1 in 100.
1 in 38.
Those three sets of numbers all purport to show the rate of autism in today's world.
The 1 in 88 figure is the Centers for Disease Control and Prevention's official estimate of how many children in America have an autism diagnosis. It is significantly higher than the agency's 1 in 110 figure from 2006, and a far cry from the 1 in 10,000 rate used
The CDC's escalating figures have been cited by many as evidence of an "explosion" in autism rates.
The 1 in 100 figure is based on a 2011 study in Britain that diagnosed adults as part of a national neuropsychiatric health screening. Importantly, it found that the 1 percent autism rate had stayed steady for decade after decade.
The 1 in 38 number is from another 2011 study, in which researchers from Yale and George Washington universities looked at thousands of children in a suburb of Seoul, South Korea, and determined that 2.65 percent of them had autism -- and the vast majority had not been formally diagnosed with the disorder.
The sharply different estimates add more fuel to the debate over how common autism is, and whether it is on the rise. To show what a difference the varying estimates would make in Allegheny County alone, which has 1.2 million residents, the number of autistic people could range from 12,000, using the British figure, to more than 32,000, using the South Korea study.
It is no wonder, then, that leading autism researchers have conflicting opinions on how widespread the disorder is.
Those who believe it's growing
Among the scientists who believe autism is showing a real increase in America are Thomas Insel, the director of the National Institute of Mental Health; David Amaral, research director at the MIND Institute at the University of California at Davis; and Peter Bearman, a social sciences professor at Columbia University.
In an interview earlier this year, Dr. Insel acknowledged that up to half of the increase in autism numbers may be due to such factors as a broader diagnosis of who has autism and greater awareness of the condition, but he said that "without being able to explain the other half [of the growth in numbers], most of us are left with the conclusion that there is a real increase."
Some have argued that autistic people were simply given other labels in the past, from schizophrenia to retardation.
But Dr. Insel said that today, "there is a large proportion of people with this disorder who are so disabled that it is unthinkable to me that they weren't detected in the past. If you have a 12-year-old in diapers who is head banging and has no language, it's hard for me to believe these kids existed in 1980 and were not being labeled."
David Amaral also thinks the autism increase is real, but he declined to put a number on it.
"My sense from talking to lots of people who have been in the field for a long time is that there really has been an increase in the number of people with autism, but there are so many confounds -- the change in the criteria, the fact we're identifying people with much more subtle changes, and the fact that when I was a kid they might have just been called odd or quirky."
Dr. Insel's perspective is based largely on the work of Mr. Bearman, who has done several studies trying to tease apart the factors causing the increase in autism rates in California. His research has found both social and biological factors that account for 50 to 60 percent of the increase in the caseload there.
In one study, he noted that autism diagnoses go up in areas where affected families live near each other. After ruling out environmental causes, he concluded that the mere fact of knowing a nearby family with an autistic child raised the chances that another family would get a diagnosis.
Social conditions can also cause fluctuations in autism numbers, as he found in studying the state's Hispanic population. In that case, autism numbers among Hispanics dropped after California passed a 1994 referendum restricting non-citizen immigrants from using public services, and again after the 2001 Patriot Act, when some Hispanics feared they might be deported. The decreases suggest Hispanic families simply weren't seeking autism services during those periods.
On the biological side, the Columbia professor found that parents over the age of 35 have a higher risk of giving birth to autistic children. He believes part of the increase may also be linked to in vitro fertilization techniques.
"I think our work does show that the idea that the rate of autism has been consistent across populations and across generations seems unlikely."
In the UK, a different result
While the Bearman and CDC studies rely on government and school district autism numbers, Terry Brugha of the University of Leicester in England took a different approach.
His team of researchers screened a random sample of about 7,400 people throughout Britain for possible psychiatric disorders. More than 600 people then consented to a more detailed interview, and based on that, his group determined that the rate of autism was about 1 in 100 in all age groups, including those born in the 1930s. Many of the older autistic people had never received a diagnosis. The ratio of males to females with the disorder was 6 to 1, even higher than the 4 to 1 figure often cited in the United States.
"I don't think the [autism] rates are going up," Dr. Brugha said after doing the study. "Whatever is going on is something that has been going on for a long time, not something recent like changes in vaccination schedules or diets."