Editor—The study by Stenninger and colleagues showed that a blood glucose concentration of less than 1.5 mmol/l is associated with long term neurological dysfunction.1 I am interested to know why the authors chose to define hypoglycaemia thus. They cited our 1988 study,2 saying “that many neonatal units diagnose neonatal hypoglycaemia at concentrationsexceeding (my italics) 1.5 mmol/l.” Shouldn’t the word have been less than rather than exceeding?

The authors may not be aware of our 1996 study,3 showing a significant change in the definition of hypoglycaemia among paediatricians and in neonatal textbooks published between 1986 and 1992, compared with that used between 1965 and 1986. In 1992 among the 420 neonatologists in the UK who would maintain blood glucose concentrations of >2 mmol/l, 78% said they would do so for term babies and 87% for preterm babies, compared with 34% and 22%, respectively, cited in our 1988 study.2 Furthermore, in 1992 the percentage of paediatricians who preferred to maintain blood glucose concentrations of 2.6 mmol/l or more was three times as high as that cited in 1986.

In view of the epidemiological4 and neurophysiological data5 showing adverse effects associated with a blood glucose concentration of less than 2.6 mmol/l, it would have been useful to know what results would have been obtained had the authors compared the babies with blood glucose at this level with those babies with a blood glucose of ⩾ 2.6 mmol/l, and whether any of the babies had recurrent hypoglycaemia.

Would the authors care to recommend what they would consider a safe blood glucose concentration for neonates born to diabetic mothers?