Recovery and Prevention

Fig. 21.1

Hypothetical data describing the course of child development in typical children and in children with non-regressive autism. The lower rate of development in autism produces a gradually increasing overall amount of developmental delay as time passes.

Since the amount of developmental delay increases over time, it follows that one would want to intervene at the earliest possible time in order to have the least amount of deficit to remediate. For example, if one begins intervention at age two instead of three (on this hypothetical chart), then one has 12 months of delay to remediate, as opposed to 18 months if intervention begins at age three. Our concept of prevention is based on the possibility of beginning intervention before the degree of delay is sufficient to warrant an ASD diagnosis and, therefore when the size of the delay is far smaller than it is when intervention typically begins. Beginning intervention at a chronological age of 18 months is not unreasonable and has become increasingly common in EIBI , as will be discussed below. Beginning at this time would require remediating only 9 months of delay—that is, a hypothetical reduction of 50 % in the amount of remediation, compared to beginning intervention at 3 years of age. But if intervention can begin at 18 months of age, there is no reason to think it could not begin at 15 or 12 months of age—perhaps even 8 or 10 months of age .

It is also worth noting that intervening with particularly young children often entails working on more basic skills, which in some cases should be easier to teach. For example, the play skills that a 2-year-old must learn to “catch up” to her typically developing peers are relatively simple (e.g., parallel functional pretend play), versus those that are displayed by a typically developing 5-year-old (e.g., imaginary and sociodramatic play). In short, if one begins intervening when a child is a very young age, one has fewer and more basic skills to teach.

Another possible benefit of very early intervention is the potential for remediating skill deficits before a significant amount of challenging behavior has been learned and reinforced. It is commonly believed that a large portion of challenging behavior that is displayed by children with autism occurs because of a lack of other, more appropriate means of communication (Carr and Durand 1985). If the child is taught successful communication and social interaction skills very early on, then it may be possible to prevent the development of challenging behavior. This, in turn, would likely make treatment more efficient because there would be little or no need to spend the first several months decreasing challenging behavior, as is often done when intervention starts at 3 or 4 years of age .

The hypothetical data in Fig. 21.1 imply that preventing autism should be more efficient than treating it. Research on predictors of successful outcomes for children with autism has indicated that a younger age at intake is strongly correlated with better outcomes (Perry et al. 2011; Harris and Handleman 2000), and it seems reasonable that this same basic logic should stand for prevention, too. In this light, prevention is less of a categorical concept and is probably better conceptualized as a continuum, where the earlier one starts intervention, the less intervention is needed. If one starts it before the diagnosis is made and removes clinically significant impairment before the child is old enough to receive the diagnosis, then one has “prevented” autism. From a purely behavioral perspective, this is no different than behavioral intervention at any other age, except that there is less work to do and it is therefore likely to be more efficient. Additionally, providing early intervention prior to a diagnosis will result in reduction of the “red flag” symptoms commonly used to identify children who may qualify for a diagnosis. These “red flag” symptoms (e.g., poor eye contact, stereotypical patterns of play behavior, etc.) are behaviors that are subject to intervention and can be treated early.

It is important to note that the data depicted in the figure are hypothetical, and it is not known whether the true function describing developmental delay in autism is linear, nor is it known what degree of skill development is possible for any given child. And of course, every child with ASD is different, with some presumably being more severe than the hypothetical data depicted in the chart and some less severe. In addition, beginning EIBI at such a young age has not yet been evaluated in rigorous scientific research, so a large number of variables warrant discussion, several of which we address in detail below .

Research on Autism Prevention

No published studies of which we are aware have specifically set out to prevent autism via very early behavioral intervention . However, one case study described results of very early behavioral intervention for a toddler at high risk for autism. Green et al. (2002) reported the case of a little girl, Catherine, who received an “at risk” diagnosis of autism based on multiple screenings indicating communication and language delays, as well as stereotypic patterns of behavior. The parents of this child sought professional evaluations at the first signs of these delays because of their prior experience with their first child, who had received a diagnosis of autism and was receiving behavioral intervention. At the early age of 14 months, Catherine began an intensive in-home behavioral program consisting of 1:1 direct instruction for 25–36 h per week. At the age of 4 years and 5 months, Catherine completed her 1:1 in-home instruction but continued to have monthly follow-up observations in her preschool classroom. At the age of five, Catherine entered a general education kindergarten classroom without a diagnosis of autism, an individualized education plan (IEP), or a classroom aide, and she did not meet diagnostic criteria for an autism spectrum diagnosis. This case study represents a critical first step in research on using very early behavioral intervention for the prevention of autism, but much more research using valid experimental designs is still needed.

A highly controversial 2004 paper proposing a purely behavioral etiology of autism also contains a discussion of the prevention of autism (Drash and Tudor 2004). Space does not permit a discussion of the main thrust of the Drash and Tudor paper—that autism is caused solely by parent-child interactions—nor is one needed. It will suffice to say that such an idea is bordering on irresponsible in that it is essentially a return to the notion that autism is caused by bad parenting. We find this notion plainly absurd and not even worth discussing. The authors do specifically state that “Our analysis in no way attempts to blame parents” (p. 60), but it seems plainly obvious that denying the contribution of any factor other than parent behavior will carry that extremely negative implication for many parents .

Aside from the controversial aspects of the argument, the Drash and Tudor paper is one of the very few existing papers that discusses the possibility of preventing autism through very early behavioral intervention, and it makes an important practical point: Regardless of the genetic contributions to the etiology of autism, the only level at which we can intervene now or any time soon is the level of behavior—environment relations. The authors give specific recommendations for how this might be done. They propose that within the first 18 months of life or less, at-risk children should be identified, and parent–child interactions should be modified in order to encourage the development of adaptive forms of child communication and decrease avoidant behavior on the part of the child. Furthermore, they describe several case studies in which this was done and report that development for all the children was corrected in a much shorter period of time than is typically required of EIBI for children already diagnosed with autism. Like the Green case study, these were uncontrolled case reports, and further replication with sound experimental designs is still needed .

Identifying Who Should Receive Preventive Intervention

The first obvious roadblock to preventing autism via EIBI is simply detecting at a very young age which children will later be diagnosed with an ASD. Although warning signs can be observed at very young ages, the vast majority of diagnosticians are highly hesitant to provide an ASD diagnosis before 2–3 years because diagnostic evaluations were not validated with younger children (Crane and Winsler 2008). Research in this area has advanced significantly, but results still vary dramatically across studies. Retrospective studies have been published that reviewed home videos of typically developing children and children who later received a diagnosis of autism, indicating that some deficits can be observed as early as 4–6 months for motor anticipation (Brisson et al. 2012), and on average around 7 months for social attention, affective responsiveness, and prelinguistic vocalizations (Crane and Winsler 2008). Another retrospective video analysis has reported that some sensory-motor and social symptoms may categorize children later diagnosed with autism from those later diagnosed with developmental delays and children of typical development at 9–12 months (Baranek 1999). These symptoms include poor visual orientation/attention to nonsocial stimuli, prompted or delayed response to name, excessive mouthing of objects, and aversion to social touch. Similar results were found by Osterling et al. (2002) in a retrospective video analysis of 1-year-old infants later diagnosed with autism versus intellectual disabilities showing that the children who were later diagnosed with autism rarely looked at others or showed an orienting response when their names were called as compared to children later diagnosed with intellectual disabilities.

Although several autism screening tools exist, very few were developed and have been validated for children younger than 2 or 3 years old. One promising screening tool is the Checklist for Autism in Toddlers (CHAT; Baron-Cohen et al. 1992) which includes screening items for children as young as 18 months. The CHAT includes nine items asked to the caregivers and an additional five items that require direct observation in the home. To validate the accuracy of the CHAT, Baron-Cohen et al. (1992) administered the checklist to 41 children considered at high risk for autism based on genetic predisposition and determined that four children consistently failed items assessing gaze monitoring (e.g., looking in the direction of a caregiver’s gaze), protodeclarative pointing (e.g., pointing at objects to direct another’s gaze toward the object), and pretend play. These behaviors develop in typically developing children around 14 months of age and appear to be distinctively deficit among children with autism. All four children in the sample received a later diagnosis of autism, whereas the remaining 37 children did not fail more than one of these three items and none received a diagnosis of autism. Extending on this initial analysis, Baron-Cohen et al. (1996) administered the CHAT to 16,000 children in Great Britain, and 12 were identified as being at high risk for autism based on failing the three critical items from the initial analysis. Of the 12 children identified as at risk, 10 later received a diagnosis of autism and the remaining two were diagnosed with developmental delays, demonstrating that the CHAT was a fairly accurate indicator of autism at 18 months. A follow-up study investigating the total number of diagnoses on the autism spectrum from all 16,000 participants identified that 94 children had a diagnosis of either autism or PDD at 7 years. Using less stringent criteria to assess which items on the CHAT were most likely to indicate which children may have been identified at 18 months, only 38 % of the 50 cases would have been flagged at 18 months. Although the sensitivity is quite low, the specificity, or likelihood that the instrument will not falsely predict that a child has autism, is quite high (97.5 %) because, out of the identified cases, very few did not actually receive a diagnosis. The concern with using this instrument is that a child may not be flagged on the screening at 18 months, thus missing the opportunity to receive intervention until later, when greater deficits are apparent.

The requirement for direct observation necessary to administer the CHAT makes this screening tool less likely to be used despite its potential benefits of detecting early signs of autism for some children exhibiting symptoms at 18 months. Additionally, some behavioral deficits may not present during a single observation. Because of these potential limitations, a modified version of the CHAT was developed that can be conducted during normal visits to a family pediatrician by relying on parent report of current behaviors. The Modified Checklist for Autism in Toddlers (M-CHAT: Robins et al. 2001) includes 23 items that require a caregiver to provide a yes or no response. The authors validated the M-CHAT by administering it to 1,293 children between 18 and 30 months of age resulting in 58 cases identified as at risk for autism. A full evaluation was then conducted with the children identified as at risk and resulted in diagnoses of autism or pervasive developmental disorder (PDD) for 39 of these children. The remaining 19 children were found to have other language or global delays but did not meet full diagnostic criteria for autism or PDD NOS. Based on these findings, the M-CHAT may be a valuable screening tool that can be easily administered during a child’s 18-month pediatric visit, but follow-up research is necessary to determine the number of children who were not identified by the M-CHAT and later received a diagnosis. Although both the CHAT and M-CHAT are promising early screening tools, they are both designed for children at least 18 months of age. With research suggesting that red flag symptoms may be observed within the first year of life, there is a need for much earlier screening tools to identify children who may benefit from even earlier intervention and the possibility of preventive intervention.

Only gold members can continue reading. Log In or Register to continue