Cooling Treatment for Babies Born a Few Weeks Early: A Study Finds It Does Not Help

What a large United States trial of body cooling in babies born at 33 to 35 weeks with a birth-related brain injury means for families

What This Study Found

A large American study has tested whether cooling therapy — a treatment that protects the brains of full-term babies who were deprived of oxygen around birth — also helps babies born a few weeks early, at 33 to 35 weeks of pregnancy. Among 168 such babies, cooling did not improve survival or development by around two years of age, and there are signs it may have made outcomes worse. Doctors now have real evidence where, until recently, they had almost none.

Why This Question Mattered

Sometimes a baby's supply of oxygen and blood is interrupted shortly before, during, or just after birth. When that happens, the brain can be injured, and in the first hours and days the baby may be unusually sleepy or floppy, may not feed, and may have seizures. Doctors call this condition neonatal encephalopathy — "encephalopathy" simply means the brain is not working normally. It is frightening for families precisely because the injury is invisible: the baby looks whole, and yet something has gone wrong in the one organ nobody can see.

For most of the twentieth century, there was no treatment at all. Care meant supporting the baby's breathing, blood pressure, and blood sugar, controlling seizures, and waiting. That changed in 2005, when two large trials showed that deliberately lowering a newborn's body temperature by about three and a half degrees Celsius for three days could reduce the chance of dying or being left with serious disability. One, the Cool Cap trial, cooled the head using a special cap and was run across 25 hospitals in the United Kingdom, the United States, New Zealand, and Canada [1]. The other, run by the United States National Institute of Child Health and Human Development (NICHD) Neonatal Research Network, cooled the baby's whole body to 33.5 °C for 72 hours [2]. The Total Body Hypothermia for Neonatal Encephalopathy Trial (TOBY), across 42 hospitals in the United Kingdom, Hungary, Israel, Finland, and Sweden, confirmed the benefit [3], and a review pooling all the evidence made cooling standard care in wealthy countries [4]. The idea is that cooling slows the chemical chain reaction that continues to damage brain cells for hours after the original oxygen shortage — buying the brain time.

Here is the gap. Every one of those trials enrolled only babies born at 36 weeks of pregnancy or later. Babies born even slightly earlier were left out on purpose, because their bodies control temperature less well, their brains are at a different stage of development, and they face other complications of prematurity. But babies born at 33, 34, and 35 weeks also suffer oxygen deprivation at birth, and their parents and doctors still had to decide something. Over the years, many hospitals began offering cooling to these slightly premature babies anyway, outside the evidence — a practice described in hospital reports that could show it appeared broadly safe in the short term, but could never show whether it actually helped [5], [6]. Families in that situation were, in effect, being offered a treatment nobody had tested for their baby. That is the gap this new trial set out to close.

How the Study Was Done

The trial, called the Preemie Hypothermia for Neonatal Encephalopathy trial, was run at 19 hospitals across the United States belonging to the NICHD Neonatal Research Network [7]. Babies were enrolled between July 2015 and September 2020, and the last child completed follow-up in December 2022.

To join, a baby had to be born between 33 and just under 36 weeks of pregnancy, be less than six hours old, and show clear signs both of oxygen deprivation — measured through blood acid levels, or through a difficult birth combined with low Apgar scores or a prolonged need for breathing support — and of moderate or severe brain dysfunction on a standardized neurological examination. That examination is based on a system first described in 1976 and used worldwide ever since, checking the baby's alertness, muscle tone, posture, and reflexes [8]. Because a baby born at 34 weeks naturally behaves differently from one born at 40 weeks, the doctors adjusted parts of the examination and required that the baby's level of alertness be clearly abnormal.

In total 168 babies were randomly assigned — like a coin toss, so that neither the family nor the doctor chose — with 88 receiving cooling and 80 receiving carefully maintained normal temperature [7]. Cooled babies were held at 33.5 °C for 72 hours using a cooling blanket, then warmed slowly, half a degree per hour. The comparison group was not simply left alone: their temperature was actively kept at 37.0 °C, and if a baby became too warm, staff were required to cool them back down, because fever is known to be bad for an injured brain. Parents knew which group their baby was in, but — importantly — the specialists who assessed the children nearly two years later did not, so their judgments could not be influenced.

The main question was how many children had died or had a moderate or severe disability by 18 to 22 months of age, assessed through developmental testing, a neurological examination, and checks of movement, vision, and hearing. Because this situation is uncommon at these weeks of pregnancy, the researchers could not gather thousands of babies. They therefore used a statistical approach called Bayesian analysis, which answers the question "how probable is it that this treatment helps?" rather than producing the familiar single "p-value." Results are reported as probabilities.

What the Results Showed

Complete two-year results were available for 152 of the 168 children. Death or moderate-to-severe disability occurred in 29 of 83 cooled children (35%) and 20 of 69 children kept at normal temperature (29%) [7]. Put plainly: cooling did not help. The analysis put the probability that cooling was beneficial at just 26% — meaning a 74% probability that it made things worse.

The difference came mainly from deaths rather than from disability. Eighteen of 83 cooled children (22%) died, compared with 9 of 69 (13%) in the comparison group, giving an 87% probability that cooling increased the risk of death. Among children who survived, disability rates were very similar in the two groups: 11 of 83 versus 11 of 69. These are small numbers, and the researchers are careful to say the ranges of uncertainty around them are wide — but the balance of probability points away from benefit, not toward it.

Side effects during the three days of treatment were mostly similar between the groups. The one clear difference was that cooled babies needed a breathing machine for longer. The researchers also noticed something specific to these smaller babies: 32 of the cooled infants dipped below 32 °C, colder than intended — for less than an hour in 26 of them — because premature babies are less able to hold their temperature steady. Partway through the study, the independent safety committee required extra precautions to prevent this. Even when every one of those babies was set aside from the analysis, cooling still showed no meaningful benefit.

What This Means for Families

The researchers' conclusion is plainspoken: cooling begun in the first six hours "did not reduce death or disability" in these babies, and they state that cooling "is not indicated" for babies born at 35 weeks or earlier with this kind of brain injury [7]. If your baby is born a few weeks early with a birth-related brain injury, the current best evidence is that a cooling blanket is not the right answer — and that is genuinely useful information, because it spares babies three days of an intensive treatment that does not appear to work for them.

What the trial does support is careful, deliberate temperature control: keeping the baby at a normal 37 °C and never letting a fever go unchecked. That, along with breathing support, seizure treatment, and early brain imaging, is what good care looks like for these babies today.

If your child was cooled at 34 or 35 weeks in the past, this result is not a judgment on the decision that was made. Your medical team acted on the best information available at the time, which was thin. That is exactly the problem trials like this one exist to fix.

What Researchers Are Working On Next

Related studies in full-term babies have already tested whether cooling deeper or for longer works better — it does not [9] — and whether starting after six hours still helps, with only limited signs of benefit [10]. For preterm babies, the next steps include brain scan results from this same trial, which may show whether cooling changed the pattern of injury even without changing outcomes, and a closer look at whether the 35-week babies behaved differently from the 33-week babies. Independent commentators have asked whether cooling helps at these ages at all [11] and have urged real caution before cooling any baby under 36 weeks [12]. International expert groups that write newborn resuscitation guidance continue to base their recommendations on the trials in full-term babies [13]. Meanwhile, researchers are searching for treatments that could protect the developing preterm brain in ways that cooling apparently cannot.

References

  1. Gluckman PD, Wyatt JS, Azzopardi D, et al. Selective head cooling with mild systemic hypothermia after neonatal encephalopathy: multicentre randomised trial. The Lancet. 2005;365(9460):663–670. doi:10.1016/S0140-6736(05)17946-X17946-X) ↩
  2. Shankaran S, Laptook AR, Ehrenkranz RA, et al.; National Institute of Child Health and Human Development Neonatal Research Network. Whole-body hypothermia for neonates with hypoxic-ischemic encephalopathy. New England Journal of Medicine. 2005;353(15):1574–1584. doi:10.1056/NEJMcps050929 ↩
  3. Azzopardi DV, Strohm B, Edwards AD, et al.; TOBY Study Group. Moderate hypothermia to treat perinatal asphyxial encephalopathy. New England Journal of Medicine. 2009;361(14):1349–1358. doi:10.1056/NEJMoa0900854 ↩
  4. Jacobs SE, Berg M, Hunt R, et al. Cooling for newborns with hypoxic ischaemic encephalopathy. Cochrane Database of Systematic Reviews. 2013;2013(1):CD003311. doi:10.1002/14651858.CD003311.pub3 ↩
  5. Rao R, Trivedi S, Vesoulis Z, et al. Safety and Short-Term Outcomes of Therapeutic Hypothermia in Preterm Neonates 34–35 Weeks Gestational Age with Hypoxic-Ischemic Encephalopathy. The Journal of Pediatrics. 2017;183:37–42. doi:10.1016/j.jpeds.2016.11.019 ↩
  6. Kim SH, El-shibiny H, Inder T, El-Dib M. Therapeutic hypothermia for preterm infants 34–35 weeks gestational age with neonatal encephalopathy. Journal of Perinatology. 2024;44(4):528–531. doi:10.1038/s41372-024-01874-x ↩
  7. Faix RG, Laptook AR, Shankaran S, et al. Whole-Body Hypothermia for Neonatal Encephalopathy in Preterm Infants 33 to 35 Weeks' Gestation: A Randomized Clinical Trial. JAMA Pediatrics. 2025;179(4):396–406. doi:10.1001/jamapediatrics.2024.6613 ↩
  8. Sarnat HB, Sarnat MS. Neonatal encephalopathy following fetal distress. A clinical and electroencephalographic study. Archives of Neurology. 1976;33(10):696–705. doi:10.1001/archneur.1976.00500100030012 ↩
  9. Shankaran S, Laptook AR, Pappas A, et al. Effect of Depth and Duration of Cooling on Death or Disability at Age 18 Months Among Neonates With Hypoxic-Ischemic Encephalopathy: A Randomized Clinical Trial. JAMA. 2017;318(1):57–67. doi:10.1001/jama.2017.7218 ↩
  10. Laptook AR, Shankaran S, Tyson JE, et al. Effect of Therapeutic Hypothermia Initiated After 6 Hours of Age on Death or Disability Among Newborns With Hypoxic-Ischemic Encephalopathy: A Randomized Clinical Trial. JAMA. 2017;318(16):1550–1560. doi:10.1001/jama.2017.14972 ↩
  11. Spahic H, Zoubovsky SP, Dietz RM. EBNEO Commentary: Is Therapeutic Hypothermia Beneficial to Infants Born Between 33 and 35 Weeks Gestation? Acta Paediatrica. 2025;114(7):1742–1743. doi:10.1111/apa.70098 ↩
  12. Sewell E, Malhotra A, Gunn AJ. Therapeutic hypothermia in infants <36 weeks gestation: proceed with caution. Pediatric Research. 2026. doi:10.1038/s41390-026-04912-w ↩
  13. Liley HG, Weiner GM, Wyckoff MH, et al. Neonatal Life Support: 2025 International Liaison Committee on Resuscitation Consensus on Science With Treatment Recommendations. Pediatrics. 2026;157(1):e2025074766. doi:10.1542/peds.2025-074766 ↩