Is autism now more common?
Thirty years ago autism was thought of as a rare disorder. Nowadays, we hear something about this condition almost every week, and most of us either know someone who’s considered to be on the autistic spectrum or who has a relative with the problem. Does this mean that we are more aware of the condition, or has there been a real increase in the number of sufferers?
According to official statistics, the numbers have increased. The current edition of the Diagnostic and Statistical Manual of
Mental Disorders ( DSM-5, published in 2013) puts the prevalence of autism spectrum disorder (ASD) at about 1 per cent of the population. In the DSM-4, published 19 years earlier, the prevalence of autistic disorder, as it was called, was estimated at between 0.02 and 0.05 per cent. This represents a 20-fold increase. What could account for this?
First, there are the changes that have been made to the diagnostic criteria. Previously, someone would be diagnosed with autism only if they demonstrated a “marked impairment” in social interactions and communication as well as in language development, and showed a “marked restriction” in interests and activities. If they appeared to have severe difficulties in social interactions and extremely restricted interests, but showed no significant delays in language development, they were diagnosed with Asperger’s disorder, which was considered a separate diagnosis. In the DSM-5, these two conditions are amalgamated within autism spectrum disorder.
Although the prevalence of Asperger’s disorder is hard to estimate, I suspect this accounts in large part for the apparent rise in the number now diagnosed.
The age when symptoms must first be noticed has also been relaxed. The DSM
4 states that impairments must be obvious before three years of age, whereas the DSM-5 requires symptoms to be present “from early development”. The DSM-5 also allows for a diagnosis to be made retrospectively, when the sufferer is an adult. To achieve this, clinicians must rely on the memories of parents and other carers to determine when symptoms first appeared. This makes me slightly uncomfortable. There’s a great deal of research – particularly Elizabeth Loftus’s experiments in the Seventies concerning eyewitness testimony – that shows how malleable our memories can be, and how subject they are to error depending on our current beliefs and on the questions we’re asked. In another change, the
DSM-5 allows “subthreshold cases” to qualify for a diagnosis. This means that when a clinician is unsure whether someone meets the criteria for ASD, they’re now more likely to diagnose the condition.
Finally, there’s greater awareness of autism, so ASD symptoms are now more likely to be picked up. Therefore, it’s unclear whether the higher rates reflect a real increase in ASD, or whether the increase is down to a relaxation of the diagnostic criteria coupled with an increased awareness of the condition. In either event, these changes are a good thing if they allow us to identify sufferers earlier and if, as a result, we can offer help sooner.