Neonatal encephalopathy has long been treated as a condition that can be sorted into neat clinical categories: mild, moderate or severe. A new perspective in Pediatric Research argues that this familiar language may conceal a far more complicated biological reality. In “Beyond definitions: measuring severity in neonatal encephalopathy,” José Arnáez examines why describing the condition is not the same as measuring it—and why the distinction could influence treatment decisions, clinical research and the long-term care of affected infants.
Neonatal encephalopathy is a syndrome of disturbed neurological function during the first days of life. Newborns may show abnormal consciousness, poor muscle tone, weak reflexes, seizures, difficulties with breathing or problems feeding. Hypoxic-ischemic injury, caused by insufficient oxygen and blood flow to the brain, is one important cause, but infection, inflammation, metabolic disorders, stroke, genetic disease and other conditions can produce similar signs. For that reason, the term “hypoxic-ischemic encephalopathy,” or HIE, should not automatically be applied to every infant with neonatal neurological dysfunction.
The urgency of assessing severity increased after therapeutic hypothermia became standard care for selected infants with moderate to severe HIE. Cooling the newborn to approximately 33–34 degrees Celsius for about 72 hours can reduce secondary brain injury when it is initiated within the first six hours after birth. Yet the criteria used to determine eligibility are based largely on early clinical observations, blood-gas measurements, neurological examinations and evidence of a difficult birth. These indicators are valuable, but they are snapshots taken during a rapidly changing process. A newborn who appears severely impaired immediately after delivery may improve quickly, while another with initially subtle signs may deteriorate as cellular injury evolves.
Clinical grading systems are therefore central to neonatal care, but they have limits. Common examinations assess consciousness, spontaneous activity, posture, muscle tone, primitive reflexes, autonomic function and seizures. These domains help clinicians distinguish degrees of encephalopathy, but they are influenced by gestational age, medication, sedation, pain, respiratory support, temperature and the infant’s general medical condition. Premature infants may display neurological signs that differ from those of term newborns, and an examination performed shortly after resuscitation may not reflect the infant’s eventual neurological state.
The article’s central message is that severity should be understood as a multidimensional and time-dependent measurement rather than a single label. The brain injury process can include an initial period of energy failure, a temporary phase of partial recovery and a later wave of secondary damage involving excitotoxicity, oxidative stress, inflammation, mitochondrial dysfunction and cell death. Because these processes unfold over hours and days, severity cannot always be captured accurately by one examination at one moment. Repeated assessments may reveal a trajectory, showing whether neurological function is recovering, remaining stable or worsening.
Electroencephalography and amplitude-integrated electroencephalography offer one way to observe this trajectory. These techniques record the brain’s electrical activity and can identify abnormal background patterns, reduced variability and seizures that may not produce visible movements. Electrographic seizures are particularly important because newborns can have “electroclinical dissociation,” in which abnormal electrical activity continues even after outward signs disappear. A persistently severely abnormal background pattern may indicate substantial cerebral dysfunction, whereas normalization over time can provide evidence of neurological recovery. However, interpretation requires expertise, and electrical patterns can be affected by anticonvulsant medication and other treatments.
Brain imaging adds another layer of information. Magnetic resonance imaging, especially diffusion-weighted imaging and magnetic resonance spectroscopy, can reveal patterns of injury in the basal ganglia, thalamus, watershed regions and white matter. Diffusion abnormalities may emerge early as injured cells lose the ability to regulate water movement, while later scans can show the established distribution and extent of damage. Magnetic resonance spectroscopy can provide metabolic information, including changes in lactate and other chemical signals associated with impaired energy production. Imaging is not a perfect crystal ball, however: timing matters, findings can evolve, and the relationship between a scan and an individual child’s developmental outcome is probabilistic rather than absolute.
Biomarkers from blood, urine or cerebrospinal fluid are also being investigated as tools for refining severity assessment. Molecules such as neuron-specific enolase, S100 calcium-binding protein B, glial fibrillary acidic protein and ubiquitin C-terminal hydrolase-L1 may rise when neurons or glial cells are damaged. Other candidates reflect inflammation, oxidative stress, mitochondrial injury or disruption of the blood-brain barrier. The attraction of biomarkers is clear: a rapid laboratory test could complement bedside examination and potentially identify hidden injury. But biological markers can be influenced by organ failure, infection, prematurity and the timing of sample collection. Most remain research tools rather than stand-alone tests for making high-stakes decisions.
Arnáez also highlights a fundamental problem in neonatal research: severity is often defined differently from one study to another. One investigation may use the initial neurological examination, another may classify infants according to cooling eligibility, and a third may rely on MRI, seizure burden or developmental outcomes at 18 or 24 months. These definitions are related, but they are not interchangeable. If studies enroll different populations under the same label of “moderate encephalopathy,” comparisons become difficult and the apparent effectiveness of treatments may be distorted. More consistent severity frameworks could improve clinical trials by allowing researchers to determine which infants benefit from an intervention, at what stage and under which biological conditions.
A more precise approach would combine several forms of evidence: perinatal history, early and repeated neurological examinations, continuous brain monitoring, neuroimaging, laboratory biomarkers and carefully selected developmental outcomes. Such a model would not replace clinical judgment or reduce an infant to a numerical score. Instead, it would recognize that neonatal encephalopathy is a dynamic syndrome with different causes, injury patterns and recovery pathways. The most useful measure of severity may ultimately be a profile that describes the infant’s current neurological state, the likely mechanism of injury and the direction of change over time.
This shift could also improve communication with families. A label such as “mild encephalopathy” may sound reassuring even though some infants with initially mild signs later develop seizures, developmental delays or learning difficulties. Conversely, an early severe presentation does not make an individual outcome inevitable. Families need explanations that acknowledge uncertainty while describing the evidence emerging from examinations, monitoring and imaging. Prognosis should be updated as new information becomes available, rather than being fixed by a single early assessment.
The broader significance extends beyond neonatal intensive care. Better measurement could support trials of treatments that complement cooling, including strategies aimed at reducing inflammation, stabilizing mitochondrial function, controlling seizures or promoting repair. It could also help distinguish infants with genuine hypoxic-ischemic injury from those whose encephalopathy has another cause, ensuring that therapies are tested in biologically appropriate groups. The challenge is to develop tools that are reliable across hospitals, practical during the narrow window when treatment can help and validated against meaningful long-term outcomes.
Neonatal encephalopathy is therefore moving from a diagnostic label toward a measurable process. The categories of mild, moderate and severe remain useful for urgent clinical decisions, but they are only the starting point. A combination of serial examinations, brain activity monitoring, imaging and molecular evidence may provide a more accurate account of what is happening inside the newborn brain. By looking beyond definitions, clinicians and researchers may be better equipped to identify injury early, tailor treatment and understand why infants with apparently similar beginnings can follow very different developmental paths.
Subject of Research: Measuring severity, prognosis and biological progression in neonatal encephalopathy.
Article Title: Beyond definitions: measuring severity in neonatal encephalopathy
Article References:
Arnáez, J. Beyond definitions: measuring severity in neonatal encephalopathy. Pediatr Res (2026). https://doi.org/10.1038/s41390-026-05378-6
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s41390-026-05378-6
Keywords: neonatal encephalopathy, hypoxic-ischemic encephalopathy, neonatal brain injury, therapeutic hypothermia, neuroimaging, electroencephalography, biomarkers, neonatal neurology, prognosis, newborn health
Tags: biological complexity of neonatal neurological conditionsclinical classification of neonatal encephalopathyhypoxic-ischemic encephalopathy diagnosisimpact of severity measurement on neonatal treatmentimportance of nuanced evaluation in neonatal carelimitations of traditional severity categorieslong-term outcomes of neonatal encephalopathymeasuring neonatal brain injuryneonatal encephalopathy severity assessmentneonatal neurological dysfunctionneonatal seizure and breathing disorder assessmentrole of therapeutic hypothermia in neonatal brain injury
