The rise in the number of autism cases is not evidence of an epidemic, but shows that schools are diagnosing autism more frequently, researchers said Monday.

The study showed that the number of children classified as mentally retarded or learning disabled declined between 1994 and 2003, the same time as the rise in autism cases.

The author of the study, appearing in Monday's issue of Pediatrics, said this suggests a switch in diagnoses, not an epidemic of autism.

"Many of the children now being counted in the autism category would probably have been counted in the mental retardation or learning disabilities categories if they were being labelled 10 years ago instead of today," said researcher Paul Shattuck of the Waisman Center at the University of Wisconsin at Madison, in a statement.

Shattuck said his study does not disprove the existence of an epidemic of autism cases, but it does show that data from school special education program is not sufficient to prove one exists.

The data from the U.S. Department of Education is sometimes used to support claims of an autism epidemic.

Some claim use of vaccines containing thimerosal, a preservative that contains mercury, is behind the epidemic. A U.S. health panel found no link in 2004 between vaccines using therimerosal and autism.

In the U.S., autism was added to the list of categories identifying school children with special needs in 1994. The number of cases of autism has increased since then, but the rise varied widely from state to state.

The average prevalence of autism among six- to 11-year-olds enrolled in special education programs increased to 3.1 per 1,000 in 2003 from 0.6 per 1,000 students in 1994.

At the same time, the prevalence of mental retardation fell by 2.8 per 1,000 pupils and the prevalence of learning disabilities fell by 8.3 per 1,000 pupils.

In states such as California, there was no drop in the number of children labelled mentally retarded to offset the rise in autism cases.

In a commentary to the study, also published in Pediatrics, an autism researcher challenged Shattuck's analysis, saying no one can know whether a child born earlier or later would have been classified differently.

"At best, analyses of this type are merely trying to determine if trends in one classification have the potential to offset those in another," wrote Craig Newschaffer of Johns Hopkins Bloomberg School of Public Health in Baltimore.

Newschaffer said more research is needed into the roles played by genetics and environmental triggers in autism.