Corticosteroid exposure remote from term
In neonates exposed to ACS before 34 weeks, remote from birth, our findings build on a recent cohort study of pregnancies ‘diagnosed as having threatened preterm labour at some point’ who nevertheless delivered from 37+0 weeks 22. In this, the 27.4% of pregnancies that received ACS were compared to those who did not. The mean gestational age at ACS administration was 32.2 weeks (SD 3.3). The incidence of treated hypoglycaemia was more than twice as high in exposed neonates, although this relationship was not examined using adjustment for covariates. Our adjustment suggests this is causal.
Similarly, Raikkonen et al 5 demonstrated that exposure to maternal ACS was significantly associated with mental and behavioural disorders in children, especially when born at term (HR 1.47 (95% CI 1.36 – 1.69)). In this cohort the vast majority of term pregnancies received ACS before 34+0 weeks.
The overall benefit of ACS in preterm neonates is unequivocal2 but is greatest in the most preterm neonates23. Whilst developmental delay is reduced in those born under 34 weeks 24, no overall effect on neurodevelopment delay has been detected in RCTs of antenatal corticosteroids in pregnancies at risk of (as opposed to having a) preterm birth 2.
Given their effect on mortality, any potential disadvantage of ACS will be outweighed by the benefits where very preterm birth occurs. Therefore public health initiatives appropriately encourage corticosteroid administration. However if more neonates are exposed to corticosteroids, the proportion that go to term will increase, and any adverse effects will then become more common. The challenge is both better sensitivity and better specificity in predicting preterm birth: identification of those at most risk is improved using point of care tests, but even these have a low specificity 25.