How would that work? If both the maternal drug use and the ADHD outcome are the result of a third familial confounding variable, then it's reasonable to expect

How would that work? If both the maternal drug use and the ADHD outcome are the result of a third familial confounding variable, then it's reasonable to expect that this third variable also leads to paternal drug use. This would be corroborated by the fact that siblings of kids whose mothers took Paracetamol during pregnancy also experienced higher ADHD. As always in these things, effect sizes are small, confounding variables are huge, and much much much larger socio economic factors that we as a society could actually do something about are routinely ignored.