Between 2000 and 2016, estimates increased two-fold for autism with intellectual disability and five-fold for autism without intellectual disability.

“Autism is a heterogenous disorder, which makes it challenging to study the etiology—but it is also important to allocate community resources based on needs,” Josephine Shenouda, DrPH, MS, explains. “Currently, the best predictor of functional outcomes for children with autism is intellectual ability. Tracking autism over time by intellectual ability helps show changes in autism expression over time, identify disparities among subgroups in the population, and helps us prepare and allocate resources in our community.”

For a study published in Pediatrics, Dr. Shenouda and colleagues assessed the prevalence of autism spectrum disorder with and without intellectual disability among 8 year olds in the New York/New Jersey metropolitan area. The cross-sectional study design included data from 2000 to 2016.

“In our region, we saw autism estimates of approximately 1% in 2000 that rose to 3% by 2016, but there were variations, with some community estimates exceeding 5%,” Dr. Shenouda says. “That led to the question of why, and we saw that in areas with high estimates, we are identifying more children with autism without intellectual disability. That question led to this study. We wanted to know if the increase over time in autism was equally distributed among children with autism with and without intellectual disability.”

Estimates Increase for Autism Without Intellectual Disability

The study team identified 4,661 8-year-old children (81.4% male) with autism spectrum disorder (ASD), including 1,505 with intellectual disability (32.3%) and 2,764 without intellectual disability (59.3%). Nearly half of the study population (45.4%) identified as non-Hispanic White, followed by Hispanic (26.4%) and Black (20.3%).

“We found that there was a steeper rise over time among children with autism without intellectual disability,” Dr. Shenouda says. “While estimates increased two-fold for autism with intellectual disability, estimates increased five-fold for autism without intellectual disability between 2000 and 2016.”

She also noted the disparities observed in the detection of autism without intellectual disability. “Specifically, our findings show lower autism estimates among Black children compared with White children and among children residing in socially disadvantaged areas compared with children residing in affluent areas.”

According to Dr. Shenouda, the researchers expected different patterns in estimates over time among children with autism with and without intellectual disability, but not the large differences based on socioeconomic factors (Table).

“A child with autism without intellectual disability in an affluent region has an 80% greater likelihood of being identified with autism and receiving services compared with a similar child in a community of low wealth,” she says. “We also saw that Black children with autism without intellectual disability were 30% less likely to be identified compared with White children. These findings can stem from multiple factors, including access to service as well as cultural barriers.”

Universal Screening ‘Essential’ to Identify Autism Earlier

According to Dr. Shenouda, the disparities observed as likely the result of lower access to services among certain populations. “It is likely many of these children are not receiving services and not coming to a provider’s attention at an early age. Early screening for autism is essential to identify children early, and children from underserved communities are the ones likely to benefit the most.”

She also noted that the best way to address rising rates of autism, and to have an impact on the disparities seen in autism identification, is through universal screening for the condition among all toddlers.

“Future research should focus on developing objective metrics to assess the needs of children with autism and predict future outcomes,” Dr. Shenouda notes. “Currently, intellectual ability remains the best predictor of functional outcomes, but other metrics need to be developed.”