WEDNESDAY, July 6, 2005 (HealthDay News) -- When calling for more research and funding, autism advocates often cite the increasing numbers of children being diagnosed with autism.
But a new study suggests the way those numbers are estimated may not be a reliable indicator of the true prevalence of autism.
The study, which appears in the July issue of Pediatrics, examined frequently cited autism statistics gathered from data from the U.S. Department of Education (USDE), and concluded that such statistics don't paint a true picture of autism prevalence in the nation.
According to study author Dr. James Laidler, part of the blame for this discrepancy lies in the fact that each school district has its own criteria for defining autism, and those criteria aren't consistent from school district to school district.
"USDE data is not designed and wasn't developed to track autism prevalence, and can't be used to track autism prevalence," said Laidler, who is in the department of biology at Portland State University in Oregon. "I'm not saying there is no autism rise. We can't say whether it's going up, down or staying the same; we need to come up with a better way to track autism."
Autism is a developmental disability, and those with the disorder tend to have problems with socialization and communication, according to information from the CDC. They may also engage in repetitive behaviors, and have trouble with changes in their routine. As many as four in 10 people with autism spectrum disorders may not speak at all, according to the CDC.
The cause of autism is still unknown, though researchers believe the disorder is probably caused by both genetic and environmental factors.
It's not clear how many children are affected by autism. Estimates range from between one in every 166 U.S. children to one in every 500 U.S. children, the CDC reports.
To assess the validity of autism prevalence statistics, Laidler looked at USDE data from 1993 to 2003. He chose to begin in 1993 because that's when school districts were first mandated to report autism as a separate disorder.
At first glance, it appears as if the prevalence of autism did increase exponentially during this time period, Laidler said. In 1993, less than five children per 1,000 were reported as autistic in USDE data. By 2003, that number was more than 25 per 1,000.
But, Laidler questions whether rates of the disorder are actually increasing that quickly, or if a greater public awareness and broader definition of autism is driving the numbers upward. Also, he speculated that inconsistency from one state's definition of autism to another could account for some of the increase.
Laidler cited the difference between the state of Oregon's definition for autism and the one used in the neighboring state of Washington.
"Washington has a fairly strict criteria compared to Oregon, and their autism prevalence is one-third that of Oregon's, which makes no sense," said Laidler. "Whether autism is caused by genetic or environmental exposure, Washington and Oregon aren't that different."
Another trend that suggests the USDE data can't be relied upon for tracking autism prevalence is that the rate of autism keeps going up within each birth-year cohort. For example, even among 16-year-olds the rate of autism is still rising, according to USDE data. That just doesn't seem logical, Laidler said.
Craig Newschaffer, director of the Center for Autism and Developmental Disabilities Epidemiology at Johns Hopkins Bloomberg School of Public Health, said that he and his colleagues also pointed out this increase by birth-year cohort in a study appearing in the March issue of Pediatrics.
What this means, he said, is that "kids aren't being recognized early enough. There's no new incidence at 10 [or 13, or 15, etc.] -- that kids are still acquitting the classification into these older ages is troubling."
"The authors are right that these data are not collecting for the purpose of estimating prevalence," Newschaffer said.
"Nobody disputes that more kids have the diagnosis and the label, but how much is due to increasing diagnostic and labeling tendencies, and how much is due to risk?," he added. "There are strong views on both sides, but not enough evidence yet to support either belief."
To that end, Newschaffer said the CDC is currently gathering autism data from 14 states to get a better handle on the true prevalence of autism. While this approach will also have its limitations, he said, "it should be an improvement."
More information
To learn more about autism, visit the National Institute of Mental Health.
