This short set of summaries refers initially to further evidence to the effect that autism and ASD cannot be regarded as “unitary” conditions.  Instead, the causes are multi-factorial and given cases may reflect a unique pattern and interaction of aetiological and developmental features.


Reference is also made to adult outcomes, with the possible implication for early and targeted intervention.


The final section describes a study which found largely positive social and emotional adjustment among the siblings of children with autism, despite the likely adversities, but the critical issue, again, may be the availability of early and varied support for the child in question and the family.


M.J.Connor                                                                                                     July 2004





A current study has confirmed the growing view of autism as a multi-factorial disorder such that different children, all legitimately diagnosed with autism or ASD, may present a range of aetiological pathways or events.


The study, by Glasson et al (2004), identified a number of perinatal factors which are associated with an increased risk of autism, along with obstetric complications, also linked to autism, which are thought to be an expression of underlying genetic factors.


The findings emerged from a large population-based sample focusing upon those individuals diagnosed with an autistic spectrum disorder within a specified geographical area of Western Australia, using prospectively gathered data comparing autistic subgroups and siblings.


The researchers noted that, compared with normally-developing controls, the children diagnosed with autism were more likely to be first-born and to have parents towards  the higher extreme of the age range within the total sample of those involved in the study. 

Further, the target children were more likely to have been associated with threatened abortion, epidural anaesthetic use, labour induction, and a labour duration of less than 1 hour.


The obstetric complications included foetal distress, the need for caesarean births, and low APGAR scores.  The frequency of complications was greater among cases of autism than among cases of pervasive developmental disorder (not otherwise specified) or of Asperger syndrome.  


The authors concluded that there was a very limited probability that single factors or events gave rise to autistic disorders; and that the aetiology of autism or ASD supports the hypothesised emphasis upon genetic influences (even if, in some cases, non-genetic influences could have been involved in the causal chain ).


This kind of conclusion is supported by the work of Belmonte et al (2004) who argued that the wide variation in the observable manifestations of autism, and the levels of severity of the condition among the individuals so-diagnosed, indicate the involvement of multiple predisposing factors.


The structure and dynamics of neural systems may be influenced by a host of genetic, neuro-chemical and environmental factors.  Further, the initial alterations in neural properties and functioning may influence, in turn, activity-dependent development so that complex behavioural features may eventually be noted which are many steps removed from the original aetiological roots.


These authors summarise much of the symptomatology of autism as reflecting a disturbance in neural information transmission, but the implication concerns the many and various reasons how, where, and why such disturbance may come about.


Still on the subject of causal pathways, Barclay (2004) describes a study by Hornig et al where there was found to be an increased risk of autism-like damage in mice from the use of thimerosal (the mercury-based preservative found in some vaccines).  However, the Safety Review Committee of the American Institute of Medicine responded by restating the view that there is no scientific evidence to link autism or ASD in children either with vaccines containing thimerosal or with the MMR vaccine.


It is further noted that, since 2001, all commonly used vaccines have been available without thimerosal, and the MMR vaccine has never contained this preservative.


This Institute of Medicine report is the most recent of a series of similar reports, and was said to be the final report on the matter of vaccine safety, designed to counter all residual concerns that mercury in vaccines could be linked to an increased risk of autism. 

The data on which the report was based have emerged from 5 large-scale epidemiological studies in America, Britain, Sweden, and Denmark since 2001 which have consistently refuted any thimerosal-autism association.


The Institute representatives also argue that the study on animal subjects cannot be generalised to human casework.  There is no animal model for autism, and there is as yet no clear picture of what causes autism or what are the aetiological routes and pathological mechanisms.


Meanwhile, there is a concern in some quarters that the marked increase in the number of children diagnosed with autism over the last 20 years or so can hardly be explained by genetic factors alone. 

One hypothesis from the authors of the quoted study above continues to link pathology to exposure to low-dose thimerasol via disturbances in immune responses.


Nevertheless, the Institute report is said to have been finalised after considering the findings from the animal study, and it notes how the hypothesis about autoimmune reactions and autism can hardly be sustained in the face of a lack of evidence for autoimmune mediated damage in the brains and central nervous systems of individuals diagnosed with autism.


Their conclusion emphasises that pursuing some hypothesised link between the neurotoxicity of mercury and the developmental pathways of autism will not be productive (at least not unless specific autistic subtypes are identified with a particular and empirically-supportable biological basis).


The recommendation is that the funds available for autism research should be channelled to the most promising areas of investigation, of which the link with vaccines does not appear to be included. 

Meanwhile, it was restated that some hypothetical risk associated with exposure to thimerosal is far outweighed by the very real risk of serious illness.


Adult Outcomes


The longitudinal study completed by Howlin et al (2004) examined the progress in a


sample of children with autism and followed up over a 20 year period from the time of initial diagnosis. The sample was drawn from paediatric referrals to a developmental clinic.


Here again, there seems to be an implication for early diagnosis as far as possible, and for clarifying the particular profile of strengths and weaknesses and styles in a given individual in order to target the intervention strategies most effectively.


The 61 boys and 7 girls in the Howlin et al study were included because they met the criteria of IQ at or above the 50 mark while showing evidence of stereotypical behaviour, abnormal preoccupations, and resistance to change. Onset of symptoms had to be observed before the age of 3 years.

Reassessment included behaviour, social functioning, the level and range of autistic-specific symptoms, and cognitive-linguistic performance.  The average age at follow up was 29 years.


The observations indicated that most of the individuals remained very dependent upon family or residential support.  Only 8 of the sample were thought to be independent.

There continued to be, in most cases, a continuation of the stereotypical and obsessive symptoms along with poor social and communicative skills.


The authors described how the probability of more positive outcomes was associated with the higher initial IQ scores (at or above 70) but there was an implicit acknowledgement of the heterogeneity of the number and type and severity of symptoms among the sample ….. perhaps supporting the above-described view of autism as a multi-factorial disorder.


The authors also noted that ritualised behaviours often interfered greatly with day to day functioning ….. with possible implications for determining the nature and target of early intervention.  Meanwhile, it was recognised that adequate or independent functioning was the more likely when support was available from the family, and/or social services, and/or employing organisations.


The Adjustment of Siblings of Children with Autism


The study by Pilowsky et al (2004) set out to examine the risks of, or resilience to, the effects of genetic liability and the environmental factors involved in having a brother or sister with autism.

This focus of interest was stimulated by the converging evidence for a significant genetic component to the source of autism such that siblings, sharing some of the genetic loading, might experience certain problems, possibly expressed as a less severe pattern of characteristic autistic behaviours.


Reference was made to an earlier study by the lead author and her associates which found that language abnormalities – core symptoms of autism – were not found in the siblings of children with autism, with the conclusion that, on current evidence, language performance would not be seen as a behavioural marker for genetic liability for autism.


Meanwhile, it is recognised that growing up with a brother or sister with these severe problems could have an impact upon the social and emotional adjustment of other children.  The autism may be reflected in challenging, aggressive, self-injurious, and ritualistic behaviours, alongside deficits in communication, which need to be faced and accommodated not only by parents but also by the sibling(s).  The latter may have to face atypical family organisation and parental stress, and accept a lower than full share of parental time and attention.


Previous studies have found evidence for the negative impact of these circumstances, including social dysfunction and affective disorders in significant numbers of the siblings of children with autism, with an incidence well above what would be expected in a population-based sample of children.

For example, Bagenholm and Gillberg (1991) identified a greater prevalence of behaviour problems, depression, and isolation among this group than among the siblings of children with learning disabilities (most commonly Down syndrome)  and among those with normally-developing children.


However, there are references to other studies which have found that the siblings of children with autism were well-adjusted and displayed positive self-concepts compared to samples of siblings of children with other disabilities or of those with normal development.


These current authors suggest that the discrepancies may reflect the choice of comparison groups, so that, for example, a comparison group of siblings of children with Down syndrome may not be ideal given that this syndrome is usually a matter of a sporadic chromosomal anomaly with no genetic implications for other children in the family. 

Their own hypothesis was that the siblings of autistic children would be more likely to show social and emotional deficits that siblings of children with language delay or mental retardation given the greater genetic vulnerability and the environmental circumstances.


Samples of around 30 children each, representing siblings of autistics, language delayed children, and children with mental retardation of unknown genetic aetiology were the participants.  Measures for assessment included ability scales, and rating scales for emotional adjustment, social skills, and behaviour problems. 


The first finding was that more siblings of the autistic children were themselves diagnosed with autistic spectrum disorder although the small numbers in each sample precluded statistically-significant differences.


On measures of socialisation and behaviour, most of the autistic sibling group (nearly 90%) were found to be well functioning and reasonably well adjusted.  However, there was the suggestion that the severity of the disabled child’s condition was related to sibling adjustment.


Family size was also positively associated with the extent of delay in the siblings’ social skills, with the speculation that this may reflect familial stress and the more limited opportunity of the parents concerned to model social cues and responses. However, this pattern was no different for the siblings of autistics than for the other groups, with the suggestion from the authors that the facilities for children with autism and their families, such as early and extended schooling, and support groups, are more available for the children with autism and their families than for their counterparts, and the potentially higher level of stress associated with autism is compensated by the enhanced support facilities.


There were no significant effects associated with the birth order or gender of the children in a family containing a child with a developmental disorder.


The authors concluded that, albeit on the basis of a small scale study, the siblings of children with autism are largely, and perhaps a little surprisingly, well adjusted in the light of the genetic, environmental, and social stresses that may exist.

They also accept that the situation may be different when another sibling has a disorder on the autistic spectrum or of any other kind.  Also, the study did not include a control group of children from families with no child with a disorder, and the inclusion of this comparison might have amplified what difficulties did exist in the target groups.


It is noted, then, that having a sibling with a disability on the autistic spectrum does not presuppose social or emotional difficulties among other children in the families in question ….. although one might highlight the point about the potentially critical role of early support facilities, reinforcing the perceived need for rapidly available intervention and advice to follow the initial diagnosis of ASD thus not leaving the family in a state of uncertainty.



                              *          *          *          *          *          *          *



M.J.Connor                                                                                                     July 2004       




Bagenholm A. and Gillberg C.  1991   Psychosocial effects on siblings of children with autism and with mental retardation.   Journal of Mental Deficiency Research  35  291-307


Barclay L.  2004   IOM report - mouse study continue debate on vaccine-autism link.   Medscape Medical News  June 2004.


Belmonte M., Cook E., Anderson G. et al  2004    Autism as a disorder of neural information processing.   Molecular Psychiatry  9(7)  646-663


Glasson E., Bower C., Patterson B., deKlerk N., Chaney G., and Hallmayer J.  2004   Perinatal factors and the developmental of autism.   Archives of General Psychiatry  61(6)  618-627


Howlin P., Goode S., Hutton J., and Rutter M.  2004    Adult outcome for children with autism.   Journal of Child Psychology and Psychiatry  45(2)  212-229


Pilowsky T., Yirmiya N., Doppalt O., Gross-Tsur V., and Shalev R.  2004   Social and emotional adjustment of siblings of children with autism.  Journal of Child Psychology and Psychiatry  45(4)  855-865



This article is reproduced by kind permission of the author.

© Mike Connor 2004.

Back to NAS Surrey Branch Welcome Page