Delays in how the brain processes danger and safety may underlie the chronic anxiety that plagues so many people with autism, according to the early findings of two related studies presented at the International Meeting for Autism Research (IMFAR) today.
If confirmed with further research, the results could guide how therapists tailor treatments for anxiety in patients who also have autism.
![]()
Autism Speaks is a major funder of IMFAR. For more daily coverage, click here.
Addressing a major debility in the autism community
Anxiety is a major – frequently debilitating – problem in the autism community. By adulthood, around half of people with autism have been diagnosed with an anxiety disorder such as obsessive-compulsion, a phobia or panic disorder. Many more may struggle with chronic but undiagnosed anxiety, studies suggest.
“Among adults with autism, these anxiety disorders frequently prove more disabling than the direct symptoms of autism,” says study co-author Mikle South, a psychologist at Brigham Young University. These disorders can also be particularly difficult to treat in patients with autism.
“So we went into this study looking for possible difference in what causes anxiety in someone with autism versus anxiety in someone unaffected by the disorder,” Dr. South explains.
Discerning danger
In the first of two studies, the Brigham Young investigators used functional magnetic resonance imaging (fMRI) to track brain activity during a classic “fear training” exercise. Participants included 21 young adults (ages 18 to 29) who had autism and a control group of age- and IQ-matched adults not affected by the disorder.
As each participant lay in the fMRI, he or she watched a screen displaying a random series of blue and yellow squares. (See figure 1 above.) A computer program randomly assigned one or the other color to be followed by startling puff of air to the neck – not every time the color flashed, but often enough to create an association.
This “fear acquisition” segment of the exercise was followed by a “fear extinction” portion. The participant again saw a series of randomly arranged blue and yellow squares. But neither color was followed by the startling puff of air.
All the while, the investigators were tracking activity in two parts of the brain crucial to rapidly discerning threatening situations – the amygdala and the insula. (Figure 2 below.)
In the non-autistic group, the amygdala and insula responded strongly to the “threat” color early in the first half of the exercise. This is the expected response when the brain is learning to recognize a new “danger.” The brain activity in these two regions then faded away across the extinction portion of the exercise – as it became clear that no more disquieting puffs of air were in store. (Red bars in figure 3 below.)
Right place, wrong time?
By contrast, a very different pattern arose among the participants affected by autism. Their brains’ danger-recognition centers responded weakly at best when introduced to the new danger cue. (Blue bars in figure 3 below.) Then, during the fear-extinction part of the exercise, brain activation continued to remain strong – even increase – in the amygdala and insula, when just the opposite should be happening.
“When the danger-recognition response finally came, it was too late,” Dr. South says. “I think what we’re seeing suggests that many people with autism can’t easily discriminate when to feel safe. In that situation – when we don’t know if a situation is safe or not – our natural default is for fear. That’s just instinct to help keep us alive.”
Implications for treatment
The findings suggest possible approaches for tailoring anxiety interventions for those with autism. “If people with autism don’t rapidly process when it’s okay to feel safe, perhaps we should be looking at how to help them better understand their environment,” Dr. South proposes.
By way of example, Dr. South explains how, as a psychologist he often finds himself helping patients with autism recognize that a given situation makes them anxious. This may start with helping them recognize the physical cues that he or she is upset (racing heart, sweaty palms, etc.). Next, the person may need help recognizing what it is about the situation that is triggering the anxiety. It might be a background noise that others don’t notice. Or perhaps a blinking light in the corner.
“Once we help them identify what’s startling them, then we can help them learn that they don’t need to be afraid of it,” he explains. “It may be that those with autism need more external help processing their internal reactions to their environment.”
Neural networks for anxiety?
In their follow-up study, the Brigham Young team found weaker connections between the “fear recognition” centers of the brain (the amygdala and insula) and the frontal cortex. The frontal cortex is the area of the brain responsible for controlling emotions and making decisions.
These pathways between brain regions consist of bundles of high-speed nerve cells and form a network of communication superhighways.
“It’s the job of the frontal cortex to put the brakes on anxiety,” Dr. South explains. “It tells us, ‘hey, that’s just a garden hose and not a snake coiled in the grass.’”
Using another brain imaging method – called diffusion tensor imaging – the investigators found that these white matter pathways to the frontal cortex were, on average, half as large in the participants with autism as in those unaffected by the disorder (figure 4 at right).
“It may be that we need to pay more attention to building these pathways – beginning in early childhood – to help those with autism to better integrate sensory information from their environment,” Dr. South comments. Already many occupational therapists embrace such an approach, which they call sensory integration training, he notes.
For more on sensory integration through occupational therapy, see
“Sensory Integration: Changing the Brain through Play.”
“These insights on how anxiety may arise and persist in people with autism may prove very helpful,” comments Paul Wang, Autism Speaks’ head of medical research. “It may be that we can modify existing behavioral therapies for anxiety in ways that improve their effectiveness for people on the autism spectrum.”
Read the investigators' scientific abstract, as presented at IMFAR here and here.