The incidence of autism may be much higher than previously thought in the United States and elsewhere in the world, according to a rigorous, comprehensive study of the condition conducted in South Korea, researchers reported Monday.
In the first study to take a broad-population look at the prevalence of autism spectrum disorders — types of autism ranging from severe symptoms to the milder Asperger's syndrome — researchers found a rate of 2.64% among South Korean children. That's 1 in 38 children, a rate far higher than the estimate of 1 in 110 children for the U.S. by the Centers for Disease Control and Prevention.
The study, being published Monday in the American Journal of Psychiatry, suggests that, under rigorous examination, many more children may be affected than previously suspected.
The study "is different in the sense that they are screening the entire population of children" including those who have never been flagged with a potential problem, said Geraldine Dawson, chief scientific officer of the research and advocacy group Autism Speaks and an autism researcher at the University of North Carolina at Chapel Hill. "It raises a question, I think, of whether we are underestimating the prevalence in the U.S. as well as elsewhere."
The five-year study, funded partly by Autism Speaks and led by Dr. Young-Shin Kim of the Yale Child Study Center, differed significantly in methodology from earlier autism-prevalence studies. This likely accounts for the dramatically different findings, Kim said.
Previous studies assessing population-wide autism rates typically focused on high-risk populations — such as classrooms of special education students. In contrast, the study conducted in South Korea assessed more than 55,000 children, ages 7 to 12, not only from special education classrooms and mental health service organizations but also regular schools.
Using several diagnostic techniques and measures to evaluate the children, the study found that the rates of autism spectrum disorder among the children in special education and mental health services programs were similar to estimates elsewhere in the world — from 0.6% to 1.8% of the population.
But when students in regular schools were part of the assessment, the prevalence rate rose to 2.64%.
Children in regular schools are more likely to be higher functioning and thus undiagnosed. The study in South Korea found, for example, that many of the children diagnosed with autism spectrum disorder who were in regular schools "looked very different" from autistic students in special education classes, Kim said. Many had Asperger's syndrome, a milder form of the condition. These children tended to have normal intelligence but poor social skills.
Kim said children in regular American classrooms should be included in future studies to get a better measure of the incidence of autism spectrum disorders. If that were done, she added, the prevalence of the condition in the U.S. and other countries also would be in the range of 2% to 3%.
The study was conducted in South Korea so that scientists could test whether diagnostic criteria could be applied in a variety of cultures, and because South Korea has robust health and education systems that allowed for this type of study.
The conclusions are more far-reaching, Dawson said.
"This study clearly confirms that autism is a significant, global, public health concern that transcends cultural, ethnic and geographic boundaries," she said. "We do need to do this type of study in the U.S. … Until we do, we won't know what the population prevalence is."
Laura Schreibman, a veteran autism expert and psychology professor at UC San Diego, said the findings, if confirmed, were "scary" but should be replicated. It is possible, she added, that there are more undiagnosed cases of children with mild autism in South Korea than elsewhere.
"Is there something about the Korean culture? Are they less likely to come forward because of stigma associated with this?" she said. "How much can we generalize this finding to the worldwide population?"