People see improvement even after taking placebo.

I haven't read the paper yet but on first glance it looks like an observational study. I know it's cliche but correlation does not imply causation and observations are just that, drawing conclusions after facts.

It's still an interesting study though...