Objectives
Pregestational diabetes may lead to excessive fetal growth which is associated with perinatal complications such as shoulder dystocia, perinatal mortality and caesarean delivery. These complications are closely related to the maternal glycemic control (1-6).
Previous studies have shown a significant association between maternal HbA1c at second and third trimester and fetal macrosomia (7-9). However, such association was not demonstrated in early pregnancy or at the time of conception.
This study aims to investigate the association between birthweight and HbA1c in type 1 diabetes (T1DM).
Methods
This is a retrospective cohort study, including all singleton pregnancies with T1DM at Aalborg University Hospital between January 2010 and December 2019. Data on maternal characteristics, fetal and perinatal outcomes were collected from patient records.
HbA1c values were included preconceptionally and from each trimester of pregnancy; gestational week 8-12, 18-22 and 30-34.
Birthweight deviation was calculated using the Scandinavian reference by Marsál et al (10), and large for gestational age (LGA) at birth was defined as birthweight deviation ≥22% of the expected for gestational age.
The association between HbA1c and birthweight deviation was investigated by linear regression, and the prediction of LGA was investigated by logistic regression and receiver-operating characteristic (ROC) curves.
Results
A total of 243 pregnancies complicated by T1DM were included in the study. The predictive performance of HbA1c in relation to LGA was increased as pregnancy advances, with significant values only for the second (OR=1.03, p=0.03, AUC=0.604) and the third trimester (OR=1.04, p<0.01, AUC=0.669).
To explore this association further HbA1c was plottet against birthweight deviation (Figure 1). Linear regression was performed stratified on HbA1c using 60 mmol/mol as a cutoff value, and a significant difference between the coefficients was demonstrated (Table 1). For HbA1c values <60 mmol/mol there was a positive correlation between HbA1c and birthweight deviation. However, for HbA1c values ≥60 mmol/mol, the correlation between HbA1c and birthweight deviation was non-significant or even negative.
Conclusions
This study demonstrates a positive correlation between maternal glycemic control and birthweight when HbA1c is < 60 mmol/mol. However, in very dysregulated diabetic pregnancies, birthweight may even be negatively correlated to HbA1c. One could hypothesize, if this negative correlation could be a result of placental dysfunction, thus further research is needed in this area. These correlations likely explain the rather poor performance of HbA1c in the prediction of LGA especially in early pregnancy.