Abstract
This study explored the potential of circulatory serum metabolite profiles to increase understanding of the physiology of feed efficiency and identify biomarkers to predict residual feed intake (RFI) in lactating Holsteins. Serum metabolite profiles were compared in high (n = 20) and low RFI (n = 20) cows at early, mid, and late lactation stages. The low RFI cows had decreased (P < 0.05) concentrations of dodecanoylcarnitine, dodecenoylcarnitine, dodecanedioylcarnitine, tetradecanoylcarnitine, succinic acid, trimethylamine N-oxide, betaine, and increased concentrations of p-Hydroxyhippuric acid, hydroxysphingomyeline C16:1, phosphatidylcholine diacyl C40:6, and glutarylcarnitine at early lactation. A similar comparison at mid lactation stage showed altered serum concentrations of 26 metabolites that fall into the categories of acyl carnitines, glycerophospholipids, biogenic amines, amino acids, and organic acids. At late lactation, fewer sets of metabolites were significantly affected by RFI grouping. Receiver operator curve analyses identified p-Hydroxyhippuric acid as the top biomarker at early lactation and acetylornithine at mid and late lactation. Models based on sets of serum metabolites in early, mid, and late lactation stages predicted RFI with a validation coefficient of determination of 0.54, 0.68, and 0.64, respectively. This study demonstrated the potential of circulatory serum metabolites as biomarkers and predictors of RFI in lactating dairy cows.