An interventional study is ideal, but unlikely to happen.

I would have been happy if the authors just tried to factor in other characteristics like socioeconomic status, education level, etc. It looks like they didn't do that but instead just observed an "association" which isn't terribly interesting or informative.