Disclaimer: I have not had time to read many of the linked papers and links yet. I have checked out "Beggars in Spain" from my local library and look forward to reading it.
I'm interested in this topic because I seem to require a lot of sleep. Without having read all, or even most, of the links, it seems like much of this content focuses on the *cognitive* impacts of sleep/not sleeping. My need seems to be more focused on athletic performance and my immune system.
I have a high training volume of running/skiing/climbing and notice that I perform at a much lower level when I don't sleep ~8 hours. The heart rate & sleep data I collect on myself seems to confirm this subjective feeling. I also tend to get sick quickly if I maintain my training volume with consistent low sleep over the period of a few days or a week. Friends that are professional runners or otherwise semi-professional athletes that I know seem to focus on sleep as a core part of their training.
I wonder if you might synthesize any of your readings that relate to this aspect of the topic?
Let's start with the immune system part. The immune system uses a lot of energy, particularly when sick. Sleep deprivation means you're burning more calories which means there's less energy to support your immune system. I think this is mostly mediated by stress hormones. After a night of bad sleep, cortisol levels are higher which supresses immune response. At the opposite end of the spectrum, when you're sick the immune system has more energy demand and you release hormones that make you sleepy/achy to reduce calorie expenditure.
I'm less certain about the loss of training performance after sleep deprivation. Less rest means reduced glycogen stores? But that would imply that eating lots of carbs in the morning would fix the problem by the afternoon or evening. There's also the fact that healing and growing muscles is an immune-mediated process so immunosupression would mean slower recovery. Countering all this, you'd expect adrenaline to increase performance in some areas.
There's probably a psychological component involved in perceived effort. It's interesting to note that ~50% of insomniacs get enough sleep, but remember their sleep as being disrupted. So perhaps some sort of CBT-i would help athletes too.
This addresses the "metabolite clearance" theory of sleep mainly through the glymphatic/tau/amyloid story: sleep physically flushes protein aggregates out of the brain, so without enough sleep you get neurodegeneration. The counterarguments offered are reasonable against that version. Neurons in the heart and lungs function constantly, REM sleep involves high brain activity, insomniacs don't seem to get more dementia, short sleepers might get less, narcoleptics get Alzheimer's at normal rates.
But there's a different version that these counterarguments don't have the same evidential strength against: reactive oxygen species (ROS) clearance. ROS are produced as a byproduct of mitochondrial energy production in all cells. They accumulate during wakefulness because waking metabolism is expensive. The body's main intracellular antioxidant for neutralizing ROS is glutathione, a small molecule made from three amino acids: glutamate, cysteine, and glycine.
A Drosophila study https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.2005206 bears directly on this. Multiple short-sleeping mutant lines (with mutations in unrelated genes) were all more vulnerable to oxidative stress. Forcing normal flies to sleep more made them survive oxidative stress better. And reducing ROS specifically in neurons made the flies sleep less.
This is compatible with the temporal niche thesis. If sleep originally evolved for niche specialization, and the body then co-opted the resting period for ROS clearance, then how much sleep is needed is partly a function of how fast you can clear ROS. That rate depends in part on glutathione availability, which depends in part on substrate supply.
Let's look at the specific counterarguments one by one as applied to ROS hypothesis:
1. Neurons in the heart and lungs function constantly.
ROS clearance isn't about neurons needing rest. It's about whole-organism oxidative damage accumulating faster during wakefulness than the antioxidant system can neutralize it. Cardiac neurons have a relatively stable metabolic load; the brain's load fluctuates with wakefulness.
2. REM involves high brain activity.
Glutathione-dependent ROS neutralization is a chemical process, not a function of neural quiescence.
3. Brain clearance is reduced during sleep.
This is about glymphatic flow (physical flushing of interstitial fluid). Glutathione operates inside cells.
4. Insomniacs don't get more dementia / short sleepers get less / narcoleptics get Alzheimer's at normal rates.
These are about protein aggregation, not oxidative stress. The ROS model predicts an interaction between sleep duration and antioxidant substrate availability. A short sleeper with adequate glutathione substrates could clear ROS in less time. The Drosophila result (reducing ROS → less sleep needed) is consistent with this.
Glycine is rate-limiting for glutathione production https://pmc.ncbi.nlm.nih.gov/articles/PMC9879756/ in many contexts, and many people may have suboptimal glycine intake for maximal glutathione synthesis. This would limit ROS clearance rate during sleep, meaning they need more sleep to achieve the same degree of clearance.
This is super interesting thank you. If I'm understanding correctly, gluthianone clearance is one of the things that became accidentally load bearing during sleep. Supplementing glycine (and thus producing more gluthianone) would address this and reduce sleep need somewhat. I'll add this to my list of interesting things for sleep!
Personally I feel that I require a lot of sleep, making it always super difficult to wake up in the morning, so this sort of research is very interesting to me.
No self-experimentation for me then. It would be really cool we could gain an extra hour or two in the day. Not that 100% of my hours are optimized already, but living in a world that's designed for 8 hours of sleep, when you could get by with 6 without negative side effects would be a major advantage.
I don’t need to sleep *significantly* longer, but it’s completely impossible for me to function on less than 7 hours of sleep, and absent intervention I seem to need a constant 8.5 hours. I know people who are functioning on 6, while still apparently living a good life, so there’s a lot of variance.
I think the barrier between what is considered a disease/condition, and what’s just a normal trait is how much of a problem it causes for the person who has it. Needing an extra hour of sleep than what Americans get on average takes an additional hour from my day, but most people already waste much of their day mindlessly browsing TikTok or whatever, so I’m not really at a disadvantage. At least sleep leaves me feeling refreshed and equipped to handle life, whereas vegetating in front of screen usually makes you feel terrible. I wouldn’t consider it a condition any more than any other trait that’s within a normal range of human experience.
9+ hours would be a problem though. Fortunately I don’t need that much.
I generally don't comment so I apologize if this comes over as rude but you seemed very lazy in doing research on the other theories of sleep and the negative, wide-spread health consequences that people face when they lose sleep. Sure, dementia and altzheimer's incidence isn't higher but all-cause mortality is.
I'm happy to look at research on this, but I'd want to see you lay out an argument for why those explanations are *better* than the temporal niche explanation. Sleep certainly has other functions, but those other functions don't require us to sleep exactly 7 hours per night.
The difficulty here is that I don't believe there *is* a primary reason for why sleep exists. I'm not deriding the notion that energy conservation nor the need to match metabolic output with opportune feeding/mating/rearing moments are important in terms of evolution, they certainly are and one only needs to look at migratory birds or see faring mammals to see that sleep is a malleable thing.
However, just because a frigatebird in the midst of its migration can sleep only an hour a day using only half of it's brain does not mean that one could suppress a **human's** sleep response without demerit.
If sleep were as simple and as biased as you suggest to function **predominately** as a regulator for energy consumption then I ask you to give your hypothesis as to why in humans that are not naturally predisposed to requiring less sleep we see that sleep deprivation weakens the immune system, increases the risk of CVD, makes people drowsy and irritable, reduces people's cognitive ability and working memory, slows muscle recovery and growth, potentially causes metabolic dysfunction, etc.
I am willing to give you citations for all of those effects and more if you feel so inclined but in no uncertain terms there are clear downsides to a lack of sleep and I believe that the onus is on you to explain how you could tangibly see one pathway towards not only sleep reduction but also the remediation of harm that such a sleep reduction would, probably, cause in people.
Disclaimer: I have not had time to read many of the linked papers and links yet. I have checked out "Beggars in Spain" from my local library and look forward to reading it.
I'm interested in this topic because I seem to require a lot of sleep. Without having read all, or even most, of the links, it seems like much of this content focuses on the *cognitive* impacts of sleep/not sleeping. My need seems to be more focused on athletic performance and my immune system.
I have a high training volume of running/skiing/climbing and notice that I perform at a much lower level when I don't sleep ~8 hours. The heart rate & sleep data I collect on myself seems to confirm this subjective feeling. I also tend to get sick quickly if I maintain my training volume with consistent low sleep over the period of a few days or a week. Friends that are professional runners or otherwise semi-professional athletes that I know seem to focus on sleep as a core part of their training.
I wonder if you might synthesize any of your readings that relate to this aspect of the topic?
Ah this is a good aspect to cover.
Let's start with the immune system part. The immune system uses a lot of energy, particularly when sick. Sleep deprivation means you're burning more calories which means there's less energy to support your immune system. I think this is mostly mediated by stress hormones. After a night of bad sleep, cortisol levels are higher which supresses immune response. At the opposite end of the spectrum, when you're sick the immune system has more energy demand and you release hormones that make you sleepy/achy to reduce calorie expenditure.
I'm less certain about the loss of training performance after sleep deprivation. Less rest means reduced glycogen stores? But that would imply that eating lots of carbs in the morning would fix the problem by the afternoon or evening. There's also the fact that healing and growing muscles is an immune-mediated process so immunosupression would mean slower recovery. Countering all this, you'd expect adrenaline to increase performance in some areas.
There's probably a psychological component involved in perceived effort. It's interesting to note that ~50% of insomniacs get enough sleep, but remember their sleep as being disrupted. So perhaps some sort of CBT-i would help athletes too.
If I may ask, what is your day job? I’m curious what your professional /academic background is in given the topics you cover.
I would prefer not to say! But you should assume I'm not an expert in ~everything I write about here.
I actually tend to write less about day job topics than others because this blog is more of an outlet for things I'm not supposed to be working on.
This addresses the "metabolite clearance" theory of sleep mainly through the glymphatic/tau/amyloid story: sleep physically flushes protein aggregates out of the brain, so without enough sleep you get neurodegeneration. The counterarguments offered are reasonable against that version. Neurons in the heart and lungs function constantly, REM sleep involves high brain activity, insomniacs don't seem to get more dementia, short sleepers might get less, narcoleptics get Alzheimer's at normal rates.
But there's a different version that these counterarguments don't have the same evidential strength against: reactive oxygen species (ROS) clearance. ROS are produced as a byproduct of mitochondrial energy production in all cells. They accumulate during wakefulness because waking metabolism is expensive. The body's main intracellular antioxidant for neutralizing ROS is glutathione, a small molecule made from three amino acids: glutamate, cysteine, and glycine.
A Drosophila study https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.2005206 bears directly on this. Multiple short-sleeping mutant lines (with mutations in unrelated genes) were all more vulnerable to oxidative stress. Forcing normal flies to sleep more made them survive oxidative stress better. And reducing ROS specifically in neurons made the flies sleep less.
This is compatible with the temporal niche thesis. If sleep originally evolved for niche specialization, and the body then co-opted the resting period for ROS clearance, then how much sleep is needed is partly a function of how fast you can clear ROS. That rate depends in part on glutathione availability, which depends in part on substrate supply.
Let's look at the specific counterarguments one by one as applied to ROS hypothesis:
1. Neurons in the heart and lungs function constantly.
ROS clearance isn't about neurons needing rest. It's about whole-organism oxidative damage accumulating faster during wakefulness than the antioxidant system can neutralize it. Cardiac neurons have a relatively stable metabolic load; the brain's load fluctuates with wakefulness.
2. REM involves high brain activity.
Glutathione-dependent ROS neutralization is a chemical process, not a function of neural quiescence.
3. Brain clearance is reduced during sleep.
This is about glymphatic flow (physical flushing of interstitial fluid). Glutathione operates inside cells.
4. Insomniacs don't get more dementia / short sleepers get less / narcoleptics get Alzheimer's at normal rates.
These are about protein aggregation, not oxidative stress. The ROS model predicts an interaction between sleep duration and antioxidant substrate availability. A short sleeper with adequate glutathione substrates could clear ROS in less time. The Drosophila result (reducing ROS → less sleep needed) is consistent with this.
Glycine is rate-limiting for glutathione production https://pmc.ncbi.nlm.nih.gov/articles/PMC9879756/ in many contexts, and many people may have suboptimal glycine intake for maximal glutathione synthesis. This would limit ROS clearance rate during sleep, meaning they need more sleep to achieve the same degree of clearance.
I wrote this up as a full essay exploring the implications: "Is fever a symptom of glycine deficiency?": https://www.lesswrong.com/posts/87XoatpFkdmCZpvQK/is-fever-a-symptom-of-glycine-deficiency The fever question is the main thread, but the sleep/ROS/glutathione mechanism is the foundation.
This is super interesting thank you. If I'm understanding correctly, gluthianone clearance is one of the things that became accidentally load bearing during sleep. Supplementing glycine (and thus producing more gluthianone) would address this and reduce sleep need somewhat. I'll add this to my list of interesting things for sleep!
Personally I feel that I require a lot of sleep, making it always super difficult to wake up in the morning, so this sort of research is very interesting to me.
I looked up Orexin supplements to see if this was something you could order on Amazon, and it seems like the only product advertising itself with "Orexin" is a placebo, considering it has "Neanderthal Extract" as one of the ingredients, which gave me a chuckle: https://ibspot.com/us/products/starmel-orexin-plus-150ml-for-anorexia?variant_id=1928146
No self-experimentation for me then. It would be really cool we could gain an extra hour or two in the day. Not that 100% of my hours are optimized already, but living in a world that's designed for 8 hours of sleep, when you could get by with 6 without negative side effects would be a major advantage.
Yeah my hope is that even if orexin agonists fall short for most people, it will at least help people with abnormal sleep schedules.
One thing I find odd is that there isn't a recognized condition for sleeping longer than most people.
I don’t need to sleep *significantly* longer, but it’s completely impossible for me to function on less than 7 hours of sleep, and absent intervention I seem to need a constant 8.5 hours. I know people who are functioning on 6, while still apparently living a good life, so there’s a lot of variance.
I think the barrier between what is considered a disease/condition, and what’s just a normal trait is how much of a problem it causes for the person who has it. Needing an extra hour of sleep than what Americans get on average takes an additional hour from my day, but most people already waste much of their day mindlessly browsing TikTok or whatever, so I’m not really at a disadvantage. At least sleep leaves me feeling refreshed and equipped to handle life, whereas vegetating in front of screen usually makes you feel terrible. I wouldn’t consider it a condition any more than any other trait that’s within a normal range of human experience.
9+ hours would be a problem though. Fortunately I don’t need that much.
I generally don't comment so I apologize if this comes over as rude but you seemed very lazy in doing research on the other theories of sleep and the negative, wide-spread health consequences that people face when they lose sleep. Sure, dementia and altzheimer's incidence isn't higher but all-cause mortality is.
I'm happy to look at research on this, but I'd want to see you lay out an argument for why those explanations are *better* than the temporal niche explanation. Sleep certainly has other functions, but those other functions don't require us to sleep exactly 7 hours per night.
The difficulty here is that I don't believe there *is* a primary reason for why sleep exists. I'm not deriding the notion that energy conservation nor the need to match metabolic output with opportune feeding/mating/rearing moments are important in terms of evolution, they certainly are and one only needs to look at migratory birds or see faring mammals to see that sleep is a malleable thing.
However, just because a frigatebird in the midst of its migration can sleep only an hour a day using only half of it's brain does not mean that one could suppress a **human's** sleep response without demerit.
If sleep were as simple and as biased as you suggest to function **predominately** as a regulator for energy consumption then I ask you to give your hypothesis as to why in humans that are not naturally predisposed to requiring less sleep we see that sleep deprivation weakens the immune system, increases the risk of CVD, makes people drowsy and irritable, reduces people's cognitive ability and working memory, slows muscle recovery and growth, potentially causes metabolic dysfunction, etc.
I am willing to give you citations for all of those effects and more if you feel so inclined but in no uncertain terms there are clear downsides to a lack of sleep and I believe that the onus is on you to explain how you could tangibly see one pathway towards not only sleep reduction but also the remediation of harm that such a sleep reduction would, probably, cause in people.