Insulin infusion with parenteral nutrition in extremely low birth weight infants with hyperglycemia.

From Nov. 7, 1983, to Nov. 6, 1986, all infants with birth weight less than or equal to 1000 gm admitted to Oregon Health Sciences University who had persistent hyperglycemia and glycosuria were treated with graded insulin infusion while energy intake was increased to at least 100 kcal/kg/day (419 kilojoules/kg/day). The records of these infants were reviewed to define the clinical characteristics of infants likely to develop hyperglycemia and to see whether insulin administration would allow goals for energy intake to be met. There were 76 surviving infants; 34 received insulin and 42 did not. Treated infants were smaller (767 +/- 161 vs 872 +/- 98 gm; p = 0.0004), were more immature (26.8 +/- 1.4 vs 27.7 +/- 2.0 weeks; p = 0.0115), and required mechanical ventilation longer (28 +/- 19 vs 17 +/- 15 days; p = 0.0196). There were no significant differences between the groups at 3, 7, 10, or 14 days for intravenously administered glucose or for total nonprotein energy intake at 3, 7, 10, 14, 28, or 56 days. Treated infants achieved an intake of 100 kcal/kg/day (419 kilojoules/kg/day) at 15 +/- 8 vs 17 +/- 11 days and regained birth weight at 12 +/- 6 vs 13 +/- 6 days (NS). There was no difference in percent change from birth weight at 7, 14, 28, or 56 days. Treated infants had a glucose concentration of 195 +/- 60 mg/dl (10.8 +/- 3.3 mmol/L) while receiving 7.9 +/- 3.0 mg/kg/min (43 +/- 17 mumol/kg/min) of glucose at the start of insulin infusion on days 1 to 14. Insulin was given for 1 to 58 days. The initial dose was 40 to 100 mU/gm of dextrose infused (57 to 142 nmol/mol) and then gradually decreased. Less than 0.5% of blood glucose values were 25 to 40 mg/dl (1.4 to 2.2 mmol/L). We conclude that insulin infusion improves glucose tolerance in extremely low birth weight infants and allows hyperglycemic infants to achieve adequate energy intake similar to that of infants who do not become hyperglycemic.