Corticosteroid exposure after 34 weeks and prior to planned caesarean birth
Our findings concur with a large RCT of pregnancies from 34+0 weeks at risk of late preterm birth8. The relative risk of any hypoglycaemia in the corticosteroids (2 doses of 12mg of betamethasone) arm was 1.60 (95% 1.37–1.87). With a 15% incidence in the placebo arm, the definition included much less severe hypoglycaemia than in our analysis. Importantly, 84% of neonates were born before 37+0weeks, so the findings are less relevant than for our term cohort. Further, in the trials assessing corticosteroids specifically prior to planned early term caesarean birth 10-12, hypoglycaemia has not been recorded. Our data importantly addresses this specific group and highlights the risk of severe hypoglycaemia in these term neonates. A particular issue is pregnancies complicated by diabetes, on which existing literature is limited26-29. In the UK, recommendations state diabetes should not be a ‘contraindication to antenatal corticosteroids for fetal lung maturation’ 3. Our numbers preclude the assessment of effects of ACS administration before 34 weeks in infants born at term. However, we demonstrated that infants exposed to ACS at or after 34 weeks had a higher incidence of severe hypoglycaemia compared to those not exposed (31% vs 3.8%, aOR: 5.76, 95% CI 2.28-14.52). Whilst respiratory benefits are likely to remain, this risk should be considered alongside that of poorer maternal glucose control.
The relationship between SGA and ACS needs to be addressed. McKinzie et al 22 found a higher risk of SGA among term pregnancies who had received ACS before 34+0 weeks; we did not (Table 1). We postulate that ACS may, and this may vary in different units, be given more in pregnancies where SGA is suspected and so the relationship is not causal. This is supported by our finding of increased SGA in pregnancies exposed after 34 weeks.
The positive correlation between corticosteroid-to-birth interval and neonatal glucose level, albeit as a univariate analysis, pictorially demonstrates the increased risk in those who benefit least. It may also shed light on the mechanism. Temporary maternal hyperglycaemia frequently follows corticosteroid administration and this could cause fetal hyperinsulinemia, but this mechanism has not been proven30. That hypoglycaemia is more common even many weeks later suggests that longer term metabolic changes are induced.