This has been "known" for years with autism, based on observational data. Normally, I don't put much into observational data, since it can be so wrong, but in the autism/paternal age case, you have a HUGE, HUGE leap in likelihood of autism based on father's age. A man in his 40's is thousands of times more likely to have an autistic child than a man in his 20's, as a I recall from previous observational data.