Road to "Recovery": What Does it Mean to Lose an Autism Diagnosis?

Marina Sarris
Interactive Autism Network at Kennedy Krieger Institute

Date Published: January 27, 2016

Such a simple gesture: waving good-bye. Toddlers see their parents do it, and repeat it. But children with autism don't instinctively imitate other people, not usually. Jake Exkorn was one such toddler. In a home video, Jake is being taught to wave bye-bye by an autism therapist; she grabs his arm and waves it for him, again and again. He seems wooden, remote.1, 2

Then, fast forward about 15 years, there's Jake as a young adult, being recorded by another video camera. This time, he smiles easily, warmly. "When I see the [old] videos it amazes me that at one point in my life, I had autism, and it makes me realize how fortunate I am to recover," he says.2 Jake is among a minority of people who "lose" their autism diagnosis. He doesn't have many memories of having autism, just those videos.

Nothing inspires more interest, debate, and even concern, than the idea that someone could "recover" from a brain-based condition like autism. The notion fuels an industry that offers treatments, both proven and dubious; triggers both hope and sadness; and fuels political and academic controversy. Is recovery real, and is it good?

Studies have estimated that from 3 percent to 25 percent of children with autism lose their diagnosis.3 Some have wondered: did those who recover really have autism at the outset? Did they lose their diagnosis but still have autistic symptoms, even subtle ones?

Researching The Loss of Autism Diagnoses

Psychologist Deborah Fein PhD has probably done more to document this phenomenon than anyone. She first noticed that some of the children she was following no longer met the criteria for autism. "I was definitely very surprised," she recalled. Like most clinicians, she believed autism was a lifelong condition. For her research, she sought out other children who also left their autism diagnosis behind.

She and her co-researchers meticulously documented the diagnoses of 34 children with autism, and their loss of those diagnoses and autistic symptoms. They tested their communication, reading comprehension, academics, language, and social abilities, even their ability to recognize faces. They compared them to two other groups, 44 people with "high-functioning autism," and 34 people who developed typically, called controls. High-functioning autism usually refers to autism with average or "normal" intelligence.

Were people in the lost-diagnosis group really free of autism, or were they "pretending to be normal,"7 to borrow the words of one author with autism?

By all measures, this group seemed to be functioning no differently than people who never had autism.4, 5, 6 They even did well with daily living skills.4 Daily living skills can befuddle people with autism who have average and above average intelligence, according to research on teens in the Simons Simplex Collection project.8 The lost-diagnosis group also did not differ from controls on measures of repetitive and ritualistic behavior, a symptom of autism.9

The Baggage of "Recovery"

But Dr. Fein avoids the word recovery to describe her lost-diagnosis group. "'Recovery' carries so much baggage," Dr. Fein said in an interview. "When you say recovery, it conjures up a period of normal development, then they got a disease, and now they've recovered," she said. Instead, this group reached an "optimal outcome," or OO for short.

She's not alone in shying away from the word recovery. In a different study, other autism researchers used the phrase "very positive outcome" to describe a group of teenagers who had shed their diagnosis and become full participants in the "social world."10

Recovery is often associated with a groundbreaking 1987 research report by psychologist O. Ivar Lovaas. Dr. Lovaas said that nine of the 19 children who underwent his form of Applied Behavior Analysis (ABA) therapy essentially "recovered" from autism. The children had received treatment during "most of their waking hours." They had "normal intellectual and educational functioning," and were "indistinguishable from their normal friends" by first grade, according to his research report.11 But he acknowledged that "certain residual deficits" could remain, perhaps faint whispers of autism hidden from parents and teachers.

Dr. Fein's research group has looked for remnants of autism in the group that no longer qualifies for the diagnosis. In a 2016 study, the researchers used a test not available when the Lovaas research was published: functional Magnetic Resonance Imaging (fMRI). An fMRI uses magnets to measure someone's brain activity while he's doing something; these scans provide a window into how the brain works. The optimal outcome children and adults may look "generally indistinguishable" from others, but do their brains, originally wired for autism, function differently?

For that study, researchers led by Dr. Inge-Marie Eigsti performed fMRI scans on three groups as they read short sentences and answered questions by pressing a button.12 They tested 16 people with "optimal outcomes" (OO), 23 with high-functioning autism, and 20 typically-developing peers, the controls. Would the scans of the OO group be identical to the control group?

Surprisingly, the answer was no. Their scans resembled those of the autism group in some ways, but in others, they differed from both the control and the autism groups. The OO group appeared to be using different parts of their brains for the sentence comprehension tasks. Researchers say they may have learned to compensate for their early autism by using new pathways to process language.

"It's a confirmation that they really were autistic when they were little," Dr. Fein said. Early intervention may have helped them function typically, but "they do not use the same areas of their brain as their peers who never had autism." She cautioned this study, like all new research findings, should be repeated to see if others get the same results.

Who Loses a Diagnosis?

Families often want to know how children achieve optimal outcome. Was there something different about them?

Research suggests that people who lost their autism diagnosis often:

  • Were diagnosed younger, such as before 31 months of age,13, 14
  • Had better cognitive abilities,10 and milder symptoms, particularly in the social arena, as young children,4, 14, 15
  • Had a larger reduction in repetitive behaviors (such as hand-flapping or finger-flicking) between ages 2 and 3,10
  • Underwent earlier, and more intensive autism intervention,10 especially therapies based on Applied Behavior Analysis.5, 11

ABA-based therapies have a history of effectiveness, but Dr. Fein is quick to point out that they are not a magic pill. Families should not assume that their children would have lost their diagnosis if only they had received ABA therapy, she said. "It is sad when a parent of 15- or 20-year old says, 'If only I had done ABA when he was 2, he'd be in a different place.' This is not necessarily the case at all." More seems to be involved in outcome than just ABA, she said. Some children may have forms of autism that are more likely to subside with treatment or maturation.

"I've seen kids get high-intensity, high-quality ABA therapy and not make the progress that other kids did," she said. It can be hard to predict, with some children ending up in far different places than she would have expected based on how they seemed at 3. "I evaluated one child who I thought would make good progress, but he didn't, even though he was getting very good therapy. Now he's a young adult with severe autism and limited language."

Some communities and families do not have access to ABA, so she has co-authored a book to help parents use ABA principles at home with their children, The Activity Kit for Babies and Toddlers at Risk.

Another Perspective of Outcome Research

Outcome research can determine which therapies work and underscore the importance of early diagnosis and intervention. But some say it carries a downside. "Autism is a natural part of the human condition and not something to recover from or eliminate," according to a statement by the Autistic Self Advocacy Network (ASAN), a disability rights group run by autistic people.

"Words like recovery and optimal outcome do reinforce a really damaging idea, as the obvious corollary is a suboptimal outcome, which is a terrible thing to call a person," explained Julia Bascom, deputy executive director of ASAN. "There's this idea that being non-autistic, non-disabled is better than being autistic and being disabled, which isn't true. We connect that mindset with a lack of research on what services and supports work best for autistic people across their lifespan, and on how to diagnose adults," she said.

"People think that if you can cure autism, if it's a sickness that people can recover from, then money would be better spent on research into how to prevent autism, rather than on supporting people across their lifespans," said Ms. Bascom, who is autistic.

Ms. Bascom says ASAN hears from adults who may not meet criteria for a current autism diagnosis but who "have clinical depression, social skills issues, and sensory integration problems. That doesn't sound like a recovery," she said.

Dr. Fein and her fellow researchers did not find more depression in the optimal outcome group, but they did report more ADHD and phobias than in the control group.16

In an interview, Dr. Fein said losing one's diagnosis and autistic symptoms is only "one kind of optimal outcome," but there are many kinds. She points to Temple Grandin PhD, the animal scientist with autism. "She writes and speaks about autism, and has made many contributions in her [scientific] work, and she's still autistic. Having a good life while still having autism is a form of optimal outcome. It just doesn't happen to be what we're studying."

Additional Resources: 


  1. Donvan, J., Zucker, C., Dawson, D. & Valiente, A. (October 6, 2015). How a child with autism became 'his own man' after treatment. Retrieved from
  2. Corcoran, K. (October 7, 2015). University freshman diagnosed with severe autism as a child recovers to lead a normal life after years of experimental therapy. Retrieved from Includes video uploaded from University of Connecticut's Youtube account at
  3. Helt, M., Kelley, E., Kinsbourne, M., Pandey, J., Boorstein, H., Herbert, M., & Fein, D. (2008). Can children with autism recover? if so, how? Neuropsychology Review, 18(4), 339-366. doi:10.1007/s11065-008-9075-9
  4. Fein, D., Barton, M., Eigsti, I. M., Kelley, E., Naigles, L., Schultz, R. T., . . . Tyson, K. (2013). Optimal outcome in individuals with a history of autism. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 54(2), 195-205. doi:10.1111/jcpp.12037 [doi]
  5. Orinstein, A. J., Helt, M., Troyb, E., Tyson, K. E., Barton, M. L., Eigsti, I. M., . . . Fein, D. A. (2014). Intervention for optimal outcome in children and adolescents with a history of autism. Journal of Developmental and Behavioral Pediatrics, 35(4), 247-256. doi:10.1097/DBP.0000000000000037 [doi]
  6. Troyb, E., Orinstein, A., Tyson, K., Helt, M., Eigsti, I. M., Stevens, M., & Fein, D. (2014). Academic abilities in children and adolescents with a history of autism spectrum disorders who have achieved optimal outcomes. Autism: The International Journal of Research and Practice, 18(3), 233-243. doi:10.1177/1362361312473519
  7. Willey, L. H. (1999). Pretending to be normal: Living with asperger's syndrome. Philadelphia, PA: Jessica Kingsley Publishers.
  8. Duncan, A. W., & Bishop, S. L. (2015). Understanding the gap between cognitive abilities and daily living skills in adolescents with autism spectrum disorders with average intelligence. Autism: The International Journal of Research and Practice, 19(1), 64-72. doi:10.1177/1362361313510068
  9. Troyb, E., Orinstein, A., Tyson, K., Eigsti, I. M., Naigles, L., & Fein, D. (2014). Restricted and repetitive behaviors in individuals with a history of ASDs who have achieved optimal outcomes. Journal of Autism and Developmental Disorders, 44(12), 3168-3184. doi:10.1007/s10803-014-2182-y [doi]
  10. Anderson, D. K., Liang, J. W., & Lord, C. (2014). Predicting young adult outcome among more and less cognitively able individuals with autism spectrum disorders. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 55(5), 485-494. doi:10.1111/jcpp.12178 [doi]
  11. Lovaas, O. I. (1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. Journal of Consulting and Clinical Psychology, 55(1), 3-3-9.
  12. Eigsti, I., Stevens, M. C., Schultz, R. T., Barton, M., Kelley, E., Naigles, L., . . . Fein, D. A. (2016). Language comprehension and brain function in individuals with an optimal outcome from autism. NeuroImage: Clinical, 10, 182-191. View article
  13. Wiggins, L. D., Baio, J., Schieve, L., Lee, L. C., Nicholas, J., & Rice, C. E. (2012). Retention of autism spectrum diagnoses by community professionals: Findings from the autism and developmental disabilities monitoring network, 2000 and 2006. Journal of Developmental and Behavioral Pediatrics, 33(5), 387-395. doi:10.1097/DBP.0b013e3182560b2f [doi]
  14. Turner, L. M., & Stone, W. L. (2007). Variability in outcome for children with an ASD diagnosis at age 2. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 48(8), 793-802. doi:JCPP1744 [pii]
  15. Blumberg, S. J., Zablotsky, B., Avila, R. M., Colpe, L. J., Pringle, B. A., & Kogan, M. D. (2015). Diagnosis lost: Differences between children who had and who currently have an autism spectrum disorder diagnosis. Autism: The International Journal of Research and Practice, doi:1362361315607724 [pii]
  16. Orinstein, A., Tyson, K. E., Suh, J., Troyb, E., Helt, M., Rosenthal, M., . . . Fein, D. A. (2015). Psychiatric symptoms in youth with a history of autism and optimal outcome. Journal of Autism and Developmental Disorders, 45(11), 3703-3714. doi:10.1007/s10803-015-2520-8 [doi]
These archived articles were originally published as part of the Interactive Autism Network (IAN) research project. 
The project is closed and no longer accepting participants.

More about Autism Research

Center for Autism and Related Disorders

Subscribe to news and updates

<< Back to the Archives