Bronchopulmonary dysplasia (BPD) remains one of the most persistent complications affecting infants born prematurely, but clinicians still lack a reliable way to identify which babies are most likely to develop the disease during the earliest stages of neonatal care. A new study led by Gu, Yang, Shi and colleagues explores whether metabolic changes in preterm infants could provide an early warning system, while also accounting for a factor that can strongly influence newborn biology: gestational age.
Published in the Journal of Perinatology, the study focuses on gestational age-specific metabolic biomarkers for predicting BPD after standardized enteral feeding. The researchers used the 2019 definition developed by the National Institute of Child Health and Human Development Neonatal Research Network, commonly known as the NRN 2019 definition. By combining this clinical framework with a consistent feeding protocol, the investigators sought to reduce variation between infants and uncover biological signals more directly linked to BPD risk.
BPD is a chronic lung disorder that primarily affects very preterm infants whose lungs are still developing at birth. These infants may require oxygen supplementation or respiratory support for prolonged periods. The condition is influenced by several interacting processes, including immature lung structure, inflammation, infection, oxygen exposure, mechanical ventilation and nutritional status. Because these factors evolve rapidly after birth, a prediction tool that works at one gestational age may not perform equally well in infants born earlier or later.
The study’s central idea is that metabolism may reveal this changing risk before the clinical signs of BPD become fully established. Metabolites are small molecules produced or modified during normal cellular activity. They include amino acids, lipids, sugars, organic acids and other compounds involved in energy production, membrane formation, oxidative balance and immune regulation. In newborns, the concentrations of these molecules can reflect the combined effects of organ maturation, oxygen exposure, nutrition and disease. Detecting coordinated metabolic shifts could therefore offer a molecular snapshot of how an infant is responding to the stresses of premature birth.
The emphasis on standardized enteral feeding is particularly important. Nutrition is not simply a background variable in studies of premature infants; it can directly reshape the metabolome. Human milk, donor milk and formula differ in their protein, lipid and carbohydrate composition, while changes in feeding volume and tolerance can alter energy availability and intestinal metabolism. If infants receive markedly different nutritional regimens, it becomes difficult to determine whether a metabolic signature reflects impending lung disease or differences in feeding. A standardized protocol provides a more controlled biological context for identifying candidate biomarkers.
Gestational age introduces another layer of complexity. A baby born at 25 weeks and a baby born at 31 weeks may have very different baseline metabolic profiles even when both are clinically stable. Their lungs, liver, immune systems and intestinal tracts are at different developmental stages, and the same metabolite concentration may have different implications in each group. Rather than assuming that one universal biomarker pattern applies to all premature infants, the researchers investigated whether predictive signatures should be tailored to specific gestational-age ranges.
This approach reflects a broader shift in neonatal medicine toward precision risk assessment. Traditional prediction models often combine clinical variables such as birth weight, respiratory support, oxygen requirement and infection. Metabolic data could potentially add a biological dimension to these models, helping distinguish infants who are temporarily vulnerable from those undergoing processes more likely to culminate in BPD. However, a metabolic marker would need to be tested across independent populations and shown to improve prediction beyond information already available at the bedside before it could be adopted clinically.
The use of the NRN 2019 definition also matters because BPD classification has changed over time. Different studies have used varying thresholds for oxygen dependence, respiratory support and timing of assessment, making results difficult to compare. Applying a contemporary, standardized definition gives the reported biomarker analysis a clearer clinical endpoint. It may also help future researchers determine whether a metabolic signature predicts any BPD, more severe forms of the disease or specific patterns of respiratory support.
The investigators’ findings are presented as a step toward earlier and more individualized prediction, rather than as an immediate diagnostic test. The value of the work lies in linking three elements that are often considered separately: developmental maturity, nutritional exposure and metabolic physiology. If the identified signatures are validated in larger and more diverse neonatal cohorts, they could eventually support algorithms that estimate BPD risk at an early stage and update that estimate as an infant’s condition changes. Such tools might guide closer monitoring, inform discussions with families and help researchers select infants for preventive trials.
The study also highlights the biological difficulty of predicting BPD. The disorder is not caused by a single pathway or molecule, and its clinical expression varies widely. Metabolic signatures may reflect several mechanisms at once, including mitochondrial energy stress, altered lipid metabolism, inflammation and impaired tissue development. The challenge will be to separate a reproducible disease signal from the normal metabolic turbulence of premature birth. For now, gestational age-specific analysis and standardized feeding offer a promising strategy for making that distinction more precise, bringing neonatal prediction closer to a model in which treatment and surveillance are tailored to each infant’s developmental biology.
Subject of Research: Gestational age-specific metabolic biomarkers for the early prediction of bronchopulmonary dysplasia in preterm infants following standardized enteral feeding.
Article Title: Gestational age-specific metabolic biomarkers for early prediction of bronchopulmonary dysplasia following standardized enteral feeding
Article References: Gu, P., Yang, Y., Shi, J. et al. “Gestational age-specific metabolic biomarkers for early prediction of bronchopulmonary dysplasia following standardized enteral feeding.” Journal of Perinatology (2026). https://doi.org/10.1038/s41372-026-02850-3
Image Credits: AI Generated
DOI: 10.1038/s41372-026-02850-3
Keywords: Bronchopulmonary dysplasia, premature infants, preterm birth, metabolic biomarkers, metabolomics, gestational age, neonatal medicine, enteral feeding, early prediction, precision medicine
Tags: chronic lung disease prediction in preterm neonatesearly detection of BPD in preterm infantsearly prediction of bronchopulmonary dysplasiagestational age-specific neonatal biomarkersinfluence of gestational age on neonatal biomarker accuracymetabolic biomarkers for neonatal lung diseaseneonatal clinical biomarkers for lung developmentneonatal metabolic profiling for BPD riskneonatal respiratory support and biomarker indicatorsperinatal factors affecting BPDpreterm infant lung development biomarkersstandardized feeding protocols for preterm infants
