A University of Kansas professor has monitored the number of children diagnosed with autism from the years 2000 to 2007 under the Disabilities Education Act. Jason Travers, assistant professor of special education, who co-authored the study, found a disparity in the number of minorities identified with autism and the number of Caucasian children identified.

"That's a pretty alarming number," Travers said of the CDC figure. "I wanted to see if there were differences in these rates. Previous research had found that African-Americans were over-identified. But the data I was looking at showed they were under-identified. This was during an era when autism prevalence rates were increasing across the board."

According to the Centers for Disease Control, one in 68 children have autism. Having previously studied autism, Travers and colleagues — colleagues Michael Krezmien of the University of Massachusetts-Amherst, Candace Mulcahy of Binghamton University and Matthew Tincani of Temple University — decided to study the differences in diagnosis.

The authors noticed that the number of Hispanics identified increases in all U.S. regions, excluding: Kentucky, Louisiana and the District of Columbia. They believe part of the problem is the fact that the criteria for diagnosis varies from state to state. And while there was an increase in identification for Hispanics (and African-Americans), it was a much smaller rate seen than the overall population and what was expected by the CDC.

"Nearly every state that had proportional representation of students in 2000 underidentified black and Hispanic students in 2007," the authors wrote. "Although there is no firm epidemiological evidence that race is predictive of autism, we found substantial racial differences in the ways U.S. school identify students with autism."

The study, published in the Journal of Special Education, highlights a greater issue than just identification numbers. The bigger picture is if Hispanics do not have access to services that will allow proper diagnosis.

"I'm not convinced we thoroughly understand this problem in special education right now," Travers said. "I think what's needed is advanced statistical models that can more accurately identify predictors associated with identification."