Yes. BUT !"If you are testing to see if thimerosal causes autism in those with an allergy to mercury," R5 #1,140
It's neither illegal nor unscientific to study risk to an entire population.
You are correct that this is determined by study design, the test population selection criteria.
Serendipitous test results are not unknown to science." you don't allow 98% of the test patients to be without this allergy." R5 #1,140
I know of nothing intrinsically wrong with your study design.
But it's not the only possible study design, and may not be the optimal study design for an autism link.