We love Science News for a reason. Best part of the analysis:
"A total of 1,156 pairs of twins fit the criteria for the study, but the researchers ended up assessing only 202 pairs. Although the study is large by twin-study standards, the low participation rate could be a problem. “That is too low of a response rate to marketing survey research, much less an epidemiological study,” says Edwin Cook, a child psychiatrist at the University of Illinois at Chicago."
Specifically: _too low to base marketing on_. *snicker*
I wish I saw even a gesture towards the cultural constraints on receiving a diagnosis. If you can actually get a high level of participation and use a diagnostic instrument, then you can eliminate (or drastically reduce) the cultural constraints on diagnosis. But in addition to a low level of participation, I don't think they even used a diagnostic instrument; I think they just asked what diagnoses the kids (some of whom are adults now) already had (or had not) received. It's even worse when you realize the twins were born between 1987 and 2004, a time frame in which we _know_ that the diagnosis rate has changed.
ETA: While I'm here:
Also a not good study (and potentially a career killer for the lead author if this thing gets the wrong kind of publicity). They compared SSRIs to tricyclics and didn't find the effect with tricyclics (never mind that the risk is really small even if it does exist, and much more easily explainable by comorbidity between ASDs and depression than anything else -- that is, mum has undxed ASD herself, is depressed because being aspie and pregnant is NOT a good combination, takes SSRIs for the depression, and has a kid with autism. Which she was going to have anyway). But _why_ didn't they find an effect with tricyclics?
"Dr. Wiznitzer, who is also the neurology liaison to the Autism Subcommittee for the American Academy of Pediatrics .. "We're only talking about 20 children who had exposure to any antidepressant, and that's an awfully small number on which to make a firm conclusion. Instead, I'd say that if this data is true, this would only be a signal that would suggest you need to study this further to determine why.""
I've got no opinion on whether one should or should not take medication for depression as a general rule; I think this is something that needs to be worked out on a case by case basis. But I sure hope that omg my kid might have autism does not come in on the don't-take-the-drug side of the decision making process. I'm pretty sure that if you have a kid on the spectrum, years of untreated depression leading up to that will only make things worse.