We already know that most of the increase, even in diagnosis of level 3 autism, is due to diagnostic shifting/switching (as explained pretty well here: https://youtu.be/BdpSfrD3Nzs). We also know there are reasonably well supported risk factors that are increasing in prevalence; e.g. father aged >50 at the time of birth.

> general cognitive decline

According to whom? The Flynn effect famously says the opposite: https://en.wikipedia.org/wiki/Flynn_effect