Here is the latest, excerpt:
Examining three years of birth records and pesticide data, scientists from the Public Health Department determined that the Central Valley women lived within 500 meters, or 547 yards, of fields sprayed with organochlorine pesticides during their first trimester of pregnancy. Eight of them, or 28%, had children with autism. Their rate of autism was six times greater than for mothers who did not live near the fields, the study said.
There is some attempt to look at larger numbers:
The scientists collected records of nearly 300,000 children born in the 19 counties of the Sacramento and San Joaquin river valleys. Of those children, 465 had autism. The scientists then compared the addresses during pregnancy to state records that detailed the location of fields sprayed with several hundred pesticides. For most pesticides, no unusual numbers of autism cases were found, but the exception was a class of compounds called organochlorines.
I am mostly skeptical. There are plenty of autistic children, here and abroad, who were never exposed to these chemicals. Agricultural valleys in California don’t seem to have especially high rates of autism. Should I believe "you either get autism from your parents, or in some cases, from pesticides"? Or "pesticides are a potent epigenetic trigger"? Or should I believe "they found 465 cases out of 300,000 when they should have found many many more. Reporting of autism is biased in that sample, and the reporting bias is somehow correlated with certain kinds of agricultural activity."?
So far I’m sticking with the latter. Four points of note: a) the study author is appropriately cautious, b) he does try to adjust for regional diagnosis center, c) I very much wish this study had a map of incidence, and d) is it really so difficult for the author to discuss what it means that the study is restricted to just one part of California?