I've read that in cave studies when isolated from any direct indication of time people will naturally sleep longer every other day like this. However, I have severe sleep issues that leave me disabled and the main issue seems to be that I have a non-24 hour circadian rhythm and can't sleep outside the rotating times but in some ways my body is still synchronized to the sun (I do much worse when awake at night). So I worry you could also potentially end up with severe circadian issues unless you are completely blind, although if you always wake up at the same time whenever you do sleep that would hopefully help prevent issues. Sleeping in late some days makes circadian issues more likely.

Taking more than 20 minutes to get to sleep is not that unusual as I understand it. I think 30 minutes minutes is quite common even in people who sleep fine. 4-5 hours is very short though. One of my parents also slept little and there is also dementia on that side of the family (I have some memory issues as well already). I don't know how much previous gnerations slept but I suspect there may be not so benign genetic short sleep issues as well. At least some issues may have susceptibility and a triggering event, with tech making the triggering event more likely. That would be another question about benign familial short sleep, if it increases the susceptibility to more severe issues.

> stay up until exhaustion.

I was like that, and among other remedies (like actually fixing the problems in my life that bother me), my latest last-resort is brisk walking to exhaustion, even very late in the evening. If it's 10:30 and I know I'm in a bad mood to sleep, I just go for a walk in the park for 2 hours and I fall asleep very fast when I return.

It's not ideal and I plan to switch to running, but I'm at a bad place where my cardiovascular fitness is not good enough and if I run for 30 minutes to exhaustion my heart rate stays elevated for more than one hour and running is just too activating.

Walking for 2 hours seems very wasteful right now, but it's what it takes to calm me down enough to fall asleep.

Sleep starting late + driving to exhaustion + occasional night skipping then catchup sounds almost like DSPD:

https://en.m.wikipedia.org/wiki/Delayed_sleep_phase_disorder

That one sleeps 4-5h per night also sounds like one light/deep/REM sleep cycle, of which there would typically be two in an idealised 8h-ish "night".

There are hypotheses out there that a single 8h block of sleep is a myth - possibly socially induced by industrialisation and 3x8 shifts - and that waking up in the middle of these 8h is a normal thing.

"waking up" here may mean anything from near-consciousness sleep to actually waking up and possibly do shit for an hour then going back to bed; or even for some, splitting sleep in two 4-hour blocks spread around the 24h day.

In parent's post it may be that one 4h block is done, then social pressure (work/child schedules) pushes one to continue with this one-block schedule.

It may even look appealing as triggering sleep through exhaustion superficially appears to help with sleep in the short term. Long term it gets exhausting but this perceptually becomes the new normal, especially when the symptoms of this kind of sleep deprivation aren't that obviously tied to sleep habits, the naively expected ones being masked by external pressure (schedule), habit bias (normalisation of deviance through repetition), exogenic (alcohol, caffeine, nicotine) or endogenic (adrenalin, cortisol, dopamine).

The main problem is then that by the time symptoms are impacting the situation is deeply anchored; worse, because the root cause is non-obvious it is often misdiagnosed.

Don't ask me how I know.

Also, not a physician, just saying: take care.