RESUMO
OBJECTIVE: To explore the prevalence of hyperglycemia and the associations between nutritional intakes, hyperglycemia, insulin treatment, and mortality in extremely preterm infants. STUDY DESIGN: Prospectively collected data from the Extremely Preterm Infants in Sweden Study (EXPRESS) was used in this study and included 580 infants born <27 gestational weeks during 2004-2007. Available glucose measurements (n = 9850) as well as insulin treatment and nutritional data were obtained retrospectively from hospital records for the first 28 postnatal days as well as 28- and 70-day mortality data. RESULTS: Daily prevalence of hyperglycemia >180 mg/dL (10 mmol/L) of up to 30% was observed during the first 2 postnatal weeks, followed by a slow decrease in its occurrence thereafter. Generalized additive model analysis showed that increasing parenteral carbohydrate supply with 1 g/kg/day was associated with a 1.6% increase in glucose concentration (P < .001). Hyperglycemia was associated with more than double the 28-day mortality risk (P < .01). In a logistic regression model, insulin treatment was associated with lower 28- and 70-day mortality when given to infants with hyperglycemia irrespective of the duration of the hyperglycemic episode (P < .05). CONCLUSIONS: Hyperglycemia is common in extremely preterm infants throughout the first postnatal month. Glucose infusions seem to have only a minimal impact on glucose concentrations. In the EXPRESS cohort, insulin treatment was associated with lower mortality in infants with hyperglycemia. Current practices of hyperglycemia treatment in extremely preterm infants should be reevaluated and assessed in randomized controlled clinical trials.