New federal figures indicate that autism prevalence among young children is on the rise.

The number of 4-year-olds with the developmental disorder increased from 1 in 75 children in 2010 to 1 in 59 kids in 2014, according to data published late last week in the Center for Disease Control and Prevention’s Morbidity and Mortality Weekly Report.

The findings are the latest to emerge from the CDC’s Autism and Developmental Disabilities Monitoring Network. Traditionally, the network tracks prevalence among 8-year-olds using a review of health and educational records in certain geographic areas across the country. More recently, the agency has expanded its surveillance work to look at 4-year-olds and 16-year-olds.

Advertisement - Continue Reading Below

The report out this month is among the first to detail the CDC’s findings on 4-year-olds. It is based on information collected by researchers at sites in Arizona, Colorado, Missouri, New Jersey, North Carolina, Utah and Wisconsin in 2010, 2012 and 2014. Some sites collected data in all three years, but others only participated part of the time.

Prevalence varied dramatically across the states studied. At the high end, New Jersey had a rate of 1 in 35 children with autism in 2014, nearly three times more than Missouri, which recorded the lowest rate. The researchers said that differences in record keeping as well as the availability of evaluation and diagnostic services in different areas could play a role.

Like in previous reports, autism was more common among boys than girls. Kids diagnosed at age 4 were significantly more likely than 8-year-olds to have co-occurring intellectual disability, the study found, suggesting that those with more severe symptoms are flagged sooner.

Children with autism in some states were far more likely than others to receive their first evaluation by age 3. And, in what experts called a concerning finding, there was no improvement over time in the age of first evaluation at the three sites which collected data across 2010, 2012 and 2014.

“Despite our greater awareness, we are not effective yet in early detection,” said Walter Zahorodny, an associate professor of pediatrics at Rutgers New Jersey Medical School who led the New Jersey portion of the study.

“Children who are evaluated for autism early — around their second birthday — often respond better to treatment than those who are diagnosed later,” Zahorodny said. “However, it appears that only the most seriously affected children are being evaluated at the crucial time, which can delay access to treatment and special services.”

For the study, researchers relied on the criteria for autism outlined in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders. The definition was significantly altered in the manual’s fifth edition, which was adopted in 2013. When children studied in the 2014 data collection were evaluated using both definitions, the CDC report indicates that 20 percent more qualified for an autism diagnosis under the old criteria compared to the new version.

Future reports from the CDC’s autism surveillance network will rely exclusively on the DSM-5 criteria.

The CDC said it will continue to peg its official autism prevalence estimate using data from 8-year-olds. Those figures — most recently released last year — also put autism prevalence at 1 in 59 kids.

The information collected on 4-year-olds can inform efforts to identify children early and help providers anticipate service needs, the agency said.

Alycia Halladay, chief science officer at the Autism Science Foundation, said it’s interesting that the data on 4-year-olds and 8-year-olds both appear to be converging on a prevalence rate of 1 in 59 kids. But, she said the report shows that much more work remains.

“What is interesting is that the 4-year-olds were more likely to have behavioral issues and intellectual challenges, suggesting that more subtle cases of ASD without these signs may not be noticed until 8 years of age in the school system,” Halladay said. “What is most unfortunate is the percent of kids getting an evaluation before 3 years of age did not move from 2010 to 2014. This is something researchers need to focus on.”