A new study may provide some intriguing new insight into autism, after finding that jaundice in newborns appears linked to a higher risk for autism.

The study, which appears in the journal Pediatrics, looked at more than 700,000 children in Denmark.

It found that babies who developed jaundice were 67 per cent more likely to be diagnosed with autism during early childhood.

The study authors say the study only linked jaundice and autism; it didn't show that jaundice caused autism. They suggest there could be a number of factors that predispose babies to both the development of jaundice and autism.

Newborns with jaundice develop yellowish skin and yellowy eyes, due to an excess of bilirubin, a byproduct of the breakdown of red blood cells. The condition is seen in 60 per cent of infants born to term and in about 80 per cent of babies born prematurely and is usually thought to be the result of an immature liver.

For most infants, the condition resolves within the first week of life, though it sometimes requires treatment with phototherapy, using lights to help break down bilirubin in the skin.

Treatment of severe cases is important since prolonged exposure to high bilirubin levels is toxic to the brain and can cause lifelong developmental problems, such as cerebral palsy.

For this study, scientists looked at data on almost 734,000 live births in Denmark between 1994 and 2004. Among them, about 35,000 were diagnosed with neonatal jaundice. During the study period, 1,721 children were diagnosed with a psychological development disorder, and 4,257 children died.

Being exposed to jaundice for children born full-term resulted in a 56 to 88 per cent increased risk of acquiring a psychological development disorder and a 67 per cent increased risk for autism.

The risk of autism was higher if the mother had had previous children, or if the child was born between October and March. The risk for autism disappeared if the child was a firstborn child or was born between April and September.

The authors suggest the seasonal difference may be due to different levels of exposure to sunlight, which has an effect on jaundice, or due to infections.

The difference in risk in firstborn versus subsequent children could be due to different levels of antibodies in women who have had multiple pregnancies.

It could also reflect different levels of access to health care in the first days after delivery.

In Denmark, women with healthy term newborns who have already had children are discharged soon after delivery. Women having their first child remain in the hospital for three to four days, and so jaundice may be diagnosed while the infant is still in the hospital.