But I could certainly imagine that a mother with autistic traits could be someone who takes painkillers more often than the average person.
But I could certainly imagine that a mother with autistic traits could be someone who takes painkillers more often than the average person.
Though I'd expect that if aspirin did have an affect that it'd change the prevalence or severity of autism in children having genes related to autism.
There'd be a first order correction fornthe likelihood that aspirin is causitive by controlling for increased ibuprofen and tylonol usage as well. The second order correction would be whether autistic people were more likely to use aspirin over ibuprofen or tylonol, etc.
Nevertheless, comparing the observed correlations of Tylenol with aspirin and ibuprofen would be the first thing to check. Seems unlikely to me that the OP's suggestion could be controlled for that way. I'll be curious if Kennedy's report checks those basic things.