Traditional management of mothers with type 1 diabetes stresses good glycemic control during pregnancy to reduce macrosomia and tight control during labor and delivery to minimize maternal and infant hypoglycemia. Taylor and colleagues studied predictors of macrosomia and neonatal hypoglycemia in 107 pregnancies in insulin-dependent mothers attending a specialist clinic in a British medical school.

They collected data on all singleton pregnancies in type 1 diabetic mothers between 1994 and 1999. All mothers were seen monthly by the specialist team until 28 weeks' gestation, then every two weeks until 36 weeks' gestation, and then weekly until delivery. More frequent visits occurred as needed, and a nurse specialist maintained regular telephone contact with each patient. Insulin requirements and glycosylated hemoglobin (HbA_1c) levels were monitored at each visit. Serial ultrasound scans were performed from 19 weeks' gestation. Delivery was planned for 38 to 39 weeks' gestation, and maternal blood glucose was controlled during labor using intravenous glucose, insulin, and potassium infusions. Infants were fed as soon as possible after delivery, and blood glucose levels were measured before the second feeding.

The mothers were 17 to 40 years of age (mean: 28.6 years), and 44 were primigravidas. The mean gestation at entry to the study was 10 weeks. Delivery occurred at a mean gestation of 258 days, with 30 percent before 259 days. The cesarean delivery rate was 49.5 percent, with 19 elective and 34 emergency deliveries. At first visit, the mean HbA_1c was 8 percent, and throughout pregnancy it was 7.2 percent. Insulin requirements remained constant until approximately 20 weeks, then steadily increased until 30 weeks. An optimal insulin regimen was developed for each woman. The mean blood glucose level in labor was 113 mg per dL (6.3 mM per L). In 50 infants, blood glucose levels were less than 45 mg per dL (2.5 mM per L). A significant correlation was found between neonatal hypoglycemia and high maternal glucose levels during labor.

When maternal blood glucose levels during labor were greater than 144 mg per dL (8 mM per L), neonatal hypoglycemia usually occurred. No other measures of maternal glycemic control, including HbA_1c during pregnancy or control levels at any period of pregnancy, were related to neonatal blood sugar levels. Neonatal macrosomia and birth weight were not related to maternal glycemic control at any stage of pregnancy or to maternal insulin requirements.

The authors conclude that neonatal hypoglycemia is related to maternal hyperglycemia during labor but not during pregnancy. The threshold effect appears to be around the upper limit of target ranges in current use 144 mg per dL, and the authors advocate close monitoring of blood glucose levels during labor to avoid the dangers of maternal hypoglycemia (and exhaustion) versus maternal hyperglycemia causing neonatal hypoglycemia. Conversely, tight glycemic control as monitored by HbA_1c levels does not appear to influence macrosomia. They suggest as explanation that HbA_1c reflects the average glycemic control whereas intermittent episodes of hyperglycemia could stimulate fetal hyperinsulinemia, leading to macrosomia. These peaks would not necessarily be detected by HbA_1c monitoring.