Over the past three decades, the number of children diagnosed on the autistic spectrum has soared from 1 in 2,500 to 1 in 166. That huge rise in diagnoses is attributable in part to a greater understanding of the disorder, leading to more accurate detection. But there’s been lots of debate in the medical community about whether autism is perhaps triggered by environmental factors such as childhood vaccines or perhaps toxins, which could explain some of the rise. Now a new study from Cornell University suggests another possibility: Autism may be triggered by television. The study did not include direct evidence correlating television watching with autism but rather linked it through other factors. Authors Michael Waldman and Sean Nicholson and research assistant Nodir Adilov found a rise in autism incidence in certain counties in Oregon, California, Washington and Pennsylvania starting in 1980 that mirrored a rise in cable television subscriptions. Waldman, the Charles H. Dyson professor of management and professor of economics at Cornell’s Johnson Graduate School of Management, talks to Media Life about the controversy surrounding the paper, what type of TV might be harmful, and whether other disorders could be triggered by television.

How did you come up with the idea of autism and TV watching being tied together?
 
Well, I saw a bunch of news reports that said rates have been rising, and I like working on puzzles. It sort of dawned on me that there’s been an increase in young children watching TV. I looked at the medical literature and what they had concerning that, and there was nothing there.

It seemed to be something that should be looked at, so I decided I would take a look and see what was there.
 

Generally speaking, there's been a lot of controversy about what causes autism and environmental triggers. What sort of reaction are you hearing from this paper?
 
It’s all over the board. A lot of parents are upset, they think it’s junk science.

It’s not. We may be right or may be wrong, but it’s not junk science.

We also got an invitation from a medical journal, and then we’ve had every reaction in between. My personal perspective is I find it seems to be an obvious thing that should be looked at, but why the medical community doesn’t I’m not quite sure.

 
Is there any particular genre of television that seemed more potentially harmful to kids (cartoons, cable, game shows, etc.)?
 
We didn’t test for that, but I’ve gotten emails from people in the field that made comments along those lines.

Certainly, from anecdotal evidence I’ve been hearing, if our hypothesis turns out to be correct, I wouldn’t be surprised to find that not all TV watching is equal from this standpoint. For example, one email from a pediatrician thought big screen TVs could be problematic.

 
What should parents take from this study?
 
The last graph of the paper discusses that. The American Academy of Pediatricians says there should be no TV watching under age of 2, and from 2-5 it should be limited.

We’re suggesting that until someone nails this down, we should pay attention to that. My sense is that the pediatricians should be letting people know. Most parents are letting their kids watch a reasonable amount of television at this point.

 
Did you run across other medical problems that may be triggered by television (besides obesity)?
 
No, although we got an email from a medical professional, and he suggested we co-author a study with him where we use our techniques and measure to see whether ADHD might be triggered by TV viewing. But we haven’t done that yet.

Did you actually measure time children spent watching TV and how it correlates to autism diagnosis?
 
No, we didn’t do that directly, the nature of our tests are standard economic statistical techniques which allow you to test to see if findings are consistent, but it doesn’t definitely tell you if there’s a connection.
 

Please summarize your paper. What is the tie between autism and TV watching? How much evidence do you have to back it up?
 
Basically, there’s no large data sets on early childhood TV viewing that would allow you to test our hypothesis directly, so we employed a standard statistical technique called natural experiments.

Basically what happens is you identify a variable correlating with the variable of interest, in this case TV, and see what other variables there are.

We show that precipitation [rain, etc.] and early childhood TV watching are correlated, if early childhood TV watching causes autism, then there’s the assumption that precipitation should cause autism. So then we go to three states, Washington, Oregon and California and we run various tests, and it’s pretty clear that there’s a strong correlation between precipitation and autism.

From a standpoint of an experiment, it says there’s some correlation between precipitation and autism. We thought [the connection] could be TV viewing, but could be something else.

In another one, we looked at cable percentages. Again, this time in California and Pennsylvania between 1970 and 1990, and again, with the TV-as-trigger hypothesis, we found a positive correlation between cable and autism.

This is not definitive evidence, and certainly I’ve seen suggestions on what could be driving our results. But there’s an awful lot here that we think indicates it should be tested directly.