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 0.99

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 0.62

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 0.88

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 1.00

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 1.00

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 0.51

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, 0.69

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, 1.00

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 0.99

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. 0.80

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 0.83

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

Summary

Transcript