This paper offers some information on the prevalence of autism across a population as a whole, indicating the appropriateness of the spectrum or continuum concept.

Reference is then made to the continuing issue of head size as an early diagnostic indicator, and to the possibly greater than chance association of Duchenne Muscular Dystrophy with Autism.

The final section is concerned with the impact of autistic symptoms upon the parents/ carers/teachers, where the major source of stress seems to be located within externalising behaviours.


M.J.Connor                                                                                                March 2006



The Prevalence of Autism


The survey completed by Posserud et al (2006) begins with references to Kanner’s original (1943) description of autism and to Asperger’s (1944) description of autistic psychopathology which was later to be designated as Asperger Syndrome.

It remains subject to debate whether Autism and Asperger Syndrome can be validly and meaningfully differentiated into two separate conditions, and a compromise view is that both may be included within a wider autistic spectrum since, despite some variation in the clinical presentations, both involve weaknesses in the triad areas of social functioning, communication, and imagination.


The authors go on to highlight the marked increase over recent years in the numbers of children diagnosed with Autism, albeit with the probability that this reflects as much as anything ongoing changes in diagnostic criteria, a heightened awareness of autistic symptomatology, and a widening of the concept of autism to include the  spectrum disorders.


However, the question is raised whether individuals with autism/ASD are outside the developmental parameters relevant to the population as a whole, or if they should be seen as occupying the lower extreme of a continuous distribution of (autistic) social and behavioural traits into which all individuals can be fitted. 

There are few existing data about the prevalence of autism across whole age groups

(as opposed to data about the number of formally-diagnosed cases); but what studies there have been – such as that of Constantino and Todd (2003) – have indicated that it is quite common for individuals drawn from non-clinical samples to show some autistic symptoms and 1 or 2% experience significant numbers of symptoms.


Meanwhile, reviews and epidemiological studies consistently demonstrate that more males than females are diagnosed with autism to a ratio of 3 or 4 to 1.  It appears that boys are more susceptible (and one hypothesis holds autism to be a reflection of an  extreme form of “male brained-ness”); but it remains possible that girls are under-diagnosed if autistic traits are somehow more concealed within female behavioural “styles” which tend to be more social and if special or idiosyncratic interests are less marked.

In any event, in the general population, too, there are more males than females who show some signs of autism.


Agreement between parents and teachers in respect of identifying or rating behavioural signs and symptoms tends to be moderate at best and can be quite low, possibly because there are genuine differences in behaviour in the different settings of home and school, or differences between the informants in their attributions of significance to the behaviours and in the direct impact upon themselves of those behaviours. 


The aim of the study by Posserud et al themselves was to gain an estimate of the distribution of autistic features in a total population of children using the Autistic Spectrum Screening Questionnaire (Ehlers et al 1999); and to examine the influence of sex and gender in the parents and teachers who completed the ratings, the extent of agreement between ratings, and the representativeness of the children concerned via a comparison of teacher ratings made on children where there was parental consent for the inclusion of such information in the general study and the (anonymous) ratings made on children where there was no such parental consent.


The ASSQ was included within the screening aspect of the Bergen Child Study which is an ongoing longitudinal whole-population examination of all the 9430 children attending 2nd to 4th grade classes (7-9 year olds) in all state and private mainstream and specialist schools during 2002.  The aim is the maintenance of reliable prevalence data for a range of mental health problems in a total population, with particular reference to risk factors, protective factors, and comorbidity.

This questionnaire has 27 items to be rated on a 3-point scale (0 - not true, 1 - somewhat true, 2 - certainly true) so that higher scores indicate a higher degree of autistic symptomatology.  A score of 21 points on the teacher version and 18 points on the parent version have been deemed high and a cut-off point for a significant risk of autism.

The teacher form was completed for 9155 children (97%), and 7007 parents agreed to participation in the study of whom 6691 (71%) completed their questionnaires


Access was also gained to data emerging from the use of the Strengths and Difficulties Questionnaire (Goodman 1999) and reference was made to the DSM-IV criteria for Autism and for Oppositional Defiant Disorder.

Information was obtained via interviews about learning difficulties, obsessive and compulsive behaviours, tics, eating and sleeping habits, health, selective mutism, hypoactivity, and the extent of use made of specialist services.


The results confirmed that, for the majority of children, there were no reports of autistic symptoms.

Where symptoms were reported, as indicated by the increasing questionnaire scores. There was no clear demarcation point between non-autistic and (possibly) autistic cases; instead, there was a smooth transition between the numbers showing no symptoms, through those showing some symptoms, and those showing many symptoms.

Boys obtained higher scores than girls in both the teacher and parent ratings, but the difference was greater in the teacher reports.  In respect of high scoring boys and girls, the ratio in the parental reports was around 2:1; in the teacher reports 5:1; and in the anonymous reports completed by teachers 7-8 : 1.


The overall correlation between teacher and parent scores was .48 for boys and .34 for girls.  High agreement was observed when identifying high scoring boys, but a little lower when identifying high scoring girls.

Meanwhile, despite the recommended cut-off points of 21 or 18, the authors observed a plateau point in both parent and teacher ratings of 15, and this was taken as a rough indicator of the start of higher risk status.  Examining the combined scores from parents and teachers and the teacher scores for the anonymous group, it was estimated that a total of 5+% of children in this total population of 7 to 9 year olds would be high scoring on parent and/or teacher ratings.


In their discussion and summary, the authors emphasised that the distribution of autistic symptoms was continuous, but with a lengthy tail of high scorers. 

There was a low to moderate agreement between parent and teacher reports; but nearly 3% of children gained high scores according to both sets of ratings.


High scorers had considerable psychiatric pathology as reflected in the rates of contact with mental health professionals; and there was a clear effect of anonymity on symptom load in that the children concerned showed significantly higher scores than the officially-identified children. 

Parents reported more symptoms than teachers; and gender differences among the children were reflected by higher scores generally for boys than girls, especially among teacher reports.


It was argued that the present study gave the opportunity to compensate for participation bias. 

While it might have been assumed that parents with a major concern would be more likely to return a questionnaire, the current evidence was that there was a greater incidence of likely high scorers among children in the anonymous group … with the implication that prevalence in fully participating samples could be underestimates.


It was also noted that parent and teacher reports showed no great agreement, especially when rating girls …. with the possible further implication that difficulties among girls are under-recognised by teachers while they may be a little over-stated by parents.  Further, only a small number of high scorers were identified by both parents and teachers with the possibility that situational context may be the issue and/or that there are real differences in behaviours observed at home and at school.
The advice to emerge from this finding was that diagnostic decision-making should be informed by reports from different sources such that family and school observations are both available.


In sum, Posserud et al suggest that, on this evidence, autistic features are more common than has been generally acknowledged ….a finding reinforced by their observation that participation bias (parental reluctance to have their child included in survey data) may lead to an underestimation of real incidence of ASD.  


Identification (The Issue of Head Circumference)


As a follow-up to the notes completed earlier in this series of papers (“ Autism and Head Circumference ” – December 2004), one can refer to the work by Hazlett et al (2005) who examined, via medical records, the pattern of growth in head circumference of a sample of boys (N= 11) with idiopathic autism, and compared these findings against observations from a sample of boys with learning difficulties and a further sample of boys with no recognised developmental disabilities.

Further magnetic resonance imaging data were available from a separate and larger

(N=62)) sample of children with autism and 164 controls with no known disability.

All the participants, whose mean age was approximately 2-5 at the time of evaluation, underwent intellectual assessment and neuro-developmental assessment.


The authors reported that the groups could not be differentiated in respect of head circumference at birth, but the children with autism demonstrated a faster rate of growth in this aspect from the age of 12 months and continuing to the age of 36 months. 

Detailed MRI data from all the groups indicated that total brain volume, and both grey matter and white matter volume, were significantly greater in the autistic group than in the children with learning/developmental difficulty or delay, but no greater than what was observed in the control children.  Total white matter volume of the cerebrum was much greater in the autistic group than in either of the other two groups, although overall volumes did not differ. 

The differences were most marked in the temporal lobes whose mean volume was greater by 8% in white matter and 4.5% in grey matter compared to typical children. Compared with children with learning/development difficulty, the differences were even more striking with percentage differences of 9.4% and 10.6% respectively.


One reasonable implication is that observations of the growth in head circumference over these early years can indeed provide some indications of at-risk status or diagnostic pointers.




From their review of the incidence of Duchenne Muscular Dystrophy (DMD) in the population at large compared to the incidence among children with autism,  Wu et al (2005) formed a hypothesis that DMD and autistic spectrum disorder or pervasive developmental disorder co-exist with a greater than chance probability.


DMD is an X-linked recessive degenerative muscular disorder characterized by deficient dystrophin protein in the muscles.  It has a prevalence of 1 in 35000 males, whose initial symptoms include delayed motor milestones and muscular weakness, and it is usually identified by the age of 5 years.

Boys with DMD often have learning difficulties which may be severe, impacting particularly upon language skills, and individuals affected typically have verbal abilities significantly poorer than those of siblings.


The authors go on to cite a prevalence figure for autism of around 2 in 1000

(and one presumes that this figure refers to relatively severe difficulty and does not reflect the full range of spectrum disorders), and, in the light of the prevalence of DMD at 1 in 35000 males, the probability of the co-existence of the two conditions in one individual should be extremely low.

However, the identification of eight such cases in a clinic in an American among a total of 158 cases of DMD was thought to be suggestive of an association.  In any event, the probability of identifying from 158 cases as many as 8 with this comorbidity was calculated to reflect a chance probability of 0.006. 


The authors further argued that this did not reflect some artefactual or biased circumstance since all children with DMD in this American state are referred to this particular clinic.  They accordingly rejected the null hypothesis that DMD and autism or pervasive development disorder occur together no more often than by chance.


In their discussion, the authors recognise how some might argue that learning difficulty and autism are often associated and that the DMD –Autism association simply reflects a subset of individuals who have DMD and learning difficulty.  However, they counter this by arguing that ASD and learning difficulty are not interchangeable categories; and go on to suggest that the DMD-Autism association is probably not anything new but only recently recognised as a result of the increased awareness of autistic disabilities (and that, previously, the individuals concerned would simply have been identified with the [severe] learning difficulty).


They also dismiss the possible argument that the association simply reflects a language weakness which is somehow inherent in DMD by stressing that a language weakness is clearly differentiable from an autistic status which is identified via observed difficulties according to social and behavioural criteria as well as language criteria.


The authors ponder over a  number of possible mechanisms for this association, including some impairment or alteration of a specific area within the dystrophin gene so that the resulting protein is dysfunctional, or a mutation in the dystrophin gene linked to some impairment in a neighbouring gene accounting for dual areas of disability.


Their conclusion reiterates the point that the DMD –Autism association appears highly unlikely to have arisen by chance and they recommend ongoing genetic and protein analyses not only to continue to explore the nature of the relationship (and to identify those individuals at risk for the joint problems) but also to gain further information about genetic features which could clarify the neurobiological mechanisms operating in the aetiology of Autism. 


Autistic Behaviours and Carer Stress


The paper by Lecavalier et al (2006) refers to the unique challenges faced by the parents or carers of children with developmental disabilities, and to the greater frequency of more negative outcomes than is the case with typically developing children.   In particular, and alongside additional financial burdens, there will be restrictions on daily routines and on social activities, and greater risk of stress.


They go on to review a number of findings which suggest that the parents or carers of children with autism are likely to experience a greater degree of stress than parents of children with other disabilities; and there is converging evidence that a positive correlation exists between the severity of symptoms and the level of this stress.

Some studies (eg  Baker et al 2002) have held that behaviour problems are the most significant predictors of parental stress; and that, once the influence of behaviour problems had been taken into account, no further variance was explicable in terms of level of mental development.


Behaviour problems are common in children and young people with ASD although there has been little attention given to the relationship between the behaviours of this particular population and parental stress …. with the exception of a study by Hastings (2003) who equated the degree of problem behaviour with parental stress and anxiety to find correlations of .58 and .49 respectively between severity of behavioural ratings and maternal and paternal self-reported stress.


The aim of the study by Lecavalier et al was to examine the correlates of caregiver stress in a large sample of children and adolescents with ASD, with a particular interest in differentiating the effects of level of functioning from those of behavioural difficulty, and in determining the stability of this relationship over time.


The participants were the parents and teachers of 293 children and adolescents with ASD.  83% of the ASD sample were male, and the average age was 9 years (SD 3.4).

Mothers provided 86% of the parental ratings.  Parental average age was 39 years, and 48% were graduates.

93% of the teacher ratings were provided by female teachers. The average age was 37.5 and the average length of teaching experience was 10.2 years.


The measures included the Parental Stress Index (Abidin 1995), and the Index of Teaching Stress (Greene et al 1997) where teachers rate on a 5-point scale a series of items to give an overall indication of self doubt, loss of satisfaction, etc.

Ratings of child/adolescent behaviours were also completed.


The outcomes showed that both parent and teacher ratings linked their experience of stress with behaviour problems more than with any other characteristic in the children and young people.

Conduct problems and lack of pro-social behaviours were the most strongly associated with caregiver stress, and the bulk of the variance was explained in terms of disruptive and rule-breaking behaviours.


The authors commented that the similarity of correlations between perceived behaviours and stress observed in the teacher and parent groups was notable, especially as different stress measures were used.  The two sets of ratings did not exactly coincide over the perceived nature and severity of behaviour problems, but there was clear agreement that certain behaviour problems (involving disruptive episodes) were more stressful than others.


With regard to the second question (about stability of the behavioural problems and the associated stress in the caregivers), it was found that only the parental reports of behaviour problems and stress were consistent over time.

There was also the observation that behaviour problems and the caregiver stress exacerbated each other.

The teacher ratings were less indicative of stability of experiences/feelings, nor was there evidence for the above kind of transactional model of stress.


These differences were attributed to a number of hypothesised issues including the observations that teachers spend only a finite amount of time with the children and young people, that school circumstances change more than home circumstances in terms, for example, of levels of classroom support, and that teachers may well have more experience and training with regard to ASD than parents.


Meanwhile, while restating that externalising behaviours in the children and young people are most strongly linked to parent and teacher stress, the authors took some encouragement in that such behaviours are amenable to behavioural or pharmacological intervention. 

They acknowledge that replication studies are required, but conclude on the basis of these current findings that there does seem to be a probability that certain core ASD symptoms are directly linked to caregiver stress, with the implication of a need to identify factors which can reduce this stress…. especially those factors which are open to manipulation.


Much of the above reinforces earlier work on this topic by Sloper (1999) who described the actual or potential stress that may operate among the parents of children with special needs, and who noted the two-way effects in that the demands of looking after the children can place a strain upon the parents and carers which can impair general family functioning and, in turn, inhibit the cognitive or social or behavioural development of the children.

What emerged from a review of studies completed at that time was a need for professionals not only to focus upon the child in terms of direct therapeutic intervention or care, but also to take into account the needs of the family as a whole for support.


Sloper’s review noted, too, that a particular source of stress among parents is any externalising and disruptive behaviours, including sleep problems in the children, which could be reduced in the presence of informal support from other members of the (extended) family, or of more formal support from various agencies such that the parents feel that they have a degree control over events in being able to select from a range of support options.

What also matters is availability of information about the child’s needs and the implications of the disability, the consistency of relationships between families and professionals ideally represented by one key worker acting as the first point of contact, needs-led interventions, and a holistic approach  ….. all of which are more likely to be realised under the current working practices based upon common assessment frameworks and multi-professional teams.


                                 *          *          *          *          *          *


M.J.Connor                                                                                                March 2006




















Abidin R.  1995   Parenting Stress Index Professional Manual.   Odessa, Florida : Psychological Assessment Resources Ltd.  


Baker B., Blacher J., Crnic K., and Edlebrock C.  2002   Behaviour problems and parenting stress in families of 3-year old children with and without developmental delay.   American Journal on Mental Retardation  107   433-444


Constantino J. and Todd R.  2003   Autistic traits in the general population.   Archives of General Psychiatry  60  524-530


Ehlers S., Gillberg C., and Wing L.  1999 A screening questionnaire for Asperger Syndrome and other high functioning autistic spectrum disorders in school age children.   Journal of Autism and Developmental Disorders 29  129-141


Goodman R.  1999   The extended version of the Strengths and Difficulties  Questionnaire as a guide to child psychiatric caseness.   Journal of Child Psychology and Psychiatry  40  791-799


Greene R., Abidin R., and Kmetz C.  1997   The Index of Teaching Stress.   Journal of Schoological Psychology 35  239-259


Hazlett H. Poe M., Gerig G. et al  2005   Magnetic resonance imaging and head circumference study of brain size in autism : birth through age 2 years.   Archives of General Psychiatry  62(12)  1366-1376


Lecavalier L., Leone S., and Wiltz J.  2006   The impact of behaviour problems on caregiver stress in young people with autistic spectrum disorders.   Journal of Intellectual Disability Research  50(3)  172-183


Posserud M-B, Lundervold A., and Gillberg C.  2006   Autistic features in a total population of 7-9 year old children assessed by the Autism Spectrum Screening Questionnaire.   Journal of Child Psychology and Psychiatry  47(2)  167-175


Sloper P.  1999   Models of service support for parents of disabled children.   Child : Care, Health, and Development  25(2)  85-99


Wu J., Kuban K., Alldred E., Shapiro F., and Darras B.  2005   Association of Duchenne Muscular Dystrophy with autism spectrum disorder.   Journal of Child Neurology  20(10) 790-795  


This article is reproduced by kind permission of the author.

© Mike Connor 2006.

Back to NAS Surrey Branch Welcome Page