Hello. My name is Eric Fombonne.
I'm a Child Psychiatrist working at
Oregon Health and Science University in Portland in Oregon in the US.
I am a Director of Autism Research in
the Department of Pediatrics and at the Institute for Development and Disability.
Today's talk will cover three main areas.
I want to review what we know about the most recent research on the prevalence
of autism spectrum disorders after having given a brief historical perspective,
then I will review the interpretation of time trends in the rates of autism.
So what do they mean? How can we understand them?
Finally, I'll give a brief overview of studies which are currently connected in
the world and give a summary of what the field
is at today in terms of international studies.
Throughout my talk, I will use ASD or autism as equivalent terms.
The epidemiology of autism started in the UK with
the first seminal study by Victor Lotter in 1966.
During that time, the studies were
employing a definition of autism which was very restrictive,
very narrow, and basically followed
the initial descriptions by Kanner of a syndrome which was quite severe clinically.
In those initial studies,
investigators employed Kanner's definition or ICD-9 definition,
and they were concentrating on a severe phenotype
usually diagnosed by local clinicians with expertise in the field.
There was no use of particular standardized tools to diagnose autism.
The way cases were ascertained in those studies was
basically relying on existing facilities,
both medical or educational,
and investigators tried to identify in
those facilities the children who were already diagnosed with autism.
In those first prevalence studies,
the range was between 4-10 per 10,000,
and it was at the time considered to be a rare condition.
Then in 1980's and in the 1990's,
the field evolved in the sense that the definition of autism broadened,
we used progressively a larger concept of autism.
Our diagnostic criteria were also adapted to these new ideas
and became much broader in their ability to capture children with autism.
We started to recognize that autism could occur in
children who had normal language development or near normal language skills,
and also had normal levels of intelligence.
During that time, we shifted from a conceptualization of autism,
which was extremely deviant from normal development to
a syndrome which had some degree of overlap with normal development.
For instance, IQ could be normal language could be normal,
and we employed increasingly a dimensional view of autism which was supported by
a new generation of diagnostic instruments such as
the Autism Diagnostic Interview or the Autism Diagnostic Observational Schedule,
the ADOS and the ADI,
which were developed at that time and were first
employed in the epidemiological study in the late 1990's.