Restricted, repetitive behaviors in autism refer to an assemblage of behaviors defined by their topographical similarity across contexts, inappropriateness, behavioral rigidity, and repetition. In the Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition (American Psychiatric Association 2000), criteria for repetitive behavior can be met by a person exhibiting at least one of the following: “(a) encompassing preoccupation with one or more stereotyped and restricted patterns of interest that is abnormal either in intensity or focus; (b) apparently inflexible adherence to specific, nonfunctional routines or rituals; (c) stereotyped and repetitive motor mannerisms (e.g., hand or finger flapping or complex whole-body movements); or (d) persistent preoccupation with parts of objects.” This symptom domain has been conceptually and empirically grouped into at least two categories—“lower order” and “higher order” behaviors (Szatmari et al. 2006; Turner 1999). Lower order motor actions [i.e., stereotyped movements, repetitive manipulation of objects, and repetitive forms of self-injurious behaviors (SIBs)] are characterized by repetition of movement, and “higher order” behaviors (i.e., compulsions, rituals and routines, insistence on sameness, and circumscribed interests) are characterized by a rigid adherence to some rule or mental set (e.g., needing to have things “just so”) (Turner 1999). What becomes clear on examination of these diagnostic criteria and individual behaviors is that they are very broad, ranging from repetitive movements of the body to more cognitively mediated symptoms such as intense interests or hobbies. Although no single type of repetitive behavior may be specific to autism, previous studies have found that repetitive behaviors in autism are characterized by a pattern of elevated occurrence and co-occurrence of these behaviors (Bodfish 2010; Bodfish et al. 2000).

Previous research has demonstrated that children with autism demonstrate unusual responses to sensory stimuli to a greater degree than their peers with other types of developmental disabilities or typical development (Baranek et al. 2006; Rogers et al. 2003; Watling et al. 2001). Although these unusual behaviors have been categorized in different ways, one current empirically supported conceptualization focuses on three sensory features in autism: hyporesponsiveness, hyperresponsiveness, and sensory seeking (Ashburner et al. 2008; Ben-Sasson et al. 2009; Boyd et al. 2010; Liss et al. 2006). Hyporesponsiveness is characterized by an absence of the expected response to a stimulus (e.g., failing to orient to a novel sound in the environment), a delayed response, or a higher response threshold (e.g., only orienting to a novel sound when it becomes more intense). Hyperresponsiveness is characterized by an exaggerated behavioral reaction, aversive response, or effort to avoid a sensory stimulus. Sensory seeking behaviors are actions that perpetuate or intensify a sensory experience, such as staring intensely at flickering lights or sniffing objects. These patterns are not mutually exclusive and may co-occur in individual children across sensory modalities (Baranek et al. 2006; Liss et al. 2006).

Repetitive behaviors comprise one faction of the triad of clinical symptoms that characterize autism spectrum disorder (ASD), often presenting concurrently with social and communication symptoms. Higher order behaviors, specifically the presence of restricted or narrow interests, odd object attachments or unusual object preoccupations (e.g., fascination with ceiling fans), are thought to be more unique to autism in comparison to lower order types of repetitive behaviors (e.g., self-injury; Lam et al. 2008). In contrast to repetitive behaviors, sensory features in autism are considered secondary or associated symptoms of the disorder. Aberrant sensory features are presumed to be less specific and universal in autism than the triad of core features, although there is some evidence that the hyporesponsive sensory feature is more associated with autism (Baranek et al. 2006).

In three prior studies, researchers reported associations between sensory features and repetitive behaviors in autism (Boyd et al. 2009, 2010; Gabriel et al. 2008). Gabriel et al. (2008) used parent report measures to examine the association between these atypical behaviors in a sample of 70 children and adolescents with ASDs. The researchers found consistently high levels of sensory features and repetitive behaviors in a subgroup of the sample. Similarly, Boyd et al. (2009, 2010) found the same relationship in a group of school-aged children (6–17 years of age) with higher functioning autism. These studies demonstrated that sensory symptoms and repetitive behaviors often co-occurred in autism. However, both studies only used parent report measures of sensory and repetitive behaviors. Observational methods are needed to validate these clinical phenomena and their interrelationships. In a latter study, Boyd et al. (2010) used a combination of parent report and direct behavioral observation measures to examine the relationship between children’s sensory features and repetitive behaviors (N = 109). The researchers found that the hyperresponsiveness sensory feature was most associated with the expression of repetitive behavior in autism. Examining the relationships between the expression of specific types of sensory symptoms and repetitive behaviors provides a more comprehensive understanding of the behavioral phenotype in autism, and better informs clinical practice and the selection of therapeutic interventions.

Based on research, it appears well established that brain abnormalities (Langen et al. 2010; Tommerdahl et al. 2008) as well as genetic factors (Hus et al. 2006; Lam et al. 2008; Tadevosyan-Leyfer et al. 2003) are involved in the expression of repetitive behaviors and sensory features in autism. However, there is also research that points to the importance of environmental (Ashburner et al. 2008) and family context (Bagby et al. 2012; Smith et al. 2008) in the expression of these behaviors in individuals with autism. For example, Ashburner et al. (2008) found that for children with autism, difficulties with sensory processing (e.g., inability to filter auditory information) were a more significant predictor of school academic achievement than IQ scores. Smith et al. (2008) found that the parents of individuals with autism can impact their expression of repetitive behaviors into the adolescent and adult years, with mother–child dyads having a higher relationship quality, more warmth, and praise associated with reductions in these symptoms over time. Thus, assessment of and interventions for repetitive behaviors and sensory features in autism must take into account the transactional effects between the child’s environment, including individuals in that environment, and any manifestation of these behaviors.

There are different topographies and types of repetitive behaviors expressed in autism, and they often require differential intervention approaches. The field of applied behaviors analysis has informed most of the evidence base on interventions for repetitive behaviors (Boyd et al. 2012). Presently, there is more evidence to support the use of practices to treat lower order behaviors (i.e., self-injury and stereotypy) but less evidence to guide the selection of intervention practices for higher order behaviors (e.g., insistence on sameness). Therefore, the evidence presented for higher order behaviors can be thought of as promising practices. Any selection of intervention practices for repetitive behaviors should still be guided by thorough assessment of these behaviors and individualized to the child and its family. Finally, the selection of evidence-based practices should be guided by a combination of the research base and practitioners’ use of their professional judgment about context and values (their own and those of the family) in the application of such practices (cf. Buysse and Wesley 2006).

Methodologies used in the identification and measurement of repetitive behavior and its severity primarily fall into three categories: observational measures, standardized assessments, and informant report. These approaches can involve identification of the repetitive behavior and its frequency, duration, and intensity. Direct behavioral observation often involves coding the child’s behavior in real time, and is used quite often in the identification of problem behaviors. One of the issues with direct observation is that it can be hard to capture behavior that is context-bound, meaning that the child may only display the behavior in specific contexts (e.g., home and not school) and under certain environmental circumstances (e.g., only when watching a favorite TV show). Even with stereotypical motor movements, Schultz and Berkson (1995) found direct observation to be only 33  % consistent with teacher report. By using video, more continuous data can be collected and behaviors can be reexamined; however, the limitation is that coding of videos is time consuming and often not practical for clinicians. Standardized assessments often use direct observation or video to rate the child’s response to a specific situation or task (e.g., Autism Diagnostic Observation Schedule; Lord et al. 2000), but these measures tend to focus on determining the sheer presence of the repetitive behavior rather than the specific type and severity. Rating scales such as the Repetitive Behaviors Scale—Revised (Bodfish et al. 1999) and Pervasive Developmental Disorder Child Yale-Brown Obsessive Compulsive Scale (Scahill et al. 2006) provide greater specificity with regard to the topography and severity of RRB; however, they often require subjective rating from an informant. It is recommended that a multimethod (e.g., observational and rating scales) and multiinformant (e.g., clinician and parent) approach be used to assess repetitive behaviors in autism and obtain a better understanding of the role of environmental context in the expression of these behaviors.

There are at least three distinct theoretical or conceptual models that can inform or have already informed intervention practices for repetitive behaviors in autism. As mentioned previously, the field of applied behavior analysis (ABA) has informed the bulk of the intervention literature for repetitive behaviors. However, research on environmental deprivation and cognitive models of repetitive behavior also can inform intervention selection. These theoretical and conceptual models are briefly described.

The category of lower order repetitive behaviors is comprised of two types of behaviors—stereotypy and self-injury. The majority of the intervention research has focused on the treatment of behaviors.

Perhaps surprisingly, one of the more replicated findings in the intervention literature is the effect of physical exercise on the stereotypic behavior of individuals with autism (see Lang et al. 2010). The intervention often involves the individual engaging in a vigorous exercise routine (e.g., jogging, roller skating) prior to participating in a subsequent task or activity that has been associated with stereotyped behavior (Kern et al. 1984). It is not fully understood why engaging in physical exercise leads to subsequent, even if fleeting, reductions in repetitive behaviors. Lang et al. (2010) proposed two reasons—first, subsequent reductions in repetitive behavior may be an artifact of fatigue; and secondly, that engaging in physical exercise may provide the individual access to the same intrinsic reinforcer as engaging in stereotypic behavior. As Lang et al. point out, it is likely that the first hypothesis can be ruled out because excessive fatigue is counterintuitive to the concurrent increases in appropriate behaviors that have been found following vigorous exercise routines (Kern et al. 1982; Powers et al. 1992). Another possible explanation for the effects of physical exercise on stereotypy is arousal theory. It has long been postulated that individuals with developmental disabilities, including autism, engage in stereotypic behavior to attenuate or intensify their arousal levels, depending upon the environmental circumstances (Turner 1999). Perhaps engaging in physical exercise decreases the need to engage in stereotypy or SIB to modulate one’s level of arousal. Still, as Rogers and Ozonoff (2005) point out, the evidence is mixed on the role arousal plays in the sensory or repetitive behaviors of individuals with ASD.

Finally, environmental enrichment has been used to decrease stereotypic behavior. This class of interventions involves providing the individual free access to appropriate, competing sources of reinforcement, such as preferred objects (Rapp and Vollmer 2005). In comparison to environmental enrichment, skill enrichment entails teaching the individual more adaptive skills (e.g., social initiation skills); with the thought being that an increase in appropriate skills will offset the need to engage in behaviors that do not serve a clear purpose or goal. Through a series of single-case design studies, Loftin et al. (2005, 2008) demonstrated that the use of peer-mediated intervention strategies, direct social skill instruction for the child with ASD, the use of self-monitoring strategies, or some combination thereof, led to increases in children’s social interactions with peers and concomitant decreases in stereotypic behavior (Lee et al. 2007; Loftin 2005; Loftin et al. 2008).

Higher order behaviors include the repetitive behaviors of obsessions and compulsions, circumscribed interests, and the child’s insistence on sameness. As with lower order repetitive behaviors, an ABA perspective has informed much of the intervention research on higher order repetitive behaviors. One exception to this is the intervention research on obsessive and compulsive behavior in individuals with autism, which can trace some of its roots to cognitive-behavioral therapy (CBT).

In two published case studies, CBT was used to treat “repetitive behavior-like” symptoms of individuals with ASD (Lehmkuhl et al. 2008; Reaven and Hepburn 2003). Both studies involved traditional elements of CBT/ERP—(a) psychoeducation session(s) with caregivers or children to help reframe their thinking about why the individual engages in obsessive-compulsive behaviors; (b) the development of a stress hierarchy that involves listing symptoms of OCD from least to most distressing to identify the individual’s “transition zone” (i.e., the area where the individual has had some success in inhibiting OCD symptoms); and (c) homework assignments that involve the individual practicing ERP techniques outside of the clinic sessions (March and Mulle 1998). In addition, modifications to traditional CBT techniques were included to accommodate symptoms of autism; for instance, the use of written schedules to denote the sequence of activities that would occur during the therapy session. Currently, it appears that CBT/ERP could be a promising treatment for individuals with ASD who have a comorbid diagnosis of OCD. However, given the cognitive components (e.g., cognitive reframing) involved in CBT it appears more applicable to individuals with intact cognitive abilities, thus, the application of CBT to individuals with autism who have cooccurring intellectual or language disabilities is indeed an area for future research.

The field of ABA has made significant and lasting contributions to the evidence base for the treatment of repetitive behavior in autism. Although the science of behavior provides a wealth of knowledge, there are limitations with the current state of the evidence. The primary limitation is that most of the research has focused on lower order behaviors, such as stereotypies or self-injury, thus, less is known about the effectiveness of these strategies or other therapeutic approaches to treat higher order repetitive behaviors.

As noted in previous sections, sensory features associated with autism may include hyperresponsivity, hyporesponsivity, and sensory seeking behaviors, and many children present with mixed patterns of sensory processing. In order to optimize intervention, one must understand the effect of sensory processing on behavior, conduct a thorough and appropriate assessment, use a valid theoretical model to guide intervention planning, and use intervention strategies that are evidence-based, family-centered, and can be implemented within everyday environments and routines.

Assessment of sensory processing patterns is based on a combination of skilled observations and the report of parents and/or other caregivers who know the child well. Skilled observations of the child’s responses to a variety of sensory experiences may take place in natural and/or clinical settings, but should include a combination of the two whenever possible. In addition, observations should occur in more than one instance, if possible, in order to discern whether the child’s responses reflect true patterns of sensory processing or are the result of isolated or specific circumstances. Observations should include not only the behavior of the child, but also the sensory, physical, social, and temporal aspects of the environment and of the activities in which the child is engaged during that observation period. There is currently no standardized observational tool for use in natural environments, but clinical observational measures include the Test of Sensory Function in Infants (DeGangi and Greenspan 1989) and the Sensory Processing Assessment (Baranek 1999; Baranek et al. 2007).

The report of parents and other adult caregivers who know the child well offers a historical perspective on responses to various sensory experiences, and contributes significantly in determining the sensory processing patterns of the child. Standardized caregiver report measures such as the Sensory Profile (Dunn 1999), the Infant-Toddler Sensory Profile (Dunn 2002), the Sensory Processing Measure (Kuhaneck et al. 2007), the Sensory Processing Measure—Home Form (Parham and Ecker 2007), and the Sensory Experiences Questionnaire (Baranek et al. 2006) are often used in addition to the skilled observations of an occupational therapist to assess a child’s patterns of sensory processing. This thorough assessment process is important not only to gather information about the child’s sensory processing differences, but also to ascertain when those sensory processing differences are not the cause of a particular child’s behaviors. For instance, without thorough assessment, assumptions may be made that a child reacts aversively to get its hair cut because of the tactile sensations on his head and around his face, when in fact it is actually a combination of other aspects of that routine, rather than sensations inherent in haircutting, that are difficult for him to handle.