I was under the impression we knew shit-all about what caused autism?
You can still study the degree of influence that genes have even if you have no idea what the mechanism is.
Twin studies have shown that it is highly heritable, especially if you broaden the definition of an "autism trait" a lot to include traits that are not really problematic. (So, siblings might become a computer geek that is a bit socially awkward)
But it can still be true that drinking wheat juice, hugging lambs and eating only raw foods would eliminate every case of autism. All we can know from studies like is that it is true that it is highly heritable under the conditions that are typical now.
It might not seem as likely though...