Abstract
Background and aims: Recent data suggest that the use of insulin to maintain intensive glycaemic control amongst surgical ICU patients can improve morbidity and mortality. The value of this procedure in non-surgical patients is not known. Current insulin therapy for non-surgical patients in many ICUs aims to keep plasma glucose below 9 mmolll. The effect of this insulin therapy protocol on the catabolic response of critical illness, characterised by increased glucose production, increased lipolysis and proteolysis is unknown. Materials and methods: A prospective study was conducted in seven critically ill non-surgical patients (6M:1F, age 64±2.72 years; BMI 24.77 ± 0.77 kg/m2) within 36 hours of their admission to the ICU. Patients with diabetes mellitus, pancreatitis, oral steroid use within 1 month of entering the ICU, or liver disease (LFTs > twice normal range), were excluded. All patients were receiving 20% dextrose intravenously to provide 25kcal.kg-lday-l. Insulin was infused at a variable rate to maintain plasma glucose below 9 mmollL. Glucose production rate (Ra) and rate of uptake (Rd), glycerol Ra (a measure oflipolysis) and leucine Ra (a measure of proteolysis) were measured with a 3 hour primed infusion of [6,6- 2H2]glucose (l70mg, 1.7mg.min-I), [2H5]glycerol (0.15mg/kg, 0.61mg· kg-lhc1) and [l-l3C]leucine (1 mg/kg, 1 mg.kg-lhcl). Steady state sampling was performed at 150 to 180 minutes. Results are compared with fasting values from an age and weight matched healthy control group. All data presented are mean ± SEM. Results: The mean APACHE II score on the day of study was 15.43 ± 1.87. Mean plasma glucose at steady state was 7.95±0.73mmol· L-l. The mean glucose infusion rate was 22.83 ± 0.74 f.lmol.kg-lmin-l. The average insulin infusion rate was 4.31 ±0.73 U.hr-l which achieved plasma insulin concentrations of 655.21 ± 181.38 pmol.L-I. Endogenous glucose Ra was decreased (2.24±3.02f.lmol.kg-lmin-l, p<0.03) and glucose Rd was increased (25.08±3.07f.lmol.kg-lmin-l, P<O.OOOl) compared to fasting glucose Ra/Rd in the control subjects (8.7 ± 0.39 f.lmol.kg-lmin-l). Glycerol Ra and concentration (1.66±0.23f.lmol.kg-lmin-1 and 46.72 ± 12.25 f.lmol.L-1 respectively) were reduced (p<0.02, P=0.07) compared to the control subjects (2.89 ± 0.47 f.lmol.kg-lmin-l and 70.03 ± 8.53 f.lmol.L -1). Plasma NEFA concentrations were very low at 0.10 ±0.03 mmol.L-1 compared to fasting values in the control subjects (0.71 ±0.09mmol.L-I, P<O.OOOl). Leucine Ra was 2.16±0.31 f.lmol.kg-lmin-l, which was not significantly different to fasting control subjects (1.71 ±O.ll f.lmol.kg-lmin-l, P=O.ll). Conclusion: In non-surgical ICU patients, the use of insulin to maintain glycaemic control below 9 mmolll was effective in suppressing glucose production rate, increasing glucose uptake and decreasing lipolysis, but did not suppress proteolysis to levels below that seen in fasting control subjects. If full suppression of proteolysis is to be of benefit in this catabolic group of subjects, then higher doses of insulin will be necessary.