It began, as so many things do, with me being a dick.

Mike is related to a person who has become a champion of the vaccine-autism link, and when I discovered this I sent an email along the lines of "I mock this and hope you have fun chatting about this at family gatherings." What followed was a very interesting back-and-forth. I consider the vaccine-autism theory to be roughly on par with the 9-11 Was an Inside Job theory in the intellectual hierarchy.

Before I go into any details, let's be emphatic about two things up front to avoid wild accusations at the end: Mike was not arguing in favor of the vaccine-autism link and I was not arguing that autism is made up or nonexistent. Are we clear on that? Great.

What Mike argued is that a steady rise in cases of autism is cause for concern. While the vaccine link appears to have no empirical support (but plenty of Hollywood celebrity support!) there is a non-trivial increase in children with autism in the last decade and it requires an explanation. The existence of substantial statistical noise – which was my counter-argument and which I will describe momentarily – does not negate the potential existence of an underlying trend.

My response was lengthy but centered around what I feel is a key semantic point: it's inaccurate, until solid evidence can be provided, to say that autism is on the rise. The diagnosis of autism is what is on the rise. I believe that the rise in diagnoses is as likely to be attributable to the following two factors as to a legitimate increase in the occurence of autism.

First, autism is relatively new in the context of medical issues. It hasn't been on the radar screen of the general public or the non-specialist medical community for more than a decade or two. I doubt that many people had even heard the term prior to the mid-1990s. So I believe that one valid hypothesis is that doctors and parents, spurred by successful public awareness campaigns, now diagnose cases that would not have been diagnosed in 1970. To prove that autism is on the rise, someone needs to convince us that the kids diagnosed autistic today are not the same kids who were called "slow" or "learning disabled" or "retarded" prior to 1980.

Second, the downside of increased public awareness of the disease is the inevitable hypochondria and hysteria that set in with panicky parents. After 10001 Oprah segments about autism, some parents become convinced that their child has this new, fashionable disease and, seeking to fulfill their own martyr complex, shop around for a doctor who will agree. You may think this is a poor argument, but anecdotally I am convinced that it is some part of the increase. It exists. To what extent, I cannot say. But there are parents out there who operate like this. The hysteria can also affect well-intentioned school psychologists or medical professionals who practically fall all over themselves in a rush to diagnose autistic every child who stacks up his toys or fails to make eye contact for a few minutes. As prior experience with social panics about psychological illnesses (ADHD, depression, etc) has shown us, over-reaction leading to over-diagnosing is a legitimate concern.

Of course the "marketing" of a new medical problem often involves our friends in the pharmaceutical industry; drug companies are pushing autism diagnoses just like they pushed depression and ADHD. They've been pushing the idea of an "autism spectrum", i.e., not really autism but close enough that we can start prescribing drugs for it. Like doctors were encouraged to throw fistfulls of pills at people with even the mildest depression symptoms, they are now being encouraged to stick the autism label on any child whose behavior even hints at behavior outside of a narrowly-defined idea of normality. There is a widespread public perception that drug companies wouldn't get involved because autism treatments are non-pharmaceutical. That is false. More than 50% of children diagnosed autistic are put on antipsychotics, an incredibly powerful and expensive class of drugs, despite the fact that no medical evidence proves that drug treatments work.

If something really is causing more children to develop autism, I certainly hope that we discover what it is quickly. I have no doubts at all about the seriousness of the problem. Autism, depression, ADHD, and other mental illnesses are real and they are serious. However, the subjective nature of psychological disorders means that over-diagnosis is very easy. So before we get all twisted up about a rise in autism I think we should make sure that we are dealing with a rise in autism rather than a rise in diagnosing it.

What do you think? I'm afraid that I didn't do justice to the other person's argument here, but let's be clear about the fact that I consider it an entirely reasonable one. Given the amount of environmental contaminants and chemicals that end up in our bodies these days it is in no way inconceivable that something is causing autism and causing more of it than ever before.