Autism can be spotted as early as age 2, but the median age in the report was 41/2. That means children are missing out on key preschool services that could improve their functioning.
Boyle said there also is a need for more research into what causes the disorder, which is thought to result from a combination of genetic and unknown environmental factors.
"More is understood about autism than ever before," she said, "but these numbers are an important reminder of the need for answers."
The rates rose 30 percent between 2008 and 2010. They have more than doubled both nationwide and in New Jersey since the first report was done in 2000. The CDC estimates that 1.2 million people under the age of 21 are on the autism spectrum.
Walter Zahorodny, an epidemiologist and psychologist at Rutgers New Jersey Medical School who directed New Jersey data-collection for the report, said it should put to rest the argument over whether autism diagnoses are increasing because of greater awareness or because more children have the disabling condition.
"It's a true increase," he said. "It's a change of great magnitude. It's silly to go on debating that." He expects the numbers to climb higher before they plateau.
Jennifer Pinto-Martin, an epidemiologist in the University of Pennsylvania School of Nursing who worked on previous versions of the report, is not so sure. While the CDC has been using the same definition for many years, she said, changing attitudes have made it easier to get the diagnosis. Stigma has abated, and parents may seek the label so a child can get extra help.
Both researchers agreed that one reason New Jersey's numbers are high - one in 28 boys fit the criteria for autism, as did one in 133 girls - is the state has especially good services and record-keeping.
The CDC's Boyle sidestepped the "why" question by saying there is evidence that some of the increase is due to changes in diagnosis without elaborating on what is responsible for the rest of it.
In successive reports over time, a growing proportion of children characterized as autistic have been of normal intelligence, suggesting that milder cases are responsible for much of the overall increase. Thirty-two percent were average or above in 2002, compared with 46 percent in 2010.
"Some of these kids are very mildly affected," Pinto-Martin said. "That doesn't mean they don't have it. That means in the past no one gave them a label."
Zahorodny said most New Jersey students were in the most seriously affected part of the spectrum. "These are not quirky kids for the most part," he said.
The report used a team of experts to evaluate data on children in parts of 11 states, including the New Jersey counties of Essex, Union, Hudson, and Ocean. Twenty percent of the children said to have autism had not previously received a diagnosis. Ten years ago, that rate was 30 percent, a sign of more aggressive diagnosis in the community, Boyle said.
In all the regions, the rate per 1,000 children has risen from 6.7 in 2000 to 14.7 in 2010.
Pennsylvania was not included in the most recent report. Pinto-Martin said it dropped out because researchers were unable to get data from schools, which are used to augment information from health-care providers. The four states included in Thursday's report that did not have education data all reported lower rates than states that provided access to more information.
Pinto-Martin is now focusing on what causes autism. Risk is known to increase with parental age, premature birth, and maternal immune dysfunction.
Zahorodny said New Jersey's well-educated populace may have relatively large numbers of older parents. In theory, that could explain a small portion of the excess.
The disorder is characterized by communication problems, obsessional interests, and repetitive movements. Diagnosis is based on symptoms, not a medical test. Symptoms for people who are on the autism spectrum can range from mild to severe.
The official description of autism changed last year when a new version of the Diagnostic and Statistical Manual of Mental Disorders, the DSM-5, was published. The new CDC report used the old definition. Some autism advocates are concerned that the new definition will reduce the number of children with the diagnosis, leaving fewer kids eligible for special services.