Insulin-mediated glucose uptake by individual tissues during sepsis.

Gram-negative hypermetabolic sepsis has been previously reported to produce whole body insulin resistance. The present study was performed to determine in vivo which tissues are responsible for the sepsis-induced decrease in insulin-mediated glucose uptake (IMGU), and whether that decrease was related to a change in regional blood flow. Vascular catheters were placed in rats and sepsis was induced by subcutaneous injections of Escherichia coli. Insulin action was assessed 20 hours after the first injection of bacteria by the combined use of the euglycemic hyperinsulinemic clamp and the tracer 2-deoxyglucose (dGlc) technique. Insulin was infused at various rates in separate groups of septic and nonseptic rats for 3 hours to produce steady-state insulin levels between 70 and 20,000 microU/mL. Rats were injected with [U-14C]-dGlc 140 minutes after the start of the euglycemic hyperinsulinemic clamp for the determination of the glucose metabolic rate (Rg) in selected tissues. The maximal response to insulin was decreased 30% to 40% in the gastrocnemius, and in the red and white quadriceps. The former two muscles also showed a decrease in insulin sensitivity. However, the insulin resistance seen in hindlimb muscles was not evident in all muscles of the body, since IMGU by abdominal muscle, diaphragm, and heart was not impaired by sepsis. The basal Rg by skin, spleen, ileum, and lung was increased by sepsis, and was higher than the insulin-stimulated increases in Rg by these tissues in nonseptic animals. Cardiac output was similar in septic and nonseptic rats and did not change during the infusion of insulin. Under basal conditions, sepsis appeared to redistribute blood flow away from the red quadriceps and skin, and increased flow to the liver (arterial), lung, and small intestine. When plasma insulin levels were elevated, hepatic arterial blood flow was increased, and flow to the red quadriceps and skin was decreased in nonseptic animals. Hyperinsulinemia did not produce any consistent change in regional blood flow in septic animals. The results of this study indicate that a decrease rate of IMGU in muscle is primarily responsible for the whole body insulin resistance seen during hypermetabolic sepsis, and that the impairment of insulin action in skeletal muscle is not dependent on fiber type or to changes in blood flow.