Thanks for your comment, Dan. Sleep studies are typically done under lab conditions (in sleep labs). If fragmentation of sleep is part of a study it is artificially induced, i.e. the subjects are awakened when they would otherwise continue to sleep. That situation does not replicate the "natural" patterns of sleep fragmentation that people encounter "in the wild" so to speak. On the other hand, observational studies that correlate these natural sleeping patterns with health outcomes are of an associative nature only. They do not allow causal inference.

I believe that sleep is a highly individual phenomenon, which requires individualized monitoring as I do using the N-of-1 method that I discussed in several previous posts.