University of Gothenburg
Bar chart with autism rising, red
Photo: Imagentle/

The increase in autism prevalence

Sebastian Lundström's latest blog entry

[Posted on 29 August, 2017 by Sebastian Lundström]

Over the last decade we have seen a remarkable increase in the number of individuals diagnosed with autism. Studies from Europe, America and Asia indicate prevalence rates of around 2.5%. This is a fifty-fold increase compared with the first prevalence study that was carried out in the mid-1960’s. The primary reason behind this change is probably how the diagnostic criteria have changed. The different diagnostic systems (Diagnostic and Statistical Manual of Mental Disorder, also known as DSM, and International Classification of Diseases, abbreviated as ICD) have broadened the definition of autism in such a way that more people can be included under the autism “umbrella”.

During the mid-2000’s, a number of studies, some of which originated at the GNC, showed the prevalence of autism to be around 1%. There were no revisions made to the diagnostic systems during this timeframe and even so, the prevalence increased. Studies from the GNC have shown that there has been no increase in the symptoms that autism diagnoses are based on, but that clinical diagnoses of autism have increased remarkably regardless of this fact. How should we interpret these findings? There are a number of different factors that have all likely contributed to the increase, both individually and in tandem with one another:

(1) Autism is significantly more debated today than it was 10-20 years ago: above all there is a great wealth of (good and bad) information about autism online, but newspapers and TV discussion panels are also showing interest in the topic. More people thus become aware of the symptoms, which in turn leads to more people becoming inclined to seek out diagnosing facilities.

(2) Today we also have access to more knowledge about autism within the institutions that actually diagnose autism. Studies have shown a phenomenon referred to as “diagnostic substitution”, which means that a person who should have been given one kind of diagnosis at a given time would have been given a different kind of diagnosis at another time. For example, the increase in autism diagnoses has coincided with a decrease in different diagnoses of intellectual disabilities and learning problems. It is hard to determine exactly why diagnostic substitution occurs, but it is probably a combination of “different” diagnoses having many overlapping features, based on behavioural observations and scoring scales, and the current zeitgeist.

(3) The GNC has expressed concern that LSS (The Swedish Act concerning Support and Service for Persons with Certain Functional Impairments) might be misinterpreted. There is a concern that the legislation is applied in such a way that only individuals with the “correct” diagnoses (which at the moment include autism) have the right to be granted support. This is not unique to Sweden; in a newly published study from Australia, where clinicians were asked about their diagnostic practices, approximately 20% claimed that they had diagnosed people with autism even though these individuals had not reached the diagnostic threshold. One of the stated reasons was the desire to ensure the affected individuals get access to supportive measures. This phenomenon is further supported by data from a study currently nearing completion at the GNC, where we see that significantly fewer symptoms are needed to establish an autism diagnosis today compared to 10 years ago.

In conclusion, we can confidently state that the increase in the number of autism diagnoses that we have seen over the last decade can be explained through societal changes rather than an increase in underlying biological/environmental factors.

One might consider whether it is misleading to talk about ONE kind of autism prevalence when autism is likely the product of tens of thousands (a guess as good as any other) of different things and might manifest very differently from person to person. More accurate would likely be to talk about SEVERAL different kinds of autism prevalence. Given the large symptomatic overlap between “different” diagnoses, another way forward might be to not simply “count” symptoms or specific diagnoses but rather to quantify the dysfunction and suffering that these things bring about?

[This is a blog. The purpose of the blog is to provide information and raise awareness concerning important issues. All views and opinions expressed are those of the writer and not necessarily shared by the GNC.]