Blood sampling and hemolysis affect concentration of plasma metabolites.

Two experiments were carried out to reveal and quantify plasma metabolites that are sensitive to hemolysis and animal stress due to the blood sampling procedure (vein puncture vs. catheter). In Exp. 1, 48 sows were fed 4 diets either once (0800 h) or twice daily (0800 h and 1500 h) in a crossover design and blood was collected after restraint via vein puncture 1, 4, 11, and 23 h after morning feeding. Plasma samples were categorized as without or with minor or major hemolysis [clear (n = 218), yellow (n = 97), or red (n = 37)] upon centrifugation. Plasma NEFA (P < 0.001) was lower in hemolyzed samples but plasma propionate, caproate, isovalerate (P < 0.001), and isobutyrate (P < 0.05) were higher in hemolyzed samples. Plasma glucose and lactate were the only metabolites that were not affected by hemolysis. Interactions with hemolysis and other fixed effects were not found (P > 0.05). In Exp. 2, a subset of samples from 24 sows fed twice daily in Exp. 1 was combined with data obtained from 30 sows sampled using jugular vein catheters. All sows in Exp. 2 were fed twice daily (0800 h and 1500 h) and blood samples collected repeatedly 1, 4, 11, and 23 h after morning feeding (other conditions were similar as in Exp. 1). Plasma isobutyrate (P < 0.001), NEFA (P < 0.01), and acetate (P < 0.05) were lowered and plasma caproate (P < 0.001), glucose (P < 0.01), lactate, and isovalerate (P < 0.05) were elevated in samples obtained via vein puncture as compared to via vein catheters. Plasma insulin, propionate, and butyrate were not sensitive to the blood sampling procedure. In conclusion, blood sampling procedure and hemolysis affect the measured metabolite concentrations and should be considered or accounted for when comparing results within and between experiments.