Higher prevalence of autism in Taiwanese children born prematurely: a nationwide population-based study.

The reported prevalence of autism in preterm and full-term children varies partially because of small sample sizes. Moreover, little is known about the specific factors that contribute to the risk of autism in preterm children. We aimed to compare the prevalence of autism in preterm and full-term children and to identify neonatal risk factors for autism in preterm children using a large national health system database. We analyzed data from 1078 early preterm (<28 weeks of gestation or birth weight<1000 g), 28,947 later preterm (28-36 weeks), and 1,104,071 full-term (≥ 37 weeks) children who were 8-11 years old in 2009. The descending order of prevalence was early preterm (2.2%), later preterm (1.3%), and full-term (0.6%). The prevalence of autism was approximately 2-4 times higher in preterm children than in children born at full-term. The male-female ratio (4:1) in preterm and full-term children was not significantly different. Most of the children were first diagnosed with autism between 3 and 6 years old. Preterm children with autism were not diagnosed earlier than were full-term children. Regression analysis showed that male gender, a very low birth weight, and neonatal cerebral dysfunction were risk factors for autism in the preterm group. We conclude that autism is more prevalent in preterm children. Preventing extremely preterm birth and significant early brain insults may be helpful in reducing the risk of autism in preterm children.