University of Edinburgh

Autism and Visual Impairment: Taking the Issues Forward in the Context of Multi-disciplinary Working

presented by Dr Anne O'Hare
Consultant Paediatrician

Autism and Visual Impairment Conference, Edinburgh, 27 March 1996


This paper considers the issues around autism and visual impairment and how they can be taken forward in the context of multi-disciplinary working, by looking at the following main categories; recognition of dual impairment, understanding how it has arisen and amelioration of its effects.

The optimum management of children with autism and a visual impairment has to take place in a multi-disciplinary setting as it is almost inevitable that children with these complex sensory and communication difficulties will have met with a number of different professionals when the problems come to light. Sighted children with communication difficulties presenting to our tertiary level hospital speech clinic were already known to other professionals in 75% of cases. (O'Hare & Quew 1996). Even in sighted children, it is difficult to be clear as to whether autism is the basis of the child's behavioural presentation.

We therefore have to reconcile these diagnostic difficulties within the group of children with autism, whilst also recognising that there is a critical role for visual stimulation for the development of communication to such an extent that there is a greater adverse effect in infancy of blindness than there is of a hearing impairment (Preisler 1995).

Definition of autism

The following 2 definitions of autism stress how it is a term for a constellation of features in a child with disturbed communication. Gilberg suggests that "autism is best defined as a behavioural disorder with multiple aetiologies". Many of the tools and checklists developed to diagnose autism in sighted children can only be interpreted in the light of an understanding of the range of normal development and thus a recognition that certain behavioural features are inappropriate for the child's developmental level. This concept is captured in Aitken (1991) definition of autism when he viewed it as "a condition currently best defined using severity cut-off on a number of inter-related social and biological dimensions".

Although the early diagnosis of autism in the visually impaired child is fraught with pitfalls, there may be advantages accruing from an early diagnosis and thus the opportunity for early intervention. The advantages of early diagnosis of autism can encompass the following (Gilberg 1992): "discovery of a treatable underlying condition, the identification of a genetic disorder requiring genetic counselling, the opportunity to help families out of a viscous circle of self-doubt and to help deal with practical problems such as sleep disturbance, the institution of appropriate treatment and education for the child and the opportunity to allow siblings to be better informed so that they can better understand and accept the child's strange behaviour."

However the potential pitfalls and diagnostic confusions when interpreting communicative behaviour when considering a diagnosis of autism are compounded by the presence of a visual impairment. When one looks at the predictors of autism retrospectively by analysis of video tapes of affected children's first birthday parties, (Osterling 1994) one can see the difficulties there would be in translating these features into the visually impaired population. Most of the work looking at early recognition of children with autism has taken place because there is good evidence that early intervention is beneficial and it is thus a worthwhile undertaking. In Osterling's study the number of times a child looked at others was the best single predictor of the child's later diagnosis of autism. When combined with showing, pointing and failing to orient to name, 91% of cases of childhood autism were correctly classified. Imitating and seeking contact were also helpful indicators. Communicative behaviours in following directions and babbling were not helpful. It was noted in this retrospective study that self-stimulation, failure to orient and covering ears was more pronounced in children who subsequently turned out to have autism.

When one considers these features which are so pertinent to the diagnosis of autism in children who are sighted, one can immediately see the problems in translating these whole scale to the visually impaired population. The early social and emotional development of blind children is somewhat different from sighted peers because of the great importance of vision in allowing the child to share attention and to read intention from others. Although we need to recognise the very significant competencies that blind infants have in developing communication and social empathy with their carers, we also need to understand how this can, in some instances when visually dependent, be delayed. This was shown by Troster and Brambring (1992) in their study of the early social and emotional development in blind infants. This study revealed that 9 month old blind infants exhibited a restricted repertoire of expressive reactions although most 9-12 month old blind infants reacted with pleasure to interactive games with tactile stimulation. Nevertheless, they still had problems relating to a game with an object. Most 12 month blind infants showed spontaneous signs of affection although very few initiated contact and by 18 months none of the blind children used conventional or referential gestures and most showed difficulty in responding to simple commands.

Failure to communicate

One of the most powerful indicators of autism in our population of children attending our tertiary speech clinic was that of failure to communicate in any sense with peers. Many of the other children that we were seeing in the speech clinic had very severe problems with their understanding and use of speech and were thus unintelligible to their peers and likely to have great difficulty following the language in play. Nevertheless they showed a desire to participate in play with their peers. Here again we have a diagnostic pitfall for blind children when we come to look at their behaviour with peers who are sighted in nursery situations (Preisler 1993). Her observations of blind children in nurseries with sighted children revealed that the blind child seldom participated in sighted children's play or initiated contact with the sighted children. They never spontaneously took part in symbolic or role play although they had all developed pretend play by 3-4 years. In group activities with the teacher, if the context became too abstract the blind child would engage in stereotypies.

A further dilemma both in recognising and in understanding the nature of the communicative difficulties in children with autism and/or visual impairment, is the interpretation of their language development. Although the language development of blind children is different from those of their sighted peers, this can be seen in quite obvious behaviours such as the stage of echolalia which can be more protracted for blind children (Elstner 1983) one must avoid concluding that the language is disordered and not of communicative use to the blind child as this may be spurious. Vocal verbal behaviour is the main recognisable referential behaviour available to a blind child (Preisler 1994) and from around the age of 4 years, the blind children at a sighted peers nursery interacted more in the group and this interaction was often based on words games or alliteration. There was seldom exchange of ideas or meaning. Thus it is important not to misinterpret how the child is using their language to communicate.

Nevertheless one also has to consider that within the pre-school blind multiply-handicapped population, there is a high incidence of language disorder of the order of 33.8-49%. Almost half of multiply-handicapped blind children are without language (Elstner 1983).

Visual function

In order to understand the nature of the child's difficulties in communication, we need to be clear about the child's visual function and this aspect of assessment is again best taken forward through a multi-disciplinary base. The incidence of developmental regression is higher in children who are blind and who have a neurological visual impairment. Children who have a neurological visual impairment can exhibit behaviours which on the surface look similar to those exhibited by children with autism and when trying to understand the nature of the child's behaviour and communication it is important to remember this and this can only really be best appreciated by marrying the experience between professionals who work with visually impaired children and those who work with children with autism. Children with central neurological visual impairment exhibit impaired visual attention and lack of visual curiosity. If they have bilateral striate cortex damage they can tend to look away from people, objects or events if they are reaching towards them. They also pay attention to one toy or part of a toy rather than a variety and so one can see how this could be confused with the behaviour seen in children with autism (Jan et al 1987).

It would seem that developmental setback is more common in children who are blind. Many children with neurological visual handicap have additional handicaps and these often take the form of learning difficulties. The incidence of autism in children with moderate to severe learning difficulties is high and it would therefore not be surprising if the dual impairments coexisted. Thus it is important to be quite clear about a child's level of visual acuity and there are a number of issues that need to be recalled when assessing visual acuity in children with cortical visual impairment (Birch and Bain 1991). In their study they showed that 58-68% of children with low vision who failed to fixate or track on clinical testing nevertheless have a measurable visual acuity on forced choice preferential looking techniques (FPL acuity). Whilst the Ophthalmologist's judgement of visual impairment based on ophthalmoscopic appearance, clinical examination refractive error and ocular motility generally agreed with the FPL acuity, in 47% of patients judged to have severe visual impairment the FPL showed only moderate to mild acuity deficits. Children with the largest acuity deficit tended to show the greatest improvement in vision. Forced choice preferential looking acuity has wide application and even in a study where 60% of children had neither understandable speech or could clearly point, 99% had success on FPL acuity cards (Hertz and Rosenberg 1992). When we recall that even very low vision greatly improves the infant's opportunities to engage in interpersonal communication and to share meaning (Preisler 1995) we can begin to understand why we might need to consider that a child has an autistic communication disorder if they seem to have adequate visual acuity.

Being able to read the mentalistic significance of the eyes, appears to be an important development to help the young child understand the thought processes and intentions of another individual. Although we do not know what minimal level of visual acuity one requires to display these skills, it is probably helpful to consider Baron-Cohen's findings (Baron-Cohen et al I 995) and consider how they might be affected by blindness. He showed that by 13 months normal young children understand the laws of vision in that the eyes are for seeing and that for a person to see something, their eyes need to be open and their line of regard needs to be unobstructed. Children with autism also understand this basic fact. By 14 months normal children gaze monitor and by 18 months they know that an object an adult is looking at whilst naming it is the one the adult is referring to, whereas autistic children do not seem to either appreciate this or display the fact that they appreciate it. By 3-4 years normal children know that if a person looks upwards and away they are thinking, whereas autistic children do not seem to know this. These are interesting findings although the cause and effect with respect to the autistic child's deficit in understanding the mentalistic significance of the eyes is as yet unclear.

Issue of regression

Related to both understanding the process of the communication difficulties and the issue of amelioration, we need also to look across the disciplines to what we know about the whole issue of regression in both sighted and blind children. As yet the prognosis for children who are blind and who regress is not entirely clear although it does appear that there are broadly 2 patterns of recovery, one of which is very poor (Cass et al 1994). The issue of regression also has to be considered when we relate it back to the optimum practice of early recognition of autism. Even within sighted children 1:5 parents will report normal development in their autistic child until the age of 18 months. Thus in the retrospective analysis of first year birthday parties carried out by Osterling, there was no difference in the behaviour of such children compared to normals at this stage. There is also a pattern of regression well recognised in slightly older pre-school children and if this occurs after the age of 3 and thus takes them away from the former definition of infantile autism, they have been described as displaying a condition called disintegrative psychosis. This is a pattern of regression which occurs in pre- school children who formerly have normal development including normal language acquisition. It has features indistinguishable from severe autism except for the age of onset. There is no identified organic pathology and it produces a low level of functioning which remains virtually static over a 10 year follow up period (Hill and Rosenbloom 1986).

Sometimes a clear explanation can be established for the onset of autistic regression and this clearly had to be considered in children with a visual impairment as it may have a direct therapeutic indication such as the autistic regression which has been described in relation to limbic pathology of the brain and epilepsy (Deonna et al 1993).

A dilemma that relates to both sighted children and blind children who undergo developmental regression, is understanding whether this is following an innate problem in brain development, or whether it is environmentally determined, or whether it is an interplay of both. Amygdalectomy of the brain in monkeys shows no immediate effect in infancy and abnormal social behaviour only develops subsequently and thus this is an illustration of the fact that there can be a latent period between the actual damage to the brain and the onset of the behavioural manifestations. However Preisler (1993) also showed in her study of blind children in nursery with sighted peers, that 2 out of 9 children regress markedly in their first year at the nursery.

Final thoughts 

Finally some thoughts to leave you with, with respect to the amelioration of the difficulties experienced by children with autism and visual impairment. Unfortunately token and sign language can be difficult for autistic children to cope with and does not compensate for the problems that they seem to have in understanding spoken language. Clearly this presents a challenge when one is dealing with children with dual impairments. Children who have very poor visual function from neurological visual damage often have good preservation of colour identification and this might be able to be harnessed to help children with the dual impairments.

Finally, although there is insufficient time within this lecture to address the points in detail, one needs to work in a multi-disciplinary forum to understand the potential role of a range of interventions that might be helpful for a child with autism and thus may be helpful for the child with the dual impairment. By examining the potential interventions with other professionals and with the parents, one can help determine the place of pharmacotherapy such as vitamin B6 and magnesium supplements, anti-convulsants, Naltrexone and neurolectics as well as the role of physical exercise therapy and education.


Aitken, K (1991), Annotation Examining the Evidence for a Common Structural Basis to Autism. Developmental Medicine and Child Neurology, 33: 930-938.

Aitken, K (1991), Diagnostic Issues in Autism -Are we measuring the emperor or another suit of clothes? Developmental Medicine and Child Neurology, 33: 1015-1020.

Baron-Cohen, S, Campbell, R Karmiloff-Smith, A, Grant, J, and Walker, J (1995), Are Children with Autism Blind to the Mentalistic Significance of the Eyes. British Journal of Developmental Psychology, 13: 379-398.

Birch, EE and Bain, MC (1991), Forced Choice Preferential Looking Acuity of Children with Cortical Visual Impairment. Developmental Medicine and Child Neurology, 33: 722-729.

Cass, H, Sonksen, P, McConachie, H (1994), Developmental Setback in Severe Visual Impairment. Archives of Disease in Childhood, 70: 192-196.

Deonna, T, Ziegler, A, Moura-Serra, J and Innocenti, G (1993), Autistic Regression in Relation to Limbic Pathology and Epilepsy (Report of 2 cases). Developmental Medicine and Child Neurology, 35: 158-176.

Elstner, W (1983), Abnormalities in the Verbal Communication of Visually Impaired Children in (Ed) Mills, AE Language Acquisition in the Blind Child - Normal and Deficient. Croom Helm, Chapter 3, 18-41.

Gilberg, C (1984), Autistic Children Growing Up - Problems During Puberty and Adolescence. Developmental Medicine and Child Neurology, 26: 122-129.

Gilberg, C (1992), Autism and Autistic-like Conditions in (Ed) Aicardi, J Chapter 26 : Diseases of the Nervous System in Childhood. Clinics in Developmental Medicine. 115-118 Mackeith Press Oxford Blackwell Scientific Publications Limited 1295-1320.

Hertz, BG and Rosenberg, J (1992), Effect of Mental Retardation and Motor Disability on Testing With Visual Acuity Cards. Developmental Medicine and Child Neurology, 34: 115-122.

Jan, JE, Groenveld, M, Sykanda, AN and Hoyt, CS (1987), Behavioural Characteristics of Children with Permanent Cortical Visual Impairment. Developmental Medicine and Child Neurology, 92 : 571-576.

O'Hare, AE and Quew, R (1996), The Identification of Autism in Children Presenting to a Speech Clinic - Implications for Practice. Submitted for publication.

Osterling, J and Dawson, G (1994), Early Recognition of Children with Autism - A Study of First Birthday Home Video Tapes. Journal of Autism and Developmental Disorders, 24-3: 247-257.

Preisler, GM (1993), "A Descriptive Study of Blind Children in Nurseries with "Sighted Children". Child Care Health and Development, 19 : 295-315.

Preisler, GM (1995), "The Development of Communication in Blind and in Deaf Infants - Similarities and Differences". Child Care Health and Development, 1: 79-110.

Troster, H & Brambring, M (1992), "Early Social and Emotional Development in Blind Infants". Child Care Health and Development, 18: 207-227.

Biographical Detail

Dr Anne O'Hare is a Consultant Paediatrician in Community Child Health, Royal Hospital for Sick Children (RHSC), Edinburgh and Senior Lecturer, Department of Child Life and Health. Works as a consultant to speech clinic at RHSC which sees children with autism and consultant to The Royal Blind School.