In recent years, there has been a reevaluation of the upper gestational age limit recommended for administration of antenatal corticosteroids. Seven randomized clinical trials (RCTs) published between 2010 and 2022 have evaluated the effects of expanding the limit from 34 to 37 weeks鈥� gestation. The largest of these was the Antenatal Late Preterm Steroids (ALPS) trial published in 2016.¹ In this trial, the administration of antenatal corticosteroids yielded a 20% relative reduction in the primary outcome (respiratory morbidity). The absolute reduction in the primary outcome was only 2.8% (from 14.4% to 11.65%), and the effect of antenatal corticosteroids was essentially limited to a reduction in the transient tachypnea of the newborn. There was no discernable benefit of corticosteroids on mortality, respiratory distress syndrome, necrotizing enterocolitis, or intraventricular hemorrhage. Conversely, corticosteroids were associated with a 40% increase in neonatal hypoglycemia (from 15% to 24% in the treated group). Subanalysis suggested that the benefit on respiratory morbidity declined with increasing gestational age, being limited to infants born before 35 weeks鈥� gestation, whereas the incidence of neonatal hypoglycemia increased over the same interval.²

The results of the ALPS trial are in line with a 2020 Cochrane Review³ that analyzed trials performed between 1972 and 2018 and suggested a lack of benefit for mortality or intraventricular hemorrhage and only a modest improvement in the risk of respiratory distress syndrome (risk ratio, 0.75; 95% CI, 0.60- 0.95). Two more recent RCTs from India, not included in the 2020 Cochrane Review, showed no benefit for neonatal morbidity and mortality (in a 4-hospital study sponsored by the World Health Organization [WHO] and published in 2022)⁴ or the need for respiratory support (in a single-hospital study in 2024).⁵ Currently, another WHO study in low-resource countries is evaluating the effect of late preterm antenatal corticosteroids on mortality and respiratory distress syndrome, also comparing the outcomes of a lower dose of betamethasone to the standard dose of dexamethasone.⁶

Ultimately, the principal tradeoff is the benefit-risk ratio of late preterm corticosteroids administration. The seemingly small number needed to treat of 35, calculated from the ALPS trial data, may actually be unfavorable considering the nature of the benefits and risks. To prevent one case of transient, nonlethal morbidity, 35 fetuses could be exposed to antenatal corticosteroids with potential long-term consequences yet to be elucidated. At 34 to 36 weeks鈥� gestation, the fetal brain is highly vulnerable because oligodendrocytes are growing at their fastest rate and large parts of the cortex and cerebellum are still developing.⁷ Exposure to corticosteroids during this interval has been reported to decrease the levels of neurotrophin-3, a neuronal growth factor critical for brain development.⁸ To their credit, the ALPS investigators initiated a long-term follow-up study of the neurodevelopmental outcomes in exposed children.

In this issue of JAMA, Gyamfi-Bannerman et al⁹ present the results of this follow-up study evaluating the neurodevelopmental outcomes of 949 children of ALPS trial participants at a median age of 7 years. There were no statistically significant differences between the betamethasone and placebo group in the primary outcome鈥攁 General Conceptual Ability Score measuring cognitive function (adjusted relative risk, 0.94; 95% CI, 0.73-1.22). No differences were observed in any of the secondary outcomes either. Secondary outcomes included verbal ability, nonverbal reasoning, spatial ability, and gross motor function assessed by certified staff, as well as other metrics assessed by parent-reported questionnaires. The authors interpret the data as reassuring with regard to the administration of antenatal corticosteroids in the late preterm period.

These reassuring results derived from a high-quality study and, most importantly, given the RCT follow-up study design, were not susceptible to bias by indication or timing of exposure. They still have to be considered within the context of several limitations: First, a well-recognized impediment in follow-up studies is loss to follow-up. The neurodevelopmental follow-up studies included in the 2020 Cochrane Review³ registered up to 40% loss to follow-up. The 60% loss to follow-up in the Gyamfi-Bannerman study is 50% higher and causes the study to be underpowered. Furthermore, loss to follow-up does not often occur completely at random; it may be affected by the intervention and/or outcome. A large loss to follow-up compromises inferences, and analysis methods may not fully compensate for such missing data.

Second, the report by Gyamfi-Bannerman et al provides data on midterm outcomes of neurodevelopment but does not address the long-term effects on finer measures such as school performance scores, or the consequence of increased cortisol reactivity. It cannot be taken for granted that the outcomes of finer measures will necessarily follow those reported for gross deviations from normal.¹⁰ Furthermore, unwanted intrauterine programming outcomes secondary to excessive antenatal corticosteroid exposure may only emerge later in life, and may even only become manifest in combination with another exposure (ie, a 鈥渟econd hit鈥�). Thus, although this report provides valuable and encouraging data, it can only engender partial reassurance.

Third, despite strict protocol specifications intended to avoid term deliveries, 16.4% of the ALPS participants still delivered at term. It is well documented that the rate of term delivery after the administration of antenatal corticosteroids for anticipated preterm delivery is between 40% and 50% due to a combination of liberal antenatal corticosteroids ordering practices and a dismal ability to predict preterm delivery. With late preterm births constituting 74% of total preterm births, it is expected that the rate of term deliveries after the adoption of late preterm corticosteroids will significantly increase. There is ever-accumulating evidence for potential long-term adverse health effects of antenatal corticosteroids in term-born neonates.¹¹ The available randomized outcome data on late preterm antenatal corticosteroids cannot be further stratified by gestational age at delivery due to missing information, and the questionable stratification based on a postrandomization event (delivery date) is an additional confounding factor. Several observational studies have found an association between antenatal corticosteroids and worse neurodevelopmental and cognitive outcomes, especially in children who had been born late preterm and term, findings echoed in a recent meta-analysis.¹² Although the level of reliability of evidence derived from observational studies is limited due to the increased potential for confounding, the number of such reports, their consistency, their different geographic provenance, and the quality of reporting make their complete dismissal impossible.

The contribution of the well-conducted ALPS trial and the scientific probity of the intermediate follow-up study by Gyamfi-Bannerman et al⁹ are undeniable. What is still at issue is the direct translation of randomized trial evidence into clinical practice. Can a subgroup analysis of 160 children who had been born at term in an RCT follow-up study cancel out the concerns raised by observational data on almost 700鈥�000 infants?¹² The dissemination of the ALPS trial results rapidly influenced practice and professional guidelines and led to a 4-fold increase in the rate of antenatal corticosteroids administration for late preterm deliveries in the US by 2018. Concomitantly, the number of term-born neonates exposed to antenatal corticosteroids more than doubled.¹³ Of note, however, there is a large regional variation in the adoption of this practice in the US, with an overall high level of nonadoption (51.8%).¹⁴ Outside of the US, very few settings have adopted late preterm corticosteroids. Many obstetrical societies, which had included late preterm antenatal corticosteroids in their guidelines by 2018, have subsequently reversed the recommendation. The WHO, the International Federation of Gynecology and Obstetrics (FIGO), the World Association of Perinatal Medicine, the German, Austrian, and Swiss societies of gynecology and obstetrics, the European Consensus Guidelines and, as of June 2023, the Society of Obstetricians and Gynaecologists of Canada, all strongly recommend that the upper limit for routine antenatal corticosteroids administration be maintained at 34 weeks鈥� gestation because of current uncertainty regarding the risk to benefit ratio after 34 weeks. The Royal College of Obstetricians and Gynaecologists extends the upper limit to 34 weeks, 6 days of gestation.

It remains to be seen if the report by Gyamfi-Bannerman et al will be perceived as sufficiently reassuring to reverse the current limited uptake of late preterm antenatal corticosteroids, either internationally or within the US.

Corresponding Author: Alex C. Vidaeff, MD, MPH, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Baylor College of Medicine, Texas Children鈥檚 Hospital Pavilion for Women, 6651 Main St, Ste F1020, Houston, TX 77030 (vidaeff@bcm.edu).

Published Online: April 24, 2024. doi:10.1001/jama.2024.2228

Conflict of Interest Disclosures: None reported.