The statistics appear to tell an alarming story: 1 of every 100 American children has an autism spectrum disorder, according to newly released results from two federally funded research studies.

Just two years ago, the government estimate was 1 in 150—and the new findings have some worried that we are in the midst of an unprecedented autism epidemic. But a growing awareness of the disorder and a broadening in the official definition make such a claim hard to prove.

Just two years ago, the government estimate was 1 in 150—and the new findings have some worried that we are in the midst of an unprecedented autism epidemic.

Until about 20 years ago, the devastating diagnosis of autism applied only to the most severe cases, people who were largely incapable of normal social interaction and were often institutionalized. But starting in the late 1980s, the official definition expanded to include a spectrum of conditions, from classic autism to milder forms, such as Asperger’s syndrome. The range of people considered to have autism now includes not only the most severely disabled but also many others whose impairments do not prevent them from getting jobs, going to college, even getting married. At the same time, growing awareness of autism by parents, teachers, and doctors means the disorder is recognized more often, and that also adds to the numbers.

All these changes make it impossible to compare present rates with those of a generation ago, so no one can say with certainty that the prevalence of autism has increased. But even with a standard definition, it has been difficult for researchers to figure out exactly how widespread autism is now. Numbers change as surveys become increasingly sophisticated.

The gold standard at the moment is a Centers for Disease Control surveillance study of 8-year-olds who have received the diagnosis of autism at 11 sites around the country. Although complete results aren’t expected until later this year, the CDC released preliminary numbers at the same time as another study was published this week in the journal Pediatrics. The methodology of that study was very different; researchers from the Health Resources and Services Administration conducted a telephone survey of 78,000 parents of children age 3 to 17, asking whether they had ever been told that their children had an autism spectrum disorder.

Both the CDC and Pediatrics results are noteworthy because these two dissimilar studies came up with essentially the same number—1 in 100. That finding roughly conforms to autism rates reported in other industrialized countries, which means the 1 in 100 could be the most accurate figure so far. Defining the extent of the problem would be a huge step forward.

But even if 1 in 100 holds up, it doesn’t answer the question of whether more people have autism now than in the past, as many advocates believe. “The concern here is that buried in these numbers is a true increase,” said Dr. Thomas Insel, director of the National Institute of Mental Health, at a press conference last week. And if there is an increase, what’s behind it? Scientists don’t know what causes autism, although research is ongoing into possible environmental hazards or genetic explanations.

Even without evidence of an increase, there are clearly a lot of children who need help. “We’re going to have to think very hard about what we’re going to do for the 1 in 100,” Insel said, adding that the Obama administration is making autism research a priority.

Although the range of therapies for children with autism has grown substantially, there is still no cure. But in the Pediatrics study, parents whose children had been diagnosed with autism were asked whether their child still had the disorder, and nearly 40 percent said no—an intriguing finding the researchers could not explain. In some cases, children diagnosed at a very young age eventually may have been found to have other problems instead.

Indeed, diagnosis can be difficult. It’s based not on a blood test or an X-ray, but on specific observation of behaviors such as a lack of social interaction, poor eye contact, and rigid, repetitive movements (rocking is common). If the behaviors change as children get older, the diagnosis may change, as well. In other cases, children who clearly need special attention may be given an autism diagnosis—even if it’s not conclusive—because then the youngsters will be eligible for particular educational and social services.

The Pediatrics study also reported that boys are four times more likely than girls to be diagnosed with autism, a finding consistent with other research. But race also seems to play a role. In the Pediatrics study, autism was diagnosed most often among whites, while parents of black children were more likely to say their children no longer had the disorder. It’s unclear why these differences exist. At the moment, the only certainty about autism is that there are many more questions than answers.

Barbara Kantrowitz is staff editor at the Hechinger Institute on Education and the Media at Columbia University Teachers College and a contributing editor at Newsweek.