The trigger for the exchange is a secondary analysis of the BeNeDuctus trial, conducted by Hoornenborg and colleagues, which examined extremely preterm infants who received paracetamol for pain relief. According to the letter, that analysis found a significantly increased risk of moderate-to-severe bronchopulmonary dysplasia, a chronic lung condition of prematurity, as well as longer requirements for respiratory support among infants exposed to the drug. Because the sample originated from a randomized controlled trial, the finding carries more weight than the observational studies that have fueled suspicion of paracetamol for the past two decades. The letter’s authors commend the trial team while pressing the point that even these researchers continue to call for yet more studies rather than recommending a change in practice.

Bronchopulmonary dysplasia remains one of the most consequential outcomes in neonatology. It arises when the lungs of infants born extremely early fail to develop normally, often after weeks of mechanical ventilation and oxygen therapy, and it predicts years of respiratory illness, rehospitalization and impaired growth. Any drug that measurably raises the risk of moderate-to-severe disease would represent a serious threat to a vulnerable population, which is precisely why the letter frames the question in such stark terms. The authors ask, in effect, how much evidence is enough before the burden of proof shifts from those who suspect harm to those who assume safety.

The biological plausibility of a paracetamol-lung link rests on the drug’s metabolism. Paracetamol is processed primarily in the liver, but a fraction is converted by cytochrome enzymes into a reactive metabolite, N-acetyl-p-benzoquinone imine, which depletes glutathione, the cell’s principal antioxidant defense. The developing lung expresses xenobiotic-metabolizing enzymes, meaning that pulmonary tissue can locally generate toxic metabolites from circulating paracetamol. In a preterm infant, whose antioxidant systems are immature and whose lungs are already under oxidative assault from oxygen therapy and ventilation, additional glutathione depletion could plausibly tip the balance toward arrested alveolar development. Pharmacokinetic studies in very preterm infants have documented measurable exposure not only to the parent compound but to all of its metabolites at standard intravenous doses, underscoring that these tiny patients handle the drug differently from adults.

This mechanistic concern is not merely theoretical. Preclinical work summarized in the pulmonary physiology literature has explicitly called for animal models to understand why the perinatal lung appears uniquely susceptible to acetaminophen, the American name for the same molecule. Meanwhile, a cohort study of extremely preterm infants detected acetaminophen and its xenobiotic metabolites in human milk and reported associations with both bronchopulmonary dysplasia and retinopathy of prematurity, another disease of oxidative stress. There have even been reports of paracetamol overdose in newborns and infants proving life-threatening, a reminder that the therapeutic window in this population is narrower than commonly assumed. Together, these strands sketch a coherent biological story in which a drug once considered benign for the lung may interfere with redox-sensitive developmental processes at exactly the wrong moment.

The broader epidemiological record adds a second, parallel line of concern. Systematic reviews and meta-analyses stretching back more than a decade have consistently reported associations between paracetamol use in pregnancy and wheezing illness or asthma in offspring, and more recent syntheses of prenatal exposure studies have reached similar conclusions. Research in high-risk birth cohorts has also examined early-life acetaminophen exposure in combination with genetic variation in glutathione S-transferase enzymes, finding evidence that the drug’s oxidative burden may interact with inherited detoxification capacity to shape asthma risk into adolescence. Population studies of childhood asthma prevalence continue to list early paracetamol exposure among the modifiable risk factors under scrutiny, and even lactation, long considered a safe channel, has been questioned in the pediatric literature.

What makes the new trial-based analysis so important is the hierarchy of evidence it occupies. Observational studies of drug exposure are notoriously vulnerable to confounding by indication: children who receive more paracetamol tend to be sicker, and sicker infants develop more lung disease regardless of which painkiller they receive. A secondary analysis nested within a randomized controlled trial does not eliminate that problem entirely, since randomization in the parent BeNeDuctus trial was not designed to allocate paracetamol, but it does allow investigators to control for illness severity far more rigorously than most registry or cohort studies can. The fact that a signal for moderate-to-severe bronchopulmonary dysplasia and prolonged respiratory support survived that scrutiny is what lends the letter its urgency.

Yet the letter’s authors are careful not to overclaim causation, and the trial investigators themselves stopped short of a practice change, calling instead for further research into paracetamol’s respiratory effects in preterm infants. That restraint reflects genuine clinical dilemmas. Preterm infants undergo painful procedures, from cannulation to surgery for patent ductus arteriosus, and untreated pain carries its own documented harms, including stress responses that destabilize cerebral blood flow. Paracetamol has become attractive precisely because it avoids the renal, gastrointestinal and platelet side effects of nonsteroidal anti-inflammatory drugs. An overview of systematic reviews on paracetamol’s efficacy and safety concluded that while the drug offers modest analgesic benefit, its safety profile in vulnerable groups deserves continued attention, a conclusion that now reads differently in light of the neonatal data.

The dilemma is sharpened by trends in neonatal care. A recent analysis of fifteen years of respiratory and hemodynamic management in very premature infants documented evolving approaches to bronchopulmonary dysplasia, and observational multicenter registries have compared conservative and medical strategies for managing the ductus arteriosus, a condition for which paracetamol has itself been trialed as a closing agent. In other words, paracetamol is not merely an incidental exposure in the neonatal unit; it is increasingly embedded in treatment protocols. If the drug independently worsens lung outcomes, the implications extend from analgesia guidelines to the very protocols designed to protect the preterm lung, and clinicians would need to weigh analgesic benefit against a respiratory cost that previous risk-benefit calculations never included.

For now, the letter functions as both a summary and a challenge. It gathers the mechanistic, pharmacokinetic, preclinical and epidemiological threads into a single argument that the evidence base has grown substantial enough to demand action beyond another round of observational studies. The authors, who report no competing interests and no external funding, frame their appeal around the families and clinicians they serve, and their central question, how much evidence is enough, is one that guideline committees, regulators and neonatologists will now have to answer explicitly. Whether the answer comes from adequately powered randomized trials of analgesic strategy in extremely preterm infants or from a precautionary shift in prescribing habits, the era in which paracetamol was assumed to be harmless to the developing lung is drawing to a close.

Subject of Research: The association between paracetamol exposure and bronchopulmonary dysplasia in extremely preterm infants

Article Title: Paracetamol and the developing lung: how much evidence is enough?

Article References: Verd, S., Boo, S., & Costa-Orvay, J.-A. (2026). Paracetamol and the developing lung: how much evidence is enough?. Pediatric Research. https://doi.org/10.1038/s41390-026-05532-0

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05532-0

Keywords: paracetamol, acetaminophen, bronchopulmonary dysplasia, preterm infants, neonatology, respiratory support, glutathione, oxidative stress, asthma risk, BeNeDuctus trial, Pediatric Research, drug safety

The clinical stakes are considerable. Bronchopulmonary dysplasia, defined by the need for supplemental oxygen or respiratory support at a corrected gestational age of thirty-six weeks, affects a substantial proportion of infants born before twenty-eight weeks of gestation and is associated with long-term respiratory morbidity, neurodevelopmental impairment, and rehospitalization throughout childhood. Its pathogenesis is multifactorial, involving inflammation, infection, hyperoxia exposure, impaired alveolar and vascular development, and genetic and environmental interactions. Because no single therapy has proven reliably protective, clinicians scrutinize every drug that reaches the developing lung for possible harm, and the analgesics and anti-inflammatory agents used in the first weeks of life are no exception.

The story of these drugs in the neonatal intensive care unit begins with the ductus arteriosus itself. Increased blood flow across a patent ductus arteriosus can be associated with increased pulmonary blood circulation and morbidity, a hemodynamic burden that physicians have tried to relieve since the latter part of the twentieth century by closing a hemodynamically significant ductus in the hope of improving outcomes. The pharmacological workhorses of that effort have been cyclooxygenase inhibitors, chiefly indomethacin and ibuprofen, which reduce prostaglandin signaling and allow the vessel to constrict and close. Both drugs have been used extensively for ductal closure, but both carry side effects and safety concerns in preterm infants, including effects on renal perfusion, gastrointestinal integrity, and platelet function, concerns that have driven a decades-long search for alternatives.

That search produced one of the more memorable accidental discoveries in modern neonatology. As recounted by Sharma and Bhandari, Hammerman and colleagues observed that the use of acetaminophen in preterm infants led to closure of the patent ductus arteriosus, a finding reported in case series that surprised many observers. The mechanism is plausible: acetaminophen acts on prostaglandin synthesis through pathways distinct from classical cyclooxygenase inhibition, and the ductus in the preterm infant remains sensitive to even modest reductions in prostaglandin activity. Following those initial case reports, multiple trials demonstrated that acetaminophen was as effective in closing the ductus as the two established non-steroidal anti-inflammatory drugs, ibuprofen and indomethacin. Crucially, those studies did not look only at closure rates; they also examined associated morbidities, and bronchopulmonary dysplasia emerged as the outcome of greatest interest.

The evidence on that outcome, however, has been reassuringly consistent so far. A large multicenter trial conducted by Gupta and colleagues, published in the New England Journal of Medicine, showed no significant difference in the incidence of bronchopulmonary dysplasia between groups of infants treated with ibuprofen and those given placebo. A retrospective study by Jensen and colleagues compared one group of infants receiving acetaminophen with another receiving ibuprofen or indomethacin and found that the incidence of death or grade 2 to 3 bronchopulmonary dysplasia was similar in both groups. These findings suggested that neither class of drug, when used for ductal closure, clearly worsens the risk of chronic lung disease, at least within the populations and dosing regimens studied.

The BeNeDuctus trial added an important dimension to this debate. This large multicenter randomized non-inferiority trial compared expectant management of a significant patent ductus arteriosus, meaning careful observation without immediate drug treatment, against early treatment with ibuprofen. The trial showed no difference in the incidence of bronchopulmonary dysplasia, necrotizing enterocolitis, or death between the two management strategies. That result was consequential in its own right, because it suggested that routine early pharmacological closure of the ductus may not be necessary for all affected infants. But it also created an unexpected opportunity: some patients in both subgroups received acetaminophen, not for ductal closure but for analgesia, meaning the trial inadvertently enrolled a cohort of extremely preterm infants exposed to the drug for reasons unrelated to their ductus.

Hoornenborg and colleagues exploited that opportunity in their secondary analysis, asking whether paracetamol given as an analgesic was associated with an increased risk of bronchopulmonary dysplasia in extremely preterm infants. The question matters because acetaminophen is among the most widely administered medications in neonatal intensive care, used for procedural and postoperative pain and increasingly considered for ductal closure. If the drug carried even a modest risk to the developing lung, the population-level consequences could be substantial given how many preterm infants receive it. Sharma and Bhandari frame their commentary around exactly this tension between therapeutic utility and pulmonary safety.

Concern about acetaminophen’s effects on the lung is not merely theoretical. Laboratory studies referenced in the commentary have shown that toxic acetaminophen exposure can induce distal lung endoplasmic reticulum stress, proinflammatory signaling, and emphysematous changes in the adult murine lung, and that the developing murine lung appears susceptible to acetaminophen toxicity in ways independent of the hepatic damage for which the drug is best known. Investigators have also raised the possibility that acetaminophen exposure in early life could have lifelong consequences for respiratory health. On the other side of the ledger, ibuprofen has its own pulmonary profile: animal work has documented vascular and pulmonary effects of ibuprofen on neonatal lung development, and recent studies indicate that inhibition of the mitochondrial integrated stress response can ameliorate ibuprofen-induced endothelial dysfunction in models of neonatal hyperoxia-induced lung injury. Both drug classes, in other words, have biological plausibility for lung effects that clinical trials have so far not translated into clear harm.

Interpreting this literature requires attention to the difficulty of the outcome itself. Bronchopulmonary dysplasia is now understood as a syndrome with multiple pulmonary phenotypes in the preterm infant, and its diagnosis has been refined through evidence-based approaches that grade severity by the level of respiratory support required. Newer tools, including magnetic resonance imaging phenotyping and life-course approaches to lung function monitoring after preterm birth, are reshaping how researchers detect subtle drug effects that survival-based or oxygen-based definitions may miss. The commentary’s authors, who have previously argued that bronchopulmonary dysplasia research must move from the bedside to the bench and back again, emphasize that postnatal inflammation, corticosteroid effects on lung development, and gene-environment interactions all complicate any attempt to isolate the contribution of a single analgesic or anti-inflammatory drug.

Where does this leave clinicians and families? The accumulating trial evidence, from the Gupta ibuprofen trial, the Jensen retrospective comparison, the BeNeDuctus trial, and now the Hoornenborg secondary analysis, points toward a consistent conclusion: neither acetaminophen nor the classical non-steroidal anti-inflammatory drugs, as used in contemporary neonatal practice, has been shown to increase the incidence of bronchopulmonary dysplasia or death in extremely preterm infants. At the same time, Sharma and Bhandari’s commentary underscores that the question of pulmonary safety has not been closed as definitively as the ductus itself. Preclinical signals of lung toxicity, the heterogeneity of bronchopulmonary dysplasia phenotypes, and the possibility of long-term respiratory consequences all argue for continued vigilance. As randomized trials of prophylactic acetaminophen for ductal closure and expectant management strategies continue to mature, the developing lung remains the outcome against which every neonatal drug must ultimately be judged.

Subject of Research: Pulmonary safety of acetaminophen and NSAID treatment for patent ductus arteriosus in preterm infants

Article Title: Does use of acetaminophen and/or NSAIDs increase the risk of developing bronchopulmonary dysplasia?

Article References: Sharma, R., & Bhandari, V. (2026). Does use of acetaminophen and/or NSAIDs increase the risk of developing bronchopulmonary dysplasia?. Pediatric Research. https://doi.org/10.1038/s41390-026-05526-y

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05526-y

Keywords: bronchopulmonary dysplasia, acetaminophen, NSAIDs, patent ductus arteriosus, preterm infants, BeNeDuctus trial, neonatology, ibuprofen, indomethacin, Pediatric Research, lung development, neonatal intensive care