Autism : Current Issues 33


These summaries are linked simply by the fact that the original articles were found during the same literature trawl, so that a range of issues is covered.

The first section refers to attachment with the finding that children with autism can display positive attachment behaviours albeit developing over a longer time scale and showing some qualitative anomalies.

Temporal processing deficits are described, and the issue of shifting or disengaging attention is revisited (with an implication that the two behaviours are distinct).

Evidence is cited for anomalies among an autistic population in voice perception.

The final section refers to brief abstracts of presentations at recent conferences, and includes evidence about empathising and systemising, the significance of adult imitation of child behaviours for social progress, musical cognition, and the benefits of exercise for improving behaviour.


M.J.Connor                                                                                          September 2004


Autism and Attachment


The paper by Rutgers et al (2004) provides a review of evidence concerning the nature of attachment or attachment security in children with autism.


Their starting point was the recognition of the significance of early and secure attachment to a parent or caregiver (usually the mother) for positive child development, including the confidence to face new experiences, and to build (and to expect to build) mutually fulfilling personal relationships. 


Attachment, as described by various authors, including Cassidy and Shaver (1999), is seen as the affectional bond established between an infant and the mother, with associated attachment behaviour evoked in the infant when (s)he feels any sense of distress or threat. This takes the form of seeking to remain close to the mother until the anxious feelings have been dissipated and reassurance established, at which point the former play activity is resumed.

(The quality of an attachment relationship can be assessed experimentally in a procedure developed by Ainsworth et al [1978] where the infant is placed in a room with the mother or carer, and, in one instance, with a second and unfamiliar person.  The mother leaves the room, and the behaviour of the infant is observed, including the reaction when the mother returns after a short period of separation).


There is evidence that the child with secure attachment can use the attachment figure as a base from which to explore the environment, and her return is usually marked by an active quest for interaction and contact, and any (mild) distress is rapidly settled …. there is a balance between the attachment and exploratory behaviours.

The insecure-avoidant child shows minimal response to the leaving of the attachment figure, or to her return when the child may even avoid contact, preferring to concentrate on available toys.

The insecure-resistant child remains preoccupied with the attachment figure, and appears highly distressed by her leaving, and seeks contact on her return albeit uncertain how to respond to the renewed contact and not being comforted even after a prolonged period.


The relevance to autism is clear in that the earliest description of this condition (Kanner 1943) emphasised the failure to form emotional contact with other people.

This point has been consistently reinforced by, for example, Rutter (1978) who referred to the distinctive nature of the social and communication impairments which could not be explained in terms of a learning disability.

Children with ASD may show a failure to seek or enjoy physical contact, and an indifference to such approaches, alongside a lack of eye contact or of responsiveness to verbal or non-verbal communications.


Meanwhile, the first DSM differentiation between childhood schizophrenia and autism was based upon the characteristically autistic failure to develop normal attachment behaviour.


Nevertheless, the review by Rutgers et al (op.cit) highlights research evidence that children with autism and ASD can and do show attachment behaviours towards the attachment figure when they experience distress. 

For example, Buitelar (1995) showed that children with autism discriminate between parent and stranger and direct more attachment and social behaviours towards the parent than towards the stranger.  Further, children with autism are similar to normally-developing controls in respect of an increase in proximity-seeking behaviour when reunited with the parent or carer after a separation.


The problem is that the scoring system for the attachment assessment using the Strange Situation procedure is validated for normally developing children up to the age of 21 months, and existing studies involving children with autism have invariably used participants above 21 months in both chronological and mental age. There is also some reluctance to use the Strange Situation procedure with this population of children if they are already known to be easily upset by changes of routine and the loss, even temporary, of familiar figures from their immediate environment.


Accordingly, the authors set out to investigate whether children with autism are similar to controls in terms of establishing a secure attachment relationship; and whether children with the more loosely diagnosed pervasive developmental disorders have more secure attachments than those diagnosed as autistic.  There was also the question whether security of attachment has a positive correlation with mental development, and whether one might support existing views, such as those of Shah and Wing (1986), that the social deficit in autism is directly related to the level of general cognitive functioning.


The findings from their systematic literature search confirmed the impairment in (reciprocal) social interactions, but highlighted a number of studies which demonstrated clear attachment behaviours in children with autism, such as distress and an active search for the mother during a period of separation, and showing a strong preference for the mother over the unfamiliar person following the reunion.

On the other hand, there were studies which reported atypical behaviour among the children with autism in, for example, more contact with the unfamiliar person, the non-response at reunion, and the inability to be comforted by the mother through physical contact.


With regard to the quality of attachment, four studies were cited as indicating clear differences between the autistic and control samples in the security of attachment, as reflected in the lack of approach to the mother after reunion or the limited proximity behaviour and even avoidance of contact.

However, a further six studies were identified where the proportion of securely attached children in an autistic sample was not observably low or significantly lower than that observed in a control sample.


Meanwhile, there was found to be converging support for the view that there is a gradual shift from insecurity to stronger attachment with age, and the older the child, the greater the probability of a secure attachment.


The authors sought to summarise much of their observations by citing the availability of considerable evidence for attachment behaviours in children with autism; although there are also commonly identified impairments in responsiveness and contact seeking.


In their discussion, Rutgers et al restate the evidence that children with ASD (and PDD) do show signs of attachment insecurity despite the impairments in social interaction.  There may be some qualitative differences in behaviours, but diagnosis with ASD does not necessarily preclude the development of  a secure attachment relationship.  Reference was made to a composite figure of 53% of all the children with ASD or PDD studied who did show positive attachment.

Nevertheless, ASD or PDD is still associated with a probability of less secure attachments than common among normally-developing children. There appeared a trend that the autistic children with the lower mental development showed more signs of insecurity.


The authors acknowledge the difficulty of gaining clear conclusions because of the heterogeneity of the samples of children acting as participants in the various studies, but they cite the suggestion of Yirmiya and Sigman (2001) that the children with ASD show their weaknesses most clearly in any situation which requires a clarity of self image compared to an awareness of the other person …. ie a developed Theory of Mind. 

Compensation appears to be a matter of enhanced sensitivity towards the needs of the children, including initial signs and symptoms of anxiety and insecurity, with the possible implication that the promotion of secure attachment behaviour can provide a protective factor against a tendency to shy away from (social) experiences and a positive influence upon social development.


Temporal Processing and (High-Functioning) Autism


The study by Szelag et al (2004) begins with a recognition that autism is characterised by severe anomalies in social behaviours and by impairments in one or more areas of cognitive functioning.  Such areas include language or perception or memory, etc..


They go on to highlight the significance of sensory processes and the characteristic absent or unusual responses to sensory stimuli, whether auditory or visual.


However, they suggest that equally significant in the way of explanatory features of the observed autistic “style” is the range of deficits in attention which will inhibit the processing of information necessary for normal and purposeful behaviour.

Such deficits include a lack of spontaneous looking, a slower shifting of attention, limitations upon selective attention, and a low capacity for filtering stimuli or for controlling gaze.


In particular, they cite a number of authors, including Courchesne et al (1994), who suggest that problems in switching attention from one focus to another result in a fragmented view of the world and a failure consistently to link cause with effect. The net outcome may be a kind of photographic record with no real discrimination between the relevant and irrelevant details.


A further school of thought has it the most significant behavioural impairments in autism are a matter of dysfunction in memory processes. 

The work of Minshew and Goldstein (1993) suggests that autism may be thought to have some similarity to amnesic syndrome, and that the particular deficits observed among autistic individuals include poor free recall and poor recognition memory. 


Finally, reference is made to a common and yet further view held by various authors, including Siegel (1998), that autism is principally the outcome of a deficit in communicative ability or some severe developmental language weakness. 

While children with high functioning ASD may be seen to have basic linguistic competences, there are still a number of communicative deficits. 

For example, the content of the communications may be stereotyped or perseverative, failing to shift from a given topic, and the language output may be marked with odd vocabulary or use of phrases, and noted by an unusual prosody.  There may also be semantic-pragmatic weaknesses, echolalic tendencies, and poor use and understanding of complementary non-verbal information.


The current authors (Szelag et al) summarise their review findings by arguing that the existing evidence does not pinpoint a primary source of the various deficiencies associated with autism.


However, they continue, research on normally-developing participants has indicated that a common and essential component of all these cognitive functions is temporal information processing, with the possible implication that the constant theme in autistic deficits is a lack of temporal integration such that information gathered over time is not linked, and successive bits of information are not combined into a meaningful perceptual gestalt.  


To explore this hypothesis, the authors completed assessments on a number of children with autism (mean age 12-6) and control children matched for age and gender.   The IQs of the control children ranged from 95 to 145, and of the autistic sample from 82 to 102.


In the study, the participants were exposed to two types of stimuli … an auditory tone, or a visual stimulus projected onto a screen.  The procedure was that the child would be exposed to the sound or sight for a given duration, and the stimulus would then be presented a second time with the requirement upon the child to press a key and switch off the stimulus when (s)he considered that the presentation time had matched the original presentation time.


The results indicated that the control children matched what has come to be seen as typical in that they were quite accurate in matching duration times up to 2 or 3 seconds, but they tended to underestimate longer durations.


However, the autistic sample were severely impaired in their ability to reproduce the time intervals.


The authors sought to explain this weakness in terms of a number of features of time judgement, including arousal, motor skills, memory, attention, and emotion : and they cite evidence that, in autism, all or any of these functions could be deficient, leading to the deviant time estimates.


One hypothesis (Treisman et al 1994) suggests the existence of a  “cognitive pacemaker ”, the ability to produce pulses at a specific rate by which to mark and measure the passage of time. 

However, this facility may be influenced by a number of factors, such as arousal, so that estimates may become inappropriately longer or shorter than the actual interval.


An alternative hypothesis implicates inadequate memory functioning as the causal deficit behind inaccurate estimates of time intervals.


The autistic sample showed a consistent pattern in that estimates of short time durations were longer than the model, and estimates of long durations were shorter than the model.  

This was thought to reflect an “ attentional gate ” where the crucial factor is the amount of attention the participant allocates to the task.  Durations of stimuli that are not given much attention appear longer because less information is transmitted to memory. Where there is a longer time duration, the autistic participant may be relatively impatient, anxious for the moment when the stimulus will end; and, because of this inability to delay a response, the stimulus is switched off more quickly than is appropriate.


In any event, the authors conclude that they may still be unable to pinpoint the cognitive weaknesses that most significantly underlie autism, but they are able to argue that the limitations in temporal information processing are clearly evident among an autistic sample and may contribute to a number of these cognitive deficiencies. 


Shifting Attention


The piece of research by Landry and Bryson (2004) has looked further at the style of visual attention among young children with autism, noting that unusual gaze behaviour may be a clear diagnostic indicator.


They describe how parents, looking back over the child’s developmental progress, may often refer to an emptiness of gaze, and a lack of shared attention by which to engage in early “social” games; and specific evidence, such as that of Hermelin and O’Connor (1967), noted anomalous visual performance such as a failure to switch attention to and fro between two adjacent stimuli as is the pattern among normally developing children.  A kind of tunnel vision , an over focus upon a stimulus, has also frequently been described.


The authors summarise much work in this field by arguing that visual orienting is of particular interest given its perceived role in cognitive development and in regulating emotional states, and they make particular reference to the facility of disengaging attention as a basic means of regulating emotional distress.

This facility falls within the second part of a tripartite attentional system, which involves alertness and sustaining of attention, the moving of attention and selection of a further focus, and the mobilisation of (cognitive) efforts for some action.


The authors then cite existing studies about disengaging or shifting attention which commonly adopt the procedure devised by Posner (1988) where children are assessed for their speed of detecting and reacting to stimuli which are projected on each side of a fixation point.  The evidence has it that children with autism consistently have  difficulty with these attentional shifts, especially when the movement is in a right to left direction; and there is a suggestion that there may be two distinct weaknesses … a deficit in disengaging attention from an existing stimulus and a deficit in shifting the attention to another stimulus.


The question arises whether a deficit, or deficits, of this kind can only be observed among children with autism; and whether there is indeed a confounding of disengaging and shifting attention.


Accordingly, Landry and Bryson (op.cit) set up a study whereby participant children were to be placed in front of three computer screens. The child is presented with a visual image on the central screen, and then eye movements are monitored as a second stimulus is presented on one or other of the adjacent screens. A particular question concerned the effect upon child performance from leaving on or switching off the initial central stimulus.

The participants included 15 children with autism with a mean age of 5-6; 13 children with Down syndrome with a mean age of 5-5; and a control group of typically developing children with a mean age of 3-6.The three groups were matched for verbal and non-verbal mental ages, but the average IQ of the control group was higher given the younger chronological age.


Observation of the responses indicated that the children with autism had great difficulty in disengaging from one of two competing stimuli.  It was common for the children in this group to remain stuck on the first of the two stimuli for the whole of the entire 8-second period of a trial.

This was held to be similar to what is observed among much younger typically-developing children (aged 2 or 3 months) :  but, by 4 months, ease of disengagement is usually well established.


This disengagement deficit was observable to at least some degree in all the children with autism, with no relationship between disengagement ability and the level of mental age or language.


The authors note that this finding is discrepant from what has been found by with adult participants with autism who find disengagement to be much less of a problem.

The authors speculated that the discrepancy may reflect the different age groups involved, and that disengagement may be a particular problem at an early age but remits with time.  Secondly, the current task was specific in employing measures of overt attention in the form of eye movements.

Thirdly, it was pondered whether earlier work might have failed discriminate the disengaging and the shifting of attention.


In any event, the authors concluded that there was not only a striking impairment in disengagement among the autistic sample, but also a subtle deficit in shifting attention. Relative to both the Down sample and the control group, the children with autism were slow in their shift of attention when it did occur, leading to the further speculation that this may reflect a problem in the programming or implementation of eye movements.


Meanwhile, the observed ease of disengagement or shift among the Down sample led to some question whether the children were meaningfully engaged in the first place. 


By contrast, the children with autism performed in such a way as to support the view that autistic attention is characterised by an over-focus and a narrow breadth of field.


Voice Perception


On a linked theme, concerned with perception and processing, one notes the work of Gervais et al (2004) who produced evidence that the cognitive response to voices among individuals with autism is differentiable from that observed among non-autistic individuals.


These researchers used MRI procedures to examine brain functioning in 5 men with autism (mean age 26 years) and in 8 control men without autism (mean age 27) as they listened to vocal or non-vocal sounds.


The responses in the control group involved increased activation in a particular part of the brain (the superior temporal sulcus) when they listened to vocalisations.  This was held to support the view that humans have evolved a specific mechanism for processing voices.

However, in four of the five participants with autism, there was no difference in responses when listening to vocal sounds or non-vocal sounds. 


Meanwhile, the autistic participants recalled far fewer of the vocal sounds than the control participants, but performed similarly when it came to remembering non-vocal sounds … raising the question whether the anomalous brain response is the cause or consequence of the apparently limited regard for voices.


The authors suggested that findings of this kind are a complement to existing evidence that autism is associated with an impairment in the capacity to infer emotional meaning from voices, with the speculation that the lack of activation of the superior temporal sulcus in voice perception is part of the general dysfunctioning of the social neural network.


Extracts from the Proceedings of the British Psychological Society (2004)


The first of the items concerned with autism and ASD identified in the current set of conference contents is the presentation by Lawson et al who discussed empathising and systemising in Asperger Syndrome.

These authors studied a sample of adult males diagnosed with Asperger Syndrome, comparing them with a sample of non-autistic males, and a sample of non-autistic females.


Results of a task involving a demonstration of empathy indicated higher scores for the control females than for the control males, but with both these groups scoring higher than the Asperger group.

On a task involving systemising, the female group scored more poorly than both male groups, whose scores did not show significant differences.


The conclusion was that autism spectrum disorders do involve a deficit in empathising but a normal level of systemising ability. The authors also held that this finding is in line with the “ extreme male brain” view of autism that locates autistic spectrum disorders at an extreme point along a continuum of cognitive styles which can incorporate the population as a whole.


The presentation by Deeney concerned imitation (by parents/carers of the children’s actions) and the impact upon the social behaviour of preschool children with autism.
A video record was kept of the child’s performance during a play session when the adult was asked to imitate everything the child did, including movement and gestures and sounds.  A variety of toys was available, and play sessions of ten minutes each were alternated between “standard” interactions and those involving the close imitation.

Results indicated an increase in social behaviours and in receptiveness among the children … with support for the view that relationships involving children with autism are likely to be fostered when the adult joins the child’s world as opposed to expecting or assuming that the child will join the adult’s world (a strategy which is exploited more fully in the SonRise approach).


Musical cognition is explored by Heaton who highlighted the possibility of preserved or even enhanced abilities in certain areas among autistic individuals despite the characteristic triad of difficulties among such a population.  Musical abilities are a common case in point with, for example, evidence for absolute pitch or pitch memory among some (savant) individuals with autism.  However, there is no such enhancement of ability in tasks which require the integration of musical features (eg chords) into musical structures (eg melodies).

This finding may be explicable in terms of weak central coherence with its characteristic anomalies in local and global processing.

Meanwhile, it is found that the affective component of music may not be accessed by people with autism, although children can perform at musical emotion-identification tasks as accurately as non-autistic children.

The implications included the progression of autistic difficulties with age, with the possibility of building upon intact skills by early intervention with specific and structured activities, along with a further recognition of marked individual variation among individuals all diagnosed with autism.


Finally, one notes an unusual study by Golby and Donohue who described the limitations in success of strategies to treat autism using pharmacological, dietary, behavioural, or physical approaches. They examined the effectiveness of exercise upon the behaviour of an adolescent autistic boy, diagnosed at the severe end of the spectrum.  Following a baseline measure of behaviour, a 4-week exercise intervention was implemented.  Three exercise sessions per week were set, for up to 20 minute each, using a static cycle. Subsequent behaviours were monitored and rated, with preliminary evidence suggesting the link between this regime and a reduction in the frequency or intensity of maladaptive behaviours … albeit with an acknowledgement of possible weaknesses from using a single case design.


                                       *          *          *          *          *                                              M.J.Connor                                                                                        September 2004




Ainsworth M., Blehar M., Waters E., and Wall S.  1978   Patterns of Attachment.  Hillsdale, New Jersey : Erlbaum Publishing


Buitelar J.  1995   Attachment and social withdrawal in autism.   Behaviour  132  319-350


Cassidy J. and Shaver P.  1999   Handbook of Attachment Theory.  New York : Guilford Publications


Courchesne E., Townsend I., Akshoomoff N., et al  1994   Impairment in shifting attention in autistic and cerebellar patients.   Behavioural Neuroscience  108  848-865


Deeney C.  2004   Does imitation increase the social behaviour s of children with autism ?   Presentation to the Scottish Branch Undergraduate Conference, University of Strathclyde.  March 2004


Gervais H., Belin B., Boddaert N. et al  2004   Abnormal cortical voice processing in autism.   Nature Neuroscience  7  801-802


Golby J. and Donohue C.  2004   The effects of exercise on the behaviours of an autistic adolescent male.  Presentation to the BPS Annual Conference, Imperial College.  April 2004


Heaton P.  2004   Musical cognition in autism.   Award for outstanding doctoral research : presentation at the BPS Annual Conference, Imperial College.  April 2004


Hermelin B. and O’Connor N.  1967   Perceptual and motor discrimination in psychotic and normal children.   Journal of Genetic Psychology  110   117-125


Kanner L.  1943   Autistic disturbances of affective contact.   Nervous Child  2  217-250


Landry R. and Bryson S.  2004   Impaired disengagement of attention in young children with autism.   Journal of Child Psychology and Psychiatry  45(6)  1115-1122


Lawson J., Baron-Cohen S., and Wheelwright S.   Empathising and sytemising in adults with and without Asperger’s syndrome.   Presentation to the Cognitive Section Annual Conference, University of Reading  September 2003


Minshew N. and Goldstein G.  1993   Is autism an amnesic disorder ?   Neuropsychology  7  209-216


Posner M.  1988   Structures and functions of selective attention.  In Boll T. and Bryant B. (Eds)  Clinical Neuropsychology and Brain Function.   Washington DC : American Psychological Association.


Rutgers A., Bakermans-Franenburg M., van Ijzendoorn M., and Berckelaer-Onnes I.  2004   Autism and attachment : a meta-analytic review.   Journal of Child Psychology and Psychiatry  45(6)  1123-1134


Rutter M.  1978   Diagnosis and Definition.  In Rutter M. and Schopler E. (Eds)  Autism : A Reappraisal of Concepts and Treatment.  New York : Plenum Press


Shah A. and Wing L.  1986   Cognitive impairments affecting social behaviour in autism.  In Schopler E. and Mesibov G. (Eds)   Social Behaviour and Autism.   New York : Plenum Press


Siegel D.  1998   Evaluation of high functioning autism.   In Goldstein G., Nussbaum P., and Beers S.  (Eds)  Neuropsychology.   New York : Plenum Press


Szelag E., Kowalska J., Galkowski T., and Poppel E.  2004   Temporal processing deficits in high functioning children with autism.   British Journal of Psychology  95  269-282


Treisman M.  Cook N., Naish P., and MacCrone J.  1994   The internal clock.   The Quarterly Journal of Experimental Psychology  47A  241-289 


Yirmiya C. and Sigman M.  2001   Attachment in Autism.  In Richer J. and Coates S.  (Eds)  Autism : Putting Together the Pieces.   London : Kingsley Publishing  



M.J.Connor                                                                                          September 2004


This article is reproduced by kind permission of the author.

© Mike Connor 2004.

Back to NAS Surrey Branch Welcome Page