Comparison of glucose concentration and glucose absorption from the GI-tract in pigs in whole blood and in plasma

The present investigation was undertaken to compare glucose absorption from the gastro-intestinal tract quantified in either whole blood or plasma using the arterio-venous differences and portal blood flow measurements. Pigs were surgically modified with catheters in the portal vein and the mesenteric artery and a flow probe around the portal vein to allow assessment of nutrient absorption using Fick's principle. In Exp. 1, six sows (mean LW = 202 ± 28 kg) were fed three diets containing 517, 216 or 225 g/kg DM of starch, respectively. In Exp. 2, six female pigs (mean LW = 59 ± 2 kg) were fed three different diets with similar contents of starch (470–506 g/kg DM). The diets in both studies differed regarding amount and solubility of fibre. Blood samples were collected repeatedly 0–10 h after morning feeding. Glucose was measured in whole blood using a glucometer (Accu-Chek®) and in plasma using a biosensitive electrode (Exp. 1) or a standard colourimetric method (Exp. 2). In general, glucose measured in whole blood was 7–11% lower than in plasma at low glucose levels (3.5–5 mM), whereas the methods agreed well at high glucose levels (10–14 mM). Evaluation of the regression lines between glucose measured in whole blood (x-variable) and in plasma (y-variable) gave the following equations: Y = 0.974 X + 0.55, n = 583 (Exp. 1) and Y = 0.949 X + 0.65, n = 655 (Exp. 2). The net absorption of glucose found in blood was in accordance with that found in plasma: Y = 1.012 X , n = 281 (no intercept, P = 0.79; Exp. 1) and Y = 0.954 X, n = 316 (no intercept, P = 0.97; Exp. 2). It was concluded that glucose measured in blood and plasma were slightly biased whereas net glucose absorption in blood and in plasma were unbiased.