Insulin acutely suppresses glucose production by both peripheral and hepatic effects in normal dogs.

To determine whether the predominant effect of insulin in suppressing tracer-determined glucose production (Ra) is hepatic or peripheral, we infused insulin peripherally (PER) and portally (POR) at both low (0.75 pmol.kg-1.min-1) and high physiological rates (2.7 pmol.kg-1.min-1) during euglycemic clamps in normal dogs. We also infused insulin peripherally at one-half these rates (1/2 PER) to match the peripheral insulin levels in POR and thus obtain a selective POR vs. 1/2 PER difference in hepatic insulin levels. At the high-rate insulin infusion, peripheral insulin levels were greatest with PER (PER = 212 +/- 10 pM, n = 5; POR = 119 +/- 5 pM, n = 6; 1/2 PER = 122 +/- 5 pM, n = 6). Calculated hepatic insulin levels were greatest with POR (POR = 227 +/- 13 pM, PER = 206 +/- 19 pM, 1/2 PER = 123 +/- 8 pM). High-dose PER yielded a greater suppression of Ra than POR (79 +/- 18 vs. 56 +/- 6%, P < .001). Ra was only suppressed by 45 +/- 6% with 1/2 PER (P < 0.01 vs. POR on 6 paired experiments). Free fatty acid (FFA) was suppressed by 57 +/- 8% with PER and only by 33 +/- 5 and 37 +/- 2% with POR and 1/2 PER, respectively. The low-dose PER and POR yielded an equal Ra suppression (PER = 46 +/- 9%, POR = 43 +/- 4%). Only 1/2 PER was associated with a lower suppression of Ra (36 +/- 8, P < 0.05 vs. POR). FFA showed similar suppression in all three groups (approximately 25%). Using both insulin infusion rates, the percent Ra suppression per unit difference in peripheral insulin was approximately twofold greater than that per unit difference in hepatic insulin. These results suggest that, during euglycemic clamps without somatostatin in normal dogs, Ra suppression is mediated by both peripheral and hepatic effects of insulin and that peripheral insulin, at least at high physiological infusion rates, is more potent than hepatic insulin in suppressing Ra.