A US cohort study found that very preterm children with early cerebellar abnormalities or moderate-to-severe placental inflammation were more likely to screen positive for autism risk by age 5, highlighting potential opportunities for closer developmental surveillance.

Study: Early autism risk factors: a cohort study of children born very preterm at 5-years. Image Credit: wutianzeri / Shutterstock

A recent study published in the Journal of Perinatology found that children born very preterm (VPT) had a higher proportion of autism-related social communication difficulties at 5 years of age than term-born children. Among VPT children, cerebellar abnormalities detected at the age when they would normally have been term babies (term-equivalent age) and moderate-to-severe histologic chorioamnionitis were associated with higher autism screening risk, along with male sex and higher social risk.

Background

Autism is a neurodevelopmental condition with the most overt symptoms being impaired communication, difficulties in reciprocal social interaction, and restrictive or repetitive interests or behaviors. Some of these characteristics, including social and behavioral difficulties, poor executive functioning, and other neurodevelopmental issues, occur more often in children born preterm.

Autism risk among children born preterm has been estimated at 7% to 28%. Those born very preterm (VPT) are exposed to multiple prenatal, perinatal, and neonatal factors that may increase autism risk. However, the prognostic importance of these factors remains unclear, especially since most prior studies have been of limited size.

The current longitudinal cohort study from a US academic medical center aimed to assess how various maternal and infant risk factors were associated with autism risk at 5 years of age in children born VPT, defined in the study as birth at 32 weeks of gestation or earlier. This might help clinicians identify children who would benefit the most from increased surveillance and early intervention.

Study design

Researchers analyzed data from a longitudinal cohort of 315 very preterm infants and 172 term-born controls at a US-based academic medical center. They did not have access to data on autism diagnoses in this cohort. However, they used data from the Social Communication Questionnaire (SCQ), which is used to screen for autism risk in children aged 4 years and older.

An SCQ score of 15 or higher was considered indicative of autism risk; this was a screening threshold rather than a clinical diagnosis of autism.

Other possible risk factors include diabetes, hypertensive disorders of pregnancy (HDP), chorioamnionitis, Apgar scores, retinopathy of prematurity (ROP), bronchopulmonary dysplasia (BPD), sepsis, brain abnormalities, and breastfeeding history.

Their association with autism risk has been demonstrated in prior research, though this does not confirm causation.

Rather than examining individual risk factors independently, the researchers used elastic net regression. This statistical approach can identify factors associated with an outcome when multiple potentially related variables are considered simultaneously.

Higher autism-related screening scores among VPT children

A greater proportion of children born VPT (8.9%) had SCQ scores associated with autism risk (15 or higher), compared with 2.9% of term-born controls. The findings therefore indicate that a higher proportion of children born very preterm screened positive for autism risk.

Broader neurodevelopmental difficulties

A higher SCQ score was also associated with poorer adaptive, cognitive, behavioral, and executive functioning, consistent with difficulties commonly observed in autism. However, the authors cautioned that difficulties in language, cognition, executive functioning, and adaptive functioning could themselves contribute to elevated SCQ scores.

Factors associated with autism screening risk

Among the VPT group, several factors were independently associated with a positive autism-risk screen. Among the 28 VPT children with SCQ scores of 15 or higher, 50% had moderate-to-severe cerebellar abnormalities, while 35.7% had moderate-to-severe histologic chorioamnionitis.

Conversely, 20% of children with moderate-to-severe cerebellar abnormalities had high SCQ scores, versus 6.8% in the no-to-mild abnormality group. Similarly, 22.7% of infants with moderate-to-severe chorioamnionitis screened positive with the SCQ, compared to 6.8% of those with no or mild chorioamnionitis.

Each one-point increase in the cerebellar abnormality score at term-equivalent age was associated with approximately 34% higher odds of screening positive for autism risk. Moderate-to-severe histologic chorioamnionitis was associated with approximately fourfold higher odds of screening positive for autism.

Male sex and higher social risk were also identified as factors associated with autism risk.

Cerebellar abnormalities and autism risk

Cerebellar abnormalities are associated with a higher autism risk in VPT, as shown by existing literature. Cerebro-cerebellar connections are key to the cognitive and socioemotional processes that underlie social intelligence.

They are also sensitive to early-life disruption, which may result in impaired recognition of mood and perspective in others and disinhibited social behavior, all of which have been associated with autism.

In addition, the cerebellum is also involved in repetitive behaviors, altered sensory perception, and cognitive ability, all of which are observed in autism. These observations provide biological context for the association identified in the study. However, the authors noted that other influences, including genetic liability, parental psychiatric history, maternal mental health, and postnatal environmental factors, could contribute to the finding.

Chorioamnionitis and autism risk

Chorioamnionitis is inflammation of the chorioamniotic membranes associated with the placenta. Histologic chorioamnionitis refers to inflammatory changes identified on examination of placental or membrane tissue after delivery.

Prior studies have reported that VPT infants with higher screening scores or a diagnosis of autism have an increased rate of chorioamnionitis. This condition results in maternal and fetal inflammation, which may affect fetal brain development.

Limitations

The study findings need cautious interpretation given that autism diagnoses were not available. Rather, autism risk was screened for. In addition, the researchers did not capture exposure to these early risk factors in term babies.

Moreover, these risk factors might reflect a more systemic or whole-brain impact caused by prematurity rather than being specific to autism. The factors may also co-occur and interact, rather than operating independently. The lack of neuroimaging at the current 5-year visit prevents the correlation of autism risk with brain maturation and neuroplasticity.

Conclusion

The findings suggest that identifying VPT children with these two risk factors (cerebellar abnormalities and moderate-to-severe histologic chorioamnionitis), both of which can be detected early in life, might facilitate earlier autism screening and developmental surveillance. The authors suggest that this could allow timely developmental interventions during a period of high neuroplasticity, potentially improving neurodevelopmental outcomes.

“Further, these results underscore the multifactorial nature of autism risk in VPT children, involving both biological and social determinants.”

Journal reference:

• Stone-Heaberlin, M., Blackburn, A. D., Tamm, L., et al.(2026). Early autism risk factors: a cohort study of children born very preterm at 5-years. Journal of Perinatology.DOI:10.1038/s41372-026-02856-x, https://www.nature.com/articles/s41372-026-02856-x