Keywords

Anxiety; ASD; autism; phenomenology; presentation; symptomatology; typical/traditional; atypical/ASD-related

Methodologies Employed in Studying the Phenomenology and Presentation of Anxiety in ASD

Quantitative studies have attempted to disentangle “traditional” anxiety (i.e., shared and commonly present in clinically anxious individuals without ASD) from more ASD-distinct anxiety, using standardized measures (i.e., Renno and Wood, 2013; White et al., 2012). A limitation of this method, however, is that the measures employed were developed for individuals without ASD and as such may be likely more useful in establishing the “shared” rather than the “distinct/ASD-related” anxiety manifestations.

To address this, a small number of quantitative studies have adapted existing measures for use specifically with participants with ASD in order to broaden the scope and range of symptomatology explored (e.g., Kerns et al., 2014; Mayes et al., 2012, 2013; Rodgers et al., 2016). Other studies have used qualitative “bottom-up” studies, in which the participants discuss open-ended questions and thus are not restricted to a potentially unrepresentative range of anxiety experiences. The themes from such analyses can then be thematically organized and established as shared or more distinct to ASD, based on existing theoretical and empirical understanding of anxiety in typically developing young people (i.e., Ollendick and Benoit, 2012).

Findings From Qualitative Studies

Ozsivadjian et al. (2012) reported on a series of five focus groups involving 17 caregivers of 7- to 18-year old cognitively able children and young people with ASD and significant anxiety concerns in the UK. Caregivers discussed triggers, settings, or situations that precipitated anxiety in their children with ASD, and their observations of their children’s somatic/physiological, cognitive, behavioral, or other anxiety presentations. Stressors and triggers identified included “traditional” sources of anxiety commonly identified in individuals without ASD (i.e., worries about social situations and being evaluated/judged; worries about not being able to meet high demands and expectations); as well as more atypical, ASD-related triggers (i.e., disruptions to routine and change; confusion about social situations; over-stimulating sensory stimuli; being prevented from engaging in preferred repetitive behaviors or circumscribed interests). Similarly, a pattern of both “traditional” and more ASD-related features of anxiety presentation emerged: traditional physiological manifestations of anxiety were consistently reported (i.e., increases in arousal and physical sensations typically associated with anxiety in individuals without ASD), while behavioral manifestations of anxiety included both “typical” anxiety-related behaviors (i.e., escape, avoidance, reassurance, safety behaviors), as well as more ASD-related presentations including increases in sensory, repetitive, and ritualistic behaviors or increases in socially inappropriate behaviors (e.g., giggling when anxious). Ozsivadjian et al.’s (2012) findings have also been largely replicated in a culturally and ethnically diverse sample of children and young people with ASD from special schools in Singapore with a range of intellectual and verbal abilities (Magiati et al., 2016).

Also of relevance, Trembath and colleagues (2012) extended these qualitative findings in Australian young adults (18–35 years old) with ASD and their caregivers. Again, both shared and ASD-related sources of anxiety were identified and these were very similar to those reported by Ozsivadjian et al. (2012) and Magiati et al. (2016) in children and adolescents. More developmentally relevant to young adulthood anxiety precipitants were also identified, including “normative” precipitants (i.e., leaving school, managing finances, public speaking, making important future decisions, news reports, and meeting deadlines) and more ASD-related triggers (i.e., anxiety about explaining their diagnosis or understanding complex social etiquette). Anxiety manifesting itself through increases in challenging and repetitive behaviors was also identified in adults, indicating that these behaviors do not necessarily lessen with age.

Finally, Bearss and colleagues (2016) thematically analyzed focus group discussions to guide the generation of ASD-specific candidate items for the development of an “ASD-friendly” caregiver-reported measure of anxiety symptoms. Forty-five US caregivers of 3–17 year old children with ASD and at least some mild anxiety discussed what anxiety looks like in their children and what situations bring about anxiety. Similar themes and subthemes to those identified in the earlier studies were identified. “Traditional” anxiety triggers identified included separation, crowds, negative (mis)interpretation of events, high academic demands, unwanted social attention, and being teased. More ASD-related triggers were unexpected changes or transitions, others’ failure to follow rules and stick to schedules, and sensory stimuli (i.e., toilets, vacuum cleaners).

In summary, four qualitative studies carried out in the UK, USA, Singapore, and Australia involving caregivers, young adults with ASD, and specialist school teachers have yielded consistent and remarkably similar findings strongly pointing towards evidence for shared and distinct ASD-related features of anxiety presentation in ASD.

Findings From Quantitative Studies

Table 3.1 summarizes key findings from both qualitative and quantitative studies with regards to shared and ASD-related presentations of anxiety.

In a study designed to address more systematically the specific question of differential anxiety presentation and diagnosis in ASD to date, Kerns et al. (2014) assessed “traditional” and distinct ASD anxiety presentations in 59 children and young people with ASD using self and parent reports and the clinician-rated Anxiety Disorders Interview Schedule Child and Parent version (ADIS-IV-C/P; Silverman and Albano, 1996). The interview was expanded with additional prompts and sections to capture anxiety symptoms that did not meet “traditional” Diagnostic and Statistical Manual (DSM) criteria, but which were distressing or interfering with the children’s development or functioning (as well as more traditional DSM-consistent anxiety disorders). Following findings from earlier studies (i.e., Leyfer et al., 2006), additional items/sections included interfering worries about change or routine disruption, worries relating to ASD preoccupations and interests, social anxiety without a clearly indicated fear of social evaluation, and unusual specific phobias not captured in traditional measures. Sixty-three percent of the participants obtained adapted ADIS-IV-C/P scores in the clinical range, of whom 31% presented with both traditional and ASD-related anxiety symptoms, 17% with traditional DSM-oriented anxiety disorders only, and 15% with impairing ASD-specific anxiety symptoms only. The four most commonly identified atypical clinically significant anxiety difficulties related to anxiety over (1) routine disruption and change; (2) social fears (without clearly articulated social rejection or evaluation concerns); (3) compulsive or ritualistic behaviors not being completed “appropriately” (without a clear indication of wishing to prevent distress or a feared outcome); and (4) unusual specific fears. Results from this study informed the development of the Autism Spectrum Addendum to the ADIS-IV-C/P, the ADIS/ASA (Kerns et al., in press), which provides a systematic approach to differentiating anxiety and ASD symptoms and specifically queries for distinct manifestations of anxiety in ASD, including fears about social etiquette and predictability, fears of change, fears related to unique sensory experiences, and fears related to circumscribed or special interests. A recent study of the ADIS/ASA in a sample of youth with ASD seeking treatment for anxiety supported the validity and reliability of this tool in identifying both traditional and distinct manifestations of anxiety in ASD (see Kerns et al., in press).

White and colleagues’ (2015) work examined the factorial equivalence of the Multidimensional Anxiety Scale for Children child and parent report (MASC; March et al., 1997) in 465 children with anxiety problems with and without ASD drawn from anxiety intervention trial studies. Although the MASC anxiety subscales had a similar latent structure in anxious youth with and without ASD at a broad level, the underlying factor structure and the relationships between factors were both similar to and different in ASD as compared to typically developing children without ASD. Four factors were identified, of which three had similar item groupings to the original MASC structure. However, the social anxiety factor items separated into two factors in the ASD group, one concerned with humiliation/rejection and the second with performance anxiety, while the MASC harm avoidance subscale disappeared altogether. Some items on the parent-reported MASC loaded onto different factors compared to the original MASC factor structure. White and colleagues (2015) speculated that this may be because although the same factors emerge, pointing towards shared experiences and presentations of anxiety, the items and factors do not relate to each other in the same way as in non-ASD children, likely because of atypical experiences and presentations of anxiety in ASD. Their findings provide further support for the view that youth with ASD also experience and express anxiety in ways that are different from those without ASD and not captured in “traditional” measures and existing factor structures. Two case vignettes (names changed to protect confidentiality) are presented at the end of this chapter to further illustrate traditional and ASD-related presentations of anxiety in youth on the spectrum.

Shared and Distinct Anxiety Presentations in ASD by Anxiety Type

Specific Phobias/Fears

A meta-analysis of 31 studies involving over 2000 children and young people with ASD by van Steensel et al. (2011) reported that specific phobias were the most commonly reported in 30% of the total sample. All four qualitative studies summarized earlier reported that caregivers spontaneously discussed fears commonly present in individuals without ASD (i.e., relating to germs, spiders, animals), a finding also reflected in studies using quantitative methods (i.e., Evans et al., 2005; Mayes et al., 2013; Magiati et al., 2016; Turner and Romanczyk, 2012). Using a semi-structured diagnostic interview, Kerns et al. (2014) also found that 30% of young people with ASD in their study reported common phobias, such as fears of dogs or spiders.

However, research also suggests variation in the quality and focus of phobias in ASD. Evans and colleagues (2005) found that children with ASD had more fears of specific situations (such as the school bus) and medical situations, but significantly fewer fears of harm/injury compared to developmentally age-matched children, children with Down’s syndrome, and chronologically age-matched children. Leyfer et al. (2006) also reported that many specific fears more commonly reported in normative samples (i.e., fear of flying, tunnels, bridges) were rarely endorsed in their sample of children in ASD, whereas relatively rare phobias of loud noises were more common. Mayes et al. (2013) also reported a high prevalence of unusual/more ASD-idiosyncratic fears in children with ASD (i.e., fears of toilets, vacuum cleaners, and other mechanical things) as well as other fears which appeared unusual in their intensity, obsessiveness, irrationality, or quality (such as fears of walking up stairs, open doors, an irrational fear of dying through bone breaking into chest, and so on). Kerns et al. (2014) also found 12% of their sample reported unusual specific fears, such as the happy birthday song, bubbles, super-markets, or running water.

Social Anxiety

Social anxiety is also a commonly reported anxiety disorder in ASD across studies (16.6% in the van Steensel et al., 2011 meta-analysis). There is clear evidence that a number of individuals with ASD, particularly more cognitively able and socially motivated individuals with ASD, do experience social anxiety as “traditionally” defined to involve heightened arousal and worry in, as well as avoidance of, social situations driven by social humiliation or performance fears (i.e., Bellini, 2004). Their motivation for interpersonal relationships may over time lead to increased social anxiety and avoidance of social situations for fear of repeated social failure, being ridiculed, misunderstood, or rejected (White et al., 2012; see also Kuusikko et al., 2008; Maddox and White, 2015).

At the same time, a key difference in the presentation of social anxiety in ASD may be that social avoidance may present in ASD without an accompanying fear of negative evaluation or worry about social performance, with more ASD-related fears possibly being driven by other social worries relating to ASD impairments, such as not understanding social “rules” and etiquette resulting in social confusion, or fears of causing offence by being too direct or honest (Gillot et al., 2001). In the Kerns et al. (2014) study, 8.5% of the sample reported social fearfulness without any explicit awareness of social judgment or negative perceived social evaluation.

Adopting a dimensional perspective to the study of ASD and anxiety symptomatology, White et al. (2012) examined the structure and construct overlap of 24 items from the Autism Quotient (AQ; Baron-Cohen et al., 2001) and the Social Phobia subscale of the Social Phobia and Anxiety Inventory (SPAI-23; Roberson-Nay et al., 2007) in 623 young adult university students without ASD (of whom 2% scored above ASD cut-off on the AQ). They found two separate, but correlated, factors: one relating to “traditional” social anxiety (i.e., anxiety about speaking in front of others, initiating conversation, entering social situations); and another relating to more ASD-distinct social difficulties (i.e., finding social situations hard, not being good at meeting new people) which included items describing preference for less social activities. Thus, both shared and distinct ASD factors were identified (for more on this, see Tyson and Cruess, 2012) which the authors posited as providing additional support for social anxiety being a “true comorbidity” in ASD, as opposed to being entirely attributable to diagnostic overlap.

Separation Anxiety

Existing literature so far suggests that the presentation of separation anxiety in children and young people with ASD is mostly similar to that of children without ASD. In factor analytic studies of existing anxiety measures developed for typically developing children but used with children with ASD, separation anxiety arises consistently as an internally reliable factor in exploratory or confirmatory derived factor structures (i.e., Evans et al., 2005, using a 69-item survey of fears and phobias developed for their study; Hallett et al., 2013a, using the parent-rated Childhood Anxiety Sensitivity Index scale; Magiati et al., 2016, using the Spence Children’s Anxiety Scale-Parent; White et al., 2015, using the MASC-Parent report), and caregivers, young people, and adults with ASD often endorse the “traditional” separation anxiety symptoms in existing scales (i.e., Gillott and Standen, 2007, using the SCAS; Hallett et al., 2013a,b, using the Revised-Child Anxiety and Depression Scale child and parent report). No qualitative or quantitative studies to date have identified any atypical separation anxiety presentations. However, caregivers and clinicians often anecdotally describe intense anxiety in children with ASD associated with unusual attachments to objects (such as bottle caps, elastic bands, rocks, or pieces of string). It is also plausible that separation anxiety may be driven by similar, ASD-related worries as those observed in social anxiety, such as “I won’t know what to say, so I need my mum/dad’” as well as more typical fears of something bad happening to themselves or a caregiver (see Vignette 1).

Generalized Anxiety

Generalized anxiety rates in ASD vary from 13.4% (Simonoff et al., 2008) to 25% in clinically anxious young people with ASD (Ung et al., 2013), and 15% in the van Steensel et al. (2011) meta-analysis. Items relating to general or excessive every day worries about school, family, the weather, health, finances, or other general worries are often endorsed by caregivers or self-report (i.e., Blakeley-Smith et al., 2012; Hallett et al., 2013a,b; Mazefsky et al., 2011; Storch et al., 2012; see Vignette 1). A generalized anxiety factor has also been identified in one factor analytic study (Evans et al., 2005; Hallett et al., 2013a,b; White et al., 2015 used measures which did not include “traditional” generalized anxiety items), although in preliminary findings from a large international pooled dataset of more than 700 6–18-year old children with ASD from the US, UK, and Singapore, Magiati et al. (under review) found that there was no “clear” generalized anxiety factor in this population, but rather a mixed social/generalized anxiety factor. These findings point towards considerable overlap in the presentation of generalized anxiety in children with and without ASD, with perhaps a more blended presentation of mixed symptomatology in ASD, which will need to be further examined in larger factor analytic studies.

Panic/Agoraphobia

In contrast to most other anxiety subtypes, where higher rates of presenting symptoms have consistently been reported in ASD, panic/agoraphobia symptoms were found to be the lowest in van Steensel et al.’s (2011) meta-analysis at 1.8%, on par with prevalence rates of panic disorder of 1–4% in neurotypical adults (Kessler et al., 2006). There is currently no evidence to suggest distinct ASD presentations of panic disorder and associated agoraphobia. It is possible that the lower rates may be due to communication difficulties verbalizing internal physiological experiences often associated with panic attacks. Sukhodolsky and colleagues (2008) found higher rates of panic disorder symptomatology in children and young people with ASD and higher IQ as compared to those with IQ < 70, which suggests that to some extent these internal experiences or the cognitions typically associated with panic disorder (i.e., the misinterpretation of a physiological experience as a sign that something dreadful will happen) either are not experienced by individuals with ASD and intellectual disabilities or cannot be verbally expressed to the same extent as other types of anxiety. Hofvander et al. (2009) also found higher rates of panic disorder in cognitively able adults with ASD (11%), suggesting that perhaps intact intellectual abilities/developmental maturity may be a prerequisite for the development of traditionally-defined panic disorder. Individuals with ASD have been reported to often present with physiological over-arousal seen in panic disorder, but without associated threatening misinterpretations (i.e., Hallett et al., 2013a,b; Storch et al., 2012).

OCD

Although Obsessive Compulsive Disorder (OCD) is no longer included in the DSM-5 anxiety disorders, it has been examined in most studies of anxiety in ASD prior to the publication of the DSM-5 in 2013 and is included as a subscale of symptoms in many existing DSM-IV informed anxiety measures. A number of studies using existing anxiety checklists or clinical interviews designed for typically developing individuals have shown that rates of endorsement of OCD symptoms or meeting criteria for a diagnosis of OCD are much higher in individuals with ASD compared to those without (e.g., 37.2%, Leyfer et al., 2006; see reviews by White et al., 2009; van Steensel et al., 2011). However, in many of these studies there were no systematic efforts to tease out ASD-related repetitive or compulsive behaviors from OCD-related obsessions and compulsions, nor efforts to examine whether the individuals felt compelled to perform certain rituals to reduce anxiety or to prevent a dreaded outcome. In other studies, where specific guidelines for distinguishing OCD and ASD were applied, lower rates have been reported (i.e., 8.2% in Simonoff et al., 2008). Some studies have reported differences in the presentation of obsessive compulsive behaviors in ASD and/or different rates or types of obsession endorsement in children with ASD as compared to those with OCD without ASD (i.e., Cath et al., 2008; McDougle et al., 1995; Ruta et al., 2010). In adults with ASD, McDougle and colleagues (1995) found that repeating, touching, and hoarding were endorsed more frequently, while cleaning, checking, counting obsessive thoughts, and compulsions were much less frequently endorsed compared to adults with OCD only, although there were differences in the two groups in terms of their intellectual functioning. Others have reported no differences (Russell et al., 2005). When considering both traditional and more ASD-related presentations, Kerns et al. (2014) reported ASD-specific compulsive and ritualistic behaviors in 8.5% of children with ASD in their sample, characterized by similar behaviors to those seen in OCD but in the absence of a clear desire to prevent a feared outcome (e.g., having mealtime or bedtime rituals or insisting that doors are closed or sleeves rolled up without articulating a feared outcome to be avoided through these rituals).

In a recent study with adults with ASD + OCD, ASD alone, OCD alone, and without OCD or ASD, individuals with diagnoses of both ASD and OCD self-reported more OCD symptoms than those with ASD alone, suggesting that despite some potential overlap in ASD and OCD symptom presentation and endorsement which may be due to challenges in disentangling the two, OCD behaviors also manifest distinctly from ASD repetitive behaviors (Cadman et al., 2015). Individuals with ASD + OCD were not significantly different to their comparison group of individuals with OCD only in checking, washing, neutralizing, or obsessing, while the 6-factor structure of the Obsessive Compulsive Inventory-Revised (Foa et al., 2002) was confirmed, pointing towards similar presentation of OCD symptoms in individuals with and without ASD. Again, the evidence from the Kerns et al. (2014) and Cadman et al. (2015) in particular point towards both shared and distinct OCD presentations in ASD, with differences in the types of obsessions and ritualistic behaviors and the presence of a clear desire to prevent a dreaded outcome.

PTSD

Similarly to OCD, Post-Traumatic Stress Disorder (PTSD) is part of Trauma and Stressor-related disorders in DSM-5. There has been relatively less focus on examining traumatic experiences and associated PTSD symptomatology in ASD (see Kerns et al., 2015b). To our knowledge, only one study has to date examined the rates and presentation of PTSD symptoms in children and young people with ASD (Mehtar and Mukaddes, 2011). They reported that 12 out of the 69 children with ASD and varying levels of average to severely impaired intellectual abilities in their sample met criteria for PTSD following a semi-structured clinical interview. They also reported a presentation of PTSD symptomatology similar to that traditionally observed in typically developing children (i.e., sleep disturbances, agitation) and children with developmental disabilities (i.e., manifesting PTSD-related distress through deterioration in social behaviors, increases in stereotyped, or challenging behaviors).

Anxiety/Fears/Worries Associated With Core ASD Symptomatology

Anxiety Around Change or Disruption of Routine/Predictability

One “hallmark” feature of ASD is insistence on sameness and related to this insistence, worries about changes in routines, rules, and novelty. Examples of related worries reported in qualitative studies include difficulty with transitions, such as getting changed for Physical Education (PE) classes or having a substitute teacher, and non-routine school events, such as sports days (i.e., Ozsivadjian et al., 2012). In Kerns et al. study (2014), 22% of the sample reported fears of change and novelty, including changing or taking new traveling routes, or changes in daily schedules (see Vignette 2).

Gotham and colleagues (2013) reported a modest significant positive association of r = .27 between anxiety and insistence on sameness, but also found that despite the relationship the two constructs were largely distinct. An interesting body of research has further examined the relationship between anxiety and Intolerance of Uncertainty (IU) in ASD. IU is considered to be a key feature of Generalized Anxiety Disorder (GAD; Dugas et al., 1998) and has been implicated in a number of other anxiety disorders in clinically anxious individuals without ASD (e.g., Carleton et al., 2012). Boulter et al. (2014) noted the resonance of this trans-diagnostic feature with some of the core characteristics of ASD and found that IU was not only significantly related to anxiety severity in a group of young people with ASD, but also accounted for increased levels of anxiety in this group. Thus, IU may be an important driving feature of anxiety functioning similarly in children with and without ASD.

Anxiety Relating to Sensory Oversensitivity

Sensory over- and under-responsivity are known to be common in young people with ASD, yet the relationship between sensory sensitivity and anxiety has received little attention. A number of the atypical fears reported in Kerns et al. (2014), Mayes et al. (2013) and the qualitative studies reviewed earlier may be sensory in origin, although in Kerns et al. (2014) such fears were identified in the absence of a generalized sensitivity to noise. Sasson et al. (2008) and Uljarevic et al. (2016) identified children and adolescents with ASD with adaptive, moderate, and severe sensory related symptomatology based on parent report and found that participants with moderate and severe sensory symptoms had significantly higher anxiety scores than those with less severe sensory symptoms.

Anxiety Presenting as Increases in Challenging or Stereotyped/Repetitive Behaviors

In all four qualitative studies, participants reported anxiety being expressed in the form of observable increases in repetitive/stereotyped or challenging behaviors. In a large clinically referred sample of 445 children and young people with ASD, Hallett et al. (2013a,b) found that items such as restlessness, tension, and sleep difficulties were most frequently endorsed at the clinical level on the caregiver-rated Child and Adolescent Symptom Inventory (CASI) Anxiety scale as manifestations of anxiety compared to other anxiety symptoms. Storch et al. (2012) also found that children with ASD and anxiety presented with more co-occurring disruptive behavior problems that further exacerbated functional impairment. Similarly, Niditch et al. (2012) reported that a positive relationship between IQ and anxiety symptoms was mediated by the presence of aggression. In an earlier study, Evans et al. (2005) found that the association between increased behavioral problems and anxieties and fears was significantly higher in children with ASD compared to the same relationship in children with Down syndrome or typically developing children (see also Farrugia and Hudson, 2006). Rodgers et al. (2012) found that children with ASD and higher anxiety displayed more repetitive behaviors that those with fewer anxiety symptoms. Spiker et al. (2012) also found that symbolic enactment of restricted interests in play was associated with greater anxiety symptoms in youth with ASD. Collectively, these findings suggest that increased externalizing/challenging or repetitive and restricted behaviors may be a way of expressing or coping with anxiety in youth with ASD resulting in a distinct anxiety presentation consistently identified in both qualitative and quantitative studies.

Reduced Verbal Reporting of Anxiety and Anxiety-related Cognitions

In Ozsivadjian et al.’s study (2012), parents reported that their children had great difficulty in telling them about their worries. In line with this, Hallett and colleagues (2013a,b), using the caregiver-rated CASI-Anxiety scale, found that items requiring verbal expression of worries (i.e., “worries about physical health”; “complains about feeling sick or heart pounding”) were rarely endorsed at the clinical level by caregivers, suggesting that questionnaires, especially carer- or teacher-reports, may fail to capture such important features of anxiety. However, when physical and social threat-related cognitions were specifically examined using self-report measures, Farrugia and Hudson (2006) found that adolescents with ASD reported more anxious cognitions and dysfunctional attitudes than clinically anxious adolescents without ASD despite equivalent overall anxiety using the CATS (Children’s Automatic Thoughts Scale). Similarly, using the same measure, Ozsivadjian et al. (2014) also found that child-rated scores on the CATS were positively correlated with anxiety symptoms, whereas the typically developing comparison group did not show this association. Elevated dysfunctional attitudes in ASD compared to those reported by typically developing children were also found by Greenaway and Howlin (2010). Kerns et al. (2014) also found that anxious cognitions (reported by youth via a questionnaire) predicted both traditional and ASD-related anxiety in their sample. These studies suggest that cognitively and verbally more able children with ASD may in fact be able to accurately report their cognitions, but possibly only with certain supports (e.g., via a structured questionnaire, which offsets difficulties with generation, verbal processing, and conversational to and fro skills typically required in clinical interviews or informal questioning by caregivers).

In the study by Ozsivadjian et al. (2012), when parents were able to make comments about their children’s cognitive processing style, some interesting patterns emerged. A few typical cognitive distortions or “thinking errors” were described, such as “I’m totally useless” or “The world’s against me”. More commonly reported were more ASD-related cognitive processes, such as a delay between a stressful experience occurring and the verbalization of the worries related to the event, and also disorganized, piecemeal, or negatively biased verbal reporting of events and worries.

Factors Associated with Traditional and Distinct Anxiety Symptomatology

Although numerous studies have examined factors associated with the presentation of anxiety in ASD (i.e., gender, chronological age, IQ, adaptive functioning, ASD symptom severity, and others; see, e.g., Dubin et al., 2015; Hallett et al., 2013a,b; Magiati et al., 2016; van Steensel et al., 2011; Vasa and Mazurek, 2015 for reviews; see also Chapter 1: Introduction), to date only Kerns et al. (2014) have explored whether child characteristics may be differentially associated with traditional versus ASD-related anxiety presentations. They found that a more anxious cognitive style, higher language ability, and hypersensitivity were positively associated with traditional anxiety, while ASD symptom severity was not. In contrast, only ASD symptom severity and an anxious cognitive style predicted ASD-related anxiety. These findings suggest two qualitatively and phenomenologically distinct mechanisms of anxiety in ASD—one which is similar in processes and presentation to the anxiety of youth without ASD, and another wherein anxiety and ASD-related vulnerabilities and symptoms may interact. Previous research examining the relationship between ASD symptom severity and anxiety has been inconsistent, with some studies finding no association (e.g., Hallett et al., 2013a,b; Renno and Wood, 2013; Simonoff et al., 2008) and others reporting a significant correlational or predictive positive relationship (i.e., Magiati et al., 2016; Mayes et al., 2011; Sukhodolsky et al., 2008). Findings by Kerns et al. (2014) suggest that there may be a stronger relationship between ASD symptom severity and ASD-related anxiety as compared to traditional anxiety symptoms, lending some insight to these inconsistent findings.

Future Directions and Implications for Research and Practice

Implications for Conceptualizing Anxiety in ASD

A number of models have been proposed, conceptualizing anxiety in ASD, summarizing current knowledge, and offering future directions for research. One of the earliest ones was that of Wood and Gadow (2010), who proposed hypothetical pathways between ASD-related stressors and anxiety (which could in turn, it was proposed, lead to an increase in ASD-related behaviors, such as social avoidance). Ollendick and White (2012) proposed a theoretically informed model of shared and unique processes of anxiety in ASD: shared processes likely include negative bias, unhelpful automatic thoughts, and physiological arousal, while more ASD-specific processes likely include social confusion, difficulties “reading” emotions in others, sensory sensitivity, negative social interactions with others, rigidity, and insistence on sameness. Kerns and Kendall (2014) also presented the first review of the differences and similarities between anxiety in youth with and without ASD and suggested the presence of both traditional and more ambiguous or ASD-related anxiety presentations with distinct phenomenologies.

Our critical review of the literature to date in this Chapter largely supports the hypotheses put forth in these models, particularly the distinction of shared and distinct presentations of anxiety in ASD and their potentially different etiological pathways. Nonetheless, future research is needed. Whereas more traditional anxiety presentations may be true co-occurring mental health conditions in individuals with ASD, reflecting “true” psychiatric comorbidity, more distinct symptoms may reflect a distinct anxiety sub-type relating to the severity and impact of the individual’s core ASD challenges or, alternatively, core symptoms of ASD (see Kerns and Kendall, 2014; Wood and Gadow, 2010).

In keeping with Kerns and Kendall (2014), the discourse relating to how anxiety problems should be conceptualized in ASD should thus potentially not be a comorbidity or ASD variance dichotomy, but rather a comorbidity and ASD variance explanation, depending on the nature of the presenting anxiety difficulties. Further studies are needed to examine factors associated with and predicting traditional as compared to ASD-related anxiety symptoms. True comorbidity of anxiety disorders with ASD will require evidence of not only phenotypical similarities, but also common pathways to traditional anxiety in individuals with and without ASD, such as cognitive factors (e.g., attentional bias), environmental factors (e.g., conditioning, adverse experiences, parenting), and genetic factors. Anxiety-like presentations specifically associated with core ASD features may, in contrast, be better predicted by ASD-specific factors, such as weaker central coherence, social impairments, and sensory oversensitivity. Direct tests of these hypotheses are needed.

Implications for the Measurement and Assessment of Anxiety in ASD

A primary challenge in the identification of anxiety in ASD is the overlap between anxiety and ASD features. Disentangling overlapping presentations requires very careful, comprehensive assessment by experienced professionals with expertise in ASD and the presentation of mental health problems in this population. Furthermore, the diagnostic identification of comorbid anxiety in ASD was not previously aided by earlier classification systems, such as DSM-IV-TR (APA, 2000), which prohibited the diagnosis of some anxiety disorders if ASD was also present. These prohibitions are no longer included in DSM-5, which instead encourages clinicians to disentangle anxiety symptoms as much as possible from ASD symptomatology when considering a comorbid diagnosis. Nonetheless, the actual process of assessing and differentiating anxiety symptoms in youth with ASD is still very much reliant on clinical opinion, in the absence of specific, widely-available measures for this target population (although see Chapter 5: Assessment of Anxiety in Youth with Autism Spectrum Disorder and Chapter 6: Cognitive-Behavioral Principles and Their Applications Within ASD for advances in the assessment of anxiety; and recommendations by Kerns et al., 2016 and Vasa et al., 2016). Existing anxiety measures developed for use with typically developing children appear useful to some extent, but are likely not sufficient to capture the full range of anxiety presentations in ASD. Data on the varied presentation of anxiety in ASD underscore the importance of evaluating the reliability, validity, and clinical utility of existing measures (i.e., see Magiati et al., 2014; van Steensel et al., 2014; Zainal et al., 2014), while also systematically piloting expanded versions of such measures that include items for the more distinct manifestations of anxiety in ASD (see Bearss et al., 2016; Kerns et al., 2014; Rodgers et al., 2016). The ADIS/ASA has demonstrated valid and reliable measurement of traditional and more distinct manifestations of anxiety in ASD, but is time intensive with regard to training and administration and thus may be most useful when careful and comprehensive description of anxiety symptoms is needed for research or clinical practice (Kerns et al., in press). Future studies may enrich our understanding of anxiety in ASD by reporting not only total or subscale anxiety scores, but also item-level analyses. As demonstrated by Hallett et al. (2013a,b), this level of detail can be extremely informative, particularly in the development or adaptation of existing measures for use in ASD.

Implications for Anxiety Interventions in ASD

Existing evaluations of anxiety interventions for young people with ASD (see Ung et al., 2015; Vasa et al., 2014 for reviews) have focused on adapting existing CBT-oriented anxiety interventions developed for children without ASD (see Moree and Davis, 2010), however less attention has been paid to examining whether different adaptations or approaches may be more effective in treating “traditional” as compared to ASD-related anxiety. It may be that traditional anxiety will respond better in interventions informed by existing traditional anxiety programs, while targeting core ASD symptomatology (especially reducing insistence on sameness or managing sensory sensitivity) may be more effective treatment targets in interventions for ASD-related anxiety problems. Future intervention studies would benefit from describing their participants’ anxiety presentations within frameworks conceptualizing the full range of traditional and ASD-distinct presentations and evaluating their differential treatment response.

In conclusion, both traditional and distinct aspects of the phenomenology of anxiety in ASD should be considered in our future clinical and research efforts to define, measure, understand, and treat anxiety more comprehensively and effectively in ASD.

Acknowledgments

We would like to give thanks to Hannah Long for her invaluable help in compiling the reference list. This work was also supported by grant funding awarded to C. Kerns (K23 HD087472).

Vignette 1—“Traditional” presentation: separation and generalized anxiety

Sophie was an 11-year old cognitively able young woman with ASD. She had a longstanding history of peer rejection at school and did not have friends. She was very reluctant to go places without her parents or allow them to go anywhere without her. If they did, she would call them repeatedly, up to 20 times per day. She carried with her when out bags containing “security items”, as she believed she might need them in case she was kidnapped or her parents were killed, which she often worried about. She often went to sleep in her parents’ bed at night due to anxiety. She worried about her performance at school, her health and her family’s health, and the war in Iraq.

Vignette 2—“Distinct” presentation: anxiety around change

James was a 4-year old minimally verbal boy with ASD and moderate intellectual disability, with strengths in visual processing. He was attending pre-school daily and had been very happy and settled there. He was using visual schedules with real photos of objects or activities to communicate with others at home and pre-school. Part way through the year, however, he demonstrated a behavioral change, becoming increasingly more irritable, and with increased echolalia and stereotyped speech. He started to show extreme anxiety when separating from his mother for pre-school. The teachers have in the last month switched to a communication system with symbols rather than photos of real objects and a new substitute teacher has been teaching his class, because his regular teacher has been unwell.