Andrew Whitehouse, The University of Western Australia

“273% Increase in Autism and We Don’t Know Why!” Thursday, April 15, 1999, Los Angeles Times.

The rising prevalence of autism is a story destined to achieve headlines, take up column inches and even bump MasterChef from the top of the water-cooler gossip list. The headline above, which appeared in the influential Los Angeles Times, is perhaps the one that started off the frenzy of intrigue.

So, is there a rise in the number of children with autism? And if so, what is causing it?

The first question is by far the easier. Yes, there has most certainly been a steady rise in the incidence (number of new diagnoses per year) and prevalence (total number of affected individuals in the total population) of autism during the past half-a-century.

The finding is well-replicated and has been observed in every country (including Australia) that has an appropriate data source to tap.

The first survey was conducted in the 1960s and produced a prevalence estimate of one individual with autism in every 2,500 people or 0.04% of the population. When I first started my research in this area in the early 2000s, the oft-quoted figure was one in every 250 people (0.4% of the population).

In the last decade, studies have seemed to continually outdo each other with higher and higher prevalence figures, ranging from the 2005 Australian figure of one in every 160 people (0.62% of the population), to a recent South Korean study, which found an astonishingly high rate of one in every 38 people (2.6%).

Of course, the true prevalence is likely to be somewhere in the middle. Currently, the most widely recognised estimate is around one individual with autism in every 100 people (1% of population).

A 25-fold increase in any diagnosis in the space of 50 years is, to put it mildly, rather worrying. But how alarmed should we be? The answer to this is related to the second question posed above: what is causing the increase in autism?

Many researchers started to explore whether there was something about the modern environment that may be causing the increase in the number of children with autism. This fire was stoked by British medical researcher Andrew Wakefield in his now infamous 1998 press conference, in which he linked the measles-mumps-rubella (MMR) vaccine to the onset of autism.

Despite the heaving amount of evidence against the MMR-autism link, and the retraction of the paper on which his tenuous assumptions were based, the drop in MMR vaccination rates triggered by this scandal remains a major cause of preventable disease worldwide.

What, then, are the possible causes for the increase in autism? The reasons are likely to be many and varied, and this is what we have come up with thus far:

Conceptual change

One of the most important discoveries in autism research over the past two decades has been that the syndrome varies along a spectrum of severity. Some 20 years ago, autism was called Infantile Autism and only diagnosed when a child demonstrated “gross deficits in language development” and “a pervasive lack of responsiveness to other people”, prior to 30 months of age.

But we now recognise that children can present with less severe autistic symptoms, which are often difficult to identify at such a young age. This understanding led to the formulation of new diagnostic categories – Asperger’s Syndrome and Pervasive Developmental Disorder – Not Otherwise Specified (PDD-NOS), both of which come under the collective banner of Autism Spectrum Disorders.

The change from Infantile Autism to Autism Spectrum Disorders has meant a greater number of children with a more diverse range of autistic symptoms are being included in prevalence counts.

But, intriguingly, studies have found the rate of the more severe form of autism (“Infantile” or “Classic” autism) has also risen over the past two decades, which suggests the broadening of the diagnostic boundaries cannot account for the entire increase in prevalence.

Diagnostic substitution

The expansion of diagnostic boundaries has meant individuals who would previously have been placed under a different “diagnostic banner” are now more likely to receive a primary diagnosis of autism. This is particularly true for the diagnoses of language disorders and intellectual disability, and the prevalence of these conditions has decreased over the past two decades as autism diagnoses have increased.

Methodological differences

As knowledge about autism has advanced, so have the techniques used to measure the prevalence of the condition. Early prevalence studies examined populations considered small by epidemiological standards (e.g. health districts), and used relatively crude diagnostic tools.

The modern method of screening whole populations (e.g. states or countries) using improved diagnostic assessments is likely to be more sensitive in identifying affected individuals.

Socio-cultural influences

Certain societal influences make it more likely for an individual to be diagnosed with autism today than in the past. These include:

• increased awareness and understanding of autism among parents and health professionals
• the formation of specific autism diagnostic teams
• a lessening in the stigma associated with a diagnosis (particularly, the dispelling of the myth that autism is caused by “cold parents”)
• the availability of governmental assistance specific to children with an autism diagnosis.

These potential explanations assume there has been no difference in the “true” rate of autism over the past 50 years; rather, what has changed is the way that we conceptualise and measure the condition.

But one should always be cautious about adopting a position based on null findings. It’s quite possible the true prevalence of autism is increasing, and we just haven’t yet identified the cause.

One potential candidate is the increase in the survival of extremely premature infants. While recent findings suggest these infants are at increased risk of autism, research in this area is at too early a stage to make conclusive statements.

What we can conclude from research to date is that there is no single environmental factor we know of that has substantially contributed to the increase in autism diagnoses.

Research into Autism Spectrum Disorders is at an exciting stage, where we can examine how genes and the environment combine to cause this condition.

If the next 20 years of research is half as fruitful as the previous 20, it’s a good bet that we will get to the bottom of the supposed “autism epidemic”.

Andrew Whitehouse, Winthrop Professor, Telethon Institute for Child Health Research, The University of Western Australia

This article is republished from The Conversation under a Creative Commons license. Read the original article.