Light Therapy for Jaundice in Very Premature Babies: Does More Time Under the Lights Cause Harm?

What a 2026 Swedish study of nearly 5,000 babies born before 32 weeks found about how long phototherapy lasts, survival and complications

A large 2026 Swedish study followed 4,970 babies born between 22 and 31 weeks of pregnancy and asked whether spending more days under phototherapy lights for jaundice made the babies more likely to die in the following weeks. It found that it did not: babies treated for six to seven days were no more likely to die than babies treated for three days or fewer. The babies who had the longest treatment did have more serious complications such as bleeding in the brain and lung disease, but this may be because they were sicker to begin with, and the researchers still recommend avoiding light treatment that is not needed [1].

Why jaundice and its treatment matter so much for premature babies

Almost every baby born very early develops jaundice, the yellow tint of the skin and eyes that comes from a natural substance called bilirubin. Bilirubin is made when the body recycles old red blood cells, and a premature baby's liver is still too immature to clear it quickly. In small amounts it does no harm, but very high levels can injure the brain, which is why medical teams watch bilirubin so closely in these babies.

The standard treatment is phototherapy: special blue light shone on the baby's skin changes bilirubin into a form the body can get rid of more easily. It is one of the most widely used treatments in newborn care, and in the past few decades it has made dangerous jaundice and the need for blood exchange transfusions much less common [2][3]. In very premature babies it is used a great deal. A study in Norway found that more than 80 in every 100 extremely premature babies received it, and the authors called for greater care in how it is used for the most fragile infants [4]. In the Swedish study described here, 95 out of every 100 babies born before 32 weeks had some phototherapy [1].

How doctors came to ask whether too much light could be a problem

For a long time, phototherapy was seen as very safe. Over the decades, though, doctors began to wonder whether the smallest, sickest babies might be affected differently. An American study from the 1980s suggested that more babies under about 2 pounds 3 ounces (1,000 grams) died after phototherapy, although sicker babies are also more likely to be given it [5].

To test the question properly, researchers in the United States organised a clinical trial through a national research network, in which extremely tiny babies were randomly assigned to gentler or stronger phototherapy [6]. Overall, the trial did not find that stronger treatment meant more babies died before going home, although a later look at the smallest babies, weighing 501 to 750 grams, suggested that about 19% more of them died with stronger treatment [7][1]. Newer LED lights, which can deliver stronger light, raised fresh questions about whether more powerful treatment was always better for very small babies [8]. Scientists had also seen that blood flow to a baby's brain speeds up when phototherapy starts, and in some babies did not return to normal afterwards [9]. Parents and doctors were left with a treatment they needed but could not be completely sure about.

What the Swedish researchers did

The Swedish team used a national database called the Swedish Neonatal Quality Register, which collects information on every baby admitted to a neonatal unit and has been checked for its accuracy [10]. By 2024 all 38 neonatal units in Sweden were sending in daily information. The researchers looked at babies born between November 2015 and December 2024 at 22 to 31 weeks of pregnancy. After removing babies who died in their first week, babies with major birth defects, and babies with missing information, 4,970 babies were left [1].

They sorted the babies into three groups by how many days in the first week of life they spent under phototherapy: up to three days (1,995 babies), four to five days (1,921 babies) and six to seven days (1,054 babies). They then looked at two main things: deaths between the 8th and 27th day of life, and serious complications including bleeding in the brain, lung disease that lasts after birth, a bowel illness called necrotizing enterocolitis, a heart-vessel problem called patent ductus arteriosus that needs treatment, and eye disease of prematurity. Because babies in the three groups were not alike, the team used statistics to account for differences such as being a twin, the mother's blood-pressure condition, the type of delivery, the baby's sex and gestational age, how well the baby was doing at birth, and how big the baby was for its age [1].

What they found

On the most important question, the news was good. Among babies with up to three days of phototherapy, 1.7 in every 100 died between days 8 and 27. Among those with four to five days it was 2.9 in every 100, and among those with six to seven days it was 4.6 in every 100. At first glance that looks like a rising trend, but the babies who had longer treatment were generally sicker, and once the researchers allowed for this the difference was gone: longer phototherapy was not linked to a higher chance of dying [1].

Complications were different. Serious complications happened in about 11 in every 100 babies with up to three days of treatment, 17 in every 100 with four to five days, and 28 in every 100 with six to seven days. Even after adjusting, longer treatment was linked with more of these complications, mainly bleeding inside the brain and long-term lung disease in babies born at 26 to 31 weeks. This was not seen in the very smallest babies, born at 22 to 25 weeks, and bowel illness and eye disease were not linked to treatment length [1].

How to read these findings

It is important not to jump to the conclusion that the lights caused these complications. Bleeding in the brain can itself cause more jaundice, so a baby who has had a bleed may need more days of phototherapy, not the other way around. Babies in the longest-treatment group were also more likely to have had low scores at birth. The study also found that babies with the highest bilirubin levels had more complications, which points to sicker babies needing more treatment rather than treatment making babies sicker [11][1]. The authors say that some of the complications may have arisen this way.

Still, it is possible that some effect of the treatment exists. Researchers have described side effects of phototherapy such as changes in body temperature, in fluid balance, in blood flow and in the time babies spend away from their parents [12]. Some of these things are also known to play a part in long-term lung disease [13]. The study had limits: it did not record how many hours a day the light was on or how strong it was, it did not follow the children as they grew up, and Sweden has excellent neonatal care that may differ from other places [1].

What this means for families

For families whose very premature baby is under the lights, the main message is reassuring. Needing phototherapy for several days does not, on this evidence, make it more likely that a baby will die, and phototherapy remains an important and effective treatment that should not be refused or stopped early when a baby needs it. You can ask your baby's team how high the bilirubin is, what level they are aiming for, and when they expect to turn the lights off. The team will also be watching for other problems, such as bleeding in the brain, because those can themselves affect how long a baby needs light.

The second message is that doctors try to use only as much light as is needed. The authors recommend avoiding unnecessary, long phototherapy in very premature babies. Phototherapy also means more time in an incubator or cot, and it can get in the way of skin-to-skin cuddles, so finishing treatment as soon as it is safe is good for babies and for bonding.

What researchers are working on next

Scientists are now testing gentler ways of giving the same benefit. In one trial of babies under 1,000 grams, turning the lights on and off in cycles, rather than keeping them on all the time, shortened the treatment substantially without a meaningful difference in the highest bilirubin level reached [14]. A review of trials on on-and-off versus continuous light found that continuous light lowers bilirubin a little more, but did not look at possible harms [15]. A larger trial of on-and-off phototherapy in babies born before 27 weeks is registered, and Sweden's national guidelines from July 2025 already recommend 30-minute on-and-off treatment for the most premature babies [1]. Researchers also want to follow children into school age to see how they do in the long run.

Summing up

This large Swedish study offers families a balanced message: longer phototherapy was not linked to more deaths in the weeks after the first week of life, it was linked to more serious complications that may partly reflect sicker babies, and the best advice for now is to give phototherapy when needed, but no longer than necessary.

References

  1. Deschmann E, Håkansson S, Söderling J, Norman M. Phototherapy, Morbidity, and Mortality in Very Preterm Newborns. JAMA Netw Open. 2026;9(5):e2614107. doi:10.1001/jamanetworkopen.2026.14107 ↩
  2. Hansen TWR, Maisels MJ, Ebbesen F, et al. Sixty years of phototherapy for neonatal jaundice – from serendipitous observation to standardized treatment and rescue for millions. J Perinatol. 2020;40(2):180-193. doi:10.1038/s41372-019-0439-1 ↩
  3. Maisels MJ, McDonagh AF. Phototherapy for Neonatal Jaundice. N Engl J Med. 2008;358(9):920-928. doi:10.1056/NEJMct0708376 ↩
  4. Mreihil K, Benth JŠ, Stensvold HJ, Nakstad B, Hansen TWR; Norwegian NICU Phototherapy Study Group; Norwegian Neonatal Network. Phototherapy is commonly used for neonatal jaundice but greater control is needed to avoid toxicity in the most vulnerable infants. Acta Paediatr. 2018;107(4):611-619. doi:10.1111/apa.14141 ↩
  5. Lipsitz PJ, Gartner LM, Bryla DA. Neonatal and infant mortality in relation to phototherapy. Pediatrics. 1985;75(2 Pt 2):422-426. doi:10.1542/peds.75.2.422 ↩
  6. Morris BH, Oh W, Tyson JE, et al. Aggressive vs. Conservative Phototherapy for Infants with Extremely Low Birth Weight. N Engl J Med. 2008;359(18):1885-1896. doi:10.1056/NEJMoa0803024 ↩
  7. Tyson JE, Pedroza C, Langer J, et al. Does aggressive phototherapy increase mortality while decreasing profound impairment among the smallest and sickest newborns? J Perinatol. 2012;32(9):677-684. doi:10.1038/jp.2012.64 ↩
  8. Arnold C, Pedroza C, Tyson JE. Phototherapy in ELBW newborns: Does it work? Is it safe? The evidence from randomized clinical trials. Semin Perinatol. 2014;38(7):452-464. doi:10.1053/j.semperi.2014.08.008 ↩
  9. Benders MJ, van Bel F, van de Bor M. The effect of phototherapy on cerebral blood flow velocity in preterm infants. Acta Paediatr. 1998;87(7):786-791. doi:10.1111/j.1651-2227.1998.tb01748.x ↩
  10. Norman M, Källén K, Wahlström E, Håkansson S; SNQ Collaboration. The Swedish Neonatal Quality Register – contents, completeness and validity. Acta Paediatr. 2019;108(8):1411-1418. doi:10.1111/apa.14823 ↩
  11. Solis-Garcia G, Raghuram K, Augustine S, et al. Hyperbilirubinemia among infants born preterm: peak levels and association with neurodevelopmental outcomes. J Pediatr. 2023;259:113458. doi:10.1016/j.jpeds.2023.113458 ↩
  12. Faulhaber FRS, Procianoy RS, Silveira RC. Side Effects of Phototherapy on Neonates. Am J Perinatol. 2019;36(3):252-257. doi:10.1055/s-0038-1667379 ↩
  13. Ambalavanan N, Deutsch G, Pryhuber G, Travers CP, Willis KA. The evolving pathophysiology of bronchopulmonary dysplasia. Physiol Rev. 2026;106(1):197-237. doi:10.1152/physrev.00042.2024 ↩
  14. Arnold C, Tyson JE, Pedroza C, et al. Cycled Phototherapy Dose-Finding Study for Extremely Low-Birth-Weight Infants: A Randomized Clinical Trial. JAMA Pediatr. 2020;174(7):649-656. doi:10.1001/jamapediatrics.2020.0559 ↩
  15. Gottimukkala SB, Lobo L, Gautham KS, et al. Intermittent phototherapy versus continuous phototherapy for neonatal jaundice. Cochrane Database Syst Rev. 2023;3(3):CD008168. doi:10.1002/14651858.CD008168.pub2 ↩