When Caleb Geary was diagnosed with autism at age 3, he had never spoken or eaten solid food.

Now 6, the boy speaks and tests at his first-grade level — progress that his parents attribute to insurance-based services at home and intensive behavioral intervention at the boy’s school in Hamden.

But they worry what will happen to Caleb’s diagnosis — and the services that have come with it — if the American Psychiatry Association’s proposal to change the definition of autism is adopted.

Lori Geary said she has already fought to get her son the help he needs. Tom Zwicker, Caleb’s father and the director of an autism center for the Easter Seals of Coastal Fairfield County, said he believes insurance companies will start requesting annual diagnostic evaluations if the definition is revised. As a result, his son — and many other children — will lose out on services to treat their conditions.

The autism community has been embroiled in a heated debate for the past few weeks over the proposal to dramatically change the criteria for autism diagnosis in the upcoming fifth edition of the Diagnostic and Statistical Manual of Mental Disorders. The DSM-5, scheduled to be published in 2013, is the first revision since 1994.

The revision would create an umbrella category known as “autism spectrum disorder” that would include traditional autism, as well as Asperger’s Syndrome, childhood disintegrative disorder and pervasive developmental disorder not otherwise specified (PDD-NOS) — which currently are considered separate disorders. A new category, social communication disorder, would also be created.

“What became very apparent is that there aren’t clear boundaries, and that they really are all on a spectrum,” said Darrell Regier, director of research for the APA. The current criteria, he said, is “fuzzy” and as a result some people have been mislabeled as autistic, while others who need treatment can’t get it because their symptoms don’t match the current criteria.

“The thing that we tried to do is be a little more clear about the different deficits that these people have,” Regier said.

But some experts worry that the revision’s main effect will be to drastically reduce the number of people who are diagnosed with autism and who now qualify for services to treat it.

Fred Volkmar, director of the Child Study Center at Yale School of Medicine, is the lead author of a study that found that 44 percent of people previously diagnosed with autism would not meet the proposed new criteria for the diagnosis. The study was based on data collected about individuals in the early 1990s.

“We went back and re-analzyed the data and recoded it,” said Volkmar, who was a member of the DSM task force committee but since has resigned. The methodology was “not perfect,” Volkmar said, “but I don’t think it’s horribly bad either.”

Periodic revisions and refinements of diagnoses are necessary, he said. “Certainly, you could make [the definition of] Asperger’s better. The problem is, how do you justify change and how do you justify major change? It’s an interesting discussion.”

According to Volkmar’s study, to be published in April in the Journal of the American Academy of Child and Adolescent Psychiatry and online later this month, about one-fourth of those diagnosed with autism would not meet the new criteria and nearly three-fourths of those with Asperger’s also would not be diagnosed. Also, 84 percent of those diagnosed with Pervasive Developmental Disorder-Not Otherwise Specified also would no longer meet the criteria.

“More and more people are doing better and better, so we have more people who are out and self-sufficient and independent,” Volkmar said. “And there’s a bit of worry that if you take away services, that that’s the group that will suffer, not just in terms of losing a label but in terms of losing services.”

“Schools have to do a re-assessment every three years. So in three years’ time they say, ‘Oh, this kid no longer qualifies.’ Is that going to be a rationale for no more services?”

‘An Uphill Battle’

Shannon Knall, a mother in Simsbury, said her son was diagnosed with autism when he was 2 1/2.

“I was on the phone every single day with Birth to Three with the state of Connecticut to get 20 hours of early intervention services — and I can tell you, with that advocacy and that constant staying on it, I’ve never gotten 20 hours, and that was eight years ago,” Knall said.

Her son, now 10 and a fifth-grader, has Asperger’s syndrome and is considered high-functioning, at least intellectually.

“I cannot at all comprehend the level of difficulty that will follow without that [autism] diagnosis,” said Knall, the Connecticut advocacy chairwoman for Autism Speaks. “It’s very scary. For us as a community, we are constantly facing an uphill battle to get whatever we can get for our kids. There is always a block in the road, and this is another one. Or could be.”

Knall also worried that a change in criteria could exclude her family and others from the benefits of a state mandate that took effect in 2010. The law requires insurers to cover specialized treatment for children with autism, such as applied behavioral analysis, which can run into the tens of thousands of dollars in a matter of months.

In Caleb’s case, Lori Geary said, she and Zwicker spent $600 a week for four hours of daily ABA therapy before insurance started paying for it last year. The outlook for her son before treatment, she believes, was grim: “I fear he’d be in a group home setting for the rest of his life.”

The boy couldn’t speak. Caleb would explode at changes to his routine. If Geary was driving and made a left turn when he expected a right, there would be “hair pulling, shoes flying, kicking the seat,” she said. Caleb’s work with a therapist involves positive reinforcement to target impulse control, stimming — repetitive movements — and other behaviors.

“It’s how we got him to talk, how we got him to eat, how we got him to sit on a stool at school,” said Geary, a senior project manager at Yale’s Center for Outcomes Research and Evaluation, which studies the effects of health care. Caleb now gets at-home therapy for two hours nearly every day for his social skills.

With the proposed change to the autism criteria, Geary said, “I’m nervous.”

Thousands Of Emails

After news about Volkmar’s study came out, Darrel Regier said his email inbox was “flooded.” He’s received 10,000 emails and counting.

“These are legitimately concerned parents who are worred that their kids are going to be dumped from these services,” said Regier, who serves as research director for the American Psychiatry Association.

The data in the Yale study is old, Regier said, and the diagnoses of the subjects were made at a time when the criteria for autism were still very much in flux. “And [the data] had a high number of very high-functioning people who are not necessarily representative of the general population” of people with autism, Regier said.

Regier cites two field studies that were recently completed that used new data and got very different results. One concludes that the new criteria would decrease diagnoses by only 5 percent and the other concludes that it would actually increase diagnoses by 1 percent. He said the data is still being prepared for publication, so he couldn’t release details of the results.

He said the DSM task force committee agreed on the new criteria 11 to 1, with Volkmar the only member to object. Regier also noted that for much of his career Volkmar has focused on Asperger’s, so he likely would be interested in keeping it as a separate diagnosis.

Services At School

The state Department of Education has been following the APA’s proposal, spokesman Mark Linabury said, and does not expect a revision to change how special services are offered in the public schools.

The state’s most recent data show that 5,866 K-12 students with autism received special education services in 2010-11 — about 1.1 percent of the total public school population in Connecticut. Among special education students, those with autism made up 9.2 percent.

While schools take into account a diagnosis of autism, they conduct their own evaluation of a student’s “functional or academic abilities” to determine whether one is eligible for special education, Linabury said. “The type of services, their frequency, intensity, duration and the personnel assigned would be driven by the student’s needs and not the clinical diagnosis.”

In West Hartford, 154 students this year get special education services because of autism, said Glenn McGrath, the school system’s pupil services director. Last year, there were 140 students, up from 128 in the 2009-10 year. According to state data, 78 town students with autism received services five years ago.

McGrath said it was too early to tell whether a redefinition would have an impact in the town schools, although he believed a child currently diagnosed on the autism spectrum “would still need supports in their educational program to address their social skill deficit.”

But Nelson Rivera, a lead psychologist in the Hartford schools whose responsibilities include students with autism, said it was possible that fewer students would receive special services.

As of Oct. 1, 229 students in the city school system had a diagnosis of autism, Rivera said, of whom 168 are considered higher-functioning on the spectrum and are placed in classrooms with non-disabled peers.

If there is a revision, the objective instruments that schools use to determine a student’s placement could be updated, Rivera said. “It’s going to make the whole area of autism more specific, less vague, in terms of symptoms and behaviors.”

Not Yet Final

The task force committee has until December before anything is finalized, and the new criteria will be subject to three independent reviews in the meantime.

“We just don’t think at this point in time that the study that Fred [Volkmar] is about to release is one with a good database to make these dramatic projections,” Regier said.

“I don’t mind using old data and doing what Fred did, but you need to call that a hypothesis-generating study,” he said. “To come out and say definitely that you’re going to lose 40 [percent] is just not a justifiable claim, based on that dataset.”

Dr. Deborah Fein, a neuropsychologist and psychology professor at the University of Connecticut, has mixed feelings about the proposed new criteria.

“The point about collapsing everything into the autism spectrum disorder, I do think that makes sense because I don’t think there’s adequate evidence that there’s a significant difference between them.”

Fein said she has her doubts that the effect would be as dramatic as Volkmar’s study suggests. “I think he was looking at a particular slice of the pie,” she said.

“The population I’m most concerned about is toddlers.”

A study that she recently worked on found that 20 percent to 25 percent of the toddlers currently diagnosed as having Autism Spectrum Disorder would not qualify under the new criteria. Many of these children would likely develop additional autism-related symptoms a few years later, as is common, and then meet the criteria.

But by that time, she said, they would have missed out on a few years of services, and early intervention is crucial in treating autism.