Can People Really Grow Out of Autism?

Emily Willingham

Let’s start with the headlines blaring the news about a recent autism study. They almost invariably use the phrase “grow out of autism,”
even though the study itself does not use that phrase or even reference
“grow” except to talk about head circumference. Instead, the authors of
the report, published in The Journal of Child Psychology and Psychiatry,
use the term “optimal outcomes” to describe what they detected in a
group of 34 people who were diagnosed as autistic when they were under
age 5.

As the study authors themselves state, this idea that autistic people
might show reduced deficits to the point of losing a diagnosis is not
new. In fact, first author Deborah Fein and colleagues cite studies
identifying frequencies of “optimal outcomes” as high as 37% among
autistic people. The lingering open questions relate to whether or not
the autistic people in these studies had received the correct diagnosis
in the first place. The only “novelty” of these latest results appears
to be confirmation that indeed, the 34 people they identified as having
an “optimal outcome” did receive an accurate diagnosis of autism in
childhood. In other words, they are confirmed to have had a
developmental disorder, a neurobiological condition called autism — yet, they “grew out of it.”

The rest? Nothing new. The people who show these optimal outcomes
tend to have started out with, as the authors describe it, “higher
cognitive functioning and somewhat milder initial symptoms.” Many of
them had behavioral interventions in childhood. The researchers point
out that the perception that everything’s all hunky-dory for the 34
people they evaluated does not rule out their having “residual
difficulties” with various aspects of autism, including executive
function–think project management–or language or social interaction.

For each participant, whose ages ranged from 8 to almost 22 years,
the researchers interviewed the parents. One of the required parental
answers for a participant to be considered for the optimal outcome
category? The parent had to report that their child/adult child “had
typically developing friends.” That question seems to imply a certain
low expectation for autistic people, many of whom I know have “typically
developing friends” despite themselves still being autistic. It
certainly suggests that for people who continue to meet the criteria for
an autism diagnosis, any pursuit of a friend of the “typical” sort
would be futile.

How did those 34 “optimal outcome” participants do on the various measures of “are you autistic or not”? Well, seven
of them–that’s >20% of the group–turned out to have “some
impairment” in nonverbal social interaction. For reasons that are
unclear, the researchers decided that these impairments were not the
result of “an autistic quality” but of “inhibition, anxiety, depression,
inattention and impulsivity, embarrassment, or hostility.” Of course,
each of those themselves could be secondary to autism. Even though this
20% showed impairment, they were retained as being “optimal outcome”
folk, those who, as the news media report might say, “grew out of”

To determine whether or not these autistic people were still
autistic, the research team administered a battery of tests; they list
eight in their paper. Of these eight, three were parent completed. One
consisted of clinician observation and scoring. One was an IQ test. One
was a handedeness test (left-handedness is more common among autistic
people), and the remaining two evaluated facial recognition and
language. ETA: None of them examined if the participants retained any of
the positive aspects of being autistic–refined sensory detection, detail orientation, pattern finding, etc.

Let’s go back to those news media stories. In one interview,
lead author Fein commented that “these people did not just grow out of
their autism.” She then goes on to credit the hard work of the parents
and therapists of “these people” for any improvements, but I’d suggest
that “these people” also did a lot of hard work–and probably still are
doing so. The thing is, no one seems to have asked “these people” about

Among the many articles covering this study, I couldn’t find a single
one featuring an autistic person commenting about the report. As you
can see from how the researchers evaluated their 34 “optimal outcomers,”
they don’t appear to have asked said optimals about how their internal
function jibes with the external results or what they do to achieve
those results. Because no one else seems to have bothered to ask
autistic people these questions, I did. I polled the autistic community
via social media, asking autistic adults what’s going on inside them
when they appear outwardly typical and asking any readers if they felt
they’d “grown out of” autism.

Their response was immediate and intense. “I don’t ‘look’ like I have
it, but I do,” responded one autistic woman, who went on to describe
how she’s learned over time about different expectations for behavior
and tried to apply those. Others describe using pattern recognition to
navigate socially, while still others report having an “optimal
outcome”-like period in later childhood but then experiencing a trough
of struggles in early adulthood as new responsibilities and expectations
arose. They wrote to me about self-monitoring, about working hard to
compensate in social situations but then experiencing crashing
exhaustion afterward. They talked about self-selecting their social
groups as adults as a way of feeling more socially at ease. The concepts
that came up again and again and again were “compensating” and

It’s not a huge surprise that autistic people with average or
above-average cognitive abilities might be able to intellectualize
social rules and algorithms and put them convincingly into practice.
Does that ability mean that they aren’t really autistic? The real crux
to answering that is this: Do we view autism only as a clinical
diagnosis based solely on behavior and outward function, or do we talk
about it as a neurobiological construct and identification, with an
understanding of the context of the hidden disability and the hard work
that those outward behaviors require?

Many conditions that we measure either directly with lab tests or
behaviorally can lie under a mask of apparent normalcy or typicality. A
woman with diabetes who maintains her blood sugar at a healthy level
through diet and medication still has diabetes. A person with
obsessive-compulsive disorder who fights successfully every second of
every day against caving to obsession or compulsion still has the
disorder. Anyone who has ever put on a public face when all they wanted
to do was stay in bed should understand something about doing the
internal hard work of compensating for a disability without showing
outward manifestations of it.

Does that capacity mean, in the parlance of the news media reports or an editorial accompanying the paper,
that the up to 25% of autistic people who can do this are “recovered”
and no longer autistic? Or does it mean, rather, that they’ve become
increasingly adept at meeting the interaction standards of the social

A version of this post was published at