To Örebro University

AIM: Neonatal jaundice may increase the likelihood of autism through its influence on neurodevelopment. We assessed whether neonatal jaundice increases the risk for autism while accounting for confounding of perinatal risk factors.

METHOD: We used a population-based cohort of 1 669 830 individuals born in Sweden between 1992 and 2010. Information on jaundice and autism was retrieved from Swedish registers. Participants were followed from the age of 3, with available follow-up until 2013. We used Cox regression to estimate the hazard ratios (HRs) relating autism to neonatal jaundice.

RESULTS: Prior to accounting for covariates such as prenatal and perinatal factors, we observed an association between neonatal jaundice and autism (HR = 1.48, 95% CI: 1.37-1.60). This association significantly weakened after adjusting for potential confounders, particularly gestational age and birth weight. After adjusting for all covariates, the association nullified (HR = 1.02, 95% CI: 0.93-1.12).

CONCLUSION: The association between neonatal jaundice and autism is confounded, in particular by preterm birth. Our results highlight the importance of accounting for perinatal confounders when investigating the link between early life factors and autism, which may not always be genetic or familial.