Predicting Autism Spectrum Disorder in Very Preterm Infants

Children born preterm are known to be at higher risk for developing autism spectrum disorder (ASD) compared with their term-born counterparts, with an estimated 7% prevalence of ASD based on a recent large prospective study and a meta-analysis.^1,2  This high prevalence is in contrast to the currently estimated prevalence of ASD in the United States of 1.8% in the general population.³  ASD has been shown to be associated with a variety of prenatal, perinatal, and neonatal risk factors, including a variety of maternal health risk factors and medications and neonatal risk factors such as seizures, birth asphyxia, and low birth weight.⁴  Previously identified perinatal risk factors for ASD specific to preterm infants include factors such as low birth gestational age and birth weight, intracranial hemorrhage, and acute and chronic lung disease (CLD).⁵ 

The article by Chen et al⁶  in this issue of Pediatrics provides the first prospectively obtained data regarding whether there is an early developmental trajectory of prematurely born children that predicts who will develop ASD. The authors tested 319 preterm children prospectively with Bayley Scales of Infant Development examinations at 6, 12, and 24 months and used group-based trajectory modeling to assess whether early-life developmental trajectory predicted autism at 5 years of age. The approach of looking at developmental trajectory has been used in other high-risk populations, such as infant siblings of children with ASD⁷  or those with a specific genetic disorder (tuberous sclerosis complex) with a high prevalence of ASD.⁸ 

The authors provide the first data revealing that although a small percentage of preterm infants who develop ASD have a similar early-life trajectory to that of term-born children, with decline in mental development from age 12 to 24 months,^9,10  the highest-risk group was identified as having low cognitive scores at 6 months, with further decline over time, allowing for early identification and intervention. The converse finding that infants with low cognitive scores who improve to >85 and those with stably high cognitive scores are at lower risk of developing ASD enables the clinician to provide reassurance to families.

Their analysis also illuminates risk factors for ASD related to preterm birth by comparing the 29 children who developed ASD with the 290 children without ASD. Notably, their study identified both nonmodifiable (eg, male sex, gestational age, and birth weight) and potentially modifiable risk factors (eg, CLD and duration of oxygen therapy) for the development of ASD. As the authors discussed, CLD is known to be a risk factor for developmental delay and cognitive impairment and/or disability, but it is unclear the extent to which the risk associated with CLD is related to brain injury and altered brain development.

One acknowledged limitation of the study was the lack of neuroimaging data, so it is unknown whether there was a contribution of identifiable brain injury to the development of ASD in their subjects. It is likely that at least some of the infants in the low cognitive score group had easily detected brain injury, such as large cerebellar injury¹¹  or cerebral injury and/or impaired brain development,^12–14  both of which are associated with low IQ and are suspected to increase the risk of ASD. Additionally, there were no data regarding genetic risk factors for ASD, which could contribute a “second hit” to risk factors related to prematurity. Numerous genes have now been identified to be associated with ASD and/or intellectual disability and could have contributed to some cases of ASD in this study. Male sex remains a strong risk factor for ASD in both preterm and term-born children, and particularly in preterm children, male sex may contribute to inherent genetic risks related to sex as well as increased vulnerability to complications of preterm birth that also increase the risk of ASD. The importance of neuroimaging and genetic data relates in part to the observation that preterm and term-born children with ASD have been shown to have important phenotypic differences. In one study, boys with ASD born preterm had higher rates of seizures, attention-deficit/hyperactivity disorder, and sleep apnea,¹⁵  suggesting a potentially different neural substrate for ASD than term-born children.

Perhaps most importantly, these findings provide an opportunity for initiating interventions in early life to mitigate ASD before the diagnosis of ASD can be definitively established. Identification of a constellation of prenatal and neonatal risk factors could help clinicians target infants at highest risk, while providing reassurance to parents whose infants are at low risk. Identification of high-risk infants with low cognitive scores at 6 months of age or those with declining scores over time could provide another opportunity to intensify early intervention services aimed at mitigating manifestations and/or symptoms of ASD. Importantly, identification of high-risk infants by neonatal discharge and/or 6 months of age could improve research into novel therapies to mitigate the manifestations of ASD, such as communication and social-emotional deficits or impairments.

ASD

autism spectrum disorder

CLD

chronic lung disease