While I don't doubt that there could be a link, it seems ridiculous that they don't (at least as far as we can tell from this article) do anything to tell if th

While I don't doubt that there could be a link, it seems ridiculous that they don't (at least as far as we can tell from this article) do anything to tell if this is causation one-way or the other. Not hard to imagine that people who are starting to suffer from chronic anxiety may try self-medication before they try the emergency room. Again, not saying there's no causation the way they imply, but basic experimental principles should tell you to try to control for this somehow, for example by looking at overall anxiety rates before and after it becomes commonly available in a given community, or comparing similar sized communities in different areas with different availability.