Association of change in haemoglobin A1c with adverse perinatal outcomes in women with pregestational diabetes.

Aims: To determine whether a net decline in glycosylated haemoglobin (HbA ) from early to late pregnancy is associated with lower risk of adverse perinatal outcomes at delivery among women with pregestational diabetes.

Methods: A retrospective analysis from 2012 to 2016 at a tertiary care centre. The exposure was the net change in HbA from early (<20 weeks gestation) to late pregnancy (≥20 weeks gestation). Primary outcomes were large for gestational age (LGA) and neonatal hypoglycaemia. The association between outcomes per 6 mmol/mol (0.5%) absolute decrease in HbA was evaluated using modified Poisson regression, and adjusted for age, body mass index, White Class, early HbA and haemoglobin and gestational age at HbA measurement and delivery.

Results: Among 347 women with pregestational diabetes, HbA was assessed in early (9 weeks [IQR 7,13]) and late pregnancy (31 weeks [IQR 29,34]). Mean HbA decreased from early (59 mmol/mol [7.5%]) to late (47 mmol/mol [6.5%]) pregnancy. Each 6 mmol/mol (0.5%) absolute decrease in HbA was associated with a 12% reduced risk of LGA infant (30%, aRR:0.88; 95% CI:0.81,0.95), and a 7% reduced risk of neonatal hypoglycaemia (35%, aRR:0.93; 95% CI:0.87,0.99). Preterm birth (36%, aRR:0.93; 95% CI:0.89,0.98) and neonatal intensive care unit admission (55%, aRR:0.95; 95% CI:0.91,0.98) decreased with a net decline in HbA , but not caesarean delivery, pre-eclampsia, shoulder dystocia and respiratory distress syndrome.

Conclusions: Women with pregestational diabetes with a reduction in HbA may have fewer infants born LGA or with neonatal hypoglycaemia. Repeated assessment of HbA may provide an additional measure of glycaemic control.