It seems to me that perhaps the data better supports that dementia is correlated with unintentional and short naps. This makes it sound like they think it is causal, which I doubt.
Yes, ironically they even admit possible reverse causal effect along with many other causal variables. Frankly, this is not exactly stellar research in my view, at least with the rather reckless insinuations not supported.