This retrospective study highlights some consequences of ACS exposure on fetal growth and on metabolic functions in term born neonates. ACS exposure in these neonates is significantly associated with a smaller HC at birth, and a smaller BH, even after adjustment for confounding factors. Moreover, neonatal complications such as hypocalcemia and eating disorders are significantly more frequent in ACS exposed newborn.
The impact of ACS exposure on cerebral growth has been previously described and our results are consistent with previous studies. Many of them, particularly in the population of preterm neonates, showed reduced HC at birth, more important in case of multiple courses of ACS (18–21). Although many studies concluded on a significant difference in HC after ACS exposure, a recent historical French cohort study moderates these results (22). Despite an important population, no difference on the rate of very small head circumference (i.e. HC < 5th percentile) was demonstrated between ACS exposed and non-exposed newborns. It can be suggested that, even if there is a restriction of HC growth secondary to ACS, the restriction of head growth stays moderate and does not lead to microcephaly.
In addition to the association of ACS exposure with a restriction of brain growth, studies showed an impact of ACS on neurodevelopment such as learning disabilities (23), mental and behavioral disorders in exposed children (14). In 2007, a study warned about a possible higher rate of cerebral palsy among exposed children and asks for further studies (24). A recent meta-analysis highlighted a significant reduction in the risk of neurodevelopmental impairment for extremely preterm neonates but suggested a higher risk for neurocognitive disorders for late-preterm births and a higher risk of mental or behavioral disorders for full term newborns exposed to ACS (25) .Regarding Intelligence Quotient (IQ), a reduction of IQ in ACS exposed children have been described in comparison with non-exposed controls (26). The mechanisms leading to these neurological impairments are not well identified, but some assumptions are made: betamethasone could impact blood – brain barrier permeability (11), the concentration (27) or the localization of certain proteins (28)such as glucocorticoid receptors.
Regarding birth height, our results suggest an impact of ACS on fetal growth. This impact on fetal length have been reported in others studies (10–25). Diguisto at al., demonstrated a higher frequency of exposed newborn with a birth length under the 5e percentile (22).
Regarding birth weight, the impact of ACS exposure is still uncertain, with a few studies conducted in term neonates (29, 30). Previous studies evidenced growth restriction on preterm neonates exposed to ACS, especially if there is more than one course of corticosteroids (19, 31, 32). Assessing the precise role of ACS exposure on fetal growth is difficult because of numerous confusion factors. Many risk factors of intra-uterine growth restriction have already been identified. First, sociodemographic factors which can be changeable or not (mother’s age, mother’s intrauterine growth, toxics or alcohol exposure, mother’s nutritional status BMI < 18 kg/m2) (33–37). Then, on the obstetrical level, multiple pregnancies or closed pregnancies are well known risk factors. In this study, we did not collect sociodemographic data apart alcohol, smoking and toxics exposure so adjustment on confounding factors is imperfect. In this study, maternal smoking exposure represents the main confusion factor, with a smoking exposure twice more frequent in the exposed group. There is an easy explanation for this difference as smoking exposure was recently identified as a risk factor for preterm delivery (38). Nevertheless, there is still a significant association between ACS exposure with HC and BH in multivariate regression analysis.
-
On a metabolic point of view, we chose to evaluate neonatal complications, which could be secondary to adrenal insufficiency. Neonatal hypocalcemia and eating disorders are more frequent in exposed term neonates compared to non-exposed controls. We didnot evidence significant difference on hypoglycemia at birth but this risk had already been notified in various studies but for preterm birth (39–41). The higher frequencies of these neonatal complications raised the question of the consequence of ACS exposure on hypothalamic-pituitary-adrenal axis function as suggested by previous studies(42, 43).
All the above-mentioned neonatal metabolic complications are influenced by weight birth, even eating disorders. Intrauterine Growth Restriction (IUGR) could be a confounding factor but there was no significant difference of newborn with growth restriction in the 2 groups of this study.
The strengths of the study are the focus on term newborn exposed to ACS, the number of patients included, calculated a priori to obtain a statistical power of 80% to evidence a difference of 0,5 cm on head circumference of newborns. For birth parameters we chose to use the AUDIPOG curves, most appropriate to our population, with an expression in centiles in order neutralize the effect of GA at birth. The AUDIPOG curves are adapted to evaluated birth parameters for a French group of patients and with frequent use for scientific studies (https://www.audipog.net). Nevertheless, our study presents some limitations. The most important one is the retrospective design of the study. However, because of the recognized beneficial effects of corticosteroids for premature neonates, it would be unethical to perform a randomized controlled trial. It would be interesting to lead a retrospective analysis but with a prospective recruitment, on a predefined period, including mothers and newborns at discharge. It could help to reduce missing data or misinformation and may allow an exhaustive recruitment of exposed patients. A second limitation is that the group sizes were too small to perform subgroup analysis (number of ACS courses, gestational age of ACS for example). In our study, due to the small number of second course (N = 10, 7.4%), subgroups analyses to study the impact of multiple courses of ACS exposure on growth parameters could not be performed. It would have been interesting to study if neonates exposed earlier in pregnancy to corticosteroids would have been more or less impacted.
This study is in line with the current trend to reduce ACS to the minimum effective dose, to prevent neonatal and post-partum complications. However, the French multicenter BETADOSE study evaluating 3244 women was not able to show a non-inferiority of a half-dose of betamethasone compared to a full dose (44). Numerous studies discuss the possibility of decreasing the empiric doses of betamethasone, but no one supports a practice change towards antenatal betamethasone dose reduction. A possible solution, to limit the prescription of ACS could be the identification of new markers improving the reliability of the risk of premature delivery, in case of premature labor.