Again this is an example of correlation, not causation.
This has also been addressed many times, there is no evidence to show that vaccinations cause Autism, and a great deal of evidence to show that vaccinations do not cause Autism.
There is also a lot of controversy about the diagnosis of Autism, in my view it is over diagnosed in children and adults, much like other behavioural disorders, such as ADHD.
Do the Amish visit mainstream doctors/psychiatrists or psychologists? I suspect not, and this correlates with a reduction in diagnosis.
In my own town, there are very few diagnosed cases of Autism, compared to the national average, but the vast majority of children are vaccinated. So there is a correlation that vaccinating children lowers the risk of an Autism diagnosis. This however is not evidence of causation, and is a weak correlation at best.
This fact could also suggest that there are fewer children being seen by health professionals that diagnose Autism. It could also mean that because we live in a relatively unpolluted environment, there are fewer diagnosed cases of Autism. Maybe people in my town eat better than the average person, there could be many reasons why the diagnosis of Autism is lower in my town. None of these suggestions or correlations prove anything.
I think most correlations are worth researching, but the focus needs to move away from correlations that have been proven wrong, and move towards finding answers for parents with Autistic children, rather than propaganda for anti-vaccination websites.