The Peter Attia Drive - #126 – Matthew Walker, Ph.D.： Sleep and immune function, chronotypes, hygiene tips, and addressing questions about his book

00:00:00.000 Hey, everyone. Welcome to the drive podcast. I'm your host, Peter Atiyah. This podcast,

00:00:15.480 my website, and my weekly newsletter all focus on the goal of translating the science of longevity

00:00:19.800 into something accessible for everyone. Our goal is to provide the best content in health and

00:00:24.760 wellness, full stop. And we've assembled a great team of analysts to make this happen.

00:00:28.880 If you enjoy this podcast, we've created a membership program that brings you far more

00:00:33.280 in-depth content. If you want to take your knowledge of the space to the next level at

00:00:37.320 the end of this episode, I'll explain what those benefits are. Or if you want to learn more now,

00:00:41.720 head over to peteratiyahmd.com forward slash subscribe. Now, without further delay, here's

00:00:48.080 today's episode. I guess this week is Matthew Walker. Many of you are probably familiar with

00:00:54.060 Matthew as back in, I think, April 2019 or so. We released a three-part series on sleep and then

00:01:00.760 did a two-part follow-up of AMAs in June and October of that year. So if you're a listener

00:01:06.640 of this podcast, you've spent a lot of time listening to Matthew Walker. That said, the

00:01:10.840 demand to have Matthew back on has been overwhelming. And there's always a lot of stuff to talk about

00:01:15.600 with Matthew. He's such an eloquent speaker on the topic of sleep. We just continue to get a lot

00:01:19.920 of questions on this. And so we figured there was no better time and place to do this than here and

00:01:24.920 now. So the interview was actually set up to be only an AMA, but one, we went very long. And also

00:01:32.600 after the fact, Matthew and I felt that some of this topic really ought to be put out there for a

00:01:37.660 more general audience. So what you're about to hear is the, obviously the regular episode for all

00:01:43.020 listeners, and it's going to be followed by an AMA for subscribers only. So in this episode,

00:01:48.720 we talk about sleep in particular in the age of COVID. What have we learned about sleep through

00:01:53.180 the lens of this pandemic? We also talk about the sleep foundations and the tips that we can use to

00:01:58.920 trigger better sleep. We dive a lot deeper into the sleep chronotypes and even discuss a test that

00:02:05.480 allows you to identify yours. We talked a little brief discussion about children and nightmares and

00:02:11.080 night terrors and things like that. And then we end the discussion talking about something that I think

00:02:15.480 a lot of people have been asking about, which are some of the errors that readers have pointed out

00:02:20.960 in Matthew's book, Why We Sleep. So as a quick refresher on Matthew, for those of you who might

00:02:26.440 not know, he's a professor of neuroscience and psychology at UC Berkeley, and he's the founder and

00:02:30.900 director of the Center for Human Sleep Science. He's earned his degree in neuroscience from Nottingham

00:02:36.360 University in the UK and his PhD in neurophysiology, also in the UK from Medical Research Council.

00:02:42.100 He's the author, as I said, of the international bestseller, Why We Sleep. And he's just an overall

00:02:47.820 awesome guy that I could talk to about sleep and race cars equally. So without further delay,

00:02:53.680 please enjoy my conversation with Matthew Walker.

00:03:01.620 Well, Matt, it's awesome to have you back.

00:03:04.960 It's a delight to be here.

00:03:06.820 Yeah, yeah. I mean, we talk so often that to put some structure to a discussion in this context,

00:03:11.660 will be fun. But there's always so much that we're talking about. At least 50% of it involves sleep.

00:03:17.420 The other 50% seems to involve our other favorite mutual discussion, which we'll try to not talk

00:03:22.320 about today.

00:03:23.220 But can we talk about motorsport to begin with? No, I'm just kidding.

00:03:27.400 Sleep it will be exclusively.

00:03:29.240 Sleep it shall be. Although I do love the book in the background behind you there. Don't think I can't

00:03:33.800 see that. Since we last spoke, which was almost a year ago, obviously the elephant in the room here

00:03:42.160 is COVID. And I'm just wondering if you have seen or have any understanding of the impact that

00:03:52.640 anything to do with COVID has had on sleep. In other words, has the world in which you pay

00:03:58.180 all of your attention? Has it been impacted by this?

00:04:02.020 It does seem to have been impacted. And we actually have some decent data. The way it's

00:04:08.060 been impacted is probably at three levels. The amount that people are sleeping, the timing

00:04:14.440 of their sleep, and then finally dreaming. Firstly, in terms of amount, probably one of the earliest

00:04:22.300 reports that I saw was from a sleep tracking company that released data of, I think, about

00:04:27.260 68,000 individuals here in the United States. And what they reported was that total sleep

00:04:34.600 time had increased across the entire country by 20%, they said, which I struggled with a little

00:04:41.880 bit. But furthermore, they broke it down on the basis of each state. And what they said was

00:04:47.520 that there was no single state that hadn't increased its total sleep time by 10%. Now,

00:04:54.480 those are just reported data from a sleep tracking company, and I didn't get the chance to actually

00:05:00.220 look at them. But then two recent peer-reviewed papers have actually come out in the same journal,

00:05:06.240 and it's a good journal, Current Biology, that seem to provide support for this. They downscale the

00:05:12.840 amount, but they add nuance. So the first paper was a paper published from a group in Europe looking

00:05:18.640 at a collection of folks throughout Europe. Firstly, what they reported was, once again,

00:05:24.780 total amount of sleep had increased by 15 minutes. They also found that something called social jet

00:05:31.020 lag had decreased. What is social jet lag? Social jet lag is the difference between the sleep that

00:05:37.500 you're getting during the week versus the weekend. So what typically happens is we don't get enough

00:05:41.840 sleep during the week, and then we binge at the weekend. And then come Sunday night, because we've

00:05:46.940 woken up, you know, 10 a.m., 11 a.m. on Saturday and Sunday, we have to drag our biological clock back

00:05:53.900 by two or three hours. It's brutal on the clock. It's like flying back and forth from San Francisco to

00:05:59.080 New York, you know, every weekend. What they found was that people were going to bed and waking up at

00:06:06.540 much more similar times during the week than they were the weekend. And therefore, that difference,

00:06:12.700 that social jet lag difference, had decreased, which is beneficial.

00:06:16.480 It almost feels like it vanished when you think back to those early days, especially of

00:06:20.660 March and April, where, you know, everybody would joke about it, which is, what day is it? If you

00:06:27.060 weren't really thinking about it, you didn't know the difference between a Wednesday and a Saturday.

00:06:31.380 They all sort of seemed like the same day, which of course would theoretically squash social jet lag to

00:06:37.680 zero. Yeah, it bleeds or it sort of softens the boundaries between those two distinct parts of

00:06:44.800 that seven-day week. And of course, I think what was happening there was that people didn't have to

00:06:50.920 wake up as early to go to work. So commutes had been removed for the large part. And then having to

00:06:59.420 wake up even earlier than that, so that you could get, you know, your kids ready for school and get

00:07:04.240 them away. Those two forces, which typically make people wake up earlier than they otherwise would

00:07:10.000 want to. And then of course, people were not coming home as late because they didn't have the back-end

00:07:16.020 commute. And those two things normally squeeze the amount of time that we have for sleep, like vice

00:07:22.020 grips in the middle of the night. In other words, modernity with, you know, getting your kids to school

00:07:26.880 and commute times has essentially compressed people having to leave the house earlier and

00:07:33.680 they come home later. That's been removed. So the opportunity for sleep has actually expanded

00:07:39.260 during the weekday. And that's what they had found too. What was interesting though, in that European

00:07:45.180 study is that overall, and I think this gets to the nuance, the quality of sleep that people reported,

00:07:51.000 some of them were reporting worse quality sleep. And I think we'll get to that too.

00:07:56.940 Sorry, just to ask more about that, Matt, does that mean people felt subjectively more fatigued

00:08:03.640 or they felt that their sleep was more interrupted and that they were waking up more or just having

00:08:09.660 bad dreams? Like what did it mean by subjectively the quality seemed worse?

00:08:14.620 We don't know which of those three or other options it was. The question is simply,

00:08:19.740 what do you feel as though the quality of your sleep was in terms of a question? So this is just

00:08:25.400 subjectively looking back retrospectively in the rear view mirror last night, what was the quality of

00:08:30.660 your sleep? And I think what you're touching on is something really important because I believe that

00:08:36.600 when we actually finally do look at the data, it's going to be more complex and there's going to be at

00:08:43.260 least two different clouds of data that we see. There's going to be one proportion, one cloud of people

00:08:49.340 people for whom sleep actually has increased and the quality of their sleep is actually better.

00:08:55.220 I think what we'll find is that there's another cloud of data for whom sleep has actually become

00:09:00.360 shorter and worse. Why is that? Because they've either lost their job, they may lose their job, or just

00:09:07.420 in general, they have a great deal of anxiety, of course, because of COVID. So I don't think the story is

00:09:14.060 quite as simple as we're seeing right now, but that's sort of what we've seen in terms of the

00:09:20.220 amount of sleep. It does seem to have changed overall, but the timing of sleep has also changed.

00:09:27.160 If you look at these studies, on average, people were actually going to bed about 30 minutes later and

00:09:33.820 waking up about 50 minutes later the next morning. So I think this is finally the revenge of the night

00:09:42.020 owls. That because modernity favors morning larks, or what we call morning types, it's designed for

00:09:52.200 morning types and it has, we chastise the night owls as being lazy, well why don't you get to bed

00:09:59.160 earlier? Because biologically they can't, that's their chronotype. And finally now with COVID and the

00:10:05.900 relaxation of these work rules, night owls are being allowed to be much more owl-like and sleeping

00:10:13.620 in harmony with their chronotype rather than trying to work against their chronotype. So not just the

00:10:21.120 amount of sleep has changed, but the timing of when we're sleeping has also changed as well.

00:10:27.920 Is that seasonally adjusted, Matt? So as we go from winter to spring to summer in the northern hemisphere,

00:10:33.720 do we tend to see movement of bedtimes and wake-up times? I know with my kids, for example, we

00:10:39.800 definitely see it because of the light. It's easier to get them to bed earlier in the winter than it is

00:10:44.460 the summer. But is that true with adults also? It is true with adults. You do see that seasonal shift

00:10:51.020 as long as there's sufficient exposure to light. But if you're an individual who essentially you wake up

00:10:59.620 under your electric lights in the morning, then you go to work and it's constantly lit in that work

00:11:05.980 environment, then you drive home and you're not getting much daylight exposure, the influence with

00:11:12.320 which the outside world impacts your sleep schedule is therefore diluted and lessened. So you have to be

00:11:18.780 exposed. And your kids are a great example of that because they're outdoors playing all day.

00:11:25.280 At school, they're given outdoor activities. So you will see some of that too. But in general,

00:11:31.000 yes, you're absolutely right. That will shift on the basis of seasons depending also on your

00:11:36.860 latitude. I mean, you know, if you're living close to the equator, you won't see that get expanded the

00:11:42.720 further you go north and south. But these findings are basically above and beyond what you would see

00:11:47.500 seasonally is what I'm understanding. That's right. So another thing I want to ask you about on

00:11:52.200 COVID. When I think back to March in particular, but probably into April, I was having dreams that

00:12:02.000 were unusual even by my standards. I'm one of those people that tends to have vivid dreams most nights.

00:12:08.460 I can usually even remember them at least as soon as I wake up. But I think I would say they were even

00:12:15.140 more heightened. Has the literature commented on that at all?

00:12:18.380 It has. And I think there is too much subjective inbound that we've had that confirms what you're

00:12:27.520 describing too. And I believe when more data comes out, that's what will happen. We will hear and find

00:12:34.160 that dreaming has increased as a consequence of COVID. Why is that? I think there's probably at least two

00:12:41.340 different reasons. First, it comes back to timing of sleep. And maybe we can at some point sort of go over

00:12:48.000 the standard sleep schedule. Well, let's do it now. I think we should assume that maybe people haven't

00:12:53.620 necessarily listened to our first episode where you described the cycles. So we'll let people fast

00:12:58.600 forward a few minutes. But for those who need a primer, maybe go through stages one through four

00:13:03.420 and REM and the cyclicality of it. So humans and most mammals, at least land-dwelling mammals,

00:13:10.140 have two principal types of sleep. On the one hand, we have non-rapid eye movement sleep or

00:13:15.960 non-REM sleep for short. And non-REM sleep has been further subdivided into four separate stages

00:13:22.040 that are unimaginatively called stages one through four, because we're a creative bunch,

00:13:27.760 as in sleep research. But on the other hand, oh, by the way, for stages one through four of deep sleep,

00:13:34.480 they increase in their depth of sleep. So stages three and four are those really deep stages of

00:13:40.680 non-REM sleep. And on your sleep trackers, that will be described as deep sleep. Stages one and

00:13:46.140 two of non-REM sleep are typically what you classify as light non-REM sleep. On the other hand, we have

00:13:52.980 rapid eye movement sleep or REM sleep. And REM sleep is typically associated with dreaming. Now,

00:14:00.200 it turns out that we dream in many different stages of sleep, but the dreams that we all think of that

00:14:06.260 are narrative, emotionally filled, bizarre, hallucinogenic, those principally come from

00:14:11.960 rapid eye movement sleep. So we have these two types of sleep, non-REM and REM. And they will play

00:14:18.200 out in this beautiful battle for brain domination throughout the night. And that cerebral war is going

00:14:25.320 to be won and lost every 90 minutes and then replayed every 90 minutes to create the standard cycling

00:14:33.160 90 minute architecture of the sleep cycle that then just repeats throughout the night.

00:14:39.280 But coming onto your question of, and this is why it's relevant, the ratio of non-REM to REM within

00:14:45.680 each 90 minute cycle as we move across the night changes. And what I mean by this is in the first half

00:14:54.040 of the night, the majority of those 90 minute cycles are going to be comprised of lots of deep non-REM

00:15:00.080 sleep stages, stages three and four non-REM sleep and very little REM sleep. But as you push through

00:15:05.260 to the second half of the night, now the brain has a shift in its taste preference for what it wants to

00:15:13.820 feast on at the finger buffet of sleep stages. And instead now it actually wants to consume much more

00:15:21.640 rapid eye movement sleep and it's lost its appetite for deep non-REM sleep. So in the second half of the

00:15:27.600 night, the majority of those 90 minute cycles are comprised much more of rapid eye movement sleep.

00:15:33.940 And this has implications in lots of different ways. And one of them comes back to COVID. If people

00:15:41.480 are naturally now on average going to bed a little bit later, but certainly sleeping in significantly

00:15:48.960 later in the morning, what's going on? It means that they're now sleeping further into that REM sleep

00:15:56.780 rich window of the night. And we spoke about REM sleep being associated with the most vivid hallucinogenic

00:16:05.260 narrative dreams that we experience. In other words, I think the first reason that people are dreaming

00:16:11.220 more during COVID is because they're sleeping in a little bit later and therefore just probabilistically

00:16:18.240 from a statistical point of view, you're much more likely to be in a dream state and have significantly

00:16:25.400 more dream cycles as a consequence. So I think that's the first reason.

00:16:31.280 That makes sense. Population wise. I know for me, that's almost not going to be the answer because

00:16:37.840 my bedtime shifted paradoxically earlier on both ends. In other words, I started going to bed earlier

00:16:44.500 and waking up even earlier, which may have just been simply anxiety. But I mean, there was a period

00:16:50.320 there where I could not sleep past 5am. And part of that is I have a, you know, a lot of patients

00:16:56.380 and people, you know, my team is mostly on the East coast. So getting up at four 30, which is seven 30

00:17:02.220 their time just seemed like the normal way to go about business. But I almost wonder like, I don't

00:17:08.960 know enough about dreaming. And I apologize because I bet we did talk about this two years ago when we

00:17:13.720 did our first go around or which was probably nearly two years ago. But is there a component of this

00:17:18.920 that is also just a way to purge negative emotion or, or, or stress in some way? And is that

00:17:24.900 potentially also a part of this equation?

00:17:28.580 Oh, that's why I love our time together and chatting. That's exactly the second reason.

00:17:32.800 And that's the reason why, even though your sleep schedule didn't change, you still were dreaming more

00:17:38.460 because one of the principal functions of rapid eye movement sleep that we now have understood

00:17:44.760 is that it seems to provide a form of overnight therapy. That REM sleep is emotional first aid.

00:17:54.320 And it's during REM sleep that we take the emotional concerns. And even sometimes in extreme

00:18:00.900 examples, some of the traumatic events that we've been experiencing and REM sleep almost acts like a

00:18:07.800 nocturnal soothing balm. And it tries to take the sharp edges off those difficult experiences so that

00:18:15.440 when you come back the next day, you don't feel necessarily as emotional as you did about those

00:18:22.980 events the day before. So in other words, it's not time that heals all wounds. It's actually time

00:18:30.400 during REM sleep and dreaming that provides this emotional convalescence. And what we also know

00:18:37.660 is that sleep is responsive. It's not a fixed system necessarily. It's what we call homeostatically

00:18:45.680 regulated, but also homeostatically reflexive. What I mean by that word salad is that if something

00:18:55.180 happens during the day that requires sleep and a specific stage of sleep, then sleep will respond by

00:19:03.340 giving the brain a little bit more of that sleep to deal with whatever happened during the day. And a good

00:19:10.440 example here is learning and memory. We've done some of these studies where we'll have people all of a sudden

00:19:15.000 go into three or four hours of really intensive textbook learning. And the next night, we see an increase

00:19:22.500 above and beyond their baseline in the amount of sleep that's needed for the fixation of those memories.

00:19:28.700 If we do a skill learning task, like a sports task, it's a different type of sleep that's needed.

00:19:34.180 We see a homeostatic response from sleep that following night to try and lock in those new

00:19:39.640 skilled memories. And it's the same for emotional experiences. If you've upped your quotient of

00:19:46.860 emotionally difficult experiences, sleep can respond with greater amounts of rapid eye movement,

00:19:53.820 sleep or dream sleep to try to help solve those emotional challenges. So I think that's the second

00:20:00.940 reason why even if you're not changing your sleep, you can still be dreaming more. Does that sort of help

00:20:07.360 reconcile that? It does. And it also brings to the front of my mind another issue, which I know I'm not

00:20:14.740 going to be the only one admitting this, which is I found myself more reflexively reaching for a glass of

00:20:23.220 wine during. And frankly, I think probably more so to this day than, than I would have historically,

00:20:30.100 right? I think I probably consume four or five glasses of wine a week right now, Matt, whereas

00:20:35.000 historically it would have been four or five glasses a month. And I would say during the height of sort of

00:20:41.060 my anxiety around some of the aspects of COVID, I was probably drinking a glass to sometimes two every

00:20:47.720 single night. And I just know from my own tracking of sleep, which is primarily based on the aura,

00:20:54.680 which I need to disclose. I'm both an investor in aura and an advisor to them. So please understand

00:21:00.860 and take anything I say about aura with a grain of salt. But I think the aura provides a very good

00:21:06.620 stream of data on metrics around sleep, particularly time in bed or time sleeping and the variables that

00:21:14.880 drive some of the interpretations of stages, namely heart rate, heart rate, variability,

00:21:19.300 temperature, and respiratory rate. And without exception, Matt, every time I drink alcohol,

00:21:26.240 especially if it's going to hit two drinks in a night, every one of those variables moves in the

00:21:31.280 wrong direction. Yeah. Heart rate variability gets crushed. Resting heart rate goes up by 10,

00:21:37.080 if not 20, 25%. Respiratory rate is up two breaths per minute. And core body temperature is up half a

00:21:44.600 degree. That coupled with what I remember of our first discussion, which is alcohol disproportionately

00:21:50.340 impacts negatively REM, creates kind of an ironic situation, which is I'm probably numbing a little

00:21:59.600 bit of my anxiety with alcohol, which is impairing my ability to have REM sleep, which is paradoxically

00:22:06.520 exactly the medicine I need. Yeah. That's exactly the interpretation I would offer. And then

00:22:12.480 it becomes a self-fulfilling prophecy. So you're right. Firstly, you know, one of the problems with

00:22:17.200 alcohol is that it's a sedative and what most people think of as the benefit of a nightcap for

00:22:23.240 your sleep is not really a benefit. Yes, you quote unquote, well, I'll say you lose consciousness

00:22:30.140 faster when you put yourself in bed. It's hard to say that you're going into naturalistic sleep.

00:22:35.880 Your sleep becomes more fragmented because the alcohol will actually stimulate the fight or flight

00:22:42.100 branch of the nervous system. It also releases wake promoting chemicals. That's why you wake up more

00:22:47.620 frequently. And then the third part is alcohol will decrease the amount of REM sleep that you get,

00:22:53.320 particularly in the middle towards the later hours. And of course, as you said there, what's happening

00:22:59.060 very understandably is that you're trying to find a way to manage your anxiety. And by the way, as a parent,

00:23:05.440 I don't even know how much of it is about anxiety around coronavirus as it is anxiety around kids or

00:23:11.480 other things that are the response to coronavirus. As you said, your job, again, your kids are going

00:23:17.420 crazy being home in front of Zoom 24 seven. I mean, it's sometimes you don't even realize what it is

00:23:23.900 that you're reaching for that drink for. But you know, my wife and I will constantly joke, like,

00:23:28.380 is it five o'clock yet? Is it five o'clock yet? Like,

00:23:30.240 it's going to be five o'clock somewhere in the world, right? Exactly. And it's weird because

00:23:35.720 neither of us are the type of people who, like, we could easily go two months without having a drink

00:23:40.500 and not notice. So there's just been something particularly challenging about the last six

00:23:46.820 months that I think have led to just more frequently that type of behavior. And I say that

00:23:52.920 not trying to be too judgmental of myself just as an observation, but I'm guessing I'm not the only

00:23:57.960 one that's gone through this. Yeah. And I have a huge degree of sympathy for that situation.

00:24:04.120 You know, everyone has this cabin fever and then you put together three kids. And as you mentioned

00:24:09.720 that too, there is this self-fulfilling prophecy where if you're downscaling the amount of REM sleep

00:24:16.360 that you get at night with alcohol, you may not be getting the necessary emotional benefit that you need

00:24:22.360 to de-risk the emotional experiences and the anxiety that's building up. So then the next day,

00:24:28.940 the anxiety is not dealt with. It's overlaid and then added to by the following day. And so you're

00:24:35.880 feeling even more anxious the following night. And then you find yourself reaching for another glass

00:24:39.880 of wine, which then takes away the REM sleep once again. And so goes the spiral of escalation.

00:24:45.740 So again, it's, I'm not trying to be at all judgmental. I'm just a scientist. I'm not here

00:24:52.120 to be puritanical about life. Life is to be lived to a degree. And I don't want to tell anyone how to

00:24:57.000 live their life. I think what we're just trying to do here is describe the consequences of these

00:25:02.000 things on sleep so that people can be informed and make their own life choices as to what they

00:25:07.260 would like to do. Now, Matt, one of the things I have the luxury of doing is looking at the

00:25:11.780 aura data of all of my patients who wear an aura ring and who are in our portal. And I will say this

00:25:18.900 when I couple that ability with the ability to speak with them, I've noticed not all people are

00:25:25.380 the same. I really think there are some people for whom alcohol does not seem to have that negative

00:25:30.160 effect. Like for example, in me, one drink doesn't really seem to have a negative effect provided it's

00:25:36.240 around 6 p.m. And I'm going to bed at 9 p.m. But two drinks does have the effect. I've seen patients

00:25:43.600 for whom two drinks has no effect. And when I've had this discussion with them about, hey, you know,

00:25:49.000 really from a, from a scientific standpoint, it would be great if you drank less for reasons X, Y,

00:25:54.060 and Z. Their counter argument is, well, the relief that comes from the alcohol, the true release of

00:26:01.140 stress may justify it. And I guess I've come to realize that every one of these things has to be taken

00:26:05.940 on a case-by-case basis. But when it pertains to sleep, would you agree with potentially my

00:26:11.280 observation that people are not all identical in this regard as well?

00:26:15.560 A hundred percent. And I think, in fact, if you do some of these genetic testing

00:26:19.620 companies out there, they will describe to you whether you're a fast metabolizer of alcohol.

00:26:26.420 You know, there's a genetic basis to it. It's the same with caffeine. So you're completely right that

00:26:31.720 there is an average adult that we will typically speak about when we speak about, you know, large

00:26:37.440 scale, be it epidemiological studies, which we're both not necessarily a fan of, or even any studies.

00:26:43.920 And we offer recommendations. This is for your average adult. But everyone is unique in that sense.

00:26:50.060 And so there absolutely will be some people, one side of the distribution, who don't show that

00:26:55.340 response to alcohol. Because maybe they're a fast metabolizer and they have a drink at 7 p.m.

00:27:00.880 And by the time it's their turn on their sleep schedule to be in the position to start receiving and

00:27:08.640 chowing down on REM sleep from that buffet, most of the aldehydes and the ketones are now, you know,

00:27:16.260 long gone and not impacting their sleep, which are the metabolic consequences of sort of alcohol

00:27:21.320 degradation. And those are the things that seem to impact your REM sleep. So I believe that there

00:27:26.680 are definitely that cluster of people. And then equally, there are folks on the other side who can

00:27:32.180 have, you know, a glass of wine with a late lunch, and it still could crush their REM sleep.

00:27:39.240 I think the take home here is that we should all go to the pub in the morning. And that way,

00:27:43.980 the alcohol is out of our system by the evening. But let's not make that statement.

00:27:48.760 Just stop drinking by noon and we think you're going to be fine. So the other thing on COVID,

00:27:54.860 if we can spend just another minute on this, is I really want to understand from you what you think

00:28:00.500 the richest sort of data are with respect to sleep and either innate or probably adaptive immunity.

00:28:08.900 So one of the things that I've thought a lot about in the past few months is the impact of nutrition,

00:28:15.460 the impact of cortisol. In fact, we wrote a piece on that a few weeks ago, and there was actually

00:28:21.860 some very elegant experimentation looking at the impact of hypercortisolemia on the adaptive immune

00:28:28.220 response to five common cold viruses. I think it's common sense that sleep improves immune function,

00:28:36.840 but can we say anything specifically or more to the point, have we learned anything new with respect

00:28:41.520 to COVID on that front? I don't think we have the data quite yet on sleep and COVID. And I should

00:28:48.640 also say, you know, for this conversation in general, I really want to be mindful of feeding into

00:28:54.480 the general anxiety that's out there in the world that we're experiencing right now. And we'll speak

00:29:00.940 probably a lot about the consequences and the impact of insufficient sleep. So I just kind of want

00:29:06.060 to maybe throw a sticky on the wall and just kind of be upfront and be sympathetic to that before I

00:29:11.700 speak about what we do know. So we don't have any data yet directly regarding a lack of sleep and your

00:29:18.040 vulnerability to COVID, or if you are infected, the severity of progression. But we definitely know

00:29:26.380 firstly, that there is a very intimate association between your sleep health and your immune health.

00:29:32.000 We know from a study, gosh, one of the classic studies in archives of general medicine,

00:29:37.000 that individuals who report getting less than seven hours of sleep a night are almost three times more

00:29:43.880 likely to become infected by the rhino virus, which is the common cold relative to those who sleep more

00:29:50.700 than seven hours. So there is some vulnerability risk there. We know from a study, perhaps that's relevant

00:29:57.280 here. There was a study, I think it was in over 70,000 women, and it was a prospective study, so a little bit

00:30:04.680 better. And what they found is that women who were sleeping five hours or less were 70% more likely to go on

00:30:12.680 to develop pneumonia, which is a component risk that we understand in the equation of COVID. We also know from

00:30:21.700 animal studies that this is just not associational, it's actually causal. There was a fascinating study

00:30:27.740 where they were selectively depriving mice of sleep, and particularly REM sleep. And what they found is

00:30:34.200 that when they were exposed to malaria, the mice that were deprived of sleep were far more likely to die

00:30:40.840 as a consequence of that malaria exposure than the well-rested mice. So there is something about survival

00:30:47.440 risk there. And then we also know from classic studies, and this to me is perhaps the most

00:30:52.960 interesting one for COVID, that if you're not getting sufficient sleep in the week before you get your

00:30:58.780 flu shot, you end up producing less than 50% of the normal antibody response, therefore rendering that

00:31:06.060 flu vaccination significantly less effective. And we have no idea, of course, right now what the ultimate

00:31:14.440 vaccine is going to be, and therefore we have no idea about its relationship with sleep. However,

00:31:19.660 I think we do need to be prospectively looking at this when the vaccine comes out and tracking sleep,

00:31:25.340 because if your sleep in the days before you get your COVID vaccination determines whether or not

00:31:32.320 it's successful, we should know that, and we should be smart about it. What about a future where,

00:31:38.960 you know, you're upcoming for your appointment, and your sleep hasn't been great? And I say, look,

00:31:44.440 doesn't look like you've been getting perfect sleep. Let's just wait a few more days. I've booked

00:31:49.240 three more appointments next week. Select the one that you want, and we'll look for a future when

00:31:55.400 you're better slept, and we'll have a significantly higher, you know, success rate. I don't think that's

00:32:01.100 coming anytime soon, but I'm just sort of, you know, this is beer and nuts talk, since we're speaking

00:32:06.640 about alcohol, as to what could be possible. But I do feel as though we need to understand what is the

00:32:13.280 relationship between sleep and your immune response to that vaccine once that vaccine is there. So does

00:32:20.980 that paint a little bit of, I know it's nascent because we don't really have direct data yet,

00:32:25.680 but I think there's enough data to tell us we should certainly be studying it.

00:32:30.860 Well, first of all, the sleep immunization, I mean, I've never heard that before. That's super cool.

00:32:35.880 I'm literally scribbling a bunch of things down as you say that, because that's also the easiest

00:32:40.240 question to study. That's a really easy study to do. And the implications of that are profound.

00:32:45.400 You know, I get asked, I don't know, somewhere between 46 and 50 times a day, Peter, when's the

00:32:50.980 vaccine going to be ready? When's the vaccine going to be ready? What's your take on the vaccine? And

00:32:53.840 you know, my take on this is nuanced. I won't go into it now, but I think part of my take on COVID

00:32:59.380 vaccines is going to be, I think it's going to take a little longer than we think. I think it's going

00:33:05.460 to probably be not quite as safe as we think, meaning not something you just want to indiscriminately

00:33:12.460 give to everybody. In other words, I think it's like any vaccine is going to have some risk.

00:33:18.960 And it's certainly possible that at least some of those vaccines, potentially really efficacious

00:33:24.240 ones may come with side effects that are undesirable, and therefore we'll have to think about who gets

00:33:28.400 them. By the way, you've said plural there. Do you think there will be multiple vaccines on offer

00:33:33.700 and multiple vaccines of different types? I do. I mean, that's my intuition is I think

00:33:38.540 there will be, you know, several approaches. I think there probably will be specific ones that

00:33:43.860 are looking, you know, just at RNA. There'll be, you know, sort of surface ones. There'll be live.

00:33:48.340 I don't know if there'll be a live attenuated one. I mean, we're working, you know, when I say we,

00:33:52.460 like I'm having anything to do with it, but I mean, I think the scientific community is working their

00:33:57.420 way up from the lowest hanging fruit and going all the way up. And that says, of course,

00:34:02.560 nothing about monoclonal antibodies and other therapies. But I guess my bigger point is it's

00:34:08.320 not like on a certain day we will have enough vaccines to give everybody, right? That's certainly

00:34:15.660 not going to be the case. So even if you just think about this through a very, you know, U.S.-centric

00:34:21.580 view and say, well, you know, when will the United States have enough vaccines to appropriately cover

00:34:27.680 its citizens, that could be years away. And then it means, well, as you identify the populations

00:34:33.180 that are most at risk, which be presumably the people you want to vaccinate first, my guess is

00:34:38.560 healthcare providers would be atop that list, but also the people that are most vulnerable from a

00:34:44.160 underlying condition perspective, anything you can do that could increase the efficacy of that

00:34:49.100 vaccination by 50%. I mean, if it turns out that the SARS-CoV-2 virus has a behavior that is comparable

00:34:56.380 to influenza with respect to this property, I mean, that's, you know, you don't get many

00:35:02.080 opportunities to 2X something. Anyway, I just hope that somewhere out there, somebody in these phase

00:35:08.400 two studies, because we're just basically in the, you know, we're at the time of this recording,

00:35:12.840 at least the phase two studies are being launched. It would really be wonderful if somebody could do a

00:35:17.460 subgroup analysis on how many soluble antibodies are you getting? You know, what type of T cell response

00:35:23.420 are you getting? And can we at least retrospectively look at the sleep quality going into that, if not

00:35:28.940 prospectively? And if anybody out there are doing those studies, it's very simple to do, you know,

00:35:34.880 we could do it with sleep trackers, but there's paper and pencil things that we could do. Feel free

00:35:39.580 to reach out to me and I'd love to sort of just provide some simple tools to do that. But what was

00:35:44.980 interesting in that immunization study, they looked at around, I think it was somewhere between,

00:35:50.380 you know, 11 to 14 days after the flu shot and they just saw this, you know, catastrophic reduction

00:35:56.460 in antibody response. But then they did something smart. They said, well, maybe it's just delayed.

00:36:01.760 Maybe they do get to that level and it just takes them a little longer. So sleep is just

00:36:06.420 slowing the true peak of that antibody response. But they kept assessing them, you know, 20 to 30 days

00:36:14.240 later and they just still saw nothing. So it really did seem to be that it had just blunted it and

00:36:19.420 you'd lost that chance and it wasn't coming back.

00:36:23.120 Wow. Yeah. That's, that's, if I learned nothing more in the next, whatever period of time we're

00:36:28.460 speaking, Matt, that was, uh, I got my money's worth on that little factoid.

00:36:32.620 And if you don't let anything more in the next coming time, then I have failed utterly miserably,

00:36:38.120 but there's my challenge.

00:36:40.300 Let's pivot away from COVID and kind of get back into some of the sleep foundation

00:36:44.220 stuff and some of the background questions. So, you know, a little less than a year ago,

00:36:48.840 we sat down and did an AMA. I don't remember how long we sat down. What I do remember is it was

00:36:54.580 at the end of a day of exhausting, but exhilarating racing in, uh, Bakersfield, California, we were both

00:37:03.040 beat. And I think we, we, we sort of cried uncle after about two hours. So we didn't get through all of

00:37:09.680 the questions. And I think that's part of the stuff I want to talk about today. Maybe we can

00:37:15.400 kind of go back and revisit some of the, you know, you gave us a pretty good primer on REM and non-REM

00:37:21.860 anything else you want to kind of go back and revisit before we jump into some of these other

00:37:27.640 things that you didn't already cover earlier today, just to get people kind of back to the

00:37:32.580 foundational level of the architecture of sleep.

00:37:35.420 I think we did a pretty good job. I know you spoke a little bit about that variability of response to

00:37:42.660 alcohol. And I just want to respect that variability. Once again, I said, you know,

00:37:47.600 that there is a 90 minute cycle on for the average adult, it seems to be about 90 minutes,

00:37:53.340 but that isn't necessarily consistent by the way. It's, you know, for some people it's shorter,

00:37:58.720 for some people it's longer and even for different species, it's very, very different. So

00:38:05.140 we humans, we have a 90 minute cycle on average, but you know, other species, birds, for example,

00:38:11.800 the sleep cycle from non-REM to REM can be maybe just two or three minutes, much shorter non-REM to

00:38:18.880 REM sleep cycle. So it's a very perplexing. We actually don't really understand what determines

00:38:27.000 the length of your sleep cycle across phylogeny, across different species. The best guesstimate

00:38:33.140 that we have right now is that it has something to do with the relative mass size of your brain.

00:38:38.840 It has a weaker correlation with body mass, but for the most part, the larger your brain mass,

00:38:44.980 the longer your sleep schedule and your relative brain mass relative to the rest of you, the organism.

00:38:52.240 But it's one of those strange things. By the way, we also don't know,

00:38:56.380 and I'm not really interested in saying, can I clarify anything? And I'm now just telling you

00:39:00.380 things that we don't understand, but I hope that's okay too, which is we have no idea why sleep plays

00:39:06.860 out in the way in which it does that I described, which is that you get most of your deep sleep in

00:39:12.900 the first half of the night and most of your REM sleep in the second half of the night.

00:39:18.040 What I mean by that is we understand what determines why you get your deep sleep early in the night,

00:39:26.120 which is largely determined by how long you've been awake and a little bit on the basis of your

00:39:31.020 circadian rhythm, whereas REM sleep is much strongly, much more powerfully determined by

00:39:36.400 the circadian time of night. But we still don't understand what the function of that skewed

00:39:43.860 seesaw balance actually is. So we understand the factors that drive that different distribution,

00:39:50.740 but we don't understand the evolutionary functional reasons as to why mother nature drives it in that

00:39:58.260 way. But overall, I think we've done a pretty good job of hopefully outlining what sleep is,

00:40:03.560 how it's structured, and its timing. I remember you saying that REM is even more under circadian

00:40:09.140 control. Is that actually the same as saying that REM occurs more often in the, as you said,

00:40:16.340 the latter part of the day? Or does it speak more towards the ability to reset that clock? And if

00:40:24.780 my question makes sense, great. If not, I'll pivot to a different question that I hope gets at the

00:40:29.600 same thing. So REM sleep is less determined by how long you've been awake during the day. That is much

00:40:38.560 more deterministic of how much deep sleep and how deep that deep sleep is going to be. And that's what

00:40:44.920 sort of almost forces the serving of a plate of deep non-REM sleep up in the first couple of cycles.

00:40:52.580 So just to be clear, Matt, and sorry to interrupt you, does that mean then that deep sleep is more

00:40:56.860 a function of how much adenosine you've accumulated, how much melatonin you've accumulated, and how much

00:41:03.580 cortisol you've suppressed? Those three things factor more into your ability to generate lots of deep

00:41:10.740 sleep? Probably a little bit less so the middle one, which is melatonin. Some of the last one,

00:41:17.280 which is cortisol. Higher cortisol typically means less deep sleep, but very much so the first one,

00:41:24.360 which is adenosine. And to explain what adenosine is, from the moment everyone listening woke up this

00:41:29.600 morning, a chemical started building up in your brain called adenosine. And the longer that you're awake,

00:41:35.580 the longer the amount of clock time that your brain has been awake, the more adenosine builds up.

00:41:41.700 And adenosine is a chemical that produces what we call sleep pressure. And the more of that sleep

00:41:48.900 pressure, in other words, think of it as sleepiness, the longer that you've been awake, typically the

00:41:53.680 sleepier that you feel. What you're feeling is the weight of that chemical pressure of adenosine.

00:42:00.600 Now, it's not a mechanical pressure in your brain, don't worry. It's a chemical pressure that's

00:42:05.400 signaling sleepiness. The more of that adenosine that there is in the brain, the more of that deep

00:42:11.720 sleep that you typically get. You're absolutely right. That makes sense now as you describe it as

00:42:17.200 that pressure. The question now comes down to one of napping, which we did touch on in our previous

00:42:23.320 episode, but I want to revisit it. Lots of times patients come to me and say, Peter, you have no idea

00:42:30.160 what a 20 minute nap does for me between two and four o'clock in the afternoon. And my historical

00:42:37.440 response to that has been, you know, gosh, you really shouldn't need a nap. I mean, if you're

00:42:42.600 sleeping correctly, isn't that actually a negative thing because it's depleting some of that adenosine

00:42:48.020 based pressure. And wouldn't you be better off letting the pressure cooker cook and driving you into a

00:42:54.820 deeper sleep when you first take your sleep? So what are your thoughts on that type of a nap,

00:43:01.840 that sort of short 20 minute, you know, late afternoon pre-dinner nap?

00:43:07.960 Naps really are a double-edged sword. And we've done lots of studies where we use naps to study the

00:43:15.200 functions of sleep and we see benefits from naps. Even naps as short as 17 minutes can produce learning

00:43:21.340 and memory benefits. So there does seem to be some, some enhancement that you get. And you can see

00:43:27.160 that from cardiovascular benefits as well. The downside of naps is that it can take away just

00:43:36.280 what you described, which is sleep pressure. And so the typical recommendation that we have

00:43:40.600 is the following. If you are struggling with sleep at night, avoid naps during the day,

00:43:47.860 because what you want to do, if you are already having problems, either falling asleep, which is

00:43:53.440 what we call sleep onset insomnia, or you can fall asleep, but you can't stay asleep, which is what

00:43:59.280 we call sleep maintenance insomnia. You want to build up as much sleep pressure, as much sleepiness as

00:44:07.340 you possibly can during the day. In fact, sometimes what we do for patients with insomnia is something

00:44:13.100 called bedtime restriction, where we'll even tell them, I know that you're, you're not sleeping

00:44:18.060 enough. And paradoxically, I'm going to tell you that you should go to bed even later. And they think,

00:44:23.720 well, that's stupid. I should be going to bed earlier, shouldn't I? Not necessarily, because what

00:44:29.040 we want to do is elongate the amount of wakefulness that you've been building up even more than is

00:44:35.060 normal. So you build up an even greater sleep pressure. And as a consequence, yes, you may only have a

00:44:42.320 six-hour sleep window that we now give you on this sleep restriction therapy, which is part of CBTI.

00:44:49.580 But that six hours may be a very constant, solid sleep. And that's, we believe, much better than you

00:44:57.380 lying in bed for eight hours and tossing and turning for two of them and only sleeping six hours. So in

00:45:03.660 both scenarios, you still get sleep six hours. But in the kind of clinical scenario, and by the way,

00:45:09.900 I'm not a clinical doctor, I'm just a scientific researcher. So I'm just describing what happens in

00:45:14.800 CBTI. At least with that delayed bedtime, it's a nice, consistent quality of sleep, because quality

00:45:22.060 of sleep matters just as much as quantity of sleep. So I'm going a little bit off topic, but to bring it

00:45:27.700 back, if you are struggling with your sleep at night, do not nap during the day. And the pressure

00:45:33.240 cooker analogy is beautiful. Keep building up all of that healthy sleepiness pressure, because when you

00:45:39.380 nap, it's like having the valve on the cooker open up and you just, you just release some of that

00:45:44.980 healthy sleepiness. And now when it comes time to go to bed, you're not going to feel as sleepy

00:45:49.580 anymore. However, on the flip side, if you are not struggling with your sleep and you can nap regularly

00:45:57.340 during the day, the advice is naps can be just fine. Because in fact, if you take a step back from an

00:46:04.360 evolutionary perspective and you study hunter-gatherer tribes, whose way of life hasn't changed for

00:46:09.220 thousands of years, they don't necessarily sleep the way that we do in modernity. In fact, particularly

00:46:15.580 during the summer months, they will typically have an afternoon nap right in the time zone that you

00:46:20.980 just described. And this is the very much the Mediterranean siesta-like behavior. And in fact,

00:46:27.760 if I stick an electrode on both of our heads throughout the day, and I monitor our physiological

00:46:33.060 brainwave activity... It'll stick much better to my head, by the way, than yours.

00:46:38.260 Well, yeah, right now, especially because I've got this terrible COVID hairdo, which I look like I

00:46:43.720 should be out of Buck Rogers and anybody who remembers that television show. But anyway, what we will see is

00:46:49.540 that somewhere between about 2 to 4 p.m., both you and I will have this kind of drop in our

00:46:55.660 physiological alertness, in our physiological level of brain activity. In other words, as if, as a species,

00:47:02.600 we're almost pre-programmed to have this enforced dip in our alertness. And so many people see this,

00:47:11.020 you know, in the afternoon meetings around the boardroom table or wherever, you sort of get these

00:47:15.220 ugly head nods that start to happen. It's not people listening to good music. It's just that

00:47:20.040 people have this sort of... They're falling asleep. They're falling prey to this, what seems to be a

00:47:25.860 genetically hardwired, pre-programmed drop in our alertness, as if maybe we should be what's called

00:47:32.180 biphasically sleeping as a species at times during the year, rather than monophasically sleeping, which is

00:47:39.320 how we do in most First World nations. In other words, one single bout of sleep at night versus

00:47:44.600 two bouts of sleep, which is how those hunter-gatherer tribes sleep, especially during the summer.

00:47:51.160 And so Matt, would they sleep long enough in that 2 to 4 window that they would get a full

00:47:56.240 90 minutes, say, and get through a full cycle? And then if so, did that mean when they did their

00:48:02.320 nighttime sleep, presumably they would stay up later and get maybe 6 hours in the evening instead of,

00:48:08.880 you know... Because people always talk about how well hunter-gatherer studies have always suggested

00:48:12.420 that, you know, 7.5 to 9 is the species-required amount of sleep on average, but you're saying it

00:48:18.980 was potentially broken into two chunks. Was that afternoon daytime chunk or daylight chunk,

00:48:24.940 call it, about 90 minutes? No, it wasn't. It was actually typically shorter than that on that basis.

00:48:31.620 But you're right, they typically are not sleeping. What we currently recommend for monophasic sleep,

00:48:37.960 which is the average adult should get somewhere between 7 to 9 hours. That's the recommended range

00:48:44.340 that we all provide. But what they would typically do is they would maybe sleep just 6, 6.5 hours at

00:48:52.240 night, and then they would make the rest up and they would get into that exact same sort of territory

00:48:58.100 by way of the nap. So they would take it in these two chunks. Now, I should note, this is very different

00:49:05.080 to something else that's been described in the literature, which is first sleep and second sleep.

00:49:10.800 That activity does seem to have occurred during our historical past. It seemed to have emerged in

00:49:17.320 the sort of Dickensian era. That was different, though. That was where people would sleep for the

00:49:23.120 first four hours, then they would wake up in the middle of the night, they would, you know, write,

00:49:27.700 they would drink, they would play music, they would make love, and then they would go back for

00:49:31.620 another four hours of sleep. Did that happen? Yes, it seems to have, based on historical writings.

00:49:37.900 But is that the way our physiology and our circadian rhythms are designed? No, it doesn't

00:49:43.960 seem to be. There's usually one study that's cited that suggests maybe we should do that. But overall,

00:49:50.280 the physiology that we know of for human beings doesn't seem to suggest that that was more

00:49:54.780 sociologically driven, rather than biologically driven. Biologically, we may be biphasic,

00:50:02.080 but very differently, according to this hunter-gatherer sort of tribe hypothesis. By the way,

00:50:07.220 the other thing that's interesting in those hunter-gatherers, and it comes on to what I think

00:50:10.900 we described before, which is this concept of midnight. Most of us don't think of what the term means.

00:50:17.420 But in those individuals, it means what the word states, which is it's the middle of the night for

00:50:24.680 them, midnight, because they usually go to bed maybe two hours after sundown, and they're asleep,

00:50:32.100 you know, by let's say 9 p.m. in the evening, or 8.30, and then they're awake, you know, just a little

00:50:38.800 bit before dawn. In fact, what's interesting, and maybe we'll get onto this in later discussions,

00:50:44.160 or we won't. But what seems to wake them up is not necessarily daylight. It's the change in

00:50:50.180 temperature, because there's a rise in temperature that seems to happen before daylight starts to

00:50:57.720 break. And it seems to be, it's the temperature change that they are much more buckled to in terms

00:51:04.520 of regulation of their sleep-wake rhythm. And so the way in which the structure of their sleep,

00:51:10.880 biphasic, is different to ours in modernity, and the timing of their sleep is different to

00:51:17.560 some of our timing. You know, for many people, midnight is the time that you think, well, I should

00:51:22.540 just check Facebook one last time, or, you know, send my last text. That's what midnight means. But

00:51:28.060 normally for a couple of thousand years, or hundreds of thousands of years, it probably meant that was

00:51:34.800 the middle of your solar cycle, and it was the middle of your sleep cycle.

00:51:40.340 I'm glad I'm closer to our ancestors. If I could go to bed at eight o'clock every night and wake up

00:51:45.400 at four, that is a perfect night of sleep for me. And by the way, you may have already mentioned this,

00:51:50.240 but just to close the loop on it, do we have every indication that our ancestors also had 90-minute

00:51:55.460 cycles? Or do we believe that there was a period in our evolution as homo sapiens when our cycle was

00:52:01.840 longer or shorter? I accept the fact that obviously when our brains were smaller, when we were pre-hominoid,

00:52:07.100 it might have been different. I don't think we know of that data yet. I think most of those studies

00:52:12.900 have used wristwatch sleep tracking rather than actually taking PSGs out there into the field.

00:52:18.600 How many people are there left in the field to study? I mean, I know that in Australia,

00:52:21.840 there are some aboriginal cultures that still allow us that opportunity. But I mean,

00:52:26.060 in Africa, are there others as well?

00:52:27.860 Yeah, there's the San in, I think that's how you pronounce it, in Namibia. There's a couple of

00:52:33.320 cultures in South America. So there are probably about four or five different cultures that have

00:52:39.340 been studied, hunter-gatherer tribes. But we don't actually know if the sleep cycle is similar. My

00:52:47.500 suspicion is that it is. But if we want to reverse and go into the time capsule further,

00:52:53.600 what seems to have happened if you use different primates as the best kind of data reversal of our

00:53:04.020 evolutionary timeline, there was something about the transition of Homo sapiens from tree to ground,

00:53:11.740 which increased the amount of REM sleep that we were getting dramatically. And in fact,

00:53:18.040 if you look at our sleep amounts relative to other primates, our non-REM sleep amounts, our human

00:53:26.200 beings, you know, Homo sapiens, somewhere in that distribution, we're pretty similar. But in terms of

00:53:33.460 REM sleep as a primate, we are a complete standout. We have almost double the amount of REM sleep that

00:53:40.460 any other primate has. And so I think, and in the book, I sort of did some lots of hand-waving and

00:53:48.740 pontificating about this. I wrote a book and in some of those chapters, I sort of speak about

00:53:54.580 what was the cause of that and what did it allow? And I think one of the benefits was, you know,

00:54:00.960 emotional intelligence. It wasn't IQ, it was EQ that REM sleep also allowed us to develop. But there was

00:54:07.480 something that happened as we moved from the trees to the ground, because orangutans, other species,

00:54:13.700 they will sleep up in trees. In fact, they will build a bed every night up in a tree. Imagine, you know,

00:54:20.160 if you said to your kids, every night, you're going to have to unpack a piece of Ikea bed furniture and

00:54:24.900 build it before you go to sleep. Every single night, you're going to have to do that. Well, that's what some

00:54:28.940 of these primates do every night. They build a new bed. But anyway, the idea is that I don't know about

00:54:34.640 the sleep cycle, but during our evolutionary past, as we transitioned down to terra firma,

00:54:42.340 we increased, there was something about that. And I think it has to do with the paralysis that

00:54:47.900 happens during REM sleep, which is obviously if you're living 20, 30, 50, 100 feet up in the trees,

00:54:53.780 and you're lying on a branch, and all of a sudden you lose all muscle tone, which is what happens when

00:54:59.020 you go into REM sleep, bad things are probably going to happen. You're probably going to be popped out of

00:55:04.180 the gene pool quite quickly that way. So when we became safe on the ground from a sheer

00:55:10.760 Newtonian physics standpoint, then all of a sudden REM sleep blossomed. And with it, I think a whole

00:55:17.460 constellation of new benefits for the hominid species. I just can't resist making one comment

00:55:24.020 about motorsport. I'm sorry, Matt. Well, I'm not sorry for you, because you're going to love it. I'm

00:55:27.560 sorry to all the other listeners. I was listening to our favorite F1 podcast last week, and they were

00:55:32.420 talking about the golden age of F1, which was of course the 80s. And someone made a wisecrack that

00:55:37.880 the chief aerodynamicist for the McLaren team in 1986 was Sir Isaac Newton. And I just couldn't stop

00:55:45.920 laughing. I was like, that is so great. And for the listener, what that really means is back in 86 was

00:55:51.640 really before you had any ground effects in cars. And so there was no aerodynamics. It was basically the

00:55:57.020 only thing that held the car on the ground was its own mass and the force of gravity. But those cars

00:56:01.520 were beasts, right? I mean, yeah, they were absolute beasts. I mean, and before that,

00:56:05.800 Colin Chapman at Lotus tried some of this stuff. If you look at some of the harebrained stuff that

00:56:10.060 they were doing, these crazy aerofoils that sat about four or five feet above the car on these

00:56:16.080 stanchions. And you never quite knew, I think, as the driver, whether it was going to make you

00:56:20.720 force you down and stick you to the road more, or whether the next report that you made was that

00:56:26.880 you were about 30,000 feet above the Arizona desert, you know, and you've sort of essentially

00:56:31.680 gained some degree of altitude. But yeah, I think Adrian Newey probably has a slightly better grasp

00:56:37.700 on things now. Well, look, it's not called the drive for nothing. So every once in a while,

00:56:43.520 we are going to have to just... I'm sorry, folks. I'm desperately sorry. No more. Yeah. All right. So look,

00:56:49.020 we have a pretty good handle on this stuff now, Matt. And on previous episodes, obviously,

00:56:52.920 we went into some pretty good depth on tricks, tips, etc. Inside our practice, we have a whole

00:57:01.560 sleep hygiene protocol. And we sort of run patients through it relatively early, especially if they're

00:57:07.500 having any sleep issues. But do you mind just sort of going through where you are now and how you think

00:57:13.680 about this? And again, I'll caveat this by saying you're not a sleep physician. So it is not your

00:57:18.760 practice to be out there treating any one individual with respect to their sleep. That's right. But

00:57:24.480 also, you're infinitely more qualified than virtually anybody to help people start to think

00:57:29.300 through what the parameters are that factor into sleep. And that's the way we kind of explain it,

00:57:34.640 which is, look, here are these five or six levers of sleep. And here's how you could move each of them

00:57:40.260 in your control. But I kind of want to hear how you think of it.

00:57:43.580 Yeah, so I think in our past episodes, we've gone through sort of the fine five main sort of sleep

00:57:50.140 hygiene tips. And I'll just quickly go through them here, because we've dealt with them before,

00:57:54.800 which is regularity going to bed at the same time waking up at the same time, getting lots of darkness

00:58:00.880 at night, because we are a dark deprived society. And I actually think I've done a bad job of describing

00:58:07.480 to the public the importance of the opposite, which is making sure you get daylight during the first

00:58:12.420 half of the day. I think that's just as important. Then temperature, we've spoken about that a little

00:58:17.640 bit. You need to get cool to get to sleep. And it's the reason that you will always find it easier to

00:58:23.660 fall asleep in a room that's too cold than too hot. You shouldn't stay in bed awake. That trains your

00:58:30.180 brain to be triggered by your bed and force you awake because you have a learned association.

00:58:35.860 So if you've been awake for 20 minutes, then get up, go and do something else,

00:58:41.020 and only come back to bed when you're sleepy is the general rule. And then finally, you know,

00:58:46.240 trying to avoid alcohol and caffeine in the afternoon and for alcohol in the evenings, as we've

00:58:51.620 said. So those are the sort of typical tips. But I think if I were to add a few others,

00:58:58.480 the other one that's absolutely critical that I probably haven't espoused enough is a wind down

00:59:03.920 routine. Because many people in society expect sleep to be like a light switch,

00:59:09.800 that we should just jump into bed and we should just turn off the light and the brain should do

00:59:15.200 something similar and go straight into sleep. Sleep as a physiological process, if you study it,

00:59:20.700 it's just not like that. Sleep is much more like trying to land a plane. It takes time to gradually

00:59:28.620 descend down onto that hard foundation of this thing that we call a stable night of sleep.

00:59:34.900 So give yourself some wind down opportunity time. Build it into your routine. For some people,

00:59:41.240 it's 15 minutes. Others, it's 30 minutes. Light stretches, meditation, putting all of your phones

00:59:47.540 and your gadgets away, staying clear of any inbound in the last 30 minutes. Whatever it takes, set it up

00:59:54.040 and then maintain it. Because if you have kids, you'll know all about this. You know, you have a routine

00:59:59.360 for the kids and you've got to try and stick to it. If you break the routine, bad things usually happen

01:00:04.520 with sleep. We're the same as adults. There's no difference. So I think that's the first thing I

01:00:10.260 would say. The second probably tweak is if you are struggling with sleep, remove all clock faces from

01:00:19.620 your bedroom. It's not going to help you to know that it's now 2.35 a.m. in the morning and you've

01:00:27.640 still not been able to fall asleep. It's only going to trigger more anxiety. And there's been

01:00:32.640 a nice study here from UC Berkeley, not from my center, but from another sleep researcher, Alison

01:00:36.840 Harvey, who looked at this in people who are poor sleepers with insomnia. Removing clock faces can

01:00:42.700 certainly help. The next thing I would say is that try to keep, of course, all of your technology

01:00:48.660 outside of the bedroom. And if you can, don't make it the first thing that you check in the morning.

01:00:55.160 Because for most of us, the first thing that we do when we wake up is that we swipe right

01:01:00.120 and this flood of anxiety just washes onto us like a tsunami. And that's problematic, not

01:01:06.700 just because it's a bad way to start your day, but because you train your brain in this

01:01:12.100 Pavlovian way that every morning, as you're tucking yourself into bed at night, every morning,

01:01:17.200 what's coming to you when you wake up is this jag of what we call anticipatory anxiety.

01:01:24.160 And it lightens your sleep throughout the night because of that expectation. So try to avoid that.

01:01:30.820 If you absolutely have to, and again, trying not to be puritanical, take your phone into the bedroom.

01:01:36.940 One of the other problems is that it causes what we call sleep procrastination, which is that you're

01:01:41.840 sitting there in bed, you've got your device and you think, well, I'll just check Amazon and order

01:01:45.820 that thing. I'll just check Facebook and I'll just, uh, I'll send that last email and you look

01:01:50.260 up and now it's 40 minutes later. The rule of thumb that a friend, a colleague of mine,

01:01:55.340 Michael Gradner said, I love this. If you're going to have your phone in your bedroom, the rule is

01:02:01.200 that you can only use it if you're standing up. And after standing up for about five or 10 minutes

01:02:07.100 of phone use, you just think, I just want to sit down. I just want to get into bed. And at that

01:02:12.000 point, the rule is you've got to put your phone away. So those would be sort of three additional

01:02:16.820 things that I would advise that can really help. I like that a lot. I would actually add just from

01:02:22.040 a personal standpoint, I've maybe three months ago instituted a little policy for myself that

01:02:28.460 I've really enjoyed, which is not looking at my phone for an hour when I wake up. Luckily I don't

01:02:35.500 sleep with my phone in my room, so that's not an issue, but it used to certainly be the first thing

01:02:40.400 I'd go and grab when I got up. And if I get up at, you know, five, I won't look at it till usually

01:02:47.760 after 6 a.m. And that's interesting. I didn't realize that the effect that that could have on my

01:02:53.500 brain as I'm coming into consciousness, knowing that I'm not going to be flooded with information,

01:02:59.500 especially the type of information I don't like, which is generally all the information on my phone.

01:03:03.260 Yeah, I'll just stick two additional pins in that. The first is, I think everyone has had this

01:03:09.420 experience in the extreme, which is where you know you've got to wake up the next day for an

01:03:13.920 interview, or you've got to wake up for a flight that you have to catch. You know for a fact that

01:03:18.800 firstly, your sleep is going to be shallow. It's not going to be particularly deep. And secondly, you

01:03:23.720 almost wake up two minutes before the alarm because you're that on edge. Well, imagine a diluted version

01:03:29.860 of that, but every single day, that's what bringing your phone into the bedroom and opening it up

01:03:34.860 first thing is all about. The second thing I would say is you're right and set manageable goals. I

01:03:42.040 really enjoy the work of a guy called BJ Fogg, who's a researcher at Stanford who looks at behavioral

01:03:46.860 change. And it's all about incremental. So if you're trying to get someone to floss their teeth

01:03:51.900 for the first time for dental hygiene, don't say, okay, here's how you floss. You need to floss before

01:03:57.200 you go to bed. You say, all you're allowed to do for the first week is floss your front two teeth

01:04:02.680 and you cannot do any more. That's it. And then we say, then to the next, then add two more teeth

01:04:10.360 for the next week and then two more. The same with sleep. If you are accustomed to doing this with your

01:04:16.000 phone, which is so understandable for all of the pressures that we understand from society and social

01:04:21.980 media, start by saying, I'm going to give myself five minutes. I'm going to firstly not check my phone.

01:04:27.420 I'm just going to wake up, brush my teeth and make whatever drink that I have in the morning.

01:04:31.900 And when I sit down, that's when I am gifted the opportunity to look at my phone and then try to

01:04:37.800 push it to 10 minutes and then try to see if you can get changed, have your shower, wash yourself up,

01:04:43.800 whatever you do in the morning, and only then open up your phone and keep pushing it longer and longer.

01:04:49.260 Make it a manageable goal. Otherwise, don't set yourself up for failure. Do it for success.

01:04:54.380 Yeah. I think the same advice is great for something like meditation where, you know, for many people

01:04:59.980 saying, hey, why don't you commit to 20 minutes of meditation a day? I mean, that's almost impossible

01:05:06.040 out of the gate. But if it's, hey, meditate every day. And if you do a three-minute meditation that,

01:05:11.400 you know, you're better off doing three minutes a day than 20 minutes once a week. I think it's the

01:05:15.160 same sort of idea, which I agree. I think it becomes a lot easier.

01:05:18.920 The last thing I would say is that none of these tips that we've spoken about, which are typically sleep

01:05:23.920 hygiene tips, are going to work if you actually have a sleep disorder. So the analogy here,

01:05:29.520 again from a colleague, if I'm an athletic sports coach and I'm giving you all of these tips for

01:05:35.200 improving your performance, but you've got a broken ankle, none of them are going to make any difference

01:05:40.100 to your performance until we actually get you to a doctor and fix the broken ankle. It's the same

01:05:44.740 with a sleep disorder. If you have, you know, insomnia or if you have sleep apnea, none of these things

01:05:50.940 that I've been talking about are going to help you. If you suspect that you have either one of

01:05:55.260 those, definitely go and see your doctor. See if you can get some kind of sleep intervention.

01:06:00.400 Matt, last time we sat down, you talked briefly about the sleep chronotypes. And I've found this

01:06:05.500 interesting for basically two reasons. One is just observing probably a migration in my own

01:06:10.720 daughter's chronotype, which is as she's getting older, she's sleeping in longer in the mornings and

01:06:17.480 going to bed later and later. And I know that as she gets closer and closer to being a teenager,

01:06:22.340 that's probably going to increase, but also just more cognizant, I think of other people.

01:06:27.580 You've alluded to it briefly earlier, which is like, there are just people who are going to

01:06:31.020 go to bed later and wake up later. And it doesn't mean that there's anything wrong with that,

01:06:35.580 or there's anything wrong with the reverse, which is people who go to bed early and wake up really

01:06:38.740 early. So kind of wondering if there's a way to actually know what chronotype you are

01:06:45.860 beyond just observation, because is it possible that the observation is not your optimal state?

01:06:54.240 In other words, if you have somebody who's sleeping a certain way because their job is imposing it,

01:07:00.080 but it's not the way that they're, you know, ideally meant to do it. Let me give a better example.

01:07:04.640 A student in college could easily drift into a later chronotype because everybody in their dorm room is,

01:07:12.600 even though they may actually be more suited to be earlier. So how can one figure this out?

01:07:19.220 So there are genetic tests that will give you a stronger sense of your chronotype. So

01:07:25.500 chronotype simply means, yeah, are you a morning person? Are you an evening person? Are you somewhere

01:07:30.960 in between? And somewhere between 25 to 30% of the population are morning types, 25 to 30% are evening

01:07:39.060 types. And then the rest of us are somewhere in between. Sleep science has then gone a little

01:07:44.480 bit further. We split it into five categories sometimes, which is extreme morning types, morning

01:07:49.820 types, middle evening types, and extreme evening types. We know that it is under strong genetic

01:07:57.380 control for two reasons. First, there is a significant degree of heritability. So if you are a morning

01:08:04.880 type, it's more than likely that one or both of your parents were morning types. Secondly, we now know

01:08:11.740 that there are a collection of genes that will determine to a degree your chronotype. And this

01:08:19.140 is why companies like 23andMe will tell you what type you are, or they will give you a probabilistic

01:08:25.120 estimation of what type you are. Last count, I think, from data from 23andMe and there's data from the UK

01:08:32.960 Biobank, there's about nine different genes that we know of right now that will contribute to your

01:08:39.080 chronotype. Most of them are what we call clock genes. This refers to the fact that these genes control

01:08:46.520 the rhythm of your circadian cycle. Now, those genes don't control necessarily the size of your circadian

01:08:56.260 rhythm. In other words, how kind of strongly active and awake you are during the day and how

01:09:01.740 deeply sort of down you go at night, these genes do something different. These genes control when

01:09:09.240 that sinusoidal wave of your circadian cycle arrives on the clock face. What I mean by that is

01:09:16.240 if you have a certain type of gene complement that makes you a morning type, your activation peak,

01:09:26.180 your peak alertness is going to arrive earlier in the day, and your awesome downstroke of your

01:09:32.680 circadian rhythm will arrive earlier in the evening. Whereas if you have a different combination of these

01:09:40.180 gene composites, then you could be an evening type. In other words, your circadian rhythm looks very

01:09:46.940 similar in its sinusoidal pattern, but when the peak and the trough of that circadian rhythm arrive on the

01:09:54.640 24-hour clock face is very different to a morning type. So we know that there is a complement of genes

01:10:03.360 that you are given at birth that will determine, on average, once you're an adult, are you a morning

01:10:09.180 type or are you an evening type? What you alluded to, however, for your daughter was something different,

01:10:15.580 which is that no matter what chronotype you are, that innate chronotype rhythm that you have,

01:10:22.620 gifted by your genes, will change as you develop from a young child to an adolescent teen to an adult

01:10:31.340 to an older adult. In other words, you go from being much more of a morning type when you're a kid,

01:10:37.240 even though you want to stay awake longer, you're found on the couch and you're carried to bed and

01:10:41.660 you're fast asleep, to then being a teen where your chronotype fast forwards in time. And this is a

01:10:48.560 problem for early school start times where you're putting kids to bed at 9 p.m. and saying you've got to

01:10:53.260 go to bed because you have to wake up at 5 o'clock in the morning to catch a 5.30 bus for a 7.15 start

01:10:59.260 time. Well, there's no point in saying that. It's not their fault that they're now in bed and it's 10.30

01:11:05.460 an hour and a half later and they can't fall asleep as a 15-year-old because their chronotype

01:11:11.560 has moved forward in time. They want to go to bed later and wake up later. Nothing they can do about

01:11:16.920 it. It's biology. And then once they become an adult, it starts to shuttle back a little bit once

01:11:22.940 they become sort of, you know, moderate age, 30 or 40. And then as we start to get older, we regress.

01:11:29.760 That's why in Florida there's something called the early bird special where people are going out for

01:11:34.860 dinner at, you know, 4 p.m. or 5 p.m. because their chronotype has regressed back and they're in bed by,

01:11:41.400 you know, 8 p.m. So there are genetic tests to come back to your question. However, you don't need

01:11:48.600 to do a genetic test to get close to understanding what you are. There is a pencil and paper method

01:11:53.980 that sleep scientists have developed and you can Google it. It's called the MEQ test. And maybe I'll

01:12:01.460 send you the link and we can put it in the show notes. Yeah, we'll put it in the show notes.

01:12:05.700 And it's a very simple link that you just click and it's a test. It probably takes about 5 or 6

01:12:10.580 minutes to fill out. And it's called the MEQ test, the morning eveningness questioner test.

01:12:18.240 And you go through, you answer a list of questions, and at the end you add up your score and it will tell

01:12:23.200 you what flavor of chronotype you are. And what's nice is that then they've pattern matched this test

01:12:29.600 and validated it against these genes, these genetic tests. And there's a pretty good correlation.

01:12:36.180 It's not too far off. So this is the poor man's version of the 23andMe test, if you want it.

01:12:42.300 And we'll do a link in the show notes and you can take it. And that's a cheaper way,

01:12:47.300 a pencil and paper method or an online method.

01:12:49.340 So the bigger question is, Matt, what do we do with this information, right? I mean, you've,

01:12:53.840 I think we did talk about this previously at one point that you have a couple who have different

01:12:59.000 chronotypes. That can be really difficult. They're going to bed and waking up at different times.

01:13:03.540 You've already alluded to the problem of children who are, I think, being subjected to potentially

01:13:09.880 too much early school, when in reality they probably would do better with later school.

01:13:15.160 Is this the kind of thing that should factor into decisions people make about when they pick

01:13:19.660 classes in college? Like, I mean, it seems like there seems to be, this seems to be one of those

01:13:24.060 things where there might be a bunch of people that go through life kind of miserable when they don't

01:13:27.540 have control over their schedule as opposed to knowing this. And so would knowing this by itself

01:13:33.800 provide a benefit to folks?

01:13:36.020 I think it would. I think it would explain a lot about why you struggle so much where other people

01:13:42.260 seem to be just these energizer bunnies and they've even got time to go to the gym for an hour before

01:13:47.420 they sit down at their office desk at 7am in the morning. And you're still on your fourth cup of

01:13:54.280 coffee desperately trying to wake up and you're struggling to make it in by 7.15 having woken up

01:13:59.280 20 minutes before. And I think the second thing that it helps with is realizing that you're not

01:14:06.860 culpable. It's sort of non mea culpa. It's not your fault. It's, you can understand that this is

01:14:13.860 genetic because a lot of night owls have gone throughout life being chastised and thinking,

01:14:19.100 well, if only I could get my act together and get to bed earlier and stop being what people tell me

01:14:25.520 I am, which is slothful or lazy. You know, I could hold down a better job or, and again, I think society

01:14:32.980 needs to be much more understanding of it and also then modifying itself in response to accommodate it.

01:14:42.220 And neither of those things I see in full serving scoop sizes in society right now. And I, that should

01:14:49.960 change. And I think COVID is interesting because if there's some positive upside to it, it may have

01:14:55.680 given us the chance for people to understand how much better that they can sleep when they're closer

01:15:01.600 in harmony with their own chronotype rather than trying to work against it. Because when you fight

01:15:07.260 biology, you normally lose. And the way you know, you've lost is often through disease and sickness.

01:15:13.200 And I think there is a version of that going on here. I also think it can sometimes explain

01:15:19.840 incorrect insomnia. Often I'll hear someone tell me if I'm sort of out in the public,

01:15:26.940 they'll say, well, I suffer from insomnia. I get into bed and I can't fall asleep and I'm wide awake for

01:15:33.120 the first hour and a half and I need to take some sleeping pills. And then I'll, I'll say, you know,

01:15:39.800 firstly, I'm not a doctor, but let me just ask you a question. If you are on a desert island with no

01:15:45.320 responsibilities, nobody to wake up for, nothing to do, and you could just sleep whenever you want and

01:15:51.080 get up whenever you want, what time would you normally go to bed? And they would say, well,

01:15:55.600 I would actually like to go to bed at midnight and wake up around eight o'clock in the morning in truth,

01:16:00.580 but I have to get up for my job. I have to wake up at six. I'm getting in bed at 10,

01:16:06.120 but I still have this terrible insomnia. And I say, well, you know, I don't know this,

01:16:10.860 but it's possible that what you could have is not insomnia, but a mismatch between your chronotype

01:16:16.600 and your working life schedule. And you may want to explore some of this. And there's a test,

01:16:21.760 an MEQ test, and I tell them about this and then go and speak to your doctor. So that's the second

01:16:26.220 implication that I think comes from the question that you asked about how it impacts society.

01:16:31.600 Once you know it, what do we do about it? And what should we be doing about it? Those are some

01:16:36.660 of the ramifications, I think. There's a question that actually came in that I think a lot of parents

01:16:40.960 might be able to relate to. And this might be outside of your area of expertise, because I know

01:16:46.040 you're not a clinician per se. And if so, if the answer is, we don't know, that's fine. The parents

01:16:50.740 ask, we have a six-year-old and they basically experienced sleep terrors, but the parents noticed

01:16:55.280 something, which is that exhaustion triggers the terrors. So when their child is actually well

01:17:01.240 rested, he doesn't get them. And they're always trying to implement great sleep hygiene. So I think

01:17:06.620 what the question is here, is there any empirical observation to support this, or is that just a

01:17:10.720 nuance of their child? I think there is good reason to explain why it happens. Firstly, night terrors

01:17:18.060 are typically where people wake up, and it's more common in children, with just this terrible sense of

01:17:23.720 dread and fear. And your kid is almost inconsolable at that stage, they're crying. But what's interesting

01:17:30.620 is that even when, you know, you have one of these night terrors as an adult, if you were to ask people

01:17:36.340 what was, you think, well, they were having a horrific dream. But if you ask someone, you know,

01:17:41.700 what was going through your mind just before you woke up, they typically say, you know, nothing much,

01:17:47.320 I don't remember anything. And in part, it's because many of these night terror events seem to

01:17:53.180 happen from deep non-REM sleep, dreamless sleep, rather than rapid eye movement sleep.

01:18:00.280 That's the first thing. Second thing is that we don't really understand too much about the brain

01:18:04.380 mechanisms, because in part, it's not a good thing to study. You know, the way we can study some of

01:18:10.020 these events in the brain is that we put you into a brain scanner, and we start scanning you while

01:18:15.140 you're having, you know, lots of these events, for example, whilst you're trying to make memories,

01:18:20.000 and we try to have you learn 100 different memories, and we scan you 100 times, and we build up a map

01:18:25.340 of what successful memory formation looks like. That's how we do these studies. But you can't stick

01:18:31.140 a child in a scanner for eight hours, and with just one event, hope to capture it and get a good

01:18:36.740 signal. So we actually don't really understand much about the real-time brain dynamics. But the reason

01:18:41.940 I bring up non-REM sleep is that, just as we spoke about before, when we are spending more time awake,

01:18:49.240 or we've had essentially sort of supercharged or extended wakefulness, when we go to bed at night,

01:18:55.980 we typically then have a homeostatic increase in the amount of deep sleep. So what could be happening

01:19:02.700 here, and again, this is just hand-waving, but one explanation would be that when the child is

01:19:08.500 exhausted, maybe they haven't been sleeping as well, maybe they missed their nap, maybe they've been awake

01:19:13.860 for typically longer than they would do otherwise, they're probably going to go into deep non-REM sleep

01:19:19.220 in a more intense fashion, or with greater amounts of deep sleep duration, which is therefore going to

01:19:25.440 mean that the probabilistic likelihood of them having one of these night terrors is increased as a consequence.

01:19:32.280 So that's sort of from a scientific perspective, but you're absolutely right. I am not a pediatric

01:19:39.280 sleep scientist. I am not an MD. I would refer anyone who has issues or questions about infant sleep or

01:19:47.080 pediatric child sleep to one of two people. The first is Dr. Craig Canapari, and maybe I can send you

01:19:54.360 guys a link and you can link to his website. He is just fantastic. He is the director of pediatric sleep

01:20:00.700 at Yale Medical School. He's written a great book that parents can read about sleep and infants and

01:20:06.180 children. You could read it in an afternoon. It's just brilliant. It's a lovely book. He's a great guy.

01:20:10.580 He's got a fantastic website with tons of resources. I wish I could keep my website like he does,

01:20:16.560 curates it beautifully. So Craig Canapari, both with C's. And then the other person is Jodie Mindell,

01:20:23.700 who is kind of like the god, you know, there's godfathers and godmothers of research. She's sort of

01:20:28.660 one of the godmothers of child and infant sleep out there in the world. So I will send you a link

01:20:34.800 there to include in the show notes. If you have any questions about child or infant sleep,

01:20:40.860 one of those two individuals is more than likely going to have an answer. They are brilliant.

01:20:46.500 I want to pivot to another topic, Matt, which is, I know you're sort of working on a second

01:20:51.600 edition to your book. Is that right?

01:20:53.300 That's right. That's correct, yeah.

01:20:54.620 And I know that over the past, I don't know, nine months or so, there's been some discussion of

01:21:02.240 errors in the first edition of the book. And anyone who's written anything knows that that's

01:21:09.400 pretty normal. Back when I used to blog a lot, I don't think I could get through a blog post without

01:21:14.380 making one mistake. And sometimes the mistakes were literally just mistakes of fact, like I

01:21:20.860 cited the wrong paper or hyperlinked to the wrong thing. And sometimes they were mistakes of

01:21:27.740 interpretation. I sort of read a paper out of context or things like that. But can you give

01:21:33.260 us a sense of where you are in the process of revising this first edition into a second edition

01:21:39.700 and what you've learned about any mistakes you've made and how you could quantify that?

01:21:44.580 Yeah, that's right. So since publication in 2017, a number of readers have reached out or notified me

01:21:53.560 of errors in the book. And when that happens, when they've reached out to me personally, I take what

01:21:59.920 they've said, I go to the book, I have a look at it, I will cross-reference it, I will see if it's an

01:22:04.560 error. If it's an identified error, I have a document that was cataloging all of those errors,

01:22:10.420 knowing I would then at some point, continuing to check with the publisher when the time is right,

01:22:16.480 we would then publish a second edition, which would make those corrections. So in total now,

01:22:22.400 and we've now completed that exercise, there've been 13 errors in the book that we've identified

01:22:29.540 that required correction. And I have to say that that's been very difficult for me to learn of those

01:22:36.380 errors, and it remains a disappointment to me.

01:22:40.420 Now, I was actually looking at the book the other day, because you'll probably don't remember this,

01:22:44.420 but you wrote a lovely inscription in my book for my son, because he, when he was really young,

01:22:49.900 loved all the moon pig. That's right.

01:22:51.700 Yeah, I remember it.

01:22:52.340 And as we're getting ready to move, you know, we're starting to slowly pack some stuff up.

01:22:57.980 So the book is, it's not the thickest book on the planet, but it's, you know, it's 300 to 300,

01:23:04.220 it's over 300 pages. Do you have a sense of how much data has made its way into that book in terms

01:23:10.360 of, like, how many statements or assertions have you made of fact?

01:23:14.040 Yeah, I have in sort of all of this. I've gone back and, you know, I haven't tried to catalog

01:23:19.560 every one, but the estimate, it's probably a little over a thousand facts, I think, which,

01:23:25.540 you know, I know gives some relativity to that statement. But, you know, nevertheless,

01:23:32.100 it just remains very vexing to me to have identified those errors.

01:23:38.420 Does my categorization make sense? I mean, are there, do you want to go through any of those

01:23:43.600 errors and talk about them in terms of, do you want to go through, like, I misinterpreted,

01:23:49.760 or I literally just made a mistake and I cited the wrong paper? I mean, how do you want to describe

01:23:54.100 some of those errors?

01:23:55.280 Yeah, I can describe sort of the seared into my brain at this point, near tattooed, I would say.

01:24:03.740 One example is that the book described an association between sleep and cardiovascular disease,

01:24:09.800 and it said that the study looked at over half a million individuals. Now, that study did evaluate

01:24:16.620 over half a million in terms of its collection of the data. But for that analysis, that specific

01:24:22.160 analysis, it wasn't over a half a million individuals that were involved in the analysis. It was actually

01:24:28.080 474,684 individuals, which is not over half a million. So that was an example of an error that

01:24:37.320 was in the book. I think another example of an error or a misjudgment was that in the opening

01:24:44.700 introduction of the book, sort of the general intro, what I'd stated is that short sleep can double your

01:24:50.560 risk of cancer. That was not specific writing. And I think it was poor writing on my behalf.

01:24:58.020 Short sleep does not double your risk of all forms of cancer. And you can probably tell me better,

01:25:03.880 at least at last count, I understood there to be probably over a hundred different major forms of

01:25:08.220 cancer. And short sleep is not associated with every single one of those. And certainly it's not

01:25:14.240 associated with the doubling of a risk of every single one of those. So what I should have said

01:25:18.900 is that short sleep is associated with a doubling of a risk of specific forms of cancer, of which it

01:25:26.060 is. And examples would be ovarian cancer or lung cancer. So that's a correction that needed to be

01:25:32.640 made. And I think that was poor judgment on my behalf. Another example that was brought up was this

01:25:39.240 really interesting relationship between sleep deprivation and depression. So I wrote, there was

01:25:45.580 many pages in the book devoted to this, which is in a subset of patients with depression, if we deprive

01:25:52.240 them of sleep for one night, we get this really interesting antidepressant benefit where it helps

01:25:58.620 them. And in fact, we've gone on to do MRI studies where we've tried to understand what is going on in

01:26:04.020 the brain to produce this antidepressant benefit, this paradoxical antidepressant benefit. So we've tried

01:26:10.120 to contribute to that work. And I discussed that work in the book too. That wasn't the issue.

01:26:13.720 What I stated in the book there was that somewhere between 30 to 40% of patients with depression will

01:26:21.520 respond to sleep deprivation. Now, two years after the book was published, there's been a meta-analysis

01:26:28.720 that's now shown that it's not 30 to 40%. In fact, in prospective studies, it's 45%. And so we've been

01:26:35.780 able to correct that in the book. I would just note one other very clear set of errors. The book at one

01:26:43.280 point, again, speaking about sleep loss and cardiovascular disease, spoke about a study in

01:26:48.900 over 4,000 individuals looking at cardiovascular disease. And in fact, that was true. The study

01:26:55.480 looked at, it was 4,652 individuals in that study. But when I described the data regarding cardiovascular

01:27:04.900 disease, it was only a subset of individuals. And I had said it was over 4,000 individuals.

01:27:10.740 It wasn't. It was actually only 2,282 individuals that were analyzed for that analysis I described.

01:27:19.400 So that was a clear error. And then the second part of that study that was also erroneously

01:27:24.760 described in the book was that I spoke about sleep loss or short sleep, I should say, being associated

01:27:31.160 with an increased risk of cardiac arrests. And they did not state cardiac arrests in that paper.

01:27:39.520 But what they were stating, the outcome measure, was cardiac events, not cardiac arrests. And you need

01:27:46.920 to be specific because what they were describing there in cardiac events was not just myocardial

01:27:52.540 infarction, heart attacks. It also included other things like angina pectoris that required coronary

01:27:58.780 bypass. So once again, there was an error in the description of the outcome measure. So it wasn't

01:28:05.580 cardiac arrests. It was cardiac events. And that's also been changed too. So those are some of the

01:28:11.880 errors that have been corrected in the book. And there've also been some misgivings, I think,

01:28:18.200 about the interpretation of certain data. And I can speak about those too. But those, I think,

01:28:23.700 are some examples. Does that give you some context? This sort of, it's not just citing the wrong paper.

01:28:29.720 There were just, you know, very clear errors regarding, you know, numbers of subjects or

01:28:34.080 the specific outcome variable. Yep. Yep. That definitely makes sense. And by the way,

01:28:39.480 this is near and dear to my heart for selfish reasons, right, is in the process of writing a

01:28:44.060 book and having a fact checker. I'm still reeling in the fact that you only had 13 errors because

01:28:50.460 admittedly, the first draft of my book is about twice as long as yours.

01:28:55.280 And maybe it has more facts in it. But there's one analyst who works for me,

01:29:01.260 part of my research team, who is the fact checker for the book. And his name is Vin.

01:29:07.860 And I think Vin is finding like 13 errors per page in my book.

01:29:13.460 I doubt that. But a lot of them are interpretive errors as well. It's like, well,

01:29:18.820 there's another way you could read this paper. And, you know, you said this, but it could have been

01:29:24.040 this. And so what were some of those interpretive errors that you think you've made?

01:29:29.560 One of the interpretive errors that people have had is regarding long sleep and its association

01:29:36.500 with mortality risk. So one of the things that is clear in the literature and that the book is very

01:29:42.460 clear about is that there is not, unlike REM sleep, actually, there is not a linear relationship

01:29:46.980 between sleep and mortality risk. It's not that, you know, the more and more you get,

01:29:51.160 the better and better things are, the lower your mortality risk. That is true up to a point.

01:29:56.260 But once you get to sort of nine hours and beyond, your mortality risk stops going down and it actually

01:30:02.240 starts to go up again. And this has led to the misinterpretation that, you know, long sleep is

01:30:08.380 bad for you. Long sleep will kill you. Long sleep is associated with a faster death. Now, at the time

01:30:15.280 in the book, I think the general modus operandi in the sleep research field was that it was because

01:30:23.040 those studies, which were large epidemiological studies, had not measured all of the different

01:30:27.840 disease outcomes in those studies. And what was happening was the following. When we get sick,

01:30:34.500 the main thing that we want to do is curl up in bed and go to sleep. Why? It comes back to sleep and

01:30:39.920 immune health. Because when you are sick, you release a whole collection of immune factors,

01:30:45.720 particularly the cytokines that are sleep inducing. In other words, it's a homeostatic mechanism for you

01:30:51.900 to get into the brain state and body state that the body knows is the best form of health insurance

01:30:59.360 policy that it's got, which is this thing called sleep. So the idea for long sleep was that these

01:31:05.220 patients were probably already sick and they were having sickness that wasn't being measured and

01:31:12.020 factored into these analyses that we weren't quantifying. And it was that additional sickness

01:31:17.460 that was missing from the analyses, but was still present that was forcing those people to sleep longer

01:31:23.820 than they would normally otherwise. They were trying to sleep themselves well, but whatever disease that

01:31:29.620 they had was still too powerful for this thing called sleep, and they still died. So it artificially

01:31:35.880 looked like long sleep was associated with a quicker death, when in fact it could have been that long

01:31:42.220 sleep was the thing that was desperately trying to save them, but it failed. That was the interpretation

01:31:46.620 of in the book. That was the interpretation at the time. And I think that is still true, but there's been

01:31:53.400 a second interpretation that's come onto the scene that again has, there's a good example of how I've

01:31:59.920 changed, I've updated my mental iOS, my operating system with a new belief, which is something

01:32:06.300 different, which is that long sleep is also associated with poor quality sleep. And poor quality sleep,

01:32:14.660 independent of length of sleep, is itself associated with mortality risk. So in other words, the lower and

01:32:22.900 lower the quality of your sleep, the higher your risk of death. Also, we know that when people are not

01:32:30.160 sleeping well, they typically try to stay in bed longer to try and get more sleep because the quality

01:32:35.940 of their sleep is so bad. And so once again, what's happening here now is that poor quality of sleep

01:32:42.860 is masquerading as long sleep, which is associated with higher risk of death. But it's not really long

01:32:51.820 sleep. It's the thing that's causing those people to stay in bed and look like they're sleeping longer,

01:32:56.920 which is the poor quality of sleep. And so that was, I think, something else that people had raised

01:33:03.840 about the book that I had not mentioned and has now also been rectified in the book and mentioned.

01:33:10.300 That's, I think, one example of an interpretation issue. I think another example is lack of clarity in my

01:33:16.920 writing. One of the statements that I made, which is actually true, which is not an error,

01:33:22.060 is that the World Health Organization had classified any form of nighttime shift work as a probable

01:33:27.740 carcinogen. They made that declaration in 2007 originally, then they re-examined the data, and in 2019,

01:33:36.700 they reasserted that same statement. If you want to get nerdy and technical, it's actually a

01:33:41.920 class 2A carcinogen, a probable carcinogen. The issue here is that it was actually,

01:33:50.080 some people Googled this and said, no, what I'm finding is that it's actually the International

01:33:54.400 Agency for Research on Cancer, the IARC, that made that statement, Matt. It's not the World Health

01:34:02.240 Organization, so that's an error. You got that wrong. But it turns out that the IARC, the International

01:34:08.360 Agency for Research on Cancer, is a part of the World Health Organization. And if you look at the

01:34:14.220 little sort of info card that was published in that press briefing, you'll see the little icon of

01:34:19.540 the World Health Organization in the top corner of that. So that's not people's fault. That's natural

01:34:26.240 for them to say, you made an error, Matt. It's what I should have said in the book is not just that the

01:34:31.520 World Health Organization. I should have said it's the IARC, which is part of the World Health

01:34:37.880 Organization, which together made the statement of it being a carcinogen. So that's another place where

01:34:46.260 I've been able to, you know, there's a long-form blog that was also published. It's over 8,000 words.

01:34:53.120 People can find it on WordPress. That describes not only all of the errors that are there in the book

01:34:58.500 that we've just discussed and all of the others, the corrections that will be made. But I also go

01:35:03.520 through any of the misgivings or the sort of the changes in interpretation that people have had

01:35:09.420 regarding what I discuss in the book. And I walk them through a collection of detailed science to a

01:35:15.780 level that I wasn't able to do in the book. And it's pretty nerdy science, but it's there because I

01:35:20.540 felt as though not only in response to this did I want to make all of the changes in the book because

01:35:27.020 I should just note, I see it as a critical part of good scholarly conduct that you correct any errors

01:35:33.520 that are identified. And I'll always hold myself to that standard. And that's what's happening in the

01:35:39.140 book right now. I've sent all of those changes to the publisher, and they should be coming out in a new

01:35:44.520 second edition of the book. But the second thing I wanted to do was sort of go into detail and clarify some

01:35:50.140 of these interpretations and at least give you all of the additional science supporting why I offered

01:35:56.260 those specific interpretations in the book that I did above and beyond simply describing the errors.

01:36:02.240 So obviously, for people who don't necessarily want to wait for the second edition to come out,

01:36:07.020 the long blog post you wrote, presumably we can easily link to that.

01:36:13.500 That's right.

01:36:14.160 Okay. So that'll be in the show notes as well.

01:36:16.840 Yeah. It's still nevertheless, you know, it's remained difficult. You mentioned you and other

01:36:21.880 authors and I said about Michael Pollan's book, Caffeine. And when I first started to realize

01:36:27.880 some of these errors, and Michael, there's a collection of, it's strange, there's a collection

01:36:32.360 of Bermuda Triangle of writers here in Berkeley. There's sort of Michael Pollan, there's Michael

01:36:37.420 Lewis. The collection of us know each other and we'll occasionally go out for dinner and

01:36:41.320 Michael's become really good and he's been very helpful to me over the years. And I wrote to him and

01:36:46.520 said, you know, I need to sort of speak with you. I didn't tell him what was happening. And I said,

01:36:52.760 can I take you out for dinner? And what's nice about Michael, because he's a food writer.

01:36:56.460 If I try to book a restaurant, you know, at a certain time, they'll say, oh, I'm sorry,

01:37:00.080 we're full. And then Michael calls up and he gets a table. So we, I took him out for dinner.

01:37:05.580 And early on in the dinner, I reached over, you know, very quietly because I was ashamed.

01:37:11.740 And, and I sort of said to him, look, Michael, that I need some help because I've identified

01:37:16.640 these, you know, errors in the book. And I was hoping that the only witness to that statement

01:37:21.380 at that point would be my salad that was sitting underneath me. And I was expecting to, you know,

01:37:26.740 see what he said. And, you know, he, he was just seemed to, he just brushed it up. He just said,

01:37:32.300 you know, well, gosh, for my probably most scientific book, which is Omnivore's Dilemma,

01:37:36.200 I don't think it was until the third edition until I ironed out, you know, all of the errors.

01:37:41.740 And I have to say for about an hour during that meal, I felt some degree of, you know,

01:37:49.460 lowered anxiety about it. And then, you know, probably at least another hour later after

01:37:56.380 leaving the restaurant, I went back to just feeling very, you know, very vexed and upset about it. So

01:38:01.880 anyway, that's the correction. Thanks for letting me speak about that, by the way. And thanks for

01:38:07.240 bringing that up. It's very good of you. Thanks.

01:38:09.760 If anything, it's made me not want to write my book.

01:38:12.320 Oh, I hope not. We all need it. We all want it. Write that book.

01:38:16.440 We'll see. Well, Matt, I'm sorry we didn't get to do this in person,

01:38:20.560 but the next one we will do in person.

01:38:22.700 Let's make sure.

01:38:23.520 Yeah, yeah. We're doing the next one in person regardless,

01:38:25.080 even if we have to sit in masks across from each other.

01:38:27.160 Well, if people, you said that, you know, people have a sort of a love-hate relationship

01:38:31.400 with sleep, a like-dislike relationship. I'll be left to see about your listeners as to

01:38:36.260 whether they want me back. I think my general personality is kind of like a dislike-dislike.

01:38:42.160 There is no like-dislike. I think some people have described my personality as the best

01:38:46.380 prophylactic known to man. So yeah, if you will have me back and if your listeners will have me

01:38:50.940 back, then I would be delighted. And if that's the case, guaranteed we do it in person.

01:38:56.000 All right, Matt. Thanks so much.

01:38:57.940 Thank you again for having me. Really appreciate it. Take care, Peter.

01:39:00.540 Thank you for listening to this week's episode of The Drive. If you're interested in diving deeper

01:39:06.040 into any topics we discuss, we've created a membership program that allows us to bring you

01:39:10.440 more in-depth exclusive content without relying on paid ads. It's our goal to ensure members get

01:39:15.940 back much more than the price of the subscription. Now to that end, membership benefits include a

01:39:21.280 bunch of things. One, totally kick-ass comprehensive podcast show notes that detail every topic,

01:39:26.720 paper, person, thing we discuss on each episode. The word on the street is nobody's show notes rival

01:39:32.220 these. Monthly AMA episodes or Ask Me Anything episodes, hearing these episodes completely.

01:39:38.480 Access to our private podcast feed that allows you to hear everything without having to listen

01:39:43.440 to spiels like this. The Qualies, which are a super short podcast, typically less than five minutes

01:39:49.160 that we release every Tuesday through Friday, highlighting the best questions, topics,

01:39:53.100 and tactics discussed on previous episodes of The Drive. This is a great way to catch up

01:39:57.600 on previous episodes without having to go back and necessarily listen to everyone.

01:40:02.460 Steep discounts on products that I believe in, but for which I'm not getting paid to endorse,

01:40:07.480 and a whole bunch of other benefits that we continue to trickle in as time goes on.

01:40:11.740 If you want to learn more and access these member-only benefits, you can head over to

01:40:15.140 peteratiamd.com forward slash subscribe. You can find me on Twitter, Instagram, and Facebook,

01:40:22.180 all with the ID peteratiamd. You can also leave us a review on Apple Podcasts or whatever podcast

01:40:28.700 player you listen on. This podcast is for general informational purposes only and does not constitute

01:40:34.260 the practice of medicine, nursing, or other professional healthcare services, including the giving of

01:40:39.740 medical advice. No doctor-patient relationship is formed. The use of this information and the

01:40:45.460 materials linked to this podcast is at the user's own risk. The content on this podcast is not intended

01:40:51.640 to be a substitute for professional medical advice, diagnosis, or treatment. Users should not disregard

01:40:58.140 or delay in obtaining medical advice from any medical condition they have, and they should seek

01:41:04.100 the assistance of their healthcare professionals for any such conditions. Finally, I take conflict

01:41:09.720 very seriously. For all of my disclosures and the companies I invest in or advise, please visit

01:41:16.300 peteratiamd.com forward slash about where I keep an up-to-date and active list of such companies.

01:41:39.720 Thank you.

The Peter Attia Drive - August 31, 2020

#126 – Matthew Walker, Ph.D.： Sleep and immune function, chronotypes, hygiene tips, and addressing questions about his book

Episode Stats

Length

Words per Minute

Word Count

Sentence Count

Misogynist Sentences

Hate Speech Sentences

Summary

Transcript