The moment I started reading, I came to the same conclusion as the paper - causality is unknown. It's very likely that kids with autism/adhd are drawn to the high-stimulus sources of entertainment more than their peers.