95% Lifestyle, 5% Genes — and the Peptide That Could Feed a Tumor
阿尔茨海默：95% 是生活方式，剩下 5% 才是基因
Louisa Nicola spent thirteen years scanning brains. The number she keeps coming back to isn’t 60 million dementia cases worldwide. It’s the share of Alzheimer’s driven by lifestyle rather than the rare deterministic genes — the good news and the bad news at once. Three hours on Shawn Ryan, distilled into a prevention playbook with a peptide warning embedded in it.
神经生理学家 Louisa Nicola 扫了十三年大脑。她反复回到的不是“全球 6000 万人患痴呆”这个数字，而是另一个——真正决定你会不会得阿尔茨海默的，不是几个罕见的“命中注定”基因，而是你日常怎么活。这是好消息，也是坏消息。三个小时的访谈，浓缩成一份预防手册，里面藏着一段关于黑市肽的硬核警告。
The Stakes
规模——先把数字摆出来
About 60 million people worldwide live with dementia today. By 2050 that count is projected to triple. In the US, around 7 million Americans currently have Alzheimer’s — on track to nearly double by 2060. The disease kills more Americans annually than breast and prostate cancer combined. Two-thirds of patients are women.
全球约 6000 万人患有痴呆症。到 2050 年，这个数字预计翻三倍。美国现在约有 700 万阿尔茨海默患者，2060 年预计翻一番。这种病每年在美国夺走的生命，比乳腺癌和前列腺癌加起来还多。三分之二的患者是女性。
95% Lifestyle, Not Genes
95% 是生活方式，不是基因
Only 3 to 5% of Alzheimer’s cases are explained by the rare deterministic mutations — presenilin 1, presenilin 2, amyloid precursor protein. The remaining cases are driven overwhelmingly by lifestyle and modifiable risk. The 2024 Lancet Commission estimates that addressing 14 such factors could prevent or delay roughly 45% of all dementia globally. Nicola’s framing is sharper still: this is not a disease of old age. It is a largely preventable disease that happens to surface in old age because the damage compounded silently across your thirties and forties.
只有 3% 到 5% 的阿尔茨海默病例，可以用三个“命中注定”的基因突变来解释——早老素 1、早老素 2、淀粉样前体蛋白。剩下的，全部由生活方式和可改风险驱动。2024 年《柳叶刀》委员会估计，调整 14 个可改因子，可以让全球约 45% 的痴呆延迟或避免。Nicola 把话说得更狠：这根本不是“老年病”。它在七十岁出现，是因为你三四十岁那十几年里，伤害一点一点累积下来——那时候才是真正决胜的窗口。
A New US Case Every 65 Seconds
美国每 65 秒新增一例
Globally: ~60M with dementia today, projected ~150M by 2050. US: ~7M with Alzheimer’s today, on track for ~13.9M by 2060. The disease kills more Americans annually than breast cancer and prostate cancer combined. The catastrophe is already here; the curve is steeply up.
全球：现在约 6000 万痴呆患者，2050 年预计 1.5 亿。美国：现在约 700 万阿尔茨海默患者，2060 年预计接近 1400 万。这种病每年杀死的美国人，比乳腺癌和前列腺癌加在一起还多。灾难早就到了，曲线还在陡升。
Estrogen Was Doing Real Work
雌激素一直在“护脑”
Two-thirds of Alzheimer’s patients are women. Estrogen is neuroprotective; menopause removes it. The hormonal cliff reshapes brain risk across the next two decades. This is also why female ApoE4 carriers, particularly those with two copies, see the steepest risk multipliers in the published literature.
阿尔茨海默患者里，三分之二是女性。雌激素本身有神经保护作用，绝经把它撤掉。这个激素断崖会在之后二十年里重塑大脑风险版图。这也是为什么女性 ApoE4 携带者——特别是两个等位基因都是 E4 的——在公开文献里跑出最高的风险倍数。
Dementia Is the Symptom Set, Not the Disease
痴呆是“症候群”，不是“一种病”
The word "dementia" comes from "demented" — slowed processing, memory decline, cognitive deterioration. It’s an umbrella term, not a diagnosis. Underneath sit several distinct diseases: Alzheimer’s dementia (most common), vascular dementia (second), Parkinson’s dementia, Lewy-body dementia, frontotemporal dementia. Same surface symptoms; different mechanisms; different treatments.
“痴呆”这个词来自“失智”——思维变慢、记忆下滑、认知能力退化。它是一把“伞”，不是某一种诊断。伞下面是好几种不同的病：阿尔茨海默痴呆（最常见）、血管性痴呆（第二）、帕金森痴呆、路易体痴呆、额颞叶痴呆。表面症状相似，机制完全不同，治法也完全不同。把它们当成一回事，方案就跑偏。
Genes Aren’t Destiny
基因不是命运书
The genetic story everyone knows: ApoE4 is the Alzheimer’s gene. About 25% of the population carries at least one copy. The story Nicola tells is more uncomfortable, and more useful: West African cohorts with very high ApoE4 frequencies show among the lowest Alzheimer’s rates measured. Lifestyle dominates penetrance.
大家都听过的版本：ApoE4 就是“阿尔茨海默基因”。大约 25% 的人至少带一个拷贝。Nicola 讲的版本更扎人，也更有用：西非人群里 ApoE4 频率特别高，但阿尔茨海默发病率反而是全球最低之一。基因把枪上膛了，是不是真扣扳机——靠生活方式。
One Copy 2–3×. Two Copies 10×. Two + Female ~14×.
一个拷贝 2–3 倍。两个拷贝 10 倍。女性两个拷贝——约 14 倍。
You inherit one ApoE allele from each parent. Variants are E2, E3, E4. One copy of E4 raises Alzheimer’s risk roughly 2 to 3×. About 2–3% of the population are homozygous (two copies); their risk runs roughly 8 to 15× across the published cohorts. Nicola cites about 14× for female homozygotes — at the upper end of replicated findings on the female-specific multiplier. Chris Hemsworth has two copies; he made a documentary about it because he doesn’t want to lose his mind.
每个人从父母各得一个 ApoE 等位基因。版本有 E2、E3、E4。一个 E4 拷贝把阿尔茨海默风险拉高大约 2 到 3 倍。约 2%–3% 的人是纯合子（两个都是 E4），他们的风险在不同人群里跑到 8 到 15 倍。Nicola 给女性纯合子的数字是约 14 倍——处在已有研究里偏高的一端，但确实在“女性专属乘数”反复被复现的范围内。克里斯·海姆斯沃斯就是双拷贝；他专门拍了纪录片，因为他不想失智。
High ApoE4. Low Alzheimer’s.
基因风险高，发病率反而低
Studies of West African ApoE4 carriers (Hendrie et al., Yoruba and Ibadan cohorts) found a high allele frequency paired with unusually low Alzheimer’s incidence — the opposite of what genetic determinism would predict. More daily activity, less processed food, fewer pesticides, harder physical living, stronger social ties. The genes loaded the gun. Lifestyle never pulled the trigger.
Hendrie 等人对西非约鲁巴和伊巴丹人群的研究发现：ApoE4 基因频率明显偏高，但阿尔茨海默发病率却异常低——基因决定论的反例。原因？日常活动多、加工食品少、农药少、生活更“硬”、社会关系更紧。基因把枪上了膛，但生活方式没扣那一下扳机。
Three Mutations Are Fate. ApoE4 Is a Tilt.
三个突变才是命；ApoE4 只是一个“偏置”
Roughly thirty to forty genes are implicated across the dementias. Three actual mutations — presenilin 1, presenilin 2, and amyloid precursor protein — do guarantee disease if inherited. Those are vanishingly rare. ApoE4 is a risk gene, not a deterministic one. Having it does not mean getting it. Not having it does not exempt you. Either way, the lifestyle levers in Chapters 3 through 6 are the same.
各种痴呆涉及大约三四十个基因。其中只有三个突变——早老素 1、早老素 2、淀粉样前体蛋白——是“遗传到就一定发病”的命运型。这种突变罕见到几乎可以忽略。ApoE4 是风险基因，不是命运基因。带着它不等于一定得；没带它也不等于不会得。不管你属于哪一类，第 3 到第 6 章里那几个生活方式杠杆，都是同一套。
The 14 Modifiable Risks
14 个可调杠杆
The 2024 Lancet Commission published the list. Fourteen factors. Modify them and roughly 45% of dementia could be prevented or delayed. The list isn’t exotic — it’s the boring everyday levers most people are pulling in the wrong direction.
2024 年《柳叶刀》委员会列出来了——十四个因子。调好它们，全球大约 45% 的痴呆可以延后或避免。清单一点也不玄：全是日常的小事，只是大多数人都在朝相反方向用力。
Lancet 2024: All Fourteen, in One Place
柳叶刀 2024：十四个因子，一次列全
Smoking · Hypertension · Obesity · Type 2 diabetes · Excess alcohol · Social isolation · Physical inactivity · Hearing loss · Depression · Low education / cognitive engagement · Air pollution · Head injury / TBI · Vision loss · High LDL cholesterol. The 2024 update added the last two; the first twelve had been on the list since 2020. The cards below dwell on the levers most people get wrong.
吸烟 · 高血压 · 肥胖 · 二型糖尿病 · 过量饮酒 · 社交孤立 · 缺乏运动 · 听力下降 · 抑郁 · 教育/认知刺激不足 · 空气污染 · 脑外伤 · 视力下降 · 高 LDL 胆固醇。最后两个是 2024 年新加的；前十二个从 2020 年起就在表上。下面挑出最容易“劝不住”那几条，单独说。
There Is No Safe Amount
不存在“安全剂量”
UK Biobank imaging studies have repeatedly found that moderate drinking — even ~7 drinks a week — correlates with reduced grey matter and white-matter changes on MRI. Not heavy drinkers. Moderate. Nicola’s position: no amount of alcohol is good for the brain. Period. The white matter is the myelinated “tree trunks” carrying signal; let it deteriorate and signal degrades.
英国 Biobank 的影像学研究反复显示：中等量饮酒——每周大约 7 杯——就能在 MRI 上看到灰质减少、白质改变。不是酗酒，是“社交饮酒”。Nicola 的立场很硬：对大脑来说不存在“安全饮酒量”，一杯都不该有。白质是大脑里包了髓鞘的“树干”，专门负责传导信号；让它退化，整套信号就跟着崩。
One Bad Night = +5% Amyloid
一晚没睡好 — 淀粉样蛋白 +5%
The 2018 PNAS study (Shokri-Kojori et al.) measured amyloid β in 20 healthy adults after a single ~31-hour night without sleep. Amyloid rose about 5%, particularly in the thalamus and hippocampus — the regions most vulnerable in early Alzheimer’s. Why? Amyloid clearance happens during deep sleep, through the perivascular space and into the lymphatic system. Skip the deep sleep, the plaques linger.
2018 年 PNAS 那项研究（Shokri-Kojori 等）：20 个健康成年人，剥夺睡眠约 31 小时一晚，淀粉样蛋白 β 上升大约 5%，最明显的是丘脑和海马——阿尔茨海默早期最先受损的两个区域。为什么？淀粉样蛋白靠深睡时通过血管周隙进入淋巴系统被清走。不让大脑进入深睡，斑块就只能堆着。
Sitting Still Is Now a Disease Driver
久坐已经被当成疾病驱动因子
Sedentary behavior has emerged as one of the largest modifiable Alzheimer’s risk factors in Nicola’s accounting — arguably bigger than alcohol, bigger than diet. Inactivity is not metaphorically harmful; it’s mechanistically harmful at the level of mitochondrial function, vascular health, and myokine release. The treatment isn’t exotic: walk daily. Lift twice a week. Push hard at least once a week. (See Chapter 5.)
在 Nicola 的清单里，久坐已经升到“最大可改风险因子之一”的位置——可以说比酒精更狠，比饮食更狠。“不动”不是个比喻意义上的伤害，是真伤：线粒体功能、血管健康、肌因子释放，全都在这一关。处方一点都不复杂——每天走路，每周两次力量训练，每周至少一次拉到呼吸都不顺。（具体见第五章。）
Quality of Close Relationships Outranks Diet
深度关系的“质”比饮食还重要
The Harvard Study of Adult Development — running 80+ years across two cohorts — found that the quality of close relationships predicted health and cognitive aging more strongly than diet, exercise, or wealth. Not number of friends. Not romantic partner. Quality — someone whose nervous system you can co-regulate with. Maintaining two or three of these across decades is a brain-health intervention as much as it is a life choice.
哈佛“成人发展研究”跨两代人、连续八十多年。结论很反直觉：决定健康和认知衰老的关键变量，不是饮食、运动或财富，而是深度关系的质量。不是朋友数量，也不一定是配偶，而是“能跟你共同调节神经系统”的人。能在几十年里维持两三个这样的关系，本身就是一项脑健康干预——它不只是“人生选择”。
The Black-Market Peptide Warning
黑市肽：那条她真正想画的红线
This is the chapter the episode was marketed around. The warning is real. The mechanism is the part most clipped versions of this conversation skip — and it’s the part that decides whether the warning is fearmongering or sound.
这是节目标题主打的部分。警告是真的——但 60 秒短视频几乎都把机制那段剪掉了，而那一段才决定“她到底是吓唬人还是讲道理”。
What People Are Actually Injecting
他们到底在打什么
The "Wolverine stack": TB-500 + BPC-157. Both injected, both bought from gray-market or black-market suppliers. Plus melanotan (tan / libido). Plus the secretagogues — sermorelin, ipamorelin — for growth-hormone release. None of these has completed a phase-3 randomized human trial. None is FDA-approved. The "gray market" form ships with a content sheet you have to trust. The "black market" form ships with nothing.
圈内俗称“金刚狼组合”：TB-500 + BPC-157。都是注射用，都从灰市/黑市渠道买。再加 melanotan（晒黑/性欲）、再加几种促生长激素分泌的肽——sermorelin、ipamorelin。这些里没有一个跑完过人体三期 RCT，也没有一个拿到 FDA 批准。“灰市”货带一张你只能选择相信的成分单；“黑市”货连那张纸都没有。
BPC-157 Builds Blood Vessels — And Can’t Tell What For
BPC-157 在长血管——但它分不清是在长给谁
BPC-157 stands for "body protection compound 157." It promotes angiogenesis — new blood-vessel formation at injury sites — in part by upregulating VEGFR2, a vascular growth-factor receptor. New vessels deliver oxygen and nutrients; healing accelerates. That’s the upside. The downside: roughly half of human cancers exploit VEGF/VEGFR2 pathways to feed themselves. Tumors signal surrounding tissue to grow vessels toward them and use those vessels to metastasize. If you systemically inject a vascularization agent and you happen to have an undetected stage-1 cancer — pancreatic, ovarian, glioblastoma, the cancers we cannot yet image early — the concern is mechanistically plausible. There is no human RCT proving BPC-157 either causes or accelerates cancer; there is also no human RCT proving it doesn’t. That gap is the problem.
BPC-157 的全称是“身体保护复合物 157”。它促进血管新生——在受伤部位长出新血管——部分机制是上调一个叫 VEGFR2 的血管生长因子受体。新血管送氧、送营养，伤愈合得更快。这是好的一面。坏的一面是：人类大约一半的癌症，都是靠 VEGF/VEGFR2 这条通路给自己“伸血管要饭”。肿瘤会让周围组织朝它长血管，再顺着这些血管转移。如果你身上某个角落正好有一颗一期、还小到现有影像学根本扫不出来的癌细胞——胰腺癌、卵巢癌、胶质母细胞瘤都属于这类——你又给自己系统性注射一个“长血管药”，机制上的担忧就成立了。目前没有任何人体 RCT 证明 BPC-157 致癌或促癌；同样也没有任何人体 RCT 证明它不致癌、不促癌。缺这两组数据，正是问题本身。
There Is No Phase-3 Trial. The Supply Exists Anyway.
没有三期，但市场照样在卖
BPC-157 has never completed a phase-3 human trial. In 2023 the FDA placed it on the section 503A list of substances ineligible for compounding — closing one of the few legal supply paths. That is why no doctor can write you a prescription for it. The supply chain that exists today exists because the formal pathway closed, not because data is pending.
BPC-157 从来没跑完过人体三期。2023 年，FDA 把它列入 503A 条款“不可配制”名单——堵掉了为数不多几条合法供应路径之一。这就是为什么医生没法给你开。今天市场上还能买到，恰恰是因为正规通道关了——不是因为“数据正在审”。
GLP-1s Are Peptides Too — And Nicola Loves Them
GLP-1 也是肽——但 Nicola 极力推荐
Semaglutide. Tirzepatide. Wegovy. Zepbound. Mounjaro. These are peptides. They have completed phase-3 trials. They are FDA-approved. They are not compounded gray-market product. Nicola’s position on them is unequivocally enthusiastic — she compares their importance to penicillin and insulin and predicts that in a hundred years textbooks will describe obesity as a disease people used to have. The blanket statement "peptides are dangerous" is wrong. The accurate statement is: gray-market peptides without phase-3 data are dangerous.
司美格鲁肽、替尔泊肽、Wegovy、Zepbound、Mounjaro——这些都是肽，但它们跑完了人体三期，拿到了 FDA 批准，不是“灰市配制”的东西。Nicola 对它们的态度是毫不含糊的“支持”——她把这一类药的意义类比成青霉素和胰岛素，并且预测一百年后教科书会说“以前人们还得过一种叫肥胖的病”。“肽就是危险”这种泛泛的说法是错的。准确的说法是：没有三期数据的灰市肽才是危险的。
The Carve-Out Doesn’t Survive a 60-Second Cut
60 秒里塞不下“例外条款”
After a sixty-second cut of Nicola’s warning circulated, Andrew Huberman messaged her to push back. The clip had compressed a phase-2 caution about melanotan into a blanket BPC-157 indictment. She corrected the record. The lesson she draws is bigger than the exchange: a clip of a three-hour conversation will reliably mistake the carve-out for the rule. Read the long version before you inject anything.
一段 60 秒的剪辑流出去之后，Andrew Huberman 发消息“纠正”她。问题是那段剪辑把她原话里关于 melanotan 二期的提醒，压缩成了“BPC-157 = 致死”。她后来公开澄清了。但她真正想强调的事比这个争论更大：一段三小时谈话的 60 秒剪辑，几乎必然会把“例外”当成“规则”。注射任何东西之前，先把长版看完。
What Actually Works
真正能用的工具
Now the part with the data. Each item below has either RCTs in humans or strong mechanism plus epidemiology. None of it is exotic. All of it is boring on the page and hard in the body.
下面是有数据的部分。每一条要么有人体 RCT，要么机制清晰加流行病学背书，没一条玄学。在纸面上无聊——在身体里都不轻松。
Lactate Is Brain Fuel, Not Waste
乳酸是脑燃料，不是“代谢垃圾”
Push your heart rate to ~90% of max and the mitochondria can’t keep up — glycolysis happens outside them, producing lactate. Lactate is not waste. It crosses into the brain and acts as a fuel source, supports struggling neurons, and aids amyloid clearance. The same hard cardio also shunts circulating tumor cells out of the bloodstream — a 2022 paper in Cell showed exercise inhibiting prostate cancer progression in men. Once or twice a week is enough.
把心率推到约 90% 最大心率，线粒体就“供不上”了——糖酵解被甩到线粒体外面进行，副产物是乳酸。乳酸不是垃圾。它穿过血脑屏障进入大脑，作为燃料给挣扎的神经元供能，还参与清除淀粉样蛋白。同一组高强度训练，还能“冲走”血液中的循环肿瘤细胞——2022 年《Cell》的一篇论文里，运动抑制了男性的前列腺癌进展。每周做一到两次就够。
Myokines Are What Pharma Can’t Replicate
肌因子——药厂搞了二十年还没搞出来
Contract a skeletal muscle hard enough and it releases small protein messengers called myokines — about a hundred of them. They travel via blood and dock at receptors across the brain, liver, pancreas, heart. Functions: synaptogenesis (new connections), neurogenesis in the hippocampus, immune support, and amyloid β clearance. Pharma has been trying to replicate myokines as injectable drugs since the early 2000s and has failed. The only reliable way to release them is to lift something heavy.
把骨骼肌收缩到一定强度，它就会释放一种叫肌因子（myokines）的小蛋白——大约一百种。它们通过血液到达大脑、肝、胰腺、心脏的受体，干的是这些事：突触生成（新连接）、海马的神经发生、免疫调节，以及清除淀粉样蛋白 β。药厂从 2000 年代初就想把这堆东西做成可注射药——到现在没成。让它们出来的唯一可靠办法，就是抬起一些重东西。
~32% Lower Alzheimer’s Risk — with a Caveat
阿尔茨海默风险降约 32%——但有个“但是”
Observational meta-analyses report roughly a 32% reduction in Alzheimer’s-specific risk and ~20% reduction in dementia broadly among statin users. Mechanism: cholesterol plaque builds in brain arteries the way it builds in heart arteries. ApoE4 carriers especially benefit, because ApoE4 disrupts cholesterol transport in the brain. Caveat: the major RCTs to date (PROSPER, HPS) showed no significant effect on cognition, so the observational signal hasn’t been confirmed in trials yet. STAREE-Mind is currently running. This is — alongside vaccines — the single most contested recommendation Nicola makes publicly. Decide with your doctor; the lipid story matters either way.
观察性 meta 分析里，他汀使用者的阿尔茨海默风险大约降低 32%，整体痴呆降约 20%。机制：胆固醇斑块堆积在脑动脉里，跟堵心血管是一回事。ApoE4 携带者尤其受益——因为 ApoE4 本身就破坏脑内胆固醇转运。但是：迄今最重要的两项 RCT（PROSPER 和 HPS）在认知终点上没跑出显著差异，所以观察研究的信号还没被干预试验确认。新的 STAREE-Mind 试验正在进行。Nicola 自己说，这是她公开推荐里最容易“被骂”的一项——和疫苗并列。和医生商量再决定；不管你最后吃不吃，“血脂”这条线都重要。
5g Saturates Muscle. 10–12g Reaches Brain.
5g 喂肌肉，10–12g 才进得了大脑
The standard 5-gram dose saturates muscle. Muscle gets first dibs and uses the whole 5g. To get creatine across the blood-brain barrier you need 10 to 12 grams daily. Once in, it boosts mitochondrial ATP production, addresses the energy crisis seen in neurodegenerative disease, and supports mood. The 2025 CABA pilot trial (Smith et al.) put 20 confirmed Alzheimer’s patients on 20g/day for 8 weeks: brain creatine rose 11%, global and fluid cognition scores improved, and some patients regained the energy to return to the gym. Single-arm and not placebo-controlled — promising signal, not yet proof.
通常的 5 克剂量只够喂饱肌肉——肌肉“最先伸手抓”，5 克全被吃掉。要让肌酸越过血脑屏障，每天要吃到 10 到 12 克。一旦进入大脑，它能拉高线粒体的 ATP 产能，缓解神经退行性疾病里普遍存在的“能量危机”，对情绪也有帮助。2025 年的 CABA 临床试点（Smith 等）让 20 位确诊阿尔茨海默患者按 每天 20 克、连续 8 周 服用：脑内肌酸上升 11%，整体和流体认知分数都改善，部分患者甚至恢复了去健身房的体力。注意：这是单臂、非安慰剂对照试验——信号好看，不等于已被证明。
2g DHA + 2g EPA — You Are Feeding the Substrate
DHA 2g + EPA 2g——你在喂大脑本身
About 70% of brain mass is fat, and most of that fat is DHA. The brain has a dedicated transporter — MFSD2A — that shuttles DHA in. In Alzheimer’s patients that transporter becomes dysfunctional. Daily 2g DHA + 2g EPA from a third-party-tested, low-oxidation source supports the supply line. Trials suggest a modest signal for plaque support. Buy from a brand that publishes oxidation values — rancid omega-3 is worse than no omega-3.
大脑大约 70% 的质量是脂肪，其中绝大多数是 DHA。大脑有一个专门转运 DHA 的“门”——MFSD2A——阿尔茨海默患者身上这扇门会失灵。每天 DHA 2g + EPA 2g，从一个第三方检测、低氧化值的品牌买，能稳住这条供应线。试验里它对斑块也有一点支持作用。一定要买公开氧化值的牌子——氧化变质的鱼油比不吃还糟。
Light + Sound at 40Hz: Mouse Promise, Human Phase 3
40 赫兹光+声：鼠子可观，人体三期进行中
MIT’s Tsai lab has shown across mouse and early human studies that 40Hz light + sound stimulation (the GENUS protocol — "gamma entrainment using sensory stimuli") substantially reduces amyloid plaques in multiple brain regions, prevents neuron death, and preserves memory in Alzheimer’s mouse models. The mechanism is entrainment: gamma brain waves (the 30–100Hz band associated with deep focus and the first oscillations to deteriorate in dementia) can mirror external 40Hz signals. Cognito Therapeutics, the MIT spinoff, is now running phase-3 human trials. Promising frontier — not yet prescriptive.
麻省理工 Tsai 实验室一系列鼠子和早期人体研究显示：40 赫兹的光加声音刺激（GENUS 协议——“用感官刺激诱导伽马同步”）能在阿尔茨海默小鼠模型里大幅减少多个脑区的淀粉样斑块、保住神经元、保住记忆。机制叫夹带（entrainment）：大脑里 30–100 赫兹的伽马波（深度专注相关、也是痴呆里最先瓦解的那段频率），可以“跟着”外部 40 赫兹的信号同步起来。MIT 衍生公司 Cognito Therapeutics 正在跑人体三期。前景看起来好——但还没到“就这么用”的阶段。
The Daily Protocol
每日方案
What Nicola actually does, and what she tells private clients to do. Specific. Numbered. Doable starting tomorrow.
下面是 Nicola 自己在做的——也是她对私人咨询客户开的处方。剂量明确，明早就能开始。
Movement, Sleep, Supplements, Vaccines
动 · 睡 · 补 · 打疫苗
Movement. 2× per week resistance training, hard enough to release myokines. Aerobic base — long zone-2 sessions for mitochondrial density. 1–2× per week zone-5 cardio at ~90% of max HR for the lactate boost. Daily walking; inactivity sits among the largest modifiable Alzheimer’s risks.
Sleep. Floor: 7.5 hours. One night at <6h raises amyloid β by ~5% in healthy adults (PNAS 2018). Train for sleep: dim lights at 8:30 pm, switch to red light if practical, no phones / email / charged conversations in the wind-down hour.
Supplements. Creatine 10–12g/day (brain dose, not muscle dose). Omega-3 2g DHA + 2g EPA from a third-party-tested low-oxidation brand. Discuss statins with your doctor — particularly if ApoE4-positive. Buy from companies that publish manufacturing standards, third-party testing, and contaminant scores.
Vaccines. Shingles vaccine after 50 (recent observational data ties it to reduced dementia risk). The standard adult schedule.
动。 每周 2 次力量训练，强度够到肌因子真的释放出来。有氧基础——长时间 2 区做线粒体密度。每周 1–2 次 5 区有氧，心率到约 90% 最大值，靠的是乳酸那条通路。每天走路；久坐已经是最大的可改风险之一。
睡。 底线 7.5 小时。一晚不到 6 小时，健康成人淀粉样蛋白 β 涨约 5%（PNAS 2018）。“为睡觉做训练”：晚 8:30 调暗灯光，能用红光更好，最后一小时不碰手机、不查邮件、不开重要谈话。
补。 肌酸 10–12 克/天（大脑剂量，不是肌肉剂量）。Omega-3：DHA 2g + EPA 2g，挑公开氧化值、第三方检测的牌子。和医生商量他汀——尤其如果你是 ApoE4 阳性。补剂只买公布生产标准、第三方检测、污染物分数的公司。
疫苗。 50 岁以后打带状疱疹疫苗（近期观察性数据显示它与痴呆风险下降相关）。其余按成人疫苗常规表来。
Things Nicola Says to Cut
她说的“别再做了”
Alcohol. No safe amount. Even moderate drinking shows up on imaging.
Marijuana for sleep. It sedates, it doesn’t sleep you. You lose deep sleep and REM.
Black-market peptides. BPC-157, TB-500, melanotan, sermorelin, ipamorelin. Wait for phase-3 data or skip.
Doomscroll content. The "brain rot" loop. Outsource trigonometry to ChatGPT often enough and the math skills atrophy. Use cognitive engagement — not its substitute.
酒。 没有安全剂量。中等量饮酒在脑影像上就能看出问题。
用大麻助眠。 那是镇静，不是睡眠。深睡和 REM 都被切掉。
黑市肽。 BPC-157、TB-500、melanotan、sermorelin、ipamorelin。要么等三期数据，要么不碰。
无限刷的“脑腐”内容。 三角函数全交给 ChatGPT 算，数学能力就真的退化。要的是“在用脑”，不是它的替代品。
Numbers Worth Knowing
值得查的几个数字
Blood pressure (target ~120/80) — continuous cuffs are now cheap. LDL and ApoB. ApoE genotype, if you want to know. Colonoscopy and standard cancer screens on schedule — exercise clears circulating tumor cells, but you still need imaging for what’s already lodged. Annual physical, fasting glucose / HbA1c, vitamin D.
血压（目标 ~120/80）——连续血压袖带已经很便宜了。LDL 和 ApoB。ApoE 基因分型——如果你想知道。肠镜和常规癌症筛查按时做——运动只能清掉“在血里跑”的肿瘤细胞，已经定居的还得靠影像查。年度体检、空腹血糖 / 糖化血红蛋白、维生素 D。
The reframe. Most of what determines whether you get Alzheimer’s is decided in your thirties and forties. The disease shows up in your seventies because the damage compounded for forty years before it became visible on a brain scan. The 95% number is the good news and the bad news at once. The agency is yours. So is the bill.
换个角度。 你会不会得阿尔茨海默，大部分在你三四十岁那十几年就已经定下来了。它在七十岁出现，是因为伤害在你看不到的地方累积了四十年。那个 95% 既是好消息，也是坏消息——主动权在你手上，账单也在你手上。