Daniel's flying home from Connecticut today, and last night was the night before the flight. Anyone with an ADHD brain knows exactly what that means. The packing anxiety loop — did I forget something, did I forget something — running on repeat while you lie there in the dark. You're not awake, exactly. But you're definitely not asleep either. It's this hazy netherworld where time passes, you might even remember snippets of the podcast you left playing, and when morning comes you feel worse than if you'd just stayed up.
So Daniel's got four questions. One: what is that weird in-between state between dreaming and being awake, and does it have a name? Two: how is it categorized in terms of sleep states? Three: why can you remember podcast episodes from the middle of the night when you were supposedly asleep? And four: why does an eight-hour night in that state feel worse than pulling an all-nighter?
He also mentioned something I want to sit with. He said it feels like it's possible to go a full eight hours in this netherworld. Not five minutes of drowsiness — a whole night of being nowhere. So what actually happens in your brain on those nights?
Sleep state misperception. That's the clinical term. It's also called paradoxical insomnia, and the paradox is exactly what Daniel's describing. You subjectively experience yourself as awake — maybe the whole night, maybe large chunks of it — but if we'd had you wired up in a sleep lab, the EEG would show sleep architecture. Sometimes normal-looking sleep architecture.
So the brain is asleep but the mind insists it wasn't.
That's the thing. It's not that you can't sleep. It's that your brain doesn't register the sleep it's getting. About five to ten percent of people with insomnia actually have this subtype. They'll come into a sleep clinic, spend the night hooked up to polysomnography, and in the morning they'll say "I didn't sleep a wink." The technician looks at the readout and sees six hours of sleep.
And they're not lying.
They're absolutely not lying. The subjective experience is real. The distress is real. What's broken is the brain's ability to accurately tag sleep as sleep. And for adults with ADHD, this is way more common. A 2023 meta-analysis in Neuroscience and Biobehavioral Reviews found adults with ADHD have two to three times higher rates of sleep state misperception compared to neurotypical controls.
So Daniel's ADHD brain is doing exactly what ADHD brains are prone to do. The hypervigilance — did I forget something — keeps the system on alert, and the system on alert doesn't believe it ever powered down.
Right. And to understand why, we need to look at what a normal night is supposed to do. A healthy night of sleep runs through four to six cycles, each about ninety minutes. N1 is that drifting-off stage, very light. N2 is still light sleep but more stable — that's where we spend about half the night. Then N3, slow-wave deep sleep, the really restorative stuff. Then REM, where dreaming and emotional processing happen.
And the deep sleep and REM cluster toward morning.
The first half of the night is heavy on N3. The second half is heavy on REM. That architecture matters because each stage does different things. Deep sleep is when the glymphatic system kicks into high gear — the brain's waste clearance mechanism. REM is when emotional memory gets processed, when the brain takes the sharp edges off yesterday's experiences.
So what happens on Daniel's pre-flight night?
The brain never gets past the lobby. It cycles through N1 and N2 repeatedly — maybe it dips toward deeper sleep briefly, but the hypervigilance yanks it back up. You're stuck in light sleep, restarting the cycle over and over without ever reaching the restorative stages. It's like a computer that keeps booting to the login screen and then crashing before the desktop loads.
And you do that six times in a row and call it a night's sleep.
And that's where the damage stacks up. A 2024 review in Sleep Medicine Reviews looked specifically at what sleep fragmentation does to glymphatic clearance. They found it can reduce it by up to sixty percent. Sixty percent less metabolic waste getting cleared out. That includes beta-amyloid, the protein that builds up in Alzheimer's. So you wake up with your brain still full of yesterday's garbage.
Which explains the headache and the fog.
And it's not just the glymphatic system. During fragmented sleep, your autonomic nervous system never fully downshifts. The sympathetic branch — fight-or-flight — stays partially activated all night. Cortisol stays elevated. Heart rate variability, which should increase during restful sleep, stays suppressed. You're essentially running a low-grade stress response for eight hours.
So you're marinating in stress hormones while also not taking out the trash. That's a terrible combination.
And here's where we get to Daniel's fourth question — why this feels worse than an all-nighter. An all-nighter triggers a completely different neurochemical state. When you stay awake, adenosine builds up in the brain. Adenosine is the molecule that creates sleep pressure — it accumulates with every hour of wakefulness and makes you feel increasingly tired. It's unpleasant, but it's a clean signal. The brain knows exactly what's happening.
It's honest exhaustion.
Honest exhaustion, I like that. With fragmented sleep, you don't get the clean adenosine buildup because you are getting some sleep. But you also don't get the restorative payoff. So you wake up with neither the benefits of sleep nor the clear signal of sleep deprivation. You just feel wrecked in a way that's hard to pin down.
And there's a third mechanism on top of glymphatic failure and autonomic dysregulation. You mentioned REM and emotional processing.
Right. Without REM sleep, the brain can't properly process emotional experiences from the previous day. REM is when the amygdala and the prefrontal cortex do this intricate dance — the amygdala reactivates emotional memories, and the prefrontal cortex helps recontextualize them, strip away some of the visceral charge. If you skip REM, you wake up with unresolved emotional residue. For Daniel, that's the packing anxiety. It never got processed. It's still sitting there in the morning, as fresh as it was at midnight.
So the anxiety that kept him awake also never got filed away, because the filing system was offline.
And that creates a compounding effect. The next night, there's more anxiety because you remember how bad last night was. It's a vicious cycle.
Let me pull on the podcast thread. Daniel said he can remember episodes he listened to halfway through the night — sometimes accurately. If he was asleep, how is that possible?
This is one of the coolest findings in sleep research recently. During N1 and N2 sleep, the brain is still processing external auditory information. The thalamus — the sensory relay station — doesn't fully close the gates during light sleep the way it does during deep sleep. So sound gets through. And not only does it get through, the brain can encode and consolidate memories from it.
So you're learning while asleep.
In a limited way. A 2022 study in Current Biology demonstrated exactly this. They had participants listen to familiar and unfamiliar audio during light sleep, and the brain showed clear memory encoding for the familiar content. Familiar voices, familiar podcast hosts, familiar show formats — those get processed and stored. Unfamiliar content mostly doesn't.
Which explains why Daniel remembers his regular shows but probably not a random episode of something he's never heard before.
The brain has built predictive models for familiar audio. It can follow the cadence, anticipate the structure, and tuck away snippets even when consciousness is offline. It's not the same as attentive listening — you're not going to ace a quiz on the episode — but you'll retain fragments.
And that probably feeds the misperception. You remember hearing the podcast, so you conclude you must have been awake.
That's exactly the trap. The memory is real. The experience of hearing it was real. But the state you were in when you heard it was sleep — light sleep, but sleep. Your brain just didn't tag it that way.
Let's get to the really weird stuff. Microsleeps.
Oh, this is where it gets fascinating. Microsleeps are brief intrusions of sleep into wakefulness — typically three to fifteen seconds. Your eyes might stay open, you might even keep doing whatever you were doing, but parts of your brain have clocked out.
Three to fifteen seconds. That's terrifying if you're driving.
It's exactly the thing that causes highway accidents where the driver has no memory of the last mile. But what's wild is what's happening in the brain during those seconds. A 2025 study in Nature Communications identified what they call local sleep phenomena. Different brain regions can be in different sleep states simultaneously.
Wait. Part of your brain is asleep and part is awake?
At the same time. The thalamus might show wake-like activity while patches of cortex show slow-wave sleep patterns. It's not a unified brain state — it's a patchwork.
So the "netherworld" Daniel described isn't just a feeling. It's literally true at the neuronal level.
It's literally true. This goes back to groundbreaking work from 2011 — Vyazovskiy and colleagues at the University of Wisconsin. They studied sleep-deprived rats and found that individual cortical neurons would go into a sleep-like off state while the rat was behaviorally awake. The rat was still moving around, still responding to stimuli, but individual neurons were taking little naps.
Zombie neurons.
That's exactly what they are. And when enough of them do it in the right brain regions, you get that subjective experience of being half-awake, half-asleep. The netherworld.
So on Daniel's pre-flight night, what's probably happening is that his hypervigilant brain keeps large swaths of cortex in a wake-like state — enough to maintain the feeling of being awake — while other regions are cycling through sleep. And the sleep regions are doing enough to fragment the night and prevent restorative rest, but not enough to register as "I slept."
The thalamus might be filtering sensory input like it's awake, so he hears the podcast. The auditory cortex might be processing it. But the prefrontal cortex — the part that does self-awareness and metacognition — might be in a sleep state, which is why he can't quite tell whether he's awake or not.
And the memory encoding regions are doing just enough to grab fragments.
Right. The hippocampus is still partially online during N1 and N2. It's not doing the full memory consolidation that happens during deep sleep and REM, but it's grabbing bits. Especially for familiar content.
There's one more piece I want to understand. Daniel mentioned that when you take a nap or get a full night's sleep, you know it happened. You fall asleep, you wake up, there's a clear boundary. With this state, there's no boundary. Why?
The boundary is created by sleep onset and sleep offset — the transitions into and out of sleep. During normal sleep onset, there's a distinct shift in brain activity. Alpha waves drop out, theta waves emerge, and you cross a threshold where self-awareness dims. During normal awakening, especially from REM or deep sleep, there's a distinct reactivation of wake-promoting systems. Sleep inertia — that grogginess — is actually the brain rebooting its networks.
And in fragmented sleep, you never get the clean shutdown or the clean boot.
You're constantly hovering right at the threshold. Dipping into N1, surfacing back toward wake, dipping again. Each transition is blurry because you're not going deep enough for the full state change to occur. It's like a light switch that's flickering instead of clicking on or off.
And the sleep inertia piece is interesting in its own right. A 2023 study in Sleep found that waking from N1 or N2 actually produces worse cognitive impairment than waking from deep sleep or REM.
Yes. That's counterintuitive, but it makes sense. Waking from deep sleep feels terrible in the moment — that heavy grogginess — but your brain has actually completed some restorative work. Waking from light sleep, your brain hasn't done anything useful and it's disoriented from the fragmented cycling. The impairment lasts thirty to sixty minutes, sometimes longer.
So you wake up feeling worse and stay feeling worse for longer.
And you have no clear memory of having slept, because the memory-encoding systems that would tag "I was asleep" were never fully engaged. So you conclude you didn't sleep. But your brain did something — just not the something it needed.
Let me circle back to the ADHD piece, because Daniel specifically mentioned it. He said the ADHD part of his brain can't let go of the packing anxiety. What's the mechanism there?
ADHD involves dysregulation of the default mode network — the brain's background chatter system — and difficulty with task-switching and attentional control. At night, a neurotypical brain can disengage from the day's concerns. The prefrontal cortex downregulates, the default mode network quiets, and sleep onset proceeds. In ADHD, that downregulation is impaired. The brain stays in a state of hyperarousal.
Hyperarousal. That's the clinical term?
It is. And it's not just psychological — it's physiological. Elevated heart rate, elevated cortisol, increased high-frequency EEG activity. The brain is literally more awake at a biological level, even when the person is trying to sleep. That hyperarousal makes it harder to initiate sleep, harder to maintain sleep, and harder to register sleep as sleep when it does occur.
The ADHD brain is primed for sleep state misperception on a good night. Add pre-flight anxiety and you've got a perfect storm.
The -analysis I mentioned — the 2023 one — found this isn't just anecdotal. The rates are two to three times higher. It's a robust finding across multiple studies.
What about the first night effect? I remember reading that people sleep worse in sleep labs on the first night because the unfamiliar environment triggers hypervigilance.
The first night effect is a perfect laboratory analogue for what Daniel experiences. In a sleep lab, the first night's data is often discarded because it's so disrupted. The brain — particularly the left hemisphere — stays more alert in unfamiliar environments. It's an evolutionary holdover. If you're sleeping in a new place, part of your brain stays on watch.
Daniel's brain, the night before a flight, is treating his own bed like an unfamiliar environment because the context — tomorrow's travel — has changed the meaning of the night.
The bed is the same, the room is the same, but the brain knows tomorrow is different, and it won't fully power down.
This connects to something Daniel said that I think is really important. He said it feels like it's possible to go a full eight hours in this netherworld state. Not a brief transition, not a few minutes of drowsiness — the whole night. Is that actually possible?
It is. And it's one of the most distressing forms of sleep state misperception. The entire night is spent in N1 and N2, with maybe a few brief dips toward deeper sleep that don't last. The person experiences the whole night as a continuous blur of semi-consciousness. Time passes — they're aware of time passing — but they never feel the distinct click of falling asleep or waking up.
In the morning, the sleep tracker might say seven hours, and the person feels like they've been hit by a truck.
That's the cruelest part. The sleep duration is technically there. The hours were spent in bed. But the quality — the architecture — was so degraded that the restorative value was near zero. It's sleep that doesn't do what sleep is supposed to do.
I've said before that sleep math can be mathematically true but biologically false. This is the perfect example.
It absolutely is. Eight hours of fragmented light sleep is biologically false sleep. The glymphatic system didn't get its cleaning cycle. The emotional processing didn't happen. The memory consolidation was patchy at best. You'd have been better off, neurobiologically, staying awake for four hours and then getting four hours of deep, uninterrupted sleep.
Even though the total sleep time is the same.
Even though the total is the same. The architecture matters more than the duration. This is why shift workers have such terrible health outcomes even when they technically get enough hours. Fragmented sleep at the wrong circadian phase is not the same product as consolidated sleep at night.
If Daniel's listening to this on the other side of his flight, having had exactly this kind of night, what's actually happening in his brain right now?
Right now, his glymphatic system is running at maybe forty percent efficiency from last night. His cortisol is elevated. His prefrontal cortex — the part that does executive function, decision-making, emotional regulation — is operating with reduced metabolic support because it didn't get its overnight reset. His amygdala is probably a little hair-trigger because REM didn't process yesterday's emotional load. And he's got sleep inertia that could last an hour after waking.
When he feels foggy and irritable and can't quite think straight, it's not weakness. It's biochemistry.
It's biochemistry. Every one of those symptoms has a specific neurobiological cause. The brain fog is accumulated metabolic waste. The irritability is unprocessed emotional residue. The cognitive sluggishness is incomplete synaptic renormalization — during deep sleep, the brain prunes and strengthens synaptic connections, and that didn't happen.
The synaptic pruning piece is underappreciated. Most people think sleep is just rest. It's actually active maintenance.
It's incredibly active. During deep sleep, your brain is physically cleaning itself — cerebral spinal fluid pulses through and flushes out waste. During REM, it's running emotional simulations, processing memories, integrating experiences. None of that is passive. It's energy-intensive work.
On a fragmented night, the brain is expending energy being half-awake while also failing to do the maintenance work. It's the worst of both worlds.
It's running the engine without going anywhere. Burning fuel, producing waste, getting no mileage.
Let me ask you something. If fragmented sleep is so damaging, why does the brain default to it under stress? Wouldn't it be better to just stay fully awake until the stress resolves?
That's a really interesting question. The answer is probably evolutionary. For most of human history, sleeping in an unfamiliar or threatening environment meant genuine danger. Predators. Rival groups. The brain that stayed slightly vigilant — that kept one hemisphere partially awake, that cycled through light sleep without going deep — that brain survived to reproduce. Deep sleep makes you vulnerable. Light, fragmented sleep is a compromise between getting some rest and staying safe.
Daniel's pre-flight brain is running Stone Age software in a modern context. There's no predator. But the brain treats the impending disruption of travel the same way it would treat a lion in the bushes.
The amygdala doesn't know the difference between a flight to Connecticut and a threat to survival. It just registers "something is different, something is coming, stay alert." And the sleep system responds accordingly.
That's both comforting and infuriating.
Welcome to having a human brain.
Hilbert: Sixty percent reduction in glymphatic clearance. You said up to sixty percent.
That's what the 2024 review found, yes.
Hilbert: It's higher. I used to count these things. Well — not glymphatic clearance specifically. But sleep fragmentation in shift workers. I ran the overnight board at WKND in Hartford for three years, early nineties. Midnight to six AM, solo operator. The station manager was a man named Gus Pirelli who was convinced his overnight guy was asleep at the board. Which I was, most nights. But the logs were perfect. Every ad played on time, every station ID at the top of the hour, every phone-in request handled.
You were doing the job while asleep.
Hilbert: For hours at a stretch. Gus installed a camera — one of those early black-and-white closed-circuit things, cost him nine hundred dollars — and showed me the footage. I'm sitting there, eyes open, cueing up carts, taking calls. But I have no memory of any of it. Four-hour gaps. The overnight log looks like someone was paying attention, but the someone wasn't me. Not the me that remembers things.
That's local sleep in action. Your motor cortex and auditory processing regions were online enough to execute the mechanics of the job. But your hippocampus and prefrontal cortex — memory encoding and self-awareness — were offline.
Hilbert: Gus kept the tapes. Said it was the most impressive thing he'd ever seen. Also said if I ever did it again he'd fire me. But I couldn't stop it. The board had this hum — a sixty-cycle ground loop thing — and the combination of that hum and the quiet of the studio at three AM, it was like a switch. I'd be there, and then I wouldn't.
Did you ever mess up? Play the wrong ad, miss a call?
Hilbert: Never. Not once in three years. That's what drove Gus crazy. He wanted to catch me making a mistake so he could justify firing me, but the work was flawless. Better than when I was fully awake, honestly. When I was awake I'd overthink the segue, talk too much over the intro. Asleep, I was efficient.
There's actually research on this — well-practiced motor sequences can be executed without conscious oversight. The basal ganglia and cerebellum can run the show while the cortex sleeps. Your radio board operation had become procedural memory. You didn't need to be conscious to do it.
Hilbert: I'd wake up at six AM with a coffee in my hand I didn't remember pouring, and the next shift guy would walk in and say "smooth night?" and I'd say "yeah, smooth" because I had no idea. I'd check the log and it was perfect. My own handwriting, perfectly legible. No memory of writing it.
That's unsettling.
Hilbert: It was my life for three years. The thing is, I felt exactly what Daniel's describing. I'd go home at six thirty AM, sleep till two in the afternoon — real sleep, deep sleep — and I'd still feel worse than if I'd pulled an all-nighter in college. The fragmented half-sleep at the board was doing something to my brain that the daytime recovery sleep couldn't fully fix.
Because the glymphatic system is most active during deep sleep, and it's regulated by circadian timing. Sleeping during the day — even deep sleep — doesn't produce the same glymphatic clearance as sleeping at night. So you were accumulating waste during your overnight shifts and never fully clearing it.
Hilbert: I lasted three years and then I quit. Took a job at a print shop. Daytime hours. Gus said I was the best overnight operator he ever had. I told him I couldn't remember being there and he thought I was joking.
Hilbert, when you were in that state at the board — did you know you were asleep?
Hilbert: No. That's the part that still bothers me. I would have sworn under oath I was awake. Tired, sure. But awake. The footage showed my eyes were open, I was moving, I was doing the job. There was no external sign I was asleep. But my brain — the part that knows it exists — was gone. The netherworld Daniel's talking about, that's exactly where I lived.
That's the core of sleep state misperception. The brain regions that generate self-awareness and episodic memory can be in a sleep state while other regions remain behaviorally awake. You can't remember being asleep because the system that would form that memory was offline. So you fill in the gap with the assumption that you were awake the whole time.
Hilbert: Three years of gaps. I filled them all in.
The brain is an extremely confident liar.
Hilbert: It's not lying exactly. It's just... completing the story with the best available information. The best available information was that I was sitting at a board playing records. Obviously I was awake. Except I wasn't.
What finally convinced you?
Hilbert: The tapes. Gus showed me forty-five minutes of footage from a Tuesday night in March, 1993. I watched myself answer the request line, write down a song title, find the cart, cue it up, play it, and log it. All while — and this is the part I still can't explain — all while my face was completely blank. Not tired. Blank. Like nobody was home. I looked at that footage and thought, that's not me. But it was.
Did you keep the tapes?
Hilbert: Gus kept them. He probably threw them out when the station switched to digital in '98. But I think about them. Especially when I can't sleep.
You know, Hilbert, your experience actually illustrates something the research is only starting to grapple with. We've been talking about sleep state misperception as a pathology — something that goes wrong. But what you did at that board was functional. You performed a complex job for hours with no errors. Maybe what we call misperception is, in some contexts, an adaptation. The brain's ability to run essential operations while the conscious self is offline.
Hilbert: I wouldn't recommend it.
No, I wouldn't either. The long-term cost is clear. But the fact that the brain can do it at all is remarkable.
Hilbert: Remarkable is one word for it. Gus had another word. He called it "spooky." Said I was like a ghost running the board. A ghost who played too much Steely Dan.
That's the real crime.
Hilbert: It was the nineties. Everyone played too much Steely Dan.
The question that sticks with me, coming out of all of this, is whether the brain defaults to fragmented sleep under stress for a reason we haven't fully appreciated. We've been talking about it as a failure mode — and for modern life, it is. But for most of human history, the ability to stay partially vigilant while still getting some rest might have been the difference between waking up and not waking up at all. The netherworld might be a feature that became a bug.
Now we have wearable sleep trackers that can show people exactly what their sleep architecture looks like. Someone like Daniel could look at his data after a pre-flight night and see that he actually got five hours of N1 and N2 sleep, even though he felt awake the whole time. Does knowing that help?
It can. For some people with sleep state misperception, seeing the objective data is actually therapeutic. It breaks the cycle of anxiety about not sleeping. You learn to trust that your body is getting something, even when your mind insists otherwise. But it's not a cure. The subjective experience is still miserable. Knowing you were in N2 doesn't make the morning fog lift any faster.
The underlying hyperarousal — the ADHD brain's difficulty downshifting — that doesn't go away just because you have a graph.
Right. The tracker tells you what happened. It doesn't change what happened. But it might change how you feel about what happened, and for some people, that's enough to break the vicious cycle. "I didn't sleep at all" becomes "I slept poorly, but I slept." That reframe matters.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop for keeping the board running — presumably while fully conscious.
If you want to send us your own weird sleep experiences or ask a question like Daniel did, email the show at show at my weird prompts dot com. Or find us at my weird prompts dot com, where every episode lives.
We'll be back soon. Until then, may your deep sleep be deep and your glymphatic system fully operational.
