Autism today is better recognized, correctly categorized, and autistics offered more accomodation/treatment options than decades ago when they were misdiagnosed then hidden away in institutions and asylums.
One of the ways to examine this question of “over-diagnosis” is to go back to the evolution of autism diagnosis itself.
In Chapter 1 “History and Evolution of the Autism Spectrum Disorders” (Chapter authors Julie K. Irwin, Jennifer MacSween, and Kimberly A. Kerns) of The International Handbook of Autism and Pervasive Developmental Disorders (Springer 2011 — http://www.springer.com/psychology/child+%26+school+psychology/book/978-1-4419-8064-9) the authors described the following historical issues:
Children exhibiting autistic behaviors/symptoms prior to the 1940s were thought to be / diagnosed with schizophrenia.
In 1943 Dr. Leo Kanner at Johns Hopkins published “Autistic Disturbances of Affective Contact”, and labeled children he was observing with similar symptoms as “autistic” [the word autism was rooted in Greek: autos + state/action = autocentric thinking and behaviors that appear to withdraw from the social world]. Autism came from a Swiss psychiatrist, Bleuler who also coined “schizophrenia.”
In 1944 a German psychiatrist named Asperger used the same autistic label to describe children exhibiting similar symptoms and noted these children exhibited different symptoms than disintegrative personalities seen in schizophrenia.
In 1979, a UK-based epidemiological study of children under age 15 yielded the “triad of impairments” that we are familiar with today in lay language: (1) deficits in social communication, (2) deficits in verbal and non-verbal communication, and (3) deficits in symbolic play.
Thus based on the above:
The appearance of over-diagnosis in autism may be attributed to:
Our ability to diagnose autism more effectively today (we have more data and observations collected since the 1940s) than before.
Our awareness of autism symptoms that allow us to “find” children who may be suffering from autism (see my answer to Education: What is ‘Child Find’?)
The fact that autism used to be mistaken as schizophrenia and therefore, misdiagnosed in the years past versus today, when autism may be more accurately diagnosed.
The evolution of the autism diagnosis to include a “spectrum” — hence the term, “Autism Spectrum Disorders” to span patients who suffer from varying degrees of the impairments that are observed in autism.
The possibility of misdiagnosis of other psychiatric as well as somatic illnesses, including ADD/ADHD, OCD, Rett’s Syndrome, Pervasive Development Disorders Not Otherwise Specified (PDD-NOS), Angelman syndrome — that share characteristics of autism spectrum disorders but are not “autism” — misdiagnoses may also count toward “autism” prevalence when in fact should not be counted
Autism did not even appear in DSM (Diagnostic and Statistical Manual of Mental Disorders or DSM) until the 1980s in DSM-III. Before this, people suffering from autism were still shunted into the schizophrenia category.
The fact that we no longer treat patients with autism the way we used to treat patients with cancer: imprisoning them in homes and from schools and from society. Would you believe in the old days patients with cancer are kept in “cancer houses” and not allowed to come out because it was thought they were contagious or beset with the Devil? Well, patients with neurological and mental illnesses (and autism is debated on whether it should be considered a mental illness) suffered similar fates. In fact, children with disabilities were excluded from schools altogether, so you’d never see them. Now that we “see” these individuals, we may perceive there is an “increasing prevalence” when in fact, this is an artifact of people no longer being hidden away from view.