Skip navigation

Calls to Action

Autism’s Rising Prevalence: What Do The Numbers Mean?

Autism Speaks epidemiologist Michael Rosanoff answers questions on how the CDC tracks prevalence and what the numbers mean in the US and globally
March 29, 2014


Epidemiologist Michael Rosanoff is Autism Speaks associate director for public health research. We asked him to explain some of the technical aspects of how the CDC tracks autism prevalence, how prevalence is tracked internationally, and what the numbers mean to communities in the United States and abroad.

Read our full coverage of the CDC’s “1 in 68” prevalence update here.

How does the CDC track autism’s prevalence, and how accurate are its estimates?
The CDC uses a surveillance system called the Autism and Developmental Disabilities Monitoring (ADDM) Network to estimate and track the prevalence of autism and other developmental disabilities across the United States. The ADDM network consists of 14 communities across the country. However, the number of communities included in a given report varies depending on federal funding. These communities are meant to represent the United States as a whole. Within this network, the CDC researchers focus on 8-year-olds based on the assumption that most children who have autism will be diagnosed by this age.

To estimate prevalence, the investigators look through educational and medical records for diagnoses of autism spectrum disorder (ASD). In other words, they’re not directly screening and diagnosing children, which can be challenging, time consuming and costly. Instead, they use an indirect, records-based approach to count autism diagnoses.

This indirectness is the main limitation of the approach, because it misses children who have autism but have not been diagnosed and/or are not receiving appropriate medical or educational services.

For this reason, the CDC is likely underestimating autism’s prevalence. By contrast, Autism Speaks has funded studies that send investigators into schools to directly assess students for autism. Our study in South Korea, for example, found a prevalence of 1 in 38. Two-thirds of these children had previously gone undiagnosed and, so, would have been missed by the CDC’s records-based counting methods.

Why is there such variability across the CDC’s monitoring sites?
Much of the variability between states appears to result from differences in the detection of autism and the availability of records on the autism-related services children receive. Based on this, we don’t think that the difference in prevalence across states is due to true differences in autism risk. This is why the investigators average the site-specific prevalence numbers into an overall estimate for the US.

What is Autism Speaks doing to improve methods for tracking prevalence? Why is this important?
When we underestimate prevalence, we underestimate the needs of individuals with autism and their families. We need more accurate and reliable measures of autism prevalence to enable public health professionals and government policy makers to develop strategies that meet the needs of the growing autism community. We also need accurate estimates of autism’s economic burden on families and on our society – to help us better understand the full picture and to develop programs that meet the magnitude of the challenge.

To improve estimates of autism’s prevalence, Autism Speaks is currently funding the South Carolina Children’s Educational Surveillance Study (SUCCESS). This study is taking place at the South Carolina ADDM site, in partnership with the CDC. It combines the CDC’s records-based approach with the direct screening and diagnosis of school children. It will give us better information on both prevalence and the barriers that prevent some children from accessing diagnostic and intervention services. We need to better understand why we’ve been missing some children who have autism and need services.

How can international epidemiology studies help us understand what’s driving the increase?
By comparing prevalence estimates across countries, we can generate clues about what causes or increases the risk of autism and how these factors may have changed over time. Conducting research in different populations also allows us to study autism prevalence in different ethnic groups, in special populations (e.g., those that are genetically isolated) and in communities exposed to different environmental conditions (e.g., exposure to certain pollutants or malnutrition). Additionally, some nations have large national health registries and public health surveillance systems that can generate new information about what causes or increases the risk for autism.

What international epidemiology studies is Autism Speaks funding and why?
In Mexico, Autism Speaks is funding Latin America’s first prevalence study, using methods similar to those in our South Korea study. This new study will do more than simply measure prevalence. It is also designed to increase awareness about autism and help improve access to autism services in Mexico.

It’s important to remember how awareness of autism’s prevalence in the United States was critical to it becoming a public health priority in our country. Similarly, families around the world will benefit when their governments understand how common autism is in their countries. Through our Global Autism Public Health Initiative, Autism Speaks is supporting such projects in more than 50 countries.

For further perspective on autism prevalence, see the transcript of Rosanoff’s Friday Twitter chat here, his FAQs on autism epidemiology here and his video chat Q&As below.




Concerned about autism prevalence? Join people around the world to Light It Up Blue for autism awareness.


Subscribe to Autism Speaks Science Digest to get all our research news,
and perspective delivered biweekly to your inbox.