The number of kids with autism seems have gone up fivefold in recent years. Parents have been worried that something artificial, like childhood vaccines, might be responsible for the surge in the disorder, which is marked by severely impaired social interactions and limited communication skills. Doctors have argued that the "increase" is due to better awareness and diagnosis, not vaccine needles. This study looked at the number of new autism cases in one Minnesota county every year from 1976 to 1997, to try to connect any spikes or drops to outside events. And the biggest connection, they concluded, was to more awareness, not to vaccines.

What the researchers wanted to know: Did autism cases go up after 1987, when a new set of diagnostic guidelines was published? Or did cases go up after certain vaccines were introduced?

What they did: In Olmsted County, Minn., more than 95 percent of all healthcare is provided through the Mayo Clinic and Olmsted Medical Center, which means there are exhaustive, complete medical records on the county population. In about 3,000 children who got any psychiatric or developmental disorder diagnosis during this time, researchers looked for signs of autism using a single common yardstick: criteria set up in a psychiatric guideline book called DSM-IV. That one source should eliminate any bias due to changing diagnostic criteria over the years.

What they found: Annual autism incidence was relatively stable up through 1987, when it was 7.9 kids in every 100,000. After that it started rising and by 1995 it was 44.9 per 100,000. But the most commonly cited vaccine suspect, measles-mumps-rubella, was introduced in 1971, well before the increase. The use of of inert ingredients like thimerosal also didn't not coincide with the increase. However, a new set of psychiatric guidelines for diagnosing autism was published in 1987, just when the upward trend started. The 1987 guidelines were more specific, so some kids previously classified with some vague problem like "mental retardation" would thereafter be called autistic. That would lead to an increase in the number of autism diagnoses.

What the study means to you: The timing suggests that the measles-mumps-rubella vaccine is not responsible for autism. Other research from Denmark similarly points to vaccine innocence. That's one less reason to avoid the shot, which can prevent three very serious childhood diseases.

Caveats: The kids were not directly examined, only their records were, so bias is still possible.

Find out more: An autism fact sheet and information page from the National Institute of Neurological Disorders and Stroke

Read the article: Barbaresi, W. J. et al. "The Incidence of Autism in Olmsted County, Minnesota, 1976–1997." Archives of Pediatrics & Adolescent Medicine, January 2005, vol. 159, no. 1, pp. 37–44.

Abstract online: http://archpedi.ama-assn.org