睡眠科学家：糟糕的睡眠究竟会对你的身体和大脑造成什么影响？Matt Walker | #167

I have a problem, Matt. My problem is that I don't know how to have a short conversation with you. Usually, it's just, “Thanks, Matt. Take care. Lovely.” No, you're so brilliant and have so much wisdom to share.

Let's not rush to judgment on that quite yet, but I will try my best.

When we're at Platinum together, our conversations go for an hour and a half or two hours, which we unfortunately can't do here. So, I'm going to jump into some zingers and some areas of clear focus. I want to start with a question a lot of folks have: sleep aids—Ambien, Lunesta—good or bad? Tell me your thoughts on those.

I've been on the public record: I've not been the biggest advocate of that class of drugs.

Can you be clear about this?

Essentially, we're now in Web 3.0 in terms of sleep medications. Web 1.0: those were the benzodiazepines—temazepam, Valium—and then came the second generation, which were Lunesta and Ambien. Both of those work by essentially tickling an inhibitory receptor in the cortex called the GABA system. It's “lights out, red light, stop traffic.” What those drugs do—they're called sedative-hypnotics—is sedation, and sedation is not sleep.

When you take Ambien, I'm not going to argue that you're awake. You're clearly not awake. But to argue that you're in naturalistic sleep, if I were to show you the electrical signature, it's not the same. We conflate sedation with sleep on those drugs.

Now, there is a new class of medications, and it's the first class of medications that, in truth, I've actually been publicly advocating for. They're called DORA drugs—D-O-R-A, small s. We'll speak a little bit about that when we actually look at some of your wonderful data.

These drugs work in a very different way. Instead of going up to your cortex and, baseball-bat-like, just knocking it out, these DORAs—and there are 3 FDA-approved medications, things like Belsomra—act like a set of elegant chemical fingers. They go down toward the brainstem and just dial down the volume button on wakefulness. Then they step back and allow the antithesis of wakefulness to come in its place, which is this thing called naturalistic sleep.

I've tried it. They're not cheap.

They're not cheap. Currently, they are not particularly well covered by insurance. It could be upwards of $350 if you don't get insurance coverage for that medication per month. That said, for some of the chronic insomnia patients that we work with, at the end of a month of horrific sleep, if I were to go see them on night 30 and say, “I could wave a magic wand right now. If you gave me $350, I could go back to the start of the month and you would have far better sleep. Do you want to take the deal?” The pain of sleep is such that it actually ends up with a yes answer.

You could say, “Well, yes, those medications increase the total amount of sleep. They're also quite good at increasing the amount of REM sleep without disrupting deep sleep.” You could argue, “Well, yes, and so what? Just because you increase sleep, it doesn't mean that that sleep is beneficial or functional. Isn't it potentially like junk DNA? It's just junk sleep.” Show me that it makes you a better individual. That's my question as a scientist.

There were some great studies. When we sleep, we wash the brain clear of all of these metabolic toxins that we build up during the day. Two of them are tau protein and beta-amyloid, which are associated with Alzheimer's disease. They did a great study with late-life adults, 65 and older. They gave them one of these DORA drugs, Belsomra, or suvorexant, and measured the amount of Alzheimer's-related protein in their blood the night before and the next morning, with and without the drug, in a placebo-controlled crossover study.

The next morning, not only did they sleep longer with this new class of medications, but the amount of amyloid and tau protein cleared away from the brain by way of sleep was greater after they had taken these medications. Whereas with Ambien, a study just came out 3 weeks ago, and it reduced it by 40%.

Wow.

In other words, it's not just that you add sleep; it's functional sleep. It's not epiphenomenal sleep.

So, a piece of wisdom there. Next question: You wake up at 3:00 a.m. Your bladder's full, or there's a sound, and you wake up. Sometimes I'll get up at 3:00 or 4:00. What do you do?

When you try to get back to sleep at that point, the reason sleep is a problem or a struggle for many of us, myself included, is that the Rolodex of anxiety begins. Once that starts turning, you start to ruminate. When you ruminate, you catastrophize. When you catastrophize, you're dead in the water for the next 2 hours.

I think, in truth, everything is twice as bad in the dark of night as it is in the light of day. So, your job at that moment, if you choose not to go down a medication pathway—which I would fully understand—is to try to distract your mind. How do you do that? You can try any one of the following 5 techniques.

Firstly, meditation. It's well proven. It is a clinically evidenced improvement in insomnia patients for those middle-of-the-night awakenings. The next is some kind of breathwork. You can do box breathing; pick your numbers. You can do that.

The third is a body scan. Start at the top of the head, close your eyes, and work through your body. Ask yourself, “Where is the tension? Where am I feeling it? Is my forehead wrinkled? Should I relax it?”

The fourth is to take yourself on a mental walk. This is a really interesting one. We did some work at UC Berkeley. Counting sheep, by the way, does the opposite. It actually makes it harder to fall back asleep because you're thinking, “This should be helping me fall asleep.”

Take yourself on a mental walk in hyperdetail. For me, I'm going to walk the dog. I open the drawer. Is it the blue leash or the red leash? I'll take the red leash. Clip them in with the right hand and open the door with my left. It's that level of detail.

What's common across all of these techniques is that they get your mind off itself. Sleep at 3:00 a.m. is a little bit like trying to remember someone's name: The harder you try, the further you push it away. As soon as you stop trying, usually the next thing you remember is your alarm going off in the morning. Why? Because you got your mind off itself.

Amazing. I love the way this man communicates. He's extraordinary, with such clarity and veracity. We're going to look at my data in a minute, but let's talk about 5 actionable steps that folks should jot down and should know inherently. I wrote a chapter in my last book on sleep, interviewing you, of course, as the great resource there. What are the 5 most important actionable steps to get the best night's sleep you can?

I'll do the standard 5 and then a couple of unconventional ones. The first is regularity. If you're going to do just one thing, go to bed and wake up at the same time, no matter whether it's the weekday or the weekend. Regularity is king because it anchors your circadian rhythm and, as a consequence, you improve the quantity and the quality of your sleep.

One of my most important things. What time do I go to sleep, guys? Anybody know? 9:30. I broadcast that. So, if somebody catches me after 9:30, it's like, “I better have been asleep.”

You'll see some of Peter's data in a second. When we looked at his data, I was stunned by how metronome-like and methodical your regularity—your deviation around, particularly, bedtime—was. We wanted to double-check the data because of this.

Yeah, my family goes much later.

Incredible. So, regularity. Second, darkness. We are a dark-deprived society in this modern era, and we need darkness at night. I think not just darkness in terms of blackout curtains or an eye mask—whatever it is, take all the electronics out.

What I'm talking about is this: In the last hour before bed, set a bedtime alarm. When that alarm goes off 1 hour before bed, I want you to turn down half of the lights in your house. Do this for the next 7 days and conduct the experiment. If it doesn't work, if you don't feel the difference, that would surprise me. It's surprising how soporific that darkness is. It encourages a sense that it's time to disengage, and I start my wind-down routine.

I think in the modern era we get what I would call junk light at night, and then we don't get enough bright light during the day. We're in this constant, low-level state of chronic anxiety light, where it's not light enough during the day and it's not dark enough during the night. It should be a beautiful sinusoidal pattern.

So, the next one is darkness. But let me just hit on that: what about blue-light-blocking glasses, like TrueDark, that Dave has, and also sleep masks? I use a Manta sleep mask, which I love.

Yeah, I love the Manta sleep mask too. No relationship with them; I think it's incredibly comfortable. I now have 3 of them. I've got one in both suitcases and one at home, and it's nonnegotiable. The blue-light glasses, I think, are interesting. The blue-light hypothesis in terms of impacting sleep—the data has been a little bit more mixed recently. It's not to say that those glasses don't work. What I think those glasses are doing, however, is essentially taking out some of the high-fidelity, true technical richness of the devices.

It's like turning down the lights, if you would. Correct.

Yeah. They do impact your sleep; it's just not necessarily through the blue light. These are attention-capture devices. They are designed to fleece you of your attention economy, and they do it ruthlessly well by activating your brain. When you use the devices, it hits the mute button on your sleepiness, so you don't think you're sleepy.

Here's the rule of thumb: if you want an unconventional tip, no matter how much I tell you about sleep, you're still going to take your phone into the bedroom. Not that I would wish that, but that's okay. If you do that, you can only use your phone standing up.

Great. Great.

After about 7 or 8 minutes in your bedroom using it, you think, “I'm just going to have a sit-down.” At that point, I'm sorry, the phone is put away. So, light—blue light—I think that helps you disengage. There's a feature on the iPhone where you can desaturate the screen; you can go black and white, or you can go largely red. I do the largely red.

Final 2 things: be mindful of alcohol and caffeine. I've changed my tune on caffeine. I say drink coffee. The health benefits are astronomical; dose and timing make the poison. After 3 cups of coffee, the health benefits go in the opposite direction. Try to cut yourself off at least 12 hours before bed—10 hours if you must.

How many folks here know whether they're a slow or a fast caffeine metabolizer? Do you know?

Right. So, you can find out; it's just your genetic screen, and that will affect you. I'm a slow metabolizer, so after 12, I'm done.

Yeah, and it's the gene to look out for. It's called the CYP1A2 gene. And if that's off the tongue, onto the floor, I've always thought it would be a great name for a punk rock band.

CYP1A2. Come on.

No. Last one.

You said food and alcohol.

Yeah. Alcohol is just not your friend. I would love to say you can get away with it. There is a dose response, and we have done some studies where, in the late morning or early afternoon, the blast radius of alcohol is far less than in the late afternoon or early evening. So, my advice—which I would never offer publicly—would be to go to the pub in the morning. That way, the alcohol is out of your system; no harm, no foul.

I should also note, by the way, that I'm being puritanical in all of these things. I'm talking about the ideal world; none of us live in that world. We live in the real world, and life is to be lived. I don't want to be the healthiest person in the graveyard, necessarily. So, deviate to some degree, but nevertheless, at least try to accommodate your knowledge so that you are making informed choices about your health.

Yeah. Being married to a sommelier makes it difficult for me not to taste the wine on occasion, but I have cut out 95% of all alcohol.

Impressive. Very impressive.

Anything last on the list? We covered those 5.

I think the final thing is you've got to have a wind-down routine. No matter what it is—whether it's stretching, listening to a podcast, listening to a sleep story, or meditation—any young parent knows that once you've got the sleep routine in place, if you deviate from it, it's a disaster. We forget that as adults: we are bound to the same edict of a wind-down routine.

Sleep is not like a light switch, and it shouldn't be like that. It's much more like landing a plane. You should take time to come down onto terra firma. Whatever it is for you, find out what it is and stick to it religiously.

Yeah. I use Audible. I listen to a book on tape, and I set a 15- or 10-minute timer to go off. Typically, I'm asleep before it ends. It's fantastic.

I mean, that's why it's like a bedtime story.

Bedtime story.

Calm, as a meditation company, was struggling a little bit. What they stumbled upon was the fact that the usage statistics on the 24-hour clock were spiking in the evening. People were self-medicating their state of insomnia. Then they realized children like to be read stories. Well, it turns out so do we adults. And next thing you know, Matthew McConaughey, Harry Styles, and a bad sleep scientist with an even worse British accent will read you a sleep story.

All right, I'm going to hand this over to you. Let's talk about Nightfall IQ Analytics. I gave you access to all of my Oura data, and you have a company that you've created which generated a beautiful report. I'm just open-sourcing my data here for you. Talk about the company, talk about my data, and talk about how the community can use it.

Yeah. What I've realized in working with concierge clients in my own sleep practice is that these wearables do a great job at sensing your data and then giving you the data in a UX. What my clients were saying is, “I don't really understand what it means, and I don't understand what to do about it.” So, we ended up creating a new company out of necessity called Nightfall IQ, and it's a deep-sleep analytics company. We're device-agnostic.

In fact, I'll show you. We ran Peter's data. Now, Peter, I know for a fact, because I've seen lots of his Oura ring data on the screen, he is a stellar sleeper.

This period of time actually spanned the L.A. fires, so obviously this is not representative. Just keep in mind, we were out of the home, in hotels, and out of my normal routine. I mean, how is the data this good? It's surprising.

Firstly, we start off with just a dashboard—a basic set of sleep metrics that will tell you how you're doing. Here, Peter, obviously because of that situation, was a little down in terms of total sleep. Then we actually give you your different sleep scores in terms of the component constituents of sleep. He's doing an incredible job in terms of his deep sleep. Even now, he's always been an exception to the rule, and I'll tell you why in just a second.

His REM sleep numbers were down a little bit, and we’ll come back to that. This is great, but it doesn’t really tell you, “Well, I got a 60 out of 100 for my REM sleep. Is that good? Is that bad for my age or sex?” Because we’ve got millions of hours of sleep data at my center, we now have a reference range. So next, we see how you stack up to your age- and sex-matched norms.

Here’s how Peter’s doing. He’s right around the average for total amount of sleep. We tried to do a chronological assessment. He’s probably about 16 to 17 years younger than his chronological age in terms of his deep sleep. I am desperately jealous and deeply envious. If he wasn’t so nice, I would dislike him immensely.

That said, when it came to REM sleep, we did see this deficiency. REM sleep, when you are under conditions of stress, is the first thing to go, and we could see it in his data here relative to age- and sex-matched norms. His sleep latency was perfect. We don’t like to see you fall asleep too quickly or take too long. In terms of the amount of wake after sleep onset, what we call WASO, this was higher. No wonder it was during that time period.

Finally, his sleep efficiency is fantastic. As long as you’re above 85%, we like to see that. The next thing that we have is that, because Peter has now got so much data, we can actually look iteratively, time after time, with a unique date-time stamp, and ask, “Based on the data, what are the trends that we’re seeing here?”

Firstly, because of the situation he was facing, we saw a huge amount of fluctuation. We wanted to bend those sinusoidal waves of high fluctuations down flat. In some areas, he was doing great. How his heart rate stayed this stable across that time period, I do not know. You need to interview him and ask what he’s doing to keep his stress under control in those conditions. That’s impressive.

What’s more interesting is that, in the analytics report that you get, we now ask something very interesting: What are the trends that we’re seeing over the past year? We saw something interesting in Peter’s data. Over about the past 6 or 8 months, there was an increase in his heart-rate variability. This is a very beneficial sign. It’s up a little bit—over 27%.

We’ll speak offline at some point in terms of finding out what the inflection moment was.

This is the power of Mona and Helen telling me to do high-intensity interval training and take my Zoom calls on my bike.

Isn’t that impressive? You don’t get to see that with your wearable, classically, but here is the data that now re-motivates a commitment to the change that you made.

The next thing that we ask is fascinating. Because we’ve got high-fidelity data, we can now cluster it with our AI analyses and say, “What is predicting your best nights of sleep, and what’s predicting your worst nights of sleep?” What’s stunning is that it’s different for different people.

What we found is that when Peter had a lower heart rate and higher heart-rate variability in the last few hours before bed, when he was well de-stressed, as it were, he had significantly more REM and significantly more deep sleep. We also found that when he had early and consistent bedtimes, he fell asleep faster and slept for significantly longer amounts of time.

By the way, it’s interesting: Coming into this summit, there were a number of times when I was literally going to sleep at 8:30 so I could wake up before the family and get work done at around 4:30 or 5:00. Those were my best sleep scores.

Yeah. Earlier bedtimes—and what it’s telling us is that you are sleeping more in harmony with what we call your chronotype, which is morning type, evening type, or somewhere in between.

We also looked at worse nights of sleep, when all of a sudden things get irregular. He had worse quality of sleep, and we saw circadian-rhythm abnormalities in things like his temperature. You don’t want circadian-rhythm abnormalities. Your body needs that 24-hour clock time for almost every enzymatic reaction.

Later bedtimes just destroyed his deep sleep. They cut into that early-night phase where we get most of our deep sleep. So what this teaches us is what we need to create as a prescription for Peter: where to lean in and where to stay away from.

We also finally looked—by the way, we reversed that analysis—not just at what predicts your best and worst nights of sleep, but at what the consequences are for you the next day after your best and worst nights of sleep. After his best nights of sleep, he was involved in about 8 to 9 more minutes of high-intensity physical activity, and he had markedly lower biomarkers of stress. He had about 30 minutes less physiologically distressed biology during the following day.

After his worst nights of sleep, he was largely immobile. He was much like an amoeba in terms of his activity—I’m just kidding you. He also had a lower overall readiness. In other words, he was more fatigued.

We offered lots of recommendations, and I’ll speed through these. We were really interested in focusing on his REM sleep and trying to improve it, and I have increased it since this report. I’m now getting—and I will show you the data—

I told you that those DORA drugs, the sort of Web 3.0, are not just good for improving your total amount of sleep. They’re also quite clever at increasing your REM sleep without decreasing your deep sleep. This was Peter’s data when we first ran his report.

We then worked with Peter and recommended one of these DORA drugs. He started the DORA drug, and now we’ve increased his total sleep time from about 7 hours to almost 9 hours. But the amazing part was what he was able to accomplish with his REM sleep. You can see here that his REM sleep score shot up to 25% from 13.9%.

By the way, let me hit on something that’s important that you taught me. If you look at your sleep over the course of the night, typically your deep sleep comes first, and your REM sleep comes in the second half, or the last third. So if you cut off your sleep—if you get woken up at 4:00 and stay awake—you’ve killed your REM.

Yeah, and that’s exactly what the first recommendation was for Peter’s REM sleep. We were saying, “Try to sleep 20 minutes later into the morning.” That’s the REM-rich phase. Remember when we said he was going to bed late at night? It decimated his deep sleep because deep sleep comes in the first half of the night.

So if you short-change the first half of the night and go to bed late, you’re taking away from the finger buffet of feasting on deep sleep. Whereas if you sleep longer the next morning, you’re going further into the REM-sleep-rich phase, so you get more time at the REM-sleep serving dish.

I loved this and found it extraordinarily useful. I need to talk to my team at Fountain. We should make this part of our baseline. What’s the cost per person for this service?

Right now, our cost dynamics are that we have 2 services. One is an elite package, where you get the full 25-page report, and that’s coming in at around $4,000. Nontrivial, I know.

The next level is the executive package. That’s really where I work with white-glove concierge clients. At that point, you come into our program for a whole year. We actually repeat it quarterly, because the important part of this service is not really the cross-sectional baseline. We then repeat it to give you time-lapse photography of what’s going on with your sleep for 2 reasons.

First, based on what we said you need to lean into, have you done it, and has it made a difference? Next, we want to nip in the bud any declines. If we’ve seen, over the last 2 quarters, a progressive decline in your deep sleep and you’re in your late 60s, we’re worried about Alzheimer’s disease risk and brain cleansing. We need to change and bend the arrow of that deep sleep back up. So that actually is closer to a $15,000 proposition.

You spend a third of your life sleeping. You should care about that. As you say, it’s the foundation upon which all the other longevity therapeutics are based.