The LA Times reports on a study based on Israeli draft records:
Men over 40 are nearly six times as likely to have an autistic child as those under 30, according to a new study that provides support for the role of genetics in the development of the disabling mental disorder.
If this is true, why didn't we know this before now? This would be a huge effect, gigantic compared to what is normally seen in epidemiological studies of unexplained diseases, and it's not all that difficult to assemble this data.
I've been complaining for years that the public health establishment hasn't taken autism seriously. It's roughly about as prevalent as AIDS, and nobody has had any idea how to avoid autism (whereas avoiding AIDS is very, very simple), but the government and the press don't care 5% as much.
P.S. Gregg Easterbrook offers a novel theory for the presumed increase in autism: television. I don't know how plausible that sounds, but he's absolutely right that we need more research into causes. Perhaps his theory could be tested using data from white South Africans, since they didn't get TV until 1976, but had modern healthcare (e.g., the heart transplant was pioneered in South Africa in 1967).