Serum unconjugated estriols in insulin-dependent diabetic pregnancies: normative data and clinical relevance

The purpose of this study was to establish normative reference values for serum unconjugated estriol (E3) concentrations in insulin-dependent diabetic pregnancy. There were 1973 samples analyzed in 63 diabetic pregnancies not complicated by fetal distress and/or neonatal asphyxia. These samples were compared with 867 samples collected in 25 pregnancies complicated by fetal distress and/or neonatal asphyxia and 103 samples collected in seven nondiabetic, uncomplicated pregnancies. E3 values were found to be significantly higher in diabetic compared with nondiabetic healthy gravid women, correlating with infant birthweight. The percent change of daily E3 values from the highest previous 3 days mean (maximum mean) was 1 +/- 27% (mean +/- SD). The 95% confidence limit for this percent change indicates that a decrease of 52% or more is theoretically seen in 2.5% of the observations. Using the traditional decrease of 40% currently used in the literature, there were false-negative and false-positive predictions of fetal distress and/or neonatal asphyxia at rates unacceptable for patient management. The use of E3 determinations in combination with nonstress test-contraction stress test did not improve the predictability of this latter test. Serum E3 determinations are of little clinical value in the management of insulin-dependent diabetic pregnant women.