Written by Christopher Kelly
March 10, 2020
[0:00:00]
Christopher: Well, Greg, thank you so much for joining me this morning. I am very excited to talk to you once again. I just published the last of a series of four articles that you wrote for Nourish Balance Thrive that we're very grateful to have, super useful when working with clients. The last article was Sleep for Athletes: Are They a Different Breed? You can find the link for that in the show notes for this episode and then also, the link to the previous podcast that we recorded as well. That stuff's so useful. Thank you so much for creating that content for me, thank you so much.
Greg: Pleasure, I'm glad that you found it useful.
Christopher: Yeah, absolutely. Can you give us a quick update? What have you been working on lately? What's new for you?
Greg: Well, I don't want to bore people with details of my personal life, but I am doing various bits and pieces at the moment. One of the projects that I've been working on recently is helping a friend formulate some nutritional products which are designed to be ergogenic for endurance athletes. I'll leave it at that for the time being.
Christopher: Oh, can't you say more than that?
Greg: Cliffhanger. Otherwise, I'm giving two talks in June, related to cancer. One of them is about how circadian system disruption may dispose people to cancer and how we can use an understanding of circadian biology to improve the treatment of cancer patients. The other is about how sleep may be an unappreciated contributor to the development of cancer and how to help people sleep better if you have cancer. Because my background isn't in cancer biology, I've spent a lot of the last month or so making it the research on those subjects and then compiling slides for those two talks.
Christopher: Do you mind me asking, how do you get funded to do that? That's fascinating to me that you can afford to spend that time. Who's paying for that?
Greg: Yes, so I'm giving a talk at Metagenics. They're a supplement company based in Australia. Some of the public speaking that I do isn't paid public speaking. Normally my transport will be covered, and my accommodation, but I'll sometimes give talks for free, depending on the circumstances. In this particular instance, I am very lucky to be paid to go out there and speak, so I'm happy putting some time aside and spending lots of time trying to put together something which, hopefully, makes sense because I know there will be smart people in the room who know a lot more about some of the nuances of cancer than I do. There's quite a prestigious lineup of speakers, myself aside, so I'll try to make sure I put my best foot forward.
Christopher: That's excellent. Well, that's fantastic, good for you. I'm really excited about that. Where is the talk, and will it be available online?
Greg: Good question. I don't know actually. My gut feeling is that it won't be. The talk is in Brisbane, and it's at the start of June. I think it's probably the first weekend of June or so. It's three days or so. I want to say it's maybe June 5th to 7th. If people are interested and they just search for Metagenics 2020 Congress, they should dig it out.
Christopher: I know we've got lots of listeners in Australia because I hear from them occasionally, but that sounds like a really good reason to go to Australia. You're going to be employing some of your own strategies to overcome the jet lag. Of course you are.
Greg: Yes, well, I'm actually flying to the States in two days, well, not two days time. I'm flying to the States tomorrow now and so I'm going to have to try and put some of them to the test. I've got a lot of travel coming up, so I need to take my own medicine.
Christopher: Excellent, glad to hear it. Well, let's get into the purpose of this interview. One of our clients sent us this fantastic article, and the article is a critique of Matthew Walker's now famous book, Why We Sleep. I have to admit I hadn't read the book until I read this article. Maybe the article had the opposite of what Alexey intended.
I listened to Matthew Walker on a couple of podcasts, and I was really put off by him and not at all encouraged to read the book. I must admit that was very much a system 1 decision. Are you familiar with Daniel Kahneman's Thinking, Fast and Slow, systems 1 and 2? Have you read that book?
Greg: Yes, I read it when it came out.
Christopher: A long time ago now
Greg: Yes, and I found it very dense, slightly hard work.
Christopher: It is dense.
Greg: Yes, and obviously he's brilliant. He's utterly brilliant, and I really enjoy listening to him speak, but I don't recall that much of the book. I remember some of the basic premises and the fact that you've got these two different ways by which you think. You've got system 1 and system 2. One of them is about knee-jerk responses and the other is about more deliberative thinking. Is that right?
Christopher: That's exactly right, yes. So, my knee-jerk response to listening to Matthew Walker talk about his new book was I don't want to read that. I think there's a lot of unknowns here, and I don't detect any humility in this man's voice. I'm thinking he sounds overconfident, and I didn't hear him say, "I don't know," even once. That, for me, is very much a turn-off.
[0:05:01]
Greg: Interesting.
Christopher: I think I've learned a lot from Josh as well, our neurologist. I know that you're familiar with his four quadrant model, and you've heard him give the How to Win at Angry Birds talk. I think there's a general, maybe people trying to explain themselves too much. Do you really need to know all the details at the source code level in order to perform this behavior? I don't think so.
You might be good enough just to say, well, humans and their ancestors have been walking around under the same light-dark cycle for millions of years and if sleep wasn't important, we probably wouldn't be doing it. Besides, it feels good, so you should sleep. That's my entire version of why we sleep. That's all you need to -- and you should resist the temptation to try and explain yourself any more than that because you're probably going to be wrong.
That's what I'm worried about with Matthew Walker's Why We Sleep, was that it would just be riddled with things that either were wrong or later turned out to be wrong. I'm starting to wonder whether that maybe is the case. First of all, that's a good question. Have you read Matthew Walker's Why We Sleep? Do you know him as well, is a second good question.
Greg: So, first question, I had a physical copy of the book, but it was rented from a library. I read the first half or so and then I had to return it, for whatever reason, and then downloaded the audio book on Amazon and finished listening to the rest of it.
Christopher: Okay.
Greg: I did so when the book came out and didn't think too much more about the book. Obviously, it's done a massive amount to bring the subject of sleep to the fore of public discourse, which I think is certainly a good thing. Then of course the Guzey article surfaced, and there has been quite a lot of discussion about the some of the potential shortcomings of the book, off the back of that.
In relation to your second question, I haven't met Matthew Walker personally. I know several people who do know him relatively well. Because I haven't met either of these people, I'm not really willing to say what I might think about them because the way that I feel about them might turn out to be wrong. For that reason, I'm very happy speaking about the claims that are made in Guzey's article and in the book.
I just want to be perfectly clear that I haven't read the book recently. I really respect some of the work that Matthew Walker has done, and I appreciate the fact that he has spent so much time doing this type of public-facing work in recent years.
I also respect the fact that Guzey went to such lengths to produce this extended article. He clearly spent a lot of his own free time doing it. I think it's a really good thing. I think that Walker has recently responded to the article in various different formats, so clearly, it's on his radar and hopefully, he'll improve the book off the back of some of Guzey's comments.
Christopher: [0:08:10] [Indiscernible]. Well, I thought that Section 6, the summary might be quite a nice outline for some questions to ask you. Maybe we should work backwards actually, from 6 through 1, rather than the other way around. Based on what you just said now, would you say that the book is indeed a scientifically accurate intervention? It's a strange word to use, intervention, isn't it?
Greg: It's a scientist word. The problem with being a scientist is that you end up using all sorts of silly words like elucidate and intervention and various other things when you could just be complaining that you also tend to speak in a very passive voice and make your work as impenetrable as possible.
Do I think it's a scientifically accurate intervention? No, I don't think it is really. I'm not sure that's how I would describe the book. One thing that's worth pointing out is the -- I think, in part, because much of Matthew Walker's research is focused on the effects of sleep loss, the book itself does take a relatively sleep loss-centric approach to explaining how sleep difficulties can lead to all sorts of maladies.
When I think about sleep, I think of several different dimensions of sleep health. I know that Matthew Walker thinks in these terms too. I don't recall the specifics of this, but I think that I've heard him say that there's another book in the pipeline which will consider some of these other factors too.
If we think about sleep health and what it is, then there's sleep duration. There's sleep quality which is quite difficult to assess, but you can think of things like sleep continuity, how your sleep continues over the course of the evening, whether it fragments, for example. You can think about sleep timing, and you can think about the variability of all of those different dimensions too. I mentioned those things specifically because they're all somewhat predictive of various health outcomes and of all-cause mortality too.
[0:10:18]
There's a great article by Daniel Buysse who is a legend in the field of sleep research, which discusses these different dimensions. We just had a really hard time coming up with what good sleep is, defining it, and actually defining what sleep is too. That might resurface later in this conversation but, yes, Walker does take quite a sleep loss-centric approach to sleep problems.
It will be nice to see a subsequent version of the book or a subsequent book which is slightly more rounded, which isn't to say that the book doesn't touch on those other dimensions too, because it is a very broad book. I found it thought-provoking at times. I really enjoyed parts of it, but as we'll get to, I do think there are various inaccuracies in it, and I think that Guzey points those out effectively.
Christopher: Actually, let's diverge from the summary for just a moment. Do you think it's helpful overall for people who are suffering from some type of insomnia? That's the criticism that we've heard from the NBT forum. Indeed Guzey talks about that in his article. Is it helpful for someone who can't sleep, knows they should be sleeping, to listen to a 13-hour monologue on Audible, talking about all the bad things that are going to happen to them when they don't sleep enough?
Greg: No, I don't think it is, in short, because what tends to happen is, we discussed on a previous podcast, the most previous podcast, is that when people sleep poorly, they tend to put more pressure on themselves to sleep well. They have sleep performance anxiety, and exposing themselves to research showing the negative consequence of poor sleep can exacerbate that.
The question, as I've mentioned to you previously, is how can we share this information in a way that supports people's ability to sleep well even if the consequences of poor sleep are so detrimental? That's an ongoing challenge that anybody who is involved in roles like ours, faces, and it's not one that I feel I have a solution to yet.
Christopher: Well, I can make a suggestion. I've learned so much about the -- I know, here's another grandiose science title -- the transtheoretical model of behavior change, which is Simon's published area of expertise. This is really a book for pre-contemplators, those are people who either don't know they have a problem or they don't care. You've got to die of something. This could be a really good book for those people.
If you're already in the action phase or the relapse phase, meaning, you've already done something about it, you already bought into the idea that sleep is important, you've already taken measures to try and improve your sleep quality or quantity; then it's probably not going to be helpful to read about all the bad things that are going to happen to you when you're already trying. Do you see what I mean?
I don't really know how you only -- probably something to do with the title, right? Maybe a subtitle is like, The Most Underappreciated Thing You're Not Thinking About, or something like -- something that appears to someone that has not really thought about the importance of sleep before.
Greg: It will be interesting to find out more about the demographics of people who have read the book, whether those people are people who are already very interested in sleep or don't know so much about sleep. Obviously, people from both categories will have read the book, but be curious to find out more about how that breakdown has taken place.
Christopher: Should we get into some of the specific mistakes that Guzey thinks that Walker made? Number five, misrepresents the National Sleep Foundation sleep recommendations and uses them to misrepresent the number of adults failing to get the recommended hours of sleep. This is Section 5. Also seems to invent the WHO's sleep recommendation. That, for me, is very easy to fact check.
That's fascinating. It's fascinating to see how science works with citations. You're now starting to see people cite Walker who invented this thing. You can't just fix it in Why We Sleep. You can't just release a second edition and say, "Actually it was the CDC that said that, not the WHO." You've got to somehow go back and fix all the things that are linking, citing the Why We Sleep. That's a real problem, right? Would you agree that that was just wrong?
Greg: Yes, I completely agree. I think that Walker has since accepted that that is the case and that he was conflating the WHO with the CDC. If you go to the CDC website, as Guzey points out, the page that declared the insufficient sleep was an epidemic, is no longer there. It's now a defunct page, so they've toned down their own language too. As you say, citing second resources is frowned upon by scientists, and it is a problem. As Guzey points out, Walker has already cited the book himself, multiple times.
[0:15:10]
There are also indications in the article with academic misconduct, would be one way to put it. For example, he points out that Walker seems to have basically double published in both Neuron and The Lancet, which seems quite egregious to me, especially given that they're both very visible, high-impact journals. So, I wonder if The Lancet and Neuron were both aware of the fact that Walker would be publishing something similar in both journals. If that wasn't the case, then I [0:15:43] [Indiscernible] to imagine how somebody thinks they could do that and get away with it.
Christopher: Get away with it. That was going to be one of my questions for you, should he have even published this book in the first place and made that claim that it's scientifically accurate? Because I thought this is why they didn't like professors publishing books before they were retired was because, well, it has not been through the peer review system. Now, as you say, Walker is citing himself in a book that never went through the peer review system. Is that not a problem in itself?
Greg: Yes, I think so. It's a problem that, I suppose, is inherent to all books like this. One encouraging development is that you now have some companies that are popping up that are actively fact checking scientific books. I know Stephan Guyenet, for example --
Christopher: Yes, I was going to say Stephan Guyenet has got his -- I forgot what it's called now. We'll find his link --
Greg: Is it Red Letter something?
Christopher: Yes, that's it. You've got it. All these things, Elaine does a super job of finding all these things and linking them in the show notes, so it will surely be there.
Greg: Yes, but the problem -- and I know that Stephan has acknowledged this in public -- is that it takes a long time to get these books --
Christopher: What's the secret?
Greg: Yes, you've got to fact check everything.
Christopher: You have to really piss someone off before they'll go through it with a fine-toothed comb and produce an article of this quality.
Greg: Guzey, of course, focused more or less exclusively on the first chapter. Certainly --
Christopher: What did you think? Do you think the first chapter was particularly problematic? Because I feel like the first chapter was explicitly fluffed up to increase the importance of the book in the hope that you would invest more time in it. This first chapter seems to be very different from the rest.
Greg: Right.
Christopher: Would you agree with that?
Greg: I haven't read the first chapter and the rest of the chapters for long enough to be able to say that. I don't know. I'd be interested hearing more about your thoughts on that. One thing that I point out, and you know this at least as well as I do, and there are people better qualified to discuss this, but people tend to remember things at the start of things and things at the end of things, so-called primacy and recency effects.
So, if you're going to write an article or a book, you want to put the information you want people to remember, at the start and in the end. So, if there is a bit of embellishment at the start and embellishment at the end, it might be done so with a view to really trying to drill home the importance of the subject. So, I don't know is the short answer. If it does seem different from the rest of the chapters then I would view it as being understandable in the light of those effects.
Christopher: Right. It's just the tone is different. It's almost as if an editor or a publisher said, "You know what, Matthew, can you just spice this up just a bit?"
Greg: Interesting.
Christopher: I don't know. It's speculation, obviously, on my behalf, but it did feel like it was somewhat different. Having said that, even I could spot problems with the first chapter. He talks about -- he goes straight into the chronotypes like it's all done and dusted. My understanding, from listening to you and Satchin Panda and Bill Lagakos, is that chronotypes are controversial, and most likely they are an explanation or explained by the environment rather than genetics, in many cases. He goes straight into the sentinel hypothesis. Do you remember that? That's controversial as well, right?
Greg: Yes, sentinel hypothesis really has been discussed by David Samson and not many other people. It's certainly not something that we understand that well. I know that I did a podcast with Tommy maybe a year or so ago about chronotypes and exclusively about chronotypes. If people want to find out more about that subject then --
Christopher: Yes, we should definitely refer them to that.
Greg: Yes, it's quite dense, that conversation, but it contains at least as much information as 99% of people listening to this would ever want to know about chronotypes.
Christopher: I loved it. I thought it was fantastic. But then I would say that. I'm biased. What do you think about this National Sleep Foundation sleep recommendations and uses them to misrepresent the number of adults failing to get the recommended hours of sleep? Are people not sleeping enough?
Greg: I think the points that Guzey makes are absolutely spot on, and the line of logic that he follows is congruent with what makes sense to me. Interestingly, I've read his article twice. I read it once maybe two-and-a-half months ago, something like that, and then I read it again today. The article has changed since then.
[0:20:10]
Christopher: I think there's a revision history on it. Yes, he's definitely clear about the fact that it has been revised over the past, which is as it should be. I think every paper should have a revision history.
Greg: That's the problem with books. They're trapped in amber, at least temporarily. For that reason, it's much easier being critical of people in some sort of online medium than it is in the printed, written word, I suppose, because you can just go back and just edit out anything that you think was silly, in retrospect.
Christopher: Right, and that's okay as long as there's the revision history. I want to see the dates. I want to see how you changed this. That's the important thing that's missing.
Greg: Absolutely, and to be clear, I listen to various things that I have said and think, I completely change my mind about that. I'm not claiming to be immune to that by any means.
Christopher: Of course the important thing is that you're willing to say that, put your hand up and say, you know what, I'm not so sure about that. In fact, I'm probably wrong; be very specific and be the first person to hand up rather than waiting for someone else to point out that you're wrong.
Greg: Yes, but anyway, to return to your question; yes, I agree with Guzey. He mentions a couple of papers, [0:21:22] [Indiscernible], that indicate that sleep duration may have increased in recent years in some parts of the world. My impression is the data on the subjects are very mixed. For example, there's some evidence that sleep duration has declined in adolescence in recent years, but findings for adults are less conclusive. They vary a lot according to where the studies are done, and both decreases and increases in sleep duration have been reported.
Just as an example of this, David Dinges from University of Pennsylvania did some work on this a couple of years ago, and he used the American Time Use Survey data which is fantastic resource. Basically they have people report what they do, minute by minute, and they have lots of people do this. What they did was they looked at the data that were collected between 2003 and 2016. This was data for more than 108,000 people. At the same time, they looked at Google trends for monthly searches for sleep during that same time period and also looked at publications on things like sleep duration during that time period.
What they found was that the time when people reported being asleep increased by about 1.4 minutes per 24 hours on weekdays and by about 0.83 minutes per 24 hours per year on weekends. So, over the course of a year, that would translate to about 83.4 hours more sleep per year just on weekdays. The indication there is that sleep duration in the US might be increasing. Interestingly, that was found for employed people, for retired people, full-time students and also for full-time high school students.
An important point of course is that sleeplessness also increased during the survey years. It might be that people report spending more time in bed, but the quality of their sleep is declining. That's an important element that's missed in this conversation.
Christopher: Do you trust that type of self-reported data? As a data scientist, I know that when you ask people stuff like that, they either don't know or they lie. Like, how much salt did you have on your breakfast this morning? You either don't know or you lie. It has been consistent in every -- Netflix movie ratings, you ask people what type of movie they like, they lie. You have to wait until they've seen the movie, and ask them to rate it rather than -- or even better, somehow quantify it, measure it directly without asking the person their opinion at all.
Greg: Yes, and I know that Guzey states that self-reported data are notoriously unreliable. What I would say is that, yes, self-reported sleep data of course aren't ideal, but that doesn't mean that we can just disregard them as not being potentially insightful. It is a problem in sleep research that we don't have standardized cutoffs that are used to identify short sleep, for example. In some studies, maybe short sleep is five hours or less. In other studies, maybe it's six hours or less. That confuses all of the meta-analysis, for example, and it makes it hard to interpret the data.
Of course, the ways that people ask about sleep will influence people's responses. Maybe they ask them about sleep but people actually end up reporting their time in bed as opposed to their sleep. Often in these large-scale studies, sleep is not the main focus. Maybe they're basing their sleep analysis on a single-question item within a very large questionnaire. So, all of that is problematic, but I do still think that self-reported data can be very useful. I think they're actually especially useful for people who have sleep difficulties. I know that we've spoken about this a little bit, Chris, but --
[0:25:13]
Christopher: The sleep diary.
Greg: Yes, exactly. Sleep diaries can be massively useful especially for insomnia patients. Wearables and the data they present can actually make sleep worse in some people. If somebody reports not feeling like they're sleeping very well and then their devices just reinforce that notion or maybe provide data suggesting that their sleep is even worse than they thought it might be, then maybe that will put fuel on the fire.
Christopher: Right, maybe it's true that sleep has got -- the Heisenberg uncertainty principle applies where you can't measure it without affecting it.
Greg: Yes, I think that's absolutely the case. So, yes, sleep is difficult to measure. Yes, there are problems with just using self-report, but self-reported data can still be useful.
One point that Guzey makes is that typically people underestimate their sleep duration, and that is true. He cites a study, for example, that compares self-reported sleep duration with sleep which is assessed by actimetry. These are just wrist-worn devices that use movement as an indication of whether somebody is asleep or awake. The limitations of that are obvious.
Basically it's like a FitBit, but the devices that are widely used in research are actually probably slightly more archaic than some of the modern wearables because if they're just using accelerometry then they're not collecting data on pulse rate, for example, which will tell you something about sleep staging. For that reason, I think it's possible that some of the new devices might actually be better than some of the research-grade actimetry devices.
Anyway, with that said, actimetry of course has its limitations. It generally gives okay estimates of sleep duration. If anything, it tends to score people as being asleep when they're just lying in bed, still, understandably. The study that Guzey pointed out indicated that it seems to generally indicate that people will overestimate their sleep, but that study was of a single adult population. There are people who tend to underestimate their sleep. For example, people who have insomnia will often report feeling like they slept two hours when, if you assess their sleep using polysomnography, which is the gold standard, then maybe they got four hours of sleep, for example.
Christopher: Right, and Walker does talk about that in the book. A related question is, so am I right to conclude that we just don't know whether people are sleeping less than before? That was what I got from that.
Greg: I think that that's how I feel, yes.
Christopher: Okay, and then a related question, so if you're not sleeping, everyone talks about eight hours, say you're only sleeping for seven hours, do you think there's anything to be gained by forcing that extra hour of sleep?
Greg: It depends how you go about it, in short. Forcing sleep is never the way to go about improving your sleep.
Christopher: Perhaps that was a poor word choice.
Greg: Yes, but I understand what you're saying. What I will say is that there have been lots of sleep extension studies, and those studies haven't been entirely consistent, but they have generally indicated that sleep extension will have various positive effects in people who habitually restrict their sleep.
I'll just mention a couple of these but as one example, there was some work that was published in 2015 that looked at men who habitually restrict their sleep from Monday to Friday. They then visited the lab on two Fridays. At one visit, they're allowed six hours in bed for three nights, which mimics their working week; and the other visit, they're allowed ten hours in bed. That's the sleep extension condition.
They found that after their sleep extension condition, their incidence sensitivity was 45% higher, and their testosterone was higher too. There has also been work showing that sleep extension among people who habitually restrict their sleep may have some positive effects on immune function. Specifically, they looked at 24-hour plasma IL-6 levels, and they found that that marker information increased during sleep restriction but returned to baseline after sleep extension, so, recovery sleep.
There has been work looking at cognitive function in people who have obesity. That has been reported that relatively long-term, moderate sleep extension improved overall cognitive function. There has been work looking at sleep extension appetite, food intake.
One of my colleagues who did her PhD at the same time as me, [0:29:39] [Indiscernible], did some work in 2017 and found that during a sleep extension intervention, the participants slept nearly an hour longer, and they consumed less free sugar. Among the people who actively reported their diet, the people tended to consume fewer carbohydrates and less fat too.
Also, there has been some work looking at sleep extension and exercise performance. I've probably mentioned this in the previous podcast, Chris, but perhaps the most highly cited of these was done on the Stanford University men's varsity basketball team. Cheri Mah and her colleagues had these players extend their sleep for five, seven weeks, and during this time, the players slept substantially longer. Their reaction times hastened. They completed a multidirectional sprint about seven-tenths of a second faster, and their accuracy in both free throws and three-point field goals, both increased by 9%.
[0:30:34]
So, there are various lines of evidence that indicate that if you take people and you appropriately prolong their time in bed, give them more opportunity to sleep if they have been habitually restricting their sleep, for example, by waking to an alarm then they will experience various improvements.
Christopher: Yes, so you just got to the core of my question then that I was very vague about, is the -- so, if you're allowing, say, eight -- I don't know. Let's just say we're only asleep for seven hours, and your sleep efficiency is close to 100%. You wake up feeling refreshed. Is there any reason to try and extend that sleep scenario, to try and get eight hours even though you're waking up after seven, feeling refreshed? Your sleep efficiency is really good. Do you think there's any benefit there?
Greg: I think it's very hard to say at this time. It would depend, in part, on how a person's waking up. Let's say that somebody has been doing that for a long period of time. They do wake through an alarm. The alarm helps them build lots of sleep pressure during the day which consolidates their sleep overnight such that the quality of their sleep is high, but there is this residual sleep pressure because these people do curtail their sleep.
If you took that person and you prolong their time in bed by having them just stop waking to an alarm for a period of time then maybe they would have some small improvements. If you take somebody who doesn't wake to an alarm and they routinely get seven hours of sleep, the quality of their sleep is high, and you then ask them to spend longer in bed, I wouldn't imagine that that would have any beneficial effects on any of those outcomes that I mentioned.
Christopher: Okay, that's good to know. What would you say is the best way for people to assess this by themselves? I think a part of what came out of Josh's How to Win at Angry Birds talk is that the randomized control trial is the wrong way to acquire knowledge for a quadrant one game level intervention. I don't care what happened to a bunch of 26-year-old, college-aged students on the other side of the world. I need to acquire this knowledge for myself. I want to be agile. I want to know where I am and then I want to make a small change, and then I want to reassess. Do you think the wearable devices -- how would you go about, if you were working with me as a client, how would you suggest that I assess where I am so that I know what has happened once I make some small changes?
Greg: Yes, it's a really difficult problem to solve. I just want to preempt my answer by saying that. If you came to me specifically with a sleep problem that you wanted addressing, then I'd use many of the techniques that we discussed previously. I also listened to your podcast with Ashley Mason which was great, so I think those are good resources for people to refer to if their problem is sleep specifically. Now with that said, if we think about somebody's general health, then there's a massive opportunity for somebody to come out with a product which systematizes this process that you're alluding to.
Because right now, I think there are lots of people who are interested in quantifying self who are trying to go about this on their own. Maybe they don't have the requisite background to make sense of the data, and maybe they're making some inaccurate inferences for that reason. Or maybe they're engineers, and they love numbers. They really want to try and fine tune everything. They treat their bodies like a machine of sorts, and they make some progress.
With that said, what I would focus on are things that we know are both amenable to accurate measurement and predictive of various health outcomes in a way that isn't too redundant. What I mean by that is that if you look at various different markers of general health, then if you use the right type of study, you can identify them as independent predictors of certain health outcomes.
So, when we think about those in relation to general health, those will be things like self-reported health. They'll be things like VO2 max.They'll be things like maximum strength. They'll be things like some marker of glycemic control, blood pressure, a few others too.
I think the issue is that now that we have lots of different tests that people can do, genetics test, various different Omics tests from proteomics to metabolomics to transcriptomics; people will look at that and they think that it might be possible to make some inferences about what's going on based on those data. I think that the scientists who are trying to work on those issues don't feel the methods are quite ready for prime time yet. That's my impression.
[0:35:07]
So, what I would do is I would have people periodically track certain health outcomes on a regular basis. The frequency of that will be person-dependent. If it was somebody who was in good health then it could be relatively infrequently. If it was somebody who was acutely dealing with a specific issue, for example hypertension, then maybe it will be more regular.
I will look to standardize the time of day of the assessments, for example, the conditions preceding the assessments, including things like sleep the night before, caffeine, food intake, that type of thing. Then I'd have people go through a battery of tests which are person-specific.
With that said, there are certain things that I would probably have all people track from time to time. One of them actually, as I mentioned, would be self-reported health because, I know it sounds very crude, but self-reported health based on the untargeted [0:35:59] [Indiscernible] that have been done so far has emerged as one of the strongest, if not the strongest predictor of future health outcomes. There are very simple questionnaires that are available online that people can use. For example, there's WHO-5. There is the short form 12 and short form 36. I think that's what they're called.
Christopher: Yes, I'm familiar with those, yes.
Greg: Yes, so people can look up those. Maybe they complete those a couple of times a year, for example, and then around the same time, maybe they do some exercise testing which could be sport-specific. I'd also look to have certain exercise tests in there which are relevant to everybody. For example, that could be a measure of relative strength, your strength relative to your body weight. That might be, for example, the number of pull-ups that you can complete if you are relatively lean and relatively strong.
It might have some measure of maximal strength in there done in a safe exercise. I wouldn't necessarily have somebody do one-rep max of anything because I just don't think that's necessary unless you are competing in a strength and power sport for which that is the focus, so, powerlifting or weightlifting, for instance. I would have somebody do some sort of assessment of cardiorespiratory fitness. That needn't be a VO2 max test. There are very simple tests that people can do. Different run tests of certain distances or duration can be very predictive of VO2 max and of various health outcomes, and then things like blood pressure.
Where you guys come in, Chris, is you can use some of these more sophisticated tests, for example, blood tests which are very useful, have been used for eons. We know how to interpret those data, and we also know something about the limits of the quantification of data or the variability of the data from one sample to the next and so on, such that people can probe a little deeper if they're having specific issues. I've, of course, tracked things like body weight. I wouldn't necessarily track body fat, but I would probably have people look at certain girths. You know that, if for example, your body mass is increasing but your waist measurement is decreasing then you're probably doing things right.
Overtime, the emphasis that you place on the individual metrics that you're tracking might shift, but you should have a relatively comprehensive picture of how those things are changing. Then hopefully in years to come, some of these more sophisticated tests will produce data which are easy to act on and also will become more accessible to more people because they're more cost-effective. That was a very, very roundabout answer, Chris. I feel like you could probably have given a much more coherent one than that, based on what you do, day to day.
Christopher: No, that's great. It's really great to hear you confirm my worldview because we've been doing the health assessment questionnaire almost since the inception of Nourish Balance Thrive, and the questions that we ask our clients are the type that you'll see on the SF-36. I got the questions from the NIH PROMIS, I think it's called, databank.
I cherry picked a bunch of questions that I thought were relevant for me as an athlete, and we've not changed those questions over nearly six years now. We've shown that the fancy lab tests are actually somewhat predictable using machine learning techniques, and the data from the subjective questioning, so I think it's super helpful. Yes, people don't generally lie on those questions, so, yes, super helpful. That's really helpful. Thank you, Greg.
Greg: Yes, no worries. Actually, just to add a comment about sleep. I'd also probably track sleep for a couple of weeks at a time using a sleep diary. If you don't have any issues with your sleep then you might just do this once a year. If you're struggling with a particular sleep difficulty at the moment then you might do so for a longer period, for maybe a period of two, three months, initially, until your sleep problem is solved.
Christopher: Right, so what do you think about this very specific claim that Walker made in the book, the longer that you sleep, the longer you live? He's talking about all-cause mortality. Everything I've looked at with all-cause mortality is there tends to be some U-shaped curve or maybe a hockey stick. There are very few things in biology that are like that, where it's less is better.
[0:40:14]
The one thing I though I found was RDW but even that, a few weeks ago, somebody found a paper showing that very low RDW is not always good. RDW being red blood cells distribution. It's a measure of variability, so lower is usually better, but not even that. Even that has been -- so, what do you think about this idea that more sleep is always better?
Greg: Well, I think it will be possible to of course say that the longer you're asleep, the shorter you're alive. If you look at the cross-sectional data --
Christopher: Because the amount of -- if you count time as being asleep is not being alive then that's right.
Greg: No, but what I mean is that really points out a lot of the meta-analysis have shown that there seems to be a J-shaped curve between self-reported sleep and all-cause mortality such that there's a sweet spot in the middle which is often around seven hours of self-reported sleep at which people have the lowest all-cause mortality. Then people who report less sleep in that will have a slightly increased risk of dying from any cause, and then people who report the longest sleep will have the greatest risk of dying from any cause.
Of course that doesn't look at things at the level of the individual. That's a population-level data. I think that if you look at the level of the individual then you probably would find, the shorter your sleep, the shorter your life, if you control for all those other dimensions of sleep, which is a big if, of course. What I'm saying is, if somebody gets more sleep and the quality of their sleep remains high, then I don't see that as being a bad thing by any means. I can't imagine many situations in which it would be a bad thing.
Christopher: Right, okay, that's helpful. Yes, going on from that, this is the sort of thing that really worries me. In one of the appendices in the Guzey article, he shows figure seven and figure eight where it would appear that Walker has pulled an answer key where it's just like -- so, there was a chart, the likelihood of injury based on hours of sleep per night, and he just chopped off the five-hour bin so that the data fit his worldview. Is that not just fraudulent?
Greg: It happens a lot.
Christopher: Does that make it right?
Greg: No.
Christopher: To have your hand in the till, does it -- everyone does it.
Greg: Yes, I know, and it's frustrating. People do similar things too. They'll change the scale of the Y axis, for example, to make it appear as if the effect size is larger than it actually is. Maybe the Y axis begins at 350 and ends at 370, and it looks dramatic. Then you scale it out from zero to 370 and the effect is tiny. So, it happens all the time. It's not excusable. In that instance, I think Guzey is absolutely spot on. Chris, I didn't quite finish my train of thought on the last question just because --
Christopher: I'm sorry.
Greg: No, no worries, I probably sounded like I did anyway. The reason I say this is it does pop up later in Guzey's article, and he actually references a discussion I had with Dave Maconi in that we spoke about the Guzey article, point by point, because Dave just wanted to chat about it. It was basically a monologue of me going through the points, individually, which --
Christopher: Oh, wow.
Greg: -- which if anyone made it through, I would be exceptionally impressed and somewhat perplexed. Guzey, off the back of listening to that, then added something to the appendices in the article. He basically quoted something I said which was too defective. The mortality sleep J-shaped curve from Section 1 doesn't disprove Walker's the shorter your sleep, the shorter you life.
First, dissociation through long sleep and short life is generally considered to reflect underlying comorbidities that prolong time in bed. Second, this doesn't look at sleep, loss of sleep extension at the level of the individual person. If you look at the individual person then shorter sleep is likely to be associated with shorter life, which is effectively what I just said to you, Chris.
Guzey rightly points out that some diseases also lead to shortened sleep, and states that if one wants to make the argument, the association between longer sleep and greater risk of dying from any cause, it's not indicative of the effect of sleep. One also has to accept that the same is true about shorter sleep and higher mortality. That's fine, and I accept that.
I think the reality is that most people artificially truncate their sleep using an alarm clock. For example, [0:44:41] [Indiscernible] MT-TQ database data suggested about 80% of adults use alarm clocks before workdays, and there isn't a corresponding intervention that lots of people use to artificially extend their sleep. So, I think his point there is somewhat moot.
[0:45:00]
He also states that the level of individuals that aren't experiments that would reasonably be able to establish causality that would support the claim that sleep restriction causes long-term health problems or an increase in mortality. That may be true, but I think it will always remain true because I just can't imagine an IRB accepting long-term experimental sleep restriction of something that's ethical.
Just to finish this particular point, he then states, "You could say that it is well-known that short-term or acute lack of sleep causes stress response from the organism." This is true. However, if you make this argument, you should also consider the following argument about the dangers of exercise is valid. When we exercise, our blood pressure and heart rate would increase, stress hormones concentration in the blood rises, and muscles get damaged. Therefore, exercise is bad for your health.
Exercise demonstrates the argument to the form, X makes me feel bad in the short term. Therefore, it's bad for health is suspect. I understand what he's trying to say here, but I think it's a really weak comparison. The reason is that, yes, exercise is a stressor, and intensive exercise may acutely dispose people, vulnerable people specifically, at certain events, for example, a heart attack in people who have certain congenital disorders, but there's abundant evidence that appropriate doses of exercise will lead to adaptations that makes us more resilient to the effects of subsequent stresses. The same is simply not true of sleep.
There's no evidence that I've seen that sleep restriction or sleep disruption of other forms will lead to hormetic effects on people that don't have sleep disorders that will protect them against the effects of subsequent sleep disruption. Just as an example of this, Hans Van Dongen did a very highly publicized study in 2003 in which he allowed people varying amounts of time in bed for up to two weeks, and basically found that sleep restriction led to cumulative and dose-dependent deficits in cognitive performance in all the tasks that he looked at. Interestingly, the people didn't realize how impaired they actually were.
So, while, objectively, their performance continued to decline during sleep restriction; subjectively, they felt like initially they performed worse but then they plateaued at that level of slightly impaired performance.
Christopher: Yes, I think that Josh's four quadrant model, again, is really useful here. So, exercise is a quadrant two intervention, that is something that is evolutionarily familiar and disruptive. Whereas, sleep deprivation is not. That's just creating an environmental mismatch. We're well-adapted to exercise, and it's not surprising we get fitter when we do it. Whereas, sleep deprivation is not something that we're well-adapted -- and that may not be true for all animals. There may be other animals where they are well-adapted to sleep deprivation. I believe that Walker writes about some of those migratory birds that are well-adapted to sleep deprivation in the book.
Greg: Yes, but even in those instances, I think they have very, very short micro sleeps during those transoceanic migrations and then they might rebound afterwards. Absolutely, certain animals will be better buffered against the effects of sleep loss.
Christopher: Can I just ask you one last question, and that would be the first point from Guzey's summary in the article. Do you think that Walker completely misrepresents the relationship between sleep and longevity and between sleep and cancer?
Greg: So, sleep and longevity, we've largely discussed. What I will say is, of course we don't have those experiments in humans that could identify causality. What I will say is that I think there are plenty of reasons to think that shorter sleep will lead to shorter life at the level of the individual.
Now, with respect to cancer, I agree with him that the epidemiology doesn't generally show a strong relationship between sleep duration and cancer. This seems to be dependent on certain things. For example, a couple of meta-analysis which have reported that short sleep duration seems to associate with cancer among Asian people but not necessarily among people of different ethnicities. Of course when I say Asian people, I'm speaking about an enormous wave of people who are very heterogenous. I think you understand my point.
There's also some data that long sleep duration associated with risk of colorectal cancer. Again, I think, at the level of the individual, there are plenty of reasons to think that sleep loss will dispose somebody to cancer. Actually, this is something that I speak about or will speak about at length in my talk that I'm giving in Brisbane later this year. If you just think about insufficient sleep then one of the effects would be that it will disrupt the circadian system. I won't touch on this now, but there are plenty of reasons to think that that could contribute to cancer.
[0:50:16]
Sleep loss, of course, leads to immune dysregulation and oxidative stress. It increases activity in the sympathetic branch of your nervous system which could be problematic. It increases the concentration of various inflammatory factors which could be problematic. Of course, the longer that somebody is awake, the more reactive oxygen species they'll produce. When the production of those reactive oxygen species overcomes the antioxidant capacity of cells to eliminate those products, the result is oxidative stress, and oxidative damage could participate in carcinogenesis.
Interestingly, if you look at some studies that have been done of acute sleep loss on marks of DNA damage, then it seems to be in elderly adults, but also there's been some data done on doctors, for example, that if you deprive people of sleep acutely, the marks of DNA damage will be increased the following morning. You can think about metabolic dysregulation. I know Thomas Seyfried, for example, has written and spoken a lot about the roles of metabolic dysregulation, cancer pathogenesis, and spoken about this previously on the first podcast I did with Tommy.
If you look at all the studies ever done on sleep restriction then they found things like increased subjective hunger. On average, people seem to consume about 250 more calories per day after sleep restriction. Sleep restriction acutely leads to a small amount of weight gain, on average. It leads to reduced incidence sensitivities, change in brain activity that dispose people to make worse food choices. There are plenty of reasons to think that high levels of processed food intake may contribute to their development of cancer and the progression of cancer too.
Also, when you restrict people's sleep, they tend to make worse decisions, and some of those decisions relate strongly to their health. For example, maybe people are disposed to abusing certain substances, alcohol and stimulants and so on, cigarettes. Through those indirect pathways, I think there are plenty of reasons to think that short sleep could contribute to cancer. So, yes, the epidemiology isn't very clear, and it seems to be peppered with lots of justifies. With that said, I think that were it possible to do a long-term study of sleep restriction, we would see that sleep restriction contributed to an increased risk of cancer.
Christopher: Okay, that's very helpful. Thank you. So, in summary then, yes, you should sleep but, no, you shouldn't probably read Matthew Walker's book, at least not the people listening to this podcast. For the people listening to this podcast, I think the how-to books are going to be way more helpful, so I would refer people to the interviews that I've recorded with you and Tommy recorded with you, and then also my interviews with Ashley Mason, the four-part series that we recorded. That's going to be far more helpful for people listening to this podcast than reading Matthew Walker's slightly suspicious, perhaps scientifically inaccurate book.
Greg: Yes, I don't want to mis-characterize how I feel about the book. I think the book is a very interesting read. I think it probably has done a lot of good. The thing that I perhaps find the most frustrating about Guzey's article is that it completely disregards the potential positive impact that the book has had. It focuses exclusively on the negative effects on things like performance anxiety in people who have sleep difficulties.
I think that the reality is that a lot of people who have read the book will have done things to attend to their sleep hygiene and made some changes in their lifestyles which now support their ability to get better sleep. My guess is the book has probably had a net positive impact, so I still would recommend that people read the book.
Yes, like any book, there are flaws in it, and there are other books out there. It's interesting when you see these phenomena in which something becomes massively popular, and everybody flocks to it. In the meantime, there are all these other resources that are out there which might actually be more useful for people, but because when you walk into your Barnes and Noble and sat there, facing you, you pick it up because it's so visible.
Christopher: Yes, that's the reason why Matthew Walker just hit -- I don't think anybody even really knows the formula that sells books.
Greg: No.
Christopher: There's a lot of serendipity involved. He's got two and half thousand reviews on Amazon. That's a big deal.
Greg: Yes, it's extraordinary. I'm sure there are lots of factors, and I don't know what those factors are. That'll be --
Christopher: I don't anybody does.
Greg: No, absolutely not, but I'm sure that some of them are things that we could do as a part -- the title itself is catchy, Why We Sleep. That's very SEO-friendly. It's very shareable too. I think that the message was delivered at the right time which really is serendipity. He was riding a wave of, a surge in interest in sleep. Also, I think he is very marketable. He's a handsome guy. He's a young guy. He speaks very clearly. That has probably done a lot to help him expand his network and get the book to more people too. It's interesting that you reacted the way to Matthew that you did, Chris, because I don't think that many people do react that way.
[0:55:38]
Christopher: Yes, I think I'm a strange bird in many ways. I really wanted to avoid the ad hominem attack actually. If anything, you've done the opposite there. You praised him on his physical appearance and that he speaks clearly, which is not necessarily relevant. I'm more interested in the scientific accuracy of the claims. I mean, you could say the contrary. He seems to have lied about where his PhD came from. Anybody can fact check that too, right?
Greg: He didn't lie about that. That's Guzey's misunderstanding.
Christopher: Oh, really? So the Medical Research Council in London does issue PhDs then.
Greg: No, what I mean by that is, for some reason, Walker -- I just checked this a few minutes ago -- Walker, on his website, states that he got his PhD from the Medical Research Council. Of course, the MRC is a funding body. They fund PhD studentships and research. What that means is that, if you're at an academic institution and you need some dough to support your ability to do the work which you want to do, then you can reach out to the likes of the MRC or the BBSRC and ask for money and apply for grants.
Matthew Walker's PhD was done at New Castle Uni. It's funded by MRC, who also actually part-funded my PhD, so, thank you the MRC. I think what Guzey is getting at is, well, what he suggests, and this is not why I'm saying -- because the thing with universities is that you can be at a university which, on paper, isn't necessarily a great university, people don't look at it and think of the Harvards and the Oxbridges of the world, but for specific subjects, it might be exceptional.
Sleep is a good example of this. When you think about the best universities in the world at which to study sleep, you think of Pennsylvania and Chicago. Here in the UK, the best places to study sleep are places like Surrey University, and they're not necessarily in the same echelon as Yale and Stanford. So, that really is a moot point.
Basically Guzey says that Walker says that he got his PhD from the Medical Research Council as if it's an academic institution when what Walker is saying is he got his PhD, and it was funded by MRC. He did it at New Castle Uni. Now, of course, he's a professor in Berkeley, and he did his post-graduate training at Harvard, which is one of those universities. So, I think it's just a lack of familiarity with the academic institutions and how research is funded and carried out on Guzey's part.
Christopher: Is that not highly unorthodox though, to say that your PhD is from MRC when all they did was write you a check?
Greg: Yes, that's the kind of thing that shouldn't be on your website, especially if you're a very popular person, if you're a very celebrated person.
Christopher: I think it's super weird to think that people are not going to check this shit because the Reddit community is really good at doing that.
Greg: Yes, I think the other thing is Walker is now an exceptionally busy guy. He's constantly being asked to fly around the world to give talks. I'm sure that he has done very well for himself off the back of the book. He will take those opportunities while you have them because you don't know if they'll be there forevermore. Of course it's hard balancing that with your academic career too.
Christopher: Of course.
Greg: My guess is that it has probably been quite hard being one of Matthew Walker's students in recent times because he has been so busy with some of these other projects. Maybe things like websites, they get put to one side while you're constantly fighting jet lag and giving your next talk and so on.
Christopher: Right, yes. It's very hard for me to reconcile that. This whole, is he just sloppy or is he fraudulent? The CDC versus WHO thing, okay, maybe that's just sloppy, but, really, when you made that claim, you didn't Google it real quick? Were you running in Hyde Park when you dictated the book? Did nobody bother to fact check afterwards? Then you see this other stuff, oh, I just accidentally lost the five-hour bin on this chart. That starts to look more fraudulent. Then I start wondering about all these other things where it's not clear whether it's fraudulent or sloppy.
Greg: Yes, well, those are all perfectly fair misgivings.
Christopher: Okay, Greg, I have one final question for you. Where can people find you online, and are you working with clients?
Greg: They can find me online, I now have a website, gregpotterphd.com. I wanted gregpotter.com, but gregpotter.com was taken. I don't want to pay for --
Christopher: Who is it taken by? Who is the other Greg Potter?
[1:00:00]
Greg: His name is probably not even Greg Potter. Anyway, it's taken, and I don't want to fork out loads of money to get that domain name. Now, I sound like a bit of a plonker calling it gregpotterphd.com, but --
Christopher: You sound a bit of a plonker when you book a massage and you put PhD after your name. What is your website? It's okay to say PhD in the domain name.
Greg: You sound like a plonker when you list your neurolinguistic programming letters after your name.
Christopher: It's some credential that -- yes, I totally agree. That's cool, so, gregpotterphd.com. What can people find there? Are you working with clients?
Greg: Yes. So, there you'll find also links to my social media which are also just @gregpotterphd. That's Instagram and Twitter. Linked In is gregpotterphd, is the extension too. At the moment, it's very rudimentary, the website. It really is just who I am and what I do. It's all very self-aggrandizing, but apparently that's necessary.
Regarding what I'm working on at the moment, I do some work, one on one, helping people with their sleep. I also do some work, I did mention this at the start, Chris, but I'm spinning a few different places at the moment. I do some work with athletes and, at the moment, specifically ultra endurance athletes.
We recently did some work, my friend, Ali and I, did some work helping two guys get ready to row the Atlantic. Without boasting, they did really well. I'm sure it wasn't a reflection on us, but they did break the World Record. So, we do a little bit of work helping people perform in ultra endurance events.
At the moment, we're getting a guy get ready to try to become the youngest man to walk the South Pole later this year, and we're doing some work with some guys who are doing the YUKON 1000 too. So, I do a little bit of that work, but the one-on-one stuff really is when I have some time in my schedule for it. If you're interested in that, then please do reach out to me, but it really depends on how much availability I have at the moment.
I also get some patients sent to me from the psychiatrist in London to help them with their sleep, so my schedule is normally pretty booked, but I'm always trying to help people even if it doesn't necessarily seem that way when they go to my website. So, if you do have issues then please do reach out to me. I'll do what I can, but I'm just crunched for time so I'm sorry if I'm a bit abrupt getting back to you quickly.
Christopher: No, that's fine. That's fine. I think a lot about what is the best way for people like you to help us, and perhaps what you're doing right now in this series of blog posts that you wrote for us, is the most efficient way for you to help because you're just going to say the same things. If I get you on the phone, it's not like you're going to come up with something magically different.
Greg: Yes, it's going to be personalized, but in that instance, people who have insomnia, they can use relatively generic programs and get good results. All of the online cognitive behavioral therapy for insomnia programs that have been studied to date have really shown that. Look at Sleepio, for example, or look at Shuteye; they're quite well-accepted, and they deliver quite good results too. So, really, when you go and see a practitioner for those types of issues, it's just the icing on the cake in terms of some of the specific tweaks that you make, but you should be able to get the bulk of the results from some of those more generic programs that are out there.
Christopher: I see.
Greg: The same way, if you take somebody untrained and you put them on some generic three-day per week, whole body resistance training exercise program, they'll probably get pretty good results. Then after a bit of time, maybe they need a bit of help to progress from the intermediate stage to the advanced stage.
Christopher: Right, yes, that makes perfect sense. Well, Greg, thank you so much. Was there anything else that you wanted to add?
Greg: I don't think so. It's a big article. We could have discussed lots of things that we didn't get to, but I think that --
Christopher: I know.
Greg: We covered the main points. I just want to reiterate. I don't know these guys, and I appreciate what both of them have done in different ways. I hope that I don't seem overly critical of them or anything like that. Again, I'm by no means perfect and don't feel like I've spoken very clearly in this podcast actually, but, yes, I just wanted to end on that note. So, reach out to those guys and say, keep up the good work even if neither of you is perfect.
Christopher: Right, particular thanks to Alexey Guzey who gave up his own time to create this fantastic piece of work. I'd highly recommend that anybody go read the original article.
Greg: Yes, and thanks to you, Chris. Thanks for the opportunity because this type of thing, it does reach a lot more people than the one-on-one work, so I'm always keen to try and do this type of thing whenever I get the opportunity.
Christopher: Robb Wolf says it's six people at once rather than one person at a time.
Greg: I'll take it.
Christopher: I've stolen his joke but, yes, it's a good one. Excellent, well, thank you so much, Greg. I really appreciate you, thank you.
Greg: Thanks, Chris.
[1:04:40] End of Audio
© 2013-2024 nourishbalancethrive