Newborn jaundice: common and usually harmless, dangerous when it isn't

Newborn jaundice — the yellow tint from a build-up of bilirubin — is one of the most common conditions of the first days of life and in most babies it is harmless and passes on its own [s1]. Clinicians watch it closely for one reason: a small minority of newborns reach bilirubin levels high enough to cause permanent brain injury, and that severe outcome is what phototherapy exists to prevent [s1][s2]. Where good newborn care is available, that catastrophe is now rare; where it is not, it is not.

Why almost every newborn goes a little yellow

Jaundice becomes clinically visible once total serum bilirubin rises above about 7 mg per decilitre, and it is common enough to be the leading cause of newborn hospital readmission worldwide [s1]. Estimates put its prevalence at 25 to 50 percent of term infants, with a higher rate again in preterm babies, whose livers are even less ready to clear bilirubin [s1]. The important clinical split is between two forms: indirect (unconjugated) hyperbilirubinaemia, which can be neurotoxic, and direct (conjugated) hyperbilirubinaemia, which is rarely associated with neurological harm [s1]. It is the indirect form, at high levels, that threatens the brain — the reason the routine yellow of a healthy newborn is watched rather than ignored.

The rare, serious end: kernicterus

The danger of very high bilirubin is acute bilirubin encephalopathy and its permanent form, kernicterus — brain damage that can leave lifelong disability. A 2017 systematic review set out to measure that burden globally and found, first, how thin the data are: of 416 articles carrying at least one marker of severe neonatal jaundice, only 21 offered population-based estimates, and 76 percent of those (16 of 21) came from high-income countries [s2]. Where estimates existed, the disparity was stark. The incidence of severe neonatal jaundice per 10,000 live births was highest in the African region at 667.8 (95 percent confidence interval 603.4 to 738.5) and in the South-East Asian region at 251.3 (132.0 to 473.2), followed by the Eastern Mediterranean at 165.7 (114.6 to 238.9) and the Western Pacific at 9.4, against 4.4 (1.8 to 10.5) in the Americas and 3.7 (1.7 to 8.0) in Europe [s2]. Exchange transfusion — the emergency treatment for dangerous levels — followed the same pattern, at 186.5 per 10,000 live births in Africa and 107.1 in South-East Asia [s2].

A 2023 meta-analysis of hospital admissions confirmed the geography. Of 84 studies, 64 (76 percent) were from low- and lower-middle-income countries, and 14 percent of the jaundiced neonates in them had severe disease [s3]. Among all admitted newborns, severe jaundice ranged from 0.73 to 3.34 percent, exchange transfusion from 0.74 to 3.81 percent, and acute bilirubin encephalopathy from 0.16 to 2.75 percent — with the highest figures consistently in the African, South-East Asian and Eastern Mediterranean regions, and jaundice-related deaths reaching 1.49 percent [s3]. The severe outcome, in short, is not evenly distributed: it tracks access to timely newborn care.

What treatment does

The mainstay is phototherapy — blue light that converts bilirubin in the skin into forms the body can excrete — and its role is well established. A 2026 meta-analysis of 12 studies (8 randomised trials and 4 cohort studies) enrolling 1,144 neonates confirmed phototherapy as the cornerstone of management and quantified the gains from a more intensive delivery of it: intensive phototherapy lowered total serum bilirubin by a mean of 21.87 µmol/L more than conventional phototherapy (95 percent confidence interval 14.05 to 29.69) and shortened treatment by about 20 hours (mean difference 20.23 hours, 95 percent confidence interval 1.54 to 38.91) [s1]. The pooled trials spanned newborns of 30 to 39 weeks' gestation with starting bilirubin levels of 172 to 390 µmol/L [s1]. The review noted that the spread of effective phototherapy has coincided with a marked decline in acute bilirubin encephalopathy [s1]. For the small number of babies whose levels climb despite light, exchange transfusion remains the rescue treatment [s2].

How to read this

For most parents the reassuring reading is the correct one: newborn jaundice is expected, usually mild, and resolves without intervention [s1]. The reason it is nonetheless screened for and treated is that the rare severe case is devastating and preventable — and the global data show that where phototherapy and monitoring are available the damage is largely averted, while the burden of kernicterus and jaundice-related death is concentrated in the poorest health systems [s2][s3]. That pattern places newborn jaundice among the newborn-care questions we return to, from the newborn bloodspot screen to vitamin K at birth and safe infant sleep. This article is informational and is not medical advice; jaundice appearing in the first 24 hours, or a baby who is drowsy or feeding poorly, needs prompt assessment.