Thursday, November 06, 2008

Rainman: Does rain cause autism?

Of course not, but the results of a study recently published in the Archives of Pediatrics and Adolescent Medicine seem to indicate an association between autism and rain.

Specifically, the rate of childhood autism is higher in counties in California, Oregon, and Washington with greater precipitation.

According to the study's co-author, Sean Nicholson, who is not a psychologist but rather a public policy analyst, Washington counties west of the Cascade Mountains, get four times as much precipitation and had autism rates twice as high as those in the drier east.

The research made headlines in papers across the country, including our small community's newspaper. It makes sense too, if rain doubles the chance of autism, you'd think that's important information people should know about.

While the results of this study are interesting, it needs to be put into proper context. First of all, this is another case of correlation research. Two things being associated does not mean one caused the other.

Another cause for concern with this type of research finding is how it is that we're looking at incidence rates, not raw numbers. And this kind of number (rates) is very difficult for to keep in perspective.

If the rate of autism is doubled in rainy counties, what do suppose is the risk of any one individual developing autism in that county compared with one in a drier county? Twice as high, you'd say. But — you'd be wrong. Why? In absolute terms, about 1 in 150 people has autism. If the rate doubles, that means 2 in 150 people has autism. So, the rate of autism goes from 0.6% to 1.3% which is a lot, lot smaller than the 100% increase that gets reported.

Dang it. So much for statistical literacy.

Lately I've been paying a lot more attention to this issue since I received a special supplement to a peer-reviewed academic journal I subscribe to. The supplement, Psychological Science in the Public Interest, features a long article written by a team of very critical scientists who argue that the way health statistics are presented, oftentimes with the complicity of medical doctors, does an enormous disservice to the public and in some cases does actual harm by leading people (and doctors themselves) to make bad medical care decisions.

I'm only 3 pages into the tome, but I'm finding it very interesting and it has certainly helped me to put studies like this Rain + Autism one into proper context.

I'm not sure what public policy decision the autism study authors would make, but I wouldn't be surprised if they were downright idiotic.

Here's some evidence to support my hunch. They had presented early results from this data before as part of the Johnson Graduate School of Management's "Research Paper Series" which I'd bet is not a peer-reviewed journal. This is what they had to say about it (edited for brevity, and bolded to point out where the authors false make causal claims where the data is merely correlational):

"Autism ... one of the current theories concerning the condition is that early childhood television viewing serves as ... a trigger. Using the Bureau of Labor Statistics' American Time Use Survey, we... show... consistent with the television as trigger hypothesis, that county autism rates are also positively related to the percentage of households that subscribe to cable television. Our precipitation tests indicate that just under forty percent of autism diagnoses... is the result of television watching due to precipitation, while our cable tests indicate that approximately seventeen percent of the growth in autism... is due to the growth of cable television. These findings are consistent with early childhood television viewing being an important trigger for autism."

Now that an apparently respectable peer-reviewed journal has published this autism + rain link piffle, I'm left scratching my head and hoping that people don't monkey around and take this crap seriously.

Rain and TV viewing do not cause autism!

6 comments:

Jennifer Lommers said...

Yeah, I couldn't BELIEVE that correlation! The statistics are interesting, but they have so much more research to do!

Also - TAG - you're it!

For info, see my blog here!
JenniferLommers

Pam said...

Very interesting. I have Asperger's. I read that and I think about it in this way:
I personally believe it is genetic. Sensory issues tend to go hand in hand. I and many others I know do better in cooler climates and less sun. The sun hurts my eyes. We also tend to understand tv better than other things. I find comfort in watching it.

So. To me, it makes sense! But it has nothing to do with the actual rain OR tv.

Thanks for posting this. If it's ok with you, I'd like to link your post on my Asperger blog (http://www.thestampedenvelope.blogspot.com).

Field Notes said...

Yes, go ahead and link it, EmptyEnvelope.

Pam said...

Thanks!

yellowfish said...

I also think that some of this stuff goes way over the top (for instance the Autism-vaccine stuff drives me nuts). But, there is another light to see it in. For instance, Schizophrenia has long been shown to associate with season of birth- this seemed very weird for a long time. But, now its thought to be because having the flu or other illness during the second trimester is actually a known risk factor for schizophrenia- due to the effect of immune responses on the fetus (this isn't some kind of splinter group crackpot research either, its very much accepted among schizophrenia researchers as one of the potential environmental components). So, obviously we already know that there is a lot of research showing autism (like schizophrenia) has a big genetic component, but that does still leave room for things yet to be completely figured out.

Not that I'm totally convinced by this particular article, just to play the devils advocate and show another side.

Virginia said...

I love that you have taken the time to decode the statistical realities here. It has been a loooong time since I finished my Anthro degree and statistical analysis was never my strength. It is so easy to be fooled by numbers.

I suspect that someday we will discover that Autism and Asperger's syndrome have multiple triggers that build upon one another. Who know how all of those factors can be measured, tested and studied? Especially if the most important component is the individual's genetic code and unmeasurable personality tendencies. I'm sure it is a fascinating field to be working in - researchers will never run out of things to test.