Trying to account for a drastic rise in childhood autism in recent years, a California study has found that it cannot be explained away by statistical anomalies or by a growing public awareness that might have led more parents to report the disorder.
But the study's authors, who reported their findings yesterday to the California Legislature, said they were at a loss to explain the reasons for what they called an epidemic of autism, the mysterious brain disorder that affects a person's ability to form relationships and to behave normally in everyday life.
"Autism is on the rise in the state, and we still do not know why," said the lead author, Dr. Robert S. Byrd, an epidemiologist and pediatrician at the University of California at Davis. "The results are, without a doubt, sobering."
As diagnoses of autism have increased throughout the nation, experts and parents have cast about for possible explanations, including genetics, birth injuries and childhood immunizations. The California study found that none of these factors could explain an increase of the magnitude reported there — more than triple from 1987 to 1998.
Dr. Catherine Lord, a professor of psychology and psychiatry at the University of Michigan who is a leading authority on autism, said it was unclear whether the California findings applied to other states.
The federal Centers for Disease Control and Prevention is working in 13 states to look at the apparent increase in autism cases, said Dr. Frank DeStefano, an epidemiologist at the agency. So far, there is no reliable count of autism cases nationwide, since criteria and reporting practices vary from state to state.
The California study was prompted by a 1999 report from the state's Department of Developmental Services, which reported that the number of children with "full spectrum," or profound, autism had increased by 273 percent, to 10,360 in 1998 from 2,778 in 1987. The study did not deal with milder forms of the disorder, like Asperger syndrome.
The numbers were surprising, Dr. Byrd said. The traditional estimate was that 4 or 5 children out of 10,000 might develop autism. Instead, it appeared that 10 children in every 10,000 were seriously autistic, meaning they suffered from a brain disorder that left them unable to speak or compulsively performing repetitive motions like flapping their arms or rocking.
After the period studied, the number of autistic children continued to rise, to 18,460 cases as of July 2002, according to the California Department of Developmental Services.
In response to the study, the legislature directed the MIND Institute, an autism research center at the University of California at Davis, to investigate.
"We wondered if the increase was real," Dr. Byrd said. "Maybe we were doing a better job of finding cases. Maybe there was an increase in awareness of autism. The movie `Rain Man' was very popular."
California has a system of 21 regional centers that diagnose developmental disorders and provide services to children with them. Dr. Byrd and his team mined these centers for data.
Researchers sent questionnaires to the parents of 684 children with full-spectrum autism or mental retardation. About half were teenagers, born from 1983 to 1985; the others were ages 7 to 9, born a decade later.
If the criteria for diagnosing autism had changed in those 10 years or if the definition had broadened, the mystery would be solved, Dr. Byrd said. But the standards used to diagnose full-spectrum autism were the same in both age groups, he said.
Some people suggested that the centers might diagnose autism so families would receive more generous state assistance. But the centers have no incentive to do so, Dr. Byrd said, since they do not receive more state financing for identifying more children with disabilities.
The study also considered whether children in the older group were incorrectly classified as mentally retarded, when they were in fact autistic. But the rate of misdiagnosis was about the same in both groups, Dr. Byrd said.
Still another possibility — that large numbers of families with autistic children had moved into California — was discarded when it turned out that most children in both groups were born in California. A general increase in population accounted for about 10 percent of the rise in autism, Dr. Byrd said. The rest remains a mystery.
There also were no significant differences over time in sex, race or parental education. Parents of the older children were more likely to report mental retardation along with autism, but the finding did not explain the rising incidence.
About a third of parents in both groups reported that their children began to regress around the age of 18 months, Dr. Byrd said. They suddenly lost the ability to say words and stopped making eye contact. Many parents blame vaccinations thatare given around 18 months; until recently some vaccines contained a mercury-based preservative that some people believe can cause brain damage in young children. The study found no evidence that the vaccine was the culprit, Dr. Byrd said.
Nevertheless, more parents of younger children reported constipation and vomiting, which they attributed to complications from the measles vaccine. Wheat allergies were also more frequent. But none of these differences fully explain the increase in autism cases in California, Dr. Byrd said.
Parents in the study were asked what might have caused their child's autism. Nearly half the parents in both groups said they did not know. A third blamed genetics; smaller numbers cited immunizations, birth injury or environmental factors.
"You can't explain an increase of this magnitude on genetics," Dr. Byrd said. "Something else is happening."
"We know autism has a strong genetic component," said Portia Iversen, a founder of Cure Autism Now, a research and advocacy group in Los Angeles formed by parents of autistic children. "But we don't know what in the environment is interacting with genes to contribute to this huge increase in cases."