabolic disease incidence in the transition period through tailored genetic strategies. Blood BHB was the most heritable trait, regardless of the lactation stage (with h² of 0.13 and 0.08 in early and whole lactation), whereas blood NEFA was the least heritable trait, with h² not significantly different from zero. Blood BHB was positively genetically correlated with NEFA, whereas glucose was negatively correlated with BHB, NEFA, and urea. The milk fat-to-protein ratio was correlated with BHB, NEFA, and cholesterol, whereas a negative correlation was calculated between lactose content and BHB and between SCS and BHB. Estimated breeding values of sires with at least 20 daughters with phenotypes available were extrapolated for a posteriori evaluation of the observed performance. The progeny of the top 5 sires exhibited a lower incidence of hyperketonemia compared with the other cows, with only 2.16% of cows having BHB concentrations above the conventional threshold (1.20 mmol/L). Conversely, the prevalence of hyperketonemia was 5 times higher in the offspring of the bottom 5 bulls (10.55% of cows with BHB above the threshold of 1.20 mmol/L). These findings suggest that, despite the low h² estimates, there is room to identify animals with low or high genetic merit for traits linked to metabolism. Therefore, selection toward healthier and metabolically resistant cows is pursuable, with MIR-predicted blood traits being potential auxiliary traits for selective breeding.