Be careful gleaning anything with email surveys like one. There could be confounders. E.g.

High IQ -> More $ -> More likely to see Dr for diagnosis

I didn't read the whole thing though--did anyone w/ the time decide to do a deep dive on the study and see how they controlled?