Insulin therapy in very-low-birth-weight infants.

To the Editor: Beardsall et al. (Oct. 30 issue)1 report the results of the Neonatal Insulin Replacement Therapy in Europe (NIRTURE) trial, which investigated the effects of a fixed-dose insulin (0.05 U per kilogram of body weight per hour) and glucose (20% dextrose) infusion as compared with usual care in very-low-birth-weight infants in the neonatal intensive care unit (ICU). Beardsall and colleagues, as well as Kashyap and Polin, in their accompanying editorial,2 interpret the negative results of this study as evidence against blood glucose control in premature babies. We do not agree with this interpretation. Pronounced hyperglycemia was present in the two study groups (average glucose level, 6.2 mmol per liter [112 mg per deciliter], despite insulin infusion, in the intervention group, and 6.7 mmol per liter [121 mg per deciliter] in the control group), as compared with the much lower normal (fasting) glucose levels in healthy newborns who are less than 4 weeks old (1.7 to 3.3 mmol per liter [31 to 59 mg per deciliter]) or in infants who are 4 weeks to 1 year old (2.2 to 5.0 mmol per liter [40 to 90 mg per deciliter]). Hence, the study intervention actually increased the glucose load and concomitant hyperinsulinemia in the presence of hyperglycemia, a combination that has been shown to be deleterious in patients as well as in animal models of critical illness.3-6 In contrast, achieving strict (ageadjusted) normoglycemia throughout an ICU stay, both in adults and in pediatric patients, has been shown to lower mortality in the ICU and to prevent severe infections and organ damage.3,5,6