AUTISM : CURRENT ISSUES 27
This set of notes begins with a further illustration of the marked increase in incidence of autism over the last 10 or 15 years which cannot be explained by changes in diagnostic practices or criteria or by population "migration".
There is then a section concerned with predictors of outcomes, with implications for the significance of early language skills, and for the differentiability of autism from Asperger syndrome.
An early intervention, with a focus upon joint attention training, is described, with the conclusion that enhancing this skill may have far-reaching and positive implications in some or many cases for social awareness and social and communication skills generally ... albeit with the need still to work towards early identification and to recognise the variability of response among children with autism to any given approach.
The final section describes the current agenda of the All Party Parliamentary Group on Autism which is concerned with the issue of decision making on the part of individuals who might lack the necessary cognitive capacity to understand or to select from different options.... while ensuring that lack of capacity is not confounded with a lack of opportunity to communicate.
M.J.Connor June 2003
As a confirmation of the observed trend whereby the rates of (diagnosed) autism are increasing, one notes the recent study carried out in California which described a doubling of cases of autism over the last 4 years.
This study was organised by the California Department of Developmental Services (study author : Dr. Ronald Huff) and involved the tracking of cases of diagnosed cases of autism referred to 21 regional centres where children and their families receive specialist services.
It was reported that the 97% increase in caseload during the 4 years beginning December 1998 overshadows the rises of between 35% and 49% in new cases of mental retardation, cerebral palsy, and epilepsy ( ...... although the current writer - MJC - would ask whether these latter rises might not also be seen as striking, and whether they were consistent with reasonable expectations and prevalence trends ??).
These figures relating to autism were released during May 2003, and the authors commented that the increase does seem real, especially as the data follow up and reinforce the pattern observed between 1987 and 1998 of a sharp rise in autism cases across the state. Reference was made to the typical pattern during the 1970s and mid-1980s of around 200 new cases of autism each year, in contrast to new cases counted in thousands over the next decade.
In a parallel study carried out at the Neuropsychiatric Unit of the University of California (study author : Dr. Marian Sigman), it was reported that the dramatic increase in the cases of autism across the state were not due to changes in population or reporting criteria. It was not a matter of families moving to California because of the perceived value of the services offered, nor of any shift or watering-down of diagnostic criteria.
The question that remains is why there is this sharp upturn in the rates of autism (and the research team at the Department of Development Services refer to preliminary findings whereby heightened levels of a naturally occurring protein have been found more frequently than chance in the blood of neonates who subsequently are diagnosed with autism. This pattern is being investigated with findings to be reported in 2006.)
Predicting Outcomes in High Functioning Children with Autism or Asperger Syndrome
The recent research completed by Szatmari et al (2003) starts by highlighting how the long term outcome of autism is largely clear ... ie, that impairments appear to extend into adulthood although there may be significant improvements in autistic behaviours, and an increase in social and communicative performance over time (see, for example, Nordin and Gillberg 1998).
The two variables most commonly identified as early predictors are IQ and language. Children with IQs in the higher part of the range and those who have developed functional language before the age of 6 years have a better prognosis than children without these characteristics.
However, the authors point out that a problem with these studies is that the children with autism were not all identified and studied at the same and very early stage (an "inception" sample), with the possibility that those children who showed a very rapid improvement or deterioration were not included within study samples. Accordingly, results will suffer from a systematic bias and the validity of predictor variables will be undermined.
Other methodological issues include the emphasis in existing studies upon children with autism whose functioning is towards the lower end of the spectrum with implications for problems in generalising findings to higher functioning children.
It is reported by Szatmari et al, for example, that they have not identified any research findings concerning the predictive ability of IQ and language performance in children with Asperger syndrome.
They have shown in a previous study of their own (Szatmari et al 2000) that children with Asperger syndrome have a better 2-year outcome that children with autism, but cannot determine whether IQ and language are equally important predictor variables for this group.
They also ponder whether, given the better language abilities in Asperger syndrome, performance level in language would be as effective in predicting outcomes as it is in autism.
There is also the question whether IQ and language predict several aspects of outcome separately (such as moderation of autistic symptoms, and social and communication skills); and it is relevant to note existing evidence that adaptive behaviour, in respect of socialisation or daily living skills, and autistic symptoms, such as lack of eye contact or echolalia or rituals, do seem to represent two independent dimensions in pervasive developmental disorder.
Whether or not the predictive power of these two variables remains the same over a long time scale, or apply only to the short term, is also not known; and there is, finally, the fundamental question whether IQ and language are independent predictors of outcome since children who are the most verbal tend to have the higher IQs.
In their current study, therefore, Szatmari et al sought to gain an assessment of the extent to which measures of language skills and nonverbal skills, taken in an inception sample of young high functioning children with autism and Asperger syndrome, predict outcome two years later and six years later.
Participant children were drawn from all those being assessed or receiving therapy at one of six specialist centres in Ontario.
During the enrolment period, the authors excluded children without a full and formal diagnosis of pervasive developmental disorder, those whose non-compliance precluded valid assessment, and those who fell below the IQ criterion (68 on the Leiter Scale or 70 on the revised Stanford-Binet Scale); and the resulting sample comprised 68 higher functioning children of whom 47 met the criteria for autism and 21 for Asperger syndrome.
Mean age at inception was 65 months for the autistic sub-sample, with a range of 48 to 87 months; and 68 months for the Asperger sub-sample, with a range of 52 to 80 months.
The assessment instruments included the Autism Behaviour Checklist and Vineland Adaptive Behaviour Scales (to be completed via an interview with the parents); Autism Diagnostic Interview; Leiter Performance Scales; Beery Visual-Motor Integration Test; Stanford-Binet Intelligence Scale; Test of Language Development; Comprehension Test from the Reynell Language Scales; and the McCarthy Oral Vocabulary Test. (If required, references to the authors/publishers of these tests can be obtained from the present writer - MJC).
Following initial assessments, the families were contacted after about 2 years (mean of 26 months) and again after about 6 years (mean of 78 months) in order that follow up assessments could be implemented and outcomes explored. The outcome measures focused upon parental assessment of social and communication skills as reflected in the Vineland Adaptive Behaviour Scales, and a composite picture of autistic symptoms as reflected by the Autism Behaviour Checklist (ABC).
The results highlighted the marked variation among individuals, but the generally low Vineland but high ABC scores were held to be consistent with the kinds of impairments and range of symptoms seen in children with high functioning pervasive developmental disorder.
However, it was found that the predictor variables continued to be strongly correlated with the outcome measures at both follow-up assessment occasions, suggesting that the significance of the predictor variables was maintained over time. The predictive ability was strongest for the communication outcome, but weaker for the autistic symptoms' outcome.
In their discussion, the authors noted that their results support findings from other studies that early language and nonverbal skills are important predictors of outcome in adaptive behaviour in communication and socialisation and, to a lesser extent, in autistic symptoms. Further, the predictive ability appears stable over time for these outcomes.
However, some qualifications were raised. Firstly, there appears to be much overlap among predictor variables, and, secondly, outcome is not a unitary construct .... ie... one may predict outcome fairly well in terms of communication skills, but less well in terms of social skills and even less well still when it comes to autistic symptoms.
Meanwhile, the strength of the prediction differed among the children with autism and those with Asperger syndrome for all three outcomes on at least one of the follow-up assessments; and the language factor was found to be a better predictor of outcome for children with autism than it was for children with Asperger syndrome on both follow-up assessments. Thus, it was held that the determinants of outcome in children with autism are not the same as those in the children with Asperger syndrome.
As far as practical implications are concerned, the authors suggested that this distinction between autism and Asperger syndrome may be useful in highlighting how working on language skills may improve outcomes more dramatically in children with autism; but it is also noted that nonverbal skills do predict outcome in autistic symptoms among children with pervasive developmental disorder generally and among those with Asperger syndrome in particular.
What underlies this predictive power of nonverbal skills is not clear, but it was speculated that early nonverbal skills may be linked with executive function abilities at a later age; or that higher scores on nonverbal tests give an indication of better attentional skills which, in turn, allow the children in question to take more benefit from early intervention strategies.
In any event, there is support for the emphasis in early interventions on language skills, as long as it is recognised that a focus on language alone may not have an impact on social skills or autistic symptoms, especially among the children with Asperger syndrome. "Outcome" is a multi-facetted concept, and an improvement in one area does not predict a similar amount of improvement in another area.
The authors also argue that the finding of differentiability between children with autism and those with Asperger syndrome in terms of predicting outcomes suggests that it is simplistic to regard Asperger syndrome as a mild form of autism. Rather, the two conditions represent different albeit overlapping developmental courses which are influenced by similar but not identical prognostic factors.
This thinking would suggest that children with pervasive developmental disorders are not easily differentiated in early infancy, but those who develop meaningful language by 3 years of age (and come to be diagnosed with Asperger syndrome) follow a different developmental pathway compared to those children who remain non-verbal at that age. Hence, the predictive ability of language in the Asperger group is limited compared to the autistic group. The children with autism may join the developmental pathway followed by the Asperger children as they develop fluent language, but this becomes less probable as the delay in language development becomes longer term.
The children who do not develop fluent language fall further behind the rest of the cohort over time even if they appeared higher functioning and to have adequate nonverbal skills at inception; and the authors conclude that the aims of continuing early intervention studies include an identification of the mechanisms for moving children from one developmental pathway to the more efficient pathway (or even away from the pervasive developmental disorder routes altogether).
Joint Attention Training for Children with Autism
The work by Whalen and Schreibman (2003) may provide an illustration of one early intervention by which to enhance the developmental opportunities and outcomes among children with autism.
They restate how autism is characterised by profound deficits in social behaviour, language, imitation, and play skills; but they note how joint attention deficits have come to be seen as one potential underlying link among these developmental anomalies.
Joint attention is defined as " the ability to coordinate attention between an object and a person in a social context " (Adamson and McArthur 1995) and has been associated with the fluent acquisition of language or play or social skills in normally-developing children.
Given that autism may not be diagnosed until the child is 3 or 4 years old, it is very important that research identifies diagnostic indicators prior to language emergence so that appropriate intervention may be initiated at a younger age ... and joint attention may be significant as a very early social behaviour, and deficits in this aspect may be apparent before language acquisition.
Studies on the capacity of children with autism to respond correctly to "protodeclarative" joint attention stimuli from others (ie the use of gestures to suggest sharing some object rather than simply to request the object) have shown that children with autism may not readily follow the speaker's gaze and may make some incorrect link between the word and the reference object.
Similarly, there is much evidence to show that protodeclarative initiations, such as pointing and showing, are absent or markedly limited among children with autism.
Specific work cited by the present authors (such as that of Mundy 1995) has suggested that protodeclarative initiating and responding may have similar social functions but are maintained by different motivational factors.
It may be that responding to joint attention overtures is maintained by some extrinsic and tangible reward, while initiating joint attention is a function of intrinsic motivation.
It is noted that, while early interventions have shown success in dealing with deficits in language or social behaviours, etc., there has been little study of the means of enhancing joint attention. However, it would seem logical to help children with autism to respond to joint attention bids; and, subsequently, to work towards their initiation of joint attention.
The authors describe how Discrete Trial Training (DTT) and Pivotal Response Training (PRT) have both been used effectively for teaching children with autism, but they have not previously been used to work on joint attention.
(DTT involves breaking a skill into the smallest component parts and using mass trials and high levels of reward to teach the steps until the full skill is mastered.
PRT emphasises child motivation by offering choices and reinforcing attempts while practising and consolidating previously mastered skills. An example is that of learning colours where a blue car is shown the child and if the child provides the correct response in saying "blue", (s)he is allowed to have the car to play with for a time. This is contrasted with DTT where a child might be shown a flashcard with a blue shape on it, and saying "blue" leads to some reward, such as something to eat, which is not directly related to the task or situation).
In their own study, Whalen and Schreibman worked with a sample of children with autism matched by normally developing children (although the chronological ages of the autistic group were higher - around 4 years - than that of the control group - between 2 and 4 years - in order to equate social behavioural level).
The children were given a baseline assessment which included standardised measures of intelligence and language, and ratings on the Childhood Autism Rating Scale, and Gilliam Autism Rating Scale. There were also behavioural assessments to gain a measure of the children's ability to respond to joint attention stimuli, the generalisation of such behaviour to different settings, and social behaviour such as turn-taking.
Joint attention training involved components from DTT and PRT involving clear prompts, interspersing already mastered tasks with unlearned tasks, allowing the child to select the target activity, modelling the desired behaviours, contingent reinforcement after the child's response whether prompted or unprompted, and direct response-reward linking (so that, for example, if a child followed the adult's gaze to a given object, the child would be allowed to play with that object as a reward).
Joint attention training included such actions as placing the child's hand on a different toy while (s)he was playing with a first toy and observing whether there was any manipulation of, or looking at, the second toy; subsequent training followed a similar protocol except that the second toy would simply be tapped , or simply shown to the child. Eye contact was reinforced along with following the adult's shift of gaze.
Initiation training involved, for example, observing whether the child shifted gaze from a toy to the adult (as an indicator of willingness to share) for which there would be a reward; and prompting was used in terms of holding the child's hands on the toy and gently moving the child's head so that eye contact was made. Similarly, various toys or pictures were provided as opportunities for pointing, with prompts offered as required, and access to the toys pointed at as reward.
Results gained from these studies were consistent with existing data in that children with autism showed some ability to respond correctly to the joint attention bids of adults prior to training, but this ability was shown to be limited when compared to the performance of normally developing children. Further, time alone did not bring about improvements in the skills of responding to, or initiating, joint attention.
However, the children with autism all made significant gains in responding to joint attention stimuli following the initial phase of training; and the timing of the changes after a period of simple observation without direct intervention supported the conclusion that it was the training rather than developmental maturation that brought about the positive progress.
The changes were also found to generalise to different and unstructured settings.
The authors held that teaching children to respond to joint attention may be important for increasing their social awareness. By teaching a child to respond to what someone else is doing in a social-communication context, the child is also taught that other people have social intentions and that a response is expected. It may also provide the means of teaching children with autism to shift their attention from one object to another (thus reducing the attention-shift deficit which is regarded as possibly underlying many of the social weaknesses observed in these children).
The autistic sample, with one exception, also showed a positive change from the initially severe limitation in coordinated attention and joint attention initiations as a result of the training. These improvements, too, were shown to generalise to other settings or individuals, including the parents with whom the child spends the majority of the waking day.
The lack of improvement in the one child was attributed to the low cognitive capacity and to the very limited language performance observed at baseline assessment, although it was acknowledged that the critical factor might have been the lack of efficacy in what was used as reinforcers .... highlighting again the issue of marked individual variation among all the children described as autistic in terms of their own characteristics and their response to intervention.
However, decreases in some of the skills, such as coordinated gaze shifting, were noted between the post-intervention assessment and the longer term follow up after 3 months. The implication suggested the need for "maintenance" trials, ie continuing to practise the skills once mastered, and ensuring that parents can be trained in means of maintaining the behaviours once learned. It was also speculated that the decrease might be attributable to inadequate intrinsic motivation, especially as the training involved extrinsic rewards (ie being allowed to play with a toy).
The authors also commented that significant changes in the child's joint attention skills emerged once the children were taught to initiate joint attention, perhaps indicating that responding to others' joint attention bids can be maintained by a relatively simple stimulus-response relationship, while the capacity to initiate joint attention is a true precursor of social awareness.
What also mattered was the finding that "naive" observers of the children reported that, after the attention training, the children appeared more ordinary and that behaviour was less idiosyncratic, reinforcing the view that joint attention deficits may be an important part of the reason why many children with autism are described as aloof or withdrawn (and, in such ways, different from their normally developing peers).
The authors concluded that joint attention skills can be taught, and that the logical next step would be that of observing the behaviour of the child with autism in his or her natural setting to determine the situations when, or the individuals with whom (parents or siblings or peers or teachers), the child is most likely to adopt joint attention behaviours, with a view to sharing the training procedures with a range of others while continuing to highlight with greater accuracy those children who are likely to benefit the most from such approaches.
All Party Parliamentary Group on Autism
It is interesting to report on the ongoing deliberations of this group (and to record its existence) which is dealing with the issue of decision making and mental capacity.
The briefing dated February/March 2003 noted that there is no universally accepted definition of mental capacity. Meanwhile people with autism cover a wide range of competences and capacities, and their ability to make decisions may depend on a variety of factors such as the precise setting and their current mood.
Further, there is no clear and consistent legal guidance about the decision-making process as it involves people who are thought to lack the cognitive scope to understand the implications.
This places family members, carers, and support professionals in an uncomfortable position in wanting to act in the individual's best interests but in being uncertain what the law permits them to do.
The example is given of situations where medication can be administered to a person without his or her direct consent; but, largely, there are no guidelines concerning what is permitted to ensure that an individual gets necessary treatment, leading to professionals' anxiety over the possibility of being accused of inappropriate practice.
With regard to autism, reference is made to the key principle whereby all adults have the right to make their own decisions and are assumed to have the ability to do so, unless it is proven otherwise. Some people with autism may need help to be able to understand the issues about which they are expected to make a decision, but the need for help does not remove their right to decide.
There is an acknowledgement that it is not possible to set down principles which will provide for all individuals with autism given the breadth of types and levels of difficulties; however, it is recognised that communication difficulties may be a common and central issue in ensuring that an individual understands a situation and can make clear his or her response.
It might appear as if the individuals cannot make a choice when, in fact, they are quite capable if given the means and opportunity to express themselves, and a trusting relationship within which to do so.
The critical matters involve presenting information in a way that the individual can understand, and offering support or enhanced time to ensure adequate processing of the information.
In their representation to the All Party Group, a spokesperson for the Making Decisions Alliance (a group of organisations representing people who may need support in this area) held that legislation is required to define what precisely is meant by " best interests ", and who decides what in which circumstances, thus to protect both the vulnerable individuals and their carers.
* * * * * *
M.J.Connor June 2003
Adamson L. and McArthur D. 1995 Joint attention, affect, and culture. In C. Moore and P. Dunham (Eds) Joint Attention: Its Origins and role in Development. Hillsdale : New Jersey
Mundy P. 1995 Joint attention and socio-emotional approach behaviour in children with autism. Development and Psychopathology 7 63-82
Nordin V. and Gillberg C. 1998 The long term course of autistic disorders. Acta
Psychiatrica Scandinavica 97 99-108
Szatmari P., Bryson S., Streiner D., Wilson F., Archer L, and Ryerse C. 2000 Two-year outcome of pre-school children with autism or Asperger syndrome. American Journal of Psychiatry 1576 1980-1987
Szatmari P., Bryson S., Boyle M., Streiner D., and Duku E. 2003 Predictors of outcome among high functioning children with autism and Asperger syndrome. Journal of Child Psychology and Psychiatry 44(4) 520-528
Whalen C. and Schreibman L. 2003 Joint attention training for children with autism using behaviour modification principles. Journal of Child Psychology and Psychiatry 44(3) 456-468
© Mike Connor 2003.
Back to NAS Surrey Branch Welcome Page