NELSON, ROGER L.

We reviewed the records of pregnant diabetic women attended at Mayo Clinic during the years 1950–79 and matched each nonaborted diabetic pregnancy with two control pregnancies. Abortions occurred in 49 of 277 diabetic pregnancies (17.7%). The relative risk of other adverse outcomes of pregnancy (± 95% confidence interval) in comparison with controls was as follows: perinatal mortality, 7.0 (3.6–13.8); major congenital birth defects, 6.3 (2.8–14.2); and respiratory distress, 9.0 (4.5–17.9). Diabetic ketoacidosis resulted in fetal death in 5 of 18 cases. Patients with mean fasting blood glucose levels of ≤ 140 mg/dl had a perinatal mortality of 81 per 1000 births and an incidence of respiratory distress and congenital birth defects of 18.9% and 9.9%, respectively. Patients with mean blood glucose levels > 140 mg/dl had corresponding rates of 306 per 1000, 20.4%, and 11.2%, respectively. Hydramnios occurred almost exclusively in diabetic women, and its presence was associated with a twofold increase in the frequency of perinatal death. Analysis of diabetic pregnancy outcome by decade showed no appreciable decrease in the occurrence of perinatal mortality, congenital birth defects, or respiratory distress over the 30-yr period.

Glucose homeostasis in healthy subjects is characterized by postmeal glucose increases of about 40 mg/dl, peaks at about 45 min, decreases close to antecibal levels 1 h after the peak, and no spontaneous oscillations until the next meal. Diabetes is characterized by progressive loss of glucose homeostasis from stable to unstable, which is directly proportional to loss of insulin secretory reserve. Degree of instability of diabetes in ambulatory subjects within a 24-h period can be expressed as mean amplitude of glycemic excursion of M-value and between two successive 24-h periods as mean of daily differences of blood glucose. Stable diabetic persons have lower values, which are closest to those of nondiabetic persons, and unstable diabetic persons have higher values. The mean diurnal blood glucose level is a measure of glycemic control. The failure to restore glycemic patterns of diabetic to those of nondiabetic persons is largely due to the failure of subcutaneously administered insulin to mimic the pattern of insulinemia of healthy subjects.