Objective To determine the incidence of hyperglycaemia in VLBW infants and to look at the factors which contribute to this.

Methods 29 VLBW <1250 g infants, were included in this retrospective analysis during a 6 month period. Outborn infants were excluded. Infants were divided into two groups, those with hyperglycaemia as defined by a glucose value of >12 mmol/l on two consecutive occasions and those with normoglycaemia. Results were compared using Student’s t-tests.

Results 16 (41%) infants had glucose levels >12 mmol/l on two consecutive occasions. Day 2 amino acid intake was significantly less in the hyperglycaemic group (1.2 vs 1.8 g/kg/d, p = 0.002). Energy intake on day 2 was less in the hyperglycaemic group (38 vs 52 kCals/kg/d, p = 0.0001). The hyperglycaemic group were in a protein catabolic state for longer (2.6 vs 3.5 days, p = 0.03). Glucose infusion rates during the first week of life were lower in the hyperglycaemic group (7.2 vs 8.3 mg/kg/min, p = 0.002). Mean triglyceride levels were higher in the hyperglycaemic group (2.65 vs 2 mmol/l, p = 0.027), and overall lipid intake in the first week was lower in hyperglycaemic group (1.69 vs 1.93 g/kg/d, p = 0.022).

Conclusion The exceptionally high rate of hyperglycaemia (41%) in this population would suggest that the current definition may not be appropriate. Smaller and more immature infants are more likely to be hyperglycaemic and hyperlipidaemic according to current accepted reference ranges which limits advancing PN at this critical time. We believe that the current reference ranges for this gestational age and birth weight population need to be re-evaluated.