Place · Level 3 · Supplement
α-GPC + CDP-Choline (Citicoline)
胆碱衍生形态 · 乙酰胆碱前体 · 卒中后认知部分证据 · 健康人证据弱 · ⚠️ α-GPC 卒中风险信号 (Lee 2021 韩国全国 1200 万人, 总卒中 aHR 1.43; 有适应证混杂)
Last updated
Story path
Chapter 1
Two choline derivatives
Two choline derivatives
α-GPC (alpha-glycerylphosphorylcholine) plus CDP-choline (citicoline, cytidine 5′-diphosphocholine) are two choline-derived nootropic supplements that complement the atlas choline story.
Recap (from the atlas choline story):
Choline: essential nutrient (AI men 550 / women 425 mg/day)Functions: phosphatidylcholine synthesis + acetylcholine (ACh) neurotransmitter + one-carbon metabolism (BHMT pathway) + hepatic lipid transport
α-GPC:
Chemistry: glycerylphosphocholine, naturally present in breast milk and brainHigh brain bioavailability (crosses the blood–brain barrier: The 'security gate' on brain vessels that blocks most substances in blood from entering the brain. well)Content: 40% choline + glycerophosphate + partial direct acetylcholine precursorA prescription drug in some European countries for dementia / post-stroke cognitionSold OTC as a nootropic in the US / China
CDP-choline (Citicoline):
Chemistry: cytidine diphosphate choline, an intermediate in adenosine triphosphate: The cell's universal energy currency — almost everything that costs energy spends it. / phospholipid synthesisGood brain bioavailabilityContent: 18% choline + 32% cytidine + phosphatePrescription drug in Japan / Europe: Somazina, etc.OTC: 'Cognizin' (a commercial high-purity form used in most cognitive RCTs)
Differences from plain choline (Choline bitartrate / Phosphatidylcholine):
Plain choline: whole-body supplementation (liver / heart / brain all use it), best cost-effectivenessα-GPC / CDP: more brain-targeted, but expensivePhosphatidylcholine (PC, lecithin): also good bioavailability, but more of a membrane component than a nootropic pathwayTMAO risk: high-dose choline (any form) is processed by gut bacteria into TMA, oxidized in the liver to TMAO; some studies link this to cardiovascular risk (covered in the atlas choline + TMG story), with large individual variation
Why atlas covers this section:
The nootropic market is growing fast, and α-GPC + CDP are mainstream single productsClinical evidence is mixed: B-tier for post-stroke cognition and dementia; weak for healthy-person nootropic effectsHeavy marketing around 'memory / focus / brain protection' — atlas goes through the evidence claim by claimNew warning: Lee 2021 (a 12-million-person South Korean retrospective cohort) links α-GPC use to higher stroke risk, with a dose-response by duration; but the people prescribed it were already declining cognitively, so confounding cannot be ruled out
Recap (from the atlas choline story):
Choline: essential nutrient (AI men 550 / women 425 mg/day)Functions: phosphatidylcholine synthesis + acetylcholine (ACh) neurotransmitter + one-carbon metabolism (BHMT pathway) + hepatic lipid transport
α-GPC:
Chemistry: glycerylphosphocholine, naturally present in breast milk and brainHigh brain bioavailability (crosses the blood–brain barrier: The 'security gate' on brain vessels that blocks most substances in blood from entering the brain. well)Content: 40% choline + glycerophosphate + partial direct acetylcholine precursorA prescription drug in some European countries for dementia / post-stroke cognitionSold OTC as a nootropic in the US / China
CDP-choline (Citicoline):
Chemistry: cytidine diphosphate choline, an intermediate in adenosine triphosphate: The cell's universal energy currency — almost everything that costs energy spends it. / phospholipid synthesisGood brain bioavailabilityContent: 18% choline + 32% cytidine + phosphatePrescription drug in Japan / Europe: Somazina, etc.OTC: 'Cognizin' (a commercial high-purity form used in most cognitive RCTs)
Differences from plain choline (Choline bitartrate / Phosphatidylcholine):
Plain choline: whole-body supplementation (liver / heart / brain all use it), best cost-effectivenessα-GPC / CDP: more brain-targeted, but expensivePhosphatidylcholine (PC, lecithin): also good bioavailability, but more of a membrane component than a nootropic pathwayTMAO risk: high-dose choline (any form) is processed by gut bacteria into TMA, oxidized in the liver to TMAO; some studies link this to cardiovascular risk (covered in the atlas choline + TMG story), with large individual variation
Why atlas covers this section:
The nootropic market is growing fast, and α-GPC + CDP are mainstream single productsClinical evidence is mixed: B-tier for post-stroke cognition and dementia; weak for healthy-person nootropic effectsHeavy marketing around 'memory / focus / brain protection' — atlas goes through the evidence claim by claimNew warning: Lee 2021 (a 12-million-person South Korean retrospective cohort) links α-GPC use to higher stroke risk, with a dose-response by duration; but the people prescribed it were already declining cognitively, so confounding cannot be ruled out
机制 · 先回到普通胆碱
要看懂它们强在哪, 先回到普通胆碱。胆碱是身体离不开的一种必需营养素 (适宜摄入量男 550、女 425 mg/天), 它在体内身兼数职: 拼成细胞膜的磷脂、造出一种叫乙酰胆碱 (ACh) 的神经信号分子、参与甲基搬运的一条代谢旁路, 还帮肝脏把脂肪运出去。补脑说的主要就是乙酰胆碱那份活——它是大脑里管记忆、也指挥肌肉发力的信使。两种形态 · 各有各的打法
α-GPC (甘油磷酸胆碱) 和 CDP-胆碱在补脑这件事上各有打法。α-GPC 本来就存在于母乳和大脑里, 分子里 40% 是胆碱, 还有一部分能直接当乙酰胆碱的原料; 它穿过血脑屏障 (脑和血液之间那道筛子) 很顺, 所以进脑快。在欧洲一些国家它是处方药, 用于痴呆和中风后的认知恢复, 在美国、中国则当补脑品在货架上卖。CDP-胆碱 (又叫 citicoline) 是身体合成磷脂时的一个中间产物, 只含 18% 胆碱、另配 32% 胞苷; 它进脑也不错, 在日本和欧洲是处方药, 保健品版本多用一个叫 Cognizin 的高纯度型号, 大多数认知试验用的就是它。性价比 · 和便宜形态比
那和便宜的普通胆碱 (bitartrate、卵磷脂里的磷脂酰胆碱) 比, 值不值得多花这笔钱? 普通胆碱是全身通用的补法, 肝、心、脑都用得上, 最划算; α-GPC 和 CDP 是把胆碱更精准地送进脑, 代价是贵。有一点三种形态都躲不开: 胆碱吃多了, 一部分会被肠道细菌转成 TMA、再在肝里氧化成一个叫 TMAO 的代谢物, 它在一些研究里和心血管风险挂钩, 而且人和人差别很大 (atlas choline 和 tmg 故事细讲)。安全 · 一个撇不清的疑点
真正让这一岛值得停下来的, 是一个新冒出来的安全信号: Lee 2021 追踪了韩国 1200 万名 50 岁以上的人, 发现用过 α-GPC 的人日后中风更多 (总卒中 aHR 1.43), 而且用得越久信号越强。但先别急着下结论——在韩国 α-GPC 本就是开给认知已经下滑的人的, 而认知下滑本身就是脑血管出问题的前兆, 这笔账观察数据分不清到底是药还是病。Chapter 2
Clinical evidence by indication
Clinical evidence by indication
Evidence tiering by indication:
1. Acute ischemic stroke (citicoline)
ICTUS Trial (Dávalos 2012 Lancet, N = 2,298): citicoline 2000 mg/day × 6 weeks did not show functional improvement vs placeboEarly small trials were positive; this large RCT was negativeSubgroup analyses showed signal in some mild patients / elderly women2024 clinical positioning: not recommended for acute stroke, but some hospitals still use it for chronic recovery
2. Vascular cognitive impairment / dementia
IDEALE 2013 + COBRA 2015 and other RCTs: moderately positiveCiticoline has B-tier evidence in vascular cognitive impairmentCiticoline + antiplatelet drug combination: some studies show synergy
3. Alzheimer's disease (AD)
α-GPC + donepezil: some positive studies (Amenta 2014)Solo evidence is weakDoesn't replace standard AChE inhibitors
4. Traumatic brain injury (TBI)
COBRIT Trial (Zafonte 2012 JAMA, N = 1,213): citicoline × 90 days was negative (no functional improvement)Early small trials were positive; large RCT was negative
5. Healthy-person nootropic (attention / memory / anxiety)
Most are small trials (N < 100) with mixed resultsConant 2024 review: some trials show moderate effect, but heterogeneity is large and funding mostly comes from the supplement industryB-C tier evidence
6. Exercise performance (α-GPC)
Bellar 2015 + Marcus 2017: a single 600 mg α-GPC dose briefly raises strength training power by 5–15%Mechanism: raises ACh, improves neuromuscular conductionB-tier evidence (exercise performance)
Important warning: Lee 2021 α-GPC and stroke signal
South Korea NHIS national insurance data, 12 million adults aged 50+ (108,877 α-GPC users); use ascertained 2006-2008, outcomes followed 2009-2018Total stroke aHR 1.43 (95% CI 1.41-1.46); ischaemic 1.34; haemorrhagic 1.37Dose-response by duration (vs <2 months): 2-6 months 1.13; 6-12 months 1.18; >12 months 1.36Hypothesized mechanism: TMAO / platelets — but that pathway was never measured in this studyKey limitation: in South Korea α-GPC is prescribed for cognitive impairment, and cognitive decline is itself a marker of cerebrovascular disease. The authors themselves note users were older with more comorbidities and may already have had subclinical atherosclerosis. That is confounding by indication; observational data cannot separate drug from diseaseSo: this is not proof that α-GPC causes stroke. It is that the benefit for healthy users was always weak, which makes an unresolved signal enough to settle the decision
Connections to other atlas pathways:
TMAO risk (covered in choline / TMG): α-GPC and CDP also route through gut bacteria into TMA → TMAOVLDL / hepatic lipids: plain choline works mainly in the liver; α-GPC / CDP skew toward the brainMethylation (one-carbon metabolism): choline → TMG → BHMT alternate path (atlas TMG L4)
1. Acute ischemic stroke (citicoline)
ICTUS Trial (Dávalos 2012 Lancet, N = 2,298): citicoline 2000 mg/day × 6 weeks did not show functional improvement vs placeboEarly small trials were positive; this large RCT was negativeSubgroup analyses showed signal in some mild patients / elderly women2024 clinical positioning: not recommended for acute stroke, but some hospitals still use it for chronic recovery
2. Vascular cognitive impairment / dementia
IDEALE 2013 + COBRA 2015 and other RCTs: moderately positiveCiticoline has B-tier evidence in vascular cognitive impairmentCiticoline + antiplatelet drug combination: some studies show synergy
3. Alzheimer's disease (AD)
α-GPC + donepezil: some positive studies (Amenta 2014)Solo evidence is weakDoesn't replace standard AChE inhibitors
4. Traumatic brain injury (TBI)
COBRIT Trial (Zafonte 2012 JAMA, N = 1,213): citicoline × 90 days was negative (no functional improvement)Early small trials were positive; large RCT was negative
5. Healthy-person nootropic (attention / memory / anxiety)
Most are small trials (N < 100) with mixed resultsConant 2024 review: some trials show moderate effect, but heterogeneity is large and funding mostly comes from the supplement industryB-C tier evidence
6. Exercise performance (α-GPC)
Bellar 2015 + Marcus 2017: a single 600 mg α-GPC dose briefly raises strength training power by 5–15%Mechanism: raises ACh, improves neuromuscular conductionB-tier evidence (exercise performance)
Important warning: Lee 2021 α-GPC and stroke signal
South Korea NHIS national insurance data, 12 million adults aged 50+ (108,877 α-GPC users); use ascertained 2006-2008, outcomes followed 2009-2018Total stroke aHR 1.43 (95% CI 1.41-1.46); ischaemic 1.34; haemorrhagic 1.37Dose-response by duration (vs <2 months): 2-6 months 1.13; 6-12 months 1.18; >12 months 1.36Hypothesized mechanism: TMAO / platelets — but that pathway was never measured in this studyKey limitation: in South Korea α-GPC is prescribed for cognitive impairment, and cognitive decline is itself a marker of cerebrovascular disease. The authors themselves note users were older with more comorbidities and may already have had subclinical atherosclerosis. That is confounding by indication; observational data cannot separate drug from diseaseSo: this is not proof that α-GPC causes stroke. It is that the benefit for healthy users was always weak, which makes an unresolved signal enough to settle the decision
Connections to other atlas pathways:
TMAO risk (covered in choline / TMG): α-GPC and CDP also route through gut bacteria into TMA → TMAOVLDL / hepatic lipids: plain choline works mainly in the liver; α-GPC / CDP skew toward the brainMethylation (one-carbon metabolism): choline → TMG → BHMT alternate path (atlas TMG L4)
临床 · 急性中风与脑外伤
先说中风。急性缺血性中风刚发作时用 CDP-胆碱救急, 曾被寄予厚望, 但迄今最大的一个试验 (ICTUS, N = 2,298, 每天 2000 mg 连吃 6 周) 没能让病人恢复得更好, 只在个别轻症、老年女性的亚组里看到点影子; 现在的定位是急性期不推荐, 少数医院还拿它用于慢性康复。脑外伤是同一个剧本——最大的那个试验 (COBRIT, N = 1,213, 吃 90 天) 同样没测出功能改善。临床 · 痴呆与血管性认知障碍
痴呆和血管性认知障碍这块稍微亮一点。CDP-胆碱在血管性认知障碍上攒到了 B 级证据 (中等强度), 几个试验偏正面, 和抗血小板药合用时有人看到协同; 在阿尔茨海默病上, α-GPC 配着标准的胆碱酯酶抑制剂 (多奈哌齐一类) 一起用有些正面结果, 但它单独顶不了标准药, 只能算辅助。证据 · 健康人补脑与运动表现
把它当补脑品给健康人吃, 才是营销主战场, 偏偏证据最薄。这里的试验大多不到 100 人、结果两边倒, 综述也点明异质性大、经费多来自补剂行业, 只够 B-C 级。相对硬一点的反而是运动那一块: 训练前单次吃 600 mg α-GPC, 短时间内力量输出能涨 5-15%, 机制是把乙酰胆碱 (ACh, 指挥肌肉发力的信使) 顶上去、让神经到肌肉的传导更利索, 这算 B 级证据。警告 · 韩国全国队列的数字
绕不开的还是那条中风警告。Lee 2021 扒了韩国全国医保里 1200 万名 50 岁以上的人 (其中 108,877 个 α-GPC 使用者), 前几年看谁在吃、后面十年看谁中风: 用过的人总卒中风险高出约四成 (aHR 1.43, 95% CI 1.41-1.46), 缺血性 1.34、出血性 1.37, 而且吃得越久越明显 (2-6 个月 1.13、6-12 个月 1.18、超过 12 个月 1.36)。怎么读 · 适应证混杂这个坑
这个信号该怎么读? 有人猜是 TMAO (肠道细菌把胆碱转出来的那个代谢物) 或血小板在作怪, 可这条通路这个研究根本没测。更要紧的是一个绕不过的坑: 在韩国 α-GPC 是开给认知障碍病人的, 而认知下滑本身就是脑血管在报警——原文作者自己都写了, 用药的人更年长、合并症更多, 可能早就有看不见的动脉硬化。这叫适应证混杂: 观察数据永远分不清, 到底是药害了人, 还是同一个病既让人吃药、又让人中风。所以真正的结论不是 α-GPC 一定害你中风, 而是——健康人本来就没多少好处可拿, 犯不着为一个撇不清的信号去赌。接回 atlas 的其它通路
顺带把它接回 atlas 别处的通路: 高剂量的胆碱不管哪种形态, 都会经肠菌走 TMA 再变 TMAO (choline / tmg 故事讲过); 论护肝、运脂主要靠普通胆碱, α-GPC 和 CDP 更偏向脑; 而甲基化那条旁路, 是胆碱经 TMG 交给 BHMT 这个酶完成的 (见 atlas tmg L4)。Chapter 3
Form comparison
Form comparison
Comparing six choline forms:
Purity and quality:
Cognizin (CDP-choline commercial brand, Kyowa Hakko, Japan): used in most RCTs, purity assuredAlphaSize (α-GPC commercial brand, Chemi Nutra, US): soy-derived, used in some RCTsThird-party certification (USP / NSF / Informed-Choice) is the quality floor
Dosing:
CDP-choline: 250–500 mg × 2/day (nootropic); 2000 mg/day (stroke)α-GPC: 300–600 mg/day (nootropic); 1200 mg/day (dementia, divided doses)Exercise performance: α-GPC 600 mg 30–60 minutes pre-trainingFood (eggs + organ meats + fish): plain choline is sufficient; no need for α-GPC / CDP
Time to effect:
Exercise performance (α-GPC): single dose, effect within 30–60 minutesCognitive nootropic: subjective feeling at 1–4 weeks, objective testing at 4–12 weeksPost-stroke cognition: assess at 3–6 months
Stacking with other nootropics:
Caffeine + L-theanine: synergy (covered in the atlas caffeine-l-theanine story)Bacopa monnieri: Ayurvedic, B-tier memory evidenceLion's mane: neurotrophic factor hypothesis, debunked in Batch VIModafinil / Adderall: prescription drugs, not supplements; stacking with supplements is high-risk
Safety:
Common side effects: nausea / headache / insomnia / tachycardia (at high dose)TMAO association: long-term high doses are worth watchingWorsened depression: α-GPC raises anxiety / irritability in some peoplePregnancy / lactation: data lacking (food choline is safe; skip supplements)Children: not recommended (except for clear medical indication)
Contraindications / warnings:
Bipolar depression: α-GPC / CDP may trigger or worsen mania (caution)On SSRI / anticholinergic drugs: interaction riskHigh cardiovascular risk: watch the Lee 2021 stroke signalStop 1 week before surgery: bleeding risk
'Focus miracle drug' marketing compared point by point:
'1 week of taking it gives a huge brain boost' marketing is mostly placebo effectReal effect is mild with large individual variationThe real levers for focus are sleep, exercise, nutrition, less screen time, less stress — no supplement can compete
| Form | Choline % | BBB | Price | Main use |
|---|---|---|---|---|
| Choline bitartrate | 41% | Medium | $ | Whole-body supplementation when food is insufficient |
| Phosphatidylcholine (PC, lecithin) | 13% | Medium | $$ | Membrane health / liver / gut |
| CDP-choline (citicoline) | 18% | High | $$$ | Post-stroke cognition / nootropic |
| α-GPC | 40% | High | $$$ | Dementia adjunct / exercise performance / nootropic |
| N-acetylcysteine (NAC) | N/A | High | $ | Not a choline form, but often stacked with choline |
| DMAE | N/A | High | $$ | Old nootropic, weak evidence |
Purity and quality:
Cognizin (CDP-choline commercial brand, Kyowa Hakko, Japan): used in most RCTs, purity assuredAlphaSize (α-GPC commercial brand, Chemi Nutra, US): soy-derived, used in some RCTsThird-party certification (USP / NSF / Informed-Choice) is the quality floor
Dosing:
CDP-choline: 250–500 mg × 2/day (nootropic); 2000 mg/day (stroke)α-GPC: 300–600 mg/day (nootropic); 1200 mg/day (dementia, divided doses)Exercise performance: α-GPC 600 mg 30–60 minutes pre-trainingFood (eggs + organ meats + fish): plain choline is sufficient; no need for α-GPC / CDP
Time to effect:
Exercise performance (α-GPC): single dose, effect within 30–60 minutesCognitive nootropic: subjective feeling at 1–4 weeks, objective testing at 4–12 weeksPost-stroke cognition: assess at 3–6 months
Stacking with other nootropics:
Caffeine + L-theanine: synergy (covered in the atlas caffeine-l-theanine story)Bacopa monnieri: Ayurvedic, B-tier memory evidenceLion's mane: neurotrophic factor hypothesis, debunked in Batch VIModafinil / Adderall: prescription drugs, not supplements; stacking with supplements is high-risk
Safety:
Common side effects: nausea / headache / insomnia / tachycardia (at high dose)TMAO association: long-term high doses are worth watchingWorsened depression: α-GPC raises anxiety / irritability in some peoplePregnancy / lactation: data lacking (food choline is safe; skip supplements)Children: not recommended (except for clear medical indication)
Contraindications / warnings:
Bipolar depression: α-GPC / CDP may trigger or worsen mania (caution)On SSRI / anticholinergic drugs: interaction riskHigh cardiovascular risk: watch the Lee 2021 stroke signalStop 1 week before surgery: bleeding risk
'Focus miracle drug' marketing compared point by point:
'1 week of taking it gives a huge brain boost' marketing is mostly placebo effectReal effect is mild with large individual variationThe real levers for focus are sleep, exercise, nutrition, less screen time, less stress — no supplement can compete
数字 · 六种形态逐项对比
6 种胆碱形态对比:| 形态 | 胆碱 % | BBB | 价格 | 主用途 |
|---|---|---|---|---|
| 胆碱 (Choline bitartrate) | 41% | 中 | $ | 全身补充, 食物补不够时 |
| 磷脂酰胆碱 (PC, Lecithin) | 13% | 中 | $$ | 膜健康、肝 / 肠道 |
| CDP-choline (Citicoline) | 18% | 高 | $$$ | 卒中后认知 / nootropic |
| α-GPC | 40% | 高 | $$$ | 痴呆辅助、运动表现 / nootropic |
| N-乙酰半胱氨酸 (NAC) | N/A | 高 | $ | 不是胆碱形态, 但常和胆碱搭 |
| DMAE | N/A | 高 | $$ | 老 nootropic, 证据弱 |
质量 · 商业型号与第三方认证
纯度 + 质量:Cognizin (CDP-choline 商业品牌, Kyowa Hakko 日本): 大多 RCT 用这个, 纯度有保证AlphaSize (α-GPC 商业品牌, Chemi Nutra 美国): 大豆来源, 部分 RCT 用第三方认证 (USP / NSF / Informed-Choice) 是质量底线
剂量 · 各用途怎么吃、多久起效
剂量:CDP-choline: 250-500 mg × 2次/天 (nootropic); 2000 mg/天 (卒中)α-GPC: 300-600 mg/天 (nootropic); 1200 mg/天 (痴呆, 分次)运动表现: α-GPC 600 mg 训练前 30-60 分钟食物 (鸡蛋 + 内脏 + 鱼): 普通胆碱足够, 不需要 α-GPC / CDP
起效时间:
运动表现 (α-GPC): 单次, 30-60 分钟起效认知 nootropic: 主观感觉 1-4 周, 客观测试 4-12 周卒中后认知: 3-6 个月评估
搭配 · 和别的 nootropic 叠加
与其它 nootropic 叠加:Caffeine + L-theanine: 协同 (atlas caffeine-l-theanine 故事)Bacopa monnieri: 阿育吠陀, B 级记忆证据Lion's Mane: 神经营养因子假说, Batch VI 拆穿Modafinil / Adderall: 处方药, 不是补剂; 与补剂叠加风险大
安全 · 副作用与禁忌
安全性:常见副作用: 恶心、头痛、失眠、心动过速 (高剂量)TMAO 关联: 长期高剂量值得关注抑郁恶化: 部分人 α-GPC 增加焦虑、烦躁怀孕、哺乳: 缺数据 (食物胆碱安全, 补剂跳过)儿童: 不推荐 (除明确医学指征)
禁忌、警告:
抑郁双相: α-GPC / CDP 可能诱发、加重躁狂 (慎)服 SSRI / 抗胆碱药: 互动风险CV 高危: 警惕 Lee 2021 卒中信号手术前停 1 周: 凝血风险
误区 · 专注神药营销逐条比对
专注神药营销逐条比对:喝完 1 周脑力大提营销, 主要是安慰剂效应真效应温和, 个体差异大真正提专注的杠杆是睡眠、运动、营养、减屏幕、减压力, 任何补剂都比不过
Chapter 4
Decision tree
Decision tree
α-GPC / CDP-choline decision path:
Q1: what's your goal?
A. Acute stroke (emergency phase):
Not recommended (ICTUS / COBRIT large RCTs negative)Standard stroke treatment (thrombolysis / thrombectomy + antiplatelet + secondary prevention) takes priority
B. Chronic post-stroke recovery / vascular cognitive impairment:
CDP-choline 500–1000 mg/day can be discussed with a neurologistB-tier evidence, assess at 6 months
C. Alzheimer adjunct (alongside an AChE inhibitor):
α-GPC 600–1200 mg/day can be discussed with a neurologistDoesn't replace donepezil / memantine
D. Healthy-person nootropic (focus / memory):
Minimal benefitTry 4–8 weeks looking at subjective + objective improvement; stop if no changeCDP-choline 250 mg × 2/day or α-GPC 300 mg/dayCost-effectiveness: compared with caffeine + L-theanine + sleep + exercise, the latter wins
E. Exercise performance (strength training):
α-GPC 600 mg pre-trainingB-tier evidence, can tryDon't take high doses every day long-term (Lee 2021 stroke signal)
Q2: can you replace it with food / plain choline?
With adequate egg yolks + organ meats + fish, most people have enough choline1–2 g/day choline from food + 1–2 g plain choline bitartrate as a boostReserve the brain-targeted forms (α-GPC / CDP) for special indications
Q3: warnings + contraindications:
High cardiovascular risk + long-term use: Lee 2021 stroke signal — use cautionBipolar depression / anxiety: worsens in some peoplePregnancy / lactation / children: skipOn SSRI / anticholinergic drugs: discuss with physicianTMAO concern + cardiovascular disease: monitor
Q4: cost-effectiveness + realistic expectations:
¥100–300/month (α-GPC or CDP, mid-tier brand)vs prescription nootropics (modafinil etc.): prescription drugs are stronger but carry dependence + side effectsvs lifestyle: sleep + exercise + nutrition are stronger nootropics
The most-overlooked nootropic levers (atlas position):
1. Sleep 7–9 hours — the single strongest
2. Exercise — BDNF + cerebral blood flow
3. Limit screen time + reduce social media — attention training
4. Diet (low UPF + Mediterranean) — chronic inflammation ↓
5. Mindfulness / meditation — attention + working memory
6. Limit alcohol
7. Manage stress — chronic stress damages the brain
8. Learning new skills / social engagement — neuroplasticity
These levers are zero-cost and zero-side-effect, with combined effect far exceeding any nootropic supplement.
Closing the loop with other atlas stories:
choline + tmg-betaine L3 + L4 (upstream choline metabolism)caffeine-l-theanine (synergistic nootropic)vitamin-b12/methylation + folate/one-carbon L4nervous/neurotransmitters L4 (ACh + the synapse big picture)depression-anxiety + sleep-apnea + insomnia (brain health foundations)
Atlas position: α-GPC and CDP are the two largest choline-derived molecules in the nootropic market. The real indications (post-stroke cognition / dementia adjunct / exercise performance) have B-tier evidence; the healthy-person nootropic case is weak, and long-term use carries a cardiovascular warning. Atlas doesn't sell panic and doesn't sell perfection — it lays the information out and lets the reader decide.
Q1: what's your goal?
A. Acute stroke (emergency phase):
Not recommended (ICTUS / COBRIT large RCTs negative)Standard stroke treatment (thrombolysis / thrombectomy + antiplatelet + secondary prevention) takes priority
B. Chronic post-stroke recovery / vascular cognitive impairment:
CDP-choline 500–1000 mg/day can be discussed with a neurologistB-tier evidence, assess at 6 months
C. Alzheimer adjunct (alongside an AChE inhibitor):
α-GPC 600–1200 mg/day can be discussed with a neurologistDoesn't replace donepezil / memantine
D. Healthy-person nootropic (focus / memory):
Minimal benefitTry 4–8 weeks looking at subjective + objective improvement; stop if no changeCDP-choline 250 mg × 2/day or α-GPC 300 mg/dayCost-effectiveness: compared with caffeine + L-theanine + sleep + exercise, the latter wins
E. Exercise performance (strength training):
α-GPC 600 mg pre-trainingB-tier evidence, can tryDon't take high doses every day long-term (Lee 2021 stroke signal)
Q2: can you replace it with food / plain choline?
With adequate egg yolks + organ meats + fish, most people have enough choline1–2 g/day choline from food + 1–2 g plain choline bitartrate as a boostReserve the brain-targeted forms (α-GPC / CDP) for special indications
Q3: warnings + contraindications:
High cardiovascular risk + long-term use: Lee 2021 stroke signal — use cautionBipolar depression / anxiety: worsens in some peoplePregnancy / lactation / children: skipOn SSRI / anticholinergic drugs: discuss with physicianTMAO concern + cardiovascular disease: monitor
Q4: cost-effectiveness + realistic expectations:
¥100–300/month (α-GPC or CDP, mid-tier brand)vs prescription nootropics (modafinil etc.): prescription drugs are stronger but carry dependence + side effectsvs lifestyle: sleep + exercise + nutrition are stronger nootropics
The most-overlooked nootropic levers (atlas position):
1. Sleep 7–9 hours — the single strongest
2. Exercise — BDNF + cerebral blood flow
3. Limit screen time + reduce social media — attention training
4. Diet (low UPF + Mediterranean) — chronic inflammation ↓
5. Mindfulness / meditation — attention + working memory
6. Limit alcohol
7. Manage stress — chronic stress damages the brain
8. Learning new skills / social engagement — neuroplasticity
These levers are zero-cost and zero-side-effect, with combined effect far exceeding any nootropic supplement.
Closing the loop with other atlas stories:
choline + tmg-betaine L3 + L4 (upstream choline metabolism)caffeine-l-theanine (synergistic nootropic)vitamin-b12/methylation + folate/one-carbon L4nervous/neurotransmitters L4 (ACh + the synapse big picture)depression-anxiety + sleep-apnea + insomnia (brain health foundations)
Atlas position: α-GPC and CDP are the two largest choline-derived molecules in the nootropic market. The real indications (post-stroke cognition / dementia adjunct / exercise performance) have B-tier evidence; the healthy-person nootropic case is weak, and long-term use carries a cardiovascular warning. Atlas doesn't sell panic and doesn't sell perfection — it lays the information out and lets the reader decide.
决策路径 · 有临床适应证的三格
α-GPC / CDP-choline 决策路径:Q1: 你的目标?
A. 急性卒中 (急救期):
不推荐 (ICTUS / COBRIT 大 RCT 阴性)标准卒中治疗 (溶栓、取栓 + 抗血小板 + 二级预防) 优先
B. 卒中后慢性恢复、血管性认知障碍:
CDP-choline 500-1000 mg/天 可与神经科医生协商B 级证据, 6 个月评估
C. 阿尔茨海默病辅助 (服 AChE 抑制剂同时):
α-GPC 600-1200 mg/天 可与神经科协商不替代 多奈哌齐、美金刚
决策路径 · 健康人与运动表现
D. 健康人 nootropic (注意力、记忆):效益最小试 4-8 周看主观 + 客观改善, 不改善则停CDP-choline 250 mg × 2次/天 或 α-GPC 300 mg/天性价比: 与 caffeine + L-theanine + 睡眠 + 运动相比 — 后者更高
E. 运动表现 (力量训练):
α-GPC 600 mg 训练前B 级证据, 可试不要每天大剂量长期 (Lee 2021 卒中信号)
替代 · 食物、警告与性价比
Q2: 你能用食物、普通胆碱替代吗?充足蛋黄 + 内脏 + 鱼, 大多数人胆碱足够食物 1-2 g/天胆碱 + 用 1-2 g 普通胆碱 bitartrate 补强偏脑型 (α-GPC / CDP) 留给特殊适应症
Q3: 警告 + 禁忌:
CV 高危 + 长期使用: Lee 2021 卒中信号, 慎抑郁双相、焦虑: 部分人加重怀孕、哺乳、儿童: 跳过服 SSRI / 抗胆碱药: 与医生协商TMAO 担忧 + CV 病: 监测
Q4: 性价比 + 现实期望:
¥100-300/月 (α-GPC 或 CDP, 中等品牌)vs 处方 nootropic (modafinil 等): 处方药效更强但有依赖 + 副作用vs 生活方式: 睡眠 + 运动 + 营养是更强 nootropic
杠杆 · 最被错过的那几件
最被错过的 nootropic 杠杆 (atlas 立场):1. 睡眠 7-9 小时 — 单一最强
2. 运动 — BDNF + 血流
3. 限屏幕 + 减社交媒体 — 注意力训练
4. 饮食 (低 UPF + 地中海) — 慢性炎症 ↓
5. 正念、冥想 — 注意力 + 工作记忆
6. 限酒
7. 管理压力 — 慢性应激损脑
8. 学习新技能、社交 — 神经可塑性
这些杠杆 0 成本、0 副作用, 综合效果远超任何 nootropic 补剂。
Nootropic landscape · A + B tier (real evidence)
The nootropic (brain enhancer) industry was ~$5B globally in 2024, most of which is over-marketed supplements. Atlas gives a real evidence tiering:A-tier (strong evidence, genuinely useful):
Caffeine 50–200 mg: focus + reaction time + alertness; the single strongest evidence (atlas caffeine-l-theanine)L-theanine 100–200 mg + caffeine: synergy (reduces coffee anxiety + boosts focus)Adequate sleep 7–9 hours: the single strongest nootropic, no supplement can replace itExercise (30 min moderate aerobic): BDNF + cerebral blood flow + neuroplasticityMeditation / mindfulness: working memory + attention (Tang 2015 meta)
B-tier (moderate evidence, benefits some people):
Citicoline (CDP-choline) 500–1000 mg: post-stroke cognition + dementia adjunct (covered above)Bacopa monnieri 300 mg: memory consolidation (Kongkeaw 2014 meta), AyurvedicLion's mane (Hericium erinaceus) 1–3 g: neurotrophic factor induction, mild cognitive signal (Mori 2009)Rhodiola rosea 200–400 mg: reduces fatigue + performance under stress (Panossian 2010, atlas adaptogens)Ashwagandha 300–600 mg: stress + anxiety + sleep (Chandrasekhar 2012)Glycine 3 g at bedtime: subjective sleep improvement → indirect nootropicomega-3 EPA + DHA: long-term cognitive protection signal (not acute)
C + D tier + focus ranking + atlas stance
C-tier (weak evidence / mostly commercial):α-GPC: weak healthy-person evidence + stroke signal (Lee 2021)Piracetam / Aniracetam / Oxiracetam: old nootropics, prescription in Europe / China, weak clinical benefitModafinil: prescription drug, REM sleep deprivation / narcolepsy; off-label use is popular but with real dependence + cardiovascular riskPhenibut: GABA analog, real addiction + withdrawal, atlas refuses to recommendDMAE: old nootropic, weak evidencePRL-8-53: a single 1978 trial, never replicated
D-tier / marketing (not recommended):
'Brain gold': mostly omega-3 + gimmick packaging'Multinutrient / Sinala' multi-ingredient products: weak signal'Lumosity / brain games': fined $2M by the FTC in 2016, no evidence for 'brain improvement''NMN / NR': anti-aging marketing, cognitive benefit RCT negative (covered in atlas nmn-nr section)
'Focus / attention' ranking (by real effect):
1. Sleep + exercise + limit social media + meditation ← zero-cost + zero side effects + strong effect
2. Caffeine + L-theanine ← cheap + immediate effect
3. Environment: silence + single-task + reduce notifications
4. Diet: Mediterranean / low UPF + adequate protein
5. (Optional trial) B-tier nootropic: Bacopa / Rhodiola / Citicoline
6. (Clinical indication) prescription drug: after ADHD diagnosis
Atlas position: 'find a supplement that makes me focus' is itself the wrong question. Attention, memory, and thinking capacity are outputs of overall brain health, not products of a single neurotransmitter. Getting the four foundations of sleep, exercise, nutrition, and mental health to 80% returns far more than any nootropic.
On 'cognitive decline worries + wanting a nootropic for anti-aging':
The levers that genuinely prevent cognitive decline: BP control, glucose control, quitting smoking, exercise, education, social engagement, hearing aids, reduced alcohol, treating depressionLivingston 2024 Lancet: combining 14 lifestyle factors can cut dementia risk by 45%No nootropic supplement comes close to this number
闭环 · atlas 立场
与 atlas 其它故事的闭环:choline + tmg-betaine L3 + L4 (上游胆碱代谢)caffeine-l-theanine (协同 nootropic)vitamin-b12/methylation + folate/one-carbon L4nervous/neurotransmitters L4 (ACh + 突触全景)depression-anxiety + sleep-apnea + insomnia (大脑健康基础)
Atlas 立场: α-GPC / CDP 是 nootropic 市场最大的两个胆碱衍生分子。真适应症 (卒中后认知、痴呆辅助、运动表现) 有 B 级证据; 健康人 nootropic 证据弱, 长期还有 CV 警告。不卖恐慌也不卖完美, 把信息给齐, 让读者自己决定。
References · 6
- Alvarez-Sabín, J., & Román, G. C. (2011). Citicoline in vascular cognitive impairment and vascular dementia after stroke. Stroke, 42(1 Suppl), S40-S43. 10.1161/STROKEAHA.110.606509
- Lee, G., Choi, S., Chang, J., Choi, D., Son, J. S., Kim, K., Kim, S. M., Jeong, S., & Park, S. M. (2021). Association of L-α glycerylphosphorylcholine with subsequent stroke risk after 10 years. JAMA Network Open, 4(11), e2136008. 10.1001/jamanetworkopen.2021.36008
- Kongkeaw, C., Dilokthornsakul, P., Thanarangsarit, P., Limpeanchob, N., & Norman Scholfield, C. (2014). Meta-analysis of randomized controlled trials on cognitive effects of Bacopa monnieri extract. Journal of Ethnopharmacology, 151(1), 528–535. pubmed.ncbi.nlm.nih.gov/24252493
- Panossian, A., Wikman, G., & Sarris, J. (2010). Rosenroot (Rhodiola rosea): traditional use, chemical composition, pharmacology and clinical efficacy. Phytomedicine, 17(7), 481–493. 10.1016/j.phymed.2010.02.002
- Chandrasekhar, K., Kapoor, J., & Anishetty, S. (2012). A prospective, randomized double-blind, placebo-controlled study of safety and efficacy of a high-concentration full-spectrum extract of Ashwagandha root in reducing stress and anxiety in adults. Indian Journal of Psychological Medicine, 34(3), 255–262. 10.4103/0253-7176.106022
- Livingston, G., Huntley, J., Liu, K. Y., et al. (2024). Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet, 404(10452), 572-628. Adds high LDL cholesterol and untreated vision loss for 14 modifiable factors totaling ~45% of dementia risk. 10.1016/S0140-6736(24)01296-0