For all of my research career so far (limited as it is), investigations into the potential effectiveness of a gluten-free and casein-free (GFCF) diet on some of the core and peripheral symptoms associated with autism has formed the vast majority of the work. It is true that the studies I have been involved with in this area have, on the whole, suggested that such a diet (in part or whole) might be a useful intervention for some people with autism (you can read about them here and here). It is also true that for some people with autism, such a diet has no effect whatsoever as described by other studies reporting no significant effect (here and here). Cumulatively these studies are also reflective of the 1000s of reports in books and posted on the web about the success / non-success of dietary intervention.
Taking into account the evidence base and all the methodological issues that have been discussed many times on such diet studies, I suppose the best position to take on whether a GFCF diet could help or not is to say, yes, potentially in some cases, but not universally and at the moment no-one really knows how or why (see the Research Autism opinion here).
I was therefore quite surprised to read a recent posting from the very prestigious Mayo Clinic on the topic of GFCF dietary intervention for autism; the headline reading: there is no evidence that special diets are an effective treatment for autism. More surprising was that the article included as references our recent ScanBrit study and also the most recent Cochrane Review of dietary intervention for autism from 2008 which concluded that more research is required.
Well, as per the information about this blog, science and real-life do not generally speaking deal in absolutes, but rather probability. For example, one could argue that if I jumped out of an aeroplane without a parachute or without my parachute opening, I would absolutely, without a doubt, die. The reality however is that a small number of people do survive such a "certainty" (see link to the Bonus Day Club) as a consequence of many different factors and effects; hence our use of probability over the absolute. It is indeed a shame that such a dead-cert line was adopted in the recent Mayo Clinic article; certainly from a group who should know a little about science, evidence-based practice and probability. I note that there has been some discussion about this on other blogs and websites; some people in agreement with the Mayo Clinic line, others slightly more hesitant to accept such an absolute account.
I think the bottom line is to remember the parachute.