Here's the thing about Daniel. The man has been a night owl his entire life. He's told us this. And then fifteen months ago Ezra arrived, and the whole architecture of his day fell apart.
And he wrote in with a interesting two-part question about it.
He did. So Daniel's first question is this. Since Ezra was born, he can't bounce between states anymore. He works in flow states, that's how he operates, and the natural interruptions of parenting are hard enough on their own. But now his home office is also a shared nursery, subdivided by a curtain. So he's trying to get everything done, including setting up and organizing the new space, on a schedule that doesn't disrupt Ezra.
And the specific thing he can't do anymore.
The specific thing is this. Hannah has a bedtime routine for Ezra that involves making the room pitch black. If Daniel participates at, say, eight in the evening, there is essentially no chance he emerges from that dark room and recovers the mental state and momentum to resume serious work. He calls it a problem of circadian flexibility, though he's not sure that's the right term. He knows about context switching and executive function from our ADHD conversations. What he doesn't know is whether this is just more of the same, or whether there's a discrete time-of-day component. Once the brain has started transitioning toward sleep, does reversing direction and returning to cognitively demanding work become harder?
That's the mechanistic question.
Then the second one is partly him asking us to tell him he's not being ridiculous. Because he's been typing things like, and I'm quoting him here, "Is it okay for a grown man in his thirties to go to bed at ten PM?" Or even eight PM. He says it carries this strange psychological weight. Somewhere in his head, going to bed at eight or nine is something old people do. Even when he's tired and it would make practical sense, it almost weirds him out. Hannah has already made this move, by the way. She goes to bed very early specifically to minimize the sleep she loses to middle-of-the-night wake-ups. So his question is what we actually know about when people go to bed at different ages and stages of life, particularly parents of young children. Is eight, nine, ten PM unusual for a man in his thirties. And is there any reason not to embrace an earlier schedule if it fits the practical realities of parenting better.
Two questions. One mechanistic, one epidemiological. And the research has something substantive to say about both.
Let's start with the mechanism, because that's where his intuition is closest to right and also most incomplete.
Here's the framing that makes the whole thing click. Sleep isn't governed by a single system. It's governed by two processes interacting. There's a homeostatic process, which is sleep pressure building up across the day, adenosine accumulating in the brain the longer you're awake. And there's a circadian process, an internal clock that modulates how receptive you are to sleep at any given hour. By evening, both of those are pointing the same direction. The sleep pressure is high and rising, and the circadian system is entering its sleep-permissive phase.
So they compound.
They compound. And that's why the flip toward sleep is self-reinforcing once it starts. But here's the part Daniel is intuiting without having the vocabulary for. There's a window earlier in the evening where the circadian system is doing the opposite. It's actively promoting wakefulness.
The second wind.
The clinical term is the Wake Maintenance Zone. It's a window of roughly two to three hours that occurs just before dim-light melatonin onset, and during that window sleep propensity is suppressed. What's counterintuitive is that cognitive performance on some tasks actually improves during it, despite the fact that you've been awake longer and your sleep pressure is higher.
So you're further into sleep debt and performing better.
Right. That's the alerting effect. And this is why the timing of a wind-down matters so much. If Daniel is starting a bedtime routine at eight in the evening and his natural melatonin onset is late, which is exactly what you'd expect from a lifelong night owl, then eight PM is landing inside his wake maintenance zone. He's not fighting a habit. He's fighting a wake signal his own brain is generating.
That reframes the problem entirely.
It does. It's not that he can't wind down. It's that he's attempting to wind down during the one part of the evening when his circadian system is most opposed to it.
Which would explain why it feels so different from ordinary task-switching resistance.
Now the second half of the mechanism. Say he does it anyway. Say he goes into the dark room, does the routine, and starts descending toward sleep. What happens when he comes back out?
Sleep inertia.
It's the transitional state of reduced alertness and impaired cognitive performance that you get on waking. And two things exaggerate it. Prior sleep loss, and the time of day at which you wake. Daniel is a parent of a fifteen-month-old. He has prior sleep loss. And he's emerging at a time of day when his circadian system is pushing him toward sleep rather than away from it.
So he's getting the worst version of it.
He's getting a compounded version. And the dark room matters here too. A pitch-black room plus a calming bedtime routine is essentially inducing the onset of that state deliberately. He's not waking from a nap. He's waking from something his body has started to read as the beginning of the night.
How long does it take to shake off?
That's the part that should change how he schedules his evening. There's a review of sleep inertia countermeasures that found nothing has successfully dissipated it within the first fifteen minutes after waking. Nothing. Not cold water, not caffeine, not bright light, not any of the standard interventions people reach for.
Fifteen minutes.
Fifteen minutes minimum, and that's the floor. There's a study on a thirty-second high-intensity exercise burst, and it did reduce subjective sleepiness. People felt more awake. But it did not improve cognitive performance or move core body temperature.
So they felt better and performed the same.
Which is the trap. The subjective feeling of recovery isn't the same as the cognitive capacity being back. And for someone trying to resume serious work, subjective wakefulness is worse than useless. It convinces you you're ready when you're not.
There's a practical implication buried in all of this.
There is. The executive-function framing suggests the solution is to push through. The circadian framing suggests the solution is to not start the wind-down unless you're done working. That's a scheduling answer, not a willpower answer.
So Daniel's intuition that he should modify his schedule rather than fight the transitions isn't him giving up. It's him reading his own biology correctly.
His instinct is well-founded. And I want to be precise about the verdict on his first question, because he framed it as a binary and it isn't one.
Give me the verdict.
It's both, but the circadian component is real and discrete. The context-switching framing explains the cost of making the switch. The circadian framing explains why the switch becomes progressively harder and eventually near-irreversible as the evening advances past melatonin onset. Those are two different mechanisms stacked on top of each other.
The first one explains the price of the ticket. The second one explains why the door eventually locks.
And they're separable. If Daniel were doing this at two in the afternoon, he'd pay the context-switching cost and recover. At eight PM inside his wake maintenance zone, with sleep inertia on the other side, he's paying both costs and the second one is the one that's actually stopping him.
So the problem isn't willpower. It's timing.
And that's not a comfortable answer for someone who's used to working through resistance, because it means the resistance isn't the thing to work through. It means the sequence is wrong.
Let's go to the second question, because that's where the data get funny.
The data are funny. Daniel is typing "is it okay for a grown man in his thirties to go to bed at ten PM" into ChatGPT, and the answer from the American Academy of Sleep Medicine's twenty twenty-four survey is that he's in a very crowded room.
Numbers.
Two thousand and six US adults, and among those who have a set bedtime, here's the distribution. Eight percent go to bed at eight PM. Twenty-one percent at nine. Thirty-one percent at ten. Twenty-five percent at eleven. And twelve percent at midnight or later.
So twenty-nine percent of adults with a set bedtime are in bed by nine.
Twenty-nine percent. Nearly a third.
And the age bracket.
That's the part that should end the self-consciousness entirely. Among twenty-five to thirty-four-year-olds with a set bedtime, thirteen percent report an eight PM bedtime. Twenty-two percent report nine PM. Thirty percent report ten PM. Twenty-two percent report eleven. Seven percent midnight or later.
Add the first two.
Thirty-five percent. Thirty-five percent of men and women Daniel's age with a set bedtime are in bed by nine PM. His eight PM is unusual only relative to his own history. It's not unusual relative to his cohort.
And fifty-five percent of adults overall are setting a bedtime of ten or eleven.
Which is the modal adult behavior. Ten PM isn't early. Ten PM is the center of the distribution.
Now the chronological piece. What happens to chronotype as people age?
There's a large analysis of American Time Use Survey data, fifty-three thousand six hundred and eighty-nine respondents, and it found that chronotype lateness peaks at about eighteen and a half years for women and nineteen for men. Then it shifts steadily earlier across the lifespan. Average mid-sleep for people between twenty-five and sixty-five ranges from nearly four AM down to just under three AM.
Mid-sleep being the midpoint between falling asleep and waking.
Right. And there are objective data from NHANES that show the same curve. Median sleep midpoint peaks around age twenty, roughly four AM on weekdays, then declines to a minimum around age forty, about two thirty in the morning, and then stabilizes.
So there's a genuine biological shift happening across exactly the decades Daniel is living through.
There is. But here's the caveat that matters more than the average. The Fischer paper found that extreme early and late chronotypes exist at every age. Every age. The range of chronotypes across individuals spans almost ten hours.
Ten hours.
Which means Daniel being a night owl in his thirties is entirely normal, and someone his age going to bed at nine is entirely normal, and both of those people can be standing next to each other at the same party.
The average tells you nothing about the individual.
The average tells you what the population is doing. It doesn't tell you what's healthy for one person. And this is where the second half of Daniel's question gets interesting, because he's not just asking whether it's normal. He's asking whether there's a reason not to embrace it.
What does the parental sleep data say?
The parental sleep data say his situation is not imagined and not unusual. There's a prospective study of first-time fathers, and it tracked disturbed sleep from the third trimester through two months postpartum. In the third trimester, twenty-nine point six percent had disturbed sleep. At two months postpartum, that was forty-four point seven percent. And among men who were already sleeping poorly before the baby arrived, seventy point six percent still had disturbed sleep at two months.
Nearly half of first-time fathers.
Nearly half. And the predictors were parenting stress and depressive symptoms. Not the baby's temperament in isolation. The load.
So Hannah's strategy. Go to bed early to bank sleep before the interruptions start.
It's the rational response to the actual problem, and it's consistent with what the literature implies. Infant night-waking is unpredictable. You can't schedule around it. But you can reduce the proportion of your total sleep that gets eaten by interruptions if you start earlier. Going to bed at nine when you know you'll be woken at one and four is a different proposition than going to bed at midnight and getting woken an hour in.
You're front-loading the sleep you can actually get.
And Daniel's framing of this as a fallback position, as a kind of defeat, is the thing the data should correct. This isn't lowering your standards. It's matching your behavior to your constraints.
Now there's one more piece on the biology side that supports the same conclusion.
The asymmetry of circadian shifting. The body clock can re-advance by roughly one to one and a half hours per day under optimal conditions. And shifting later is easier than shifting earlier.
So for a night owl trying to move earlier, he's pushing against the grain of his own system.
He's doing the hard direction. Someone who wants to stay up later is working with their biology, not against it. Someone who wants to move everything an hour earlier is fighting for every minute of that shift.
Which means the answer to "should I fight or accommodate" isn't a tie. The biology has a preference.
It does. And the epidemiological data back it up. There's no demographic or health argument against an eight or nine PM bedtime for a man in his thirties. The "old people" thing is a cultural narrative, not a fact about bodies.
Let me put the two questions back together, because I think Daniel is actually asking one question in two parts.
Go ahead.
The first part establishes that reversing an evening wind-down is hard for circadian reasons, not just attentional ones. The second part establishes that modifying his schedule is both normal and supported. So the practical conclusion is that he should stop trying to treat his evening as a recoverable resource and start treating it as a boundary.
That's the read. And the reason it's hard for him specifically is that he's been a night owl his whole life. His identity is built around the late evening being his best working hours. Giving that up feels like losing something, not adjusting something.
Which is why he's typing questions into ChatGPT instead of just going to bed.
Which is exactly why. The data don't resolve the identity problem. They just remove the factual objections to the change.
Let me flag the limits of what we actually know here, because I don't want to overstate it.
Please.
There's no single study that directly tests reversing an initiated wind-down to resume cognitively demanding work. The mechanism is well-supported by adjacent literature, the wake maintenance zone, sleep inertia, the two-process model, but nobody has run that specific experiment.
Agreed. And the term "circadian flexibility" isn't an established scientific term. The closest established concepts are sleep inertia, the wake maintenance zone, and circadian phase. Daniel invented a useful phrase, but it doesn't map onto a body of literature with that name.
And the parenting-specific bedtime distribution. Is there a dataset?
No. The AASM and NHANES and ATUS datasets aren't segmented by parental status. So the claim "parents of young children go to bed earlier" is plausible and it's consistent with the sleep-loss literature, but no dataset directly measures it. That specific statistic doesn't appear to exist.
Which is a real gap, given how many people are living through exactly this.
It's a real gap. The mechanism is well understood and the population-level data are well understood, and the intersection of the two is basically unstudied.
There's a dimension the data can't capture, though.
What's that.
The part where you know all of this, and you're still standing in a living room at nine thirty while everyone else is still there.
Hilbert: Nine fifteen. That was the time I left my sister's wedding.
Hilbert: I had a shift the next morning. Bakery, up at three forty-five. I told my sister I was going, and she said, of course, and I drove home. But I sat in the car for about twenty minutes before I started it.
Why?
Hilbert: I don't know. That's the honest answer. I had done the arithmetic. Six hours if I left then, four and a half if I stayed for the cake. I knew the number. And I still sat in the car.
Was this the bakery in...
Hilbert: It was the one on the corner with the broken awning. I worked the ovens there for two years. And the thing I figured out during those two years was that the hardest part of an early bedtime isn't the bedtime. It's the leaving.
The social cost.
Hilbert: The leaving. Falling asleep was never the problem. I was tired enough. The problem was being the first one out, every time. After a while people stop inviting you to the late things, and they don't mean anything by it, they just know you won't come. So you get the afternoon invitations and not the evening ones.
Did you ever try to stay?
Hilbert: Once. I stayed until eleven. I was useless the next morning. Not dangerous, just slow. And the whole time I was there I was doing the math in my head about the shift. So I wasn't really at the party anyway. I was at the party in body and at the bakery in my head.
So you stopped.
Hilbert: I stopped going to the late things. And after a while I stopped missing them.
Hilbert: It is what it is. The people who matter figure it out. My sister still invites me to everything. She just knows I'll be gone by nine.
Did you ever resent it? The schedule, I mean.
Hilbert: No. I liked the work. Four in the morning is a good time to be awake if you're the only one. There's nobody asking you anything. You just do the thing. The resentment, if there was any, was about the leaving, and that's not the schedule's fault. That's just how it goes.
That's the part the research can't get to.
Hilbert: It's a fact about people, not about bodies. You can prove that going to bed early is normal. You can't prove that it feels normal when you're the one doing it in a room full of people who aren't.
And that's the thing Daniel is actually wrestling with, underneath the two questions.
Hilbert: Probably.
The research can tell him he's not eccentric. It can't tell him the leaving gets easy.
It doesn't, according to Hilbert.
Hilbert: It doesn't get easy. It gets normal. Those are different things.
So where does that leave Daniel.
With the answer to the first question being "it's both, and the circadian part is the part you can't argue with," and the answer to the second question being "you're not unusual, and there is no reason not to, and the reason it feels strange is social rather than biological."
And Hilbert's addition, which is that the social part is real even when you know the data.
The social part is real. The data don't dissolve it. They just clarify that the friction isn't coming from biology. It's coming from the room.
There's one more thing worth flagging, which is what the research doesn't answer.
The understudied part.
No study directly tests what Daniel described. Nobody has run the experiment where you start a wind-down, stop it midway, and try to resume cognitively demanding work. The pieces are all there, sleep inertia, wake maintenance zone, the two-process model, but the specific phenomenon he's describing isn't a named construct in the literature.
And no parenting-specific bedtime distribution exists. Which means as more people navigate a night-owl chronotype colliding with an infant sleep schedule, they're doing it without a direct evidence base. The adjacent evidence is strong. The specific evidence isn't there.
Which is a strange place to be. You can tell Daniel with high confidence that the mechanism is real and the epidemiology is reassuring, and you still can't point him to the study that addresses his exact situation.
You can't. The best you can say is that the inference is well-supported and the direct test hasn't been run.
One forward-looking thought, then. The next time somebody builds a cohort study, they should segment by parental status. Because the intersection of circadian biology and early parenting is a very large population living inside a very small evidence base.
And they should measure the social cost while they're at it. That's the part Hilbert identified that nobody puts in a questionnaire.
Thanks to Hilbert Flumingtop, our producer, for keeping the show running and for the story about the car.
This has been My Weird Prompts.
If you want to send us something, email us at show at my weird prompts dot com. We read everything.
We'll be back soon.
See you tomorrow.
