Poe Links REM Norepinephrine Shutdown to Brain Memory Wipe

Poe将REM睡眠期去甲肾上腺素停工与大脑记忆清除联系起来 在REM睡眠期间，蓝斑核放电的完全沉寂是清除短期记忆存储并将情绪创伤与生理痛苦解耦所必需的。 Gina Poe博士详细解析了不同睡眠阶段的具体生物学功能，解释了慢波睡眠如何在前半夜促进大脑废物清除、蛋白质合成以及生长激素释放。她深入阐述了睡眠纺锤波与REM睡眠如何协同工作，将短期记忆从海马体转移至新皮层，构建创造性认知图式，并彻底清空短期存储通路。此外，对话还强调了REM睡眠期间蓝斑核的完全静默如何作为一种夜间治疗机制，将创伤记忆与身体的本能恐惧解耦。 睡眠以约90分钟为周期循环运行；在Stage 2睡眠中，睡眠纺锤波（sleep spindles）与K-复合波推动海马体与大脑皮层对话，将短期记忆转移并固化至分布式长期存储网络中。 (2:23) 入睡后的首个慢波睡眠周期会触发大量生长激素集中脉冲释放并加速突触蛋白合成，同时通过神经元同步胀缩如“舱底泵”般冲洗清除代谢废物；熬夜推迟入睡将直接错过这一修复窗口。 (4:34) 白天学习新知识会提高当晚的睡眠纺锤波密度，促使大量钙离子涌入皮层神经元远端树突，将新记忆缝合进既有认知图式（schemas）中，而REM睡眠期间的P波则激活大脑多区域以激发创造力。 (22:24) 快速眼动（REM）睡眠是蓝斑核完全停止放电的唯一阶段，彻底切断去甲肾上腺素使大脑能在不引发战或逃心身应激反应的前提下处理情绪记忆，充当内源性创伤修复机制。 (25:47)

Neurobiologist Dr. Gina Poe and Andrew Huberman outline how distinct sleep stages drive physical neural repair, schema consolidation, and overnight emotional regulation. They emphasize that circadian alignment and lowering pre-sleep sympathetic arousal are critical for hormone release, waste clearance, and trauma processing. Slow-Wave Sleep and Metabolic Brain Clearance (16:02) Sleep begins with metabolic restoration, where the first 20 minutes rapidly replenish adenosine back into cellular ATP, explaining the efficacy of short naps. During early deep slow-wave sleep (Stage 3), neurons execute synchronized cycles of firing and silence. Poe explains that sodium influx during firing pulls water into neurons, causing them to swell, while subsequent silence causes them to contract. This rhythmic expansion functions as a biological bilge pump, flushing fluid through brain tissue while glial cells vacuum up cellular debris and misfolded proteins. Spindles, Dendritic Plasticity, and Memory Consolidation (22:24) Memory processing relies on Stage 2 transitional sleep, defined by 10–15 Hz sleep spindles and K-complexes that coordinate communication between the thalamus and the neocortex. Poe notes that baseline spindle density correlates with general intelligence and increases directly following daytime learning. Citing research by Julie Seibt and Anita Lüthi, Poe describes how spindles trigger substantial calcium influxes into the distal dendrites of cortical neurons, physically integrating short-term hippocampal memories into long-term cortical networks. As demonstrated in rodent studies by Sidarta Ribeiro, memory traces step across cortical regions with successive sleep cycles. Alcohol disrupts this process by suppressing REM sleep and degrading the Stage 2 spindle windows required for consolidation. Locus Coeruleus Silence and Emotional Trauma Processing (18:23) The locus coeruleus, the brain's primary source of norepinephrine, modulates alertness by firing at approximately 2 Hz during wakefulness and 1 Hz during slow-wave sleep. Healthy REM sleep is the sole state in which the locus coeruleus shuts off completely (0 Hz). Poe explains that this total silence allows the brain to downscale outdated synapses and clear its short-term novelty storage for subsequent learning. Furthermore, this noradrenergic silence enables REM sleep to function as endogenous trauma therapy: the amygdala processes emotional memories while the autonomic fight-or-flight response remains inhibited, decoupling memory content from visceral panic. In post-traumatic stress disorder, persistent locus coeruleus activity during REM prevents this uncoupling, continually reinforcing traumatic fear responses. Sleep Inertia, Motor Paralysis, and Nighttime Awakenings (10:06) Normal sleep cycles last approximately 90 minutes. Healthy REM sleep enforces absolute motor paralysis (atonia) to prevent the enactment of narrative dreams, whereas sleepwalking originates as a hybrid state out of deep slow-wave sleep. Poe compares waking mid-cycle to opening an unfinished washing machine, causing profound grogginess known as sleep inertia; waking at the conclusion of a cycle leaves the individual clear-headed. Poe also notes that waking up once per night to urinate is a normal physiological occurrence influenced by fluid intake rate, while commercial sleep trackers remain only roughly 70% accurate when identifying specific sleep stages. Poe and Huberman conclude that sleep quality depends on strict temporal alignment rather than total duration alone, as delaying sleep onset skips non-negotiable early-night slow-wave clearance windows and growth hormone surges. They emphasize entering sleep in a calm state to suppress locus coeruleus activity, though consumer tracking devices still face technical limitations in accurately measuring these stage transitions.

神经生物学学者 Gina Poe 博士指出，睡眠是一个由严密节律调控的主动重塑过程，不同周期的脑电特征与神经化学环境协同完成了脑脊液排毒、记忆转录固化以及情绪创伤剥离，且这些关键功能高度依赖于稳定的入睡时机与自主神经状态。 慢波睡眠的代谢排毒与激素节律 (0:00) 整夜睡眠由数个约 90 分钟的周期交替推进，前半夜主要由阶段 3（N3）的深层慢波睡眠主导。在此阶段，全脑大量神经元经历高度同步的放电与沉寂：放电时钠离子内流促使细胞轻度吸水膨胀，沉寂时则回缩，这种群体性律动如同脑部的“舱底泵”（bilge pump），驱动体液冲刷神经组织，协同神经胶质细胞搬运并清除日间积累的代谢废弃物与错误折叠蛋白。入睡最初 20 分钟内，大脑还会迅速将累积的腺苷转化重构为三磷酸腺苷（ATP），这构成了短时小憩迅速恢复体能的分子基础。 在内分泌方面，首个慢波睡眠周期会激发机体集中释放单次大剂量的生长激素脉冲，并同步触发脑内蛋白质合成，以物理铺设突触基础设施。Poe 强调，由于细胞分子钟受褪黑素节律严密锚定，推迟 90 分钟以上入睡并非简单顺延排毒与激素释放，而是会直接错过该生物学窗口，因为后半夜睡眠将直接被 N2 浅睡与 REM 睡眠占据。内分泌学证据表明，集中脉冲式释放与持续低剂量释放的生理效果具有质的差异；保持固定就寝时间也因此成为衡量中老年神经系统健康的重要指标。 纺锤波与神经网络的记忆图式构建 (2:23) 记忆巩固的核心发生在非快速眼动第 2 阶段（N2），其特征为 10 至 15 Hz 的睡眠纺锤波与 K-复合波。Poe 将海马体比作临时内存（RAM），而新皮层则是长期存储的硬盘；纺锤波代表丘脑与新皮层之间的直接对话，介导了短期记忆向皮层网络的转录。Sidarta Ribeiro 的大鼠追踪研究证实，记忆痕迹在随后的周期中逐步从海马体迁移至初级投射皮层、次级区域乃至全皮层。Julie Seibt 与 Anita Lüthi 的实验进一步揭示，纺锤波会引发皮层神经元远端树突发生大规模钙离子内流，驱动显著的结构可塑性，将新近习得的体验编织进大脑既有的认知图式（schemas）之中。 个体的基础纺锤波密度与一般智力呈现显著相关，且日间进行新技能学习后当晚的纺锤波密度会适应性上升。进入睡眠中段后，REM 比例增加，源自脑桥的桥脑-膝状体-枕叶波（P 波/PGO 波）向皮层释放大量谷氨酸，在缺乏外部感官输入的情况下随机激发远端脑区，构成夜间概念重组与顿悟的生理底层。相反，睡前饮酒会直接抑制 REM 睡眠并严重干扰 N2 纺锤波，在酒精被肝脏彻底代谢前，记忆转录与图式整合会被系统性阻断。 蓝斑核静默与快速眼动期创伤解耦 (18:23) 大脑中负责合成去甲肾上腺素的核心核团——蓝斑核（locus coeruleus），在清醒维持警觉时通常以约 2 Hz 频率发放脉冲，突发刺激下的放电激增可触发快速注意力转移或恐慌。Poe 强调，快速眼动（REM）睡眠是蓝斑核完全静默（0 Hz）的唯一生理时期。这一递质绝对归零的状态承担着突触修剪职责，使大脑能够抹去并重置短期“新颖性编码通道”（宛如格式化临时 U 盘），为次日获取新信息腾出神经容量。 与此同时，REM 睡眠构成了天然的夜间创伤修复机制。在健康的 REM 睡眠中，尽管杏仁核等情绪中枢高度活跃且伴随生动的剧情梦境，但去甲肾上腺素的彻底关闭与骨骼肌张力丧失（atonia）使大脑得以在完全缺乏生理应激唤醒的环境下复盘记忆。该过程将经历的情理事实与出汗、心悸等外周战逃反应解耦，使患者能理性回忆创伤而不再诱发恐慌。而在创伤后应激障碍（PTSD）中，蓝斑核在 REM 期异常持续发放，导致恐惧反应每夜被反复强化。因此，Poe 建议睡前通过温水浴或深呼吸抑制交感神经张力，警惕在情绪未平复时盲目回溯创伤。 周期完整性与睡眠行为的日常管理 (12:28) 睡眠周期的物理完整性直接影响醒后机能。中途强制打断一个约 90 分钟的周期会引发严重的“睡眠惰性”（sleep inertia）。Poe 将其比喻为洗衣机运转中途强行开门：衣物浸透水和洗涤剂而无法立即穿用；而在 REM 睡眠末期自然苏醒则能保证头脑清爽。梦游等异常行为源自高唤醒阈值的深层慢波睡眠，是一种介于睡眠与局部清醒之间的混合状态；而健康的 REM 睡眠则通过主动的主动运动麻痹保护机体免受剧烈梦境驱使。 在日常调控层面，夜间自然醒来排尿一次属于正常生理规律而非病理征兆。控制补水行为（傍晚小口慢饮而非大口牛饮）能显著减少夜尿打扰；若夜间醒来，只要在 1 小时内重新入睡且整体睡眠窗口可调整，便不会造成持续伤害。此外，Poe 提醒当前消费级可穿戴监测设备判断具体睡眠阶段的准确率仅在 70% 左右，宜作为日常作息趋势的大体参考，而非绝对的生活干预标准。 本期节目确立了睡眠作为动态神经工程系统的核心地位：慢波阶段提供物理层面的代谢废料冲刷与蛋白构建，而 N2 与 REM 阶段则完成记忆编码的分布式归档与情绪脱敏。其论述的核心局限与现实考量在于，睡眠益处高度依赖于体内生物钟阶段的严密对齐，无法单纯依靠白天的随意补觉来弥补生理窗口的错失；同时，商用睡眠监测设备有限的分类精度，也要求大众应更多依托规律作息和主观状态评估，而非过度依赖算法读数。

[2:23] Dr. Gina Poe: Stage 2 sleep is really cool because that has something called sleep spindles and K-complexes... It's a unique time when our hippocampus—the sort of like the RAM of our brains—writes it to a hard disk, which is the cortex. [12:28] Dr. Gina Poe: I liken it to a washing machine cycle... You don't want to open the washing machine and try and function, put them on and wear them around while they're soaking wet and full of soap. You have to wait until the cycle is through.

[0:00] Welcome to Huberman Lab Essentials, [0:01] [music] where we revisit past episodes [0:04] for the most potent and actionable [0:06] science-based tools for mental health, [0:08] physical health, and performance. [0:11] I'm Andrew Huberman and I'm a professor [0:13] of neurobiology and opthalmology at [0:15] Stanford School of Medicine. And now for [0:17] my discussion with Dr. Gina Poe. Dr. [0:20] Gino Poe, welcome. [0:21] >> Thank you. [0:22] >> I've really been looking forward to this [0:24] conversation. I know that many people [0:25] are going to be excited to learn about [0:27] your work as it relates to sleep, as it [0:30] relates to problem solving, creativity, [0:32] and a number of other important topics. [0:34] To start things off, I would love for [0:36] you to educate us a bit about this thing [0:38] that we are all familiar with and yet [0:40] very few of us understand, which is [0:42] sleep. Could you describe the various [0:44] phases of sleep that exist, what [0:46] distinguish them, and perhaps frame this [0:49] within the context of what would a [0:51] perfect night's sleep look like? [0:53] >> All right. So sleep is really different [0:55] from wakefulness and in fact can't be [0:57] replaced by any state of wakefulness [1:00] that we've been able to come up with so [1:01] far. Our brain chemistry is completely [1:04] different and in the different stages of [1:06] sleep which there are is nonREM and REM [1:08] are the two major states of sleep. Those [1:10] two states are entirely different from [1:12] one another too. And even within nonREM [1:14] there are three states. Stage one, which [1:17] is what you slip into when you first [1:19] falling asleep. It's dozing. There's [1:21] kind of an interesting rhythm that goes [1:23] on in the brain. It's kind of a fast [1:24] gamma rhythm. And then there's stage [1:27] two, which is a really cool state. We [1:29] sort of used to ignore sleep researchers [1:32] because it was a transient state between [1:34] wakefulness and the deep stage three [1:36] slowwave sleep, which is the most [1:37] impressively different, which is when [1:39] big slow waves sweep through our brain. [1:41] And now we've realized that it cleans [1:43] our brain. Um, one of the things that [1:45] those big slow waves do is cleans our [1:47] brain and does other really important [1:48] things to restore us from a day of [1:51] wakefulness. And then REM sleep, which [1:53] is the most popular because that's where [1:55] we have the most active dreams. When you [1:57] wake up someone out of REM sleep, [1:59] they'll almost always report having [2:01] dreamed something really bizarre. That's [2:03] called REM sleep, rapid eye movement [2:05] sleep. So those are the four states of [2:07] sleep of human sleep. And we cycle [2:09] through them every 90 minutes or so. And [2:12] then we start over again. And we have [2:15] about five of those per night for a [2:16] perfect night's sleep. Four or five, [2:18] something like that. So a perfect [2:20] night's sleep is 7 and 1 half, 8 hours. [2:23] And [2:23] >> what about the sleep where we are [2:25] lightly asleep [2:27] >> and we might have a dream that has us [2:30] somehow thinking about movement or that [2:31] we jolt ourselves awake. That often [2:33] happens early in the night, right? [2:34] >> Yeah. That's the first stage, stage one [2:37] and stage two of sleep. And stage two [2:39] sleep is really cool because that has [2:43] something called sleep spindles and K [2:45] complexes. And what sleep spindles are [2:47] are a little of activity that's 10 to 15 [2:51] hertz in frequency. It's a conversation [2:54] between the phalamus and the cortex. The [2:57] phalamus is the gateway to consciousness [2:59] and the neoortex, you know, processes [3:01] all our cognition. And if you wake up [3:03] out of that state, you will often report [3:05] a dream like a hallucination style [3:08] dream. It won't be a long dream report [3:10] like you have out of REM sleep, but it [3:12] will be some hallucination state. [3:14] >> Are they quite different than the [3:16] patterns of sleep and dreaming that [3:18] occur later in the night or toward [3:19] morning? [3:20] >> There is some evidence that the first 4 [3:23] hours of sleep are very important for [3:26] memory processing. And in fact, if [3:28] you've learned something new that day or [3:31] have experienced a new sensory motor [3:34] experience, then your early sleep dreams [3:38] will incorporate that experience much [3:40] more than the later sleep dreams. Later [3:42] as that memory gets consolidated from [3:45] the early structures which are the [3:47] hippocampus deep in the temporal lobe to [3:49] the cortex in a distributed fashion that [3:53] memory seems to move from that [3:56] hippocampus to the cortex and also the [3:58] dreams that incorporate that memory also [4:00] move later in the night. So [4:02] >> there was a great study by Sedarto Rayo [4:06] who studied the consolidation of [4:08] memories from the hippocampus to the [4:10] cortex in a rat across the period of a [4:13] full day sleep because rats sleep in the [4:16] daytime. And he found that each [4:18] subsequent rim sleep period moved that [4:22] memory from the hippocampus to the f [4:25] first area that projects to it and then [4:27] the second area and then the third area. [4:29] and you can see the memory moving um [4:32] through the throughout the sleep. [4:34] >> There's a number of different hormones [4:36] associated with the different stages of [4:38] sleep. We know that melatonin is a [4:40] hormone [4:41] >> of night time [4:41] >> of nighttime that makes us sleepy. What [4:44] about growth hormone release? When does [4:45] that occur during sleep? [4:47] >> So growth hormone release happens all [4:48] day long and all night long. But the [4:51] deep slow wave sleep that you get, the [4:53] very first sleep cycle is when you get a [4:56] big bolus of growth hormone release. and [4:58] in men and women equally. And if you [5:02] miss that first deep slow wave sleep [5:04] period, you also miss that big bolus of [5:07] growth hormone release. And you might [5:09] get ultimately across the day just as [5:12] much overall growth hormone release. But [5:14] andologists will tell you that big [5:16] bololises do different things than a [5:19] little bit eaked out over time. There's [5:21] also a big push to synthesize proteins. [5:24] So that's when the protein synthesis [5:27] part that builds memories for example in [5:30] our brain happens in that first cycle of [5:33] sleep. So you don't want to miss that [5:35] especially if you've learned something [5:36] really big and needs needs more synaptic [5:39] space to encode it. [5:40] >> How would somebody miss that first 90 [5:42] minutes [5:43] >> depriving themselves? Yeah. So, so let's [5:45] say I normally go to sleep at 10:00 p.m. [5:48] >> and then from 10 to 11:30 would be this [5:50] first phase of sleep and that's when the [5:52] growth hormone big bololis of growth [5:54] hormone would be released. Does that [5:55] mean that if I go to sleep instead at [5:58] 11:30 or midnight that I miss that first [6:00] phase of sleep? Why is it not the case [6:02] that I get that first phase of sleep [6:04] just simply starting later? Every cell [6:06] in our body has a clock and all of those [6:08] circadian clocks are synchronized and so [6:11] our cells are ready to respond to that [6:15] growth hormone release at a particular [6:17] time and if we miss it and it's a time [6:19] in relation to melatonin also. So if you [6:22] miss it, yeah, you might get some growth [6:25] hormone release, but it's occurring at a [6:27] time when that your clock has already [6:28] moved to the next phase. And so it's [6:31] it's just a clock thing. So what this [6:33] means is that we should have fairly [6:35] consistent bedtimes in addition to [6:36] fairly consistent wake times. Is that [6:38] right? [6:39] >> Exactly. And in fact, one of the best [6:41] markers of good neurological health when [6:44] we get older is consistent bedtimes. [6:47] >> What other things inhibit growth hormone [6:49] release or other components of this [6:51] first stage of sleep? Are there things [6:53] that I perhaps do in the preceding hours [6:55] or the preceding day like ingest [6:57] caffeine or alcohol that can make that [7:00] first stage of sleep less effective even [7:02] if I'm going to sleep at the same time? [7:03] >> Alcohol definitely will do that because [7:05] alcohol is a REM sleep suppressant and [7:07] it even suppresses some of that stage [7:09] two transition to REM with those sleep [7:11] spindles. And those sleep spindles, we [7:13] didn't talk about their function yet, [7:14] but they're really important for moving [7:17] memories to our cortex. It's a unique [7:19] time when our hippocampus the sort of [7:22] like the RAM of our brains writes it to [7:25] a hard disk which is the cortex and [7:27] there it's a unique time when they're [7:29] connected. So if you don't want to miss [7:31] that you don't want to miss REM sleep [7:33] when is also a part of the consolidation [7:35] process and alcohol before we go to [7:37] sleep will do that until we've [7:39] metabolized alcohol and put it out of [7:41] our bodies it will affect our sleep [7:44] badly. What about the second and third [7:47] 90minute blocks of sleep? Is there [7:48] anything that makes those um unique? [7:51] What what is their signature? Besides [7:53] the fact that they come second and third [7:54] in the night, [7:55] >> there's more and more REM sleep the [7:57] later the night we get. There's also a [8:00] change in hormones. You know, the growth [8:01] hormone and melatonin levels are [8:04] starting to decline, but other hormones [8:06] are picking up. So, it is a really [8:09] different stage that you also don't want [8:11] to short change yourself on. And I think [8:14] that's the stage many studies are [8:16] showing that those are the times in [8:18] sleep when the most creativity can [8:20] happen. That's when our dreams can [8:22] incorporate and put together old and new [8:25] things together into a new way and our [8:28] schema are built during that time. So [8:31] that's when your computer of your brain [8:33] is opening folders and comparing [8:35] documents seeing if is there anything [8:37] the same? Are these two documents look [8:39] very much the same? that there's a [8:41] little bit of difference and it can it [8:43] can link those conceptually. So that [8:45] that's probably one of the origins of [8:47] creativity is finding things that are [8:49] related maybe just linked a little bit [8:52] and you can find that link and [8:54] strengthen it if it you know makes your [8:57] schema interesting and different. [8:59] >> Many people including myself tend to [9:01] wake up maybe once during the middle of [9:03] the night to use the restroom. I've [9:05] tried to drink less fluid before going [9:07] to sleep. I've heard also the impulse to [9:09] urinate is also dictated by how quickly [9:11] you drink fluid, not just the total [9:12] volume. So, I've switched to um sipping [9:14] fluids more slowly for my last beverage [9:16] of the day, which seems to help. Is [9:18] there any known detriment to this middle [9:21] of the night waking or should we [9:23] consider it a normal feature for some [9:25] people's sleep architecture? [9:26] >> Yeah, I think we shouldn't worry about [9:28] it. Actually, sleep is really incredibly [9:30] well homeostatically regulated. And so [9:33] really don't worry about how much you're [9:35] sleeping as long as you're not [9:37] intentionally depriving yourself of [9:38] sleep by doing something really [9:40] rewarding and exciting because even that [9:42] is stressful to your body and deprivives [9:44] you of a lot of the things we're talking [9:45] about. It's absolutely normal to wake up [9:48] at least once in the middle of the night [9:49] to go to the bathroom. And as long as [9:51] you can get back to sleep in a [9:53] reasonable amount of time or even if it [9:55] takes you an hour, don't worry about it. [9:57] as long as you have a lifestyle that [9:59] allows you to then make up that sleep [10:01] either the next morning or the next [10:03] night um or going to bed a little [10:05] earlier. [10:06] >> What is unique perhaps about the [10:08] architecture of dreams and sleep in the [10:11] let's say the last third of the night or [10:13] the or the second half of the night, [10:14] >> right? Yeah. In the second half of the [10:15] night, we have longer REM sleep periods [10:18] and those are considered the deepest [10:21] sleep. Even though slowwave sleep, big [10:24] slow wave is considered deep. It is [10:26] deep. [10:26] >> Yeah. They call slowwave sleep deep [10:28] sleep and REM sleep rapid eye. But now [10:30] you're telling me that REM sleep is [10:31] actually the deeper sleep. [10:32] >> The reason why you call slowwave sleep [10:34] deep sleep is because it's difficult to [10:36] arouse people out of that state. And [10:38] when you do arouse them out of that [10:40] state, they're most often confused and [10:42] just want to go back into sleep and can [10:44] go back pretty easily. If you arouse [10:47] someone out of REM sleep, they're more [10:49] likely to report something that was [10:50] really kind of almost like wakefulness. [10:52] It's it was so vivid. And maybe one of [10:54] the reasons why REM sleep is deeper is [10:57] especially in adults and older people [10:59] that deep slowwave sleep goes away. So [11:02] it's not as deep. It's not as big. The [11:04] slow waves aren't as large which is [11:06] probably problematic but we are not [11:08] sure. And so then REM sleep becomes the [11:11] deepest stage. [11:12] >> We are paralyzed during REM sleep. [11:14] Correct. [11:15] >> Yes. normally paralyzed and that's [11:16] really good because that's the time when [11:19] we're actively dreaming storyline dreams [11:22] and we could hurt ourselves. We're [11:24] actually really cut off from the outside [11:26] world in terms of responding to say this [11:30] table or window or a door and so [11:33] different from sleepwalking which is out [11:35] of slow sleep. Out of slow sleep that [11:37] sleepwalking is a mixture between sleep [11:39] and wakefulness. You can cook a full [11:41] meal, um, drive your car while you're in [11:44] deep, slow sleep. It's scary because you [11:46] never know what you're going to do. You [11:48] don't have voluntary voluntary control [11:50] over it. You have no conscious control [11:51] over it, but you can actually safely [11:54] navigate um some situations in [11:57] sleepwalking and actually have a [11:59] conversation. Although it may not make [12:00] much sense when you're sleeptalking, in [12:03] REM sleep, you're not processing the [12:06] outside world. and instead um when [12:08] you're acting out your dreams, you could [12:10] be doing things like walking through a [12:12] plate glass window or falling off of, [12:15] you know, down a stairs um things like [12:17] that. So, you really want your muscles [12:20] to be inactivated during REM sleep, [12:23] otherwise you will act out those dreams [12:25] and really hurt yourself or your bed [12:27] partner. So, as people start to approach [12:31] morning or the time when they normally [12:32] would wake up, I've heard that it's [12:34] important to, if possible, complete one [12:37] of these 90minute cycles prior to waking [12:40] up. That is, if you set your alarm for [12:42] halfway through one of these 90minute [12:44] cycles that come late in the night of [12:46] sleep that it can lead to um rather [12:50] groggy patterns of waking. [12:52] >> It's called sleep inertia. When we wake [12:53] up out of the wrong state, I liken it to [12:56] a washing machine cycle. This 90 minute [12:58] cycle is like a washing machine cycle. [13:00] And the first part is to add water, [13:02] right? Then your clothes are soaking [13:03] wet. You don't want to open the washing [13:05] machine and try and function, put them [13:07] on and wear them around while they're [13:09] soaking wet and full of soap. So you [13:11] have to wait until the cycle is through [13:14] before you can well actually [13:16] put it in the dryer, too. Um before you [13:19] want to wear them. So you can function. [13:22] It just takes a little while for those [13:24] clothes, that brain to dry out so you [13:26] can actually function well. But it's [13:28] better to wait through the whole cycle [13:30] is complete. So that's why you want to [13:33] set that 90 90 minute alarm clock. And [13:36] again, that's around 90 minutes because [13:39] the first stage of sleep, the first [13:41] cycle of sleep is actually a little [13:42] longer, more like 105, 110 minutes. But [13:46] then the second ones and third ones, [13:48] they get sort of shorter and shorter as [13:49] the night goes on. And in the last few [13:52] cycles, you're just doing the N2 REM [13:55] sleep cycle, which takes less time. And [13:57] if you wake up out of REM sleep, there's [13:59] usually no problem cognitively. [14:01] >> Are you a fan of of sleep trackers? [14:04] >> Yeah. [14:04] >> Yeah. Do you use one? [14:05] >> I have one on. I don't live my life by [14:08] them because the best ones right now are [14:11] about 70% effective at [14:13] >> staging your sleep. So 70% it's okay. [14:16] It's okay. But um take it with a grain [14:18] of salt is what I'm saying. tell us a [14:20] little bit more about the wash out that [14:22] occurs in the brain during sleep and [14:24] perhaps if there are any ways to ensure [14:25] that it happens or to ensure that it [14:27] doesn't happen and obviously we want [14:29] this to happen. [14:30] >> Yeah, we talked about the circadian [14:31] clock and how certain things happen at [14:34] certain times. Well, one of the things [14:35] that happens when we're awake and [14:37] talking to each other is that there's a [14:40] lot of plasticity. there's something [14:42] that I'm learning from you today and [14:43] you're learning from me and that changes [14:46] our synapses and it changes the way our [14:48] proteins are going to be folded and [14:52] changed during sleep. This process [14:54] actually uses a lot of ATP, the power um [14:58] structure, the fuel of the brain. It [15:00] unfolds also proteins while we're doing [15:02] this, while we're using them. And so [15:05] during that first part of the night, uh, [15:07] when we first fall asleep, in the first [15:08] 20 minutes or so, we're building that [15:11] adenosine back into ATP. And that's, uh, [15:14] probably why power naps are called power [15:16] naps because we're actually rebuilding [15:19] the power. And then we're also cleaning [15:23] out through the deep slow waves of slow [15:25] wave sleep. We're cleaning out all those [15:27] misfolded proteins, unfolded proteins, [15:29] and other things that get broken down [15:32] and, you know, need to be rebuilt when [15:35] we're asleep because of its use during [15:38] wakefulness. So I liken that to, you [15:41] know, having a big party during [15:43] wakefulness and you need all those [15:44] partygoers to leave in order to do the [15:46] cleanup. What happens when a neuron is [15:48] firing is that it expands. The membrane [15:52] expands a little bit. It becomes more [15:53] translucent. That's how we know, one of [15:55] the ways we know that neurons expand [15:57] when they fire. And so every action [15:59] potential, the membrane expands a little [16:02] bit as sodium brings water into the [16:04] cell. And then when they're silent, they [16:06] contract. And so in during slow waves, [16:09] the cool thing is that the reason why [16:11] you can measure them is that all the [16:12] neurons at the same time, not all of [16:14] them, but a good portion of them are [16:17] firing at the same time and silent at [16:19] the same time. And so you think about [16:21] that as contracting and expanding all at [16:23] the same time. It's kind of like a [16:25] billagege pump of the brain. So that can [16:27] pump out. GIA are also really important [16:30] for this in terms of cleaning up debris [16:33] and transferring it to where it needs to [16:35] go. So, um, so I think of it actually as [16:38] a billagege pump cleaning out our brain. [16:41] >> Here you're talking about literally an [16:43] expansion and a contraction of the [16:44] neurons in unison [16:46] >> and pushing the fluid through [16:48] >> cleaning out any misfolded proteins or [16:50] debris that might occur [16:52] >> on the basis of these metabolic [16:53] pathways. [16:55] >> And the consequence of that is to to [16:57] what to leave the brain in a state of [17:00] more pristine action for the next day. [17:02] Is that right? [17:03] >> Yeah. You think of it again like a party [17:05] and if you don't clean up after that [17:07] party and you try and hold another one [17:08] the next day, it's going to get more [17:10] clogged. You know, it's people have a [17:12] harder time moving around and enjoying [17:14] themselves. And um if that builds up day [17:17] after day, you know, it's going to be [17:18] cognition that would be the partygoers [17:21] moving around becomes [17:24] hard. [17:25] >> And so this um build pump that you [17:27] describe is associated with the big slow [17:29] waves of slowwave sleep. So this is [17:32] going to occur more or less in the first [17:33] third of the night. Is that right? [17:35] >> That's right. [17:35] >> And is this similar to the case with [17:37] growth hormone where if you go to sleep [17:39] later than you would normally, you miss [17:42] the wash out. You It's not You don't [17:44] delay it. You miss You miss the wash [17:46] out. [17:46] >> That's right. That's right. So if you go [17:48] to sleep at 1 or 2 in the morning, your [17:50] sleep is still going to be dominated by [17:52] N2 and REM sleep, not by slow wave [17:55] sleep. So you need to you need to get [17:57] that first bit of sleep. Would a caveat [18:00] to that be if somebody normally goes to [18:02] sleep at 1 or 2 a.m. and wakes up at [18:04] 10:00 a.m. if that's their normal sleep [18:06] cycle? [18:06] >> Yeah, it should be okay. Somebody would [18:09] want to do a sleep study with people who [18:11] do that normally and see if also the [18:13] melatonin release is later and the [18:15] corticostone rise that's happens [18:17] normally in the morning also happens [18:19] later. So if everything shifted good, [18:23] >> I'd love for you to tell us about this [18:25] incredible structure in the brain which [18:27] is the locus ceruius and hopefully tell [18:30] us a little bit about its relationship [18:31] to epinephrine aka adrenaline. [18:34] >> Yeah. So um the locus curillus is filled [18:37] with neurons that have in them [18:39] norepinephrine which is the brain's [18:41] version of epinephrine or adrenaline. [18:43] It's also called noradrenaline. And what [18:46] it does is it just like adrenaline in [18:48] the rest of our bodies, it helps prime [18:49] us to respond to our environment. So [18:53] when locus rules neurons fire and fire [18:56] in a burst, we can switch our attention [18:59] and they will fire in a burst if for [19:01] example a loud noise happens in the [19:03] middle of your concentrating on [19:04] something. It fires and it helps you [19:06] switch your attention to that thing and [19:08] then learn quickly from it. So it's [19:10] really important in a stress response. [19:12] It helps us do quick one trial learning. [19:15] But just tonic levels are signature of [19:18] wakefulness and alertness. So too much [19:21] is panic with the locusillis activity. A [19:24] burst is switching attention. And then [19:27] tonic levels are sustained constant [19:30] attention. And then when we go to sleep [19:32] the locus slows and come goes from about [19:36] on average um two hertz to about one [19:39] hertz one cycle per second tonically and [19:42] then when we go into REM sleep it's the [19:44] only time when it shuts off completely [19:46] and it appears that that complete [19:49] silence is really really important for a [19:52] number of things and the main thing that [19:54] I think it's important for is the [19:57] ability to erase and break down synapses [20:00] that are no longer working for us. So [20:02] they encode things that are false now or [20:05] they are encoding things that we um [20:08] learned in the novelty encoding pathway [20:11] in of our of our brain that have now [20:14] been consolidated to other pathways and [20:16] so we need to now erase them from the [20:18] novelty encoding pathway and that is [20:20] really really important for being able [20:22] to continue to learn things all of our [20:24] lives. So like erasing that RAM um or [20:27] that I don't know what do you call those [20:29] discs that you stick into computers that [20:31] >> oh thumb drives thumb drives erasing [20:33] your thumb drive. So that thumb drive is [20:34] what you carry around all day long and [20:37] then during sleep you write that thumb [20:39] drive to the cortex to the long-term [20:41] memory structures and you need to [20:43] refresh that thumb drive and that's what [20:45] happens during REM sleep when the locus [20:47] realis is off and so if we're not able [20:49] to do that we fill up that RAM um really [20:53] quickly or that thumb drive really [20:54] quickly and we're not able to learn new [20:57] things even memories that are years past [21:00] if you're never able to downscale that [21:02] novelty encoding structure and purge it [21:05] from that traumatic memory, it will stay [21:07] fresh and new and then become [21:10] maladaptive. [21:12] >> What approaches are you aware of that [21:14] can um turn down the output of locus [21:18] ceruius during these phases of sleep and [21:21] allow late stages of sleep each night to [21:24] have their maximum positive effect. [21:26] >> Is there anything that I can do besides [21:27] avoiding um serotonin or noradinergic [21:30] compounds? Well, I would also avoid [21:33] anything just prior to going to sleep [21:35] that might excite those systems. So, a [21:38] lot of novelty, stress inducing video [21:40] games. Try and enter sleep with as much [21:46] calm as you can. So, maybe deep [21:47] breathing exercises. That's a beautiful [21:50] way to calm your sympathetic fight [21:52] orflight system is deep breathing. And [21:55] if there's a way you can make your [21:57] sympathetic nervous system calm down [21:59] before you go to sleep, might free for [22:02] you meditation or deep breathing [22:03] exercises. Might be for some a warm bath [22:06] or a comforting book. Nothing too [22:08] exciting, but also nothing too boring [22:11] perhaps. Just something right in the [22:12] middle which makes you feel happy and [22:15] calm is what you should do. And if you [22:18] instead go to sleep while you're anxious [22:20] or hyped up, then your sleep could [22:23] become maladaptive. [22:25] >> I'd love for you to share with us a [22:26] little bit more about um these spindles [22:28] that have come up a few times. [22:30] >> The density of our sleep spindles, the [22:31] number that we produce per minute is [22:34] well correlated with our intelligence in [22:36] the first place. And that no matter what [22:38] your intelligence is and no matter what [22:40] your sleep spindle density is, if you [22:42] learn something during the day and [22:44] increase your sleep spindle density, [22:46] it's really almost perfectly correlated [22:49] with our ability to consolidate that [22:50] information and and incorporate it into [22:53] the schema that we already have in our [22:55] brain. So if you try and learn something [22:57] new, even if your sleep spindle density [22:59] at baseline is great, if you don't [23:01] increase your sleep spindles that night, [23:03] you're not going to, you know, use sleep [23:05] to really incorporate it. One of the [23:07] things we now know through some great [23:09] studies by Julie Seep and Anita Luthie [23:12] is that sleep spindles are accompanied [23:14] by an incredible [23:17] plasticity out in the distal dendrites, [23:19] the listening branches of our neurons [23:22] that listen to other cortical areas. So [23:25] there are proximal dendrites in our [23:27] neurons that listen to the external [23:30] world and are conducted through the [23:32] phalamus and then there are distal [23:33] dendrites which listen to an internal [23:36] kind of you know conversation that's h [23:38] happening in our brains. It's kind of [23:40] you know our internal state really and [23:43] during sleep spindles that's when those [23:45] distal dendrites are able to best learn [23:49] from other cortical areas and from the [23:51] hippocampus. It is during sleep spindles [23:53] that the hippocampus and the cortex are [23:55] best connected and when that plast [23:58] incredible plasticity can happen. So [24:00] when I talk about schema, that's a [24:02] cortical cortical thing. There's big [24:04] surges of calcium into those distal [24:06] dendrites and where plasticity happens [24:09] in just huge amounts during that sleep [24:12] spindle stage of sleep which is N2 [24:14] stage. There's another excitatory event [24:17] that comes all the way from the brain [24:18] stem and projects everywhere in our [24:20] cortex which is called PGO waves which [24:23] we should generalize to P waves because [24:25] they come from the ponds and go to the [24:27] phalamus and then the cortex happens all [24:29] over the brains and that is where [24:31] glutamate which is a major excitatory [24:33] neurotransmitter involved in learning [24:35] and plasticity is being released in big [24:38] amounts also in those distal dendrites. [24:40] Pwaves and spindles work together to [24:43] cause plasticity and sew our schema [24:45] together which could be the origins for [24:48] insight and creativity. Now when P waves [24:50] were first discovered, it was thought to [24:53] be random because this small area that [24:56] generates P waves all over the brain [24:58] projects all over the phalamus and [25:00] causes P waves all over. And you don't [25:02] measure P waves all over the brain at [25:04] the same time. In fact, it's just seems [25:06] sporadic and random. And Pwaves also [25:09] happening even more during REM sleep, [25:11] rapid eye movement sleep. So that's why [25:14] people think that REM dreams are so [25:16] random is because these P waves are [25:18] random and they could generate dreams [25:22] because they're an internal source of [25:25] excitation that kind of replaces the [25:27] outside world during our dream state. [25:30] And so these Pwaves, if they are random, [25:33] could be the underlying reason why [25:34] creativity can happen there is because [25:36] we're randomly activating, co-activating [25:39] different things in our brain that we [25:41] can then sew together. Um, but it might [25:43] not be as random as we think. So that's [25:46] a caveat there. [25:47] >> So locus ceruius is suppressed. So we [25:49] can't release norepinephrine. We can't [25:51] act out our dreams. This almost starts [25:53] to sound like a little bit of a built-in [25:56] um while sleeping trauma therapy. Mh. [25:59] >> So please if there's anything about [26:02] locust ceruius and and dreams and that [26:05] can help people basically extinguish [26:08] traumas or traumatic features to to real [26:10] life events and we definitely want to [26:11] know about them. [26:12] >> Yeah. Yeah. Well, I think one of the [26:15] things that people thought might help [26:17] after a trauma like a school shooting or [26:19] whatever you know car accident is to [26:21] talk about it and in but in fact that [26:24] ended up being counterproductive. And I [26:26] think one of the reasons why it was [26:27] counterproductive is because it didn't [26:29] take you them back down. It brought them [26:32] up and continued to reactivate the [26:34] emotions of it, but then didn't [26:36] emphasize the safety fact that it's over [26:39] or help them work through how they might [26:41] avoid it again in the future to calm the [26:44] sympathetic nervous system down again [26:46] before they went to sleep. And and in [26:48] any none of these studies has sleep ever [26:50] been considered, but to me that's the [26:53] key part is bringing down your [26:55] sympathetic nervous system before you go [26:57] to sleep so that your sleep can be [26:58] adaptive, your locus can shut off like [27:01] it normally does or should do and then [27:04] able to erase the novelty of it. The [27:08] other thing that I just mentioned a [27:10] minute ago was that the emotional system [27:12] is highly activated in REM sleep. And [27:14] that's definitely true. And that might [27:16] seem counterproductive in terms of, you [27:18] know, the nightmares and and how to help [27:22] REM sleep be a therapeutic thing rather [27:24] than reinforcing the emotionality of the [27:27] trauma. And I think the key to that [27:29] again is the absence of norepinephrine. [27:32] So even though the emotional system is [27:35] in high gear, without norepinephrine, [27:38] you can actually divorce those highly [27:41] activated emotions from the cognitive [27:45] parts of the memory that you have just [27:48] written out in that N2 stage of sleep [27:50] when the sleep spindles are going. So [27:52] you've just now consolidated the [27:54] information that you'll need to survive [27:56] and to you know to make that adaptive [27:59] and now you need to divorce from that [28:02] schema and from that semantic parts of [28:05] memory the emotional part because [28:07] whenever you remember something it's [28:09] fine if you remember the being emotional [28:12] at the time but you don't want to bring [28:15] back and sew into that memory all of the [28:17] same emotional systems. You don't want [28:18] to bring back, you know, the heart rate [28:21] changes and the sweating and and all of [28:23] that. You want to be able to remember [28:25] all the parts of it and even remember [28:27] that you were traumatized and that you [28:29] did cry and that you did have, you know, [28:32] your heart was racing. But when you're [28:34] talking about it years later, you don't [28:36] want to have to relive all that. [28:37] Otherwise, who who would ever want to [28:40] recall a traumatic memory because you're [28:42] basically putting yourself through the [28:43] same trauma, which is what people with [28:45] PTSD have. They don't want to recall [28:48] this traumatic memory because it's [28:49] reliving it like it's just happening [28:51] again. So that's what we're thinking is [28:54] that the emotional parts are not able to [28:57] be divorced because the norphan system [29:00] is not downscaled during REM sleep. And [29:02] so that REM sleep serves to instead [29:06] reinforce and in fact amplify the [29:08] emotions because your your emotional [29:10] system is up. Look, Serilla says, "Hi, [29:13] re- sewing in every night the [29:16] emotionality of those memories and with [29:18] the memory itself." [29:20] >> I must say, you've taught us a [29:22] tremendous amount in um in a relatively [29:24] short amount of time about the [29:26] architecture of sleep, the different [29:28] phases, such a wealth of information, [29:30] and much of it that's actionable for [29:32] people. So, I want to say thank you for [29:34] taking the time to sit down and have [29:35] this conversation that so many people [29:37] are sure to benefit from. I know I speak [29:39] for everybody when I say thank you so [29:42] much. [music] [29:42] >> Thank you so much.