Results released from a recent study conducted by the journal Pediatrics revealed that autism could potentially affect 2.5% of children in the United States. This is significantly higher than the 1.7% estimated by the CDC using 2014 data. Thomas Frazier, chief science officer of the advocacy organization Autism Speaks attributes the discrepancy to "... methods that are a bit more liberal and inclusive than the CDC's methods…[but] they [CDC’s numbers] are likely a bit conservative."
Furthermore, the study in Pediatrics was based off of parent survey data, which, unlike the CDC report, is not validated by health and education records. Other discrepancies may be attributed to the ages of the children included in the report, their geographic location, and even the years the study was conducted. While the prevalence of autism has been rising for years, it will remain difficult to pinpoint an exact number of children affected, in part because autism is so difficult to diagnose.