The study, published in The Lancet Regional Health – Europe and led by researchers from the Neonatal Critical Care Section of the European Society for Paediatric and Neonatal Intensive Care (ESPNIC), examined guidelines from eleven European countries: the United Kingdom, Italy, Spain, Switzerland, Germany, the Netherlands, France, Norway, Portugal, Slovakia and Turkey. A multidisciplinary panel of experienced clinicians independently scored each document using two of the most widely accepted appraisal instruments in medicine, AGREE-II and the RIGHT statement, in a blinded and anonymous process designed to eliminate bias between raters.

The panel itself was deliberately unconventional. Alongside neonatologists with at least a decade of experience in academic cooling centres, it included a paediatric intensivist, an obstetrician and an adult intensivist, reflecting the perinatal origins of the condition and the broader critical care experience with temperature control. None of the raters had been involved in producing any of the guidelines under scrutiny, and each evaluated the documents without knowing how their colleagues had scored them. Inter-rater agreement was strong, with intraclass correlation coefficients of 0.831 for quality and 0.909 for reporting completeness, lending statistical weight to the consensus findings.

The results were striking. On the AGREE-II instrument, which assesses six domains including purpose, stakeholder involvement, rigour of development, clarity, applicability and editorial independence, the German guidelines scored highest overall and were judged to be of medium quality, while the Norwegian guidelines scored lowest. But the most damning finding concerned the domain widely regarded as the most important: rigour of development. Not a single guideline achieved a high score here, and even the German document fell short of the fifty per cent mark. In practical terms, this means that none of the eleven documents demonstrated a transparent, systematic process for searching, appraising and grading the underlying evidence.

Reporting completeness, measured with the RIGHT statement across seven domains and thirty-seven items, was equally troubling. Total completeness ranged from just twenty to forty-nine per cent, and no guideline achieved satisfactory global completeness. The French guidelines performed best at the domain level, while the Swiss guidelines fared worst. Many guidelines lacked even basic information: most did not state whether they had undergone peer review, and the majority were silent on their consensus methodology and on how recommendations were graded. Only the French and Italian guidelines had been published in full as peer-reviewed articles, which the authors suggest may partly explain the generally moderate-to-low quality of documents that never faced academic scrutiny.

Beneath the methodological scores lie clinically consequential discrepancies. The guidelines disagree on the very thresholds used to decide which babies should be cooled, prescribing different cut-offs for base excess, pH and lactate, with some even advising clinicians not to trust cord blood gas values. They weigh the neurological examination differently, giving variable importance to the components of the Sarnat score or substituting different scoring systems, and they interpret amplitude-integrated EEG findings in inconsistent ways. These are not academic quibbles: the eligibility decision determines whether a newborn receives a complex, resource-intensive therapy that can alter the trajectory of an entire life.

The inconsistencies widen sharply in the grey areas that clinicians encounter daily. For late preterm infants, recent randomised trial data suggest that cooling may be associated with white matter injury and increased likelihood of death or disability, yet the guidelines offer divergent and weakly supported advice. For babies transferred to a referral centre after the crucial first six hours of life, a common scenario depending on geography and transport logistics, evidence suggests cooling may still help, but guidance again varies. Alarmingly, none of the eleven documents addressed the use of therapeutic hypothermia in babies requiring extracorporeal life support, leaving an entire category of the sickest patients without any documented direction.

The authors argue that this heterogeneity cannot be explained by the evidence itself, because the original randomised trials that established the therapy’s effectiveness used essentially similar inclusion criteria. Instead, local custom and differing interpretations have accumulated into recommendations that were never tested in dedicated trials. The consequences ripple outward: patients receive different care depending on where they are born, intensive care beds and healthcare costs are affected, and medico-legal exposure grows when practice diverges between centres and countries for reasons that the available evidence does not justify. The American Academy of Pediatrics has meanwhile reaffirmed that cooled newborns should meet the inclusion criteria of the published trials, underscoring how far some European documents have drifted.

The study team is careful to acknowledge its limits. The sample was a convenience set of guidelines from countries with the largest ESPNIC membership and does not cover every European nation, and healthcare organisation, reimbursement and regionalisation differ across jurisdictions in ways that may shape national documents. Still, the central message is unambiguous: no guideline evaluated should be applied directly without critical appraisal and modification. The authors call for national bodies, or better still continental organisations, to rebuild neonatal cooling guidance using the strict methodological standards already adopted in adult critical care medicine, and to commission studies in the grey areas where evidence is thin.

For clinicians at the bedside, the findings are a call to scepticism rather than despair. Therapeutic hypothermia remains firmly evidence-based and profoundly valuable, but the documents that translate that evidence into practice have not kept pace with the standards the field now demands. Reading guidelines critically, understanding where they rest on solid trials and where they rest on local habit, may be the most important skill a neonatal intensivist can bring to the cooling cot, at least until Europe’s next generation of guidelines earns the trust that babies and their families deserve.

Subject of Research: Quality and completeness of European clinical practice guidelines for neonatal therapeutic hypothermia

Article Title: Clinical practice guidelines for neonatal therapeutic hypothermia across European Society for Paediatric and Neonatal Intensive Care (ESPNIC) member countries: a multidisciplinary blind comparison

Article References: De Luca, D., Neri, C., Alonso-Ojembarrena, A., Ippolito, M., Milan, A., Piastra, M., Sanchez-Luna, M., Schettler, K. F., Zanin, A., Ghi, T., & Cortegiani, A. (2026). Clinical practice guidelines for neonatal therapeutic hypothermia across European Society for Paediatric and Neonatal Intensive Care (ESPNIC) member countries: a multidisciplinary blind comparison. The Lancet Regional Health – Europe, 71, Article 101874. https://doi.org/10.1016/j.lanepe.2026.101874

Image Credits: AI Generated

DOI: 10.1016/j.lanepe.2026.101874

Keywords: therapeutic hypothermia, neonatology, hypoxic-ischaemic encephalopathy, clinical practice guidelines, ESPNIC, AGREE-II, RIGHT statement, guideline quality, perinatal asphyxia, neonatal intensive care, Europe, evidence-based medicine

Perivascular spaces are, in healthy brains, almost invisible. They are the cuffs of cerebrospinal fluid that accompany penetrating arteries and veins into the brain’s substance, and in a typical newborn the amount of fluid within them is too small to register on clinical imaging. When these spaces balloon and multiply, however, they show up on T2-weighted magnetic resonance images as bright, linear or round dots, and on FLAIR sequences as dark counterparts of the same shapes. In adults, a heavy burden of enlarged perivascular spaces has been linked to cerebral small vessel disease, cognitive decline, and poor glymphatic clearance of toxic proteins. Until now, almost nothing was known about what these structures mean in the earliest days of life, particularly in babies whose brains have just endured the twin assaults of oxygen deprivation and the inflammatory firestorm that follows it.

The new study, published in the World Journal of Pediatrics, enrolled 319 newborns admitted to the hospital’s neonatology department between January 2020 and September 2023, all transferred within a week of birth. Using strict diagnostic criteria based on Apgar scores and umbilical artery blood gas analysis, the researchers divided the infants into a mild asphyxia group of 229 babies and a severe group of 90. Each child underwent brain magnetic resonance imaging on a 3.0 Tesla scanner, along with amplitude-integrated electroencephalography to capture the electrical rhythms of the injured cortex, and a battery of laboratory tests ranging from inflammatory markers to coagulation profiles. Two experienced radiologists, blinded to all clinical information, counted the enlarged perivascular spaces in the basal ganglia across four to six axial sections, producing a quantitative burden score for every infant.

The results were unambiguous. In the severe asphyxia group, lactate levels and C-reactive protein concentrations were significantly elevated compared with the mild group, and the number of enlarged perivascular spaces was also markedly higher. Across the whole cohort, the count of these spaces correlated positively with diastolic blood pressure and lactate, and negatively with Apgar scores at one, five, and ten minutes. More importantly, when the researchers built multivariable linear regression models to disentangle independent predictors, the number of enlarged perivascular spaces remained significantly associated with the severity of brain injury as graded on a simplified magnetic resonance score, with a P value of 0.0001 across all infants, 0.0248 in the mild group, and 0.0114 in the severe group. This held true even after accounting for gestational age, blood pressure, base excess, Apgar scores, and coagulation measures such as thrombin time and the international normalized ratio.

Intriguingly, the association was selective. The perivascular space count predicted the structural magnetic resonance score but not the amplitude-integrated electroencephalography grade, a measure of acute cortical electrical function. The authors interpret this dissociation as evidence that enlarged perivascular spaces capture structural, glymphatic damage rather than transient functional suppression. Electroencephalographic activity can be temporarily blunted by metabolic derangement and then recover, whereas the magnetic resonance image visualizes established injury to brain parenchyma. A marker of perivascular structural integrity, the researchers argue, aligns more naturally with the latter. This distinction matters clinically, because it suggests the marker adds a dimension of information that neither routine electrophysiology nor blood-based biomarkers of inflammation and neuronal injury can currently provide.

The mechanistic story behind the finding centers on the glymphatic system, the recently characterized network through which cerebrospinal fluid pulses along perivascular channels to flush metabolic waste from the brain, driven in part by the water channel protein aquaporin 4, which is polarized to the endfeet of astrocytes. Asphyxia triggers oxidative stress and a cytokine storm that disrupt this polarization, causing perivascular spaces to enlarge and glymphatic clearance to falter. The consequence may be self-reinforcing: as clearance fails, neurotoxic metabolites accumulate, further amplifying inflammation and astrocytic injury in a vicious cycle of progressive damage. The authors are careful to note that their observational design cannot establish causality; the enlarged spaces may actively contribute to injury or may simply be an epiphenomenon of the same upstream pathology. Either way, their count offers a non-invasive readout of a process that has been nearly impossible to monitor in living newborns.

Perhaps the most provocative result concerns sleep. Because mild asphyxial injury can damage the pineal gland, the source of melatonin, sleep disturbance has long been recognized as a sequela of perinatal asphyxia. The Suzhou team followed 258 of the 319 infants, using monthly parental questionnaires over six months, adapted from the Brief Infant Sleep Questionnaire and the Paediatric Sleep Questionnaire, to score seven sleep problems including prolonged sleep latency, insufficient sleep, irregular schedules, snoring, breathing difficulties, and frequent night wakings. After excluding age as a confounder by stratifying children into three age bands, the researchers found no relationship between perivascular space count and sleep in the full cohort or in the mild group. But in the severe asphyxia group, the correlation was significant: a higher number of enlarged perivascular spaces predicted more severe sleep disturbance, with a Spearman coefficient of 0.3365 and a P value of 0.0061.

That severity-specific pattern, the authors suggest, points to a mechanism beyond direct pineal injury. In severely asphyxiated infants, glymphatic damage may impede the clearance of toxic proteins, perpetuating brain injury and indirectly degrading sleep architecture, whereas in mild cases the perivascular spaces tend to shrink over time as the infant brain repairs itself, explaining the absence of a sleep correlation. The correlation is moderate rather than overwhelming, and the researchers acknowledge that infant sleep is multifactorial and cannot be reduced to a single biomarker. Still, the finding hints that a routine scan performed in the first week of life could flag babies at risk of sleep problems before parents notice anything wrong, opening a window for early behavioral and medical interventions during a period when sleep is critical to brain development.

The study has limitations that the authors confront directly. Ethical constraints precluded a healthy control group, though published data confirm that enlarged perivascular spaces are rare in healthy neonates, and the single-center design, moderate sample size, and reliance on parent-reported sleep questionnaires all warrant caution. The team calls for multicenter prospective cohorts incorporating diffusion tensor imaging along the perivascular space, an advanced technique that quantifies glymphatic flow directly, along with objective sleep monitoring and longitudinal follow-up. If those studies confirm the present findings, the humble perivascular space, long dismissed as an incidental dot on adult brain scans, could become a standard element of neonatal risk stratification, helping clinicians decide which asphyxiated newborns need the most aggressive neuroprotective care and which families should be counseled and monitored for the sleep disorders that may otherwise surface silently in the years after a difficult birth.

Subject of Research: Enlarged perivascular spaces as imaging biomarkers of perinatal asphyxia-induced brain injury and later sleep quality

Article Title: Correlation between perinatal asphyxia-induced brain injury, number of enlarged perivascular spaces, and sleep quality in later life

Article References: Correlation between perinatal asphyxia-induced brain injury, number of enlarged perivascular spaces, and sleep quality in later life. (n.d.). https://doi.org/10.1007/s12519-026-01093-8

Image Credits: AI Generated

DOI: 10.1007/s12519-026-01093-8

Keywords: perinatal asphyxia, enlarged perivascular spaces, glymphatic system, neonatal brain injury, MRI biomarker, sleep quality, hypoxic-ischemic encephalopathy, newborns, aquaporin 4, neonatology, amplitude-integrated EEG, World Journal of Pediatrics