Interview: Penelope Lewis, Author Of 'The Secret World Of Sleep' Think of everything your brain processes in a single day: your breakfast, a stain on a book cover, a meeting at work. If you remembered all those things, your brain would reach capacity. Author and neuroscientist Penelope Lewis says sleep helps sort through the memories that are worth keeping.

Of Neurons And Memories: Inside The 'Secret World Of Sleep'

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript


This is FRESH AIR. I'm Terry Gross. If you're tired, consider that one of the reasons you'll want to hear this interview. It's all about sleep. We'll get some advice about what to do if you're having a hard time falling asleep. We're also going to hear about what scientists think is happening in our brains when we're sleeping and why it's so hard to be in a good mood or form lasting memories without adequate sleep.

My guest Penelope Lewis is the author of "The Secret World of Sleep: The Surprising Science of the Mind at Rest." She's a neuroscientist who directs the Sleep and Memory Lab at the University of Manchester in England. Penelope Lewis, welcome to FRESH AIR. You run a sleep lab. Just give us a sense of what your lab focuses on.

PENELOPE LEWIS: My lab is particularly interested in how memories change while we're asleep. People might think that memories, once they're formed, remain static, and it's just, you know, kept like a library book on a shelf, and then you can pull it out when you want to, but actually that's not the case. Memories evolve constantly. They change, and a lot of that evolution seems to happen while you're asleep.

GROSS: So give me some kind of clue about how when we feel like we're doing nothing, when we feel like we're completely turned off, except for dreaming, and we don't dream all the time, that something's happening in the brain that's actually helping us remember things.

LEWIS: Well, there are different ideas about what it might be that's happening during sleep that helps us remember, and also critically, it doesn't necessarily always help us remember exactly what happened. Sometimes it helps us to recombine information, kind of extract out the gist, the main points from, you know, lots of different experiences. So it's not necessarily always strengthening.

But when it comes to strengthening, we know that the neural responses in your brain that are associated with things that you've recently experienced are spontaneously replayed or we say reactivated while you're asleep. So that means, supposing you're learning to play the piano, and that means you're moving your fingers a lot, and that's associated with responses in motor areas of your brain associated with fingers, then those areas will become active again while you're asleep.

And that replay or reactivation is we think what's responsible for the strengthening. So it's kind of like your brain is rehearsing stuff without you knowing while you're asleep.

GROSS: Let me tell you something that happens to me and see if this relates to science as you know it. (Chuckling) I do a lot of my research for interviews the night before the interview, but I don't actually write my questions until the day of the interview because I find after doing all the research, like I can't, I can't synthesize it yet. I have to just, like, let it sit for a while before I can synthesize it into what does this mean, what questions do I really want to ask, what order do I want to ask them in.

And when I - if I had a good night's sleep, when I wake up it's just suddenly clearer to me. I can see what really matters, what do I want out of this interview. Does that make scientific sense?

LEWIS: That's actually a perfect example of what I meant when I said sometimes it's not necessarily just about strengthening, but it's about extracting out the gist or maybe the main points. So it's interesting that you say, you know, you've read the material the night before, but you don't really - you haven't maybe kind of synthesized it all together, and you can't quite work out what you want to ask, what are the important things, what's really interesting. That's exactly the kind of thing we think sleep helps with, as well.

GROSS: How can the brain replay memories while we sleep without us being aware of it? And you could argue, well, some of that is being played out in dreams, but a lot of it isn't.

LEWIS: Well, I suppose that goes down to a kind of philosophical question about awareness, really. The brain is doing all kinds of things that we're not aware of all the time. So why should we be aware of it?

GROSS: OK, that's compelling.


GROSS: I guess I can't argue that. And the only reason why scientists think that the brain is replaying memories during sleep is because of memory exercises you've conducted, some of which proceed even while the person's sleeping.

LEWIS: There's lots of really strong evidence that memories are replayed during sleep. So probably the most compelling evidence comes from rats, actually. This work is usually done by getting rats to run through a maze, and there are cells in the hippocampus of the rat brain - and actually we have them in humans as well, but we don't tend to measure them in this way, by sticking electrodes in - there are cells which respond to specific places, and they're called place cells.

So as a rat runs along say a linear track, you're going to have a series of cells firing in the same order every time it runs down that track, so it'll be Cell A then B then C then D because they respond to specific spots on the track. The kind of research that supports replay in sleep shows that if you record a rat's sleep before it practices on this track and then again after, then what you find is that during the sleep after practicing running down this track, you'll find those cells firing, A then B then C then D, much more often than it would have happened before.

So this is suggesting that the memory of running down the track is being replayed.

GROSS: You write that sleep provides a sort of spring cleaning for the brain. What kind of cleaning are you talking about?

LEWIS: Well, you clean up after you've made a mess, right, and so that's exactly what sleep seems to do for the brain. So looking at what happens with neurons in the brain, it looks like across the day while we're busy doing things, you know, experiencing things, seeing things, hearing things, learning things, processing different kinds of information, the neurons, the connections between neurons in our brain get strengthened a lot because they're trying to retain all this information.

And an awful lot of it is garbage, it's stuff that you don't want to remember and you don't care about, you know, what you had for breakfast or the color of a stain on the cover of a book or something. It's really not useful or interesting. And the problem is that if you keep storing all this stuff, at some point you reach capacity, and you can't keep storing more.

And so what happens during sleep and specifically during the deep stage of sleep that we call slow wave sleep is that all of those synapses get downscaled again. So where they've been strengthened up, it's all proportionately downscaled, and that's why we compare it to spring cleaning.

GROSS: Sometimes when you're faced with a difficult decision, somebody will say to you, like, well, sleep on it. That's actually good advice.

LEWIS: Yeah, I think that's excellent advice. In a way, this area of research is just one of - you know, it's one of these many areas of psychology where what we've wound up doing after a lot of very expensive research and careful experiments is just showing that some of those old anecdotal bits of advice that we got from our grandparents are actually very sensible.

GROSS: Yeah, but I always just assume it's because, like, you think more clearly when you're not tired, and you're more forgetful and think less clearly when you are tired. But you're finding out that there's actual - it's not just a question of being tired or not, there's actual activity in the brain that's doing all this work when we sleep.

LEWIS: Yes, so it's not just a question of being tired or not. I mean, the two processes that we've talked about so far, the two things that are thought to be important for strengthening memories or at least making them easier for you to retrieve in a way that's useful for you, and one of them is the replay, so that's a very active process, and the other one is this downscaling.

And I suppose you could argue about whether that's active or not because it is a reduction, actually, in activity.

GROSS: While you're sleeping, your brain kind of helps you retain certain memories. It also helps clean out ones you don't need. So does that connect with the dreaming process at all? Has the research you're doing helped you understand what dreams are?

LEWIS: I think that's a really fascinating question. The research I do specifically hasn't because we haven't looked at dreams at all in my research yet. But other people, for instance Bob Stickgold and Erin Wamsley at Harvard, are doing that kind of work, and it's really interesting what they're finding.

So their research suggests that dreams might actually represent part of the information that is replayed. So I said earlier that memories are replayed during sleep, and that might be why they get strengthened, or they get protected against decay. What Bob Stickgold and Erin Wamsley think is that dreams represent kind of the tip of the iceberg.

So while your brain might be replaying lots of different memories, not all of them will come to consciousness as dreams, but some of them will. And so the dreams can give us an idea of which information is being replayed. And the work that they've done that shows that has - what they've done is they've trained people on mazes, so kind of a computerized, computer game maze, and then let people sleep and then asked them afterwards, did you have any kind of mentation during that sleep about the maze, and then tested their performance on running through the maze from Point A to Point B; they'd have different points each time, to see whether it's improved across the sleep epoch.

And what they find is that people who say that they had some kind of thoughts or, you know, dreams about that maze while they were asleep are the ones who improved the most across that epoch, whereas if you just think about it while you're awake, it isn't associated with any improvement.

GROSS: In my dream world, the things that happen in my dreams often reflect on what happened to me through the day but not a kind of real version of it. It's like a totally surreal, choppy, episodic, barely sensical version of what happened during the day. Is there any scientific explanation for why dreams are usually so surreal?

LEWIS: Yeah, there actually is a hypothesis about why dreams are so surreal. We can't say it's an absolute definite correct explanation. But the hypothesis is that essentially during some stages of sleep, and particularly REM sleep, which is when you tend to have your most emotional dreams and your most vivid dreams, the communication between different parts of the brain is disrupted such that it's difficult for you to bind together the different elements of memories.

So imagine that you had a memory of what you had for breakfast this morning. It should include information about the food but also the place where you were, who was there, what it tasted like, what it smelled like, all those things. And the way this information is represented in the brain is typically distributed.

Some of it will be in the areas associated with smell, some in taste, some in vision areas, et cetera. And so in order to call that a complete memory that makes sense, you need to bind those things together, and for that you need a structure called the hippocampus.

But it turns out that in REM sleep, the communication between the neocortex, where these things are represented, and the hippocampus is disrupted. So it's difficult to bind them together. And we think that that's why the dreams you have in REM sleep are particularly fragmentary compared to dreams you might have in other stages of sleep where that line of communication is actually stronger.

GROSS: One thing that happens is what you describe as dream amnesia. Even when we remember our dreams, they tend to evaporate pretty quickly. Is there an explanation for that?

LEWIS: Yeah, so there has been a proposed explanation for that, and it relates to this idea that the hippocampus is not communicating well with the neocortex while you're asleep, because the hippocampus is not only important for binding together the different elements of a memory, it's also the area where we encode memories initially. So when we learn something new, it seems to involve the hippocampus straight away, and the hippocampus has to be in communication with the neocortex in order for that to happen.

And so if that line of communication is disrupted, as it is in REM sleep, for instance, it's difficult for you to learn something new, we think, and so this is the possible explanation of why when you dream something, you tend not to remember it because it's not being encoded, it's not being inscribed into your hippocampus in the way that it would if you were awake.

GROSS: If you're just joining us, my guest is Penelope Lewis. She's the author of the new book "The Secret World of Sleep," and she is the director of the Sleep and Memory Lab at the University of Manchester in England.


GROSS: So by talking about the importance of sleep and how sleep helps us remember things and synthesize what we've learned, I feel like you're putting the pressure on us to make sure we really get an adequate amount of sleep because you're also telling us we're not going to perform well if we don't sleep well.

And, you know, sleeping isn't always the easiest thing. It seems like it should be so restful and relaxing, but, you know, a lot of people have a hard time falling asleep or staying asleep. You know, why is it sometimes so difficult to sleep?

LEWIS: I don't really know the answer to that, but I could talk about some of the things that can prevent people from sleeping. There are different ideas about it, but one idea is just the obvious one, that your brain gets overactive, and then it's hard for that activity to calm down, as it needs to do when you fall asleep.

Another one is if the place where you're sleeping is not well set up for sleeping, and so if it's, you know, really bright light, or if it's too warm, if you don't have a comfortable bed, all these kind of things are obvious things that would keep you awake.

Another big one is shift work, so trying to sleep at the wrong time of day. And this is important because our bodies have a 24-hour cycle, and this controls all kinds of things like body temperature and release of various hormones and things like this. And part of this influences when our body is ready to sleep.

Even if we might feel tired at different times, the correct time in the circadian cycle is going to be the most healthy time for us to sleep, and we're going to get the longest and most refreshing sleep at that time. So if you're doing shift work, and you're having to sleep at odd times, then you're disrupting that circadian cycle.

There are also people who have actual problems with their circadian cycle. So it might be slightly off what it is for everyone else. So they might want to fall asleep much later at night or, you know, in the day, and so if they're trying to sleep at normal times that might be a problem for them, as well.

GROSS: Since you've made it clear that it's really important to get a good night's sleep if you want to remember and synthesize things, if you're having trouble sleeping, it's going to be even more tempting now to take a sleeping pill, you know, because sleep is so important. But do sleeping pills interfere at all with what is supposed to be happening neurologically while we sleep?

LEWIS: Well, there are lots of different kinds of sleeping pills. So it would depend very much on what you are using. But it's certainly the case that some of them really do. And what some of them do is suppress some of the stages of sleep. So sleep isn't actually just a homogenous state, it isn't like our brain switches off and stays off all night, and that's it. It actually doesn't switch off at all.

It goes through a series, a very highly structured series of different states. And it's important that you get all of these and that they come in the right order. And what some sleeping pills do is suppress, either partially or completely, some of those stages. And the one that most commonly gets suppressed is REM sleep. So that's the stage that you have a lot of towards the end of the night.

Probably most people are familiar with REM, meaning rapid eye movement, and the idea that actually you can see people's eyes moving around quickly under the lids while they're asleep. That's REM sleep. And yeah, some sleeping pills completely suppress that so you don't get any.

GROSS: And is that a very important stage of sleep?

LEWIS: Well, we think so. (Chuckling) That's - we think that all of the stages of sleep are probably important, but one of the, you know, most interesting research questions, at the moment, is around what are the different stages doing, what are they important for, how do they act together. And I think the most current thinking is that all stages are important and they sort of reinforce each other. So it's like different steps in a processing line, which, you know, you're doing different kinds of processes on this thing.

One has to happen first and then the next and the next, and they have to happen in the right order. But we're not quite sure exactly what's happening at each step.

GROSS: Do antidepressants have any effect on the sleep cycle?

LEWIS: Yeah, so antidepressants often do. Again there are different kinds of antidepressants, but a common kind is called SSRIs. So these are selective serotonin reuptake inhibitors. And they tend to influence the sleep cycle in the way that I've just described, by suppressing REM, although this does vary depending on exactly, you know, which drug you are taking.

GROSS: So do you know what the long-term impact would be on your sleep and on the work that's done during sleep?

LEWIS: If you suppress REM completely?

GROSS: Well, if you were on an antidepressant for a long time that interfered with sleep.

LEWIS: Yeah, if you were on an antidepressant that suppressed REM sleep, which many of them do, then - it's a really good question, actually. What would the long-term effect be? One effect seems to be that you don't feel depressed anymore. So that's got to be a good thing, and that's why people take them.

We don't know if that has anything to do with the suppression of REM. There is one hypothesis that I find quite interesting, which is that maybe actually in some people who are depressed, it's a good thing to suppress their REM because REM is associated with specifically strengthening emotional memories and in many cases negative memories.

And so there's an idea that it might be that in people who have too much REM, as many depressed people do, these negative memories are getting overstrengthened, and that can actually be quite damaging and quite pathological. And if people are just selectively strengthening negative things more than positive things, this might help to hold them in a depressive state and prolong their depressive epoch.

So we don't know this - so this is a hypothesis. We don't know this for sure. But one idea is that actually these antidepressant drugs which suppress REM may be partially functioning by disrupting that cycle and preventing the overconsolidation of negative memories.

But it's very important that people realize this is a hypothesis, and it's something that scientists are just starting to test now.


GROSS: Something that you could get over the counter that allegedly helps you sleep is melatonin, which is - melatonin is a hormone that your body naturally secretes, but it's also sold in tablet and capsule form over the counter. What do you think about melatonin?

LEWIS: Melatonin is a really interesting hormone. It's naturally secreted a few hours before you should naturally fall asleep. It's actually - when you are exposed to bright light, particularly bright blue light, like sunlight, it's broken down. And so, in natural circumstances, you know, the sun has gone down and it's getting dark, your melatonin starts being secreted and when that gets to a high enough level it makes you feel sleepy and you fall asleep.

So why do people buy it over the counter and take it? It's usually if they have problems with jet lag, so they've moved to a different time zone and they actually need to fall asleep earlier than they would naturally and they're trying to reset their circadian clock so that it's in tune with the new time zone where they are. And you can do that by taking melatonin a few hours before you go to sleep.

GROSS: When it comes to getting a good night's sleep, part of the advice you give is check the temperature of the room. And you think it's important that the room be cool enough. And, you know, speaking for myself, if the room is not cool enough I'm going to have trouble sleeping. But why is that? Why do I want it cooler at night than I do during the day?

LEWIS: Well, one reason is that your body temperature naturally cools down when you fall asleep. And so if you're trying to get yourself to fall asleep then a really handy trick for making that happen is artificially cooling your body temperature down. So if your room is too warm it's going to be harder for your body to start cooling down, you know, and so that's going to maybe keep you awake a little bit. If it's cool, then when you stop being active and you lie down, you'll cool down. If you're having trouble sleeping, then it's even a good idea to kind of heat yourself up before you go to bed. So you could have a hot bath, a hot shower or put your feet in a hot foot bath so that you heat up your temperature and then you go to bed in a nice cool room and your body temperature cools down. And that tricks your body into thinking oh, I'm falling asleep because this is what happens when I fall asleep.

GROSS: With alcohol you sometimes get really drowsy after you've been drinking - even, you know, for a lot of people a glass of wine. But at the same time, if you have a glass of wine not long before you go to sleep, you might be very drowsy and then find yourself waking up in the middle of the night feeling it's impossible to go back to sleep. What's going on?

LEWIS: Oh, so this is an interesting one. There are different ideas about what's going on with alcohol. So it definitely makes people feel drowsy and puts them to sleep if they've had, you know, a certain amount of alcohol. But exactly as you've described, people do tend to then wake up and have trouble getting back to sleep. One idea is that the reason you have trouble getting back to sleep is actually because your body is, a few hours after breaking down the alcohol, is going through some kind of withdrawal. And that's why people think this "hair of the dog" idea works, you know; the next day when you have a hangover, just having a little tiny bit makes you feel better because your body is going through a withdrawal from the alcohol that it's had.

GROSS: So if you can't sleep, and if it's three in the morning and you're just lying in bed thinking and worrying because you're not getting asleep, and now worrying more because you've told us how important sleep is - thus making it even harder to fall asleep - so are you better off just like lying there and faking it? Or are you better off just saying, I give up. I'm going to go watch TV or check my email or finish the novel I'm reading? What do you think?

LEWIS: I think it's pretty clear that if you've been lying there for a while and you've been unable to get back to sleep, then it's not a good idea to just keep lying there. It's a much better idea to get up and go and do something else and come back later. So, but if you do this it's really important to avoid anything that's going to really wake you up. So you wouldn't want to go and have a cup of coffee. You - obviously, you wouldn't want to go and eat something really sweet. And you wouldn't want to go and turn on any really, really bright lights. So, you want to...

GROSS: Does that include a television - staring at a television?

LEWIS: That does include a television.

GROSS: Or a screen?

LEWIS: Or a computer screen - unless you either have a filter over it that would filter out the blue wavelength of the light, which are the ones that influence your circadian system and would cause you to wake up. Or on some computers you can get apps, you can get programs that will do the same thing, so they'll control which wavelengths of light actually come through. So yes, the best thing would be to get up and maybe have a hot shower and then try getting back into bed, that's one thought or, you know, quietly read a book or yeah, look at something on a screen that's been filtered.

GROSS: Now you advise eating your last meal of the day, I think, four to five hours before you go to bed, but then having a light snack - how long before you go to bed?

LEWIS: About an hour before you go to bed.

GROSS: And why is that helpful to have a snack?

LEWIS: Well, it could be helpful for two reasons. Firstly, we all know how hard it is to fall asleep if you're hungry, so you want to avoid that. But secondly, there are certain foods which contain proteins which actually promote sleep. And so and what they are is proteins that get broken down and processed to form neurotransmitters that promote sleep. And so if you have a snack that contains those, it can actually help you to sleep.

GROSS: And what foods have that function?

LEWIS: Well, there are lots of different foods, but things like bananas, turkey. I think tuna fish was one. There are a whole range of very common foods that contain these proteins.

GROSS: A lot of people drink caffeinated beverages because it helps them, you know, snap awake in the morning and they think it helps them work better and be more focused and have more energy. What does caffeine do in our bodies that makes us feel like we, you know, that it's helpful?

LEWIS: What it does to make us feel more alert. So what caffeine actually does is it binds up a receptor in your brain for something called adenosine. And any biologist in the audience may have heard of a little molecule called adenosine triphosphate, which is basically energy. It's a form of energy. And the way that we access that energy is by breaking off a phosphate and turning it into adenosine diphosphate. And so during the day, as we're active and we're doing lots of things, this adenosine diphosphate builds up because we've used more and more of the energy and it actually binds onto receptors in the brain and it helps to make us feel tired. And what caffeine does is it binds up those receptors before the adenosine diphosphate can get there. And so it prevents - even though this stuff might be floating around in our system and we should be tired, the caffeine prevents us from feeling that. So it's kind of a temporary protection against feeling tired.

GROSS: How has the research that you do on sleep affected your ability to sleep, how much you sleep and how important sleep is to you?

LEWIS: I think that by working on sleep I've become very aware of how important it is and it's caused me to always make time for it - well, try to make time for it. And I think also learning some of the tricks about how to facilitate sleep has really helped a lot. Like many sleep scientists, I've had problems with insomnia. I think that that often is one of the things that gets people interested in this field and I don't have that problem anymore. So I think, yeah, genuinely understanding it has dealt with the problem.

GROSS: What works for you to get to sleep?

LEWIS: For me, it needs to be really dark. I'm terrible. I go around and cover up, you know, any lights on any appliances. I've got blackout curtains on all my windows. And I put a towel over my head and tuck it in.

GROSS: You put a towel over your head?


LEWIS: Tuck it in every night. Yeah.


GROSS: OK. With all the work that you're doing to blacken the room, you still need to do that.

LEWIS: Absolutely. I think that the towel actually, even if it is perfectly dark in the room already, the towel for me has become associated with falling asleep. So it's kind of like putting a hood on a bird or something. But for my body to be, you know, feel like now it's time to fall asleep and I'm ready for it, just feeling that kind of terrycloth against my face is the signal. So that's really effective for me.

GROSS: What are you most hoping to learn from the research that you're doing about sleep?

LEWIS: I think there are a number of things that we'd really like to know. One thing is, we'd like to understand what sleep is really for. So we spend a third of our lives asleep as humans. That's a huge proportion. That's more time than we spend doing anything else, so it's got to be important for something. And so, part of the purpose of the research I'm doing is to understand that, what it's important for. But another part of it is for understanding something about memory. So memory is fundamental to the way our brains work. It's fundamental to our personalities. And there are an awful lot of unknown questions about memories - not necessarily how they are formed, but how they change over time. How do we extract just information? How do we go from an experience to having that experience integrated with everything else that you know about the world? All of these are really important questions and I think that sleep holds the key to some of those questions.

GROSS: So do you think - like how does sleep help us synthesize our memory?

LEWIS: So this is a very topical area of research at the moment. And what we think, so for instance, I have put forward a theory or a hypothesis which suggests that actually when we replay memories while we're asleep, if you can imagine each memory that we're replaying as a circle and imagine that you replay a whole series of them, so some of these circles are going to overlap. And if those circles actually represent neural activity in the brain that's getting strengthened when it's replayed, then the areas of overlap are strengthened more than the other areas.

GROSS: I want to thank you so much for talking with us about sleep.

LEWIS: Oh, thank you very much. It's my pleasure.

GROSS: Penelope Lewis is the author of the new book "The Secret World of Sleep." You can read an excerpt on our website,

Copyright © 2013 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.