Background and Aims Children born preterm face an increased risk of autistic disorders.

We examined whether the association between preterm birth and risk of autistic disorders could be explained by pregnancy complications or neonatal morbidity.

Methods Swedish, population-based, case-control study including 1216 cases with autistic disorders born between 1987 and 2002, and 6080 controls matched with respect to gender, birth year, and birth hospital. Associations between gestational age and autistic disorders were assessed and adjusted for maternal, birth and neonatal characteristics.

Results Compared with infants born at term, the unadjusted odds ratios (ORs) for autistic disorders among very and moderately preterm infants were 2.05 [95% CI: 1.26–3.34] and 1.55 [95% CI: 1.22–1.96], respectively.

In analyses controlled for maternal, pregnancy, and birth characteristics, ORs were reduced to 1.48 [95% CI: 0.77–2.84] and 1.33 [95% CI: 0.98–1.81], respectively.

Adding also neonatal complications to the analyses, ORs were 0.98 [95% CI: 0.45–2.16] and 1.25 [95% CI: 0.90–1.75], respectively.

Reductions in risks of autistic disorders related to preterm birth were primarily attributable to preeclampsia, small-for-gestational age birth, congenital malformations, low Apgar scores at 5 minutes, and intracranial bleeding, cerebral edema, or seizures in the neonatal period. Neonatal hypoglycemia, respiratory distress, and neonatal jaundice were associated with increased risk of autistic disorders for term but not preterm infants.

Conclusions The increased risk of autistic disorders related to preterm birth is mediated primarily by prenatal and neonatal complications that occur more commonly among preterm infants.