Mapping autism research on behalf of the steering group from The National Autistic Society (NAS) and Parents Autism Campaign for Education (PACE) Funded by the Baily Thomas Charitable Foundation
Research activity in the UK and the rest of
Conclusions: integrating findings from the
mapping exercise with other recent reviews
Mapping autism research Identifying UK priorities for the future
Published by The National Autistic Society 2004
The authors of the report are grateful to the staff of The National Autistic Society (NAS) and of Parents
Autism Campaign for Education (PACE) for supporting the mapping exercise and the publication of this report,
and to the Baily Thomas Charitable Foundation for funding the study. They are especially grateful to Virginia
Bovell, David Potter and Su Thomas who initiated the project and whose enthusiasm, encouragement and
friendship sustained us throughout the work and enabled us to complete what we set out to achieve. Any
errors, inaccuracies or biases are the responsibility of the authors.
Autism† is more common than was previously recognised (Wing & Potter, 2002). This has led to renewed
efforts to summarise the state of the science and identify priority areas for future research by scientific and
practitioner bodies in the UK and the USA (Institute of Medicine, 2001; Medical Research Council, 2001;
National Academy of Sciences, 2001; National Autism Plan for Children, 2003; Novartis, 2003; Public Health
There is increasing recognition that discussion and dissemination of research findings and identification of
priority areas for future research should involve a dialogue between scientists, policy-makers, people with
autism and their families. In the absence of such a dialogue there is a risk of misunderstanding and a
breakdown in communication. For instance, the public cannot evaluate the relative merits and opportunities
of scientific findings, nor why particular programmes of research are being pursued, unless research is
disseminated in an open and accessible manner.
A different type of risk is threatened if the research and policy-making communities set de facto research
priorities (in terms of awarding grants to specific programmes) without taking a comprehensive overview of
the field. This should include taking account of the views of the non-academic community in the
identification of appropriate goals for research and taking advantage of their insights into the challenges that
autism presents to individuals and families based on their experience of autism ‘all day every day’.
In this context, Parents Autism Campaign for Education (PACE) and The National Autistic Society (NAS)
initiated an exercise to ‘map’ the state of the autism research field in the UK. This exercise was developed and
conducted in partnership with an academic base – the Institute of Child Health (ICH). The mapping exercise
Gather the views of scientists and the non-academic community regarding priorities for
Identify the pattern of research funding for autism in the UK
Compare current research activity in the UK to that in the rest of the world
Identify emerging themes and priority areas for future research in the published
This report summarises the most important findings to emerge from this exercise. No one source of
information can determine future priorities. However, the strategy of combining information from different
sources and discussing the findings of the mapping exercise with the different stakeholder groups should help
determine appropriate research priorities and goals over the coming years. The findings of this report will
prove to be a useful guide to funding agencies and government departments with respect to current strengths
and weaknesses of the UK research field. It is intended to help them identify emerging areas of research in
which academics and those affected by autism see likely benefits, in order to set effective funding priorities.
†Autism is now considered to be a spectrum of related conditions (‘autism spectrum disorders’ including Asperger syndrome and pervasivedevelopmental disorder) that vary in severity. For simplicity the term ‘autism’ will be used throughout the report.
The current report makes a unique contribution by gathering views about priorities from scientists and from
people with autism and their families; by contrasting research activity and funding in the UK to that in the
rest of the world; and by integrating information about research activity, research funding, views on future
priorities, and emerging themes in the published literature in one document.
Autism spectrum disorder is the term used to describe a range of behaviourally defined neurodevelopmental
conditions. They are characterised by impairments in social interaction, social communication and language
development and a restricted repertoire of interests, behaviours and activities. Sensory abnormalities and
unusual interests in some sensations are common. A lack of imaginative play indicates an underlying difficulty
with generation of ideas that is highly relevant in the development of understanding of, and thinking about,
other people and other situations. All of these characteristics can be seen in varying degrees of severity
(World Health Organisation; ICD-10, 1993). As a developmental condition the manifestation of autism for any
one individual will vary across the lifespan and also with maturation, the effects of different environments
and due to specific interventions and treatments. One view held by many scientists and individuals with
autism and their families is that these characteristics are shared in varying degrees of severity with the
‘neurologically typical’ population.
Autism affects millions of people worldwide, yet many gaps remain in our understanding of the condition. This
has led to renewed efforts by scientists, practitioners and affected families to summarise the state of the
science and identify priority areas for future research. Unique information is added by this mapping exercise
adds to the dialogue in the UK between funding agencies, government agencies, researchers and people with
autism and their families. The overview is crucially important as it will enable the research and policy-making
communities to support research relevant to the views not only of scientists but also of affected families.
Gathering views from scientists, people with autism and their families enables us to highlight
where agreement about areas of need exists and where opinions differ.
Information from research funding agencies provides for the first time a picture of the
distribution of research funding in the UK.
No previous reviews have systematically compared autism research in the UK to that in the rest of
Priority areas emerging from a literature review and suggestions from experts in the field in a
comprehensive range of research areas are collected together in one document.
Views of researchers and the non-academic community aboutfuture priorities
Scientists and the non-academic community were in agreement that research into the causes of
autism and the effectiveness of different interventions were priority areas.
However, at the level of the specific type of research required into causes and interventions, there
was some difference in the opinions of researchers and non-academics.
The current exercise does not allow us to determine the reasons underlying the divergence of views
between researchers and non-academics where these occurred. This topic should be a focus for
future qualitative research studies.
Few people in either sector gave high priority to research on families and services. Attitudinal
research is needed to identify the reasons for this, given the potentially beneficial impact of high
quality research into the effectiveness and acceptability of services.
The major UK and USA government and charitable bodies that fund research on autism were asked to send
summaries of the projects they supported between 1996 and 2000.
In comparison with the major USA government and charitable funding agencies, proportionally less
research into the causes of autism is funded in the UK.
More research in all areas of science is funded in the USA than in the UK and research funding for
autism is co-ordinated across government agencies in the USA.
Family and services research receive only a small proportion of research funding in both the UK
Research activity in the UK and the rest of the world
A systematic search of the published literature in a five-year period (1996-2000) enabled the number of
active autism researchers in the UK to be compared to that in the rest of the world.
UK researchers accounted for 16% of active autism researchers on the main scientific databases.
One area of significant weakness in the UK is research into interventions. The proportion of
researchers evaluating interventions in the UK was a third of that in the rest of the world.
Little research is published on family issues and services for people with autism even though this
Priority areas identified from a review of the published literature
These are summarised in Appendix A of the main report. They extend recommendations on existing knowledge
and research gaps made in other recent expert statements.
Joint working There is increasing recognition that identification of priority areas for future
research, evaluation of funding proposals and dissemination of research findings should jointly
involve scientists, policy-makers, people with autism and their families. Consensus view Researchers and non-academics agree that a greater emphasis should be placed
on research into both environmental and genetic contributions to the causes of autism and into
well- controlled intervention studies. Research into differing opinions needed There is some divergence of views between researchers
and non-academics in terms of the specific areas of research that are needed. It is important to
understand these differences to increase the consensus on research priorities. Qualitative research
studies should investigate this divergence of views. New mechanisms required Mechanisms need to be developed to incorporate the views of
researchers, policy makers, people with autism and their families into future research and
Overlooked areas deserve investigation Research into effects on families and research into
services are overlooked areas. Given the practical implications for government agencies in the
education, health, social care and employment fields, and for the voluntary sector, the reason for
these being such low priorities deserves further investigation. Areas of weakness identified in the UK field These were research into interventions (where the
proportion of active researchers was low) and research into the causes of autism (where the
proportion of research studies funded was low). Regular monitoring of research activity At the moment, there is no overview of research activity
in the autism field. It would help in monitoring UK research if there was a systematic annual
(a) research funding for autism in the UK from both government and charitable sources and
Electronic network of scientists Using current information technology, this mapping exercise has
demonstrated that it is possible to identify the active autism research community in the UK. It may
be useful to establish an electronic network of scientists in the autism field to disseminate research
findings, collaborate with parent organisations and publicise research funding opportunities.
Future Activities • Models for policy and research strategy Initiatives from the USA (in particular the Interagency
Autism Coordinating Committee) and Canada (in particular the Canadian Autism Intervention
Research Network) provide models for further development of policy and research strategy
involving all stakeholders in the process.
The findings of this report will prove to be a useful guide to funding agencies and government departments
with respect to current strengths and weaknesses of the UK research field. It is intended to help them identify
emerging areas of research in which academics and those affected by autism see likely benefits, in order to
set effective funding priorities. Clearly, no one source of information can determine future research priorities.
However, the strategy of combining information from different sources and discussing the findings of the
mapping exercise with the different stakeholder groups should help in the decision-making process.
We are encouraged that many of the conclusions from the mapping exercise are in agreement with other recent
reviews prepared by government departments and agencies, as well as researchers, in the UK and the USA†.
This newly documented information will contribute to the dialogue between research funding agencies,
scientists and people with autism and their families about future research goals. The report also contains
important conclusions about the dissemination of research findings and the translation of knowledge into
†The report is intended to be complementary to other recent summaries of the field (Institute of Medicine, 2001; Medical Research Council, 2001;National Academy of Sciences, 2001; National Autism Plan for Children, 2003; Novartis, 2003; Public Health Institute of Scotland, 2001).
This mapping exercise gathered views from scientists and people with autism and their families regarding
future research priorities, collated information about funding for autism research in the UK, and made a
systematic comparison of research activity in the UK compared to that in the rest of the world. It also
includes a comprehensive literature review of the current state-of-the-science in autism.
In order to organise and summarise the information collated from scientists and the non-academic
community, funding bodies and the published literature, a classification system based on that used by the
NAS Information Centre was employed. The largest research categories were broken down further into sub-
categories to document areas of research in more fine detail. 1. Causes/Aetiology
An event or series of events that result in an individual developing autism. a. Genetic causes
Genetics research including the broader autism phenotype, the inheritance of autism,
chromosomal abnormalities and candidate gene causes. b. Environmental causes
All factors other than genetic susceptibility, including obstetric and perinatal factors,
gastrointestinal abnormalities and immunological causes. 2. Epidemiology/Prevalence
The number and distribution of cases of autism in a population. 3. Diagnosis and Assessment
All aspects of the identification of autism, from initial screening to the specific diagnostic assessment
and tools that are used to measure the degree of impairment. 4. Symptoms
Signs that indicate a possible disorder or disease. a. Neurology
Research investigating brain abnormalities in autism using examinations of brain tissue,
structural and functional imaging techniques, animal models and neurochemical investigations. b. Neuropsychology
Research investigating how abnormalities in cognitive and developmental processes and
c. Behavioural
All aspects of behaviours that are associated with autism e.g. stereotyped and repetitive
behaviours, sleep difficulties and toilet training. d. Associated medical and psychiatric disorders
Medical disorders and conditions that are associated with autism e.g. tuberous sclerosis and
epilepsy, and associated psychiatric disorders e.g. anxiety and depression. 5. Intervention/Treatment a. Psychoeducational
Including behavioural interventions (e.g. ABA), communication-and language-focused
interventions (e.g. PECS), play-based approaches and educational programmes (e.g. TEACCH). b. Psychopharmacological
Psychopharmacological medicines, that is, prescription drugs used to treat autism. c. Biomedical
Including non-mainstream, complementary treatments such as vitamin therapy, chelation
therapy, secretin, diets and sensory integration approaches. 6. The Family and Services
Including a wide range of studies associated with the family function, effects on parents and siblings
etc. All research into services available to individuals with autism (e.g. supported work schemes,
community support and families’ experience of diagnostic and intervention services). Research on policy
and financial matters was coded under this section.
Gathering views of people with autism and their families regarding future research priorities
A list of parent and voluntary organisations was created whose membership included parents and other
relatives of individuals with autism, individuals with autism, and professionals with an interest in autism. It was
based on those autism organisations known to the NAS Information Centre. The list contained the following:
Families for Early Autism Treatment (FEAT) (now known as the Schafer Report)
Parents Autism Campaign for Education (PACE)
Parents for the Early Intervention of Autism in Children (PEACH)
Independent Living on the Autism Spectrum web mailing list
These organisations were contacted and asked to relay our request for information to their mailing list. In
some cases the message was forwarded to email server address books, whilst in others our call was reproduced
in newsletters or other publications. The call for information was also posted on parent organisation websites
(including Allergy induced Autism (AiA) and Autismconnect) and in Communication, the NAS publication.
People were asked for their views on the future priority areas for research. This approach was couched as an
opportunity for the non-academic community to have ‘their say’ about the future direction of autism research.
In total 123 responses were received from parents of people with autism (N = 98), advocates and
professionals working in the field (N = 17), international societies (N = 6) and respondents who identified
themselves as individuals with an autism spectrum disorder (N = 2). As these subgroups are not large enough
to compare, for example, parent vs. professional responses, data will be presented for the responses received
from the non-academic community combined. For comparison of these responses with those of responses
received from researchers we will use the shorthand notation ‘parents’ (as the predominant respondents) to
refer to this group in the data presentation.
We have no mechanisms to test how representative the responses we received are of the opinions of the
broader non-academic autism community. One limitation of the current exercise is that the particular section
of the non-academic community whose views were sought may be mainly parents of people with autism,
perhaps predominantly of young children, although we did not systematically enquire as to this information.
Their concerns and priorities might differ from those of parents of adults with autism, professionals working
with adults with autism, and adults with autism. Any future research that solicits views regarding priorities
for research and service development in the field must include parents and carers of adults with autism and
Many of the responses received from the non-academic community expressed their thanks at being given the
opportunity to have their say on future research priorities, and also how important they thought the mapping
project was. Below are two examples of responses received from parents.
“Our 3 and a half-year-old son was recently diagnosed with autism. We are still suffering from
information overload about the disorder, but from our preliminary reading of the subject so far, I would
suggest that the main areas of research should focus on:
Fundamental brain structural changes and brain chemistry in autism, including genes responsible
An evidence-based approach as to which early interventions really do make a difference to long
term outcome. The Lovaas approach needs restudying with bigger numbers.
Specific interventions such as types of speech therapy need to be continuously evaluated.
Rather like the UK Leukaemia trials, it might be an idea to look prospectively at a specific problem
“I consider that the biochemical differences in our children need to be fully identified and investigated
as a matter of urgency and strategies to help our children maximise their potential development – e.g.
the role of diet, which medicines/supplements could help, what are the role of Essential Fatty Acids,
From this base it should be possible to develop a protocol for testing and appropriately treating autistic
children immediately upon diagnosis (or preferably just upon suspicion of autism). Such research could
build upon that currently being done in the US by the Defeat Autism Now (DAN) group of doctors.”
Active researchers in the autism field were identified for two purposes. First, in order to gather their views
about future research priorities. Second, to assess the amount of research activity in different fields
internationally and then to compare the profile of UK research activity to that in the rest of the world. The ISI
(Institute for Scientific Information®) database was used as the source of information regarding peer-reviewed
published research. ISI is the largest, most up-to-date online database of scientific research. ISI is made up of
three multidisciplinary databases: the Science Citation Index, the Social Sciences Citation Index, and the Arts
and Humanities Citation Index. Each database holds the abstracts of the core journals (over 8,100) in addition
to research articles, book reviews, editorials, letters and biographical material. Coverage in each of the
citation indexes extends back to 1981 and approximately 22,000 new articles are added each week.
A systematic search was conducted on ISI for literature published in a five-year period (1996-2000) with a
keyword of autis* (covering autism, autistic, autist etc.). This search yielded 3,107 results. That is, just over
three thousand scientific articles about autism had been published in the past five years (approximately 10
per week). The information from this search was exported to Reference Manager® software. The first authors
on each of the database items were sorted into alphabetical order. Where an author had multiple entries as
the first author, these were deleted, so that researchers did not appear as first author more than once in the
database. In these cases, the topic of the article retained was selected to be most broadly representative of
the work of the particular researcher. This reduced the list of researchers to 1,336. Many groups of researchers
alternate the first authors on papers, so this helped to increase the number of different first authors.
The index of research activity we chose was at the unit of ‘researcher’. An alternative index would have been
the absolute number of publications in each area of research. These alternative indexes of research activity
have different strengths and weaknesses. A pragmatic decision was taken to use the ‘researcher’ and not the
‘publications’ unit of research activity since this would require fewer units to be coded and added to the
By viewing each abstract, non-autism research that had been incorrectly identified by ISI (for example, ‘Some
dinoflagellate cysts from the Kimmeridge Clay Formation in North Yorkshire and Dorset, UK’) was excluded,
leaving a final total of 1,222 active autism researchers. The list of researchers was married with information
provided by the NAS Information Centre on researchers active in the field of autism to ensure no significant
researchers had been omitted – none had. Contact details (email, fax or address) were identified for 972
researchers by searching online for their research institution. Many (predominantly junior) researchers had
moved from the research institution that was listed on ISI at the time their research was published, which
meant that it was not possible to identify a current contact address for all active researchers identified.
All 972 researchers were contacted and a request was made for information on the topics of current research
they were involved in, on agencies that had funded their autism research over the past 5 years, and their
views on priorities for autism research in the next 10 years. Publicity about the mapping exercise was
published in the main autism journals (Journal of Autism and Developmental Disorders, Autism: TheInternational Journal of Research and Practice, Focus on Autism and other Developmental Disabilities) and the
inaugural IMFAR (International Meeting For Autism Research) conference programme (San Diego, California,
November 2001), to increase awareness about the project and hopefully to increase returns from our request
A total of 207 responses were received from the researchers contacted, representing a response rate of those
autism researchers approached of 21% and an overall response rate of 17% of all active autism researchers
identified from the ISI databases. Of these responses, 23 were from researchers who did not provide the
requested information. This included some researchers sending reference lists of all their published work, CVs
and links to their websites. This left 184 analysable responses. Given the high cost (and likely low return) of
re-contacting non-responders this was not considered a good use of time.
The information received from the researchers was categorised by topic area, coded and entered onto a
statistical database (SPSS®). The published literature from the search identifying all active autism researchers
on the ISI database (N = 1,222) was coded in the same way. This provided a way of categorising the topic
areas of autism research published in the past five years, which allowed us to identify any bias in the research
areas of researchers who had responded to the call for information. A comparison of research topic areas
between the researchers who responded to our enquiry and all autism researchers on ISI revealed that our
sample of researchers was fairly representative (data not shown).
Major UK government and charitable bodies that fund autism research were identified from the names of
funding bodies frequently mentioned in acknowledgements of papers, from sources of research funding
identified by the UK-based researchers who responded to our survey, from a search of web-based NHS
funding summaries (National Research Register), and from our experience of reviewing grants for UK-based
In order to have a comparison to this data on UK-based research funding we sought information from two
(not completely comprehensive) sources of autism research funding in the USA. First, as there are several
well-established USA autism charities that fund significant amounts of research, the two largest were
approached: the National Alliance for Autism Research (NAAR) and Cure Autism Now (CAN). Second, the USA
National Institutes of Health (NIH) was approached as it funds a significant amount of autism research across
its many divisions (e.g. National Institute of Child Health and Human Development (NICHD), National Institute
on Deafness and Other Communication Disorders (NIDCD), National Institute of Mental Health (NIMH),
National Institute of Neurological Disorders and Stroke (NINDS)). It also has a centralised system for indexing
its research funding and a co-ordinating committee for autism research.
The following UK-based government and charitable organisations that have funded research into autism were
identified, as were the above-mentioned USA organisations:
Biotechnology and Biological Sciences Research Council (BBSRC)
Economic and Social Research Council (ESRC)
National Health Service (NHS)/Department of Health (DoH) Research and Development
National Alliance for Autism Research (NAAR)
Each of these organisations was contacted and a request was made for them to send copies of their annual
reports from 1996-2000 i.e. overlapping with the review of research activity from the ISI database.
Information regarding the title and aims of each autism research project funded in the five-year period was
extracted from the reports. The research information from funding agencies was categorised in the same way
as the information obtained from the ISI researcher database. Separate entries were not made for a single
project for every year the project received funding when the lifetime of the grant was greater than one year
(i.e. each individual project was only counted on one occasion).
The summary of research projects funded by UK-based funding agencies is not exhaustive but every effort has
been made to systematically collate this information. Organisations not responding to our initial request were
contacted a second time. Some (for example, those known to have funded autism research from
acknowledgements in publications or from information sent by researchers who responded to our survey) did
not respond to this second request. For some funding agencies full information for the five-year period was
not available. For yet others, only the individuals and institutions funded and not the topic of the research
projects funded were detailed in the material sent in response to our request. Because the topic of these
research projects could not be categorised according to our system, information on these projects was not
included in the database and was therefore excluded from the following summary.
Thus, the research summarised under-represents the total number of UK autism research projects funded in
the five-year period and we cannot tell what bias was introduced through this process. These limitations also
apply in part to the information we received from the USA funding agencies. Information on the amount of
research spend, as opposed to the number of projects funded, was variable across the responses received and
was sometimes hidden by a summary of spend across autism and non-autism studies. For the purposes of this
mapping exercise, therefore, the total number of projects funded is the unit count.
Recent literature was reviewed using the same categorisation system. The exercise was not a systematic
review but rather involved careful reading and integration of recommendations in recently published reviews,
as well as reading of the most important original sources. The topics covered overlapped with other recent
state-of-the-science reviews cited previously on page 3. For the purposes of the mapping exercise, the focus
was on identification of consensus views regarding priority areas for future research, likely to bring significant
gains in knowledge and understanding. Some of the recommendations involved the extension and re-focusing
of ongoing research enterprises, whilst others involved more ‘high-risk’ or ambitious but potentially beneficial
Within each category area, feedback and comments were solicited from senior academic researchers in the
field (see Appendix A). The reviewers were asked to consider the following in their comments:
If they agree with the suggested future directions
If further future directions should be added.
However, the content of the literature reviews and the summary recommendations contained in this report –
and any inaccuracies or biases represented therein – represent the views and opinions of the authors of this
The priorities for future research identified via the literature review are shown in Appendix B.
Figure 1 shows the distribution of research on autism published in the last five years as indexed by the ISI
author count, classified into the 6 major category areas: causes, epidemiology, diagnosis and assessment,
symptoms, intervention and family and services. Figure 1 – Global research activity in autism 4% Family and Services 18% Causes 20% Intervention 2% Epidemiology 4% Diagnosis 52% Symptoms
By far the largest proportion of research activity was in the area of symptoms (52%). There were four
subcategories within this section: neuropsychological, neurological, comorbid disorders and behavioural
The breakdown of research at the sub-category level in the area of symptoms shown in Figure 2 overleaf
reveals that nearly half of research into symptoms (48%) was directed at investigating the neuropsychological
symptoms of autism. Approximately one fifth was in the area of neurological symptoms (21%) and comorbid
disorders (19%) and just over one tenth (12%) was in the area of behavioural symptoms. In our coding
scheme, the neuropsychological symptoms category included research in cognitive psychology (e.g. theory of
mind, executive functions), communication, socialisation and language. These aspects form the core deficits of
autism, which may explain why a high proportion of research into symptoms of autism focused on these
Furthermore, many of these studies were small, group experimental studies requiring less time and resource
than other areas of work demanding greater capacity in terms of multidisciplinary collaboration (for example,
in research into comorbid disorders), multi-site collaboration (for example, in genetic research) or time (for
example, in longitudinal intervention designs).
After research on symptoms of autism, research on intervention (20%) and causes (18%) were the next most
common areas of research (Figure 1). Looking at the subcategories of research into causes, approximately
two-thirds of research into the causes of autism was genetic (63%) and one-third environmental (37%),
which included all aetiological research other than genetic research (e.g. birth complications, immunological
causes etc. see Figure 3). Looking at the subcategories of intervention research, 42% was psychoeducational,
34% was psychopharmacological and 24% was biomedical (Figure 4). Figure 2 – Breakdown of research on symptoms Figure 3 – Breakdown of research on causes 12% Behaviour 48% Neuropsychological 37% Environmental 19% Comorbid 63% Genetic 21% Neurology Figure 4 – Breakdown of research on intervention 23% Biomedical 62% Psychoeducational 15% Psychopharmacological
A relatively small amount of research has been published in the last five years on the topic of diagnosis and
assessment (4%) and epidemiology (2%) (Figure 1). These categories are relatively specific compared to
‘umbrella’ categories such as causes, symptoms and interventions. In the case of epidemiology, the low index
of research activity also reflects the large and time-consuming nature of such research studies.
The category of family and services formed only 4% of research activity into autism during the last five years.
This is a surprisingly low figure given the large number of topics covered by this category. For example, it
included studies investigating the effects of autism on siblings and parents, educational services, services for
adults, families’ experience of diagnostic and interventions services and economic and policy-based issues.
Research activity is dominated by research into the symptoms of autism, although research into
causes and intervention are also well represented.
A low proportion of research was being conducted into a wide range of family, service and policy
issues. Given the change in our understanding of the prevalence of autism considerably more
research into these diverse topics will be required.
Research priorities identified by people with autism and their families
The 123 parents, professionals and people with autism who responded to our survey identified 248 priority
areas for future research (as individuals could prioritise more than one area). Only half of the parents
identified a country of origin. Of those who did, nearly one half were from the UK (44%) and a similar
proportion were from North America (45%). Only a few respondents identified themselves as being from non-
Figure 5 – Non-academic community identified priorities 4% Family & Services 32% Intervention 42% Causes 13% Symptoms 2% Epidemiology 7% Diagnosis
Figure 5 shows the distribution of the parent identified research priorities for future research. Almost half of
the respondents (42%) identified research into the causes of autism as a priority and one third (32%)
Research into symptoms of autism was identified as a priority by only 13% of respondents. Looking at the
subcategories of responses (data not shown), three-quarters of parents and professionals (75%) identified
research into environmental causes as a priority area and one quarter (25%) identified research into
When considering the types of intervention research, parents and professionals suggested that research into
biomedical interventions was the greatest priority (56%), followed by research into psychoeducational
interventions (39%). Only a minority of respondents called for more work into psychopharmacological
interventions (5%). Surprisingly, parents did not make research into families and services a high priority (4%).
Parents and professionals who responded identified research into the causes of autism as the main
priority, followed by research into interventions.
Parents particularly called for research into environmental causes and they did not see research into
psychopharmacological interventions as a priority.
Research priorities identified by researchers
Judging the representativeness of academics who responded to our survey
When we compared the countries of origin of autism researchers identified on ISI and the countries of origin
of the researchers who responded to our request, a fairly similar pattern emerged. Just under half of the
responses from researchers (45%) came from academics in North America, one quarter (24%) came from the
UK and 17% from Europe (not including the UK). Thus, there was a slightly higher percentage of responses
from the UK compared to the proportion of ISI-identified UK researchers (16%). This probably reflects an
English language bias and perhaps greater knowledge of and interest in the mapping project in the UK.
A very similar pattern was revealed, too, when we compared the research topics studied by those researchers
who responded to the total ISI breakdown of research activity (data not shown). At the very least, these
comparisons indicate that the responses received did come from a fairly representative sample of countries
and from researchers engaged in a fairly representative range of research activity. However, we are not able to
judge the representativeness of the respondents in comparison to all autism researchers worldwide in other
Researcher future priorities in comparison to current research activities
One comparison of interest is that between the current topics researchers were investigating and those they
identified as priorities for future research. This can be taken as one index of the academic community’s views
as to what shift in research focus is required in the field. Figure 6 shows the distribution of current research
topics of academics who responded to our survey alongside their suggested priorities for future research. The
186 respondents to our survey were classified as being active in 307 research areas (as individual researchers
could be active in more than one area) and identified 331 priority areas for future research (as individual
researchers could prioritise more than one area). Figure 6 – Researcher identified priorities
Current research area (N=307)
Future research priorities (N=331)
We expected that researchers would tend to nominate the area in which they themselves worked as a priority
area for future research, as this has an internally consistent logic in terms of motivation and interests.
However, there were a number of significant discrepancies between current topics studied and researcher
Most notably, researchers indicated that less research into symptoms of all types was needed in future than
they themselves were currently undertaking (27% compared to 45%) and that more research into causes than
they themselves were currently undertaking was required (29% compared to 18%). Less strong trends were
seen for researchers suggesting that more research on diagnosis and assessment (12% vs. 7%) and on
interventions (28% vs. 21%) was required than they themselves were currently undertaking. Researchers did
not identify research into families and services as a high priority (3%).
Researchers indicated that significantly more research into the causes of autism is required, as is
The researchers who responded to our survey indicated that less research into symptoms of all
types is needed in future than is currently being undertaken.
Comparison of ISI identified research activity, researcher and parent priorities
Figure 7 compares the distribution of current research activity indexed by ISI identified researchers to the
future priorities for research identified by scientists and by the non-academic community, respectively. In
order to make the comparison across these 3 sources of data simpler only the largest categories of research
are included, namely causes, interventions and symptoms. The remaining research categories are combined
into an ‘other’ category (epidemiology, diagnosis and assessment, family and services). Figure 7 – Comparison of ISI activity and researcher and non-academic community priorities
The pattern of findings revealed that neither researchers nor parents and professionals think that the current
distribution of research activity (by our ISI researcher index) into autism is correct. Both groups agreed that
more research into the causes of autism was required and that more research into interventions was also
Conversely, both researchers and parents indicated that less research into symptoms of autism was required.
In both of these patterns the shift from ISI to researcher priorities to parent and professional priorities
showed some linearity: researchers felt there should be more research than is currently conducted into causes
and parents/professionals felt this to a stronger degree.
The same pattern held, though less strongly, for intervention research. The reverse effect was seen for research
into symptoms where the difference in emphasis between researcher priorities and parent priorities was
stronger than that between ISI and researcher priorities. The only other more minor trend was that
researchers had a higher priority in the ‘other’ category. This was accounted for by a call from researchers but
not parents for more research on diagnosis. Figure 8 – Comparison of ISI and researcher and non-academic community priorities on causes Figure 9 – Comparison of ISI activity and researcher and non-academic community priorities on intervention
Researcher: biomedical ISI: psychopharmacological
Once researcher and parent/professional priorities were examined at the subcategory level it is clear that
there was somewhat less consensus than at first it appeared. However there was still considerable overlap in
some areas. Figures 8 and 9 show the same ISI research activity vs. researcher identified priorities vs. parent
and professional identified priorities for the subcategories of the causes and intervention categories. Figure 8
indicates that whilst researchers and parents agree that there should be more research into the causes of
autism they disagree on what likely cause should be investigated. Researchers’ priorities were more in line
with the current breakdown of research activity into causes, with 6 in 10 identifying aetiological research in
the area of genetics as a priority, although nearly 4 in 10 identified research into possible environmental
causes as a priority. This breakdown is reversed for parents who felt that environmental causes were the main
priority for future research over genetic causes (75% compared to 25%).
This pattern of findings might have been expected given the high profile given in the media and in the non-
academic community (e.g. on the internet and in parent organisations) to perceived but unproven associations
between autism and environmental agents including the MMR vaccination and thiomersal. However, it is
interesting to note that as many as 38% of scientists identified environmental causes as worthy of further
investigation. Whether this is based on a belief that the environment may be contributing to autism, or
whether it is out of a desire to disprove certain environmental hypotheses, is something that might be worthy
of further investigation through attitudinal research.
In the area of interventions, the overall agreement between researchers and parents that more research on
interventions is required also masks some differences of views at a more detailed level (see Figure 9). Parents
thought the level of current research (42% as indexed by ISI researcher activity) into psychoeducational
interventions was about right (39%) but researchers placed this as the main priority (64%) for research into
interventions. Researchers may be most aware of the dearth of well-controlled treatment trials of
psychoeducational approaches that could be addressed by future studies.
Conversely, whilst researchers identified research into biomedical treatments as a lower priority (12%) than
the current level of activity (24%), for parents it was the highest priority in the intervention category (56%).
This may relate to the above concerns of parents regarding perceived environmental causes of autism and
interventions relating to diet or secretin that have received wide publicity. A different picture again emerged
from the responses on psychopharmacological interventions. Researchers placed this as a priority slightly
below the current level of research in this area but few parents (5%) felt this was a priority. We do not know
if this was a general view or one formed by parents who responded following personal experience of
Whilst there was apparent agreement at the main category level between researcher and
parent/professional identified priorities, with both calling for more research into causes and
interventions, this disguised some divergence of views.
Researchers felt that the current balance of research into genetic and environmental causes was
appropriate, whilst parents argued for more research into environmental causes. However, nearly 4 in
10 researchers also identified research into possible environment factors as a priority.
Parents thought that research into biomedical interventions was a priority and did not favour
research into psychopharmacological approaches. Over half of researchers argued for more research
into psychoeducational approaches, as did nearly 4 in 10 parents.
The current exercise does not allow us to ascertain what underlies this divergence of views but this
topic could form the focus of a future qualitative and attitudinal research projects, for example via
discussion forums involving funding, researcher and non-academic stakeholder groups. It is
important to understand these differences to increase the consensus on research priorities.
From the searches and requests for information made a total of 46 UK funded research projects into autism
were identified. From some funding agencies full information for the 5 year period was not available. From
others only the individuals and institutions funded but not the topic of the research projects were available so
these projects were excluded. The total of 46 therefore, under-represents the true level of funded projects
Figure 10 – Breakdown of UK research funding 7% Family & Services 17% Causes 17% Intervention 4% Epidemiology 2% Diagnosis 53% Symptoms
Figure 10 shows the distribution of funded UK autism research projects in the years 1996-2000 across the
research category areas. The largest proportion of funding is concentrated within the symptoms category
(53%). Intervention (17%) and causes (17%) research are the next largest areas of autism research funded in
the UK. (See overleaf for funding comparisons with the USA.)
In order to contrast government (NIH) and charitable (CAN, NAAR) allocation of funds into autism in the USA
these figures are presented separately. Figures 11 (in terms of proportions) and 12 (in absolute numbers of
projects) compare the number of projects in each of the topic areas funded in the UK to the number of
projects in each area funded by the USA funding bodies. Figure 12 shows that both the USA charitable and
government sources funded a significantly higher number of projects than the UK funding agencies.
Figure 11 – Proportion of funded research projects UK vs USA charities vs NIH
UK (N=46)
NAAR/CAN (N=124)
NIH (N=437) Figure 12 – Number of research projects funded UK vs USA charities vs NIH
UK (N=46)
NAAR/CAN (N=124)
NIH (N=437)
Figure 11 shows that both the USA charities surveyed (40%) and NIH (26%) funded a higher proportion of
studies into the causes of autism compared to UK funding agencies (17%). Conversely, the proportion of
projects funded into interventions was greater in proportional terms in the UK than in the USA. This contrasts
to what is reported below: that there are proportionally fewer intervention researchers in the UK compared to
the rest of the world from the ISI researcher index. This discrepancy is accounted for in part by the
intervention studies and school/programme evaluations funded by The Shirley Foundation (3 out of 8 studies).
In terms of absolute number of projects funded (Figure 12) there was still very considerably more intervention
research funded by NIH than by all of the UK funding agencies combined. All UK and USA sources funded the
largest proportion of studies in the area of symptoms of autism, reflecting the predominance of this research
in our ISI index of research activity. The area of family and services represented a low proportion of funded
studies by the UK, USA charity and NIH funding agencies (varying between 3% and 7%).
Proportionally more research into the causes of autism is funded in the USA both by autism charities
Although proportionally the level of research activity on intervention in the UK appears to be
healthy, this in part reflects the effect of one charity only. In terms of the absolute number of
intervention projects funded the activity in the USA considerably outstrips funding in the UK.
Research into symptoms of autism continues to secure the greatest proportion of funding in the UK
The proportion of funded research in the area of family and services is low in both the UK and the
Country of origin of ISI identified autism researchers
Figure 13 shows the country of origin of ISI identified autism researchers. The largest proportion of autism
researchers originated in North America (the USA and Canada; 53%). ISI identified Europe (excluding the UK)
(22%) and the UK (16%) as the next largest places of origin of ISI identified researchers. Only a small number
of autism researchers from other continents were identified on ISI.
However, one important caution is that the ISI database has a clear bias to English-language research and
many and perhaps most non-English language academic journals are not included on the databases (except
where non-English articles are published on ISI with English Abstracts). For the purposes of the current
exercise we only had the resource to categorise English language abstracts and the breakdown of research
activity is at the level of ISI published research rather than a true reflection of the global picture. Figure 13 – Country of origin of ISI researchers 3% Australia 1% Middle East 1% South America 22% Europe 53% North US
Comparison of research activity in the UK vs. the rest of the world indexed by ISI identified autism researcher topics
In order to examine whether there are differences in research activity in the UK compared to other countries
whose researchers publish in journals included on the ISI database, we examined the same category
breakdown for UK based academics (N = 195) and non-UK based academics (N = 1,027). Figure 14 shows this
comparison.The proportion of research activity in the UK was fairly similar to that in the rest of the world
across the causes, diagnosis and assessment, epidemiology, and family and services categories, but some
differences in research activity in different areas emerged.
The most striking finding was this: in the non-UK research community (23%) compared to the UK research
community (8%) almost three times the level of research activity focused on interventions (in all intervention
subcategory types - psychoeducational, psychopharmacological and biomedical; data not shown). Other less
striking differences were that research into various symptoms of autism is higher in the UK than in other
countries (59% compared to 50%). The predominance of research into symptoms in the UK may in part reflect
the fact that some areas of cognitive neuropsychology of autism (e.g. theory of mind, central coherence)
originated from academics in the UK and that the UK is still strong in these areas (Medical Research Council,
One area of significant weakness in the UK is research into interventions. The proportion of
researchers evaluating interventions in the UK was a third of that in the rest of the world. This held
true across all types of intervention research.
The level of research activity in the UK is relatively strong, with UK researchers accounting for 16%
of active autism researchers on the main scientific (English language) databases. Figure 14 – Comparison of UK vs rest of the world research activity
Non UK (N=1,027)
UK (N=195)
from the mapping exercise with other recent reviews
The mapping exercise was undertaken to contribute to the dialogue between funding agencies, government
agencies, researchers and individuals with autism and their families regarding priority areas for research into
autism. The report is intended to be complementary to other recent summaries of the field†. However, it adds
Gathering views from scientists and people with autism and their families enables us to highlight
where agreement about areas of need exists and where opinions differ.
Information provided by research funding agencies provides for the first time a picture of the
distribution of research funding in the UK.
No previous reviews have systematically compared autism research in the UK to that in the rest of
Priority areas emerging from a literature review and suggestions from experts in the field in a
comprehensive range of research areas are collected together in one document. What are some of the most important findings to have emerged?
Proportionately less research into interventions (of all types) is conducted in the UK compared to the
rest of the world, although interventions were not under-represented in terms of funded projects
(largely due to the efforts of the Shirley Foundation)
Research into the causes of autism and research into interventions are seen as priorities both by
scientists and by the non-academic community. There are clear emerging themes in the literature in
both areas, for example the need to understand the contributions of both genetic and environmental
factors in the causation of autism and the need for more rigorous and well-controlled studies of
Some areas of disagreement or divergence emerged in terms of the views of parents/professionals
and researchers regarding the types of research into causes and interventions that they saw as
priorities. However, there was also considerable overlap in the priorities set by researchers and
The present exercise cannot identify the reasons for the divergence of views identified and these
should be explored in future attitudinal and qualitative research studies. Researchers and research
funding agencies need to consider the lessons from such research for the dissemination of research
findings and the influence this has on community opinions (for example through the general media).
Innovative approaches to some of these issues have been attempted in both the USA and Canada.
– The Canadian Autism Intervention Research Network (CAIRN; www.cairn-site.com) is a
partnership between researchers, parent organisations, professionals and policy makers. It aims
to develop research priorities, conduct evidence-based research and to communicate findings
†(Institute of Medicine, 2001; Medical Research Council, 2001; National Academy of Sciences, 2001; National Autism Plan for Children, 2003; Novartis,2003; Public Health Institute of Scotland, 2001).
(for example by providing web-based summaries of recently published research in an
accessible and authoritative form for the lay and professional community alike).
– In the USA an Interagency Autism Coordinating Committee (IACC) has been established to
coordinate autism research and other efforts within the Department of Health and Human
Services (DHHS). The National Institute of Mental Health (NIMH) has been designated the lead
agency for this activity. The IACC reports on its activities to Congress. The committee's primary
mission is to facilitate the efficient and effective exchange of information on autism activities
among the member agencies, and to coordinate autism-related programmes and initiatives.
Public members of the IACC helps bring to DHHS the concerns and interests of members of
Clear areas that were under-researched emerged. In particular there were a broad range of issues
regarding service delivery (and effectiveness), accessibility and choice, the effects on families and the
community in general, financial implications of the increased recognition of autism and the
subsequent demands on services. When autism was considered a rare condition such issues appear
largely to have been left off the research map but this has now changed with the increased
recognition of the likely true prevalence.
Efforts need to be made to encourage researchers in collaboration with research funding agencies,
families, professionals and individuals with autism to develop high quality research programmes into
the effectiveness and acceptability of services.
Another area that appears to largely have been overlooked and one that was not systematically
ascertained in our indexing of research activity was the lack of a solid research base with respect to
adulthood and autism. This may in part reflect views we received from the non-academic community,
the majority of whom may have been parents of children.
However, there is a clear consensus in the literature and from expert opinion that, despite isolated
examples of good practice, many worthwhile areas have been little studied to date. This by no means
exhaustive list includes: the epidemiology of autism in adulthood; the prevalence, identification and
treatment of associated (e.g. psychiatric) comorbidities; service use; employment opportunities and
development of appropriate living arrangements; life-stage transitions for individuals and families
and issues regarding quality of life and community integration.
Another output from this mapping exercise is a more practical opportunity for the field. The use of
new technology (the internet, electronic databases, email etc.) was important in allowing us to
efficiently and systematically combine information and to contact a wide number of individuals and
At the moment no overall overview is held about research activity in the autism field. Using current
information technology, this mapping exercise has demonstrated that it is possible to identify the
active autism research community in the UK. There may be opportunities in the UK to use such
technologies to monitor and systematically review research into autism in a number of ways. The
– The establishment of a UK autism research database or email network.
– Ongoing monitoring of research activity in the UK as indexed by funded projects or research
outputs in terms of published peer-review papers.
– Collation of information from both government and charitable sources of research funding in the UK.
The findings of this report will prove to be a useful guide to funding agencies and government departments
with respect to current strengths and weaknesses of the UK research field. It will help them identify emerging
areas of research in which academics and those affected by autism see likely benefits, in order to set effective
funding priorities. Clearly, no one source of information can determine future research priorities. However, the
strategy of combining information from different sources and discussing the findings of the mapping exercise
with the different stakeholder groups should help the decision-making process.
We are encouraged that many of the conclusions from the mapping exercise are in agreement with other
recent reviews prepared by government departments and agencies, as well as researchers, in the UK and
This newly documented information will contribute to the dialogue between research funding agencies,
scientists and people with autism and their families about future research goals. The report also contains
important conclusions about the dissemination of research findings and the translation of knowledge into
References
Institute of Medicine Immunization Safety Review Committee (2000) Immunization Safety Review: Measles Mumps Rubella Vaccine and Autism. Chicago, IL: American Medical Association.
Medical Research Council (2001) MRC Review of Autism Research - Epidemiology and Causes. London: MRC.
National Initiative for Autism: Screening and Assessment (NIASA) (2003). National Autism Plan for Children. Report produced in collaboration with theRoyal College of Paediatrics and Child Health, the Royal College of Psychiatrists, and the All-Party Parliamentary Group on Autism (APPGA). London: NAS.
National Research Council (2001) Educating Children with Autism. Committee on Educational Interventions for Children with Autism. Division ofBehavioral and Social Sciences and Education. Washington, DC: National Academy Press.
Novartis (2003). Autism: Neural basis and treatment possibilities, Novartis Foundation Symposium 251, (pp. 10-19). Edited by M. Rutter. Chichester: Wiley.
Public Health Institute of Scotland (2001) Autistic Spectrum Disorders - Needs Assessment Report. NHS Scotland.
Wing L & Potter D (2002) The epidemiology of autistic spectrum disorders: Is the prevalence rising? Mental Retardation and Developmental DisabilitiesResearch Reviews, 8, 151-161.
World Health Organisation. (1993). Mental Disorders: A Glossary and Guide to their Classification in Accordance with the 10th Revision of theInternational Classification of Diseases: Research Diagnostic Criteria (ICD-10). Geneva: WHO.
Appendix AScientists who commented on the literature reviews or other aspects of the report
For the purposes of the literature review, within each category area feedback and comments were solicited
from senior academic researchers in the field. The reviewers were asked to consider the following in their
If they agree with the suggested future directions
If further future directions should be added.
Other scientists were asked to provide comment on early drafts of this report.
However, the content of the literature reviews and the summary recommendations contained in this report –
and any inaccuracies or biases represented therein – represent the views and opinions of the authors of this
report. We are grateful to the following scientists who found time from their busy schedules to contribute to
Prof. Simon Baron-Cohen, University of Cambridge
Prof. Deborah Fein, University of Connecticut
Prof. Susan Folstein, Tufts University, Boston
Prof. Eric Fombonne, McGill University, Montreal
Prof. Catherine Lord, University of Chicago
Prof. Joseph Piven, TEACCH, University of North Carolina, Chapel Hill
Prof. Isabelle Rapin, Albert Einstein College of Medicine, New York
Prof. Sally Rogers, MIND Institute, University of California Davis
Prof. Sir Michael Rutter, Institute of Psychiatry, London
Prof. Paul Shattock, Sunderland University
Prof. Tristram Smith, Washington State University
Dr. Philip Whitaker, Leicester Educational Psychology Service
Dr. Lorna Wing, National Autistic Society
Recent literature was reviewed using the same categorisation employed for the ISI (Institute for Scientific Information®)research activity, and researcher and parent views on priority areas. The exercise was not a systematic review but ratherinvolved careful reading and integration of recommendations in recently published reviews, as well as reading of the mostimportant original sources. The topics covered overlapped with other recent state-of-the-science reviews cited in the mainreport. For the purposes of the mapping exercise, the focus was on identification of consensus views regarding priority areas forfuture research, likely to bring significant gains in knowledge and understanding. Some of the recommendations involved theextension and re-focusing of ongoing research enterprises, whilst others involved more ‘high-risk’ or ambitious but potentiallybeneficial research designs. Within each category area, feedback and comments were solicited from senior academicresearchers in the field (see Appendix A). However, the content of the summary recommendations contained in this report –and any inaccuracies or biases represented therein – represent the views and opinions of the authors of this report. 1. Causes/aetiology a. The genetics of Autism Although it is well established that autism is highly heritable, it appears likely that several overlapping sets of predisposing genes result in overall susceptibility in a complex and as yet not understood mechanism1,2,3. Whether (and which) environmental risk factors contribute additional susceptibility remains to be determined. A small proportion of cases of autism are due to single gene disorders and chromosome abnormalities. Linkage, association and candidate gene studies have not yet led to the identification of a susceptibility gene. Priorities for future research:• Further progress will be made in the genome-wide search for susceptibility genes by further delineation of the ‘core’ and
broader phenotype and novel approaches to testing for susceptibility for specific components of the autism phenotype inisolation (e.g. language delay, regression, insistence of sameness, associated abnormalities e.g. sensory responses, head circumference).
• Similar strategies need to be employed across a range of familial designs, including large proband studies in combination
with parent and sibling studies, studying multiplex families and the rare large kindred lineages that have been identified.
• Statistical power to accurately identify susceptibility loci will be improved by meta-analysis or combination of the
• The potential to test for the presence of a high-risk genotype could significantly influence genetic counselling to
• Once susceptibility genes have been identified, identification of the downstream biochemical processes, their effect on
neuronal migration and delineation of the pathway to altered central nervous system development and the known neuropathophysiology of autism will begin.
• Developmental (as opposed to static) animal models of autism can then be developed. • Genetically sensitive research designs that control for genetic effects will be necessary to investigate environmental risk
factors that may be associated with autism. b. Environmental risk factors Whilst many environmental factors have been suggested in the scientific literature and amongst the lay community to be associated with autism (implicitly the suggestion is that this association is causal) the empirical evidence-base is at present insufficient to draw firm conclusions4. Certainly, to date only rare environmental risk factors have clearly been demonstrated to play a role in the pathogenesis of autism (e.g. congenital rubella, perinatal cytomegalovirus5). Critical methodological issues limit the conclusions that can be drawn. These include: sample bias, lack of empirical data to support key steps in the proposed explanatory argument, lack of appropriate controls, poor diagnostic information, small sample sizes, lack of details about laboratory methods and unconfirmed measurement of purported abnormalities6,7. However, as a consequence of significant public concern regarding several of these purported mechanisms (e.g. MMR, thiomersal) research is ongoing that may provide more secure answers. Priorities for future research:• The possible role of environmental toxins and variations in neuropeptides and the functioning of the immune system in
brain development and outcome should be explored using animal models.
• Reports of gastrointestinal abnormalities should be independently replicated and the relevance of the findings to the
broader population of individuals with autism studied using non-invasive techniques in epidemiological samples.
• The effectiveness of possible interventions including gluten- and casein-free diets and immunnomodulatory therapies
should be tested by rigorous randomised control trials. One obstacle to this research will be the identification of the subgroup of individuals for whom these therapeutic approaches may be helpful (and ethical given potential side effects). However, as many parents opt to take up diet regimes it may be possible to use randomised control research designs (e.g. crossover placebo or waiting-list) designs using ‘opt-in’ convenience samples.
• High quality investigation of many putative factors is required in order to build a secure evidence base. This should take
place in nationally recognised laboratories using large, representative, well-diagnosed and described sample, with appropriate controls. 2. Prevalence and epidemiology It is clear that autism is more common than was previously recognised. There is agreement that changes in the conceptualisation, content and application of the diagnostic criteria (especially to individuals with average or above average IQ), and more rigorous methodology in recent prevalence studies, are likely to account for all or a large part of the putative increase8,9,10. However, at this point in time a real increase of some size cannot be definitively ruled out. One difficulty in interpretation of the literature is that different methods of counting cases (prevalence figures, service use databases and registers) have been used to argue for apparent increases in incidence over time9. The majority of studies have examined prevalence in child and adolescent populations, taking advantage of the integrated school and child healthcare systems. Consequently, little is know about prevalence in adults. Priorities for future research:• In depth, prospective studies of total population cohorts from birth are required in order to marry prevalence data with
onset and disease course data. However, the relative rarity of autism may make this prohibitively expensive.
• Future prevalence studies should adopt a symptom as well as a syndromic approach. • Dimensional or trait approaches to identifying the prevalence of individual characteristics of autism (e.g. the broader
phenotype of social impairment) in populations might complement case-based studies.
• When biological or genetic markers for a subgroup of individuals with autism are identified these should be included in
prevalence designs. This is critical since autism is not a unitary disease entity or disorder but an end phenotype of a number of complex, distinct and overlapping aetiologies.
• The nesting of biological and genetic research designs within future epidemiological studies will provide information to
better answer questions about heterogeneity of presentation and aetiology.
• Studies examining prevalence across the life course should be planned, despite the significant methodological and
3. Diagnosis and assessment For many years the age of diagnosis of autism was unacceptably late. However, progress has recently been made in the earlier identification of children with autism and many children are now first identified in the pre-school period11,12. Whilst it has been possible to pick up unrecognised cases using screens, no instrument has yet proved sufficiently robust to recommend universal screening13,14. However, autism screens can play an important role in improved surveillance for autism and other developmental disorders. A multidisciplinary approach to diagnostic assessment is required. The information necessary for a diagnosis includes a detailed developmental history, parents’ descriptions of the everyday behaviour and activities of the child, and direct assessment of the child's social interaction style, communicative and intellectual function. If possible the direct observation should include two contexts. The use of structured interview assessments (e.g. Autism Diagnostic Interview-Revised (ADI-R)15; Diagnostic Interview for Social and Communication Disorders (DISCO)16 and observation schedules (e.g. Autism Diagnostic Observation Schedule-Generic (ADOS-G)17, help systematise the range and depth of information collected. Developmental assessments are important to identify possible problems and assist in an accurate diagnosis. They also provide an objective description of a child’s abilities and deficits, which can be useful in deciding on appropriate intervention. There is an emerging consensus on the range of appropriate medical investigations, depending on presentation. Priorities for future research:• Further studies are required to develop and test screening instruments to determine whether universal population
screening for autism and other developmental disorders is feasible and appropriate. Whatever the outcome of these studies, more work will be required to determine how and to what extent such screening instruments can enhance ongoing surveillance to detect autism.
• Further work on the validation and limitations of formal diagnostic schedules including the ADI, the ADOS and the
• Earlier diagnosis and rising recognition as reflected in prevalence rates have significant implications for diagnostic and
• Primary healthcare practitioners including general physicians, health visitors, playgroup, kindergarten and nursery school
staff need to be made aware through training which early signs of possible autism warrant further investigation and what to do and say.
• Specialist child development services need to develop a referral strategy for autism and a multidisciplinary
approach to the differential diagnosis from other, often complex, developmental conditions (NIASA, 2002).
• In the UK, the National Initiative for Autism Screening and Assessment sponsored by the Royal Colleges of Paediatrics &
Child Health and Psychiatry12 should provide an impetus to greater provision of co-ordinated diagnostic, therapeutic and educational services for pre-school children with autism and their families. 4. Symptoms a. Neurobiological basis of autism Brain abnormalities in autism have been studied via post-mortem examination of brain tissue, structural and functional brain imaging techniques, neurophysiological, neurochemical assays and animal models. Although it is clear that brain development and organisation is abnormal in autism, the pattern of findings across these diverse methodologies is complex, and several potentially important findings have either not been replicated or have been inconsistent across studies, although this may reflect differences in the techniques used and samples studied. Post-mortem studies conducted to date have identified several abnormalities: brain weight is increased in the majority of children (supported by structural imaging findings) and there are decreased numbers of Purkinje cells18,19. Findings from functional imaging studies with typical adults and adults with autism have identified abnormalities in the amygdala and fusiform face processing areas20,21. Important clinical signs associated with autism, particularly the loss of language skills seen in the one third of children who regress, are not well understood22. Several abnormalities in neurotransmitter systems have been reported, including serotonin and the cholinergic and the GABAergic systems, although again findings across different studies are sometimes inconsistent23. Experimental lesion studies in non- human primates have provided leads regarding the possible role of brain structures (e.g. the medial temporal lobe, the amygdala) in autism pathology24,25. Priorities for future research:• There is a need for further co-ordinated work to produce replicable findings that will produce clearer and better-
established models of the brain pathology that underlies autism. In particular attempts should be made to corroborate findings using functional, structural and post-mortem studies.
• Post-mortem studies should include the study of neurotransmitters and well as other neuropeptides and neurotrophins
• Further elucidation of neuropathology in autism must be linked with findings in genetics, the identification of possible
environmental causes and psychopharmacological approaches to intervention. Specifically the identification of genes important in the development of specific brain regions or neuronal cell population that are abnormal in autism should be a target.
• Future functional and structural imaging studies need to include sufficient sample sizes in order to relate findings to
participant characteristics such a as age, IQ, symptom profile etc. This has implications for the study of young children and the need for automated methods.
b. Neuropsychology Abnormalities in brain structure and function of individuals with autism affect behaviour through the abnormal development of psychological functions. Three main psychological theories have been advanced that attempt to explain the nature of the psychological symptoms of autism. These theories focus on social understanding (theory of mind - ToM), control of behaviour (executive function - EF) and detail-focus (central coherence - CC). A substantial number of studies have demonstrated that individuals with autism are impaired in theory of mind, or mentalising, abilities and more recently brain imaging studies have explored the neural basis of these cognitive processes26,27. Psycho-educational interventions to circumvent mentalising impairments have been piloted. Deficits in emotion recognition and response may be related to mentalising deficits, perhaps sharing in common some social orienting or social-affective reward system impairment28. Within the executive, or action control domain, individuals with autism have most difficulty with set shifting and inhibiting prepotent response, where working memory and inhibitory control are simultaneously required29. Our understanding of ‘weak central coherence’ in autism is less well developed, and findings from experimental studies more inconsistent, and further work to determine possible low-level perceptual processing and attentional problems in order for a more specified account of the CC theory to be developed is required30. It is as yet unclear how ToM deficits, problems in EF and the tendency for weak CC relate to each other, both online and over development. Several studies have investigated multiple cognitive processes longitudinally in young samples of pre-school children with autism and to combine neuropsychological studies with functional imaging studies of young children may allow us to identify how differences develop over time in the organisation of brain systems that regulate social understanding, emotion
recognition, control of behaviour and perception28. Novel experimental methodologies such as eye tracking of a participant's gazewhilst watching a scene in a movie have allowed us to 'see the world through the eyes of individuals with autism31. Priorities for future research:• Neuropsychological studies should include measures of more than one domain of function and explore individual
differences in heterogeneous groups of individuals with autism and not aim solely to identify autism-pecific impairmentsin one domain of information processing.
• At both the cognitive and brain level autism is by definition a developmental disorder whose presentation changes over
time. Combining sophisticated neuroimaging techniques with experimental neuropsychological methods in longitudinal designs will allow to us to examine how the organisation and function of different brain systems develops over time.
• Most studies have investigated cognition in isolation and few have attempted to establish associations between
cognitive processing impairments and behaviour and adaptation, despite the fact that these are critical to identifying priority areas within the broad cognitive domain for future research and practice.
• In addition to mapping back from the psychological level to the brain system level, further attempts should be made in
all areas of neuropsychological investigation to develop intervention strategies that will either allow individuals with autism to part-compensate for constitutional impairments or to develop alternative means to develop adaptive social and behavioural responses and understanding. c. Behavioural symptoms Individuals with autism show a range of behavioural difficulties. These can include aggressive outbursts and other challenging behaviours and self-injurious behaviours, as well as rigid and ritualistic behaviours in relation to feeding and toileting32,33. Although some behaviours such as stereotypic body movements (e.g. rocking) are common to individuals with autism and individuals with severe developmental delay, others – in particular the difficulties encountered due to the preference for rigid routines and rituals – are more specific to autism. There is also evidence that sleep disorders are more common for individuals with autism34. The most common approaches to managing behavioural difficulties are behaviour modification procedures that are based on the premise that these are learned behaviours maintained by operant contingencies. Studies have demonstrated a moderate to high degree of success in reducing stereotyped motor movements, self-injury, or repetitive language35. Stimulus- based approaches that involve altering antecedent events to problem behaviours are now more commonly employed. Interventions developed from functional assessments appear to be more likely to result in significant behaviour reduction36. There is evidence that dopaminergic, serotonergic and opiate drugs therapies can diminish repetitive behaviours disorders, although significant potential adverse side-effects require close monitoring especially in individuals with low IQ and limited communication abilities (see Psychopharmacology section). Priorities for future research:• A large scale, epidemiologically-based study is required to determine the range, frequency and severity of behavioural
difficulties in autism and their association with severity of autism and developmental delay.
• Further evidence is required to determine the efficacy of behavioural and psychopharmacological interventions for sleep
problems, including the use of melatonin.
• Further research is required to determine the most effective combinations of behavioural and psychopharmacological
interventions for severely challenging (aggressive and self-injurious) behaviours.
• The development of cognitive-behavioural interventions to manage rigid and inflexible behaviours in individuals with
autism of average intelligence is required. d. Associated medical and psychiatric conditions Associated medical conditions have been found in between 6% to 10% of cases of autism, with higher figures for more developmentally delayed individuals37. The most common disorders include the genetic conditions fragile X syndrome and tuberous sclerosis, both accounting for approximately 1% to 2% of cases of autism. Other disorders identified in individuals with autism include phenylketonuria (PKU), Prader-Willi and Angelman’s syndrome, neurofibromatosis, William’s syndrome and cerebral palsy. Epilepsy occurs in one third of cases of autism with a bimodal onset distribution with peaks in early childhood and later adolescence38. Subclinical seizures – identified using EEG – are also found in some individuals but the treatment implications of these remain controversial. Recommendations for standard clinical assessment protocols have been developed39. Individuals with autism are at increased risk of developing psychiatric disorders, most commonly depression and anxiety40,41. Obsessive symptoms and tic disorders, Tourette's syndrome and hyperactivity are also associated with autism. Schizophrenia and psychotic disorders appear not to be more common than in the general population. Ascertainment and referral bias, the limitations of individuals with autism to report subjective mental state symptoms and the hierarchy rules adopted in the ICD
and DSM classification systems have limited the certainty and generalisability of these findings. The association betweenautism and psychiatric conditions has implications for aetiological models of autism at the genetic, neurophysiological andneurotransmitter/neurochemical levels. Both psychopharmacological and cognitive behavioural therapeutic approaches can playa role in the treatment of comorbid psychopathology (see Section 5).
Priorities for future research:• Systematic examination of whether there is an aggregation of comorbidity between autism and other medical conditions
beyond that associated with developmental delay would require considerable investigation of a large population cohort and the scientific yield might be only modest.
• Promising, specific associations – such as that found between early onset temporal lobe epilepsy (rather than the
presence of temporal tubers per se) and autism in individuals with tuberous sclerosis – may yield greater potential breakthroughs in terms of understanding pathogenesis than large population-based investigations.
• Future studies should use appropriate methodologies to ascertain accurate rates of psychiatric disorders for adults and
children, and for individuals with developmental delay and those of average intelligence.
• Clinical services should adopt more systematic approaches to the identification and recognition of psychiatric disorders,
as well as developing and evaluating appropriate intervention strategies, both pharmacological and psychological.
• Studies should investigate how psychiatric disorders in individuals with autism are related to similar disorders in their
relatives, including exploration of the potential shared neurobiology/neurochemical and genetic basis. 5. Interventions a. Psychoeducational There is increasing evidence that appropriately targeted intervention improves outcome in children with autism42. Evidence includes a large number of individual case and small case series studies, providing support for behavioural and some psychosocial interventions, targeting language, behaviour, play and social skills43,44,45. In contrast there is a relative paucity of well-controlled empirical group evaluations of intervention programmes for children with autism and even fewer randomised controlled trials42. The strongest evidence is for early, intensive behavioural interventions that have led to gains in IQ and language ability46,47. Preliminary findings from ongoing studies using education-based approaches48 have also demonstrated positive outcomes in terms of IQ gains and reductions in symptom severity. Intervention approaches that place an emphasis on the development of non-verbal social-communicative skills have also provided promising data49,50. Direct empirical evidence that early compared to later intervention has a specific positive benefit is not yet available. However, there is a consensus that developmental principles support the notion of early intervention. Regardless of the underlying approach, there is a consensus that, along with structure and an emphasis on developing communication skills, children with autism should be enrolled into programmes as early as possible42,51,52. Priorities for future research:Reports that followed recent expert meetings in the UK53 and USA54 set out research agendas to improve the evidence-base regarding psychoeducational interventions for children with autism. Amongst the priorities identified were:• One methodological issue is the choice of appropriate outcome measures, measures of treatment fidelity, identification
of the necessary and effective elements (and intensity) of multi-modal treatment approaches, and identification of moderating and mediating variables to determine ‘what works for whom’ (and why).
• Although no design other than a randomised control trial can avoid bias resulting from unmeasured confounding factors,
appropriate research strategies can take a variety of different forms and should relate to the stage of development of the evidence-base for any particular approach.
• It is essential to develop a framework within which randomised trials can be successfully conducted. This will involve
consideration of patient preference trials (e.g. stratification of randomisation), access to alternative treatments that all offer something of perceived potential value to parents and children, and the development of effective strategies for persuading families of the importance of randomisation techniques. Cost-effectiveness, health economic and quality of life measures should be included in such trials.
• There is a need for increased research into psychosocial (e.g. social skills training) and psychotherapeutic (e.g. cognitive
behaviour therapy) interventions for children, adolescents and adults with autism, in particular to increase social competence, integration and stability, and to decrease psychiatric morbidity. b. Psychopharmacological Although a wide range of psychopharmacological agents has been employed in the treatment of children and (more frequently) adults with autism, for the most part they have not been used to treat the core features of social and communication impairments. Rather, they are given to ameliorate associated symptomatology, including, poor attention and concentration,
morbid or unusual preoccupations, obsessive and compulsions or rituals, stereotyped behaviour and self-injury, excessiveanxiety, depressed mood, sleep problems and tics55,56. Amongst the best-supported drugs are those that target serotonergicfunctions (mood, obsessional and repetitive behaviours), including fluvoxamine and sertraline57,58. Methylphenidate andrisperidone have been shown to have some success in targeting hyperactivity and aggressive behaviour, although as with manyother drugs, side effects (in particular weight gain) have been noted59,60.
Priorities for future research:• One important unresolved question is the use of psychostimulant medication to reduce inattention and overactivity in
the face of clinical reports of increased agitation, aggression and insomnia. Large, double-blind, placebo-controlled trialsare required in order to determine optimum dosage and predictors of treatment response and adverse effects.
• Novel medications such as mood stabilisers (e.g. lithium) and beta-blockers require further investigation using pilot case
series trial and open label trials to test potential effectiveness.
• Longer-term trials of atypical antipsychotics and SSRIs are required to extent positive findings from shorter-term trials
and to gather longitudinal safety data.
• Specific efforts need to be made to develop and employ medications with fewer serious adverse side effects that cause a
significant proportion of individuals to terminate medication. c. Complementary approaches Parents of children with autism often turn to alternative, or complementary, treatments (those that fall outside of mainstream, statutory provision) in the hope of alleviating some of the symptoms of autism. Although much of the research published to date suffers from methodological limitations, including lack of control, small sample size, lack of independent replication, findings appear to fall into two groups. First, complementary treatments for which cases series and controlled and/or uncontrolled trials provide some indication of positive benefits. Included in this group are vitamin B6 and magnesium supplements61,62, melatonin for sleep disorders63 and casein and gluten-free diets64. Further study, where possible employing blinded randomised controlled trials, of each of these interventions is warranted. The second group is a set of interventions that have been sufficiently rigorously investigated and found to be ineffective. These are now de-recommended by most mainstream professional bodies65,66. They include auditory integration training (AIT)65, facilitated communication66,68 and secretin67. Parents need to be made aware that despite positive claims by proponents, anecdotal evidence of successes for individual children, and the willingness of parents to take-up these treatments, many have little or no firm scientific evidence to support them. Priorities for future research:• Further study of complementary treatments with some evidence of benefit (B6 and magnesium supplements, melatonin
for sleep disorders, and casein and gluten-free diets) is warranted, where possible employing blinded randomised controlled trials.
• Studies should focus on target treatment outcomes that are directly empirically related to the proposed mechanism of effect
(e.g. reduced hypersensitivity to noise in AIT; bowel problems in gluten- and casein-free diets). Such outcomes are more likelyto be revealing about the underlying mechanism of any demonstrated treatment effects and less likely to be spurious.
• Interventions should only be attempted under medical supervision to monitor potential adverse reactions. 6. Families and services Relatively little research has been conducted into family function in families with a child with autism. The majority of the work that has been conducted has studied issues regarding the diagnostic process, access to services and the psychological well- being of parents and siblings of children with autism69,70. Guidelines for the development and organisation of service involved in the diagnostic process and intended to ensure that the level of stress experienced by families is minimised have recently been published in the UK12.
A recent report by the Public Health Institute of Scotland71 assessed the needs of individuals with autism and proposed that ‘ideal’ services for these individuals should aim to deliver:• Joint assessments, involving relevant agencies, services and professionals• Active involvement of the family• Early identification• Early intervention• Provision of a range of services – developed in a multi-agency and seamless way to ensure the range of needs of people
• Sensitive management of the transition between childhood and adulthood• Individual needs of children and families should be addressed
Overall, the Scottish review noted that there was a lack of specialists working with individuals with autism (see also USANational Research Council recommendations for training). Similarly, when reviewing services in the whole of the UK, Howlin72found that provisions for individuals with autism depended on where families lived and on what services and schools wereavailable in the particular area, rather than on research-based and needs-based evidence.
Adults were particularly poorly served in terms of both diagnostic services and post-diagnostic support. A recent NationalAutistic Society survey73 found that many adults were not able to access the social services available to them. Some familieswere not being offered services as individuals with ‘high-functioning autism’ and Asperger syndrome were not considered‘disabled’. Half the adults in the sample were still living at home with their parents, often because the support available tothem outside of their home was inadequate. Many adults with high functioning autism or Asperger syndrome would have beenable to live independently if they received adequate support. The annual societal cost of individuals with autism in the UK hasbeen estimated to exceed £1 billion (based on a prevalence of 5 per 10,000), whilst the lifetime cost for a person with autismhas been estimated to be greater than £2.4 million74. Enhanced provision of services is likely to have cost benefits to society asa whole and the impact of new services should be tested via health economic modeling.
Priorities for future research:• Little is known about family relations and family stress beyond childhood and further research with families with an
adult with autism is urgently required.
• Few studies that have attempted to identify family protective factors or beneficial treatment and support approaches. • Future work should identify child and family characteristics that make families most vulnerable and the appropriate
health, education, social services and other support services to ameliorate dysfunction and poor coping.
• Further work on the economic costs to families and to society of autism are required in order to develop cost-benefit
models of intervention programmes and support services.
• Very few services for adults with autism have been evaluated. Once an evidence-base for models of good practice exists,
training schemes to disseminate such knowledge should be developed and evaluated. References 1 Folstein SE & Rosen-Sheidley B (2001) Genetics of autism: Complex aetiology for a heterogeneous disorder. Nature Reviews Genetics, 2, 943-955. 2 Jones MB & Szatmari P (2002) A risk-factor model of epistatic interaction, focusing on autism. American Journal of Medical Genetics, 114, 558-565. 3 Lauritsen M & Ewald H (2001) The genetics of autism. Acta Psychiatrica Scandinavica, 103, 411-427. 4 Rodier PM & Hyman SL (1998) Early environmental factors in autism. Mental Retardation and Developmental Disabilities Research Reviews, 4,
MOSAIC Meeting Monday, April 2, 2007 New Introductions-Welcome Kim, Michelle and Connie II The Whole Soy Story- We will probably be hearing and reading more about genetically modified food. The thought is genetically modified soy (does this include all soy even not genetically modified?) increases estrogen and as a result, there may be a link to cancer. Did someone say there was a connect
Report on Availability of essential medicines in PHCs of 5 districts of Maharashtra During the 4th phase of Community based monitoring (CBM) in Maharashtra, availability of essential medicines in PHCs has been assessed quite extensively. Availability of 28 essential medicines was assessed in the 4th phase of CBM (data collected in the period Sept. 2010 to March 2011). All selected medic