Entrez PubMed: autism and etiology
The recent press about mass profile of unusual blood chemistries statistically correlated with autism led me to review the medical literature on autism etiology. I did my best to exclude MMR immunizations; that seems relatively unlikely as a contributing factor. I was interested in mention of infectious agents.
I didn't come up with very much. The literature is a confusing stew of hypotheses, with none seeming to lead. Part of the problem, of course, is that the wide range of entities we call "autism" may represent several different conditions. Even the most "classic" autism may represent many processes with similar outcomes. It's pretty hard to sort out such an ill-defined target.
If there's any fashionable correlation with infection it works on the assumption that some autism subsets are caused by an autoimmune reaction to a common virus or insult that happens to damage brain tissue. There's also a tendency to look for common causes of both autism and celiac disease; these conditions have some superficial similarities.
If autism has an autoimmune and infectious component, and if the frequency is rising in wealthy nations, they one would wonder if it was a "hygiene disease". These are diseases thought to arise from the absence of early infections; chickenpox, for example, is a much more dangerous disease in middle-age than in childhood.
All very speculative. Progress will probably come from studying well defined familial autism syndromes that may be more homogenous.