The research, published in BMC Pediatrics by Genmiao Zhuang, Xiaohua Xie, and Jianan Chen of the First Affiliated Hospital of Xiamen University, examined 68 extremely preterm infants treated between March 2016 and January 2024. Of these, 41 received low-dose dexamethasone in addition to standard care, while 27 received standard respiratory support, nutritional support, anti-infection measures, and other conventional treatments alone. The team then asked a deceptively simple question: did the steroid change the odds of the worst outcomes, and did it trigger side effects that clinicians and parents have long worried about?

To understand why this question matters, it helps to know what dexamethasone actually does. It is a synthetic glucocorticoid, a powerful anti-inflammatory agent that mimics the stress hormones the body would normally produce. In the immature lung, inflammation driven by ventilation, oxygen toxicity, and infection is a central driver of BPD. By dampening that inflammatory cascade, steroids can in principle accelerate recovery, reduce the duration of mechanical ventilation, and give the lung a window in which to continue its development. Dexamethasone has been used for this purpose for decades, and early trials in the 1990s showed striking reductions in ventilator dependence. But those early successes came at a cost that changed neonatal practice almost overnight.

Follow-up studies of infants treated with early, high-dose dexamethasone revealed higher rates of cerebral palsy and impaired neurodevelopment, prompting professional bodies worldwide to warn against routine systemic steroid use in preterm infants. Since then, neonatologists have walked a tightrope: steroids remain one of the few interventions that can wean a ventilator-dependent baby off the machine, yet their safety profile, particularly at lower doses and later in the course of the disease, is still not fully mapped. The optimal dosage, treatment duration, and safety window remain subjects of genuine clinical uncertainty, which is precisely the gap the Xiamen team set out to address.

The study’s design was retrospective, meaning the researchers looked back at records of infants already treated rather than assigning treatments prospectively. That choice carries inherent limitations, since the decision to give dexamethasone was made by treating clinicians and may have been influenced by how sick each infant appeared. The researchers confronted this problem with statistical adjustment. Their primary analysis used multivariable binary logistic regression, controlling for three of the most important predictors of outcome in this population: gestational age at birth, birth weight, and the respiratory severity score, a measure of how much respiratory support an infant requires. By adjusting for these variables, the team attempted to disentangle the effect of the drug itself from the effect of being a sicker baby in the first place.

The results were striking. After adjustment, infants who received low-dose dexamethasone had substantially lower odds of reaching the study’s composite endpoint of severe BPD or in-hospital death, with an adjusted odds ratio of 0.26, meaning roughly a quarter of the odds compared with untreated infants, and a 95 percent confidence interval of 0.09 to 0.55 with a P value of 0.003. Among survivors, the odds of severe BPD specifically were similarly reduced, with an adjusted odds ratio of 0.27. Perhaps most clinically meaningful for day-to-day intensive care, the treated infants had nearly three times the odds of successful extubation, that is, of coming off the ventilator, with an adjusted odds ratio of 2.88 and a confidence interval of 1.20 to 7.57, reaching statistical significance at P equal to 0.024.

Then came the safety data, and here the picture grew more nuanced. Hyperglycaemia, or elevated blood sugar, a well-known metabolic side effect of glucocorticoids, occurred in 41.5 percent of the dexamethasone group compared with 18.5 percent of the control group. That is a substantial numerical difference, and one that any clinician would notice at the bedside, since steroid-induced hyperglycaemia often requires monitoring and sometimes insulin therapy in neonates. Yet when the researchers ran the adjusted analysis, the difference did not reach conventional statistical significance, yielding an adjusted odds ratio of 2.04 with a confidence interval of 0.82 to 4.97 and a P value of 0.121. The confidence interval spans the null value of one, which means the data cannot rule out either no effect or a meaningful increase in risk.

The authors are refreshingly candid about why that ambiguity persists. With only 68 infants in total, the study is small, and small studies have limited statistical power to detect differences in adverse events, particularly ones that are relatively uncommon or variable in presentation. The team also notes that long-term growth was not assessed, leaving open the question of whether the drug affects the infants’ development after discharge. And because the study was conducted at a single center, its findings may reflect local treatment protocols, patient demographics, and clinical judgment rather than universal patterns. In their conclusion, the researchers state plainly that clinically relevant increases in adverse events, particularly hyperglycaemia, cannot be excluded given the limited sample size, and that their findings require validation in larger multicenter prospective studies.

That caveat is not a dismissal; it is a roadmap. The field of neonatology has been burned before by premature enthusiasm for steroids, and the memory of high-dose dexamethasone’s neurodevelopmental toll still shapes every treatment decision. What this study adds is a carefully adjusted signal that low-dose regimens, used in extremely preterm infants with established respiratory disease, may substantially reduce the risk of severe BPD and death while improving the chances of successful ventilator weaning, all with a side-effect profile that, at least in this dataset, did not show statistically confirmed harm. The hyperglycaemia signal, though not statistically significant, is a reminder that the drug is pharmacologically active and metabolically consequential even at reduced doses, and that blood glucose monitoring should be an integral part of any treatment protocol.

For the families of the roughly one in ten babies born preterm worldwide, the stakes of this research are anything but abstract. Severe BPD can mean months in the neonatal intensive care unit, home oxygen therapy, and elevated risks of asthma-like symptoms throughout childhood. If larger, prospective, multicenter trials confirm what the Xiamen team observed, low-dose dexamethasone could take a more secure place in the standard of care, offering the sickest premature infants a better chance of breathing on their own. Until then, the study stands as a measured contribution to one of neonatal medicine’s longest-running debates: how to harness the anti-inflammatory power of steroids to protect fragile lungs without trading respiratory outcomes for developmental ones. The answer, the authors suggest, will come not from any single center’s records, but from the kind of rigorous, large-scale trials that only a coordinated research community can deliver.

Subject of Research: Safety and efficacy of low-dose dexamethasone for bronchopulmonary dysplasia in extremely premature infants

Article Title: Analysis of adverse reactions to low-dose dexamethasone in bronchopulmonary dysplasia of extremely premature infants: a retrospective study

Article References: Zhuang, G., Xie, X., & Chen, J. (2026). Analysis of adverse reactions to low-dose dexamethasone in bronchopulmonary dysplasia of extremely premature infants: a retrospective study. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07706-5

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07706-5

Keywords: dexamethasone, bronchopulmonary dysplasia, extremely preterm infants, neonatology, adverse drug reactions, hyperglycaemia, mechanical ventilation, retrospective study, corticosteroids, premature birth, respiratory severity score, neonatal intensive care