In a finding researchers are still trying to understand, autism rates in British children appear to have plateaued in the past decade, according to a new study released Wednesday.
These findings differ from U.S. data, which estimate that the prevalence of the disorder is still rising in this country.
The authors of the new study in BMJ Open, an “open-access” medical journal put out by an arm of the British Medical Association, said that it was possible that increased public awareness and expanded diagnoses for autism spectrum disorder (ASD) led to increased diagnoses in the 1990s that are now leveling off.
But the measles-mumps-rubella vaccine, which some have argued is related to increased autism rates, “was surely not the cause,” and “the actual cause remains in large part a mystery,” wrote British pediatrician Brent Taylor, researcher Hershel Jick and systems analyst Dean MacLaughlin of the Boston Collaborative Drug Surveillance Program at Boston University School of Public Health.
Autism spectrum disorder refers to a range of developmental disabilities that can cause significant social, communication and behavioral challenges. It appears in more boys than girls, and is typically diagnosed by age 8.
In the U.S., among 8-year-old children, about 11 in 1,000 will receive an autism diagnosis, or 1 in 88. This is greater than prevalence rates seen in 2002 (1 in 150 children) and 2006 (1 in 110 children), according to the most recent data from the Centers for Disease Control and Prevention’s (CDC) Autism and Developmental Disabilities Monitoring Network.
The network collects data on autism from as many as 14 states every two years, using medical records and/or school health records. Its latest data are from 2008.
The causes of autism are not understood, and recent studies have found associations between childhood autism and multiple births, preterm births, low birthweights, maternal obesity, having a full sibling with autism and having older parents. A major debate continues about the relationship between vaccines and autism.
According to the new BMJ Open paper, autism diagnoses were rare in both the United Kingdom and the United States until the 1990s, when the rates began surging.
In March 2012, the CDC reported that U.S. autism rates were continuing to grow significantly, Mr. Jick and his colleagues wrote. That CDC news “prompted us to review and update the information” about autism in Britain, using the country’s “General Practice Research Database.” The massive collection held 20 years of medical records for about 3 million persons — including validated autism diagnoses for 8-year-old children.
The data showed that while the British autism rate rose significantly in the 1990s, it “reached a plateau shortly after 2000 and has remained steady through 2010,” the researchers wrote.
As of 2010, their article said, autism prevalence rates for 8-year-old British boys was about four cases per 1,000, and 0.8 per 1,000 for British girls. This was essentially the same as in 2004.
By contrast, autism rates for 8-year-old U.S. boys rose from a range of 8.9. to 15.8 cases per 1,000 in 2004 to an average of 18.4 cases per 1,000 in 2008. For U.S. girls, rates went from 1.5 to 3.7 cases per 1,000 in 2004 to four cases per 1,000 in 2008.
Taken together, the U.S. autism prevalence rate rose 78 percent between 2004 and 2008, the CDC said, even as British rates apparently remained flat.
The new study noted that countries use different surveillance methods and diagnoses for autism spectrum disorder. However, given the sharp rises in the number of children diagnosed as autistic in other parts of Europe and North America in the 1990s, the researchers said, it was unlikely that just broadening the diagnostic criteria or publicizing the condition would have led to such large, simultaneous increases.
The CDC media office was unavailable to comment on the study Wednesday, due to the partial government shutdown.
Mr. Jick declined to comment on how the CDC might react to the study. “All I can tell you is that this whole area of research is arcane, complicated and difficult,” he said. “It all boils down to the design you use to collect the data, and the data themselves, which come from some population out there.”