Science 17 Oct 2006 10:25 pm
Does TV Cause Autism?
One of the things that I have always wondered about is what the effect of television would be on developing brains. I kinda thought that the rapid increase in the diagnosis of things like ADHD, for instance, might be related in some way. One of the hallmarks of TV is that no visual cut lasts for more than 30 seconds or so, which would seem, to me, to have a potential effect on the developing brain.
The cause of Autism has been up in the air for a while. It is known that there are genetic factors at work, and some have suggested that vaccines might be a culprit. Now, three researchers are saying that early TV exposure in children might be a cause.
There are a couple of caviats here, right off the bat. None of the researchers seem to have medical, science, or public health backgrounds - they are economists. This raises a red flag for me. But, perhaps they are looking at it from a sufficiently fresh perspective. But it also has not gone through peer review, from what I can tell.
So, I’ve downloaded the article, and read it. Here is my analysis:
They state that the current view in the field is that Autism is caused by a genetic predisposition, and an environmental trigger. They are suggesting that the environmental trigger is television.
The way that they are getting to this is pretty indirect. They are looking at county-by-county rates of cable TV subscription, and, well, precipitation. Yes, precipitation. Apparently, according to some specific studies (American Time Use Survey) young children watch TV more when it rains.
They say that 40% of the autism cases are due to increased TV watching due to precipitation, and 17% of the growth in Autism cases in the states they studied are due to increases in cable TV subscriptions.
So problem #1: the data they use to correlate TV watching with precipitation is data that is self-reporting about how people spend their time, and who else in the household is present at the time. If they were people with public health backgrounds, they would know that self-reporting is problematic. They take the self-reports basically as gospel, and don’t really ask about how the self-reports might be biased.
Problem #2: They suggest that increases in use of cable TV in homes during the 1970s and 1980s would increase the amount of time that children watched TV. But there is no data to back this up. Since they are talking about children under 3, this seems a problematic assumption. Yes, rates of autism are correlated with increases in cable TV, but its pretty weak. So a weak correlation with something that isn’t backed up is, well, useless.
Bottom line: It’s a pretty flawed study, but I guess an argument can be made that it makes it worth looking into, at least. There needs to be a serious case-control study, etc. The public health mantra is: "correlation is not causation" (oh, right, they aren’t in public health) and this paper certainly only hints at a correlation. It’s not even a slam dunk for that.
So, I think I see why it hasn’t been peer reviewed.