New government survey pegs autism prevalence at 1 in 45

National Health Interview Survey underscores gap between the number of kids diagnosed with autism and the number receiving services

November 13, 2015

A new government survey of parents suggests that 1 in 45 children, ages 3 through 17, have been diagnosed with autism spectrum disorder (ASD). This is notably higher than the official government estimate of 1 in 68 American children with autism, by the Centers for Disease Control and Prevention (CDC).

Because the new numbers come from a parent survey, they don’t replace the CDC’s 1-in-68 figure as the official estimate of autism prevalence in the United States.

However, the CDC has acknowledged that its estimate has significant limitations. It’s based on an analysis of the medical and school records of 8-year-old children at monitoring sites across the country. As such, it can miss children who are not receiving medical or special education services related to autism.

“The 1 in 45 estimate is not surprising and is likely a more accurate representation of autism prevalence in the United States,” comments epidemiologist Michael Rosanoff, Autism Speaks director for public health research. “This means that 2 percent of children in the U.S. are living with autism. The earlier they have access to care, services and treatment, the more likely they are to progress.” (Read statements from Autism Speaks and Autism Society of America here.)

The parent survey results released today are from the 2014 National Health Interview Survey. Conducted by the National Center for Health Statistics, it’s the most in-depth survey of its kind in the United States. In all, more than 12,000 parents are interviewed on family health conditions. As part of the interview, one child in each family is randomly selected to be the subject of detailed questions on health and disabilities.

In part, the 2014 survey's high autism prevalence number reflects changes in the order and wording of the survey’s questions, the authors say. They made these changes after noting that the questions on their 2011-2013 survey had likely confused parents and resulted in an underreporting of autism.

For exact wording and order of survey questions, see figure 1 below.

Indeed, the rewording and reordering of the questions produced a near doubling of reported autism diagnoses – from 1 in 80 in the 2011-2013 surveys to 1 in 45 in the 2014 survey. (See figure 2 below.)

The new numbers are more similar to those from another national parent survey – the 2013 National Survey of Children’s Health – which found an autism prevalence of 1 in 50.

Pursuing more accurate numbers
In its effort to develop better methods for identifying autism and estimating prevalence, Autism Speaks has funded studies using active surveillance methods that go beyond parent reports and record reviews. In the first such study, in South Korea, researchers went into schools to screen children for autism. They found a prevalence of 1 in 38 – with many of the children having gone previously undiagnosed.

Autism Speaks and the CDC have since collaborated to support a study testing the same active screening methods in a U.S. community. The results are expected in 2016.

“We need to better understand not only who has autism,” Rosanoff concludes, “but whether they are receiving the support they need and how we can ensure that they do receive it.”