Does Education Cause Autism?
If a strong and consistent association between autism and a single chemical or vaccine were found in ten separate clusters around California you'd expect to read about it; and you'd expect the researchers to infer that the chemical or vaccine was, in fact, the cause of autism. But what if the only strong and consistent association wasn't between autism and exposure to some substance but was instead between autism and something complex, like "high parental education"? Well, you'd probably expect a lot of frustration; and you'd be pretty sure autism wouldn't be laid at the feet of higher education.
So what should we make of "Geographic Distribution of Autism in California: a Retrospective Birth Cohort Analysis" ? Why would the children of parents with college degrees be at a 400% increased risk of autism? Why aren't chemicals to blame?
The authors concede in interviews with Scientific American that the results tend to undermine claims that industrial pollution causes autism. Yet rather than explore the possible reasons why college-educated parents might be significantly more likely to have autistic children the authors speculate that their results may be biased. Perhaps, they suggest, the less educated aren't educated enough to take their autistic children in to be diagnosed. Or maybe, they wondered, the educated are much more likely to use some unknown household chemical which does, in fact, cause autism. Scientific American hastens to add that there's an unpublished study that shows a doubling of the risk of autism in mothers who use pet flea shampoos.
But why point to a much smaller effect allegedly to be found in an unpublished paper when a much larger effect published in a peer reviewed journal is in front of you?
First, people tend to prefer simple answers with readily implemented solutions, the blame for which, and the costs of which, get imposed on someone else.
Second, epidemiology, at least as practised in the West, tends to be reductionist. The triumph of linking tobacco to smoking led many epidemiologists to believe, or at least to want to believe, that indeed there are simple solutions to big problems and that silver bullets are out there just waiting to be found.
Finally, it's becoming increasingly clear that all sorts of ailments, from cancer to obesity, are highly complex and result in large part, from lifestyle, cultural and economic choices. Take breast cancer for example. One of the very best ways to reduce the risk of breast cancer is for a woman to bear children and to begin doing so in her late teens or early 20s. Yet no one wants to say that putting off children to get a degree and start a career is a cause of breast cancer.
And so, for now, the search is still on for the usual suspects.
