Can ultrasound fetal biometry predict fetal hyperinsulinaemia at delivery in pregnancy complicated by maternal diabetes?

OBJECTIVE: The aim of this study was to determine whether there is an association between ultrasound fetal biometry and amniotic fluid insulin levels at delivery in women with pre-existing diabetes or impaired glucose tolerance in pregnancy. STUDY DESIGN: This retrospective cohort study identified 93 women who had amniotic fluid insulin levels measured at time of delivery. Standardised estimated fetal weight and fetal growth velocity were calculated from serial third trimester fetal ultrasound measurements. RESULTS: Women with pre-existing diabetes had significantly greater mean growth velocity [1.39 (95% CI: 0.43-2.23) versus 0.39 (95% CI: -01.7-0.95); p=0.04], significantly greater mean estimated fetal weight (EFW) Z score prior to delivery [2.36 (95% CI: 1.82-2.9) versus 1.38 (95% CI: 1.02-1.74); p=0.002] and greater mean birthweight centile [82 (95% CI: 0.74-0.89) versus 67 (95% CI: 58-76); p=0.02] than those with GDM/IGT. Amniotic fluid insulin levels demonstrated a similar significant difference between the pre-existing and GDM/IGT groups [20.5 (95% CI: 12.9-28.1) versus 8.5 (95% CI: 5.4-11.7); p=0.001]. An association between fetal growth and size and amniotic fluid insulin was observed in women with pre-existing diabetes. Positive likelihood ratios were 1.67 and 2.08, respectively, for the prediction of liquor insulin greater than the 95th centile in women with pre-existing diabetes. CONCLUSION: Ultrasound measures of fetal size and growth used in this study are not sufficiently accurate to predict those infants likely to be at risk from the adverse effects of fetal hyperinsulinaemia.