Prematurity and autism: understanding the gestational-age association

In a matched cohort of 1,861 children with autism and 18,610 controls, the odds of autism increased progressively as gestational age decreased. The dose–response pattern remained after adjustment, but the retrospective study cannot determine an individual child’s outcome or prove that prematurity alone caused autism. Its practical implication is vigilant, proportionate developmental screening after preterm birth.

Bottom line: most children born preterm are not diagnosed with autism. Gestational age is a population-level risk marker that can help target follow-up, not a prediction for one child.

Study design

The nationwide retrospective cohort included children born from 2011 through 2023. Autism diagnoses recorded in electronic health records were matched to ten controls on demographic and maternal factors. Researchers grouped gestational age into clinically meaningful categories and used conditional logistic regression.

What the dose–response looked like

Compared with term birth at 37–42 weeks, adjusted odds increased as gestational age decreased: approximately 1.37 at 36 weeks, 1.67 at 34–35 weeks, 2.14 at 32–33 weeks, 2.95 at 28–31 weeks, and 4.37 below 28 weeks.

The graded pattern strengthens the case that prematurity-related biology may contribute to risk. Proposed pathways include white-matter vulnerability, inflammation, oxidative stress, and altered connectivity. These mechanisms were not directly tested in the database analysis.

What the study does not show

Autism was identified from clinical diagnostic codes rather than a single standardized research assessment. Residual confounding is possible. Prematurity can co-occur with infections, pregnancy complications, genetic factors, neonatal illness, and other exposures that may influence neurodevelopment.

An association between two outcomes does not justify blaming parents, obstetric decisions, or an individual neonatal event.

Practical implication

The result supports systematic developmental surveillance for children born preterm, with intensity appropriate to gestational age and clinical history. Screening helps identify support needs earlier; it is not a statement that autism is inevitable or should be framed only as an adverse outcome.

Primary source