In vivo insulin resistance during nonlethal hypermetabolic sepsis

The present study was performed to determine whether hypermetabolic sepsis alters peripheral and hepatic insulin sensitivity and/or responsiveness. Nonlethal sepsis was produced in chronically catheterized conscious rats by repeated subcutaneous injections of live Escherichia coli. Basal glucose metabolism was determined using a primed-constant infusion of [3-3H]glucose initiated 20 hr after the first injection of bacteria. Thereafter, in vivo insulin action was assessed using the euglycemic hyperinsulinemic clamp technique. Insulin was infused at various rates in separate groups of animals for 3 hr to produce steady-state insulin levels of approximately 60, 120, 400, 2,500, and 25,000 microU/ml, and euglycemia was maintained by varying the glucose infusion rate. The sepsis-induced hyperglucagonemia was not significantly altered by the infusion of insulin and glucose. In septic rats, the dose-response curve for the insulin-induced increment in glucose utilization was shifted downward and to the right. As a result, septic rats showed a twofold increase in the ED50 value (380 vs. 190 microU/ml) and a 50% reduction in the maximal responsiveness compared with control animals, indicating peripheral insulin resistance. Septic and nonseptic animals, however, had a similar reduction in the endogenous glucose production rate as the plasma insulin concentration was increased, suggesting that there was no hepatic insulin resistance. The plasma lactate concentration increased in a dose-dependent manner in both septic and nonseptic rats as the plasma insulin concentration was raised. However, the increment in steady-state lactate concentration was consistently higher (75-220%) in septic animals at each insulin infusion rate. These results indicate that nonlethal hypermetabolic sepsis in the rat is associated with peripheral insulin resistance.