Pregnancy-Specific Glucose Management Index Predicts Preterm Birth and Pre-Eclampsia Superior to HbA1c in Women With Type 1 Diabetes Mellitus

Abstract

Aims

This study aimed to evaluate the predictive accuracy of HbA1c and pregnancy-specific Glucose Management Index (GMI) in forecasting adverse pregnancy outcomes for pregnancies with T1DM.

Materials and Methods

In this pre-specified secondary analysis of the CARNATION study, one hundred pregnancies with T1DM who used continuous glucose monitoring systems (CGMS) and had pregnancy outcomes were included. We compared the predictive performance of HbA1c and GMI in identifying composite adverse pregnancy outcomes (CAPO, including maternal death, pre-eclampsia, miscarriage, preterm birth, neonatal death, large for gestational age, congenital malformations, neonatal hypoglycemia, and admission to the neonatal intensive care unit (NICU)) among them. Log-bin regression and receiver operating characteristic curves were utilised for the analysis.

Results

The participants had a mean diabetes duration of 8.0 ± 6.2 years and experienced HbA1c of 6.1 ± 0.7% and GMI of 6.4 ± 0.6% during pregnancy. Among them, 2 (2.0%) had a pregnancy loss, 51 (51.0%) experienced CAPO, 29 (29.6%) foetuses were admitted to the NICU, 15 (15.3%) had a preterm birth, and 5 (5.1%) were pre-eclampsia. HbA1c and GMI were consistent predictors of NICU admission (AUC 0.72 vs. 0.67, p = 0.508) and CAPO (AUC 0.66 vs. 0.61, p = 0.385). GMI was more advantageous in predicting pre-eclampsia (AUC 0.80 vs. 0.49, p = 0.009) and preterm birth (AUC 0.67 vs. 0.54, p = 0.030) compared to HbA1c.

Conclusions

GMI, a measure reflecting shorter-term blood glucose control, has emerged as a significant biomarker for adverse perinatal outcomes in pregnancies with T1DM. Notably, GMI provides a more pronounced advantage in predicting pre-eclampsia and preterm birth.

Trial Registration

ChiCTR1900025955