I feel like you're making a pretty elementary subset error there...

I meant what I said. If 1/3 of X are Y, but X doesn't have anywhere near a 1/3 prevalence in the general population or in other subgroups that are disproportionately Y for separate reasons, then it's fair to say that X has a huge influence on Y.

Nor do I see a need to continue this conversation.

The proper way to end a conversation is to just end it, not to say "this is why I am right, now that I am done saying that, I'll end it".