Intensive insulin therapy delays the onset of diabetic complications but increases the risk of severe hypoglycemia. In children with type 1 (insulin-dependent) diabetes, hypoglycemia may impair normal brain development and neurophysiologic function. Tight glycemic control in children with diabetes requires extreme caution. About one half of the severe hypoglycemic episodes in children occur during the night, when self-monitoring of blood glucose levels is unlikely. Asymptomatic nocturnal hypoglycemia may cause loss of perception of neurogenic hypoglycemia symptoms, predisposing children to severe hypoglycemia. Beregszàszi and associates evaluated the prevalence of nocturnal hypoglycemia in 150 children with type 1 diabetes.

Patients in the study were under age 16 and had had type 1 diabetes for more than one year. Most of them were receiving conventional twice-daily insulin mixtures, and a small proportion were receiving more intensive therapy by means of three or more daily insulin injections. The study participants were admitted to the hospital, and blood samples were obtained hourly through a catheter for 22 hours starting at 10 p.m. Nurses recorded manifestations of hypoglycemia, such as sweating, tremor and agitation.

The overall prevalence of at least one hypoglycemic event during the 22-hour period of monitoring was 46.7 percent (70 patients). Hypoglycemia was asymptomatic in nearly one half of the children. In 48 of the 70 patients with nocturnal hypoglycemia (68 percent), the initial hypoglycemic episode occurred between 10 p.m. and midnight. Multiple episodes of hypoglycemia were noted in seven patients. The proportion of children reporting at least one hypoglycemic episode during the daytime on the day before admission was significantly higher in the group with nocturnal hypoglycemia (31 percent versus 16 percent in children without nocturnal hypoglycemia), as was the percentage of patients in whom urinary C-peptide was undetectable.

The findings suggest that the risk of nocturnal hypoglycemia is higher in children than in adults with type 1 diabetes. Hypoglycemia began most frequently three to five hours after the evening insulin injection. This is understandable in view of the pharmacokinetics of a mixture of short- and intermediate-acting insulins. Previous studies have shown that risk factors for nocturnal hypoglycemia include lower hemoglobin A_1c levels, a longer duration of diabetes and higher insulin dosages.

Four additional risk factors for nocturnal hypoglycemia in children were identified: younger age, the occurrence of two or more episodes of severe hypoglycemia after the onset of diabetes, more than one episode of hypoglycemia weekly for the preceding month, and daily insulin dosages exceeding 0.85 IU per kg. The authors found that a blood glucose threshold of 93.6 mg per dL (5.2 mmol per L) or less before dinner is highly predictive of the occurrence of subsequent nocturnal hypoglycemia.

The authors note that a snack at around 10 p.m. for all children cannot be advocated, because their study was not designed to test this approach. They state that their data suggest that a snack at 10 p.m. for all children could worsen nocturnal hyperglycemia. Further studies are needed to evaluate potential advantages of a snack at 10 p.m.