Persons with autism spectrum disorder (ASD) present with a unique array of challenges as well as strengths (Amaral et al. 2011). The therapeutic focus initially emphasizes interventions for the core problems surrounding social interactions and relationships as well as the difficulties that arise due to their restricted and repetitive patterns of behavior, interests, and activities (American Psychiatric Association 2013). In addition, common comorbid problems such as anxiety, depression, sleep disorders, and difficulties with executive functioning further complicate efforts to assist these individuals with becoming more independent. A number of behavioral and cognitive-behavioral approaches exist to treat these conditions (Durand 2014a). This chapter reviews the extant evidence base for these treatments. However, first discussed will be the range of presentations of ASD and the social nature of this disorder since this guides treatment development, implementation and outcomes.

DSM-5 significantly (and somewhat controversially) reorganized how ASD is diagnosed by combining previously separate disorders (i.e., autistic disorder, Asperger’s disorder, childhood disintegrative disorder, and pervasive developmental disorder—not otherwise specified) under a single diagnostic entity (American Psychiatric Association 2013; Lord and Jones 2012; Skuse 2012). Full discussion of the rationale for this change and the scientific and political implications is beyond the scope of this chapter. However, it is important to highlight that a diagnosis of ASD by itself has poor predictive and treatment validity. In other words, the range of functioning of persons with ASD and the developmental trajectories of the disorder are quite diverse (e.g., Fein et al. 2013; Lord et al. 2012) and therefore requires a comprehensive assessment of strengths and needs for each person affected (Steiner et al. 2012). The addition of three levels of severity included in the DSM-5 diagnosis of ASD (Level 1—“Requiring support,” Level 2—“Requiring substantial support,” and “Level 3—“Requiring very substantial support”) attempts to highlight these differences (American Psychiatric Association 2013), however their somewhat subjective nature and limited specificity provide only partial assistance for treatment design. This chapter adopts the convention of referring to persons with “mild” ASD and “severe” ASD (Durand 2014a). This division maps roughly onto the previous distinction between persons who are able to converse with others but demonstrate difficulties with pragmatics (previously diagnosed with Asperger’s disorder) and those individuals who have severe language and communication problems (previously diagnosed with autistic disorder, childhood disintegrative disorder or pervasive developmental disorder—not otherwise specified). Although it is recognized that persons with ASD manifest a broad spectrum of challenges, research on treatment typically focuses separately on these two distinct groups.

In addition to the range of symptoms influencing treatment choice and outcome, the nature of ASD directly impacts treatment design. Specifically, the disruption in typical social motivation displayed by persons with ASD provides interventionists with unique challenges. Those with ASD either lack interest in others or are impaired in their ability to navigate social interactions (Dawson 2008; Mundy 2011). Therefore, unlike those with other disorders where the motivation to enter into treatment is usually some combination of the desire to reduce personal distress and the wish to please significant others (Barlow and Durand 2014), this is often not the case among persons with ASD. At the more severe end of the autism spectrum, the lack of social motivation to participate in treatment is typically supplemented by more extrinsic motivation (e.g., the use of non-social reinforcers). Those on the mild end of this spectrum will sometimes express distress over either their social difficulties alone (e.g., not being able to make and keep friendships) or comorbid difficulties that may be the result of these social problems (e.g., anxiety, depression). However, difficulties such as in impairments with theory of mind (i.e., not being able to “read” and understand what others might be thinking or feeling) complicates the ability to apply standard approaches to intervention. As a result, social skills training programs as well as cognitive-behavioral treatments for comorbid problems need to be significantly adapted for this group (Durand 2014a).

Early intervention approaches are receiving increased attention, especially as the ability to diagnosis children with ASD improves to the point where it may be possible to identify these children as early as 6 months of age (e.g., Elsabbagh et al. 2012; Kim and Lord 2012; Veness et al. 2012). One of the first large scale studies reported on the effects of using intensive intervention using techniques from applied behavior analysis with very young children with ASD (younger than 3 1/2 years of age) (Lovaas 1987). This intervention—now referred to as Early and Intensive Behavioral Intervention (EIBI)—involved 40 h per week of educational efforts over the course of two years. During the first year, the focus of treatment was to reduce repetitive behaviors, teach the children to speak in words, increase compliance on tasks and imitate adults as well as establishing the beginnings of appropriate toy play. The second year involved teaching targeted expressive and early abstract language and interactive play with peers and was extended into the community to teach children to function within a preschool setting. In that study, it was suggested that almost half of the children “recovered” from their ASD—meaning that several years later teachers could not tell them apart from their students without ASD. Although criticized on methodological grounds, that study led the way for a growing database that points to the value of early and intensive intervention for some children with ASD (Peters-Scheffer et al. 2011; Reichow and Wolery 2009; Strain and Bovey 2011; Warren et al. 2011).

Reviews of successful early intensive behavioral intervention programs suggest a number of common features in each program including; (1) comprehensive curriculum focusing on imitation, language, toy play, social interaction, motor, and adaptive behavior, (2) sensitivity to developmental sequence, (3) supportive, empirically validated teaching strategies (applied behavior analysis), (4) behavioral strategies for reducing interfering behaviors, (5) involvement of parents, (6) gradual transition to more naturalistic environments, (7) highly trained staff, (8) supervisory and review mechanisms, (9) intensive delivery of treatment (25 h per week for at least 2 years), and (10) initiation by 2–4 years of age (Dawson and Osterling 1997; Green et al. 2002; National Research Council 2001). More recent work that blends naturalistic and developmentally-driven behavioral interventions suggests that some of these suggestions (such as the intensity of training) may need to be revised (e.g., Dawson et al. 2010; Strain and Bovey 2011).

Researchers are now examining the characteristics of children (for example, language ability, IQ) that may predict the best treatment outcomes. In addition, some emerging important work suggests that early behavioral intervention may alter the functioning of the developing brain in these children to be comparable to typically developing children (Dawson et al. 2012; Voos et al. 2013). For example, Dawson et al. (2012) assessed brain activity to objects versus faces in young children with ASD (age 48–77 months) both before and after involvement in a developmental behavioral intervention (the Early Start Denver Model [ESDM]) and compared their responses to a community based control group. Their main finding was greater cortical activation while viewing social stimuli (faces) among the ESDM treatment group and they also noted that this finding was associated with improved social behavior. As discussed later in the chapter, this type of intervention may help improve social motivation at a critical developmental period and may assist with learning an array of social skills. Overall, the treatment progress in areas such as language, cognitive abilities, and adaptive behavior seems to greatly improve for many children with ASD if intensive behavioral intervention is implemented early in life and with fidelity (Dawson and Burner 2011; Strain and Bovey 2011). These changes appear quite robust, and for many children appear to be maintained years after initial intervention (e.g., Kasari et al. 2012a).

Described next is the research on treatment for the problems displayed by older children and adults on the more severe end of the autism spectrum, followed by work with individuals on the less severe end of this spectrum. Treatment typically begins with teaching social and communication skills (Goldstein 2002; Kasari and Locke 2011; National Research Council 2001) and implementing interventions for disruptive or destructive behaviors (e.g., tantrums, aggression, self -injury) (Durand 2012). The goals of treatment are to increase the individual’s ability to be more independent and improve the overall quality of life (Myers and Johnson 2007).

Behavioral research targeting social communication skills in persons with more severe ASD dates back to the early 1960’s. At that time problems with social communication were viewed primarily as skill deficits. In other words, it was thought that individuals with ASD did not appropriately interact with others because, for reasons as yet unknown, they did not learn the necessary skills. Remediation, therefore, would involve using behavioral interventions to teach these new skills (Hewett 1965; Lovaas and Smith 1989). Intervention approaches typically involve techniques developed from applied behavior analysis (ABA), which uses principles of learning theory to encourage new behaviors by manipulating antecedents and consequences (Matson et al. 2012; Smith 2001). Some of the earliest efforts in this area were documented by Ivar Lovaas and his colleagues at UCLA. For example, they used the basic behavioral procedures of shaping and discrimination training to teach nonspeaking children with ASD to imitate others verbally (Lovaas et al. 1966). Once children could imitate, teaching speech became easier, and progress was made in teaching some of these children to use labels, plurals, sentences, and other more complex forms of language (Lovaas 1987). Initially these treatments were delivered primarily in highly structured settings with adult directed activities and artificial consequences (referred to as discrete trial training) (McEachin and Leaf 1999; Smith 2001). Recent modifications allow for a more naturalistic and developmentally driven context, with behaviors taught in typical settings, child directed activities and natural consequences. Growing evidence supports this approach for improving generalization and maintenance of learned skills (Schreibman and Ingersoll 2011).

Despite some successful outcomes using behavioral interventions to teach communication skills there are limitations with this approach. First, many children with ASD are not able to benefit from this training such that they gain functional speech skills (Carr 1979; Howlin 1981). Second, for those who do acquire speech it is often limited to speech that would lead to non-social outcomes (e.g., asking for soda). Socially oriented speech (e.g., saying “Hello” to another person) is less likely to be produced (Koegel et al. 2011; Wetherby and Prutting 1984). This observation is related to the limited interest in social interactions by this population. For example, the skills needed to engage in a social exchange (e.g., saying “Hi! How are you?”) are unlikely to be maintained given that the outcome is an increase in social attention from another person—a consequence that may not be reinforcing to many persons with severe ASD.

Currently, initial efforts with young children focuses on improving social motivation by teaching imaginative play and joint or shared attention (defined as “spontaneous seeking to share experience, enjoyment, interests, or achievements with other people” Mundy 2011, p. 151). These skills are characteristically absent or impaired in young children with ASD and are seen as pivotal prerequisites necessary for later development of important social, communicative and cognitive skills (Charman et al. 1997; Koegel et al. 2001; Poon et al. 2012). For example, longitudinal research suggests that deficits in joint attention predict language delays years later (Mundy et al. 1990; Stone and Yoder 2001; Thurm et al. 2007). The neurobiology of these skills is just beginning to be explored. One study of adults with ASD, for example, found an atypical pattern of brain activation (including in the dorsal medial prefrontal cortex (dMPFC) and right posterior superior temporal sulcus (pSTS)) in these adults during a joint attention task as compared to neurotypical adults during the same activity (Redcay et al. 2013).

A number of studies now show that many young children with ASD can be taught these skills through naturalistic behavioral approaches (Kasari et al. 2012b; Lawton and Kasari 2012; Wong and Kasari 2012). In one study, 58 3–4 year olds with ASD received either joint attention intervention, symbolic play intervention or were in control group (Kasari et al. 2006). Intervention involved a combination of behavioral (e.g., prompting behaviors, reinforcement) and naturalistic procedures (e.g., using child interests to create learning opportunities) to encourage either joint attention or symbolic play. Sessions were conducted 30 min daily for 5–6 weeks. The results demonstrated the success of this limited intervention in improving these skills in young children with ASD (Kasari et al. 2006). Importantly, a follow-up of these children demonstrated that their improved ability to engage in joint attention and symbolic play predicted their spoken language ability 5 years after the initial training (Kasari et al. 2012a). This finding is significant since having functional spoken language at age 5 is highly predictive of positive social and cognitive outcomes (Billstedt et al. 2005; DeMyer et al. 1973; Venter et al. 1992).

Although it is premature to say definitively why teaching joint attention and related skills improves later speech, it may be helpful to consider the distinction between viewing social communication problems as a skill deficit alone and considering these difficulties also as a problem of social motivation. By teaching joint attention and play skills to young children with ASD, its value may lie in pairing activities the child enjoys with the presence of and interaction with others (Mundy and Gomes 1998). If other people are viewed as positive, engaging in social interactions will result in desirable outcomes, encouraging more engagement with others (Dawson 2008; Koegel et al. 2011; Stavropoulos and Carver 2013). Increased social motivation may be the mediating factor in the role of improved joint attention leading to improved social communication skills (Paul et al. 2013).

Intervention for the communicative skills of older individuals with limited abilities focuses on teaching either spoken language or the use of augmentative communication strategies (e.g., pointing to a picture, using vocal output devices) (Ganz et al. 2012; Odom et al. 2010b; Prelock et al. 2011). Using graphic symbols (e.g., pictures of desired items or activities) to encourage communication has an emerging research base and are employed as visual aids for monitoring school and home schedules (e.g., visual supports); additionally graphic symbols assist students to make choices (e.g., pointing to one of a series of activity options), and to engage in general expressive communication (Wegner 2012). A variation of this approach teaches students to select one or more pictures/words and hand them to another person—called the picture exchange communication system (PECS) (Bondy and Frost 1994; Tincani and Devis 2011). The potential advantage of this approach is that the student can initiate communication with someone even if the other person’s attention is focused elsewhere.

A related approach uses speech generating devices (SGDs) to assist non-verbal students to communicate with others (Ganz et al. 2012). A variety of devices are available that can be programmed to generate human speech when a picture or word is pressed by the student (e.g., “Help me.”). These devices have both the advantage of being able to catch the attention of others as well as being understood by anyone. With advancing technologies and available software that can be used to easily program these devices (e.g., tablet computers), their use is increasing. Some research suggests that students with ASD may have idiosyncratic preferences for one system over another (e.g., PECS versus SGDs) (van der Meer et al. 2012), and therefore the selection of any of these non-speech approaches to teaching social communication should in part be guided by student choice.

In addition to relying on teachers to instruct students on these new communication skills, some work points to advantages for peer-mediated strategies (using same age typically developing peers as tutors or models) (Kasari et al. 2012b; Kashinath 2012; Locke et al. 2012) as well as employing parents and siblings as instructors, as well (Dawson and Burner 2011).

Many approaches to teaching social skills use a variety of techniques that exploit the tendency of persons with ASD to learn skills better through visual rather than verbal cues. For example, variations of video modeling (including video self -modeling) show students with ASD how to behave and interact in a variety of social communication situations (Bellini and Akullian 2007; Buggey 2012; Charlop-Christy et al. 2000; Nikopoulos and Keenan 2004). Visually-based stories that students can use as cues in social situations (called social narratives) are often useful when teaching appropriate behavior in these scenarios (e.g., how to wait in line in the cafeteria) (Odom et al. 2010b; Test et al. 2011). It has only been relatively recently that this work has been subjected to rigorous testing through larger N studies and randomized clinical trials and evaluated with meta-analyses. More work is needed to identify factors related to better outcomes when teaching social communication skills (e.g., presence of comorbid intellectual disability) (Hume and Odom 2011; Maglione et al. 2012; National Professional Development Center on Autism Spectrum Disorders 2011).

In addition to the social communication and interaction difficulties characteristic of persons with ASD, they also display behaviors that are relatively rigid and/or unusual (Leekam et al. 2011). The repetitive behaviors of those with severe ASD such as stereotyped movements, repetitive manipulation of objects and some self -injurious behaviors still remain a challenge for intervention efforts and the data base for effective interventions remains limited (Boyd et al. 2012). The reason these behaviors may be more intractable is that they may involve internal processes related to sensory activities. There is research suggesting that some of these repetitive behaviors may later be used by these individuals to manipulate their environment (Durand and Carr 1987), thus serving a social communicative function (Durand 1990).

A significant evidence base exists for the treatment of challenging behaviors such as aggression, self -injury and other disruptive outbursts which are more frequent among persons with severe ASD (Carr 2011). As with interventions for social communicative behaviors, interventions for challenging behavior rely mainly on techniques derived from applied behavior analysis. The general strategy involves assessing the function of the behavior (called functional behavior assessment) (Beavers et al. 2013; Loman and Horner 2013) and using a variety of techniques to support appropriate behavior and teach alternative behaviors (Durand 2012; Vismara and Rogers 2010). In one randomized clinical trial, parents of children with developmental disabilities including ASD were taught how to; (1) assess the function of their child’s severe behavior problem (including aggression, tantrums and self-injury), (2) plan for emergency situations, (3) appropriately use consequences (e.g., praise) and, (4) teach their child alternative responses that serve the same function as the challenging behavior [called functional communication training; Durand 1990, 2012; Durand et al. 2013]. This clinic-based intervention demonstrated that parents could successfully intervene with their child’s severe behavior problems and reduce them in a meaningful way at home and in the community. A large number of small N studies and a growing number of larger demonstrations document the success of these procedures to significantly improve challenging behavior in multiple settings and across a variety of behavioral topographies (Carr 2011; Durand 2012; Vismara and Rogers 2010).

There are numerous comprehensive treatment programs being used with those having severe ASD and they typically are implemented either in schools or in special clinical settings (Howlin et al. 2009; Maglione et al. 2012; Odom et al. 2010a; Peters-Scheffer et al. 2011; Reichow and Wolery 2009; Rogers and Vismara 2008). Most of these programs also integrate parents or other caregivers into the program in order to extend teaching into the home and in the community. There is a general consensus that for treatment to be optimally successful it should be carried out for a minimum of 25 h per week and across all 12 months of the year (National Research Council 2001). This is just a general guideline and the specific interventions and their level of intensity are determined individually according to the needs of the person (Wilczynski et al. 2012).

The comprehensive programs with some empirical support can be broadly divided into three categories; (1) behavioral programs using applied behavior analysis techniques, (2) behavioral programs that integrate developmental considerations to guide treatment targets and (3) programs that have a relationship-based focus (Howlin et al. 2009; Maglione et al. 2012; Odom et al. 2010a; Reichow and Wolery 2009). In their review of 30 comprehensive program models, Odom and colleagues evaluated these models based on several criteria including thorough documentation of their procedures, data on the fidelity of implementation, outcome data and whether or not there were independent replications of the program (Odom et al. 2010a). Overall they found that most of the models had little evidence for their efficacy, and for those with published outcome data, most of the studies rated low on the quality of their data. It is important to note that research evaluating techniques and programs for students with ASD is evolving rapidly and several of these programs are currently being studied in recent and ongoing randomized clinical trials (e.g., Dawson et al. 2010).

Individuals with mild ASD do not have the cognitive delays and communication skills difficulties often found in persons with severe ASD, and can—with support—perform well academically in school. However, their social difficulties and common comorbid problems (e.g., ADHD, anxiety) complicate their interactions with peers and teachers and can lead to disruptive behavior problems. A number of different behaviorally-based programs exist to assist school-aged children with mild ASD improve skills such as appropriate social interaction, problem-solving, self -control, recognizing emotions in others, expanding their often narrow range of interests, and improving their understanding of non-literal idioms (e.g., understanding that the phrase “You are pulling my leg.” is not meant literally) (Karkhaneh et al. 2010; Koning et al. 2013; Laugeson et al. 2012; Rao et al. 2008).

Although this research base is emerging, one RCT used a comprehensive package to improve social skills in this population (Thomeer et al. 2012). This program used a 5 week summer camp for intensive work on a variety of social skills including social interactions , face-emotion recognition, interest expansion, and interpretation of non-literal language. A total of 17 children ranging in age from 7 to 12 years were randomly assigned to the treatment condition in groups with one staff member for every two children. Five 70 min treatment sessions were conducted each day, 5 days per week. The sessions focused on a target skill and adapted the social skills curriculum called Skillstreaming (A. P. Goldstein and McGinnis 1997). They found significant improvements in measures of knowledge of target social skills and understanding of idioms. On the other hand, there was not a significant improvement in facial-emotion recognition (Thomeer et al. 2012). Other programs specifically teach children how to make and maintain quality friendships (e.g., the Program for the Evaluation and Enrichment of Relational Skills—PEERS; Laugeson et al. 2012; Laugeson et al. 2009). As yet, the literature in this area remains in a nascent state and awaits more evidence for the effectiveness for these types of programs to make meaningful changes in the lives of persons with mild ASD.