Listen or watch on your favorite platforms
In this episode, Dr. Matt Walker, Professor of Neuroscience and Psychology and the Founder & Director of the Center for Human Sleep Science at the University of California, Berkeley, joins Dr. Andrew Huberman. Dr. Walker is the author of the international best-selling book Why We Sleep and the host of The Matt Walker Podcast. Drs. Huberman and Walker discuss the biology of sleep, including its various stages and what specifically happens to those stages when we don’t get enough sleep. They also discuss the effects of sunlight, caffeine, alcohol, naps, hormones, exercise, marijuana, sexual activity, and various supplements on sleep. The episode consists of both basic science information and many science-supported actionable tools.
About this Guest
Dr. Matthew Walker
Matthew Walker, Ph.D., is a professor of neuroscience and psychology and founder of the Center for Human Sleep Science at the University of California, Berkley, and the author of the book, “Why We Sleep.”
Andrew Huberman:
Welcome to the Huberman Lab podcast, where we discuss science and science-based tools for everyday life.
Andrew Huberman:
I'm Andrew Huberman and I'm a professor of neurobiology and ophthalmology at Stanford School of Medicine. Today I have the pleasure of introducing Dr. Matthew Walker as our guest on the Huberman Lab podcast. Dr. Walker is a professor of neuroscience and psychology at the University of California, Berkeley. There his laboratory studies sleep. They study why we sleep; what occurs during sleep, such as dreams, and why we dream. Learning during sleep, as well as the consequences of getting insufficient or poor-quality sleep on waking states. Dr. Walker is also the author of the international bestselling book "Why We Sleep." Our discussion today is an absolutely fascinating one for anyone that's interested in sleep learning or human performance of any kind. Dr. Walker teaches us how to get better at sleeping. He also discusses naps, whether or not we should or should not nap, whether or not we can compensate for lost sleep, and if so, how to best do that.
Andrew Huberman:
We discuss behavioral protocols and interactions with light, temperature, supplementation, food, exercise, sex, all the variables that can impact this incredible state of mind and body that we call sleep. During my scientific career, I've read many papers about sleep and attended many seminars about sleep. Yet my discussion with Dr. Walker today revealed to me more about sleep, sleep science, and how to get better at sleeping than all of those papers and seminars combined. I'm also delighted to share that Dr. Walker has started a podcast. That podcast, entitled the Matt Walker podcast, releases its first episode this month and is going to teach all about sleep and how to get better at sleeping. So be sure to check out the Matt Walker podcast on Apple, Spotify or wherever you listen to podcasts. Before we begin, I'd like to mention that this podcast is separate from my teaching and research roles at Stanford.
Andrew Huberman:
It is, however, part of my desire and effort to bring zero-cost-to-consumer information about science and science-related tools to the general public. In keeping with that theme, I'd like to thank the sponsors of today's podcast. And now my discussion with Dr. Matt Walker. Great to finally meet you in person.
Matthew Walker:
Wonderful to connect. I mean, it's been too long, but I suspect it would've been a shorter time before we'd met, lest the pandemic. Thank you.
Andrew Huberman:
Thank you. Yeah. I'm delighted that we're finally sitting down face-to-face. I've been tracking your work both in the Internet sphere and I read your book and loved it. And also from the perspective of science, you actually came to Stanford a couple of years ago and gave a lecture for Brain Mind.
Matthew Walker:
Oh yeah, yeah, yeah, yeah.
Andrew Huberman:
And of course you talked about sleep and its utility and its challenges and how to conquer it, so to speak. Let's start off very basic. What is sleep?
Matthew Walker:
Sleep is probably the single most effective thing you can do to reset your brain and body health. So that's a functional answer in terms of what is sleep, in terms of its benefits. Sleep as a process, though, is an incredibly complex physiological ballet, and if you were to recognize or see what happens to your brain and your body at night during sleep, you would be blown away. And the paradox is that most of us, and I would think this too if I wasn't a sleep scientist, we go to bed, we lose consciousness for seven to nine hours, and then we sort of wake up in the morning, and we generally feel better. And in some ways, that denies the physiological and biological beauty of sleep. So upstairs in your brain when you're going through these different stages of sleep, the changes in brainwave activity are far more dramatic than those that we see when we're awake.
Matthew Walker:
And we can speak about deep sleep and what happens there. REM sleep is a fascinating time which is another stage of sleep, often called dream sleep, which is rapid eye movement sleep. That stage of sleep, some parts of your brain are up to 30% more active than when you're awake. So again, it's kind of violating this idea that our mind is dormant, and our body is just simply quiescent and resting. So I would [be] happy to just sort of double-click on either one of those and also what changes in the body as well. But it is an intense evolutionary adaptive benefit and system. That said though, I would almost push back against an evolved system. When we think about the question of sleep and what sleep is, our assumption has always been that we evolved to sleep. And I've actually questioned that, and I have no way to get in a time capsule and go back and prove this.
Matthew Walker:
But what if we started off sleeping, and it was from sleep that wakefulness emerged? Why do we assume that it's the other way around? And I think there's probably some really good evidence that sleep may have been the proto-state, that it was the basic, fundamental living state. And when we became awake, as it were, we always had to return to sleep. In some ways at that point, sleep was the price that we paid for wakefulness. And that's another way of describing what sleep is. But again, I think it sort of denies that the active state of sleep, it's not a passive state of sleep either. And then finally, you can say what is sleep across different species, and in us, human beings, and in all mammalian species and avian species as well, sleep is broadly separated into these two main types.
Matthew Walker:
And we've got non-rapid eye movement sleep on the one hand, and then we've got rapid eye movement sleep on the other. And we can speak about how they unfold across the night and their architecture because it's not just intellectually interesting from the perspective of what sleep is. It's also practically impactful for our daily lives. And I'd love to sort of go down that route too, but you navigate, you tell me. I can-
Andrew Huberman:
Let's definitely go down that route. So you mentioned how active the brain is during certain phases of sleep. When I was coming up in science, REM sleep, rapid eye movement sleep, was referred to as paradoxical sleep. Is that still a good way to think about it? Paradoxical because the brain is so active, and yet we are essentially paralyzed. Correct?
Matthew Walker:
Yeah. It really is a paradox. And where that came from was simply the brainwave recordings, that if all I'm measuring about you is your brainwave activity, it's very difficult for me sitting outside of the sleep laboratory room to figure out are you awake or are you in REM sleep? Because those two patterns of brain activity are so close to one another, you can't discriminate between them. Yet the paradox is that when you are awake, I go in there and you're sort of sitting up, you're clearly conscious and awake. But yet when you go into REM sleep, you are completely paralyzed. And that's, I think that's part of the paradox. But the paradox really just comes down to two dramatically different conscious states, yet brain activity is dramatically more similar than different. And the way I can figure out which of the two you are in is by measuring two other signals, the activity from your eyes and the activity from your muscles.
Matthew Walker:
So when we're awake, we will occasionally have these blinks and we'll have sort of saccades. But during REM sleep, you have these really bizarre horizontal shuttling eye movements that occur. And that's where the name comes from. Rapid eye movements.
Andrew Huberman:
Are they always horizontal?
Matthew Walker:
Mostly they are horizontal.
Andrew Huberman:
That's interesting.
Matthew Walker:
And that's one of the ways that we can differentiate them from other waking eye movement activity, because it's not always ... It can be sometimes horizontal, but can it also have diagonal and also vertical in that plane. But then the muscle activity is the real dead giveaway. Just before you enter REM sleep, your brain stem, which is where the dynamics of non-REM and REM are essentially played out and then expressed upstairs in the cortex and downstairs in the body. When we go into REM sleep, and just a few seconds before that happens, the brain stem sends a signal all the way down the spinal cord, and it communicates with what are called the alpha motor neurons in the spinal cord, which control of voluntary skeletal muscles. And it's a signal of paralysis. And when you go into dream sleep, you are locked into a physical incarceration of your own body.
Andrew Huberman:
Amazing.
Matthew Walker:
And why would Mother Nature do such a thing? And it's in some ways very simple. The brain paralyzes the body so that the mind can dream safely, because think about how quickly we would've all been popped out of the gene pool. If I think I'm one of the best skydivers, who can just simply fly, and I've had sometimes those dreams too, and I get up on my apartment window and I leap out ...
Andrew Huberman:
You're done.
Matthew Walker:
You're done. So that's one of the ... That's part of the paradox of REM sleep, both its brain activity similarity, despite the behavioral state being so different, and this bizarre lockdown of the sort of brain of the body itself. Now of course the involuntary muscles, thankfully, aren't paralyzed. So you keep breathing, your heart keeps beating.
Andrew Huberman:
Is this why men have erections during REM sleep and women have vaginal lubrication during sleep?
Matthew Walker:
That's one of the reasons. Part of the other reason, though there, is because of the autonomic activity. So there is a nervous part of our nervous system called the "autonomic nervous system." And it controls many of the automatic behaviors. And some of those are aspects of our reproductive facilities. During REM sleep, what we later discovered is that you go through these bizarre, what we call "autonomic storms," which sounds dramatic, but it actually is when you measure them that you'll go through periods where your heart rate decelerates and drops, and your blood pressure goes down, and then, utterly randomly, your heart rate accelerates dramatically. And what we call the fight-or-flight branch of the autonomic nervous system or the sympathetic nervous system are badly named because it's anything but sympathetic, it's very aggravating. And that all of a sudden fires up, and then it shuts down again. And it's not in any regular way. And it's when you get those autonomic storms; you get very activated from a physiological perspective; that you can have these erections, and you have vaginal discharge, et cetera-
Andrew Huberman:
But you're totally paralyzed.
Matthew Walker:
But you are still paralyzed. There are only two voluntary, two voluntary muscle groups that are spared from the paralysis. Bizarre. One, your extraocular muscles, because if they were paralyzed, you wouldn't be able to have rapid eye movements. And the other that we later discovered was the inner ear muscle. And we've got no good understanding as to why those two muscles groups are spurred from the paralysis. It may have something to do with cranial nerve, but I don't think it's that ... I think it's perhaps something more sensory related. Some people have argued that the reason the eyeballs are spared from the paralysis is because if your eyeballs are left for long periods of time inactive, you may get things such as oxygen sort of issues in the aqueous or vitreous humor. And so the eyeballs have to keep the-
Andrew Huberman:
The drainage systems of the anterior eye are made to require movement. People with glaucoma have deficits in drainage through the anterior chamber. But there I'm speculating. I'm also speculating when I ask this: I would imagine that there are states in waking that also resemble slow-wave sleep. Or rather that there are states that slow-wave sleep also resembles waking states. You've beautifully illustrated how REM sleep can mimic some of the more active brain states that we achieve in waking. What sort of waking state that I might have experienced, or experienced on a daily basis, might look similar to slow-wave sleep, non-REM sleep, if any?
Matthew Walker:
It's a genius way of thinking about it, turns the tables. I love it. We almost never see anything like the true ultraslow waves of deep non-REM sleep. So we spoke about these two stages, non-REM and REM. Non-REM is further subdivided into four separate stages. Stages one through four increasing in their depth of sleep. So stages three and four, that's what we typically call deep non-REM sleep. Stages one and two light-
Andrew Huberman:
So maybe take me through the arc of a night just so that ...
Matthew Walker:
Yeah.
Andrew Huberman:
So I put my head down. Well for you, what time do you normally go to sleep?
Matthew Walker:
So I'm usually sort of around about a 10:30 p.m. guy. And usually I'll naturally wake up sort of a little bit before seven, sometimes before 6:45 or seven. I have an alarm set for 7:04 a.m.
Andrew Huberman:
You heard it here folks. Matt Walker does use an alarm clock.
Matthew Walker:
I rarely ... I'm usually-
Andrew Huberman:
He doesn't recommend it, but he does use it.
Matthew Walker:
Yeah. I usually-
Andrew Huberman:
You're human after all.
Matthew Walker:
Oh. I am so human. And I've had my sleep issues, and I'd love to speak about that too, but it's only just in the event that, because I like to keep regularity too. You've got to keep those two things in balance, and 7:04, just because why not be idiosyncratic? I don't know why we always set things on these hard numbers. So when you-
Andrew Huberman:
So you go to sleep around 10:30. So using you as an example, because I imagine a number of people go to sleep at different times, but 10:30 is about when I go to sleep, 11 is for me. So you go to sleep at 10:30. So for that first, let's say, three hours of sleep, what does the architecture of that sleep look like as compared to the last three hours of your sleep before morning?
Matthew Walker:
Yeah. So I should note that sort of 10:30 to seven, that's just based on my chronotype and my preferential. It's different for different people. I'm not suggesting that that's the perfect sweet spot for humanity's sleep. It's just my-
Andrew Huberman:
But I imagine most people probably go to sleep somewhere between 10 p.m. and midnight and most probably-
Matthew Walker:
... Between like nine and midnight.
Andrew Huberman:
... Between 5:00 a.m. and 7:00 a.m., or 5:30 and 7:30.
Matthew Walker:
Yeah. At least in, if you look at of first world nations, that's a typical sleep profile. So when I first fall asleep, I'll go into the light stages of non-REM sleep, stages one and two of non-REM, and then I'll start to descend down into the deeper stages of non-REM sleep. So after about maybe 20 minutes, I'm starting to head down into stage three non-REM, and then into stage four non-REM sleep. And as I'm starting to fall asleep, as I've cast off from the, usually with me, murky waters of wakefulness, and I'm in the shallows of sleep stages one and two, my heart rate starts to drop a little bit, and then my brainwave pattern activity starts to slow down. Normally, when I'm awake, it's going up and down maybe 20, 30, 40, 50 times a second. As I'm going into light non-REM sleep, it will slow down to maybe 15, 20 and then really starts to slow down. Down to about 10 or eight cycles per second, eight cycle waves per second.
Matthew Walker:
Then as I'm starting to move into stages three and four non-REM sleep, several remarkable things happen. All of a sudden my heart rate really does start to drop. And I'll come back to temperature. I'm going to write temperature down because I always forget these things. Now I'm solidly in the foothills of middle age. So as I'm starting to go into those deeper stages of non-REM sleep, all of a sudden hundreds of thousands of cells in my cortex all decide to fire together, and then they all go silent together. And it's this remarkable physiological coordination of the likes that we just don't see during any other brain state.
Andrew Huberman:
That's really interesting. Having recorded from the brains of animals and a little bit from humans, I don't think I've ever seen the entire cortex, or even entire regions of cortex, light up like that.
Matthew Walker:
Yeah. It's stunning. It's almost like this beautiful sort of mantra chant or this sort of ... It's a slow inhale and then a meditative exhale. Inhale, exhale. And these waves are just enormous in their size.
Andrew Huberman:
And the body is capable of movement at this time. There is no paralysis.
Matthew Walker:
There is no paralysis, but for the most part, muscle tone has also dropped significantly at that point. And then you will, or I will, then stay there for about another 20 or 30 minutes. So now I'm maybe 60 or 70 minutes into my first sleep cycle, and then I'll start to rise back up into stage two non-REM sleep. And then after about 80 or so minutes I'll pop up, and I'll have a short REM sleep period. And then back down I go again, down into non-REM, up into REM. And you do that reliably, repeatedly. And I will be doing that. And I do do that every 90 minutes. At least that's the average for most adults. It's different in different species. What changes, to your question, is the ratio of non-REM to REM within that 90-minute cycle as you move across the night.
Matthew Walker:
And what I mean by this is in the first half of the night, the majority of those 90-minute cycles are comprised of lots of deep non-REM sleep. That's when I get my stage three and four of deep non-REM sleep. Once I push through to the second half of the night, now that seesaw balance changes, and instead the majority of those 90-minute cycles are comprised either of this lighter form of non-REM sleep, stage two non-REM sleep, and much more an increasingly more rapid eye movement sleep. And the implication that I was sort of speaking about pragmatically is let's say that I have to, and I usually never do early morning flights or red eyes just because I'm a mess if that happens. I'm not suggesting other people shouldn't-
Andrew Huberman:
I'm suggesting people not do that. Every time I've taken a red eye or I've done that, two or three days later, I get some sort of general feeling of malaise. My brain doesn't work as well. I think red eye should be abolished. For the pilots too. I mean ...
Matthew Walker:
Oh. And we can think about those-
Andrew Huberman:
... And for emergency room, I mean long shifts have been shown to lead to physician-induced errors that lead to a lot of fatalities. I mean, there are a lot of reasons why staying up too long or being up at the wrong times if you're not adapted to it is just terrible.
Matthew Walker:
The data in all of those cases, and particularly physicians too, there was some recent data looking at suicidality, and the rates of suicide in training physicians are far above the norm. And I don't suspect that their schedules are helping them. I suspect that sleep is a missing part of that explanatory equation.
Andrew Huberman:
I teach medical students and they're phenomenal. But yeah. They're under extremely challenged conditions.
Matthew Walker:
We shouldn't put them under those-
Andrew Huberman:
No. it's not optimizing performance. I have-
Matthew Walker:
Sorry. I was-
Andrew Huberman:
No, no, no. This is important. It's an important digression of ... I have one question, which is you're saying that as across the night, a greater percentage of these 90-minute cycles are going to be occupied by REM sleep as you progress through the night. I'm aware that based on work that you've done and from your public education efforts and others that we have so-called circadian forces, and we have other forces that are driving when we sleep and when we want to sleep, et cetera. Without going into the details of those, I have a simple question. The experiment is the following. Let's say, God forbid, you are prevented from going to sleep at your normal time, and you stay up for the four hours or five hours that normally you would be in predominantly slow wave sleep. If let's say, you finally get to lie down at 3:00 a.m., a time when normally your sleep would be occupied mostly by rapid eye movement sleep. Will you experience a greater percentage of rapid eye movement sleep because of these so-called circadian forces, meaning that's what's appropriate for that time?
Andrew Huberman:
Or will your system need to start at the beginning of the race that we're, as I'm referring to it that we're calling sleep? If that's not clear to anybody, basically what I'm asking is if you are forced to skip the slow wave sleep part of the night, will your system leap into rapid eye movement sleep? Or does it have to start at the beginning and get slow wave sleep first? In other words, does one sleep state drive the entry to the next sleep state?
Matthew Walker:
Great question. So there is some degree of reciprocity between the sleep state. I should note that when we drive one of those up, we often but not always see a change in the other. There are some pharmacologists that have shown an independence to that, and we've also played around with things like temperature. And sometimes you can nudge one and not seem to upset or perturb the other. But to your, I think, lovely point, the answer is it's a mix, but it's mostly the latter. Meaning you will mostly go into your REM sleep phases and be significantly deficient in your deep sleep. So just because I start my sleep cycle at 3:00 a.m. rather than at 10:30 p.m., it doesn't mean that my brain just says, "Well, I've got a program and I'm just going to run the program. And the way the program runs is that we always start with a first couple of hours of deep sleep. So we're just going to begin act number one, scene one." It doesn't do that. Now I will get some deep sleep to begin with. And part of that is just because of how sleep works based on how long I've been awake. Longer I'm awake, there is a significantly greater pressure for deep sleep. But we actually use exactly what you just described as an experimental technique, to selectively deprive people of one of those stages of sleep or the other. So we will do first half of the night deprivation, and then let you sleep the second half. So that means that you will be mostly deep sleep-deprived and you will still get mostly all of your REM sleep. And then we switch it so you only get your first four hours, which means you will mostly get deep non-REM sleep, but you will get almost no REM sleep. So in both of those groups, they've both had four hours of sleep. So the difference between them in terms of an experimental outcome is not the sleep time, because they've both slept for the same amount. It's the contribution of those different stages. Now we actually have more elegant methods for sort of selectively going in there and scooping out different stages of sleep. But that's the way we used to do it old school, was just using this timing difference.
Andrew Huberman:
And who suffers more, those that lack the early phase or those that lack the later phase of the night sleep? In other words, if I have to sleep only four hours, for whatever reason, am I better off getting the early part of the night sleep or the second half of the night's sleep?
Matthew Walker:
Depends on what the outcome measure is.
Andrew Huberman:
So that gets right to the differences between slow wave sleep and REM. I was probably misinformed, but my understanding, a very crude understanding, I should say before-
Matthew Walker:
I very much doubt it, au contraire with someone like you.
Andrew Huberman:
Which is that ... Very nice of you. But the first part of the night, the slow wave sleep is restorative to the musculature, to motor learning, and that the dream content tends to be less emotional. The second half of the night being more emotional dreams and sort of the unpairing of the emotional load of our previous day and other experiences. So in other words, if I were to deprive myself of REM, I would be hyperemotional, maybe not as settled with the kind of experiences of my life. Whereas if I deprived myself of slow wave sleep, I would feel a more physical malaise. Is that correct or is that far too simple? And if it is too simple, please tell me where I'm wrong.
Matthew Walker:
No, I think much of that is correct, and it's sort of that plus. So for example, during deep non-REM sleep, that's where we get this, it's almost a form of natural blood pressure medication. And so when I take that away from you, the next day, we're usually going to see autonomic dysfunction. We're usually going to see abnormalities and heart rate, blood pressure. We also know that during deep non-REM sleep that there is a certain control of specific hormones. For example, we know that the insulin regulation of, sort of, metabolism, meaning how will you look from a regulated blood sugar perspective versus dysregulated prediabetic look of profile. That's where deep sleep seems to matter. We selectively deprive you of that. We can see-
Andrew Huberman:
Growth hormone?
Matthew Walker:
Growth hormone is different, actually. So that's a beautiful demonstration where growth hormones seems to be more REM sleep dependent, and that's why we can come onto the effects of alcohol. And there's some really impressive, frightening data on alcohol and its disruption of sleep. But then we also know testosterone. Peak levels of testosterone happen during REM sleep.
Andrew Huberman:
So the second half of the night.
Matthew Walker:
Which is the second half of the night. So it really just means that your profile of mental and physical dysfunction will be different under both of those conditions. Which one would you prefer? I would prefer neither of them. And it really depends on what you're trying to optimize for. So it's just so ... Sleep is just so pluripotent. It's so physiologically systemic that it's almost impossible not to undergo one of those two things, just deep sleep deprivation or just REM sleep deprivation, and not show a profile that you would really prefer to avoid. And that's the reason, from an evolutionary standpoint, that we've preserved those stages of sleep. I mean, sleep is just so idiotic from an evolutionary perspective.
Andrew Huberman:
Or maybe waking is idiotic.
Matthew Walker:
Or waking is ... Well, yeah.
Andrew Huberman:
Based on your previous idea-
Matthew Walker:
Who have you been talking to? I think that comment is very specific to me. Yeah. I am normally always an idiot when waking. But I think this idea that sleep is so profoundly detrimental to us, if you were to take it at face value, you're not finding a mate, you're not reproducing, you're not foraging for food, you're not caring for your young and worst of all, you're vulnerable to predation on any one of those grounds. Sleep probably should have been selected against, but it wasn't. Sleep has fought its way through heroically, every step along the evolutionary path, and therefore every sleep stage has also survived, as best we can tell. What that means is that those are nonnegotiable. If Mother Nature had found a way to even just sort thin some of that sleep from us, there would've been fast, I'm sure, evolutionary benefits, but it looks as though she hasn't. And I'm usually in favor of her wisdom after 3.6 million years.
Andrew Huberman:
Yeah. It's incredible. I want to introduce another gedankenexperiment, another thought experiment. So in this arc of the night, slow wave sleep predominates early in the night and then REM sleep. There's a scenario that many people, including myself, experience on a regular basis, which is they go to sleep, sleeping just fine — three, four hours into it, they wake up. They wake up, for whatever reason, maybe there was a noise, maybe the temperature isn't right. We will certainly talk about sleep hygiene, et cetera. They get up, they go to the restroom, they might flip on the lights, they might not, they'll get back in bed, hopefully they're not picking up their phone and starting to browse and wake up the brain through various mechanisms, light and cognitive stimulation, et cetera. They go back to sleep. Let's say after about 10, 15 minutes, they're able to fall back asleep, and then they sleep till their more typical wake time.
Andrew Huberman:
How detrimental is that wake-up episode or event in terms of longevity, learning, et cetera? I would love to sleep the entire night through every night, but most nights I don't. And yet I feel pretty good throughout the day, some days better than others. So if you were to kind of evaluate that waking episode and compare it to sleeping the whole night through, what are your thoughts on that?
Matthew Walker:
So I think if you are waking up sort of frequently as you're describing, I would probably get your estate in order because my guess is within the next year you're going to be done for ... No, I'm kidding you. Absolutely kidding you. It is perfectly natural and normal, particularly as we progress with age. Children tend to have typically more continuous sleep. Now, it's not that they aren't waking up for brief periods of time. They are. And in fact, we all do. When we come out the other end of our sleep cycle, at the end of our REM sleep period of the 90-minute cycle, almost everybody wakes up, and we make a postural movement. We turn over because we've been paralyzed for so long. And the body will also like to shift.
Andrew Huberman:
Do we ever look around, do we ever open our eyes and look around?
Matthew Walker:
Sometimes people will open their eyes, but usually it's only for a brief period of time, and they usually never commit those awakenings to memory. Your situation, and it's my situation as well; I usually, now at the stage of life, I don't sleep through the night. I'll usually have a bathroom break, and then I'll come back. That's perfectly normal.
Matthew Walker:
That's perfectly normal. We tend to forget that in sleep science, we think of sleep efficiency, so of the total amount of time that you're in bed, how much of that percent time is spent asleep? And we usually look to numbers that are above 85% or more as a healthy sleep efficiency. So if you were to think about me going to bed, and I spend, let's say eight-and-a-quarter, eight-and-a-half hours of time in bed with a normal, healthy sleep efficiency, I still may be only sleeping a total of seven-and-a-half hours or seven-and-three-quarter hours, meaning that I'm going to be awake in total, not in one long bout, but I'm going to be awake for upwards of 30 minutes net some time.
Matthew Walker:
Sometimes that can be after a 10-minute dalliance after having gone to the bathroom, and I'm just gradually drifting back off again. Other times it will just be for a couple of minutes, and most of those you don't commit. So I think we need to stop, we don't need to get too worried about periods of time awake just because we're not sleeping throughout the night. I would love to do that too, and I remember when that used to happen, and it still happens occasionally.
Andrew Huberman:
Every once in a while. It feels great when it does happen.
Matthew Walker:
And it's a lovely thing-
Andrew Huberman:
It's a surprise, right? Like, oh my goodness, I slept through the whole night.
Matthew Walker:
It is now a surprise. Yeah, it is a surprise. But for the most part, I think we can be more relaxed about that. Where we have to be a bit more attentive though is if you are spending long periods of time not being able to get back to sleep, and usually we define that by saying if it's been 20, 25 minutes, normally that's a time when we would really say, okay, let's explore this. What's going on? Let's see what's happening. The other thing is if it's happening very frequently, so even if you're not awake for 25-minute stretches, but you're finding yourself waking up and being consciously aware that you've woken up for maybe six, seven, or eight times throughout the night, and your sleep is very, what we call fragmented.
Matthew Walker:
The great science of sleep in the past five or 10 years has been, yes, quantity is important, but quality is just as important, and you can't have one without the other in terms of a good, beneficial next-day outcome. You can't just get four hours of sleep, but brilliant quality of sleep, and be unimpaired, nor can you get eight hours of sleep but have very poor quality of sleep and be unimpaired the next day. So that's why I just sort of want to asterisk this idea of let's not get too worried about waking up and having some time awake. That's perfectly normal and natural. But if it's happening very frequently throughout the night, or those periods of time or long stretches of time, upwards of 25 minutes, then let's look into it.
Andrew Huberman:
Well, I can assure you you just helped a lot of people feel better about this waking-up episode that I and many other people experience.
Matthew Walker:
I hope so because I think it's really important that we ... I think I've been desperately guilty of perhaps, early on, being too puritanical about sleep, and I've retrospected and I've tried to explore why this was the case. It was almost sleep or else, dot, dot, dot. And at the time when I was starting to write the book, which was back in 2016, sleep was still a neglected stepsister in the health conversation of today, and I could see all of the-
Andrew Huberman:
That has certainly changed.
Matthew Walker:
And it's changing, and not because of my efforts, but because of all of my colleagues-
Andrew Huberman:
No, I would say ... Well, it's great that you give attribution to other people involved in it, of course it's a big field, but I think you've done a great service by cuing people to the importance of this state, not just for avoiding troublesome outcomes, but also for optimizing their waking state. I view sleep as this period that feels good, but we're not aware of how it feels when we're in it, necessarily. It has tremendous benefits when you're doing it well, so to speak, and it has tremendous deficits when we're not. And I think it was an important thing for you to do, to cue people to this issue, and I would say mission accomplished, that people are aware of the need for sleep. I think that knowing that waking up in the middle of the night is normal, provided it's not too frequent, is great and will also help people who may have been overly concerned about that.
Andrew Huberman:
I do want to use this as an opportunity to raise something about the so-called Uberman schedule. Not to be confused with the Huberman schedule.
Matthew Walker:
Huberman? Yeah.
Andrew Huberman:
Fortunately, no one has confused those yet. Some years ago, there was a discussion about the so-called Uberman schedule, meaning the superman schedule, so that's Huberman without an H, which I have nothing to do with. If you read your Nietzsche, this will have a subtext. But regardless, the Uberman schedule, as I understand, is one in which the person elects to sleep in 90-minute bouts spread throughout the day and night in an attempt to get more productivity and/or reduce their overall sleep need. There was a paper published recently that explored whether or not this is good or bad for us. Maybe you can just give us the take home-message on that?
Matthew Walker:
Yeah, so these Uberman-like schedules, and there's lots of different forms of that, they try to essentially pie chart the 24-hour period into short bouts of sleep, with some shorter or, well, slightly longer periods of wakefulness, then short bouts of sleep, then wakefulness. I sort of made, I think, a quip — it's almost like you're sleeping like a baby, because that's the way that babies will sleep-
Andrew Huberman:
In 90-minute naps, yeah.
Matthew Walker:
... that they all have these brief naps, then they're awake, then they're asleep, then they're awake, and to the chagrin of parents, across the night, it's basically the same. They're awake, they're asleep, they're awake, they're asleep. And that's more the schedule that these types of protocols have suggested. And there was a really great comprehensive review that found not only that they weren't necessarily helpful, but they were actually really quite detrimental, and on almost every performance metric, whether it be task performance, whether it be physiological outcome measures, whether it even be the quality of the sleep that they were having when they were trying to get it. All of those were in a downward direction. And it's not surprising if you look at the way that your physiology is programmed, if you look at the way your circadian rhythm is programmed, none of that screams to us that we should be sleeping in that way.
Andrew Huberman:
Well, I'm chuckling because we always hear sleep like a baby. This is how babies sleep, and I would say, don't sleep like a baby. Sleep like an adult. Be an adult. Get your solid eight hours.
Matthew Walker:
Yeah, it's Billy Crystal's line, who's a long-standing, suffering insomniac. He says, "I sleep like a baby. I'm awake every 20 minutes." And I think this is another one of those demonstrations that when you fight biology, you normally lose, and the way you know you've lost is disease, sickness and impairment. And I think if you sleep in accordance with the natural biological edict that we've all been given, life tends to be both of a higher quality and a longer duration.
Andrew Huberman:
Yeah, I agree. And along those lines, as a vision scientist, I've been very excited by the work on these non-image forming cells in the eye, the so-called melanopsin cells that inform the brain about circadian time of day. And I'm a big proponent of people getting some sunlight, ideally sunlight, but other forms of bright light into their eyes early in the day and when they want to be awake, essentially during the phase of their 24-hour circadian cycle when temperature is rising, and then starting to get less light in their eyes as our temperature is going down later in the day and in the evening. Are there any adjustments to that general theme that you'd like to add in any way?
Matthew Walker:
No, I think that's exactly what we recommend right now, which is try to get at least 30 to 40 minutes of exposure to some kind of natural daylight. Now, there may be parts of the world where it's-
Andrew Huberman:
You're from a rather cloudy part of the world.
Matthew Walker:
I am from Liverpool, England, and the northwest of England is not known for its beach resorts and fine weather. I remember I went back home for a trip when I'd first been out in California, and I thought, why is the sky so low? It's just constantly ... We joke that in the U.K., we usually have nine months of bad weather and then three months of winter, and that's your entire year in terms of a climate. But to come to your point, you're exactly right. Try to get that daylight. Now, it can be working next to a window, and you're getting that natural sunlight, but that natural sunlight is, even on a cloudy day in England, is usually far more potent than anything that you'll get from indoor lighting, despite you thinking, sort of from a perceptionwise, maybe they're much closer than I would think.
Andrew Huberman:
Yeah, I've been, sorry to interrupt, I've been a big proponent of, there's an app called Light Meter, which will, it's a free app, I have nothing to do with it, that will allow you to get a pretty decent measurement of the amount of light energy coming toward you. And if you hold it up to a cloudy morning where you don't think it's very bright out, kind of a dismal day, you'll notice that there'll be 1,000, 2,000, even 5,000 lux, lux just being a measure of brightness, of course. And then you can point the same light meter toward an indoor light that seems very bright and very intense, and it'll say 500 lux, and you realize that the intensity as we gauge it perceptually is not really what the system is receiving. So outdoor light is key. How do you get this natural stimulation, or I should just say light stimulation early in the day? What does Matt Walker do to get this light stimulation?
Matthew Walker:
I am no poster child, but usually I will ... if I'm working, I usually work out most days, and I shopped around and I found a gym that has huge amounts of window exposure facing to the east. This is going to sound so ridiculous. Matt Walker chooses a gym on the basis of the solar input, so he can correct his circadian-
Andrew Huberman:
I love it. There are a lot of criteria for selecting gyms; this one is actually grounded in physiology and biology, and so-
Matthew Walker:
And selfishness about my own sleep.
Andrew Huberman:
No, that's great. So you get your exercise and your light stimulation simultaneously?
Matthew Walker:
That's right. Yep, yep.
Andrew Huberman:
And so you're stacking cues for wakefulness early in the day?
Matthew Walker:
Exactly. So both exercise and daylight are wonderful cues for circadian rhythm alignment, and also circadian rhythm reset each day, and so I will use both exercise ... I mean, I'm neither a strong morning type or a strong evening type, and my preference to exercise is probably sometime in the middle of the day, probably somewhere around 1:00 a. ... 1:00 p.m., sorry, not 1:00 a.m. But I'm usually working out probably around the sort of 7:45, 8:00 a.m. time. That's usually when I'll start my workout. And there I'll start with cardio, with spin bike, facing a window, and luckily, for the most part here in California, there's usually sunlight coming through. But it doesn't matter to me, because just as you said, even when it's a cloudy day, that lux coming through of light, the intensity, is splendid.
Matthew Walker:
So I would prefer to favor my exercise, just because for efficiency too, I want to get also working on the day, I'll try to match my exercise more with my circadian light exposure than I would probably if I'm going to, do I really want to crush a workout or do I want to just make sure it's a good workout. I would prefer to work out at a different time. But I like that, because of the daylight, and we can speak about exercise timing at some point, because there's a lot of discussion around that — when is the right time to exercise during sleep, and we can sort of bust some myths there too.
Matthew Walker:
So I think you're spot-on with the suggestion. Get some morning daylight, try to get that exposure, usually at least 30 to 40 minutes. There was some great work recently coming out in the occupational health domain where they moved workers from offices that were just facing walls and didn't have any exposure to natural daylight, and then they did a time period during that study where they actually were in front of a window and working, and they measured their sleep, and their sleep time and their sleep efficiency increased quite dramatically. I'm forgetting the numbers now, but I think the increase in total sleep time was well over 30 minutes, and the improvement in sleep efficiency was five to 10%. And if you're batting an 80% sleep efficiency average, we're a bit concerned about that. But add 10% to that, and now you're in a great echelon of healthy sleepers, and all you did was just spend some time working in front of windows.
Andrew Huberman:
That's great. And probably folks might want to consider spending a little less time with sunglasses, provided they can do that safely, you know, driving, et cetera.
Matthew Walker:
Yep.
Andrew Huberman:
You're not alone with your exercise behavior and facing east. So the one and only Tim Ferriss told me recently that his morning routine nowadays consists of jumping rope while facing east to get the sunlight stimulation of the eyes. And as Matt and I both know, it has to be of the eyes, right?
Matthew Walker:
Yeah, yeah.
Andrew Huberman:
I mean, these portals are the only way to convey to the rest of the brain and body about the time of day and wakefulness. Along the lines of wakefulness, I have a number of questions about caffeine, the dreaded and beloved caffeine. I love caffeine, but I like it in relatively restricted periods of time. So I'm a big fan of waking up and — even though I wake up very groggy — allowing my natural wakefulness signals to take hold, meaning I wake up very slowly but I don't drink caffeine right away. I'll sort of delay caffeine by a little while, usually 90 minutes to two hours. And that idea came to me on the basis of my understanding of how caffeine and the adenosine receptor interact. I have a feeling you're going to pronounce adenosine differently than I do.
Matthew Walker:
No, no, no. I will go with it. I'll go with adenosine.
Andrew Huberman:
I'll try to go with your skeletal instead of skeletal, and synapse and synapse.
Matthew Walker:
And I'll be schedule and schedule.
Andrew Huberman:
There we go. But to make it really simple for folks, how does caffeine work to make us feel more alert? And does the timing in which we ingest caffeine play an important role in whether or not it works for us or against us? So maybe we just start with, how does caffeine work? Why is it that when I drink maté or coffee, which are my preferred sources of caffeine, do I feel a mental and physical lift?
Matthew Walker:
Yeah. So I'm going to suggest, counter to what most people would think, drink coffee.
Andrew Huberman:
Or maté. Is maté okay also?
Matthew Walker:
Yeah, yeah.
Andrew Huberman:
Or whatever form you enjoy, yeah.
Matthew Walker:
We'll come onto why I suggest that, but when it comes to coffee, I would say the dose and the timing makes the poison. So let's start with how caffeine works. Caffeine is in a class of drugs that we call the psychoactive stimulants. So it works through a variety of mechanisms. One is a dopamine mechanism. Dopamine, we often think of as a reward chemical, but dopamine is also very much an alerting neurochemical as well, and caffeine has some role it seems to play in increasing dopamine. But its principle mode of action, we believe, in terms of making me more alert and keeping me awake throughout the day, is on the effects of adenosine. And to explain what adenosine is, from the moment that you and I woke up this morning, this chemical adenosine has been building up in our brain. And the longer that we're awake, the more of that adenosine accumulates.
Andrew Huberman:
May I ask, is it accumulating in neurons, in glia, or in the blood vessels? And is it also accumulating in my body? Where is this adenosine coming from, and where is it accumulating?
Matthew Walker:
Yeah, so the adenosine here that we're talking about that is creating the sleep pressure is a central brain phenomenon, and it comes from the neurons themselves combusting energy. And as they're combusting energy, one of the offshoots of that is this chemical adenosine. And so as we're awake throughout the day and our brain is metabolically very active, it's accumulating and building up this adenosine. Now, the more adenosine that we have, the sleepier that we will feel, so it really is like, a sleep pressure is what we call it. Now, it's not a mechanical pressure. Don't worry, your head's not going to explode. It's a chemical pressure, and it's this weight of sleepiness that we feel gradually growing as we get into the evening.
Andrew Huberman:
May I just interrupt you again to just ask, do we know what the circuit mechanism is for that? I mean, not to go too far down the rabbit hole, but for the aficionados and for myself, we have brain mechanisms like locus coeruleus that release things — brain areas, locus coeruleus just being a brain area, of course — that release things that proactively create wakefulness. So are those neurons shutting down as a consequence of having too much adenosine, or are there areas of the brain that promote sleepiness that are becoming activated? Because you can imagine both things working in parallel. One or the other would accomplish the same endpoint.
Matthew Walker:
Yeah. And it's both. And so there are two main receptors for adenosine, the A1 receptor and the A2 receptor, and they have different modes of activating brain cells, or inactivating or decreasing their likelihood of firing. And adenosine works in this beautiful, elegant way where it will inhibit and shut down the wake-promoting areas of the brain whilst also increasing and dialing up the volume on sleep-activating, sleep-promoting [inaudible 00:50:16].
Andrew Huberman:
Wow, that's so beautiful. Fantastic.
Matthew Walker:
It's always a push-pull. Yeah.
Andrew Huberman:
And we could have a larger discussion at some point about everything. Seeing dark edges, seeing light edges. Our ability to smell or to sense pressure on the ... Everything's a push-pull in biology.
Matthew Walker:
Oh, it's great. Yeah, yep.
Andrew Huberman:
So this is another example where as I am awake longer, adenosine is released in the brain, and my wakefulness areas are being actively shut down by that adenosine, and my sleepiness brain areas, so to speak, are being promoted to be more active. Is that correct?
Matthew Walker:
That's right, and it's a very progressive process. It's not like a step function where ... And sometimes that happens occasionally, but it's usually because you've been sort of driving through, and as we'll come onto, have caffeine in the system, and then all of a sudden you just hit a wall and it just engulfs you, and you go from a zero to the one of sleepiness within a short period of time. Normally-
Andrew Huberman:
What explains the fatigue after a hard conversation? The desire to go to sleep, or the desire to go to sleep during a hard conversation?
Matthew Walker:
That's an interesting one. I think it's usually just based on personality-type interactions, and for the most part-
Andrew Huberman:
Not that I've ever experienced that before.
Matthew Walker:
No, people with you don't, but with me, they always [inaudible 00:51:28]-
Andrew Huberman:
Oh, no, no. I mean, I've experienced the desire to ... Some conversations, I'm halfway through them, and I feel like I want to take a nap. Right?
Matthew Walker:
Yeah. I would love to look at people's sleep history. We've sort of seen that time and time again. And then it could be, with folks like me, people just lose the will to live within about five minutes of speaking with me, so [inaudible 00:51:48]-
Andrew Huberman:
Not true. They hear that sleep is important, that's all, so they-
Matthew Walker:
... and that's flattery. That's great. So the way that then caffeine comes into this equation, as I was saying, it's usually kind of a linear process, or maybe it's probably closer to an exponential in terms of your subjective feeling of sleepiness. And we haven't really been able to measure that in humans, because normally we ... It's hard to actually stick something into the brain and be sucking, siphoning off stuff every couple of minutes, as you could do in animal studies, and keep asking people every couple of minutes, "How sleepy do you feel? How sleepy?" and track to see if there's a linear rise in adenosine, which then creates an exponential rise in subjective sleepiness, or what the dynamics are. But I'm kind of nerding out.
Matthew Walker:
Caffeine comes into play here because caffeine comes into your system, and it latches onto those welcome sites of adenosine, the adenosine receptors. But what it doesn't do is latch onto them and activate them because if it was doing that, then in lots of ways, it would dial up more sleepiness. It does the opposite. The way that caffeine works is that it comes in, competes with quite sharp elbows with adenosine, competitively forces them out of the way, hijacks that receptor by latching onto it, but then just essentially blocks it. It doesn't inactivate the receptor. It doesn't activate the receptor. It functionally inactivates it, in the sense that it takes it out of the game for adenosine.
Matthew Walker:
So it's like someone coming into a room, and you're just about to sit down on the chair, and caffeine comes in and just pulls out the chair and you're like, well, now I've got nowhere to sit. And caffeine just keeps pulling out the chairs from adenosine, and adenosine, even though it's at the same concentration in your brain, your brain doesn't know that you've been awake for 10 hours, 16 hours at that point when you've downed a cup of coffee, because all of that adenosine that's still there can't communicate to the brain that you've been awake for 16 hours, because-
Andrew Huberman:
But the adenosine is still in brain circulation?
Matthew Walker:
Correct.
Andrew Huberman:
So the real question is, what happens when caffeine is dislodged from the adenosine receptor?
Matthew Walker:
Unfortunate things happen. And that's what we call the caffeine crash, which is caffeine has a half-life and it's metabolized, and-
Andrew Huberman:
Do you recall what the half-life is?
Matthew Walker:
Yeah, the half-life is somewhere between five to six hours, and the quarter-life, therefore, is somewhere between 10 to 12 hours. It's variable. Different people have different durations of its action, but for the average adult, five to six hours. That variation, we understand, it's down to a liver enzyme, or a set of liver enzymes, of the class that we call the cytochrome P450 enzymes. And there are, I think, last I delved into the data, which is pretty recently, there are two gene variants that will dictate the enzymatic speed with which the liver breaks down caffeine. And that's why you can have some people who are very sensitive to caffeine and other people who say, "It doesn't affect me really that much at all."
Andrew Huberman:
These are the people that have a double espresso after a 9:00 p.m. dinner and can sleep just fine?
Matthew Walker:
Yeah. Well, and we'll come-
Andrew Huberman:
Or at least subjectively, they think they're sleeping.
Matthew Walker:
Subjectively, yeah. And we should speak about that assumptive danger too. So then the caffeine is in the system, and after some time period, it will be inactive in the system. So let's say that I've been awake for 12 hours now and it's 8:00 p.m., and I'm feeling a bit tired, but I want to push through, and I want to keep working for another couple of hours. So I have a cup of coffee. All of a sudden, I was feeling tired, but I don't feel like I've been awake for 12 hours anymore, because with the caffeine in the system, maybe only half of that adenosine is being communicated through the receptor to my brain.
Matthew Walker:
A hundred percent of the adenosine is still there. Only half of it is allowed to communicate to my brain. So now I think, oh, I haven't been awake for 12 hours. I've just been awake for six hours, I feel great. Then, after a few hours, and the caffeine is starting to come out of my system, not only am I hit with the same levels of adenosine that I had before I'd had the cup of coffee several hours ago, it's that plus all of the adenosine that's been building up during the time that the caffeine has been in my system.
Andrew Huberman:
So sort of an avalanche of adenosine?
Matthew Walker:
It is a tsunami wave, yeah. And that's the caffeine crash.
Andrew Huberman:
And it's interesting, because the caffeine crash at two o'clock in the afternoon when you have work to do is a terrible thing. But what about the person — maybe this person is me in my twenties — who says, I'm going to drink caffeine all day long, and then I want the crash, because at 9:00 or 10:00 p.m., if I stop drinking caffeine at say 6:00 p.m., and I crash, then I crash into a slumber, a deep night of sleep. Is that sleep really as deep as I think it is? Because given the half-life of caffeine that you mentioned a few moments ago, I have to imagine that having some of that caffeine circulating in my system might disrupt the depth of sleep, or somehow the architecture of sleep, in a way that even if I get eight, or who knows, even 10 hours of sleep, it might not be as restorative as I would like it to be.
Matthew Walker:
Yeah. And that is the danger, just sort of those people that you described who say, and a lot of them will speak with me too, say, "Look, I can have two espressos with dinner, and I fall asleep fine and I stay asleep," because usually those are the two phenotypes that we typically see with too much caffeine. I just can't fall asleep as easily as I want to, or I fall asleep, but I just can't stay asleep. And caffeine can do both of those things quite potently.
Andrew Huberman:
How late in the day do you think is, assuming somebody — translate this, folks, if you go to bed earlier or later; you have to shift the hours accordingly — but given somebody who typically gets into bed around 10:00, 10:30 and falls asleep around 11:00, 11:30, when would you recommend they halt caffeine intake? And these are not strict prescriptors but I think people do benefit from having some fairly clear guidelines of what might work for them. Would you say cut off caffeine by what time of the day?
Matthew Walker:
I would usually say take your typical bedtime and count back sort of somewhere between 10 to ... eight hours is probably getting a little bit close, but take back sort of 10 hours or eight hours of time. That's the time when you should really stop using caffeine, is the suggestion. And the reason is because for those people who even just keep drinking up until into the evening, you're right that they can fall asleep fine, maybe they stay asleep, but the depth of their deep sleep is not as deep anymore.
Matthew Walker:
And so there are two consequences. The first is that for me, and it can be up to ... by 30%, and for me to drop your deep sleep by 30%, I'd have to age you by between 10 to 12 years. Or you can just do it every night to yourself with a couple of espressos. The second is that you then wake up the next morning, and you think, well, I didn't have problems falling asleep, and I didn't have problems staying asleep, but I don't feel particularly restored by my sleep. So now I'm reaching for three or four cups of coffee the next morning rather than just two or three cups of coffee. And so goes this dependency cycle, that you then need your uppers to wake you up in the morning, and then sometimes people will use alcohol in the evening to bring them down because they're overly caffeinated, and alcohol, and we can speak about that too, also has very deleterious impacts on your sleep as well.
Matthew Walker:
So you're right that it's not just the quantity of your sleep or even difficulties falling or staying asleep, it can also be deep sleep. But here again, I think, I don't want to be frightening people, and I mentioned this before, I think one of the real problems that I, or mistakes that I made, because I didn't ... I'd never had much public exposure before the book, and I was so saddened by the disease and the suffering that I was seeing as a consequence of a lack of sleep in our society and the fact that it wasn't really being discussed very much; I sort of came out a little bit headstrong, more than a little bit headstrong. And I think I was perhaps too much gas pedal and too little brake, as it were. And I don't think that's the right way to approach a health message within the public sphere, and I've become much softer in how I think about these things. I have ideas about what the ideal world looks like for sleep, but I also realize that none of us live in this thing called the ideal world.
Andrew Huberman:
We certainly don't.
Matthew Walker:
So I want to be really mindful of that. And I think I've done a really bad job of being sort of too forthright, particularly for people who struggle with sleep, early on, when I would offer these sort of messages about sleep. I want to be veritical when it comes to the science. I want to be faithful to the science, but I also don't want to go out and scare the living daylights out of people, particularly people who are struggling with their sleep, because it's probably only going to make matters worse.
Matthew Walker:
So I've been beautifully schooled by learning how to be a slightly better public communicator. I'm nowhere near of the standing that you are. You're very elegant and it's very intuitive to you. I'm still with training wheels, but I'm getting a little bit better. But I just want to say that when I'm speaking about caffeine, because it sounds as though I'm very sort of overt about it. But I will come back to why I say drink coffee, but I just wanted to make that point.
Andrew Huberman:
Yeah. Well, I appreciate you making that point, and I'm sure our listeners will too. I still will stand behind my statement, which is that what-
Andrew Huberman:
Stand behind my statement, which is that what you've done for the notion that sleep is vital for all aspects of health and for performance, mental and physical and wakefulness — the message in that and the packaging it was contained in ... has been clearly net positive. People needed to be cued to this. The "I'll sleep when I'm dead" mentality is one that I had. It's one that other people have, people in a huge number of vital communities, not just your students, but also people ... The messaging that you provided and continue to provide has positively impacted the first responder community, the medical community — there's still steps that need to be taken — the military community and of course the civilian community. And so I think these adjustments about, yeah, caffeine's okay. Just restrict it to the early part of the day if you can most days. I mean, I think the law of averages; it's like the light-viewing behavior.
Andrew Huberman:
I think it is critical to view sunlight or natural, some other form of bright light early in the day. But if you miss a day, it's not that your whole system is going to dissolve into a puddle of tears. That'll happen on the second or the third day. No, I'm kidding. You get a couple days. Biology works in averages, except with respect to accident or injury. A car accident is a car accident. You don't get to have three of those before the brain damage occurs if the accident's severe enough. But with sleep behavior, these homeostatic-type behaviors, or with food, one chocolate sunday, is it going to kill you? No. Right. Every night, yeah. It's going to make you demented and kill you early. We know this. And so I think the middle ground is often a hard place to achieve. So I think you've done a phenomenal job, but I appreciate you raising these points.
Andrew Huberman:
And I think it's clear that we all need to .. that we all can and should do certain things better, including being gentle with ourselves from time to time when we deviate from these ideal circumstances. Along these lines, I do want to talk about alcohol because I think caffeine and alcohol represent the kind of two opposite ends of the spectrum. Clearly there are other stimulants; they're your adderalls and your high-energy drinks that people use. But alcohol and caffeine are the most commonly consumed stimulants and sedatives — depressants as they're sometimes called. So what happens when somebody has a glass — we always hear a glass or two of wine in the evening, or a cocktail after dinner or before dinner. How does that impact their sleep? And then we'll be sure to circle back in terms of what [are] reasonable ranges of behavior when it comes to avoiding alcohol, or if it's age-appropriate, et cetera, enjoying alcohol.
Matthew Walker:
So alcohol, if we're thinking about classes of drugs, they're in a class of drugs that we call the sedatives. And I think one of the first problems that people often mistake: alcohol is often used as a sleep aid for people who are struggling with sleep when things like over the counter remedies, et cetera, or herbal remedies have just not worked out for them. And alcohol, unfortunately, is anything but a sleep aid. The first reason that most people use it is to try and help them fall asleep.
Andrew Huberman:
And this process of this event that we call falling asleep, I have to imagine is a process.
Matthew Walker:
It is a process.
Andrew Huberman:
Like everything in biology.
Matthew Walker:
Yeah.
Andrew Huberman:
And that that process involves, in some way, as we talked about, push-pull before turning off, thinking, planning, etcetera, and turning on some sort of relaxation mechanism. I have to imagine that these two things are knobs, turning in opposite directions.
Matthew Walker:
That's right.
Andrew Huberman:
That gives us this outcome we call falling asleep. Alcohol, it seems, is helpful for some people to turn off their thoughts or their planning, right? Is that right?
Matthew Walker:
Yes, it is. And so I think if we look at the pattern of brain activity, if I were to place you inside an MRI scanner, where we're looking at the activity of your brain, and watch you drifting off, some parts of your brain will become less active, other parts will become more active. And this is the push-pull model. It's inhibition, excitation. But alcohol is quite different in that regard. Alcohol is, because it's a sedative, what it's really doing is trying to essentially knock out your cortex. It's sedating your cortex. And sedation is not sleep. But when we have a couple of drinks in the evening, when we have a couple of night caps, we mistake sedation for sleep, saying, well, always when I have a couple of whiskeys or a couple of cocktails, it always helps me fall asleep faster. In truth, what's happening is that you're losing consciousness quicker, but you're not necessarily falling naturalistically asleep any quicker.
Matthew Walker:
So that's one of the first, sort of, things just to keep in mind. The second thing with alcohol is that it fragments your sleep. And we spoke about the quality of your sleep being just as important as the quantity, and alcohol — through a variety of mechanisms, some of which are activation of that autonomic nervous system, that fight-or-flight branch of the nervous system — alcohol will actually have you waking up many more times throughout the night. So your sleep is far less continuous. Now, some of those awakenings will be of conscious recollection the next day. You'll just remember waking up. Many of them won't be, but yet your sleep will be littered with these sort of punctured awakenings throughout the night. And again, when you wake up the next morning, you don't feel restored by your sleep. Fragmented sleep, or noncontinuous sleep in this alcohol-induced way is usually not good-quality sleep that you feel great on the next day. The third part of alcohol in terms of an equation is that it's quite potent at blocking your REM sleep, your rapid eye movement sleep, and REM sleep is critical for a variety of cognitive functions. Some aspects of learning and memory seems to be critical for aspects of emotional and mental health.
Andrew Huberman:
You've described it before as a sort of self-generated therapy that occurs while we sleep.
Matthew Walker:
Yeah, it's overnight therapy. It's emotional first aid. Is certainly-
Andrew Huberman:
People that don't get enough sleep are very easy to derail emotionally. Not that one would want to do that to people, but we all sort of fall apart emotionally. I always think about almost like our skin sensitivity can be heightened.
Matthew Walker:
Yes.
Andrew Huberman:
When we are sleep-deprived, our emotional sensitivity is such that when we're sleep-deprived, such that it takes a much finer grain of sandpaper to create that kind of friction. Things bother us.
Matthew Walker:
Threshold to trigger.
Andrew Huberman:
Even online comments bother us when we're sleep-deprived and never when we're-
Matthew Walker:
I would love to say that I never look at them, except I look at-
Andrew Huberman:
Well, I should-
Matthew Walker:
Maybe everyone-
Andrew Huberman:
Here I will editorialize because the notion of not looking at comments is unreasonable to ask of any academic. Because academics, we are all trained to look at our teaching evaluations and, just like with online comments, to ignore 20% of them. No, I'm kidding. We look at them all in any event. So in terms of translating this to behavior, I don't particularly enjoy alcohol. I guess I might be fortunate in that sense, but I also have never really experienced the pleasure of drinking alcohol. I sometimes like the taste of a drink, but I never like the sensation. So I don't have a lot of familiarity with this, but many people do. And I understand that. So let's say somebody enjoys a glass of wine or two with dinner, and they eat dinner at 7:00 p.m. Is that likely to disrupt their sleep at all? Let's just, let's make this a series of gradations.
Matthew Walker:
And the answer is yes. I think once [inaudible] just looked at a single glass of wine in the evening with dinner, and I would be untruthful if I didn't just simply say it has an effect. And we can measure that in terms of-
Andrew Huberman:
Less REM sleep.
Matthew Walker:
Less REM sleep. And one of the fascinating studies, I can't remember what dose, I think they got them close to a standard illegal blood-alcohol level. So maybe they were a little bit tipsy. And yes, you see all of the changes that we just described; they sort of lose consciousness more quickly. They have fragmented sleep, and they have a significant reduction in REM sleep. But what was also interesting, because REM sleep, as we spoke about before, is a time when some hormonal systems are essentially recharged and refreshed. Growth hormone being one of them. There was well over a 50%, five zero, drop in their growth hormone release during alcohol-laced sleep at night.
Andrew Huberman:
And growth hormone is so vital for metabolism and repair of tissues. And it's not just keeping body fat low.
Matthew Walker:
This is essentially adults.
Andrew Huberman:
Essential. Along those lines, I just want to highlight the fact that this information that you're sharing, that growth hormone is released is strongly tethered to the presence of healthy amounts of REM sleep, is interesting to me because I always thought that growth hormone was released in the early part of the night.
Matthew Walker:
Well, it is released across both of those, but across the different stages. But what we also know is that when you disrupt REM sleep, there are those growth hormone consequences. So it's not an exclusive system, just like with testosterone. We can see changes throughout non-REM sleep, but if you ask when are the peak release rates of testosterone, it's right before we go into REM sleep and then during REM sleep.
Andrew Huberman:
And of course testosterone being important both for males and-
Matthew Walker:
For men and women. Yup.
Andrew Huberman:
... for females, for libido and tissue repair and wellbeing. Nobody, regardless of chromosomal, hormonal or any other background wants to have their normal levels of testosterone reduced acutely.
Matthew Walker:
No.
Andrew Huberman:
That's just a bad ... it equates to a terrible set of psychological and physical symptom.
Matthew Walker:
And the mortality risk that's associated with low testosterone is nontrivial.
Andrew Huberman:
Prostate cancer.
Matthew Walker:
Right, Exactly. So coming back to just the point on REM sleep that you mentioned regarding emotional instability, and we see that that's one of the things, one of the most reliable signatures of just insufficient sleep, doesn't have to be sleep deprivation. What we've discovered over the past 20 years here at the sleep center is that there is no major psychiatric disorder that we can find in which sleep is normal. And so I think that firstly told us there is a very intimate association between your emotional mental health and your sleep health. But when it also comes to REM sleep, I think what's fascinating is that it's not just about your emotional health.
Matthew Walker:
It's not just about your hormonal health. We've also been seeing other aspects of cognition. But then there was a report; I think it could have been about two years ago, out of Harvard; I think it was Beth Clayman's group. They found that, and they replicated it in two different large populations, if you look at the contribution of different sleep stages to your lifespan, REM sleep was the strongest predictor of your longevity. And it was a linear relationship, but wasn't sort of one of these U-shape or J-shaped curves that we often see with total sleep and mortality risk. It really was linear, that the less and less REM sleep that you were getting, the higher and higher your probability of death, and then they did ...
Andrew Huberman:
Was that death due to natural causes, or accidents? 'Cause I can imagine if you're not getting enough REM sleep, you're more likely to drive off the freeway, step off a cliff.
Matthew Walker:
Yeah, I think it was all cause-
Andrew Huberman:
Just make bad decisions about anything, relationships ... which can also be life threatening.
Matthew Walker:
Yeah, I've tried to lean into that and claim that I, with those bad relationships situations, that, oh, I just didn't have enough REM sleep last night, my darling.
Andrew Huberman:
The REM sleep defense.
Matthew Walker:
Yeah, but she's far wiser than I thought. But so they did this great machine learning analysis, and I may get these numbers backwards, but I think for every 5% reduction in REM sleep, there was a 13% associated increased risk of mortality. And I'll have to go back and check, but to me, and in the machine learning algorithm, what they ultimately spat out was that of all of the sleep stages, REM sleep is the most predictive of your longevity, of your lifespan. So we often, I hear people saying, How can I get more deep sleep? Or they sometimes say, how can I get more dream sleep? And my answer is a question: why do you want to get more of that? And they'll say, well, isn't that the good stuff? And I'll say, well, actually all stages are the good stuff.
Andrew Huberman:
It's all the good.
Matthew Walker:
Yeah.
Andrew Huberman:
Well it's like the exercise question. And it took decades for people to understand that moving around for about 150, probably 180 minutes a week at doing endurance-type work, zone two cardio-type work, it is correlated with living longer, feeling better, less diabetes, et cetera. There's really no way around it. Yeah. I mean, you can ingest metformin until the cows come home. You can take NMN, all of which I think have their place in certain contexts. I'm a big fan of the work surrounding all those protocols.
Matthew Walker:
Yeah, likewise.
Andrew Huberman:
But without getting proper amounts of movement, meaning sufficient numbers, it doesn't matter how many 12-minute exercise regimes you follow per week, you need that threshold level. And it sounds like the same is true of REM sleep and total amount of sleep. You pay the piper somehow.
Matthew Walker:
Yeah. The return on investment, I mean, to flip the coin, the return on investment is astronomical. I think of sleep. It is the tide that moves, that raises all of those health boats.
Andrew Huberman:
And the most fundamental layer of mental and physical health. Whenever people ask me, even though I'm not a physician, they'll ask me, What should I take or what should I do? The first question is always, How's your sleep?
Matthew Walker:
Great.
Andrew Huberman:
Meaning, how well do you sleep every night, and how long you sleep? I always recommend your book. I always recommend your podcast, the podcast you've been a guest on, et cetera. Who knows, maybe you'll even release your own podcast at some point soon and keep ... because I do think people need to hear from you more often. One thing, I don't want to return to the notion of public health discourse too much, but I do want to say one issue with books in general is that they can be revised, but it's more or less a one-and-done kind of thing until the next book comes out.
Matthew Walker:
Yeah.
Andrew Huberman:
One thing I like about the podcast format is that updates can be provided regularly, corrections and updates as new data come out. And so that's a wonderful aspect to this format, and hopefully the format that you'll be embracing; the world needs to hear more from you more often about sleep in its various contours, not less. And so I do have a question about drinking alcohol. Not that we want to promote day drinking, but let's say that the one or two glasses of wine or cocktail is consumed with lunch — something that isn't traditionally done nowadays — or in a late afternoon happy hour cocktail, and then one is going to sleep seven or eight hours later. Do you think that that will improve or somehow mitigate the effects of alcohol? Or if you have a drink, are you basically screwed for the next 24 hours?
Matthew Walker:
No, I think there's going to be a time window dependency. Now, I don't know of anyone who's essentially done what you and I would like, which is the time separation, dose-dependent curve, where, okay, you drink at 10:00 a.m. or at 11, 12, one, two, three, four, five, all the way up to 10:00 p.m. and estimate what is the blast radius, and is it linear or is it nonlinear. Is it such that only when you drink in the last four hours do you just hit this exponential, and it's bad, bad? Or is there some other curve that we could imagine? There would be many possibilities? But certainly, what we know is that the less alcohol and the less ... and more specifically the metabolic byproducts, aldehydes and ketones, they're the nefarious players here ...
Andrew Huberman:
And not the ketones that people are all excited about. The other ketones; the chemists know what we're referring to. But-
Matthew Walker:
Yeah, this is not about ketogenic.
Andrew Huberman:
This is not about ketogenic; they are ketone bodies that are released after ingesting alcohol that are not of the positive sort.
Matthew Walker:
Correct.
Andrew Huberman:
That a ketogenic diet might promote.
Matthew Walker:
Right.
Andrew Huberman:
Yeah.
Matthew Walker:
So I think in terms of that alcohol profile, we certainly know that as you're heading into the evening hours, once again, timing and dose make the poison. But I think it's also important once again from that public message standpoint. And thank you. I think I am leaning into the podcast consideration arena at some point, but I don't want to be puritanical here. I'm just a scientist, and I'm not here to tell anyone how to live. All I'm trying to do is empower people with some of the scientific literature regarding sleep. And then you can make whatever informed choices that you want. Now unlike you, it turns out I'm not a big drinker, and it's just because I've never liked the taste. And I'm surprised that they haven't taken away my British passport because I don't like lager or beer. But I also want to say that life is to be lived to a certain degree.
Matthew Walker:
It's all about checks and balances. So if I go out and I have an ice cream sundae — I'm not big on those either, but sure, I know that my blood glucose is not going to be ideal for another 12 hours maybe. That's just the price you pay for having some kind of relaxed, fun life. Sure. I don't want to look back on life and think, Gosh, I lived until I was 111, and it was utterly miserable. But it's all about some kind of a balance. And mine, my job is not to tell people a prescription for life, it's just to offer some scientific information.
Andrew Huberman:
All right. Think you're doing a terrific job of that. People are, I always say we have all these neural circuits, and if it's working properly, we all have a circuit that allows us to skip over information or as we wish. If the circuits between your brain and your thumbs are working, you can slide right along. You can drop to the next content however you like. I would like to ask about marijuana and CBD. This is a discussion that I think five years ago would've ventured into the realm of illegal, But now in many places, not all, medical marijuana is approved or is legal.
Andrew Huberman:
And certainly it's in widespread use. Certainly not recommending people do it. I have my own thoughts about marijuana, CBD. I've been fortunate, I suppose, that I don't particularly like marijuana or CBD. I don't even know if I've ever tried CBD. First of all, does marijuana disrupt the depth of sleep, the architecture of sleep? And if so, as with alcohol and caffeine, does when you ingest it or when it's in your bloodstream, relative to when you go to sleep, does that play an important role? So does marijuana disrupt sleep?
Matthew Walker:
Yeah, it does. And there's a pretty good amount of data on ... So we can break cannabis down into two of its key ingredients. We've got THC, tetrahydrocannabinol, and we've got CBD, and CBD is sort of the less psycho ... It's what we think of as the nonpsychoactive component. In other words, when you take CBD, you don't get high. If you take THC, you can get high. That's the psychoactive part of the equation.
Andrew Huberman:
Are both considered sedatives in the technical sense?
Matthew Walker:
No, they're not. Neither of them have that class right now. THC seems to speed up the time with which you fall asleep. But again, if you look at the electrical brainwave signature of your falling asleep with and without that THC, it's not going to be an ideal fit. So you could argue it's non-natural, but many people use THC for that fact because they find it difficult to fall asleep, and it can speed the onset of at least nonconsciousness, I guess is the best way of describing it. But there are problems with THC and there are twofold. The first is that it too, but through different mechanisms, seems to block REM sleep. And that's why a lot of people when they're using will tell me, Look, I definitely, I was dreaming. I don't remember many of my dreams. And then when they stop using THC, they'll say, I was having just crazy, crazy dreams.
Matthew Walker:
And the reason is because there is a rebound mechanism. REM sleep is very clever, and alcohol is the same way in this sense; it's the same homeostatic mechanism. Some people will tell me, Look, if I have a bit of a wild Friday night with some alcohol, maybe I'll sleep late into the next morning, and I'll just have these really intense dreams, and I thought I wasn't having any REM sleep. Well, the way it works is that it's during, in the middle of the night really, when alcohol blocks your REM sleep, and your brain is smart — it understands how much REM sleep you should have had, how much REM sleep you have not because the alcohol's been in the system. And finally, in those early morning hours when you're getting through to six, seven, eight a.m., all of a sudden your brain not only goes back to having the same amount of REM it would've had, it does that plus it tries to get back all of the REM sleep that it's lost.
Matthew Walker:
Does it get back all of the REM sleep? No, it doesn't. It never gets back all of the REM sleep, but it tries. And so you have these really intense periods of REM sleep, hence you have really intense, bizarre dreams. And that's what happens also with THC. You build up this pressure for REM sleep, this debt for REM sleep; will you ever pay it back? Doesn't seem as though you get back everything that you lost. But will you get back some of it, yes. The brain will start to devour more because it's been starved of REM sleep for so long. But one of the bigger problems with THC that we worry about is withdrawal dependency. So as you start to use THC for sleep, there can be a dependency tolerance. So you start to need more to get the same sleep benefit. And when you stop using, you usually get a very severe rebound insomnia. And in fact, it's so potent that it's typically part of the clinical withdrawal profile from THC, from cannabis.
Andrew Huberman:
And there's anxiety withdrawal. I don't ask anybody to change their behavior. Just as you said, we try and inform people about what the science says and let them make choices for themselves. People who are regular pot smokers, many will insist they're not addicted, and maybe, indeed, they don't actually follow the profile of classical addiction. I don't know. I'm guessing some do, some don't. But if you ask them, Well, what if I took away all marijuana consumption for, I don't know, two weeks? That thought scares many of them. And many of them will experience intense anxiety without marijuana, which speaks to perhaps not addiction, but a certain kind of dependency. And again, I know many pot smokers, some of whom have jobs that are quite high performing and they manage.
Matthew Walker:
Here in Berkeley, I don't know any of those.
Andrew Huberman:
Yeah, no, none of those. Right. What about CBD? I mean, we hear so much about CBD. I've, I've been a little concerned about the fact that the analysis of a lot of CBD supplements out there is confirmed, that much like with melatonin, the levels that are reported on the labels in no way, shape or form match the levels that are actually contained in the various supplements. Sometimes the levels are much higher than they're reported on the labels. Other times it's much lower. What does ingesting CBD do to the architecture and quality of sleep?
Matthew Walker:
Right now, I don't think we have enough data to make some kind of meaningful sense out of it. I think the picture that is emerging, however, is probably the following. Firstly, CBD does not seem to be detrimental in the same ways that THC is. So we can start by saying, does it create potential problems? Not of the nature necessarily that we see with THC, but the devil is a little bit in the details, from the data that we do have. And it comes onto your valid point of purity. At low dose, CBD can seem to be wake-promoting. So in lower doses, let's say sort of five or 10 milligrams — I'm trying to remember some of the studies off the top of my head there — it actually may enhance wakefulness and cause problems with sleep. It's only once you get into the higher dose range that there seem to have been some increases in sleepiness or sort of sedation-like increases.
Matthew Walker:
And that's usually, I think, above about 25 milligrams, as best I can recall from the data. And then when we look in animal models, you typically see the same type of profile too. So then the question becomes ... and now again, you just don't know about purity. It's very difficult. Although I think, and again, I'm not a user, not necessarily because I have anything against it. It's just that's not necessarily my cup of tea. There are some firms that are now doing third-party independent laboratory tests. I don't know how gamed that is. So I've got no sense of.
Andrew Huberman:
No, I think some supplement companies are quite honest and accurate about the amounts of various substances that are in other products, and some are not.
Matthew Walker:
Yeah. Yeah.
Andrew Huberman:
And I think there's just a huge range. I think the FDA is starting to explore CBD. There are ... certainly, I saw some grant announcements to explore the function of CBD. Most of the work on CBD is being done by the general public, ingesting it, and seeing how they feel. I gave it to my dog, who was, had some dementia-related sleep disturbances, and it actually created a heightened wakefulness. It completely screwed up his sleep.
Matthew Walker:
It sounds as though he just wasn't-
Andrew Huberman:
He's a bulldog, so he's going to get access to sleep; he's going to take it. Really messed him up. Took him, took it away. He did better. But that's a canine, so.
Matthew Walker:
Right and it could have been sort of dose-related too-
Andrew Huberman:
Or binders or other things that are in there.
Matthew Walker:
Correct. Yeah. But right now if we were to, and I'm not making this statement — I don't think anyone can make the statement now — but if it ends up being that CBD is potentially beneficial for sleep, how can we reconcile that mechanistically? And I think there are, to me at least, there are at least three candidate mechanisms that I've been exploring and thinking about. The first is that it's thermoregulatory. And what we found in some animal models is that CBD will create a profile of hypothermia. In other words, it calls the body, the core body temperature down. And that's something that we know is good for sleep. The second is that it's an anxiolytic, that it can reduce anxiety. And that data is actually quite strong, even with some functional imaging work that's been coming out recently, showing that one epicenter of emotion called the amygdala, deep within the brain, is quietened down with CBD. So I think that's at least a second nonmutually exclusive ...
Andrew Huberman:
That's great. That's interesting.
Matthew Walker:
... you know, possibility. I think the third is some recent data that's come out that was suggesting that CBD can alter the signaling of adenosine. So it doesn't necessarily mean that you produce more adenosine, but what it can do is perhaps modulate the sensitivity, perhaps, of the brain so that the weight of that same adenosine is weightier in its brain signal and therefore it creates this stronger pressure for sleep. So I think these are all tentative mechanisms. I think any one of them is viable. I think all three are viable together. But right now, I think, does that sort of help think through the tapestry of THC and CBD?
Andrew Huberman:
Yeah, very much. And actually it's a perfect segue from, we've talked about caffeine, alcohol, THC and CBD as sort of, we framed them, anyway, as things that done in moderation at the appropriate times are probably okay for most people. Certainly not for everybody. There will be differences in sensitivity, but that done at the incorrect times and certainly in the incorrect amount, will greatly disrupt this vital stage of life we call sleep. CBD, it seems, represents a kind of bridge to the topic I'd like to talk about next, which is things that promote more healthy sleep or somehow contribute to enhancing the architecture and quality of sleep. So I'd love to chat for a moment about the kind of grand ... the original, I should say that, not the granddaddy, but the OG of sleep supplementation, which is melatonin.
Andrew Huberman:
The so-called hormone of darkness that's inhibited by light, et cetera. Frame for us melatonin in the context of its naturally occurring form. And then I'd like to talk about melatonin, the supplement, because, as in my experience, anytime I say the word melatonin, people think about the supplement melatonin, which in itself is an interesting phenomenon, that people are so queued to its role as something you take. We often forget that this is something that we make endogenously. I'd love for you to comment in particular on ... even though we ... Without necessarily getting into its precise nanograms per deciliter values, what are the typical amounts of melatonin that we release each night? And then I'd like to compare that to what is contained in say a three.
Andrew Huberman:
And then I'd like to compare that to what is contained in say a three milligram or six milligram tablet that one might buy at the pharmacy.
Matthew Walker:
Right. Yeah.
Andrew Huberman:
So I go to sleep at night, has melatonin already kicked in before I shut my eyes and lay down my head?
Matthew Walker:
Usually yes. If your system is working in the correct way, as dusk is starting to happen. So let's say that you look at hunter gatherer tribes who aren't touched by electricity, and so they're sort of the puritanical state par excellence when it comes to electric light influence. And usually it's as dusk is approaching, that's when melatonin will start to rise. And so when you lose the brake pedal of light coming through the eyes — that normally acts like a hard brake pedal that stamps down and prevents the release and production of melatonin.
Matthew Walker:
As that light brake pedal starts to fade with dusk, then we ease off the brake pedal, and the spigot of melatonin is opened up, and melatonin starts getting released. And usually we'll see this rising peak of melatonin sometime, usually an hour, two hours later, or around. And it varies from different people around the time of sleep itself, but it's already been on the march-
Andrew Huberman:
Interesting.
Matthew Walker:
For some hours before you actually hit sleep itself.
Andrew Huberman:
Interesting. And I was always taught, and I'm assuming it's still true, that the only source of melatonin in the brain and body is the pineal gland. Is that still true?
Matthew Walker:
Yeah, it seems to be, from best that we can tell. The pineal gland, sort of meaning P-like sort of shape. Usually people say it's P-like. I think if you look at the Latin derivative, I think it's derived from pine cone, not Pea. Because, in fact, if you look at the pineal, it is more pine cone-shaped, and so it's aptly named.
Andrew Huberman:
Any human brain I've ever dissected ,or I confess I've dissected a lot because I teach neuroanatomy and have for years, I love looking at the pineal. It's the one structure in the brain that's not on both sides. It's usually pretty easy to find, and it's pretty good size. It looks like a pea.
Matthew Walker:
Yeah.
Andrew Huberman:
And it's sitting right there, and it's remarkable that it releases this hormone probably our entire lifespan and is inhibited by light. So our pineal starts to release this into the general circulation. I have to imagine we have melatonin receptors in the brain and body.
Matthew Walker:
Correct. So yeah. Essentially your brain has a central master 24-hour clock called the suprachiasmatic nucleus that keeps internal time. Now it's not a precise clock if left to its own devices, nothing that a Swiss clock maker would be proud of. It runs a little bit long and laggy.
Andrew Huberman:
It's like an American clock. There are a couple good American watches, by the way. Hamiltons are very nice.
Matthew Walker:
It's very much like a-
Andrew Huberman:
We're not famous for our timekeeping, or our punctuality, for that matter. But the Swiss are.
Matthew Walker:
It's not quite Swisslike it's more Berkeleylike, which is very relaxed, whatever. So in most adults, in the average adult, I should say, your biological clock normally runs a little bit long. It's about 24 hours and 13 minutes, I think was the last calculation. But the reason that we don't keep drifting forward in time and kind of running consistently later and later, 30 minutes by 30 minutes by 30 minutes each day, is because your central brain clock is regulated by external things, such as daylight and temperature, as well as food and activity. All of these are essentially different fingers that come along, and on the wristwatch of the 24-hour clock, we'll pull the dial out and reset it each day to precisely 24 hours.
Matthew Walker:
And I make that point because it knows 24-hour time, but it needs to tell the rest of the brain and the body the 24-hour time as well. And one of the ways that it does this is by communicating a chemical signal of 24-hour nurse of light and day using this hormone melatonin. And when it is at low levels or it's non-existent, it's communicating the message, it's daytime, and for us, diurnal species, it says it's time to be awake.
Matthew Walker:
Yet at nighttime when dusk approaches and the brake comes off melatonin, and we start to release it, then it signals to the rest of the brain and the body, look it's dusk and it's nighttime. And for us, diurnal species, it's time to think about sleep. So melatonin essentially tells the brain and the body when it's day and when it's night, and with that, when it's time to sleep, when it's time to wake, and therefore that's why melatonin helps with the timing of the onset of sleep. But it doesn't really help with the generation of sleep itself. And this is why we'll come onto what those studies of supplementation have taught us.
Andrew Huberman:
So it tells the rest of my brain and body it's time to go to sleep. It perhaps even aids with the transition to sleep, but it's not going to, for instance, ensure the overall structure of sleep, or it's not the conductor that's guiding the sleep orchestra, so to speak, throughout the entire night.
Matthew Walker:
Yeah.
Andrew Huberman:
It's more like the people that essentially take you to your seat and sit you down and give you your program.
Matthew Walker:
Right, exactly. Yeah. Sort of the far less sophisticated analogy I have is melatonin is the starting official at the hundred-meter race in the Olympics.
Andrew Huberman:
That's a better analogy.
Matthew Walker:
It calls all of the sleep racers to the line, and it begins the great sleep race.
Andrew Huberman:
Better analogy by the way. Coming from a sleep researcher of all people.
Matthew Walker:
But it doesn't participate in the race itself. That's a whole different set of brain chemicals and brain regions. Which then brings us onto perhaps the question of supplementation, which is, is it helpful for my sleep? Will I sleep longer? Will I sleep better? And if I am, what doses should I be taking?
Matthew Walker:
Sadly, the evidence in healthy adults who are not older age suggests that melatonin is not really particularly helpful as a sleep aid. I think there was a recent meta-analysis that demonstrated — it looked at all of the different sleep parameters — melatonin, and a meta-analysis, for those not knowing what that is, it's a scientific sort of method that we use where we gather all the individual studies, and we put them in a big bucket, and we kind of do this kind of statistical fancy sleight of hand, and we try to come up with a big picture of what all of those individual studies tell us. And what that meta-analysis told us is that melatonin will only increase total amount of sleep by 3.9 minutes on average.
Andrew Huberman:
Minutes?
Matthew Walker:
Minutes.
Andrew Huberman:
Not even percent.
Matthew Walker:
No. And it will only increase your sleep efficiency by 2.2%. So it really-
Andrew Huberman:
This is, as they say in certain parts of California, that's weak sauce. That's a weak sauce-effect.
Matthew Walker:
The sauce is not strong; the force is not strong in this one; when it comes to a tool that in healthy people who are not of older age, it doesn't seem to be especially beneficial. Now results can vary. Everyone is different, of course. So we're talking about the so-called average human adult here.
Andrew Huberman:
Well melatonin, in defense of what you're saying ... And also I should mention I have a colleague at Stanford, Jamie Zeitzer, who you know-
Matthew Walker:
Oh, wonderful.
Andrew Huberman:
Chuck Czeisler's lab, at Harvard Med, where he also trained — terrific sleep researcher — and I asked him about melatonin, and he essentially said the same thing that you just said, which is very little if any evidence that it can improve sleep. And yet it's probably the most commonly consumed so-called sleep aid.
Matthew Walker:
Hundreds of million dollars industry.
Andrew Huberman:
Yeah. So it, either massive placebo effect, or it's operating through some other mechanism related to quelling anxiety, perhaps?
Matthew Walker:
Yeah, that's is actually interesting. There are some studies where you do see some effects. Now again, when you do the grand average of all studies, it just doesn't seem to have an effect. But let's assume that for some people it does have an effect. Let's not again be sort of completely dismissive of that. How could it have that effect? One of the reasons that I've become a little bit more bullish on melatonin from a sleep perspective, and then melatonin, more generally, for a — maybe we can speak about this too — as a countermeasure when you're undergoing insufficient sleep; there are two different routes there. The first reason that I think it could have a sleep benefit for some people is not because it helps in the generation of sleep. We know that it doesn't. It's because it too seems to drop core body temperature.
Andrew Huberman:
There it is, temperature again. I'm fascinated these days more and more by temperature, as maybe not just a reflection of brain state and wakefulness and in sleep, but actually a lever that is quite powerful.
Matthew Walker:
I think it's both.
Andrew Huberman:
And with all the interest in ice baths and hot showers and saunas and stuff, something that we will definitely touch on, temperature variation is so key. So if melatonin is dropping body temperature by a degree or so, something that you've said before can help induce a sleepy state, maybe that's what's allowing people to get in-
Matthew Walker:
That's one possibility. I don't think melatonin by itself will drop it by, sort of, a degree, certainly not a degree Celsius. And for ordering us to fall asleep and then stay asleep across the night, we do need to drop our core body temperature by about one degree Celsius or about two to three degrees Fahrenheit. And that's why it's always easier to fall asleep in a room that's too cold than too hot. I think that that's one potential avenue that we are considering thinking more deeply about when it comes to melatonin.
Matthew Walker:
And then the other is melatonin as an antioxidant. But let me table that for now because I'll just get us sidetracked. That's what we know so far about melatonin, in terms of its supplementation benefit or lack thereof. Two final points that I shouldn't forget. One is the only population where we typically see some benefit, and it often is prescribed, is in older adults. Because as-
Andrew Huberman:
Older meaning 60 and older?
Matthew Walker:
Yeah. Sixty, 65 and older. Because as we get older, you can typically have what's called calcification of the pineal gland, which means that that gland that's releasing melatonin doesn't work as well anymore. As a consequence, they tend to have a flatter overall curve of melatonin released throughout the night. It's not this beautiful lovely peak and this bullhorn message of, It's darkness, please get to sleep. That's why older adults can have problems falling asleep or staying asleep. It's not the only reason, by any stretch of the imagination, but it's one of the reasons, and it's why melatonin implementation in those cohorts, older adults, especially older adults with insomnia, people have thought about that as maybe an appropriate use case.
Andrew Huberman:
Well, along those lines, if we were to compare dosages, do we know how much melatonin is typically released into the bloodstream per night, and can we use that as a rule of thumb by which to compare the typical amount that someone would supplement?
Matthew Walker:
Yeah.
Andrew Huberman:
I mean, typically the supplements for melatonin that I see in the pharmacy and elsewhere, and online range — anywhere from one milligram to 12 or even 20 milligrams.
Matthew Walker:
Yep.
Andrew Huberman:
My guess is that a normal night's release of melatonin typical for somebody in their twenties, thirties, forties would be far lower than that. Am I correct or wrong?
Matthew Walker:
Yeah, it's at many magnitudes lower. And this is one of the problems is that I see that too. I see typical doses are five milligrams or 10 milligrams, and of course if you're a supplement company putting 10 milligrams, versus five milligrams if that's what you're actually doing, which we'll speak about purity as well, it's kind of like the super gulp size. Nobody wants to lower price, we'll just give you more for the same price, and that's how we'll compete. So it's been this escalating arms race of melatonin concentration, and it really does not look meaningful for sleep in anyway.
Matthew Walker:
What we've actually found is that the optimal doses for where you do get sleep benefits in the populations that we've looked at are somewhere between 0.1 and 0.3 milligrams of melatonin. In other words, the typical doses are usually 10 times, 20 times, maybe more than what your body would naturally expect. And this is what we call a supraphysiological dose. In other words, it's far above what is physiologically normal. And to put that in context, imagine I said to you, I want you to eat 20 times as much food today.
Andrew Huberman:
I thought you were going to use testosterone as example. You're going to take 300 times the normal amount of testosterone. We know that would have tons of deleterious effects, be terrible.
Matthew Walker:
Yeah.
Andrew Huberman:
And yet you can do this. One thing that I'm concerned about these supraphysiological levels of melatonin is that many years ago, actually here at Berkeley when I was a graduate student, we would inject animals, which were seasonally breeding animals, with melatonin. And the consequence of that was that their gonads, either their testes or ovaries, would shrink many hundredfold or more. In other words, that they would go from having nice healthy-sized hamster testicles, what a hamster would consider healthy size, for a hamster, and they would shrink to the size of a grain of rice. So from an almond to size of a grain of rice. I had to see that only once for me to be very concerned about supraphysiological levels of melatonin.
Andrew Huberman:
And I realized that melatonin does different things in different species. We are not hamsters; we are not seasonally restricted breeders. There might be more breeding during certain seasons. I don't know those data, but nonetheless, hormones are powerful, and, sure, there is an optimal, and sometimes we see that going slightly above endogenous levels, for certain hormones, not always, can have beneficial effects, and sometimes it can have detrimental effects. I'm just concerned about taking high levels of a hormone that has effects on the reproductive axis. And that's one of the reasons why I get very concerned when I see people really getting aggressive about melatonin supplementation, taking 100, 10, 500, sometimes even 10,000 times the amount that we would normally release. That's my concern, although it's not nested in any one specific human study; I certainly don't want to see other people, and I don't want to personally, take a hormone that's known to be androgen-suppressive at high levels. Why would I take that? That's the question I ask myself.
Matthew Walker:
I think it's a very good point. And if you look at some of the evidence around melatonin's lethality, if you want to go to that extreme — for the most part it's pretty safe.
Andrew Huberman:
You mean you can take a lot of it before you die?
Matthew Walker:
Right. Exactly. Yeah. But-
Andrew Huberman:
I rather that's-
Matthew Walker:
That should not be your yardstick because you really need to think about your health, not just whether the thing is going to kill you or not as the decision matrix through which you pop a pill. And it comes onto this concern around melatonin because there was a study, I think it's one that you mentioned too, where they looked at over, I think it was at least over 20 different brands of melatonin supplements. And what they found is that based on what it said on the bottle versus what was in the capsules themselves, it ranged from, I think it was 83% less than what it said on the bottle to 478% more than what it said on the bottle.
Matthew Walker:
Now if that's a 10 milligram pill, and it's 478% more than 10 milligrams, and we're already at 10 milligrams at many tens of times more than is a physiological rather than a supraphysiological dose, we do need to be a bit thoughtful.
Andrew Huberman:
Yeah. Remember those hamsters, folks. Well, and I do appreciate the deep dive on melatonin because I think people need to understand that it's nuanced, it's a matter of dosages and timing, et cetera, and that it may have its place, as you mentioned, in older individuals. And I should mention that I'm an avid consumer of supplements that I believe in for me, and I have been for a very long time. So I'm by no means antisupplement. Some supplements I refuse to take or avoid taking, others I quite avidly take. And along those lines, I personally, and I don't know what your thoughts on this are, but there are a few things that I've personally found beneficial. I'd love your thoughts on them, and I would love it if you would tell me that everything I'm about to refer to is placebo, that would be fine. So that's what we do. We're scientists, we argue, and then we remain friends in, as it goes.
Matthew Walker:
Always.
Andrew Huberman:
So magnesium. There are many forms of magnesium. Magnesium citrate is, as we know, is a terrific laxative. Magnesium malate seems, at least from a few studies, seems to relieve some of delayed onset muscle soreness. Doesn't seem to create a kind of sedation. Two forms of magnesium that I'm aware of, magnesium bisglycinate and magnesium threonate ...
Matthew Walker:
Yeah.
Andrew Huberman:
... we believe based on the data can more actively cross the blood brain barrier. So you put it in your gut, but some of that needs to get into your brain in order to have the sedative effect. What are your thoughts on magnesium supplementation? Do you supplement with magnesium? And what studies would you like to see done if they haven't been done already?
Matthew Walker:
So I don't supplement with magnesium, but I do think threonate is interesting because of that higher capacity to cross the brain barrier and actually have a central nervous system effect. And the reason that that interests me is because the ... sleep is by the brain, of the brain and also for the brain as well as for the body. We just don't have a particularly good set of studies that have targeted exclusively threonate. We do have lots of studies that have just looked at magnesium, in general, for sleep. And overall the data is uncompelling.
Andrew Huberman:
Interesting.
Matthew Walker:
And for a while I was confused as to why, where did this come from? This kind of myth of magnesium. So I started looking back into the literature, and I've best traced it, at least as far as I can tell, to early studies showing that those who were deficient in magnesium also had sleep problems. They had other problems too, of course, but sleep problems were one of that set of sequelae that came from having lower magnesium. And when they supplemented with magnesium and tried to restore those levels, some of those sleep problems dissipated. And then that seems to have gotten lost in some game of sort of whispers around the room. And it's become translated into people who don't have sleep problems, who are healthy sleepers and who are healthy, in general, and who have healthy, normal levels of magnesium — if they take more magnesium, they will sleep better. And the data, really, there is not good.
Andrew Huberman:
Interesting.
Matthew Walker:
Once again, the only study that I've seen where magnesium did have some efficacy was in a study with older adults. I think they were 60 to 80 years old; it may have been exclusively women, now I think about it, and they also had insomnia. And in that population, you did see some benefits. And my guess is that because it's an older community as well, they were probably deficient in magnesium. So they fit the former category of simply when you're deficient, and you restore, you can help sleep sort of return to normal. But if you are not deficient, and you are healthy and you're not old, and you don't have insomnia, and you are supplementing thinking that it provides sleep, right now the data isn't supportive of that. But I just don't think we have enough threonate data to actually speak about that because it could just be a blood brain barrier issue so far with the other forms.
Andrew Huberman:
So maybe some additional studies looking specifically at threonate or bisglycinate would be useful.
Matthew Walker:
I'd love to have, yeah.
Andrew Huberman:
And magnesium's involved in so many cellular processes. You can imagine that this effect, if it truly exists is, as we say in science, in the noise. Meaning it's in the jitter of the data. But to isolate the real effect, one needs to do some more refined studies. What are some things that are of interest to you, if not things that you happen to take? These are not things that I personally take, mostly because I just haven't experimented with them. Valerian root is one. Tart cherry and kiwifruit. Tell me about valerian root, tart cherry and kiwifruit. This is new to me. I mean I've certainly heard of them, and tart cherry and kiwi sounds delicious. But what's happening with valerian root, tart cherry and kiwi. And are we talking about eating tart cherries and kiwis and valarian roots?
Matthew Walker:
Yes.
Andrew Huberman:
Or are we talking about taking them in pill form?
Matthew Walker:
Usually it's supplements, but it's also, both for tart cherries and for kiwis, it's the actual fruit themselves. Valerian often touted as a beneficial sleep aid, and lots of people swear by it too, but the evidence is actually quite against that.
Andrew Huberman:
Oh really?
Matthew Walker:
Not that it makes your sleep worse, but of at least the seven good studies that I've been able to find, and typically these are of the nature of what we call a randomized placebo crossover design. And I won't bore people with what that means. It's sort of one-
Andrew Huberman:
Good study. Solid study.
Matthew Walker:
Yeah. It's one of the sort of gold standard methods that we have, when we're looking at intervention studies such as drug studies. Five of the seven found no benefit of valerian root on sleep. Then two out of the seven, the data was just insufficient. I think it was a power issue where they just couldn't make any strong conclusions. And then I think there was, the most recent study, I think looked at two different doses of valerian, and I could have this wrong, and they just failed to find any effects, once again. But the stunning part of that paper, as I recall, they had this big table with all of the different sleep metrics that they looked at, and there were well over 25 different things that they tried to see if valerian impacted. And none of them were significant. Which stunned me because from statistical probabilities, we know if you just randomly performed 25 statistical tests, chances are, probabilistically, you'll just get one significant result by random chance. And even with random chance on their side, they still couldn't find a benefit of valerian.
Andrew Huberman:
So valerian root might be worse than nothing at all, so to speak.
Matthew Walker:
I mean, again, placebo effect, we can think about that too. And I would say that if you feel as though it's having a benefit for you, and with all of the caveats that we have with supplements, things like melatonin, purity, concentration, et cetera, maybe it's no harm, no foul. But I'm not a medical doctor and I don't tell anyone about, we have all of these disclaimers about not recommending.
Andrew Huberman:
Sure. And we'll include these. I mean I always say we're not physicians, we don't prescribe anything. We're scientists and professors, so we profess things. And it's up to people to be responsible for their own health, not just to protect us, but to protect themselves. I do want to hear about tart cherry-
Matthew Walker:
Yeah.
Andrew Huberman:
And kiwifruit. What's the story there?
Matthew Walker:
Strange, isn't it? I'm kind of a hard-nose scientist. When people some years ago started saying, oh, tart cherries, it's the thing, or kiwifruits, I was thinking, oh my goodness, this sounds a bit-
Andrew Huberman:
Been in California too long.
Matthew Walker:
Yeah, I know. Yeah. The sun has softened me some. But I thought, look, one of the things that we have to do as scientists is be as open-minded as possible. And I should not be so quick to dismiss. So I went to the literature, just started reading as much as I could about it. And there were three really good randomized placebo crossover trials with tart cherries. And what they found was that in one study it reduced the amount of time that you spent awake at night by over an hour. And then the other two studies, one of them found that it increased the amount of sleep that you got by 34 minutes. The other, it increased the amount of sleep that you got by 84 minutes.
Andrew Huberman:
Wow.
Matthew Walker:
And what's striking is that they were independent studies, I think meaning that they were from independent groups, and some of these guys and girls I know pretty well.
Andrew Huberman:
You know and trust their work.
Matthew Walker:
Right. I really trust their work too.
Andrew Huberman:
Were they ingesting actual tart cherries, or they're drinking the juice or in capsule form?
Matthew Walker:
It was juice. So in all three studies, it was juice. Although you can, I think as a supplement you can buy it in a capsule, and we've got no idea whether that changes the benefit or not. What was also interesting in, I think it was that last study where they got an increase in sleep by 84 minutes, it also decreased daytime napping significantly.
Andrew Huberman:
Wow. That's one that I could certainly make use of. I love my daytime naps, but I'd love to skip them too.
Matthew Walker:
Right. And we can speak about naps and sort of the upside and downside of that, which then made me think, well if that's the case, maybe the net benefit on sleep overall is no different. It's just that it decreases the amount of time that some people were taking to sleep during the day and giving it back to the night. But that wasn't the case. Because if you added the total amount of sleep that they were getting without tart cherries, both naps and nightly sleep combined, still when you took tart cherries, you still got to net some benefit of-
Andrew Huberman:
Interesting.
Matthew Walker:
Total amount of sleep. So far when it comes to supplements and those types of studies, they're good studies and the data looks interesting. But as a drug itself, if this was clinical drug, three studies that are somewhat small in nature and have some positive benefit, that's what we would call preliminary data of maybe a chin-scratching kind.
Andrew Huberman:
Yeah. And depending on the margins for safety, one might think, well, given that it's a tart cherry, as opposed to some pharmaceutical you need a prescription for, then some people — their threshold to experiment with supplements is quite low. Some people their threshold is quite high. I feel like there are two categories, or at least two categories of folks out there: people who hear, oh, tart cherry can improve sleep, and will run out and try it. And people who hear, well that's sounds crazy, why would I do that? But of course we have to remind people that tart cherry isn't really what we're talking about. Presumably if this is a real effect, and it sounds like it might be, that there's a compound in tart cherry-
Matthew Walker:
That's right.
Andrew Huberman:
That if we were to call it whatever, five alpha six, some molecule, if we referred to it by its technical name, then people would say, oh, that sounds like a very interesting technical way to approach sleep, but doesn't sound very natural. So both groups are a little bit misguided.
Matthew Walker:
That's right.
Andrew Huberman:
In the sense that people who think that everything that comes from naturally occurring foods, plants, et cetera, things that grow out the ground, that that's all safe, that's not true. And people that think that pharmaceuticals, if it's not evidenced with the purified molecule, then something's not of utility. Well, that's certainly not true. Somewhere in the middle I think lies the answer, which is, sounds to me like tart cherry is at least an intriguing potential sleep aid, intriguing potential sleep aid. And I'm underscoring potential. I'm certainly intrigued by it to the point where I might experiment a bit, but I'm an experimenter for myself.
Andrew Huberman:
Before I ask you about kiwi, I've had quite good results from taking something called apigenin, which is a derivative of chamomile. But in supplement form, I think I take 50 milligrams about 30 minutes before sleep, and I subjectively experience a better night's sleep, so to speak. I confess, I don't measure my sleep. I'm not a sleep tracker guy. But there are a few papers out there. They're not what we would call blue ... published in blue ribbon journals, but they have control groups and it looks somewhat interesting, and there, when I say apigenin, people get somewhat intrigued over this molecule. Chamomile has long been thought to be a sedative, a mild sedative, but a sedative.
Matthew Walker:
Yeah.
Andrew Huberman:
Do you drink chamomile tea? Do you take apigenin? What are your thoughts on apigenin,
Matthew Walker:
I don't. And I have looked into some of the data regarding sleep as well. Right now, from best I can tell, it's mostly subjective data rather than objective hard sort of sleep measures. And that's why right now it's sort of unclear, not no comment, but just unclear. Not dismissing it, because I think you and I both subscribe to the idea of absence of evidence is not evidence of absence. So keep your mind open, at least I tell that to myself.
Matthew Walker:
I think if you're finding a benefit, and you can do what I would think of if I was personally experimenting, which is both the positive and negative parts of the experiment. What I mean by that is, let's say that I now want to think about some kind of a sleep supplement. I will take some kind of baseline set of recordings for a month, and I will just gauge where I'm at, sort of supplement-free. Then I'll go on for a month to whatever I'm thinking of taking, and I don't supplement. But let's say that I want to and I experiment with that, and I feel as though, based on my metrics, be them objective from my ordering or be them subjective from whatever I'm writing down in the morning, and both are important — valid — subjective and objective. We like both in the sleep world. And I think, okay, look, it's clearly that it seems to have some kind of an effect. The key-
Matthew Walker:
... Believe that it seems to have some kind of an effect. The key thing, however, is then do the negative experiment, which is, now come off it for another month and see, do things get worse? If I can see that by directionality, then I'm starting to think, maybe I'm believing this a little bit more. That's the way I would typically approach a supplementation regiment if I were to do it, and that's just me. That's just the way my mind works. But ...
Andrew Huberman:
No, that's great. I think it's very scientific and organized in a way that allows you, and would allow other people, to make very informed decisions for themselves. I like that. I like to think in terms of manipulating any aspect of our biology that behavioral tools always are the first line of entry; then nutrition, everyone has to eat sooner or later, even if you're fasting; then perhaps supplementation; then prescription drugs; and then, perhaps brain machine interface, devices that are used to induce something. Those could be done in combination, but what concerns me is when I hear people say, "Well, what should I take?" Without thinking about their light-viewing behavior, et cetera. But of course, these things work in combination.
Matthew Walker:
Yeah. I think you're right that when it comes to sleep, there are many low-hanging fruits that don't necessarily require you to put exogenous molecules, in other words things like supplements, into your body, or use different types of drugs to help you get there. Now, when it comes to prescription sleep aids, I think I've been, again, a little bit too forthright. We know in clinical practice that there may be a time and a place for things like sleeping pills. They are a short-term solution to certain forms of insomnia, but they are not recommended for the long term. We also know that there are lots of other ways that you can get a sleep help or you can get a sleep curative profile from things like cognitive behavioral therapy for insomnia, which is a nondrug approach, a psychological one.
Andrew Huberman:
And quite effective from what I understand.
Matthew Walker:
Very.
Andrew Huberman:
[inaudible 02:06:11].
Matthew Walker:
Just as effective as sleeping pills, great data, more effective in the long term. There was a recent study published that after working with a therapist, some of the benefits lasted almost a decade.
Andrew Huberman:
Amazing.
Matthew Walker:
Now, if you stop sleeping pills, usually you have rebound insomnia, where your sleep goes back to being just as bad, if not worse. I think the same is true when we think about supplementation. There are so many things that are easy to implement when it comes to sleep that don't require venturing out into those waters. Again, we're not here to tell anyone about whether they should venture or not. That's completely your choice. All I'm saying is that if you want to think about optimizing your sleep, there are a number of ways that you can do it that don't necessarily require you to swallow anything or inject anything or smoke anything or [inaudible 02:06:59]-
Andrew Huberman:
Right, and for which the margins of safety are quite wide.
Matthew Walker:
Right.
Andrew Huberman:
That's the other one.
Matthew Walker:
Yes, sorry. Thank you for that.
Andrew Huberman:
Speaking of low-hanging fruit, I don't know how low it hangs in reality, but what about kiwi? They're delicious to me, anyway.
Matthew Walker:
Yeah, the humble kiwifruit, shouldn't be mistaken for the flightless bird of New Zealand, which is the native bird. We're talking about the kiwi, the fruit here, which those trees and shrubs are mostly Southeast Asia. Kiwifruits have been previously touted as potentially having a sleep benefit, which again got me curious and I at first threw it out. To my knowledge, there's really only one published human study that's of any value, but what they did find was that it decreased the speed of time with which it took you to fall asleep.
Andrew Huberman:
These are people ingesting the whole kiwi?
Matthew Walker:
So, it's ingesting the whole kiwi.
Andrew Huberman:
With the skin? I eat the skin and-
Matthew Walker:
Skin.
Andrew Huberman:
People cringe when they see me doing. But-
Matthew Walker:
Well, I think-
Andrew Huberman:
Oh, don't eat the skin?
Matthew Walker:
No. I think the idea is some of the good stuff, and I'll come onto this, may actually be in the skin itself.
Andrew Huberman:
Oh, wonderful. Thank you. You just helped me win a bet. I'll give you your [inaudible 02:08:07].
Matthew Walker:
Okay. Yeah, you can pay me later. By the way, the skin is use ... No, he just told me to say that. No, he did not. The skin seems to be part of this potential sleep equation. You fell asleep faster and you stayed asleep for longer, and you spent less time awake throughout the night. I just thought, well, that's one study. What can you really do with that? There is another study, however, in an animal model, which is a little bit more interesting. Once again, they found a very similar phenotype that the rats ... Oh, sorry, they were mice. The mice fell asleep faster and they also spent longer time in sleep. The sleep duration also increased. What was also interesting, mechanistically, and this is not the mechanism that I think ties together tart cherries, kiwifruit, and things like melatonin, because I think there could be one common binding mechanism — what they found in the animal study is that they could block those kiwifruit sleep benefits using a GABA-blocking agent.
Andrew Huberman:
Interesting.
Matthew Walker:
Now, GABA, which stands for gamma-aminobutyric acid, is one of the major inhibitory neurotransmitters of the brain. It's like the red-
Andrew Huberman:
A naturally occurring sedative.
Matthew Walker:
Yeah, it's the red light on the traffic light signal. Others are green light. GABA is red light. By playing around with some clever drugs to manipulate the system, they could prevent the benefit of the kiwifruit by buggering around with the GABA receptor, meaning that perhaps part of the kiwifruit benefit on sleep was mediated by the brain's natural inhibitory neurotransmitter system, called the GABA system.
Andrew Huberman:
That's exciting.
Matthew Walker:
I thought that convinced me a little bit more that maybe there's something here to read into. To be determined. Again, here is the banner, but tart cherries and kiwifruits, the data surprised me because in part, I was so preoccupied with being, I don't know, a bit purist about it and a bit snobby thinking, "Oh, come on, that's definitely not going to work." Well, [inaudible 02:10:31]-
Andrew Huberman:
Compounds are compounds. I look forward to a day when supplements are no longer called supplements because at the end of the day, whether or not something has an effect, whether or not it's a whole kiwifruit or a derivative kiwifruit, will depend on the molecular compound. As you mentioned, this potential mechanism via the GABA system, we both as scientists get excited about mechanism because when you can trace a mechanism and a pathway, it provides a rationale, a grounding for why kiwi, of all things, or tart cherry, of all things, might help increase total sleep time.
Andrew Huberman:
I'd be remiss if I didn't mention or ask about tryptophan and serotonin. I can anecdotally say when I've taken tryptophan, the precursor to serotonin, or serotonin itself, I have a horrendous night's sleep. I fall asleep very easily and I experience ridiculously vivid dreams, neither pleasant nor unpleasant, it's a mishmash, and then I wake up and I experience several days of insomnia. I've done the positive control and the negative control at all the variations thereof to confirm that, at least for me, supplementing with serotonergic agents is a bad idea for me.
Andrew Huberman:
Tryptophan is a common sleep supplement and sleep aid that's discussed. The normal architecture of sleep involves the release of serotonin, but in a very timed and regulated way. What are your thoughts about serotonin in sleep if you had to put that into a nutshell, and then why supplementing with serotonin and, or its precursor, tryptophan, might be a good or a bad idea for somebody?
Matthew Walker:
I think one of the potential dangers is that based on what's going on in your body, that can change the absorption of natural tryptophan and serotonin uptake within the brain itself. I'm always thoughtful when you're playing around with that Mother Nature dynamic as it were. The data, as you described, is a little bit all over the map. Some people say that it knocks them out. Other people say, just like you do, "It has a terrible impact on my sleep and when I stop, it's pretty bad for a couple of days. It seems to have this lingering after-effect." I think what could be happening here is we need serotonin, just as you described, to be modulated in very specific ways during the different stages of sleep. If you look at the firing of the brain epicenters where serotonin is released, and there's a bunch of them in the brain stem, what you find, and in the release of serotonin too, when we're awake, it's usually in high concentrations.
Matthew Walker:
As we start to drift off to sleep, it lowers some, but not necessarily dramatically as we're going into non-REM sleep. But then when we go into REM sleep, serotonin is shut off. One of the other neuromodulators, noradrenaline, also shut off. REM sleep is the only time during the 24-hour period where we see noradrenaline and serotonin, or norepinephrine, completely shut down. When I say serotonin, we're also talking 5-HTP ... Sorry, 5-HT. That's just its chemical name here. Speaking about serotonin or 5-HT, it's the same thing. Norepinephrine, noradrenaline, both of those need to be shut down for you to produce REM sleep. One of the other neuromodulators that then ramps up to produce REM sleep is acetylcholine. These three neuromodulators have this incredible reciprocal dance that they have for you to generate what is called a natural architecture of sleep throughout the night.
Andrew Huberman:
It's the push-pull again.
Matthew Walker:
It's a push-pull again. It's chest and back, whatever you want to think of.
Andrew Huberman:
Sure.
Matthew Walker:
That's why I think if you're trying to increase, dramatically drive up your serotonin levels at night, and that sustains throughout the night when you're trying to get into REM sleep, you could be artificially fragmenting REM sleep. Now, I don't know the data. I don't think anyone's really got good data.
Andrew Huberman:
No, I haven't been able to find [inaudible 02:14:55]-
Matthew Walker:
If you were to say, "Matt, two years' time, that's the data. Help me understand the potential mechanism, or let's design some experiments. Where would you go first?" I would say, "Let's look at the disruption of REM sleep, non-REM sleep, reciprocal regulation," because you need serotonin to be up at one time, down at another.
Andrew Huberman:
I agree with everything you said, and I'm personally never taking tryptophan or serotonin again unless there's some clinical reason for that, that I would need to do that. I want to ask about some other prosleep behaviors, but before I do that, let's talk about naps. I love naps. I come from a long history of nappers. My dad always took a nap in the afternoon. I take a 20 or 30-minute nap or I do a practice, which I took the liberty of coining "NDSR," non-sleep, deep rest, some sort of just passive laying-
Matthew Walker:
Genius.
Andrew Huberman:
Out there, feet up elevated, or I'll do yoga nidra, I'll do hypnosis, or something of that sort. But 20 or 30 minutes of that has been very beneficial for me to get up from that nap or period of minimal wakefulness, we'll call it, and go about my day quite well and also fall asleep just fine. What are the data on naps? Do you nap, and what are your thoughts about keeping nap short, meaning 20 to 30 minutes, versus getting out past 90 minutes, two hours? For you personally, naps, yay, nay or nah?
Matthew Walker:
I don't nap and I've just never been a habitual napper.
Andrew Huberman:
Is that because you don't feel sleepy in the afternoon or because ...
Matthew Walker:
I typically don't feel sleepy.
Andrew Huberman:
Okay. So, you're just hardier though.
Matthew Walker:
I wouldn't say hardier. I may be less capable of falling asleep. My sleep drive-
Andrew Huberman:
But you're not dragging through the afternoon?
Matthew Walker:
No, I don't drag through.
Andrew Huberman:
So, you don't nap because you don't feel a need to nap.
Matthew Walker:
That's right, yep.
Andrew Huberman:
That's right? Okay.
Matthew Walker:
Now, it's not that I am immune to what we call a postprandial dip in alertness. I definitely feel as though there can be this afternoon lull, where I'm not quite as on as I was at 11 o'clock in the morning. We know the physiology to that, which brings us back to whether we were designed to nap. For naps, we've done lots of different studies, and other colleagues have done these studies too. Naps can have some really great benefits. We found benefits for cardiovascular health — blood pressure, for example. We found benefits for levels of cortisol. We found benefits for learning and memory, and also emotional regulation.
Andrew Huberman:
How long are the naps typically in those studies?
Matthew Walker:
Anywhere between 20 minutes to 90 minutes. Sometimes, we like to use a 90-minute window so that the participant can have a full cycle of sleep and, therefore, they get both non-REM and REM sleep within that time period. Then when we wake them up, we usually wait a period of time to get them past what we call sleep inertia, which is that window of grogginess where you say to your better half, "Look, darling, please don't speak to me for the first hour after I've ... "
Andrew Huberman:
"Don't anything right now."
Matthew Walker:
Yeah. After the first hour of waking up, I'm just not the best version of myself.
Andrew Huberman:
Right.
Matthew Walker:
We wait for that time period, and then we do some testing. We've done some testing before and after, and we look at the change. That's how we measure what was the benefit of naps and the reason why we sometimes do 90 minutes, so that they get all of those stages of sleep, and then we correlate how much benefit did you get from the nap and how much of that benefit was explained by what REM sleep you got, what deep sleep you got, what light sleep you got. That's the only reason that we use that as an experimental tool. What we've also found is that naps of as little as 17 minutes can have some quite potent effects on, for example, learning. None of this is novel. NASA pioneered this back in the 1990s. During the missions, they were experimenting with naps for their astronauts, and what they found was that naps of little as 26 minutes improved a mission performance by 34% and improved daytime alertness by 50%.
Andrew Huberman:
Fantastic.
Matthew Walker:
It birthed what was then called the NASA nap culture throughout all terrestrial NASA staff during that time period. So, it's long been known that naps can have a benefit. Naps, however, can have a double-edged sword. There is a dark side to naps, and it comes back to our story of adenosine and sleep pressure. The longer we're awake, the more of that sleep pressure adenosine that we build up. But what I didn't tell you is that when we sleep, the brain gets the chance to essentially clear out that adenosine, and after about 16 hours of wakefulness and then after about eight hours of sleep, eight hours of sleep seems to be able to allow the brain to decrease its adenosine levels back to normal.
Matthew Walker:
And so naturally, we should start to wake up, which also aligns with your circadian rhythm. Those are two separate processes, but with about eight hours of good quality sleep, seven to nine hours for the average adult, we are free of all of that adenosine. We've evacuated it essentially out of the brain, and we wake up naturally feeling refreshed. The reason that naps can be potentially dangerous is that when you nap, you are essentially opening the valve on the pressure cooker of sleep pressure, and some of that sleepiness is lost by way of the nap. For some people and not all people, and you're a great example of this, however, if they are struggling with sleep at night and they nap during the day, it makes their sleep problems even worse.
Matthew Walker:
For people with insomnia, we typically advise against napping. The advice is if you can nap regularly and you don't struggle with sleep at night, then naps are just fine. But if you do struggle with sleep, stay away from naps. If you are going to nap, try to limit your naps. Try to cut them off a bit like caffeine, maybe eight to 12 hours, maybe not that far off, maybe seven to six hours is a good rule of thumb. Try not to nap essentially late in the afternoon. If you do take a nap and you want to maintain your ... You don't want to have that grogginess hangover that can happen after a full night of sleep for the first hour, try to limit it to about 20, 25 minutes. That way, you don't go down into the very deepest stages of sleep, which if I wrench you out of with an alarm, then you almost feel worse for the first hour after the nap because-
Andrew Huberman:
Yeah, I've definitely experienced that if I oversleep. Certainly, if the sun goes down during my nap, and I wake up and overall lighting conditions have changed, I find it very hard to jolt myself back into the evening, and it can screw me up. So, I try and keep those naps pretty brief. I should say, I'm very happy to hear you mention individual differences and why some people might want a nap and other people might not want to nap. I have a colleague, Liqun Luo, he'll be familiar to many neurobiologists, but-
Matthew Walker:
Yeah, of course.
Andrew Huberman:
Who's an absolutely spectacular scientist, member of the National Academy, Howard Hughes Investigator, just a phenom and has a ton of energy. But years ago, I learned that he always takes a nap in the afternoon, so much so that when he travels to give seminars at other universities, he will tell his postlunch host, whoever it is that he's supposed to meet with, "May I have your office for 30 minutes of our ... 30-minute discussion, or 60-minute discussion, because I like to take a nap?" He does that and then gives ... His talks are typically in the afternoon in academic culture. He describes the effect of the nap for him, these short naps in the afternoon, being so profound for his productivity. That's actually what inspired me to start feeling okay about my desire to nap. And so I think, for me, that was great vindication for those that might feel guilty about wanting a nap. But I take to heart your note about avoiding naps if you have trouble falling and staying asleep because I think that I have family members who also, if they nap, they're a wreck. They can't sleep.
Matthew Walker:
Yeah, I think it's just we've often been very pro-sleep via the sleep community. So, I think it's good to always point out these potential dark sides of any aspect. But you're absolutely right. No one should feel guilty about getting the sleep that they need. I think that's been one of the big problems in society. Society has stigmatized sleep with these labels of being slothful or lazy, and we're almost embarrassed to tell colleagues that we take a nap. I think sleep is a right of human beings, and I therefore think that sleep is a civil right of all human beings, and no one should make you feel unproud of getting the sleep that you need.
Andrew Huberman:
Well, I love that and it's an important point. I also feel that one of the best ways to beat your competition in any endeavor is to outlive them. Now that we know that sleep can enhance longevity and lack of sleep can shorten one's life, that's all the justification I need anyway. Can somebody sleep too much? Is it possible to get too much sleep?
Matthew Walker:
It's a very good question and there are probably two things to say about it. There is a condition that we call hypersomnia, but that's a mixture of things. It's where people have either a very high sleep need, or they are very sleepy during the day and they're typically falling asleep. These can happen in a variety of different clinical contexts. One of the places where we've often seen hypersomnia believed to manifest is in depression. But if you look at some of those studies, it turns out it looks more as though those people are simply reporting being in bed longer, but not necessarily sleeping longer. That fits very well with one of the profiles that we know of depression, which is anhedonia. You don't get pleasure from normally pleasurable things. So, you just don't want to go out into the world. You don't want to interact because you're depressed. So, what do you do? You just stay in bed and that-
Andrew Huberman:
Blinds closed, watching TV on the phone.
Matthew Walker:
Right, and that just looks as though-
Andrew Huberman:
Feeling miserable.
Matthew Walker:
When people say, "What time did you go to bed, and what time did you get out of bed?" The mistake made in that question is, okay, that's how much time they're sleeping, when you should have said, "What time did you fall asleep and what time did you wake up?" The answer could be very different. That's hypersomnia from a clinical context. Can you sleep too much though? One of the data points that argue is yes, that's possible, is when we look at all-cause mortality. Certainly, what you find is that using this sweet spot of seven to nine hours, when you start to drop below seven hours, there is a relationship which suggests that the shorter your sleep, the shorter your life. Short sleep in that regard predicts all-cause mortality.
Matthew Walker:
But it's again not a linear relationship like the one that we've seen with REM sleep. Once you get past nine hours, the mortality curve stops going down, and then once you get further, 10 or 11 hours, it hooks back up again. It's almost like a J-shape tilted over a little bit and reversed. So, there's this strange hook. What's going on here? Right now, sleep science has at least two nonmutually exclusive explanations for this. The first is that if you look at some of those populations, the idea is that whatever was causing them illness and took their lives was just too much for sleep to deal with. However, we know that when we get sick, one of the things that we do immediately — and there's a whole mechanism, an inflammatory mechanism, cytokine mediated – we want to sleep more. We just want to curl up in bed and sleep it off.
Matthew Walker:
The argument there is that it's not that sleep was killing people prematurely. It was that these people were calling on the help of sleep. They were calling on this thing called the Swiss Army knife of health that is sleep. But whatever it is that they were facing was just too powerful for sleep to overcome. So, it artificially looks as though more sleep leads to a higher risk of death, when sleep was actually responding to the mortality risk and it lost the battle. That's one argument. The second is that we know that sleep quality and poor sleep efficiency is a very strong predictor of all-cause mortality. When you look at people who often report sleeping long amounts, 10 or 11 hours, they typically report having very poor quality sleep.
Matthew Walker:
So because their quality of sleep is poor, they just try to sleep longer to try and get some of that back. Again, here now, it's poor quality sleep masking as too much sleep leading to this artificial huck, which looks like mortality. That's a second explanation. A third, which is more of a gedanken, which is just a thought experiment — I'm of this mentality; I don't know how many other people share this — I actually think there could be a thing as too much sleep. Physiologically, I think it's possible, but the reason I think that is because it's no different than food, water or oxygen. Can you overeat? Yes, you can. Can you undereat? Of course.
Andrew Huberman:
Or light. Light in the early part of the day, all throughout the day, wonderful.
Matthew Walker:
Right.
Andrew Huberman:
Light late in the day and at night, detrimental.
Matthew Walker:
Bidirectional. For water, can you overhydrate? Hyponatremia — it happened in the 1990s and 2000 with the ecstasy craze, where governments were saying, "You're dehydrating. You're dancing all night. Please drink water." They drank too much water, the blood electrolytes went all over the place and they were having cardiac arrests or stroke.
Andrew Huberman:
Yeah, people were dying.
Matthew Walker:
They were dying because of excessive hydration. Can you get too much oxygen? Hyperoxemia, it can cause free radical damage, which can be profoundly harmful and kill brain cells. Can you sleep too much? Which is the fifth element of the life equation alongside food, water, oxygen. Fourth, I should say. Can't even count. Clearly, I'm sleeping well. Yes, I think there could be that possibility. Are most people in danger of getting too much sleep, au contraire, if you look at the data?
Andrew Huberman:
Right. Yeah, so-
Matthew Walker:
But I don't dismiss that idea. I think it's possible.
Andrew Huberman:
That's a very thorough and very nuanced and very clear answer. It's so interesting to think that a lot of the data that are out there talking about being in bed too long, it's just trying to compensate for the actual fragmentation of sleep.
Matthew Walker:
Right.
Andrew Huberman:
What I'm coming away with is that there are many paths to this, both positive and things to avoid but the idea is to get, most nights, a similar amount, probably seven to nine hours, somewhere in there, of high quality sleep. This notion of sleep quality is going to become, I would hope, a phrase that more people think about and learn about and cultivate as a practice.
Andrew Huberman:
I'm aware of at least one company that's starting to track in their sleep monitoring device, and that's orgasm. Sex, orgasm and masturbation, topics that are somewhat sensitive, but from the perspective of biology, none of us would be here were not for sperm meets egg in some fashion, either in a dish or in vivo. What are the data as you know them to be, or maybe your lab is even doing this kind of work and exploration, about the role that sex, orgasm and masturbation play in getting to sleep and staying asleep and sleep quality? Certainly, those behaviors and those physiological events have been part of our evolutionary history.
Matthew Walker:
Right.
Andrew Huberman:
What's the story there? What can we say about this in terms of science and, dare I say, practice?
Matthew Walker:
Yeah. I mean, it's almost that caricature of in the movies, where a couple makes love and then all of a sudden, you just hear snoring, with the idea that it's somewhat somnogenic, that it's sleep promoting.
Andrew Huberman:
Well, the post-orgasmic increase in prolactin-
Matthew Walker:
Well, that's very ...
Andrew Huberman:
Is ...
Matthew Walker:
Is interesting.
Andrew Huberman:
Thought to be a naturally occurring sedative that presumably has a function in ...
Matthew Walker:
Oxytocin has also that benefit where you see a dissipation of the fight-or-flight branch of the nervous system, which has to happen for you to fall asleep. That's why we often see, here at the sleep center, a phenomenon called wired-and-tired, where people say, "Look, I am so desperately tired. I'm so tired, but I can't fall asleep because I'm too wired." Your sleep drive — you're desperately tired, it's there, but because you're wired, because you've got too much sympathetic activation, too much cortisol as well, you can't fall asleep. It's an impressive roadblock to anything like good sleep, and it's one of the principle mechanisms that we now believe, stress and physiological activation, that is the underlying cause of many forms of insomnia. But coming back to sex, the data is actually quite compelling that both, either, subjectively assessed sleep quality or objective amount of sleep, sex that has resulted in orgasm, and I think it's that latter part that typically needs to happen-
Andrew Huberman:
I would imagine so.
Matthew Walker:
Between two mutually exclusive individuals, where both are beneficial in terms of receiving an orgasm, that's-
Andrew Huberman:
Yeah. I would say, any discussion about sex, we are referring to consensual, age-appropriate, species-appropriate, context-appropriate.
Matthew Walker:
Wow, I would never have even gone to the species-appropriate, but thank you for doing that.
Andrew Huberman:
I put species in there because it's the Internet, and people will come up with all sorts of ideas.
Matthew Walker:
Wow, good point, well made.
Andrew Huberman:
I think age-appropriate, consensual-
Matthew Walker:
Yeah, consent.
Andrew Huberman:
Context-appropriate and species-appropriate covers all the bases, but if I missed any-
Matthew Walker:
Yeah. Well, let's-
Andrew Huberman:
Put them in the comment section and we'll be sure to take note.
Matthew Walker:
No, I think that's really well said and important to say. So the data is there when you look at couples who have orgasm. We've also, however, found benefits of masturbation. It's not frequently spoken about, but if you do some surveys, it turns out that people will often use masturbation as a sleep tool if they're struggling with sleep. I know this sounds almost like a strange conversation, or it's a taboo conversation, but I think we just need to be very open about all this.
Andrew Huberman:
I started off in science; one of the things I worked on early in my career, the very first topic, was the topic of early influences of hormones, estrogen and testosterone, on sexual development of the brain and body. When you are weaned in a laboratory like that, regardless of what era, you look at sex and its behaviors and its hormones and its physiologies as a scientist.
Matthew Walker:
Right.
Andrew Huberman:
To be clear, what we're doing here, we're exploring these behaviors from the ...
Andrew Huberman:
And we're exploring these behaviors from that perspective. I mean, one thing is for certain, everyone is here because a sperm met an egg, either in a dish or in vivo, as we said before, and, at least in 2021, there's no way around that fact.
Matthew Walker:
Yeah.
Andrew Huberman:
And what preceded that, is typically this act we call sex. And sometimes, hopefully, I like to think, orgasm is associated with that sexual activity. Masturbation is one dimension of that, is something that I think it can and should be discussed if in fact there are data that relate it to sleep.
Matthew Walker:
Yeah. And both of those routes seem to lead to a sleep benefit. Now, I'm not saying that it's all about the orgasm. I think as we spoke about with oxytocin, there is some degree of pair bonding, that if you have a partner and you experience an intimate, loving relationship that involves that, then you can have hormonal benefits that are sleep promoting that may not necessarily be seen if you're just engaging in the solo singular act of masturbation. So what we certainly know now ... and I am not someone to take any advice on when it comes to anything relationshipwise or sexwise, but-
Andrew Huberman:
That's a different episode of the podcast.
Matthew Walker:
Yeah, yeah, yeah, certainly. And that's not a podcast series that I'm going to be releasing anytime soon. It's going to be mostly about sleep. Although I will touch on — I'll release a podcast on sleep and sex. But that's the data that we have so far. We also know that it works both ways, though. And it's commonly the same way with sleep. Sleep and exercise, sleep and diet. How you eat can affect how you sleep, how you sleep can affect how you eat. Same with exercise, and it turns out it's the same way with sexual behavior too. So here we're talking about whether sex can help with sleep. Can sleep help with your relationship and sex? And the answer is: yes, it can. Firstly, we know, and we've spoken a little bit about this, that the reproductive hormones are under profound sleep regulation-
Andrew Huberman:
Both estrogen and testosterone.
Matthew Walker:
Estrogen, testosterone-
Andrew Huberman:
Because we hear so often about testosterone-
Matthew Walker:
FSH, as well.
Andrew Huberman:
... but women who, or I should say anyone who's interested in having higher levels of estrogen or normal, healthy levels of estrogen, I presume the data show for estrogen what the data also show similarly for testosterone.
Matthew Walker:
Testosterone.
Andrew Huberman:
If you get too little or poor quality of sleep, both sex steroid hormones, as they're referred to, testosterone and estrogen, which are present in both males and females and every variation thereof, are going to be diminished below normal, healthy levels.
Matthew Walker:
Yep, estrogen, testosterone, FSH in women, a key hormone in the regulation for ... key for conception, of course. All of these sex hormones seem to become significantly disrupted when sleep becomes of short quantity or poor quality. We also know that in women sleep disruption can usually lead to menstrual cycle disruption. We know this particularly from evidence in shift working women, where they are nighttime shift workers; they struggle with sleep during the day; often menstruation is disrupted or even becomes impaired. But we also know it works this way not just for sex hormones, but for sex itself. For example, we found that for every one hour of sleep, extra sleep that a woman gets, her interest in becoming sexually intimate with her partner increases by 14%, which is-
Andrew Huberman:
Substantial.
Matthew Walker:
... a nontrivial amount. And then the final part of, I think, this equation when it comes to sleep and sex is your relationship itself. And there's some great work here from UC Berkeley, by Professor Serena Chen. And what she found was that restless nights mean for far more brutal fights in your relationship. And they did this in-
Andrew Huberman:
And vice versa.
Matthew Walker:
... number of different, elegant ways, and vice versa as well.
Andrew Huberman:
I mean, not that I've ever had conflict in relationships, but-
Matthew Walker:
Me ... This is just data I've read, I've never experienced that at all. And they found reliably that sleep would predict higher likelihood of relationship conflict. Secondly, if you got into that conflict, the chances of you resolving it were significantly lower when the parties had not been sleeping well. Part of the reason is because when you are not well slept, your empathy goes down. So you are taking more of an abrasive stance with your partner rather than a more agreeable stance with your partner. So at almost every dimension of a human loving relationship, sleep can have a dramatic impact.
Andrew Huberman:
I think these are really important things to underscore. One of the most common questions I get, because there seems to be a community of people on the Internet that are obsessed with this — I don't know who they are because it's, all this Internet stuff is shouting into a tunnel and getting comments back and written for them; It's a very bizarre conversation, so to speak — is whether or not sexual behavior itself lowers or increases testosterone? And I went into the data which expands many decades actually, both animal studies and human studies. And it seems just to underscore this, as long as we're talking about this subject, that it does seem that sexual activity, sex between two people, does seem to increase testosterone in both.
Andrew Huberman:
There is this question about orgasm or no orgasm, ejaculation, no ejaculation, and indeed there do seem to be some effects of restricting ejaculation in males as a form of further increasing testosterone. So sex without ejaculation further increasing testosterone. But the data are not clean-
Matthew Walker:
Okay.
Andrew Huberman:
... presumably because organizing these sorts of studies and getting truth in self-reporting is probably hard to get from subjects. But everything sort of points in the direction that, provided that the relationship is a healthy one — it's consensual, it's age appropriate, context appropriate, species appropriate — that sex between two individuals does seem to increase the sex steroid hormones, testosterone and estrogen, toward healthy ranges. And what I'm hearing now, the sort of gestalt of what ... of the discussion we just had is, that too can promote sleepiness, restful states and quality sleep.
Andrew Huberman:
And I think this is an important conversation that just hasn't been held enough. I mean sooner or later, both in the U.S. and elsewhere, we're going to have to acknowledge that we are biological organisms of some sort and that we have choice in life about all these things, from supplementation to sex or no sex, et cetera. But that they have profound effects on our core biology. I mean, it's fascinating to me that the areas of the hypothalamus, the preoptic area, the supraoptic areas, those areas which the names might not mean anything to anybody besides Matt and I sitting here, but those areas sit cheek-to-jowl with each other in the hypothalamus and control sleep and sex.
Matthew Walker:
Yeah.
Andrew Huberman:
The trigger of orgasm, the appetite for food-
Matthew Walker:
Yeah, you can say-
Andrew Huberman:
... the appetite for water, for electrolytes. I mean the hypothalamus is kind of a festival of neurons with different booths for different-
Matthew Walker:
It's such a small-
Andrew Huberman:
... primitive behaviors. Right.
Matthew Walker:
... small structure in the brain, but it's the orchestrator of a vast number of our behaviors. Disproportionate in terms of its size versus effect.
Andrew Huberman:
Yeah. I don't think you can go to this hypothalamic festival without at least seeing all the booths, whether or not you decide to visit them or not.
Matthew Walker:
I love that ... master analogy.
Andrew Huberman:
So I'm glad that we've broached that conversation, and I hope people will think that we've approached it with the appropriate level of sensitivity. It's an important one that we're going to hear more about one way or the other. People are certainly thinking about this, if not engaging in these sorts of behaviors or avoiding them. So the more we can understand about the biology, the better.
Matthew Walker:
Well said.
Andrew Huberman:
And so thank you for bringing that topic up because, for the record, Matt tabled it for discussion.
Matthew Walker:
I said ... We were just like chatting outside, and I think we said something about,-
Andrew Huberman:
Sure.
Matthew Walker:
... sort of, sex and I said, "Yeah, we can go there, there's some actually interesting data."
Andrew Huberman:
Absolutely. I want to touch on just two remaining topics. One is, are there any unconventional sleep tips or things about sleep that we've overlooked? If we've covered everything, great. But we hear to keep the room cool, we hear because of this temperature phenomenon, the light aspects, the considerations about alcohol, CBD, marijuana, cognitive behavioral, tart cherry fruit, kiwis, perhaps-
Matthew Walker:
[inaudible 02:44:07].
Andrew Huberman:
Perhaps-
Matthew Walker:
Please don't put me on the hook for tart cherries and kiwi. I was just offering what I know about the ways, I'm not-
Andrew Huberman:
No, and these are considerations and whether or not people batch these things. I won't even list them off now because there are too many jokes that one could make.
Matthew Walker:
And I have no affiliation with any of these products or companies or anything so-
Andrew Huberman:
Well, I'm going to take out stock in a kiwi company. No, I'm just kidding. But the question I have is about any unconventional or lesser known things, or maybe you do things or you think about things just in a purely exploratory way as a scientist. The "what if" kind of things that ... "Yeah, what if it turns out that," and here I just, I've got a blank there for you to fill in.
Matthew Walker:
I think you're beyond the standard fare that I've dished out plenty of times of sleep hygiene, of regularity, temperature, darkness, alcohol, caffeine — and we've spoken about all of those. What are some more unconventional tips, I guess? The first one which is unconventional, along the lines of naps: if you've had a bad night of sleep, let's say that you are starting to emerge with insomnia and you've had a bad night of sleep, the advice, and I learned this from my wonderful colleague, Michael Perlis: do nothing. What I mean by that is don't wake up any later, don't sleep in the following day to try and make up for it, don't nap during the day, don't consume extra caffeine to wake you up — to try to get you through the day — and don't go to bed any earlier to think that you're going to compensate.
Matthew Walker:
And I can explain all of those things, but if you wake up later, you're not going to be sleepy until later the evening. So you're going to go to bed at your normal time and you won't be sleeping. You'll think, "Well, I just came off a bad night of sleep and now I still ... I can't even get to sleep and it's my normal time." It's because you slept in later than you would otherwise, and you reduced the window of adenosine accumulation before your normal bedtime. So don't wake up any later. Don't use more caffeine, for the reasons that are obvious, because that's only going to crank you and keep you awake the following night, or decrease the probability of a good following night of recovery sleep. Third, I mentioned don't take naps because once again that will just take ... Naps, particularly later in the afternoon, I almost liken them to snacking before a main meal. It just takes your appetite off the edge of that main meal of sleep. So don't do it.
Matthew Walker:
And then finally, don't go to bed any earlier. Resist and resist and go to bed at your normal time. What I want to try and do is prevent you from thinking, "Well, I had such a bad night last night, and I normally go to bed at 10:30; I'm just going to get into bed at nine o'clock because last night was just so bad." But that's not your natural bedtime, and it's not aligned with your natural chronotype, because presumably you kind of know something about that, or are a morning type, evening type, and you're trying to sleep in harmony, which is usually how you get best quality sleep. But you go to bed at nine, and my body is not ready to sleep at nine o'clock. But I'm worried because I had a bad night of sleep last night, so I get into bed, and now I'm tossing and turning for the first hour and a half because it's not my natural sleep window. But I just thought it was a good idea. And if I didn't know anything about sleep, I would think all of these same things too, so I'm not finger wagging.
Matthew Walker:
If I have a bad night of sleep ... And I am not immune just because I know a little bit about sleep, doesn't mean I don't have my bad nights. I do. Doesn't mean I haven't had bouts of insomnia in my life. I have. But after a bad night of sleep, I do nothing. I don't do any of those four things. I think the second tip I would offer in terms of unconventional is have a wind-down routine. Many of us think of sleep as if it's like a light switch. That we just jump into bed and when we turn the light out, sleep should arrive in that same way. Just the binary, it's on or it's off. Sleep is a physiological process. It's much more like landing a plane. It takes time to gradually descend down onto the terra firma of what we call good solid sleep at night. Find out whatever works for you, and it could be light stretching. I usually meditate for about 10 or 15 minutes before bed. Some people like reading. Try not to watch television in bed. That's usually advised against.
Andrew Huberman:
Just something that doesn't emit too much light to your eye.
Matthew Walker:
Too much light, too activating. You know, you can listen to a relaxing podcast, although we can speak about technology in the bedroom too, but have some kind of a wind-down routine. It's almost ... You wouldn't race into your garage and come to a screeching halt from 60 miles an hour. You typically downshift your gears, and you slow down as you come into the garage. That's the same thing with sleep too. So, that's the second thing. Have some kind of a wind-down routine. Find what works for you, maybe it's taking a hot bath or a warm shower, and then stick to it. We do this with kids all the time. We find out what their bedroom, sorry, their bedtime regiment is, and then we just stick to it faithfully, because we humans are the same way too.
Matthew Walker:
The third thing is a myth. Don't count sheep. There's a study done here at UC Berkeley. I didn't do this today, I wish I did — it was by my colleague, Professor Allison Harvey. And they found that counting sheep actually made it harder to fall asleep. It made matters worse. As a countermeasure to that, what they did find was that taking yourself on some kind of a mental walk — so think about a nice walk that you take in nature or a walk on the beach, or even a walk around an urban environment-
Andrew Huberman:
Just visualizing that.
Matthew Walker:
... and visualizing that, that seemed to be beneficial. The other thing about sort of that idea of shifting focus away from your mind itself, get your mind off itself, is a good piece of advice. Catharsis. You can try to write down all of the concerns that you have, and do this not right before bed but usually an hour or two before bed. Some people call it a worry journal. And to me, it's a little bit like closing down all of the emotional tabs on my browser. Because if I shut the computer down and all of those tabs are still open, I'm going to come back in the morning and the computer's red hot, the fan is going because it didn't go to sleep. Because it couldn't, because there were too many tabs active and open. I think it's the same way with sleep as well. So try to think about doing that. So just vomit out all of your concerns on the page, and if there's emotional-
Andrew Huberman:
I like that, because my 3:00 a.m. waking is often associated with me writing down the list of things that I forgot to do that I need to do.
Matthew Walker:
Yeah.
Andrew Huberman:
And once I eventually wake up from the later night, second half of the night's sleep, that stuff seems much more tractable and reasonable, but it sure would be great to get that stuff out of the way before sleep.
Matthew Walker:
Well, there's also something that I don't think people have spoken about a lot and I'd like to research it. Which is, difficulty and anxiety at night in the dark is not the same difficulty and anxiety in the light of day. And when we have those thoughts at night, it comes with a magnitude of rumination and catastrophization that is disproportionate to that which you would describe when you are awake. And I don't know what's going on about the brain and thought and emotion at the time. I've got a bunch of theories as to why. And that's why I like the idea of closing up, zipping up all of those different components. Just get them out on the page, and I at first thought, this just sounds like hooey, it sounds very Berkeley. It's kind of kumbaya, we all hold hands and walk home at the end of the day. But then the data started coming out, really good studies from good people, and they found that keeping one of those journals decreased the time it takes you to fall asleep by 50%. Five zero.
Andrew Huberman:
Amazing.
Matthew Walker:
It's well on par-
Andrew Huberman:
That's substantial.
Matthew Walker:
... with any pharmaceutical agent.
Andrew Huberman:
Oh, abso-
Matthew Walker:
Yeah.
Andrew Huberman:
I've long thought that the worries and concerns and ideas I have at 3, 4:00 a.m., I've learned to not place any stock in them.
Matthew Walker:
Yeah.
Andrew Huberman:
Because of something ... I'm glad that you might decide to eventually look at this in your laboratory because I feel like something is melted away or altered. I suspect it's in the regulation of the autonomic nervous system, that it makes sense why a concern at 3, 4:00 a.m. ought to evoke more of a panic sense than a concern sense. And certainly that's my experience although not ... I'm fortunate to not suffer from full-blown panic attacks, but everything seems worse at 3, 4:00 a.m.-
Matthew Walker:
Yeah.
Andrew Huberman:
... provided you're awake.
Matthew Walker:
And we need to sort of look into that, because if you look at suicide rates around the 24-hour clock phase, disproportionately higher rates in those middle night hours. So, now I don't know if that's positive or not, but something ... It could just be that that's the time when we're mostly lonely, and we're by ourselves, and that's the reason. So it's got nothing to do with sleep or the nighttime. I don't know. So that's the third thing. I think the fourth sort of little tip I would give that's unconventional is, remove all clock faces from your bedroom.
Andrew Huberman:
Including your phone.
Matthew Walker:
Including your phone, and resist checking it. Now I know, and I can speak about the phone too. That genie of technology is out the bottle, and it's not going back in anytime soon. So we've got to think, as scientist and sleep scientist, as to what we do with phones in the bedroom.
Andrew Huberman:
Years ago, I was a counselor at a summer camp. I worked with at-risk kids, and there was a phrase that comes to mind here. It's "Be a channel, not a dam." Because when you try and dam certain kinds of behavior, physically dam certain kinds of behavior, not morally dam — that too — it just creeps over the edge and you get a waterfall, so it has-
Matthew Walker:
Yeah, so-
Andrew Huberman:
... to be channeled. The phone and devices have to be worked with and negotiated, not eliminated.
Matthew Walker:
That's right. And think about those mindfully too. But clock faces, remove all of those, because if you are having a tough night, knowing that it's 3:22 in the morning, or it's 4:48 in the morning, does not help you in the slightest. And it's only going to make matters worse than better. So try to remove all clock faces. And I think that's one of those other tips that some people have found helpful. But those would be, sort of, some slightly unconventional, I guess, more than your stock fair of "here are the five tips for sleep hygiene tonight."
Andrew Huberman:
Those are terrific sleep tips and several of which, if not all of which, I'm going to incorporate. Matt, this has been an amazing deep dive on sleep and its positive and negative regulators that's featured-
Matthew Walker:
I hope it hasn't been too long and I don't-
Andrew Huberman:
No, this has been great.
Matthew Walker:
Please cut it down, shorten it to five minutes of meaningful stuff that I offer.
Andrew Huberman:
Absolutely not. Absolutely not ever. It is chock-a-block full of valuable takeaways. It's been tremendously fun for me to dissect out this incredible aspect of our lives that we call sleep with a fellow scientist and a fellow public educator. I want to say several things. First of all, we should say where people can find you, although it shouldn't be that difficult these days. You're very present on the Internet-
Matthew Walker:
Unfortunately.
Andrew Huberman:
... for better or for worse. I think it's wonderful that you're out there. Look, it's a public health service that you're doing. No one requires you or any other scientist to get out and share this information. My sense of you, knowing you a bit, and from following your work very closely — both your scientific work in detail and your public-facing educational work — is that you very much want the best for people. And it's an interesting thing as a scientist or a clinician to know that certain answers exist. That we don't have all the answers, but that there is a better path, there are better ways, and people can benefit in a myriad of ways. So for that, because I know that to be very genuine in you, you want the best for people and you're offering tremendous advice and considerations, and people can take it or leave it — that's the way I view it. I also want to thank you for taking the time out of your day to sit with me here and-
Matthew Walker:
Not at all.
Andrew Huberman:
... have this discussion.
Matthew Walker:
Not at all. It's a privilege, it's a delight. You and I, I think we're kind in lots of ways, and I take you as a shining example of how you can effectively connect with the public. And I know that we've had our conversations before we ever sat down together about how to think about communicating with the public, and the pros and cons of that. And I've just loved your opinions. I've been drinking it all in. And then, I think the third thing I'd like to say is, thank you for being such an incredible sleep ambassador. The series that you've released on sleep, the way that you speak about sleep, the way that you moderate and have championed sleep. It is remarkable. So thank you for just being a brother in arms in that way.
Andrew Huberman:
We are, and thanks for those words. Ninety-nine percent of what I discussed there was the work of you and your colleagues in the sleep field, so proper acknowledgement — but thank you. Where can people learn more about what you're doing currently and what's coming next? You're on Twitter.
Matthew Walker:
I am on Twitter. I typically tweet-
Andrew Huberman:
As the sleep diplomat.
Matthew Walker:
No, it's just sleep diplomat.
Andrew Huberman:
Sleep diplomat on Twitter,
Matthew Walker:
Sleep diplomat on Twitter, sleepdiplomat.com website. If you want to learn more about the science that we do here, it's humansleepscience.com. It's the Center for Human Sleep Science. You can pick up a copy of the book, if you want. It's called "Why We Sleep." If you're curious about sleep, that's one path to take and it's my view-
Andrew Huberman:
Is there another book, someday in the future.
Matthew Walker:
I think there may. Yeah.
Andrew Huberman:
Uh-huh. Great, great.
Matthew Walker:
I think there-
Andrew Huberman:
Many, many millions of people will be very happy to hear that.
Matthew Walker:
I think it's starting to take hold, and then as we discussed, I am more than kicking around the idea of a short-form podcast rather than a long-form. Not long-form because I don't have the mental capacity or the interviewing capability that someone like you has. So it will probably just be monologue, short-form. So if there is some interest, I'll probably do that as well. So those are the ways that people can find me. But overall, if you are interested in sleep, just listen to Andrew. That's the best thing I can tell people.
Andrew Huberman:
All right, now we're batting back and forth the vector of action, so to speak. But I do hope you'll start a podcast, however brief or lengthy these episodes turn out to be, because I do believe that's a great venue to get information out into the world. And we don't just want to hear more from Matt Walker. I speak for many people. We need to. The work you're doing is both influential, but more importantly, it is important work. It has the impact that's needed, especially in this day and age, where science and medicine, public health and the issues of the world, et cetera, are really converging. So I know I speak on behalf of a tremendous number of people when I just say thank you for doing the work you do and for being you, and thanks for being a good friend.
Matthew Walker:
Likewise, too. And by the way, I just going to note that it was nice that the two of us both got the Johnny Cash memo about how to dress today. It seems as though we're both kind of ... We got that same memo, which will mean nothing to people who are listening, but if you are watching the video, you'll probably see what I mean. Andrew, thank you for taking this time. Thank you so much.
Andrew Huberman:
Thanks so much, Matt. Thank you for joining me for my discussion with Dr. Matt Walker. Please also check out his podcast, the Matt Walker podcast. A link to that podcast can be found in the show notes. If you're enjoying this podcast, please subscribe to us on YouTube. On YouTube, you can also leave us comments and suggestions for future episodes and guests in the comment section. As well, please subscribe to us on Apple and on Spotify, and at Apple, you can leave us up to a five-star review. Please also check out our sponsors, mentioned at the beginning of this episode. Links to those sponsors can be found in the show notes. And last, but certainly not least, thank you for your interest in science.
Join 500,000+ subscribers to get regular emails on neuroscience, health, and science-related tools from Dr. Andrew Huberman.
You'll also get Andrew's exclusive Daily Blueprint. In it, Andrew shares his daily routine. He also shares practical tools and protocols that you can use to stay productive and maximize your health.
By submitting your email to subscribe, you agree to Scicomm Media's Privacy Policy