Or maybe dementia prevents deep sleep, which is much more logical given many diseases in general disrupt proper sleep.

I'm sick of low-quality studies which try to draw causative conclusions from correlative observations (and usually often quite weak ones).

Stop writing garbage papers.