Affiliation:
1. Department of Obstetrics and Gynecology Aalborg University Hospital Aalborg Denmark
2. Steno Diabetes Center North Jutland Aalborg Denmark
3. Department of Clinical Medicine Aalborg University Aalborg Denmark
4. Department of Endocrinology Aalborg University Hospital Aalborg Denmark
5. Department of Obstetrics and Gynecology Aarhus University Hospital Aarhus Denmark
6. Department of Clinical Medicine Aarhus University Aarhus Denmark
Abstract
AbstractObjectiveTo compare ultrasound‐assessed fetal head circumference (HC), abdominal circumference (AC), HC/AC ratio, and estimated fetal weight (EFW) in prediction of large‐for‐gestational‐age (LGA) at birth in pregnancies affected by type 1 (T1DM) and type 2 (T2DM) diabetes.MethodsThis retrospective cohort study included all women with T1DM and T2DM giving birth to singletons between 2010 and 2019 at Aalborg University Hospital, Denmark. Ultrasound scans were performed at 16, 20, 28 and 34 weeks of pregnancy. LGA was defined as birth weight deviation of 15% or greater from the expected for gestational age (≥90th centile). Prediction of LGA was assessed by logistic regression adjusted for maternal characteristics and glycated hemoglobin (HbA1c) and area under the receiver operating characteristics curve (AUC).ResultsAmong 180 T1DM pregnancies, 118 (66%) had an LGA neonate at birth. At 28 weeks of pregnancy, they were predicted with AUCHC/AC = 0.67, AUCAC = 0.85, and AUCEFW = 0.86. The multivariate analysis did not improve the predictive performance of the HC/AC ratio or AC. Among 87 T2DM pregnancies, 36 (41%) had an LGA neonate at birth. At 28 weeks, they were predicted with AUCHC/AC = 0.73, AUCAC = 0.83, and AUCEFW = 0.87. In T2DM, the multivariate analysis significantly improved the predictive performance for both HC/AC ratio and AC from 20 weeks of pregnancy.ConclusionIn T1DM and T2DM pregnancies, LGA is characterized by a general fetal overgrowth including both AC and HC. Therefore, AC and EFW perform better than the HC/AC ratio in the prediction of LGA. In T2DM, as opposed to T1DM, the predictive performance was improved by the inclusion of maternal characteristics and HbA1c in the analysis.