Many people have heard of a connection between autism and the gut microbiome. This hypothesis has been investigated for some years now, because there really does seem to be a correlation between autism symptoms and various gastrointestinal ones. And there is more and more evidence that the gut microbiome does communicate with the rest of the body, sometimes in unexpected ways, so the idea that autism (or at least its symptoms?) might be treated via the GI tract is not something that can be immediately dismissed on principle. And it has an immediate appeal in that GI approaches are generally a lot easier to realize clinically than CNS-directed ones (even if we knew what to do with the latter!)
But like all hypotheses, this one has to meet its data and it has to stand up under new experiments. Here's a new paper that addresses both of those: the authors review the previous studies on GI/microbiome/autism connections, and find that it's not as strong as one might have thought from a distance. The sample sizes are small, the statistical approaches vary in rigor, the readouts on microbial populations have been less-detailed than you might want, and the results of these studies themselves have been inconsistent. Notably, the larger and better-controlled trials in this area have definitely tended to be the ones with negative results - and the experiences over the last year or two with the coronavirus pandemic should be enough to show anyone what a warning sign that can be. The authors are frank: ". . .it appears that interest in the ASD microbiome outstrips what the primary evidence base warrants" A general alarm bell for the gut microbiome field was sounded last year, in a paper that pointed out that the rate of successful extrapolation of rodent studies to human disease in this area is far too high to be real.
The current paper goes on to present results of a new (and larger-than-usual) study of 247 children in Australia. They found that the gut microbiome profiles correlated very poorly indeed with autism diagnoses, severity, and symptoms, and they go on to propose that everyone has been getting things backwards. Instead of unusual gut microbial profiles causing autism, it seems more plausible by now that autism - with its behavioral affects such as repetitive behaviors and strong limited diet preferences - causes unusual gut microbial profiles.
It'll be interesting to see how this is received by the autism patient community, which is rather wide-ranging in what various members of it are prepared to believe (and prepared to dismiss). That's understandable, given the impact and severity of the disease. If this sort of work does end up moving gut-based approaches back down the list into the "unlikely to do any good" category, it'll be (at first glance) a loss. But if the microbiome idea really is wrong, then the real loss would be continuing to put time, money, and resources into it that could go to things with a better chance of working. This new paper finishes up with a call for people to replicate its results, and that's exactly what needs to be done. Ideally, we'd do that before continuing with more autism/microbiome work in general, but that's unlikely. But this hypothesis has a lot of issues to deal with, and the field will do itself - and autism patient, and their families - no favors by not dealing with them as soon as possible.