When I read things like this my first thought is "Post hoc ergo propter hoc"?
While I do think cannabis may induce anxiety it could very well be that those prone to anxiety are the ones more likely to consume cannabis in the first place. there is no way of knowing from this set of data which direction the association goes.
this is why double-blind studies are always critical and any conclusions drawn from data that is not double-blind in nature should be considered highly suspect.