Dr. L. Scahill, Yale University School of Medicine


The following is a modified transcription of Dr. Scahill’s talk. Any glaring omissions or errors are my own. INSAR members can listen to Dr. Scahill’s talk in its entirety via its abstract page. -SR

Fears, phobias, and worries; Toward better measurement of anxiety in ASD (Autism Spectrum Disorders)

Dr. Scahill is very interested in identifying compounds that can treat target problems in people with autism and Tourette’s syndrome and the like. But how do we get to measure the targets that we think are important?

We’ve seen in many posters [at IMFAR] that anxiety is common and adds to despair and disability in children with autism, but we need to find a way to measure it. This is an inspection of how we might go about that.

img_4541-5036379Disclosures: Dr. Scahill has consulted with companies who have paid attention to autism and Tourette’s. They have compounds that should be of interest to us that in some cases are not yet on the market.

The questions he’d like us to think about — he can’t say we’re going to answer all of them — are:

  • How common are anxiety disorders in children and adolescents with ASD?
  • Are some anxiety disorders more common than others in ASD?
  • Does anxiety look the same in ASD vs typically developing youth?
  • Should we be thinking about dimensions rather than categories?
  • Is there overlap between anxiety and ASD? Or is there a blend, a subgroup of children with autism who have anxiety as part of their syndrome?
  • Are there differences by IQ?

By doing a survey of parents, it was found that children with autism have lots of fears. He actually uses “how does the kid do at the barber shop?” as part of his clinical interviews. If a kid can make it through a trip to the barber shop, and it’s not a big event, that kid’s probably not too anxious. But then there are kids who can’t go to the barber, and can’t go to the dentist. These kids have lots of fears, and this is something the parents have been telling us about — a lot. These kids also can protest changes in routines, for instance not letting their mother turn right out of the driveway.

Manifestations of Anxiety in ASD:

  • Phobias
    • Loud noises
    • Barbers and dentists
    • Storms
    • Crowds
  • Anxiety
    • Protest changes in routines
    • Reluctance in new situations
    • Protest separation from parents
    • Avoid social situations – seen more in teenaged kids
    • Excessive tension and restlessness

What we see by looking at a number of studies, on mostly kids, is high rates of anxiety. There are several different methods of assessing this anxiety, but they all show, fairly similarly, that there is an elevated level of anxiety disorder or symptoms in kids with ASDs.

Several kids had anxiety symptoms in more than one diagnostic category. We see this in typically developing children with anxiety as well. But when we look at the rates, we see many children who have more than one disorder.

ASD Twin study in the UK (Hallett et al.) had access to data from 128 twin pairs where at least one twin had screened positive for ASD. Most of the twins has IQs above 70. The controls who screened negative for ASD were also screened for anxiety. Parents filled out evaluations, with an anxiety score from 0 to 3, in diagnostic categories — which is not the same as a diagnosis. Categories were:

  • Total Anxiety
  • Social Anxiety
  • Separation Anxiety
  • OCB (Obessive Compulsive Behavior)
  • Generalised [sic] Anxiety
  • Panic
  • Depression

There were dramatic differences in the twins with autism compared to the non-ASD controls matched for age and gender and more or less by IQ. Really dramatic differences, especially in Total Anxiety, Separation Anxiety, OCB, and Generalised Anxiety.

Initial conclusions:

  • Fair to say that youth with ASD are at higher risk for anxiety disorders. 
  • Most common:
    • Simple phobia
    • Social phobia (wide variation)
    • Separation Anxiety
    • Generalised Anxiety Disorder
  • Multiple anxiety disorders often present
  • Dimensional approach is best fit for treatment studies (i.e., inclusion and outcome)

He has mined and studied data from four large-scale multi-site trials of Risperidone [Risperdal], Methylphenidate [Ritalin], Citalopram [Celexa], and Risperidone plus Parent Training. Characteristics included age, gender, ASD or not, and IQ. He used the 20-item CASI Anxiety scale and the ABC Aberrant Behavior Checklist.

He thought repetitive behavior and anxiety would go together, but based on the results from the Citalopram study which included screening for repetitive behaviors — it doesn’t appear to be so. Though subjects had lowered irritability, which was not a surprise. Also lowered self-injury, and social withdrawal — meaning the extend to which a person responds to an initiated action from another person and the extent to which they initiate action on their own.

What he saw is that there is no real pattern, no sense that the anxiety scale is mapping so easily. As soon as he thought he might see something, another column showed that it doesn’t really matter. Scahill thinks that’s good, that means the data is measuring something different.

Top seven CASI items based on two samples from RUPP [Research Units on Pediatric Psychopharmacology] sample. They’re parent-measured, but he considers them believable.

  • More anxious in social situation than most kids
  • Acts restless or edgy
  • has difficulty falling asleep
  • Is overly fearful of specific objects
  • Upset when expects to be separated from parents
  • Is extremely tense or unable to relax
  • Is excessively shy with peers

Bottom six CASI items in both samples are items that represent a high level of inference from parents — things kids can’t tell their parents about if they’re non-verbal:

  • Worries about physical health
  • Complains about heart pounding
  • Worries that disaster will separate them from their parents
  • Has nightmares about being separated from parents
  • Complains about feeling sick when expects separation from parents
  • Worries about being left home alone or with sitter

Social anxiety. Scahill thought there was going to be a real connection between anxiety and social anxiety. Particularly in “higher functioning” kids.

  • Even vague awareness of social disability may induce anxiety in social situation
  • “Failures” in social interaction may promote social avoidance – if you’re not doing so well, it’s an aversive to be in those situations
  • Presence of social anxiety may amplify social avoidance and social disability – because even if they’re interested, kids are not putting themselves in situations where they can benefit from interaction

This caused him to want to look into social anxiety. So, he looked at a 20 country, 103,000+ person survey conducted by WHO. The WHO wanted to know: are there differences in social anxiety between developing and non-developing countries?

If you look at the DSM-IV, you’ll see that there are two parts to social anxiety factors:

1) Interaction items — more relevant to ASD

  • Meeting new people
  • Going to parties
  • Expressing disagreement
  • Talking with strangers
  • Dating situations
  • Entering an occupied room

2) Performance items — not endorsed as often

  • Writing/eating/drinking while being watched
  • Using public bathrooms
  • Speaking in meetings/class
  • Working while being watched
  • Public speaking/performance

What about differences by IQ? It doesn’t seem to matter than much based on correlations with the 20-item CASI Anxiety Scale. The good news is that is that it’s possible to identify an anxiety measure, parent-rated, that cuts across IQ.

CASI Anxiety Scale: Tentative Conclusions

  • Shows orderly distribution (across IQ range)
  • Solid internal consistency
  • Appears to be measuring a separate construct
  • Social anxiety (unfortunately) not well-covered
  • Separation Anxiety “over-represented” (8 items out of 20)
  • Six items have low rate of endorsement
  • Unclear if sensitive to change – we need to test that

Overall conclusion: anxiety disorders appear common in ASD

  • Generalized Anxiety Disorder, separation anxiety and social anxiety appear to be most common (social anxiety appears to the be variable depending on the sample and the method used to define it).
  • A dimensional approach fits best for treatment and trials — not really a conclusion, but an opinion — and he’s sticking to it, because he thinks it’s the right approach.
  • Unanswered at the moment is [anxiety as a] co-occurrence versus an ASD symptom. He’s not so sure he’s thinking of it as a co-occurrence; he’s thinking this is a problem for perhaps a large subgroup of children with ASD.
  • Anxiety in ASD vs. typically developing kids: He doesn’t think we know the answer to this, but it would be a good thing to find out. 
  • We need to know more about differences in IQ.