Autism and ASD : More Evidence about their Nature and Symptomatology


The topics covered here include sleep patterns among children with autism or ASD and among adults with autism and learning difficulties; sensory integration and perceptual experience; motor symptoms as a possible means of differentiating high functioning autism from Asperger Syndrome; and family environment (expressed emotion) and symptoms in adolescents and adults with autism.


M.J.Connor                                                                                           December 2006



Sleep Patterns


In the introduction to their research findings, Allik et al (2006) describe how existing and converging evidence has demonstrated that disordered sleep is common among individuals of all ages with pervasive disorders such as (high functioning) autism and Asperger Syndrome (HFA and AS) 

However, it is their view that there is little systematic information about the pattern of such disorders among children.


Their initial review of studies highlights the common finding that many children or adolescents with HFA or AS show co-occurring symptoms of behavioural or emotional problems including sleep abnormalities.

The effect of interrupted sleep may involve not only day-time drowsiness (and, presumably, reduced efficiency of functioning and inattentiveness) with an impact upon learning but also behavioural difficulties ….. as well as having a disruptive effect upon the sleep of other family members.


Some researchers (eg Richdale and Prior 1995) have suggested that characteristic behaviours and symptoms of HFA or AS may be directly attributable to disturbed sleep; and a study by Segawa et al (1992) showed that reducing sleep disturbance and helping children to improve the length and quality of sleep could be linked to improved social relatedness and to reduced insistence upon sameness.


Meanwhile, existing evidence shows some inconsistency in respect of observed sleep patterns among children with high functioning and lower functioning autism.  Some evidence suggests that sleep disturbance is greater or more common in the latter group, and other evidence suggests that it is the former group who are more affected.

However, there is consistency in the evidence which has shown sleep disturbance to be more common among children with autism or ASD compared to normally- developing control groups of children with the implication that sleep problems are related to some particular deficit in autism rather than to the level of intellectual functioning per se.


The authors go on to cite further evidence that treatment designed to reduce problems with the onset, maintenance, and duration of  sleep (such as by means of the administration of melatonin ….. Paavonen et al 2003)  does have positive effects upon the day time behaviour and general manner of children with AS.


It is recognised that many school-age children with no recognised developmental difficulties can show a shift in their sleep patterns such as a less regular sleep/wake cycle and a delay in the inset of sleep at weekends compared to during the school week. 

However, the current authors have identified no existing information about children with HFA or AS in respect of any variations in sleep patterns during these different parts of the week. 


Accordingly, their own study, involving samples of children in the age range 8-5 to 12-8 years with AS or HFA and age and gender matched controls, set out to gain further evidence whether disordered sleep should be seen as a characteristic part of childhood HFA and/or AS, with separate analyses of data concerning sleep variables on school days and at weekends. Separate analyses were also conduced to determine whether sleep patterns differed between children with HFA and those with AS, as well as between the combined target sample and the control sample.


The results showed generally greater difficulties in falling asleep, lower sleep efficiency, and lower ratings of the children’s sleep made by parents among the target children with HFA or AS compared to the control children.

No differences were found between children with HFA and children with AS in respect of sleep patterns.

However, while parental reports commonly indicated that the onset of sleep was a particular problem, with these children going to bed earlier and spending a longer time in bed before falling asleep, and that sleep efficiency or quality were poor, objective measures of actual sleep start and sleep end (using actigraphs) did not differentiate between the children in the target and control groups on school days.


The finding that many parents of children with HFA/AS reported sleep problems, especially with regard to sleep onset, is in line with existing evidence, although the current evidence deviates from what has gone before in noting that the sleep latency problems were mainly present on school days.


It was speculated that the longer sleep latency among the target group reflected a tendency for these children to be put to bed earlier than their counterparts and before they were significantly sleepy. 

There was the further speculation that this could reflect a coping mechanism by parents who experience some stress or fatigue in managing their children with HFA or AS and who seek maximal respite in the evenings.  Alternatively, the earlier bedtimes could reflect good parenting techniques in applying a consistent routine with a view to establishing consistent sleep patterns.

It might also simply be that the children in question showed sleepiness at their normal bedtimes but required longer time in bed before the initiation of sleep …. possibly attributable to the higher anxiety and physiological arousal associated with autistic spectrum disorders.


It was noted that children in the target and control groups showed greater similarities in the times at which they fell asleep and woke up on weekdays, with the implication that the children were comparable in their ability to adapt to school-day schedules and routines.  Meanwhile, both groups went to bed and woke up later at weekends.


Maintenance and duration of sleep (compared to reported onset latency and quality) did not differ significantly between the groups in this study, and this is not in accord with some of the existing data.  It is possible that this reflects differences in the methodologies used or in the samples of participants, with Allik et al acknowledging that their selected sample from clinically-referred children may not be representative of children with HFA or AS generally, and that parents of children with observed sleep problems might have been the more willing to take part in the study. 


The authors concluded with a summary of their findings ….. viz ….

Diary, actigraph, and parental data all showed that children with HFA/AS have difficulties in falling asleep, possibly linked to earlier bedtimes.  The target and control groups did not differ in respect of timing of sleep on school days, nor in respect of sleep duration, but there were differences in that the HFA/AS group showed this longer sleep latency as well as poorer sleep efficiency and parent-reported sleep quality.


The implication drawn concerned the need to continue research with a view to exploring the possible mechanisms underlying the longer sleep latency, including the effects of anxiety, arousal, and day-time behaviours.  


Still on the question of circadian rhythms, albeit with a focus upon adult subjects with both intellectual disabilities and autistic spectrum disorders, Hare et al (2006) note the frequency in the literature with which ASD has been linked with some disturbance in the sleep/wake cycle.


In the introduction to their own research, they cite findings that sleep problems, such as night-time waking, are observedly common in typically-developing children up to 2 years of age when there may be a prevalence rate as high as 20%, but this prevalence declines to around 4% by the age of 8 years.

In contrast, they cite evidence from Richdale and Prior (op cit) that anything up to 60-80% of children with autism exhibit sleep disorders by the age of 4 years; and further evidence from Hoshino et al (1984) that such disturbances are likely to be severe, particularly when the children concerned also show relatively severe intellectual disability (ID).

It has been argued by various authors that reduced day-time activity levels indicative of sleep problems, have also been associated either directly or indirectly with challenging behaviours among both typically-developing children and those with ASD or ASD plus ID.


Part of the problem may be a matter of deficiencies in melatonin production, and an abnormally high level of day-time melatonin levels in children and young people with autism. 


One view (Boucher 1999) held that impairments in the functioning of the “biological clock “may be the fundamental cause of key impairments in autism. 


While there is little firm evidence by which to support a proposal which goes this far, there has been a consistency of observation that abnormal sleep/wake cycles are implicated causally in the development of the anxiety and restlessness commonly linked to ASD. 

The authors also refer to the evidence for improvements in general day to day functioning and reduced reaction to change and to social demands following intervention to improve sleep onset and quality. 


Caution is counselled when assessing such findings or in drawing causal rather than correlational implications given the evidence from Hering et al (1999) that sleep functioning measured objectively by actigraph revealed little differences between children with ASD and controls (apart from earlier waking among the ASD group), even if parents consistently reported the children with ASD as having sleep disorders such as disturbed sleep and frequent waking. 


(As with the Allik et al study, one might suggest that the difference between groups may be less in the duration of sleep as measured by actigraph, but in the quality of sleep such that ASD and non-ASD individuals can still be differentiated in terms of sleep variables.)


Hare et al continue by noting how research into sleep patterns and cycles has placed heavy reliance upon questionnaires or diaries completed by parents, and they are concerned about the risk of inaccuracy or error. Their own study emphasised the use of actigraphs by which to gain a measure of sleep and activity based upon continuously recorded bodily movements over an extended time period.


The participants were 31 adults with ID living in supported residential settings, of whom 14 had also been diagnosed with ASD. 


The overall results from the analysis of actigraph data indicated no significant differences in sleep patterns and circadian rhythm functioning between those participants with ASD and those without. 


The finding that no dysfunction of this kind appears to be associated with ASD per se conflict with the above reported study findings. 

However, Hare et al themselves seek to address the inconsistency by referring to a study by Takase et al (1988) which found that older children with an ASD showed fewer sleep problems than younger children with the same disorder.  The suggested implication is that there is a developmental process at work whereby disturbances in sleep patterns may be common among young children, but the circadian anomalies remit by the time the individuals reach late adolescence or adulthood.


On the other hand, it is recognised that the findings could reflect the circumstances of this particular sample in living in specialist centres with a need for clear routines and a greater probability of experience of psychotropic medication which will influence all these individuals with ID regardless of whether they are also diagnosed with an  ASD.


(One might also suggest that it may not be entirely reasonable to seek consistency among samples of individuals with ASD and samples of individuals with significant ID and ASD.  The previously summarised study did note observable differences between higher functioning and lower functioning individuals with autism/ASD.)


In addition, the authors pondered whether their results could have been influenced by the presence of individuals in the non-autistic group who had unidentified ASD, thus bringing about a lack of between-group differences; and the potentially confounding       

factors relating to health (including epilepsy) and use of medication were acknowledged. 


Nevertheless, the conclusion held that it is important to identify sleep disturbances among individuals with ID and or ASD given the likely significance for their own well being and for that of their carers, just as it is important to assess the presence and level of any such disorders among children and young people given the likely impact upon day time activity, and educational or social functioning.


(In other words, there is the possibility that the link between anxiety and/or disturbed/disturbing behaviour and sleep disorders is bi-directional.  Anxiety and associated behavioural symptoms may have some causal significance for disturbed sleep patterns, but impairment of sleep or poor quality of sleep may have causal significance for observable symptoms of anxiety and behavioural disturbance.

This line of thought highlights again the variability of the nature and number and level of symptoms among individuals all legitimately identified with autism or ASD, with the corresponding need to gather assessment data from observations in various settings and occasions and from ratings carried out by different informants.)


Sensory Integration and Perceptual Anomalies


Iarocci and McDonald (2006) begin their review by noting that the earliest, and continuingly significant, theories of autism concern anomalous processing of sensory information.

Such anomalies have included, for example, a focus upon parts at the expense of a scanning of the whole stimulus or an avoidance of (high levels of) stimulation via any modality.


It is still the case that theories about the cause or nature of autism reflect the perceived significance of sensory idiosyncrasies as core elements of autism with an impact upon the perceptual capacities of the individuals concerned (which, in turn, will impact upon social and peer interaction skills and motives).


Their first theme is about the status of evidence concerning sensory profiles among individuals with autism.


It is commonly the case that parents of infants with autism report sensory anomalies at a very early developmental stage.  These might include either an intense preoccupation with certain stimuli, or a stressful reaction or wish to avoid others, while remaining unresponsive to what might be considered more meaningful stimuli such as the sound of their own names.

These atypical sensory and perceptual behaviours are persistent over the long term, and are observed in the absence of any actual deficits in sight or hearing; and one commonly held hypothesis links these atypical behaviours to attempts to manage the sensory environment (which may be either over-whelming or under-stimulating) by either creating or avoiding sensory events.


One practical problem described by the authors is that evidence may emerge largely from parental questionnaires, and its generalisability is questionable given the small sample sizes, the heterogeneity of the children, and the lack of data relating to a typically-developing control group of children.


They cite one study where the authors did compare 3-6 year old children diagnosed with autism or pervasive developmental disorders with age-matched controls by means of parental ratings, and were able to differentiate the groups in terms of 8 out of 10 of the elements of the Sensory Profile, including sensory seeking, emotional reactivity, inattention, etc.  

However, while these data were considered somewhat more meaningful because of the direct matching of groups, there were still problems of marked variability within the autism/PDD group and a lack of mental age matching.


A further study (Rogers et al 2003) did include mental and actual age matching, together with additional control groups of children with Fragile X Syndrome and general developmental delay. 

Significant differences among groups were observed in the number of sensory symptoms reported on the Sensory Profile  (such as tactile sensitivity, under-reactivity, or stimulus seeking, etc.), although the overall number of sensory symptoms reported among the children with autism did not differ from those reported among the children with Fragile X Syndrome even if the permutations were not the same.

The problem here was that the observed findings were not correlated to the level and nature of the symptoms of autism (as set out in diagnostic schedules) such that meaningful links between sensory information and autistic status could not be made.


The conclusion, therefore, indicated the risk of confounding actual sensory impairment with idiosyncratic sensory behaviour whose meaning and purpose may be different among individuals with autism compared to other people.


Turning to self reports, Iarocci and McDonald describe how a small numbers of high-functioning and verbal adults with autism have provided autobiographical accounts including references to sensory experiences.


Typically such reports refer to problems with the input and the processing (making sense) of incoming sensory information across all or any sensory modalities; sensory distortions; overload; synesthesia (a stimulus in one modality, eg a sound, evoking an image in another modality, eg a colour or a smell) : and difficulty in processing information impinging simultaneously in different modalities.


First hand accounts of sensory disturbances were analysed by Jones et al (2003) who found 4 basic response clusters :


The problem here is that autobiographical accounts provide only one source of information, and the experiences and reports of any given individual may change over time, may not be true of other people, and may even confound self and others’ memories for events and feelings. 


Additional sources of information are the home videos which contain a record of the actions and behaviours of young children who are later diagnosed with autism.  Evidence has been gleaned for delayed sensori-motor development, or hypotonia, or a lack of social attention and social smiling, or the lack of different facial expressions.


Further studies have shown that early behaviour among children with autism include a lack of orientation to their own name; and, compared to children with Down Syndrome, more mouthing of objects, poor visual attention, and aversion to being touched. 


The current authors believe that video-tape records can provide useful information. However, they feel it necessary to take account of the variability of the data sources such as the range of settings and contexts, difficulties in specifying the children’s ages, observer rating bias, and the mental age of the child in order to be clear that the behaviours are atypical rather than a matter of developmental delay. 


With regard to early behavioural studies, sensory and perceptual and attention anomalies were among the first issues to attract research attention, and one initial aetiological theory related sensory abnormalities to a persistent state of over-arousal as a result of a failure in those systems whose activities modulate arousal level. 

Motor stereotypies and restricted activities were thought to reflect attempts to bring about this modulation .... but, under experimental conditions enabling the amount of sensory input to be varied, these repetitive behaviours were not found reliably to be linked with increased arousal.


Meanwhile, Hermelin and O’Connor (1970) had found support for the view that children with autism could be differentiated from typically-developing or learning disabled controls in their behavioural and physiological response to visual and auditory stimuli … such as a preference for “proximal”sense receptors like taste or touch or smell over more distant receptors such as sight and hearing.


The problem with work in this field, according to Iarocci and McDonald, is the failure to define constructs unambiguously, or to specify sample characteristics, with implications for limited validity and reliability of findings. 


More recent theories include the significance of weak central coherence.  Children with autism have difficulties in integrating information across a range of contexts

(perception, attention, linguistic, and semantic) so that higher-level meaning is not attached to the various elements of input.


One particular hypothesis holds that structural abnormalities in the cerebellum underlie deficits in the systems which govern attention, with a particular weakness in the capacity to shift attention from one task to another where there needs to be a corresponding shift within the relevant modality (from visual to auditory, for example).


An alternative view has it that there is a wider neurological problem involving, for example, a deficit in executive functioning and the coordination of different sources of information, or in connectivity between different brain areas sharing responsibility for language processing.


The various theories may differ with regard to the nature of the problem (structural or functional), the sensory modality affected, the complexity of the deficits, and the processes concerned, such as neural connectivity.  However, they all support the view of atypical sensory processing as a core feature of autism. 


The authors then turn to an attempt to “operationalise” the concept of sensory integration. 


They begin by restating that meaningful perception develops from the more basic capacities of selective attention and of integrating various sources of input.  Among typically-developing individuals, this merging of information is of such an automaticity and fluency that the observer is not aware of the process involved despite the involvement of different brain areas.


This multi-sensory processing is seen by the authors as valuable construct by which to seek an understanding of the perceptual and sensory experiences and anomalies relevant to individuals with autism.  They note, in any event, that many of the current conceptualisations of autism concern weaknesses in processing and integrating input  as illustrated by the weak central coherence hypothesis, or problems in shifting attention, or inadequate connectivity.


They go on to describe how sensory signals are translated at receptor organs into neural impulses which are transmitted along modality-specific pathways to subcortical and cortical structures.   At the cortex, processing continues along specialised pathways that are still largely modality-specific. 

However, beyond these pathways, several brain areas, not designated for the processing of stimuli within any single sensory modality, may be involved.  These association areas have the role of integrating information from different sensory systems.


Various inputs converge upon the deeper layers of the sensory cortex, and many neurons located here respond to stimulation in multiple modalities; and evidence has shown that the neural responses to multi-sensory stimuli are of greater intensity than the sums of responses to individual stimuli. 

Under typical circumstances, multi-sensory stimulation leads to enhanced perception of, and enhanced responses to, objects in the environment.


This process may be evident from many everyday situations.  The example is given of watching television when one hears sounds that come from speakers alongside the screen, but the sounds are perceived as if they come from the people or objects on the screen.  Similarly, when watching a ventriloquist, it is known that the speech is coming from the person working the dummy, but the tendency is to perceive the words as coming from the dummy’s mouth. 


This kind of effect has been manipulated in laboratory trials where, for example, the experimenter presents simultaneous auditory and visual stimuli at different locations and observes reactions to these spatial differences in respect of where the individual locates the stimuli.  The individual is asked, for example, to point to the apparent location of one stimulus and ignore the other.

The pattern emerging from such studies indicates an inter-sensory bias in that vision dominates the multi-sensory interactions in spatial perception.

In other words, visual input largely determines the perceived location of a stimulus when there is also auditory or proprioceptive input; while auditory and proprioceptive information only slightly modifies the visually-perceived location of a stimulus.


Where there are discrepancies in sensory inputs, the resolution occurs in terms of the information provided by the most precise or appropriate modality.  Vision provides the most reliable spatial information while the auditory modality supplies the most precise temporal information.

To highlight the significance of integrated multi-sensory input, the example is given of face to face conversation when each person will have access to the sounds of the speaker’s voice and the sight of the speaker’s expression, body language, and lip and  mouth movements.   Most listeners process these latter articulatory movements much more than is commonly realised, so that the presence of “congruent” information can increase comprehension of speech especially in busy environments (such as a classroom).

Another example is that of the McGurk Effect where, for example, incongruent information is deliberately introduced by dubbing a different consonant-vowel sound onto a visual clip of someone articulating the original consonant-vowel sound.  The McGurk effect is seen when the incongruent visual and auditory information are combined and the observer reports either hearing what the visual information would convey or a compromise sound which is somewhere between what is indicated by the different visual and auditory information.


The authors return to the theme of autism by highlighting the apparent complexity and variety of ways in which multi-sensory processing may be achieved via the above-described form of convergence, or feedback from higher-processing areas to lower-level areas, and via networks of neural connections. 


Patterns of performance, such as perceptual anomalies observed among individuals with autism, may arise from deficits or inefficiencies in the integration and organisation of processing systems rather than simply from some impairment in individual components. 

For example, such individuals can apply global processing to some stimulus object when global processing is all that is required, but will be at a disadvantage in dealing with some stimulus which requires both global and local processing. The autistic response is likely to emphasise local processing, with attention paid to component details, and the overall view of the stimulus (the general “gist”) remains unclear.  


The authors conclude by suggesting that the challenge for ongoing research is to explore weak central coherence or atypical neural connectivity, etc., in the context of these multi-sensory and integrative needs and issues.   The goal is to gain some shared  meaning of the great range of the behavioural indicators of sensory abnormalities in autism at various ages and various levels of functioning.  Such abnormalities may occur at specific stages in development or in particular combinations and provide links to early clinical markers of autism (such as lack of gaze monitoring); and the implication involves the “borrowing”of theories and methods from cognitive neuroscience to explore multisensory processing in order to move towards a comprehensive theory of perception in autism.


Motor Signs and High Functioning Autism or Asperger Syndrome        


The work by Jansiewicz et al (2006) set out to gain further information about the range of subtle neuromotor signs indicative of autism which can act as diagnostic differentiators between children with autism and non-autistic peers.


They note that motor impairments have been widely reported in ASD, including poor coordination, awkward gait, and limited motor planning in both children and adults. 

There will be direct impacts from fine motor difficulties in respect of day to day demands such as in writing or changing clothes at any speed, and in coping with social play, such as riding a bike, or catching a ball, and sharing team games. 

They also note that motor functioning is considerable among children with autism such that they can be clearly differentiated from children experiencing other developmental or psychiatric disorders in this respect.


Some existing studies cited by these authors have attempted to use differences in motor functioning to distinguish among the heterogeneous manifestations of autistic spectrum disorder. 

For example, the review completed by Ghaziuddin et al (1992) found that around 50% of existing research studies found clumsy and uncoordinated movement patterns among children with Asperger Syndrome (AS). 


Some studies sought to determine whether or not the AS children have greater motor impairment than other sub-groups of chidren, such as those with high functioning autism (HFA) and evidence is cited that young people in both diagnostic groups are likely to show problems with running at any speed, agility, balance, bilateral coordination, limb strength and coordination, visual motor control, and dexterity.


Further evidence suggests that the motor difficulties are greater among autistic children towards the lower levels of functioning (as reflected in cognitive scores) than among children with average cognitive scores.  

One theory has it that the movement or gait anomalies commonly observed among children with autism are similar to the symptoms observed in Parkinson cases, with the possible implication that the motor impairments could be a reflection of some specific deficit in the functioning of the basal ganglia part of the motor system.

Another theory takes a contrary view and associates the difficulties with cerebellar anomalies.


However, Jansiewicz et al hold that there has been a lack of research into the full range of subtle neurological signs by which to explore where problems exist in the neurological systems; and their own study set out to provide further information. 


95 boys and male adolescents participated.  40 had a diagnosis of ASD (HFA or AS) and 55 were control participants with no diagnosed disorder.  The age range was from 6 to 17 years with a mean of 11 and standard deviation of 2.5. 


Their results showed that the HFA and AS participants had impaired performance on a wide variety of measures in a motor assessment relative to a control sample of age matched individuals. 

The target group showed greater difficulty with balance and gait, slower speed and more dysrhythmia in tasks involving timed movements of the hands and feet, thus reinforcing the findings from existing studies and extending the deficits to include problems with timed actions and with “overflow” (the extension of the hand during heel walking; or additional oro-facial or mirror movements during timed actions) as observed through the use of the PANESS (Physical and Neurological Examination for Subtle Signs).


The authors recognise that some caution is to be applied to the results given small sample size and other questions concerning the diagnostic validity of the PANESS. Nevertheless, the general pattern of results was taken as supporting the relevance of motor abnormalities in the clinical presentation of autism, and the appropriateness of investigating motor functioning among children with autism and ASD.


In respect of distinguishing HFA and AS, there has been some reference to the prevalence of clumsiness in cases of AS.  The findings by these current authors do not support this kind of differentiation since no significant differences were found between HFA and AS sub-samples on the dependent variables of the PANESS, and, instead, support those researchers who have quoted no motor differences between these diagnostic groups. 


One tangential finding was that those autistic children prescribed stimulant or other psychoactive medication demonstrated greater awkwardness of gait compared to other autistic children peers taking no medication, but no other motor differences. 

The authors speculated that continued use of stimulants could lead to residual gait disturbance; or that, more probably, children prescribed such medication have pre-existing motor disturbances associated with symptoms of impulsiveness, hyperactivity, or inattention.  (Motor difficulties are commonly identified in children with ADHD.)


The authors went on to compare the gait of autistic children not taking medication with that of control children.

A significant difference was observed between the groups suggesting that gait disturbance is associated with the autistic spectrum independently of medication effects.  Further analysis showed that a diagnosis of autism or ASD significantly predicted gait performance while the use of stimulant medication carried no such predictive significance.

An assessment was made of any difficulties that could be associated with the use of other forms of medication, but no differences were observed between children who were and who were not receiving the medication.  The conclusion held, therefore, that motor impairments observed among children with autism are not secondary to the effects of medication. 


With regard to the neural basis for the observed motor abnormalities, the authors held that the wide-ranging nature of difficulties observed during their study is indicative of some developmental anomaly or dysfunction in the systems responsible for motor activity including both frontal and sub-cortical regions, such that both the cerebellum and the basal ganglia could be implicated.

It is also considered probable in the light of MRI studies indicative of greater variety of “functional maps”and less distinct regional activation among individuals with autism compared to non-autistic controls that motor pathways are less well organised among those with autism leading to problems in generating motor responses. 


Their final speculative point is that the observed deficits are indicative of impaired acquisition of basic motor skills; and that this impairment in early motor learning underlies ongoing and greater difficulties over time in the gaining of advanced skills and in taking advantage of (motor) experience …. as illustrated by studies showing that samples of children with autism tend to display a consistent and slow reaction time in motor learning trials and no observable improvements with time and practice.


Expressed Emotion and Behavioural Problems


Moving away from neurological issues to environmental issues, the work of Greenberg et al (2006) explored how the family setting and style can impact upon the expression of the core deficits of autism.


This is not to return to that discredited view of autism as being somehow caused by inappropriate parenting strategies, when autism is properly to be seen as a complex disorder with possibly multiple causal pathways.  Nevertheless, it is reasonable to seek links between the nature or level of observed behaviours characteristic of autism and the nature of the familial environment or atmosphere. 


In particular, the authors refer to “expressed emotion“ which refers to the emotional style within the family, and includes the (spectrum) dimensions of warmth, positive comments, emotional over-involvement, criticism, and hostility.

Existing evidence cited by these authors (such as that of Beck et al 2004) indicates that mothers of a child with a developmental disability and of at least one other child without a disability showed higher levels of expressed emotion (in whatever dimension or permutation of dimensions) regarding the child with the disability than the other child(ren), and that the child with disabilities had more behaviour problems.


The purpose of the study by Greenberg et al was to investigate bi-directional effects between expressed emotion and (maladaptive) behaviours and symptoms in families of individuals with ASD.

The hypothesis held that an environment involving high levels of expressed emotion is likely to be experienced as stressful to any individual but especially to someone with autism given the existing difficulty in regulating stress responses. 


They recognise that it is not clear what mechanism underlies the association between expressed emotion and an escalation in symptoms.

One hypothesis relates to a heightened (genetic) vulnerability or sensitivity among people with autism to stress, with a high level of expressed emotion perceived as a form of psychosocial stress.  This view is supported by evidence that a reduction in expressed emotion is followed by a reduction in symptoms among persons with major psychosocial disorders.

On the other hand, there is a view that expressed emotion may also be a reaction by other family members to high levels of symptoms and behavioural difficulties in the individual with the disability.


The authors comment that unravelling the interactions will be particularly complex in the case of parents of children with autism because of the major care-giving challenges that they constantly face, such as feeling they must monitor behaviours and anticipate their child’s needs given the problems with direct communication.


The authors also suggest that where the child displays highly stereotyped and ritualistic patterns of behaviour, obsessional behaviour or resistance to change, and an insensitivity to the feelings or actions of others, expressions of frustration are understandable.

This may be particularly so if the parents feel that constant and close supervision is necessary to keep the children safe such that the autistic behaviours elicit a highly involved form of parental behaviour.


In other words, there is already an indication that the relationship between expressed emotion and symptoms is bi-directional.


In their own study involving adolescents and adults with autism, Greenberg et al controlled for age and gender and any co-morbid diagnosis of learning disability.

Their thinking was influenced by findings that family stress is greater during the  adolescence of the autistic children with some abatement of behaviour problems with age; that mother-child interactions vary according to the gender of the child; and that comorbid learning disability commonly involves greater behavioural problems and more limitations in communication and social skills which might confound the relationship between expressed emotion and other (dependent) variables.

Their hypothesis was that the individuals with autism living in family environments characterised by high (maternal) expressed emotion would display increasingly severe behaviour problems and symptoms of autism compared to individuals living in families with low expressed emotion. 


Their findings produced strong support for the hypothesis that high levels of expressed emotion would have negative effects upon the well-being of individuals with autism.  This high expressed emotion was linked to greater intensity of internalising and externalising and asocial maladaptive behaviours over time.  Similar albeit weaker effects on the symptoms of autism, such as repetitive behaviours, were noted with consequent intensification of impairments in reciprocal social impairments.


Where the expressed emotion took the form of a high level of criticism, there was a corresponding increase in internalising and asocial behaviours, and repetitive behaviours, over time. 

Their findings were also said to concur with those of Mink and Nihira (1987) who were among the first to examine the bi-directional nature of these parent-child influences. This early study found that the direction of influence was a function of the type of family.  Influence flowed from parents to child in highly cohesive families, and from child to parents in control-oriented and child-oriented families.


Greenberg et al are at pains to point out that, even if their study was concerned with problems of a kind associated with expressed emotion, the majority of the families in their sample did not display elevated levels of expressed emotion. The levels observed were actually lower than those reported among families of children with a psychiatric disorder, which was thought possibly to indicate a pattern of family strength and effective coping with the stresses that can be associated with autism.


The authors also speculated that the low observed prevalence of expressed emotion may be linked to attributions made about the control the children are able to exert over their behaviour and symptoms.  Higher expressed emotion is likely in those cases where the children with a disability are seen as having control over the symptoms, whereas the low level in this sample of families with an autistic child reflects the view that the behavioural symptoms are attributed to factors beyond the child’s control.


There was some indication that the characteristics of the individual with autism would influence maternal expressed emotion, but influence in this direction was weaker.  Meanwhile, mothers tended to have greater expressed emotion, reflected in over-involvement, with their daughters than with their sons; but over-involvement tended to be lower with a comorbid diagnosis of learning disability.  This latter finding was thought to reflect the greater involvement with support services and a consequent sharing of responsibility.


Expressed emotion tended also to be lower when the child’s autistic characteristics were more of the internalising type, including withdrawal, inattention, and self injury …. perhaps reflecting a pattern of some mutual emotional disengagement.  This issue is thought to warrant further investigation in the light of the high prevalence of depression among individuals with autism. 


Further, stability of expressed emotion was observed, suggestive of a specific personality trait; although other findings have suggested that it is a matter of reactivity to events, notably fluctuations in behaviour problems.  Yet other findings suggest that it could be both a trait and a state !


The authors acknowledge some limitations to their study such as the use of volunteer families which may have introduced some bias, and the 18 month time span assessed in the study which could be regarded as a very short part of the life of mother-child interactions.

Nevertheless, the authors confirm that autism compared to other types of developmental disability involves heightened levels of potential or actual family stress.

Further, they note the benefits of early intervention, as well as the link between gradually abating symptoms and low levels of expressed emotion particularly in the form of criticism and over-vigilance, with implications for the value of early interventions which include ongoing support for the parents in their handling of, and response to, the issues which arise.


                                  *          *          *          *          *          *


M.J.Connor                                                                                           December 2006








Allik H., Larsson J-O., and Smedje H.  2006   Sleep patterns of school-age children with Asperger Syndrome or high functioning autism.   Journal of Autism and Developmental Disorders  36  585-595


Beck A., Daley D., and Hastings R.  2004   Mothers’ expressed emotion towards children with and without intellectual disabilities.   Journal of Intellectual Disability Research  48  628-638


Boucher J.  1999   Time parsing, normal language acquisition, and language related developmental disorders.  In Perkins M. and Howard S. (Eds)  New Directions in Research into Language Development.  London : Plenum Press


Ghaziuddin M., Tsai L., and Ghaziuddin N.  1992   A reappraisal of clumsiness as a diagnostic feature of Asperger Syndrome.   Journal of Autism and Developmental Disorders  22  651-656


Greenberg J., Seltzer M., Hong J., and Orsmond G.  2006   Bidirectional effects of expressed emotion and behaviour problems and symptoms in adolescents and adults with autism.   American Journal on Mental Retardation  111(4)  229-249


Hare D., Jones S., and Evershed K.  2006   Objective investigation of the sleep-wake cycle in adults with intellectual disabilities and autistic spectrum disorders.  Journal of Intellectual Disability Research  50(10)  701-710


Hering E., Epstein R., and Elroy S.  1999   Sleep patterns in autistic children.  Journal of Autism and Developmental Disorders  29  143-147 


Hermelin B. and O’Connor N.  1970   Psychological Experiements with Autistic Children.   Oxford : Pergamon Press  


Hoshino Y., Watanabe H., Yashimo Y., Kaneka M., and Kumashiro H.  1984   An investigation on sleep disturbance of autistic children.  Folia Psychiatrica et Neurologica  38  45-52


Iarocci G. and McDonald J.  2006   Sensory integration and the perceptual experience of persons with autism.   Journal of Autism and Developmental Disorders  36(1)  77-90


Jansiewicz E., Goldberg M., Newschaffer C., Denckla M., Landa R., and Mostofsky S.  2006   Motor signs distinguish children with high functioning autism and Asperger Syndrome from controls.   Journal of Autism and Developmental Disorders  36  613-621  


Jones R., Quigney C. and Huws J.   2003   First hand accounts of sensory perceptual experiences in autism.  Journal of Intellectual and Developmental Disability  28  112-121


Mink I. and Nihira K.  1987   Direction of effects; family life styles and behaviour of TMR children.   American Journal on Mental Retardation  92  57-64  


Paavonen E., Nieminen-von Wendt T., Vanhal R., Aronen E., and von Wendt L.  2003   Effectiveness of melatonin in the treatment of sleep disturbances in children with Asperger disorder.   Journal of Child and Adolescent Psychopharmacology  13(1)  83-95   


Richdale A. and Prior M.  1995   The sleep/wake rhythm in children with autism.  European Child and Adolescent Psychiatry  4(3)  175-186


Rogers S., Hepburn S., and Wehner E.  2003   Parent reports of sensory symptoms in toddlers with autism and those with other developmental disorders.   Journal of Autism and Developmental Disorders  33  631-642


Segawa M., Katoh M., Katoh J., and Nomura Y.  1992   Early modulation of sleep parameters and its importance in later behaviour.   Brain Dysfunction  5  211-223


Takase M., Tiara M., and Sakasi H.  1988   Sleep disorders in children with autism.  Psychiatry and Clinical Neurosciences  52  181-182


This article is reproduced by kind permission of the author.

© Mike Connor 2006.

Back to NAS Surrey Branch Welcome Page