5 ms·

How Sleep Clears the Brain

127 points by graeham 13y ago

hakanderyal 13y ago
We need more research on sleep, to be able to further improve the 33% of our life (assuming 8 hours of sleep). While ideas like polyphasic sleep are cool to tinker with, the long term risks are too high IMO to 'experiment' as we know so little about sleep.
- epaladin 13y ago
  I'd not mind doing some self-experimentation with sleep, but I have a hard time drawing a line with when to stop and try to evaluate what the effect was. Changes might be so gradual as to not really being able to tell ("do I feel better now than before?"). I wish there were more objective measures for this sort of thing that didn't require a sizeable study population to be reliable. I guess we need more biomarkers for... sleep health?
parasight 13y ago
Sleep works wonders. Sometimes dozing off for 20 minutes after lunch feels like a whole night of sleep to me. It's like a hardware reset.
- JTon 13y ago
  I've heard this before and it's simply amazing to me. I've never been a napper, and I'm envious
tokenadult 13y ago
This press release from two months ago resulted in BBC reporting that was extensively discussed here on HN back then.[1] (I remember this because I had a good night's sleep last night.) As usual for news stories about science discoveries, this is a report of a preliminary primary research finding that may or may not be replicated and extended into the framework of an overall theory of human sleep. The science news cycle gives a crucial role to press releases like the release kindly submitted here,[2] so all of us who sleep will look forward to further research studies to confirm or disconfirm this preliminary finding and help us understand how this mouse study relates to human behavior. The obligatory link for any discussion of a report on a preliminary research result is the article "Warning Signs in Experimental Design and Interpretation"[3] by Peter Norvig, director of research at Google, on how to interpret scientific research. Considering carefully how the current research was done (after a good night's sleep, of course) may help suggest further research that will advance our understanding of what sleep is good for. [1] https://news.ycombinator.com/item?id=6567966 https://news.ycombinator.com/item?id=6567966 [2] http://www.phdcomics.com/comics.php?f=1174 http://www.phdcomics.com/comics.php?f=1174 [3] http://norvig.com/experiment-design.html http://norvig.com/experiment-design.html
- DigitalJack 13y ago
  Maybe this already exists, bit I see a startup opportunity for tagging a study as "of interest" and having a site build a portfolio of information related to that study and finding new studies. Feedback could be given on the relevance of returned results. Something like the app flipboard but have it contain dedicated magazines/journals that gather up and continually build out. Rather than just be a social media stream.
Raphmedia 13y ago
And I only slept 3 hours tonight. However, I've came to master my sleep cycles. Had I slept 3.5 or 5 hours, I would be dead tired right now. I wonder if it's long enough for the brain to clears itself. I tend to sleep only a few hours all month, then I get tired once a month and sleep for around 12 to 14 hours and feel good for the next 30 days or so. I'm probably killing myself. Or perhaps I've found the perfect sleep schedule for me.
- Cacti 13y ago
  If you slept "a few hours per month" you'd be dead.
  - flatline 13y ago
    There's certainly an increased risk of death from e.g. falling asleep at the wheel, but I'm not aware that sleep deprivation can kill you outright. People have gone without any sleep for extensive periods and recovered quite normally after a couple nights' sleep.
    - nmcfarl 13y ago
      It may be that people recover normally with some sleep, however if you don't get that sleep you will die (after you start to hallucinate and then that progresses to dementia). Fatal familial insomnia, is named Fatal for a reason. http://en.wikipedia.org/wiki/Fatal_familial_insomnia#Treatment http://en.wikipedia.org/wiki/Fatal_familial_insomnia#Treatme...
      - aninhumer 13y ago
        I think the other point here is that it's physically impossible to stay awake long enough to die of sleep deprivation unless you actually have a condition preventing sleep.
      - kybernetikos 13y ago
        Diseases with similar causes such as CJD also involve hallucinations and dementia and don't involve lack of sleep, so I'm not 100% convinced that it's the lack of sleep that is causing the death in these cases, although it's an interesting data point.
yawz 13y ago
So, we have a choice! We need to be asleep or anaesthetised.
kailuowang 13y ago
No wonder that, very often, solutions to problems I struggled all day just pop up in the shower next morning.
johnohara 13y ago
I know this is OT, and it's not caused by sleep deprivation, but does anyone know what NIH uses to generate their site. It seems simple and straightforward on the surface but a look at the source reveals it to be tight, templated, and well organized. Seems well designed and executed.
- CitizenKane 13y ago
  It looks like this section of their site is generated using Dreamweaver templates. Actually, cool to see that they're using Crazy Egg for clickmaps of their site. Seems like government websites often lack the usability factor.
  - johnohara 13y ago
    Thanks. Kudos to the design and development team. Nice work.
sshThis 13y ago
This is really cool, and surprisingly similar to the way some software systems are designed. Probably because I have been researching it a lot, but this reminds me of a data replication system, where data builds and builds during active periods of the day, and then at night when activity slows, the back logs are processed and cleared by morning.
pcunite 13y ago
I don't visit .GOV websites anymore. I'm sure the article was interesting.
dsego 13y ago
Not sure what it is with me, but if have no daily obligations and don't force my self to go to sleep at night, I always end up reversing my sleep schedule. Been that way even through high-school, summer vacation comes and I'm there watching tv or playing video games until 5-6am and then I crash and wake up for lunch at 1-2pm. Somehow, the peacefulness of the night is soothing, even my mind seams clearer. Also, sometimes my mind is racing, and I won't go to sleep even if my body is tired. Usually, at that time before crashing, I feel so close to having all the answers.
- simoncion 13y ago
  I'm much the same way. If I don't constantly work at it, I'll end up waking in the late evening and retiring in the early afternoon.
- irremediable 13y ago
  I'm much the same. I wonder if it's related to my screen use and artificial light. What kind of climate do you live in, out of interest?
  - dsego 13y ago
    Mediterranean.
- stephth 13y ago
  I used to behave like this, would go to bed between 2 and 5, despite not liking the consequences (shortened daylight time, city noise while I slept in, being out of sync with others). I think I've gradually fixed it. Still working on being more regular, but these days I'm able to fall asleep at 10h30. Here's what I think helped me personally: - learning about mindfulness (useful to stop/slow down the intense flow of thoughts. I recommend reading/listening to The Power of Now by Eckhart Tolle) - 500 mcg of melatonin at the end of dinner. (the chewable ones sold at Trader Joe's work for me) Required read: http://www.gwern.net/Melatonin http://www.gwern.net/Melatonin During the transition, some days I may have forced myself to get up a little earlier so that I would go to bed earlier, it helped but I avoided doing this too often because it could affect my abilities for days. You want to sleep as much as you need but still go to bed at a reasonable time, not sleep deprive yourself to sleep. Setting your clock is a gradual thing. Maybe you can try to go to bed 15 min earlier every day. EDIT: I'm not a doctor, and I didn't get help from a doctor for this. Take this advice at your own risk. But if you do get help from a doctor and you're offered a prescription for sleeping pills, please remember there may be healthier alternatives. EDIT: Forgot to mention, I think exercising helped me too. If you're not exercising because you can't find the time, you could - like me - start here: http://well.blogs.nytimes.com/2013/05/09/the-scientific-7-minute-workout http://well.blogs.nytimes.com/2013/05/09/the-scientific-7-mi...
mililani 13y ago
I wonder how this works for people who have stayed up for multiple days at a time. I know of people who have underwent sleep deprivation studies for over several days. The current record, I think, is 2 weeks. Personally, I have stayed up 3 nights in a row once to study for a chemistry exam. I've had great mental clarity and focus during the entire time. Yes, it was harder to do than fully rested. But, I have often found that I do better under pressure anyways. By the way, I also ended up getting the highest score on the exam and in the class. It was a brutal chemistry course. Only two of us got A's, and the other guy got an A-.
- shin_lao 13y ago
  Did you use any kind of drugs?

127 points by graeham 13y ago

We need more research on sleep, to be able to further improve the 33% of our life (assuming 8 hours of sleep). While ideas like polyphasic sleep are cool to tinker with, the long term risks are too high IMO to 'experiment' as we know so little about sleep.

epaladin 13y ago
I'd not mind doing some self-experimentation with sleep, but I have a hard time drawing a line with when to stop and try to evaluate what the effect was. Changes might be so gradual as to not really being able to tell ("do I feel better now than before?"). I wish there were more objective measures for this sort of thing that didn't require a sizeable study population to be reliable. I guess we need more biomarkers for... sleep health?

parasight 13y ago

Sleep works wonders. Sometimes dozing off for 20 minutes after lunch feels like a whole night of sleep to me. It's like a hardware reset.

JTon 13y ago
I've heard this before and it's simply amazing to me. I've never been a napper, and I'm envious

tokenadult 13y ago

This press release from two months ago resulted in BBC reporting that was extensively discussed here on HN back then.[1] (I remember this because I had a good night's sleep last night.) As usual for news stories about science discoveries, this is a report of a preliminary primary research finding that may or may not be replicated and extended into the framework of an overall theory of human sleep. The science news cycle gives a crucial role to press releases like the release kindly submitted here,[2] so all of us who sleep will look forward to further research studies to confirm or disconfirm this preliminary finding and help us understand how this mouse study relates to human behavior. The obligatory link for any discussion of a report on a preliminary research result is the article "Warning Signs in Experimental Design and Interpretation"[3] by Peter Norvig, director of research at Google, on how to interpret scientific research. Considering carefully how the current research was done (after a good night's sleep, of course) may help suggest further research that will advance our understanding of what sleep is good for. [1] https://news.ycombinator.com/item?id=6567966 https://news.ycombinator.com/item?id=6567966 [2] http://www.phdcomics.com/comics.php?f=1174 http://www.phdcomics.com/comics.php?f=1174 [3] http://norvig.com/experiment-design.html http://norvig.com/experiment-design.html

DigitalJack 13y ago
Maybe this already exists, bit I see a startup opportunity for tagging a study as "of interest" and having a site build a portfolio of information related to that study and finding new studies. Feedback could be given on the relevance of returned results. Something like the app flipboard but have it contain dedicated magazines/journals that gather up and continually build out. Rather than just be a social media stream.

Raphmedia 13y ago

And I only slept 3 hours tonight. However, I've came to master my sleep cycles. Had I slept 3.5 or 5 hours, I would be dead tired right now. I wonder if it's long enough for the brain to clears itself. I tend to sleep only a few hours all month, then I get tired once a month and sleep for around 12 to 14 hours and feel good for the next 30 days or so. I'm probably killing myself. Or perhaps I've found the perfect sleep schedule for me.

Cacti 13y ago
If you slept "a few hours per month" you'd be dead.
- flatline 13y ago
  There's certainly an increased risk of death from e.g. falling asleep at the wheel, but I'm not aware that sleep deprivation can kill you outright. People have gone without any sleep for extensive periods and recovered quite normally after a couple nights' sleep.
  - nmcfarl 13y ago
    It may be that people recover normally with some sleep, however if you don't get that sleep you will die (after you start to hallucinate and then that progresses to dementia). Fatal familial insomnia, is named Fatal for a reason. http://en.wikipedia.org/wiki/Fatal_familial_insomnia#Treatment http://en.wikipedia.org/wiki/Fatal_familial_insomnia#Treatme...
    - aninhumer 13y ago
      I think the other point here is that it's physically impossible to stay awake long enough to die of sleep deprivation unless you actually have a condition preventing sleep.
    - kybernetikos 13y ago
      Diseases with similar causes such as CJD also involve hallucinations and dementia and don't involve lack of sleep, so I'm not 100% convinced that it's the lack of sleep that is causing the death in these cases, although it's an interesting data point.

yawz 13y ago

So, we have a choice! We need to be asleep or anaesthetised.

kailuowang 13y ago

No wonder that, very often, solutions to problems I struggled all day just pop up in the shower next morning.

johnohara 13y ago

I know this is OT, and it's not caused by sleep deprivation, but does anyone know what NIH uses to generate their site. It seems simple and straightforward on the surface but a look at the source reveals it to be tight, templated, and well organized. Seems well designed and executed.

CitizenKane 13y ago
It looks like this section of their site is generated using Dreamweaver templates. Actually, cool to see that they're using Crazy Egg for clickmaps of their site. Seems like government websites often lack the usability factor.
- johnohara 13y ago
  Thanks. Kudos to the design and development team. Nice work.

sshThis 13y ago

This is really cool, and surprisingly similar to the way some software systems are designed. Probably because I have been researching it a lot, but this reminds me of a data replication system, where data builds and builds during active periods of the day, and then at night when activity slows, the back logs are processed and cleared by morning.

pcunite 13y ago

I don't visit .GOV websites anymore. I'm sure the article was interesting.

dsego 13y ago

Not sure what it is with me, but if have no daily obligations and don't force my self to go to sleep at night, I always end up reversing my sleep schedule. Been that way even through high-school, summer vacation comes and I'm there watching tv or playing video games until 5-6am and then I crash and wake up for lunch at 1-2pm. Somehow, the peacefulness of the night is soothing, even my mind seams clearer. Also, sometimes my mind is racing, and I won't go to sleep even if my body is tired. Usually, at that time before crashing, I feel so close to having all the answers.

simoncion 13y ago
I'm much the same way. If I don't constantly work at it, I'll end up waking in the late evening and retiring in the early afternoon.
irremediable 13y ago
I'm much the same. I wonder if it's related to my screen use and artificial light. What kind of climate do you live in, out of interest?
- dsego 13y ago
  Mediterranean.
stephth 13y ago
I used to behave like this, would go to bed between 2 and 5, despite not liking the consequences (shortened daylight time, city noise while I slept in, being out of sync with others). I think I've gradually fixed it. Still working on being more regular, but these days I'm able to fall asleep at 10h30. Here's what I think helped me personally: - learning about mindfulness (useful to stop/slow down the intense flow of thoughts. I recommend reading/listening to The Power of Now by Eckhart Tolle) - 500 mcg of melatonin at the end of dinner. (the chewable ones sold at Trader Joe's work for me) Required read: http://www.gwern.net/Melatonin http://www.gwern.net/Melatonin During the transition, some days I may have forced myself to get up a little earlier so that I would go to bed earlier, it helped but I avoided doing this too often because it could affect my abilities for days. You want to sleep as much as you need but still go to bed at a reasonable time, not sleep deprive yourself to sleep. Setting your clock is a gradual thing. Maybe you can try to go to bed 15 min earlier every day. EDIT: I'm not a doctor, and I didn't get help from a doctor for this. Take this advice at your own risk. But if you do get help from a doctor and you're offered a prescription for sleeping pills, please remember there may be healthier alternatives. EDIT: Forgot to mention, I think exercising helped me too. If you're not exercising because you can't find the time, you could - like me - start here: http://well.blogs.nytimes.com/2013/05/09/the-scientific-7-minute-workout http://well.blogs.nytimes.com/2013/05/09/the-scientific-7-mi...

mililani 13y ago

I wonder how this works for people who have stayed up for multiple days at a time. I know of people who have underwent sleep deprivation studies for over several days. The current record, I think, is 2 weeks. Personally, I have stayed up 3 nights in a row once to study for a chemistry exam. I've had great mental clarity and focus during the entire time. Yes, it was harder to do than fully rested. But, I have often found that I do better under pressure anyways. By the way, I also ended up getting the highest score on the exam and in the class. It was a brutal chemistry course. Only two of us got A's, and the other guy got an A-.

shin_lao 13y ago
Did you use any kind of drugs?