You need a significant period of time without any time-bound obligation (or even any obligation at all). Any need for an alarm or anything that you'd possibly materialise in a calendar needs to be removed.
You need a room which is pitch black at night (e.g blocks street lights, no leds), yet still allows a tiny bit of sunlight (rationale: without cues the circadian cycle extends up to 48h, there was an experiment about that)
Then, once you have that, start freewheeling, just when you feel tired go rest and when you wake up get up. It'll be a mess at first, for two reasons.
First you may be unable to properly recognise the "I need to take some rest" signal: we're trained by life to largely ignore it.
Second, you need to pay off any debt, sleep, physical, mind, that would play a role in altering your base cycle. Have fun, work out, meditate, go see a therapist even. In a nutshell, find your balance.
Once you have recovered from everything, once all biasing sources have been removed, then sleep will converge to some rhythm, which carries some error margin, so you can only observe it statistically over time. There's your baseline.
It could take months, which more often than not isn't practical to have, so from the above ideal scenario one could devise a protocol that would try to stick as close as possible to it.
Source: first principles+anecdata, sample size of 1, double-non-blind protocol; a.k.a myself digging out of a hole.
Protocol: 3 week sick leave for burnout x anxiety depression, coincidental breakup (so no SO), went from office to remote working, workout plan, 6 month therapy, statistically reliable sleep/health/performance tracking via watch allowing for outlier identification/retrospective deviation root cause analysis.
Result: 0130-0930~30min sleep, 7h45min~5min sleep duration made of two ~4h-ish blocks, 25h~30min cycle. When bound to 24h rhythm, occasional sleepless night leading to 48h day and three aforementioned blocks of sleep (~11h-ish total).
From the data it became painfully obvious that I experience delayed sleep phase disorder, took me 6 months to figure the baseline out, and 6 more to confirm, which is quite hard when you don't even know that it is a thing.
> rationale: without cues the circadian cycle extends up to 48h, there was an experiment about that
Do you have a link to this experiment? I haven't seen this mentioned anywhere. Studies referenced in the non-24 wikipedia page suggest natural circadian rhythms typically range from 24-25.5 hours for people not getting external cues.
Here's an interview on the subject, which includes the initial experiment (which had him having a slightly over 24h rhythm) and subsequent ones (where the 48h rhythm showed up)
The interesting thing about going down a cave is that it removes a ton of other possible cues like temperature variation (or side effects of trying to mechanically account for that)
My pet theory about human beings generally having slightly longer than 24 hour circadian rhythm is that it gives margin: with a 24 sun cycle and a 24.5 rhythm you naturally get a self-correcting consistent positive benefit; otherwise you'd always be right on the edge and unable to recover when things go sideways.
You need a significant period of time without any time-bound obligation (or even any obligation at all). Any need for an alarm or anything that you'd possibly materialise in a calendar needs to be removed.
You need a room which is pitch black at night (e.g blocks street lights, no leds), yet still allows a tiny bit of sunlight (rationale: without cues the circadian cycle extends up to 48h, there was an experiment about that)
Then, once you have that, start freewheeling, just when you feel tired go rest and when you wake up get up. It'll be a mess at first, for two reasons.
First you may be unable to properly recognise the "I need to take some rest" signal: we're trained by life to largely ignore it.
Second, you need to pay off any debt, sleep, physical, mind, that would play a role in altering your base cycle. Have fun, work out, meditate, go see a therapist even. In a nutshell, find your balance.
Once you have recovered from everything, once all biasing sources have been removed, then sleep will converge to some rhythm, which carries some error margin, so you can only observe it statistically over time. There's your baseline.
It could take months, which more often than not isn't practical to have, so from the above ideal scenario one could devise a protocol that would try to stick as close as possible to it.
Source: first principles+anecdata, sample size of 1, double-non-blind protocol; a.k.a myself digging out of a hole.
Protocol: 3 week sick leave for burnout x anxiety depression, coincidental breakup (so no SO), went from office to remote working, workout plan, 6 month therapy, statistically reliable sleep/health/performance tracking via watch allowing for outlier identification/retrospective deviation root cause analysis.
Result: 0130-0930~30min sleep, 7h45min~5min sleep duration made of two ~4h-ish blocks, 25h~30min cycle. When bound to 24h rhythm, occasional sleepless night leading to 48h day and three aforementioned blocks of sleep (~11h-ish total).
From the data it became painfully obvious that I experience delayed sleep phase disorder, took me 6 months to figure the baseline out, and 6 more to confirm, which is quite hard when you don't even know that it is a thing.