Why not go and research that for yourself? I don't wish to be uncivil but I've heard this question so often for the last 20 years that I now regard it as little more than concern trolling. I get it, it's hard to make meaningful quantitative statements about a behavioral disease whose clinical parameters are hard to define.
But it's a naive point and this subject has been a topic of public debate for a long time now, so the skeptical signalling gets really tedious after a while. It's about as enjoyable as if I described your work as a computational scientist as 'pushing buttons all day' or made some crack about it being an excuse to play video games, as a way of signaling a lack of regard for nerdy pursuits.
I'm by no means suggesting that autism can be an extremely serious disorder. The question is if these numbers represent an increase in serious cases.