Often Undiagnosed but Treatable: Case Vignettes and Clinical Considerations for Assessing Anxiety Disorders in Youth with Autism Spectrum Disorder and Intellectual Disability

Anxiety, Autism, and Intellectual Disability

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by social communication deficits and restricted and repetitive behaviors (American Psychiatric Association [APA], 2013). The United States prevalence rate of ASD in 2016 was 2.76% (Zablotsky et al., 2017). It was previously believed that the majority of individuals with ASD also have intellectual disability (ID), (Thurm et al., 2019), which is diagnosed when individuals shows deficits in both intellectual functioning (IQ<70) and adaptive behavior (APA, 2013). However, more recent reports estimate that only 40% of those with ASD have concurrent ID (Baio et al., 2014). This shift in reported rates for those with ASD and ID is thought to be fueled by changes in the diagnostic criteria of ASD as well as a greater prevalence of those being diagnosed with autism having average or above average IQ (Baio et al., 2014).

Importantly, individuals with both ASD and ID are at high risk for multiple co-occurring conditions, such as attention deficit hyperactivity disorder (ADHD), sleep and gastrointestinal problems, and seizures (Mannion & Leader, 2013). In addition, children with both disorders experience mental health conditions such as anxiety and mood disorders (Matson & Nebel-Schwalm, 2007). Diagnosing any co-occurring condition in individuals with ASD and ID is challenged by limitations of language and cognition in reporting symptoms, resulting in varying ranges in the literature (Mannion & Leader, 2013). Furthermore, variation in study methodology and design, measurement, and type of sample (e.g., community vs. clinical) also contributes to the prevalence discrepancies. For anxiety specifically, some studies suggest that children with ASD and ID experience less anxiety (Mayes et al., 2011; Mingins et al., 2020; Sukhodolsky et al., 2008), some studies suggest that the risk is equal (Kerns et al., 2020), and one meta-analysis suggested a heightened rate of anxiety in autistic individuals with ID (van Steensel et al., 2011). Despite the potential clinical significance of anxiety for this population, it can be challenging to diagnose, and is often overlooked.

The Challenge of Diagnosing Anxiety Disorders in Autistic Individuals

As described in the most recent 5^th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; APA, 2013), there are 6 major forms of anxiety disorder including: Separation Anxiety Disorder, Specific Phobia, Social Anxiety Disorder, Generalized Anxiety Disorder, Panic Disorder, and Agoraphobia (see Table 1). This study did not include Panic Disorder or Agoraphobia, which are low prevalence conditions in childhood (1% or lower; Beesdo et al., 2009). All of these forms of anxiety can significantly interfere with social skill development, daily functioning, and level of independence in both autistic and non-autistic children (Grondhuis & Aman, 2012; Swan & Kendall, 2016; Vasa et al., 2016).

Table 1.

DSM-IV and Distinct Anxiety Included in this Study, Descriptions, and Examples

Anxiety Disorder	Description	Example(s)
Separation Anxiety	Excessive anxiety when separating from home or attachment figures	Clings to parent and follows them around everywhere
Specific Phobia	Significant anxiety induced by exposure to the feared object or situation	Needles, dentist, heights
Social Anxiety Disorder	Significant anxiety induced by exposure to certain social situations	Speaking in front of class
Generalized Anxiety Disorder	Persistent and excessive worry about everyday things	Worrying about possible bad things that can happen in the world
Uncommon Phobia*	Similar to specific phobia but with distinct content	Fears of glasses, beards, toilets, specific sounds
Special Interest Fear*	Excessive anxiety related to a restricted or repetitive interest	Excessive worry about missing the garbage truck
Other Social Fear*	Significant fear around people without evidence of fear of negative evaluation	Anxious confusion, worry, and/or hiding around people
Fear of Change*	Anxious anticipation of and distress following changes or novelty	Worry about changes in schedule or going to new places
Other Distinct Anxiety*	Fears/worries/anxieties present but do not fit in any of the categories above	Fear of the house (see case Everly)

^*Distinct Anxiety (Kerns et al., 2017)

In addition to the DSM-5, there is a diagnostic manual for areas of mental health in those with intellectual disability, called the Diagnostic Manual – Intellectual Disability (DM-ID-2; Fletcher et al., 2018). The DM-ID-2 separately specifies whether adaptations should be made when applying DSM criteria to those with Mild ID, Moderate ID, and Severe to Profound ID. For the four anxiety disorders assessed in this study, there were typically no adaptations suggested for those with Mild or Moderate ID. However, for Severe to Profound ID, the adaptations typically include: “fear can be observed rather than subjectively described.” In addition, there are some specific examples depending on the anxiety disorder. For example, Specific Phobia includes “fear or anxiety may be expressed by crying, tantrums, freezing, or clinging.”

Notably, fears that do not align with these traditional categories have also been observed in children with ASD. These variations in the expression of anxiety, which are referred to as “distinct anxiety,” align closely with hallmark features of ASD, but are also characterized by fear and avoidance consistent with an anxiety disorder (Kerns et al., 2016; Vasa et al., 2016). Distinct expressions of anxiety are responses to unusual and specific stimuli that would not be expected to the same degree in those without autism. Table 1 includes examples of the distinct anxiety areas referred to as uncommon phobia, special interest fear, fears of change, and other social fear (Kerns et al., 2014). If impairing distinct anxiety is present, the DSM diagnosis of Other Specified Anxiety Disorder may be considered (Kerns et al., 2016) in order to denote cases that may require an anxiety-focused treatment that might otherwise be overlooked or left unmanaged.

Along with identifying distinct anxiety, symptom overlap is common in autism and anxiety, contributing to the difficulty of detecting anxiety. For example, avoidance of social situations is a common feature both in individuals with ASD and in people with social anxiety (Lecavalier et al., 2013). In youth with autism, obsessive thoughts are often reasonably construed as symptoms aligning with the child’s primary diagnosis of ASD, but in reality, these can be clinical indicators of a co-occurring anxiety disorder when they are accompanied by fear and avoidance (Helverschou & Martinsen, 2011). In these situations, “diagnostic overshadowing” or the failure to recognize mental health symptoms in the presence of a salient disorder (Mason & Scior, 2004), is one reason why the diagnosis of anxiety is challenging. Another reason why anxiety in ASD and ID might go undetected includes a belief that individuals with ASD and ID lack the cognitive capacity to worry about future events and, therein, experience anxiety. Furthermore, individuals with ASD and ID are less likely to describe their symptoms verbally due to language impairments and/or reduced emotional awareness and, when they do, the anxiety might be expressed in unconventional ways (e.g., repetitive requests to see their schedule) or about unconventional content (e.g., fear of leaves).

In response to observing idiosyncratic triggers of anxiety (i.e., distinct anxiety), Kerns and colleagues created an Autism Spectrum Addendum (ASA; Kerns et al., 2014, 2016, 2017) to be added to the gold standard Anxiety Disorders Interview Schedule for DSM-IV (ADIS-P; Albano & Silverman, 1996) to assess both clinically significant traditional anxiety according to DSM criteria, as well as distinct anxiety in youth with ASD. Both of these measures are more thoroughly described below.

In a first study (Kerns et al., 2014; IQ range=67-158, M=104.69), similar proportions of children presented with DSM anxiety disorders (48%) and distinct anxieties (46%) overall, with 31% presenting with both DSM and distinct fears, 17% with only traditional anxiety and 15% with only distinct presentations. Recently, Kerns et al. (2020) examined anxiety in 75 children (aged 9-13) with ASD (36 also with ID; IQ range≤25-170, M=77.39) using the ADIS-P and ASA. This was a longitudinal study of ASD that began when the children were toddlers and, therefore, not recruited based on anxiety symptoms. Similar to Kerns et al. (2014), 69% of children met criteria for at least one anxiety diagnosis. When comparing those with ASD and those with ASD and ID, the rates of anxiety disorders did not differ. Similar findings were observed in both groups with DSM-anxiety disorders (56% ASD vs. 47% ASD/ID) and distinct anxiety (49% ASD vs. 47% ASD/ID) (Kerns et al., 2020). However, DSM anxiety disorders in the ASD and ID group were comprised solely of specific phobias and separation anxiety disorder. This is consistent with other studies of anxiety in youth with ASD, which find that specific phobias are often the most common type of traditional anxiety (Kerns et al., 2014; Kerns et al., 2020; van Steensel et al., 2011).

Screening/Assessment Factors in Understanding Rates of Anxiety in Youth with ASD/ID

Within ASD research, estimates of elevated levels of co-occurring anxiety in youth range widely between 11-84% (van Steensel et al., 2011). Imprecise detection of anxiety disorders in children with ASD and relatively lower cognitive and verbal functioning likely contributes to this broad range. Self- and parent-report measures are often used to assess anxiety. Endorsing symptoms of anxiety in self-report questionnaires and asking parents to infer about anxiety in their children in parent-report questionnaires both require that the child have a sufficient level of expressive language and the ability to convey internal states. For example, prompts that begin with “my child worries about” or “when my child is frightened, he feels” are difficult to answer with certainty for a child with autism or ID, who may not state emotions verbally or exhibit emotions in an atypical manner. Self-report measures may be impossible to administer to children with minimal verbal abilities. In sum, many of these relatively easy to administer measures rely on the verbal abilities of typically developing children, for which the measures are designed and standardized (Lecavalier et al., 2013).

Consistent with these concerns, emerging research also suggests that parent-report questionnaires have reduced validity and precision in ASD samples, particularly those with individuals with co-occurring ID (Dovgan et al., 2019; Kerns et al., 2015, 2020; White et al., 2015). For example, the Multidimensional Anxiety Scale for Children, Parent Report (MASC-P; March et al., 1997), The Child Behavior Checklist/6–18 years, Parent Report (CBCL; Achenbach & Ruffle, 2000), and Screen for Child Anxiety and Related Emotional Disorders, Parent Report (SCARED-P; Birmaher et al., 1999) demonstrated weak sensitivity for detecting any clinically significant anxiety (MASC-P at 0.0, CBCL at 13.0, and SCARED at 9.1) when compared to a semi-structured interview like the ADIS-P, in a sample of youth with ASD and ID (Kerns et al., 2020). Notably, greater language ability has been associated with increased endorsement of DSM-consistent anxiety (Davis et al., 2011; Kerns et al., 2014; Rodas et al., 2017). Moreover, parents of children with ASD and ID may, like clinicians, find it challenging to differentiate anxiety from other problem behaviors in their children (Tarver et al., 2020), but have little opportunity to discuss or clarify their responses when using a questionnaire format. Notably, studies using semi-structured interviews, which allow clinicians to follow up and query ambiguous responses, tend to result in a tighter range of problematic anxiety in youth with ASD than those using questionnaires, potentially because they allow clinicians greater latitude to address these diagnostic challenges (Kerns & Kendall, 2012).

However, it must be noted that there certainly are limitations to conducting semi-structured interviews in clinical settings including accessing the training, the cost of the interview, and having the time to conduct the interview. In this case, gaining the knowledge to differentiate anxiety from ASD and to detect potential distinct areas of anxiety would be recommended to address the goal of more accurate assessment of anxiety in those with ASD/ID.

Observing Fear Responses

When attempting to diagnose anxiety in autistic individuals with ID, it can be useful to consider the Defense Cascade or the 6 Fs: Freeze, Flight, Fight, Fright, Flag, and Faint (Lang et al., 2016; Schauer & Elbert, 2010) as a theoretical model applicable to understanding and behaviorally observing anxiety and fear. This model suggests that Freeze is an orienting response for the person to stop and attend to the concern of threat. Flight and Fight are “uproar reactions” and particularly regulated by sympathetic activation while this arousal reaches its height at Fright (intense fear combined with tonic immobility), followed by a “shut down” or parasympathetic activation with Flag and Faint. Flight, Fight, and Freeze, the three most commonly discussed behavioral variations of the Defense Cascade, are often observable, but can be confused with other behaviors typically seen in children with ASD and ID (Bradley et al., 2014). Fright can be observed as well when a child is in intense fear and appears as though they can’t move. Flag and Faint reactions may also be highly informative when present; however, fainting is a severe, but less common symptom and flag reactions or signs of fatigue may be difficult to associate with anxiety, particularly in children.

More specifically, in terms of the common behavioral responses, “Flight” can appear like eloping, a term used to describe children with ASD who wander or run away from an environment possibly due to seeking out something preferred or avoiding a non-preferred situation/demand. For example, a child who runs out of the classroom due to fear of the Columbia pictures trailer before a movie shown in class, indicating an uncommon phobia. “Fight” appears similar to aggressive behavior towards others or self-injurious behavior and may be interpreted as a difficulty with frustration tolerance, when it could also be a manifestation of anxiety. For example, a child who is anxious in social situations and displays aggression towards the teaching staff when prompted to enter a room full of peers. “Freeze” may appear as if the child is simply not responding to a prompt or request, which occurs regularly in some children with ASD and ID for reasons that are often related to cognitive, language, or executive functioning impairments, but may also be a sign of significant fear. An example is a child who freezes and stares straight ahead or puts their head down when a worksheet is passed out due to worry about their performance. In sum, flight, fight, or freeze behavior might be interpreted as relating to other forms of problem behavior with potential anxiety being overlooked. Therefore, a careful consideration of whether anxiety is the function behind behaviors like eloping, aggression to self or others, and freeze behavior may lead to a more accurate diagnosis and, ultimately, necessary treatment recommendations.

Towards Best Practices for the Assessment of Anxiety in those with ASD and ID

Together, significant mental health needs and obstacles to assessment reveal a need for cohesive clinical considerations for assessing anxiety in youth with ASD and ID that are drawn from the literature, ongoing programs of research, and clinical experience. Recommendations regarding diagnosis of anxiety and co-occurring behavioral conditions, such as ASD, from MacNeil et al. (2009) include advocating for utilizing clinical interviews with various informants and, if possible, obtaining direct observation of patients in differing contexts. Vasa and colleagues (2016) recommend a careful evaluation of physical and behavioral symptoms suggestive of anxiety such as elevated heart rate, gastrointestinal and sleep problems, nail biting, skin picking, flat affect, and increased (or decreased) verbalizations while keeping in mind that these behaviors could also have overlap with ASD predilections such as seeking sensory stimulation. Lastly, the creation of questionnaires with a behavioral focus, that are validated for use specifically in those with ASD, is recommended (MacNeil et al., 2009) and is a growing area of psychometric research. For example, Scahill and colleagues (2019) recently created, validated, and published a 25-item measure focusing purely on behavioral symptoms called the Parent-Rated Anxiety Scale for Youth with Autism Spectrum Disorder (PRAS-ASD). In addition, Mazefsky et al. (2018) have created and validated a short and precise survey titled the Emotion Dysregulation Inventory. Both of these measures have psychometric data available in those with a wide range of functioning and cognitive ability.

Below, we include three case vignettes of traditional and distinct presentations of anxiety in children with ASD and ID, including moderate-severe ID, where participants’ language is very limited. We then provide clinical considerations for assessment of anxiety in individuals with ASD and ID that derive from the literature, our ongoing programs of research, and clinical experience working with these and other children. The overarching goal of the following case illustrations and clinical considerations is to improve the detection of anxiety in youth within this population of children, thereby leading to the provision of necessary mental health services to these individuals. To this end, we conclude the manuscript with a discussion of treatment options.

Overview of Participants in the Research Study

The three cases described below and subsequent clinical considerations are drawn from work in a larger longitudinal research cohort of intellectually diverse children with ASD and typically developing individuals that were first seen at the UC Davis MIND Institute at ages 2-3 ½ years as part of the Autism Phenome Project with follow-up assessments in early childhood (ages 5-8 years) using a comprehensive battery including autism diagnostic, cognitive, language, and neuroimaging measures. The participants were recently followed up again at the 8-12-year middle childhood timepoint with a focus on the study of anxiety. In this phase of the Autism Phenome Project, anxiety was assessed with semi-structured parent interviews (ADIS-P/ASA) in a large sample of children with ASD and a broad range of intellectual functioning (IQ range≤25-170; M=77.39). The ADIS-P/ASA was administered and scored by licensed clinical psychologists, with established research reliability in the ADIS-P/ASA. Diagnosis was confirmed via scores on the semi-structured interview (ADIS-P criteria and clinical severity ratings) and the consensus of two independent experts in the assessment of mental health in children with ASD and ID, to whom the cases were presented. Participants reported a range of socioeconomic and racial and ethnic backgrounds with participants primarily reporting Caucasian and non-Hispanic (over 70%; see Kerns et al., 2020 for full demographic characteristics). This research was approved by the UC Davis Institutional Review Board and all participants gave informed consent or assent.

To summarize, in the overall study sample, participants with ASD and intellectual impairment (defined as IQ<70), specific phobias were diagnosed in 42% and separation anxiety in 6% (Kerns et al., 2020). These were the only traditional DSM-IV anxiety disorders diagnosed in the 36 participants with intellectual impairment. Kerns and others (2020) note that these are the anxiety disorders that present earliest in development (Beesdo et al., 2009), which is consistent with the developmental status of these participants. For distinct anxieties, 19% were identified as having interfering fears of change, 14% with uncommon phobias, 7% with other social fears, and 3% with special interest fears.

An alarming statistic was that, of the 40 overall participants in the study with ASD and ID, only three (7.5%) had been diagnosed with anxiety disorders before the current study evaluation. However, when they were carefully assessed using the clinical interviews ADIS-P/ASA with expert consensus diagnosis, an additional 24 were identified with at least one clinically significant DSM or distinct anxiety problem bringing the total to 27 (67.5%). It is likely that the challenges in diagnosing anxiety in individuals with ASD and ID, including distinct expressions of anxiety, the difficulty of determining if anxiety is present from behavior alone rather than verbal report, the lack of sensitivity in traditional anxiety screening questionnaires, and potential diagnostic overshadowing, among others, played a central role in these rates considering that the majority did not have a prior diagnosis.

Case Selection and Assessments

Case study participants described in the vignettes were selected because they had a range of clinically significant traditional and distinct anxiety manifestations determined from a modified parent interview approach using the ADIS-P/ASA and expert consensus, illustrated some of the more complex diagnostic presentations including unique and distinct areas of anxiety, met criteria for ASD, had reduced language ability or were nonverbal, and cognitive functioning in the range commensurate with moderate-severe intellectual disability. Table 3 includes data regarding the IQ scores, ADOS-2 Calibrated Severity Scores (CSS; Gotham et al., 2008), and anxiety parent questionnaires of the three cases. Through these case illustrations, it is our goal to (a) illustrate approaches for conducting a more developmentally sensitive parent interview for anxiety in children with ASD and ID and (b) shine light on how accurate diagnosis of anxiety is the first major steppingstone to providing appropriate and effective evidence-based treatment for those with anxiety, ASD, and ID. Names of the children in the case vignettes were changed to protect anonymity.

Table 3.

Case Study Descriptives

Case	DAS-II GCA	DAS-II VIQ	DAS-II NVIQ	ADOS-2 CSS	MASC-P Total	SCARED-P Total	CBCL DSM Anxiety
Kylo	39	31	42	6	46	13	53
Everly	47	32	54	8	42	4	50
Amelie	42	<25	64	6	54	14	55

^*DAS-II GCA = DAS-II General Conceptual Ability;

VIQ = Verbal IQ; NVIQ = Nonverbal IQ (Elliot, 2007).

^**ADOS-2 CSS = ADOS-2 Calibrated Severity Score (Gotham et al., 2008).

Cognitive Ability

Participants were administered the Differential Abilities Scale, Second Edition (DAS-II; Elliott, 2007), to measure intellectual functioning level. The General Conceptual Ability (GCA) composite was reported as a measure of full-scale IQ. The Verbal and Nonverbal Ability Composites are reported in Table 3 due to the large discrepancy between those scores in most participants. If a participant could not complete the age-appropriate School Age version of the DAS-II, we used the Early Years version.

Autism Spectrum Disorder

ASD diagnostic status and severity was evaluated with a gold standard autism diagnostic assessment including the Autism Diagnostic Observation Schedule – Second Edition (ADOS-2; Lord et al., 2012), the DSM-5 diagnostic criteria, and expert clinical judgment. Table 3 includes a CSS for the three case examples, which provides a level of ASD severity across the range of functioning for those with ASD (Gotham et al., 2008).

Anxiety

Semi-structured Interviews

The ADIS-P (Albano & Silverman, 1996) is an established, semi-structured interview developed for typically developing children (Lyneham et al., 2007; Silverman et al., 2001). Research-reliable clinical psychologists assigned a clinical severity rating (CSR) for each anxiety disorder assessed. CSRs range from 0 to 8, with a score ≥4 suggesting clinically significant interference and the threshold for diagnosis if the other criteria are met. The ADIS-P modules administered in this study were separation anxiety disorder, social anxiety disorder, generalized anxiety disorder, specific phobia, and obsessive-compulsive disorder. The generalized anxiety disorder module was not included if the child had not ever verbally expressed thoughts or feelings to their parent. Considering that these participants were either non-verbal or had reduced language ability, we adjusted the wording of the prompts if they were verbally mediated by saying, “Has he/she indicated…,” rather than, for example, “Has he/she told you…” We relied heavily on deriving solid behavioral examples of anxiety (Table 2) with a focus on determining whether anticipatory anxiety was indicated by the child in any way. For example, a child looking distressed/anxious, pacing around, and checking many times to see if her mother’s handbag is still in the house in anticipation of her mother leaving for work. In addition, the child’s estimated developmental level was taken into consideration when evaluating whether the extent to which their fears and behavior should be considered excessive or atypical. For example, the inability to complete a sleepover at a friend’s home was not considered problematic for a child with severe developmental delays/the developmental level of a 3-year-old, whereas an inability to stay with a trusted family member or other familiar adult due to high distress in anticipation of separation and lack of reduction of distress following separation was considered problematic.

Table 2.

Behavioral Examples of Anxiety in Children with ASD and ID

Common Responses		More Nuanced Responses
• Avoidance*	• Increased repetitive behaviors pacing/rocking/vocalizations*	• Repetitive asking for comfort items/foods
• Crying*	• Irritability*	• Repetitive asking to see their schedule
• Freezing Behavior*	• Tantrums*	• “Hiding” Behavior (e.g., covering head with blanket/shirt, asking to use stroller)
• Fearful Affect*	• Disruptive Behavior*
• Clinginess*	• Aggression*	• Hair pulling/skin picking
• Sleep Problems*	• Self-injury*	• Physical symptoms (e.g., stomach aches, vomiting, headaches)

^*Examples provided by Vasa et al., 2016.

The ASA (Kerns et al., 2017) includes a number of prompts and guidelines that are added to the ADIS-P to facilitate differential diagnosis between symptoms of anxiety and those of ASD. It also captures “distinct” presentations of anxiety (Table 1) with CSRs provided for each distinct area. Furthermore, the ASA includes considerations for differential diagnosis such as assigning a rating for the child’s functioning in the following areas: friendships, social motivation, bullying, theory of mind, hypersensitivity, and perseverative thinking. These items are coded on a Likert scale of 0 to 3, where 0 representing no impairment and 3 severe impairment in each domain. These scores are used to help the clinician parse out potentially overlapping ASD and anxiety features (e.g., sensitivity v. phobia of loud sounds).

Parent-report Questionnaires

Three parent-report questionnaires were administered to determine whether clinically significant anxiety in these participants was detected by these screening measures. The results of these measures were viewed after anxiety diagnoses were determined through the clinical interview with the ADIS-P/ASA and are reported in Table 3. The CBCL (Achenbach & Ruffle, 2000) is a widely used measure of a range of problematic behaviors. We chose to specifically focus on the DSM Anxiety Scale due to our goal of determining whether the participants met criteria for DSM clinically significant anxiety. Clinically significant elevations are indicated by a T-Score of ≥70. The MASC (March et al., 1997) includes items of harm avoidance, physical symptoms, social, and separation anxiety. A MASC Total score is derived from the 39 items with a T-Score of ≥65 suggesting clinically significant anxiety. The SCARED (Birmaher et al., 1999) includes items of somatic symptoms, school phobia, and generalized, separation, and social anxiety. A total score is derived from the 41 items with a total of ≥25 indicating clinically significant anxiety.

Case Vignettes

Kylo

Kylo was a 12-year-old Caucasian (Non-Hispanic) and Asian male diagnosed with ASD at age 2 with a current verbal IQ of 31 and adaptive skills ranging from those of a four-month-old (socialization) to a four-year-old (daily living skills). At the time of the assessment, he spoke some words and used an iPad for communication training. Clinical interviewing revealed that when Kylo approached social situations, he would look anxious and uncomfortable and would cling to his parents and not leave their side. He did not engage with others in public and would actively avoid people by putting his hand up if people tried to speak to him or by putting on headphones and covering his head with a blanket when he had to be around new people. This was in contrast to his behavior around familiar individuals, illustrating that it was not general avoidance of people. When Kylo was around people, he would talk repetitively to his parents in a pressured way about his favorite food, which was French fries, even if he had just eaten. He appeared comforted by thinking about this preferred food like another child might be comforted by a favorite object, such as a blanket. In addition, this behavior only occurred when he was around other people. When he was attending mainstream classes, he displayed aggression towards teachers, staff, and peers and also engaged in self-injurious behavior. It was difficult to determine the source of his stress in the school situation at first; however, the behavior decreased once he was moved to a specialized classroom with fewer and more similar peers. This behavior was not always present for Kylo; his parents noticed an increase in this distress around people starting about a year prior. The family tried various medications to target this aggressive behavior and only noticed a decrease once they began tetrahydrocannabinol (THC)/cannabidiol (CBD), which has been thought to ameliorate anxiety; however, while CBD has been studied in adults with anxiety (e.g., Crippa et al., 2011), there are no known controlled trials that have specifically evaluated anxiety reduction in children with ASD and ID.

Likely given Kylo’s low level of cognitive functioning, most viewed his behavior as lack of social interest related to ASD or sensory sensitivity, rather than symptoms of anxiety. This case is a good illustration of why it is necessary to consider an anxiety diagnosis. Supportive of the fact that Kylo’s behaviors were not just problem behaviors or symptoms of his autism, he exhibited anticipatory anxiety (with an anxious facial expression and clinginess towards parents upon approaching social situations) and avoidance, two hallmarks of an anxiety disorder when in social situations. Furthermore, Kylo had this response even in quiet social situations when unfamiliar people were close to him, indicating it could not be solely accounted for by a sensory sensitivity to noise.

Furthermore, it is possible that Kylo’s aggression towards others and himself were examples of “fight mode” reaction with underlying anxiety to propel this behavior rather than aggressive behavior due to other reasons, such as lack of preference of social situations or sensory irritation. As is typical in those with ASD and ID, Kylo was not able to verbally indicate any fear of negative evaluation, the primary characteristic of social anxiety disorder. In addition, he scored a 3 on the theory of mind section of the ASA (most severe score indicating very poor to no awareness of the thoughts/opinions of others), suggesting that his awareness of others’ thoughts and opinions was limited, and that people were potentially confusing or unpredictable to him in a manner that made him anxious. Though Kylo’s fears were considered to be both severe and functionally impairing, they fell short of the criteria for social anxiety disorder, given that a fear of negative evaluation (criterion A) could not be established. As such, Kylo’s significant and functionally impairing fears (ADIS/ASA CSR=6) were considered to be most consistent with a diagnosis of Other Specified Anxiety Disorder: “Other Social Fears.”

As shown in Table 3, his CBCL DSM Anxiety, MASC-P and SCARED Totals were all well below the cut-offs for clinically significant anxiety, indicating the low sensitivity of these measures in this population, likely due to their reliance on verbally mediated symptoms, rather than a lack of anxiety.

Everly

Everly was an 11-year-old Caucasian (Non-Hispanic) female with a verbal IQ of 32 and diagnosed with ASD when she was 16 months old. She was speaking at a level that included single words at the present evaluation and had adaptive skills ranging from those of a one-year-old (socialization) to a five-year-old (fine motor skills). Everly and her family had lived in several different homes. After a period of several weeks coming and going without issue, Everly would begin to express high levels of distress in response to driving up to each house, including looking scared, crying, covering her ears, screaming, covering someone’s mouth if they tried to talk to her, and refusing to get out of the car. If they did get her out of the car, she would try to run away. We queried her parents in depth in order to obtain these behavioral examples and to accumulate evidence of anticipatory anxiety/fearful avoidance (e.g., looking scared and avoiding the house) and defense cascade behavior (e.g., attempting to run away aligning with “flight”) to determine whether they were associated with anxiety. To further confirm that their function was related to anxiety and not something simply to avoid activities that were non-preferred, we considered whether she preferred outings but did not want them to end. However, she did not seem distressed upon leaving activities or outings and it was only when she approached the house that she became highly anxious. Also, she would not enter the house calmly even if there were highly preferred objects or activities inside. Everly could not indicate to her family what was causing her great distress but, once it started, it would occur every time they pulled up to their house and the process to get her to enter the house often took over 30 minutes. This caused great interference for the family since they became worried about taking her out of the house for fear of what would happen when they had to come home, and if she might injure herself. The family attempted to determine the cause by walking with her to see what she was focusing on or if she could gesture to indicate the problem; however, this was not successful. Each time the family eventually explained to her that they were going to move (they waited 6 weeks before moving the first time) and took her to look at new houses. She was fine with going into new houses and the family picked a new house based on her display of positive emotion in the house. The positivity would not persist following the move, and eventually the problem behaviors would start again. The family had moved five times by the time of this evaluation at 11 years of age. It is unclear what the trigger was for her high levels of distress. Some untestable possibilities discussed with her parent were whether she saw a spider or shadow in her room that scared her, or if she had an uncomfortable sensory experience that developed into a fear. Since anticipatory anxiety, fearful avoidance, possible fight and flight mode, and high interference were present, but the topic of her anxiety/fear was uncertain, Everly’s symptoms were deemed most consistent with a diagnosis of Other Specified Anxiety Disorder: Fears related to the home (coded as “Other Distinct Anxiety” on the ADIS/ASA; CSR=7). Her CBCL DSM Anxiety, MASC-P and SCARED totals were the lowest scores of all of the participants highlighted here, indicating that the type of anxiety/fear she experienced was not captured by these brief-screening questionnaires. Furthermore, anxiety as a trigger for these symptoms and behaviors had never been mentioned to the family by any professional. So, pharmacological or anxiety cognitive-behavioral/behavioral interventions had not been previously recommended.

Everly also had a fear of people singing in person or on the radio. This was singing at a regular or even low volume rather than extremely loud volume, indicating that it was not just a sensory aversion to loud sounds. This began 2-3 years prior to this evaluation. If her family tried to take her into a store or restaurant that potentially played music, she would start looking distressed in the car before entering the place, indicating anticipatory anxiety. If they were able to get her out of the car, she would then pause before walking through the door and listen. If she heard the music, she would look scared, then scream, and shake her head no. If she wasn’t permitted to avoid going in, she would cover her ears, scream, attempt to run, and hit people to get away if they were in her way. This indicated the pattern of fearful avoidance with “fight” and “flight” mode if she was not able to avoid. This occurred as well if she heard singing on the radio. The family had stopped taking her to any public places that they knew played music due to her high levels of distress that affected her safety at times. It is important to note that the ASA has a section to assess for the level of sensory sensitivity in order to assist with differential diagnosis. Everly scored a 1 on the sensory sensitivity section (score of 3 indicates the most interference). She was not sensitive to most loud sounds or other sensory experiences other than some taste sensitivity with food, so it is just these specific sounds that she experienced with such distress. It is possible that she could be sensitive to the sound of singing regardless of volume. However, by obtaining clear behavioral examples, there was an indication of not only an aversion to singing, but also of anticipatory anxiety (e.g., checking for music in stores before going in), avoidance, and fight and flight behavior. Considering idiosyncratic or uncommon areas of anxiety were relevant with Everly since she was not scared of all loud sounds but sounds that were much more specific. Therefore, these symptoms and behavior were deemed most consistent with a diagnosis of Specific Phobia: Other Type (coded as “Uncommon Phobia” on the ADIS/ASA; CSR=6).

Amelie

Amelie was a 12-year-old Caucasian (Hispanic) female with a verbal IQ of <25 on the DAS II Early Years Protocol (Elliott, 2007) and adaptive skills ranging from those of a four-month-old (socialization) to a three-year-old (daily living skills). She was diagnosed with ASD when she was 2 years old. She had previously been diagnosed with cerebral palsy, spastic hemiplegia, and tethered cord (neuromigrational). As of the current evaluation, she had also been diagnosed with cyclic vomiting syndrome, abdominal migraines, seizures, and tic disorder. She did not display spoken language and used an Augmentative and Alternative Communication (AAC) program on her iPad. Amelie appeared anxious and fearful in situations involving separation from her family members (mom, dad, and brother). If she thought her mother was leaving (e.g., noticed her mom putting on her shoes and getting her purse), she went to her mother, held her hand, cried, stomped her feet, and asked for “mommy” on her communication device. This all occurred before her family member actually left and there were clear behavioral examples indicating anticipatory anxiety. At home, she followed her mother around and sat at the dining room table where she could see most areas of the house and looked for her family members. If a family member left, she would become distressed, start sweating, cry, bite her hand, wring her hands, and ask for them repetitively on her communication device. Amelie’s behavior of biting her hand is a sign of self-injurious behavior and can be considered to be a part of “fight” mode, although it is to herself rather than someone else. This could last up to an hour or until someone could redirect her to an activity of interest. Amelie previously slept with her mother every night but transitioned to sleeping with her service dog once she joined the family. In addition, Amelie regularly (ranging from several times throughout the day to continuously for at least an hour or more) checked that all of her family members were there by doing a head count and also looked to see that her mother’s car was in the garage, and her mother’s purse was in the house. These seemed to be ways Amelie attempted to cope with her anxiety, by looking for reassurance that her family was home, rather than compulsive checking related to obsessive-compulsive disorder. This information was derived from exploring how Amelie was currently coping as it can give clues into the anxiety a child is experiencing. If the entire family of four would go to the grocery store, Amelie would want everyone to stay together in the store. If they were in line and they forgot something, she became distressed if one person tried to leave to go retrieve it so the entire family would get out of line to go. Amelie’s mother reported that the family had eventually organized their whole life around this area of anxiety and worked to run all of her errands or do anything she needed to do while Amelie was in school. When Amelie was around other children, she wanted her mother to stay by her side and do everything the kids were doing. Amelie had a history of vomiting and stomachaches resulting in medical diagnoses of cyclic vomiting syndrome and abdominal headaches. Considering that these exact physical symptoms are in the DSM-5 diagnostic criteria for separation anxiety, it is possible that these symptoms were directly related to her anxiety and her mother wondered this as well. This was possible since Amelie appeared to be worrying about separation often throughout the day, even if separation was not imminent; however, also difficult to parse out since her anxiety did not depend on specific triggers for separation to be present. Therefore, these symptoms and behavior that significantly interfered with Amelie’s life were deemed most consistent with Separation Anxiety Disorder (ADIS/ASA CSR=7). Amelie’s CBCL DSM Anxiety, MASC-P and SCARED Totals were, again, all within the average range. A psychiatrist had recently informed Amelie’s family that she had anxiety and Zoloft had been prescribed.

Amelie also had a fear of garland or other overhead hanging decorations. The family attempted to take Amelie to go see the holiday decorations in their town. When she saw the garland, she attempted to run away and appeared scared, indicating “flight” mode and fearful avoidance. They could not convince her to go towards the decorations and her fear remained sufficiently intense that they eventually left without participating in the event. Her immediate and extended family refrained from decorating for the holidays in this way due to this fear and related interference. When Amelie’s mother decorated the playroom in their house with hanging decorations, outside of the holiday season, she refused to enter. She invariably looked away, showed a fearful facial expression, trembled, and pointed to the items as if to request to remove them. Her mother moved the hanging decorations to her brother’s room, at which point she refused to enter until they were removed. Amelie’s mother was asked specific questions to obtain the behavioral examples to confirm the presence of anticipatory anxiety, fearful avoidance, flight behavior, and an idiosyncratic fear of decorations hanging overhead. Therefore, these symptoms and behaviors were deemed most consistent with Specific Phobia: Other Type (coded as “Uncommon Phobia” on the ADIS/ASA; CSR=4).

Clinical Considerations for Assessing Anxiety in ASD and ID

1). Keep anxiety on the radar.

Always carefully consider whether a child with ASD and ID might have a significant anxiety problem that has gone undetected. Some behavioral manifestations of anxiety in this population can be easily confused with autism symptoms or problem behaviors.

2). Behavioral examples.

Use semi-structured interviews (as recommended by MacNeil et al., 2009), particularly those that are designed and validated for use in children with ASD (e.g., ADIS-P/ASA) as a guide with a central focus on obtaining behavioral examples, as suggested by the DM-ID-2. Also, while the DM-ID-2 specifies focusing more on behavioral observations of anxiety for children with severe or profound ID, we also applied this approach to all children with autism, with or without ID, since some children do not express their emotions verbally even if their overall language and cognitive level is sufficient to do so. Table 2 includes a list of behavioral examples (Vasa et al., 2016) with a potential underlying function of anxiety that can be queried as part of the interview process. It may be necessary to adjust the wording of interview probes to make them more appropriate for the child’s language level (e.g., “Has he/she indicated…,” rather than, for example, “Has he/she told you…”).

3). The Defense Cascade or 6Fs.

While the 6Fs are not specifically outlined in the DSM-5, the behaviors are mentioned in the DM-ID-2 (e.g., freezing). They are also inherent in clinically significant anxiety and fear and can often be observed directly. Notably, the behaviors in the case vignettes of aggression, self-injurious behavior, eloping, and freeze behavior do not always indicate anxiety; however, identifying when these symptoms co-occur with hallmarks of anxiety disorders (e.g., avoidance, and anticipatory anxiety) may lead to more accurate examination of anxiety disorders.

4). Physical symptoms.

As recommended by Vasa et al. (2016), consider gastrointestinal symptoms such as vomiting and stomachaches, as well as headaches, as these symptoms are often associated with anxiety. Physical symptoms of anxiety can be difficult to determine if a child cannot express them verbally. However, if there is evidence that pain or other somatic patterns are present, considering anxiety as the culprit is encouraged. Physical symptoms may prompt extensive medical workups that can be highly distressing to children with ASD and ID. Some examples we have observed, with a function eventually determined to be anxiety, include feeding tube placement due to nausea, vomiting, and reduced desire to eat, and asthma diagnosis and extensive medicine due to difficulty breathing. Therefore, while a medical workup should always occur first if physical symptoms are present, we encourage a consideration of anxiety as well if no clear medical cause is found.

5). Anticipatory anxiety/fearful avoidance.

Differentiate anticipatory anxiety/fearful avoidance from behaviors not associated with anxiety/fear (e.g., core ASD symptoms, non-preferred tasks/situations, or distress with a different underlying emotion such as frustration, anger, or sadness). Anticipatory anxiety occurs when an individual experiences increased anxiety about what they think might happen in the future (e.g., my parent might leave and not come back). Fearful avoidance is when one avoids situations/experiences that could lead to the fear (e.g., avoid separating from parent). These are two hallmarks of an anxiety disorder and can be difficult to determine if a child is not able to verbally describe these symptoms. To evaluate this, the clinician might ask the parent to describe how they can tell when the child is afraid v. angry or afraid v. disinterested and also to describe why they think the child is afraid in the particular scenario being discussed. In this way, the clinician can combine their expertise about varied signs and behavioral patterns that may reflect anxiety with the parent’s expertise regarding the unique way their child expresses their emotions. Asking the parent to differentiate emotions in this way may be critical to ensure that not all moments of distress or challenging behavior are reframed as anxiety. In reality, a child may engage in repetitive behaviors both when they are excited and upset. Anticipatory anxiety or fearful avoidance is a key component to determining whether clinically significant anxiety is present. In conducting these interviews, parents often report, “My child does not like that,” when asked about whether they are fearful of something or exhibit anxiety. It is the role of the interviewer to ask further questions to determine if there is anything that is evident in a fearful or anticipatory way.

6). Idiosyncratic coping/repetitive behaviors.

Consider whether the child does anything idiosyncratic to attempt to cope with anxiety (e.g., repetitively ask for favorite food item, check schedule often to see when parent is returning). These behaviors may serve as coping mechanisms, and in many cases, anxiety may be identified by observing how the child attempts to cope when anxious. Further, repetitive behaviors sometimes increase when the child is anxious (e.g., pacing) and other children talk more frequently about their special interest when anxious. More specifically, Spiker et al. (2011) found that symbolic enactment of restricted interests (RI; i.e., a child repetitively enacting or mimicking characters or scenes related to a RI) was significantly associated with increased anxiety.

7). Distinct anxiety.

Using the ASA (Kerns et al., 2017) as a guide, consider whether distinct areas of anxiety are present (e.g., uncommon phobia, fears of social interaction despite limited fears of negative evaluation). Since the specialized training on the ASA might not be possible for all clinicians, even becoming familiar with distinct areas of anxiety can help detect anxiety that might otherwise go overlooked.

8). Clinician training.

We recognize that some of these considerations, such as full semi-structured interviews, might not be as feasible in a fast-paced clinical setting. It is still possible to obtain knowledge through training that would lead to more accurate detection of anxiety. It is suggested that clinicians responsible for the mental health in those with ASD/ID have in-depth knowledge of the differentiation between ASD and anxiety including where there is symptom overlap; have awareness of how to obtain clear behavioral examples of anticipatory anxiety and fearful avoidance, especially in those with Severe ID, as suggested by the DM-ID-2; consider how the Defense Cascade manifests behaviorally and ways to ask about these symptoms; consider screening questionnaires carefully due to the reduced sensitivity in this population; and consider whether distinct areas of anxiety are present and need clinical attention.

Additional strategies suggested for a multi-method assessment

Other methods/approaches ideally integrated into a multi-method assessment include using newer anxiety parent questionnaires that have been specifically designed for individuals with ASD and have prompts that are more behaviorally oriented than verbally mediated (e.g., PRAS-ASD; Scahill et al., 2019); including self-report when possible either through interview or drawing; and direct observation of the child, especially in contexts that might induce anxiety (e.g., while greeting unknown people). Moskowitz, Rosen, et al. (2017) discussed using the Behavioral Avoidance Test (BAT; Hagopian & Jennett, 2008), as one type of direct observation, to systematically expose the child to the feared object/situation while assessing the child’s various displays of anxiety and avoidance. The PRAS-ASD was not yet available for the present study so results of the parent-report screening questionnaires (MASC, CBCL, SCARED) used were discussed.

Assessment as a Means to Effective Treatment

Accurate assessment of anxiety disorders ultimately has one clear aim: to guide appropriate treatment. The evidence base for cognitive-behavioral therapy (CBT) for youth with anxiety disorders is both long-standing and robust (Higa-McMillan et al., 2015). CBT for youth with anxiety targets impairing symptoms through several key components, including psychoeducation, cognitive restructuring, and exposure (Gosch et al., 2006). Exposure consists of interrupting the fear-relief cycle that occurs when youth avoid the stimuli or situations (real or imagined) of which they are afraid. CBT, fortunately, is not reserved for typically developing populations and CBT for youth with autism and anxiety is supported by a growing body of research (Kester & Lucyshyn, 2018). To date, less research has focused on CBT for youth with ID though evidence shows promising results with appropriate modifications (Blakeley-Smith et al., 2021; Hronis et al., 2017). Recent studies have investigated treatments for anxiety in ASD and ID that bear resemblance to CBT, specifically an array of behavioral interventions, which de-emphasize the cognitive component of CBT and adapt the intervention(s) to the child’s functioning level (Moskowitz, Walsh, et al., 2017; Rosen et al., 2016).

The case participants presented here did not receive treatment as part of this study, yet exposure-based strategies could clearly be applied to target their anxious symptoms. For Amelie, prescribed exposures would likely consist of gradually spending increased amounts of time without her family immediately nearby (e.g., across the room, in an adjacent room, upstairs, and so on), consistent with traditional separation anxiety exposures. Her fear of garland, which is an uncommon fear, would similarly be targeted via exposures to the feared stimuli (e.g., pictures of garland, garland being inside a bag in the room, touching garland for one second and so on). Importantly, exposure could decrease avoidant behavior, not simply as a form of compliance and habituation, but also a sense of mastery and reduction of the fear response. Indeed, reaching this aim may require particular modifications to achieve treatment goals, such as including concrete visual materials, idiosyncratic rating scales (e.g., feelings thermometer), tangible reinforcement programs, and enhanced parental involvement as compared to traditional CBT (Moree & Davis, 2010; Walters et al., 2016). In virtually all cases the integration of children’s special interests is key to building motivation and engagement. Lastly, special attention must be paid to the co-occurring attention, language, and memory deficits that may present in youth with ID (Hronis et al., 2017). Together, a high level of flexibility, creativity, and individual tailoring is necessitated.

For instance, Everly might benefit from visual materials that illustrate her fear and avoidance in a concrete manner, such as cartoons showing her tantrumming about music on the radio and her bravely listening to a song. Her mantras may consist of one word, such as “Okay!” or “Listen!” or even a preferred character (e.g., “Mickey!”) while participating in initial music exposures. Similarly, Kylo’s presentation would warrant rapid rewards, such as watching a Star Wars video intermittently during exposures. Kylo may also benefit from inclusion of comfort items during initial exposures (e.g., Yoda stuffed animal, blanket), which could be faded out gradually over time. Further, he may engage most from modeling in the form of showing two Star Wars characters talking. Such forms of treatment may significantly and positively impact these children’s lives, but identification of anxiety must occur in order for appropriate mental health referrals to be made.

Limitations

The data presented here was limited to case studies of youth with anxiety, ASD, and ID. They are uniquely useful for illustrating detailed information regarding the idiosyncratic presentation of anxiety; however, they may not be representative of anxiety in those with ASD and ID overall. In addition, our parent-report questionnaires of anxiety were not sensitive enough to detect anxiety in any of the three cases. While this highlights the discrepancy between the utility of brief screening questionnaires versus semi-structured interviews in this population, newer measures have recently been developed to help address this concern, such as the PRAS-ASD (Scahill et al., 2019). Furthermore, while we provided clinical considerations related to a thorough parent-interview process, a multi-method approach is recommended for clinical determinations.

Conclusion

An imperative future direction for our field, both within neurodevelopmental and mental health specialties, is to enhance clinicians’ ability to assess for anxiety amongst youth with ASD and ID. We have outlined a parent interviewing approach with case vignettes and clinical considerations with a focus on behavioral examples, anticipatory anxiety, and fearful avoidance, to more accurately detect anxiety that might otherwise be easily overlooked in this complex population. To this end, more training on the topics presented is needed at both the graduate training program and professional level. These children experience anxiety that can be treated but first it must be identified. Therefore, thoughtful assessment of potential anxiety is paramount to enhancing the mental health of this population of children and their families.