Autism at Age 4: Prevalence and Early Identification in Six U.S. Communities, 2016

The Early Autism and Developmental Disabilities Monitoring (Early ADDM) Network, part of CDC's ADDM Network, tracks autism spectrum disorder (ASD) among 4-year-olds. For surveillance year 2016 it followed children born in 2012 whose parent or guardian lived in the surveillance areas in Arizona, Colorado, Missouri, New Jersey, North Carolina or Wisconsin. Records from pediatric clinics, special education and early intervention programs were reviewed, and trained clinicians applied a standard case definition based on the DSM-5.

How common

In 2016, 15.6 per 1,000 four-year-olds — about 1 in 64 — had autism, up from 14.1 per 1,000 in 2014 using the same DSM-5 criteria. It was similar to the 2016 National Survey of Children's Health estimate of parent-reported autism among children aged 3–5 (19.7 per 1,000).

• By site: from 8.8 per 1,000 in Missouri to 25.3 in New Jersey. Such wide variation may reflect uneven success in community identification.
• By sex: 23.9 per 1,000 boys and 6.8 per 1,000 girls — 3.5 times as common in boys, at every site. Among 8-year-olds the ratio was 4.3 to 1.
• By race and ethnicity: prevalence was similar for white (13.2) and Black (14.3) children at every site — a narrowing of an earlier disparity. In New Jersey, Hispanic 4-year-olds had a higher prevalence (32.1 per 1,000) than white or Black children, the reverse of the pattern among 8-year-olds there; elsewhere it was 11.0 per 1,000 for Hispanic children.
• Diagnostic criteria: at four sites that reviewed at least half of their records under both editions, prevalence was 16.3 per 1,000 under DSM-5 and 18.9 under DSM-IV-TR; prevalence was lower under DSM-5 at one of those sites, New Jersey.

Intellectual disability

At the four sites with intellectual disability data for at least 60% of children, 53% of 4-year-olds with autism also had intellectual disability — from 29% in Colorado to 66% in North Carolina — compared with 33% among 8-year-olds. It rose from 2014 at two of three comparable sites (New Jersey, 45% to 53%; North Carolina, 45% to 66%), and the rise held when 2014 cases were reclassified by DSM-5, suggesting it was not caused by the change in criteria. Unlike among 8-year-olds, girls and boys aged 4 were equally likely to have it.

How early

• First evaluation: median age 26 months (22 in North Carolina to 30 in Arizona and Missouri). 84% of children with autism had been evaluated by 36 months, up from 74% in 2014 — progress toward the Healthy People 2020 goal.
• Diagnosis: 71% already had a documented autism diagnosis from a community provider, up from 58% in 2014. The median age at earliest diagnosis was 33 months, with site medians ranging from 29 to 36 months, much as in 2014. North Carolina had the youngest median (29 months) but the lowest share with a documented diagnosis (55%).
• With and without intellectual disability: median ages were almost the same — first evaluation at 25 and 24 months, diagnosis at 32 months for both.

Cumulative incidence of autism diagnoses by age, children aged 4 years compared with children aged 8 years, Early ADDM Network, 2016. CDC, MMWR.

Younger children, earlier diagnoses

Eight-year-olds in the same areas had a higher prevalence, 23.5 per 1,000. But comparing the two cohorts by age, children born in 2012 had been diagnosed earlier than those born in 2008: from 26 to 48 months, cumulative incidence was higher for the younger group, reaching 10.2 per 1,000 at 48 months versus 8.3. The pattern held at four of six sites — the curves overlapped in Arizona, and in North Carolina the older group was ahead, possibly because one county's education records were available only for 8-year-olds. It held for both sexes, though diagnoses among girls seemed to level off around 36 months while boys' kept rising. If the trend continues, prevalence among 8-year-olds may rise in the 2020 surveillance year.

What it means

More 4-year-olds with autism are being evaluated by age 3 and diagnosed by age 4, but a median diagnosis at 33 months is well above the youngest age at which autism can be identified. Evaluating children as soon as developmental concerns arise could lead to earlier diagnosis, earlier evidence-based intervention and better outcomes.

Limitations

Results depend on the records each site can review; education and early intervention records are not available everywhere. The same intellectual disability definition was used for 4- and 8-year-olds, though IQ is less stable at younger ages. The surveillance areas are parts of states chosen through competitive funding and may not represent their states or the country.

Sources

• Shaw KA, Maenner MJ, Baio J, et al. "Early Identification of Autism Spectrum Disorder Among Children Aged 4 Years — Early Autism and Developmental Disabilities Monitoring Network, Six Sites, United States, 2016." MMWR Surveillance Summaries 69(3). CDC.