Or, people with headaches produce offspring with headaches, but kids don't know they have headaches and just get angry and act out
They thought about that. Doesn't look like it.
> They failed to discern a pattern of increased medication-taking by women who were carriers of genetic variations linked to behavioral problems.
The mothers carrying known behavioral genetic markers were not more likely to take acetaminophen. If the headaches were correlated to a known marker then you would expect it to also be passed on to an offspring in a statistically significant way.
But if there's an unknown marker that correlates with pain complaints? I know of at least one psychologist who is going back to some of Jung's commentary on behavioral issues and trying to carve out a set of people who are diagnosed as 'shy' or 'ADD' but who just have heightened senses.
The gist of it is that some people have a less aggressive filter on their senses, and eventually they can't cope with the information overload and they begin to act out, not unlike an exhausted child. Depending on environment and luck, they might develop shyness, ADD like symptoms, become scarily perceptive (extreme case: Sherlock Holmes, Psych), or some combination of all three.
These people also complain about physical discomforts and pain with a higher frequency than normal people. The question of how exactly they eliminated pre-natal painkiller use of the mother as a predictor for behavioral problems wasn't addressed in either summary but might be buried in the data.
What if there's a marker that not only correlates the aspirin use and the child's behavioral problems, but subjects with the marker more strongly correlate the relationship than subjects who don't?
That kind of statistical analysis is how we find out a random stretch of DNA is a marker in the first place. All forms of human study will advance dramatically once this becomes standard practice, as indeed it must.
Of course, you're completely throwing out the fig leaf of 'anonymous' health data. But it turns out that for the vast majority of health studies, deanonymizing the data is quite practicable in most cases already. And the gain to medicine would be enormous.
From the guidelines:
Please don't insinuate that someone hasn't read an article. "Did you even read the article? It mentions that" can be shortened to "The article mentions that."