Place · Level 3 · Condition
Stroke · the one thing most likely to happen to you
脑每分钟死 190 万神经元 · 缺血和出血是相反的两种病 · 别自行吃阿司匹林 · 半暗带决定时间窗 · 症状消失更要去急诊 · 十个因素解释九成风险
Story path
- 1How badly the brain takes ischemiaHow badly the brain takes ischemia
- 2Ischemic vs hemorrhagic · oppositesIschemic vs hemorrhagic · opposites
- 3Three roads into ischemic strokeThree roads into ischemic stroke
- 4The penumbra · where the clock comes fromThe penumbra · where the clock comes from
- 5Recognize · call nowRecognize · call now
- 6Prevention is the real battlefieldPrevention is the real battlefield
Chapter 1
How badly the brain takes ischemia
How badly the brain takes ischemia
Your brain is only 2% of your body weight, yet it eats about a fifth of all the oxygen and energy you use (Raichle 2002). It stores almost no fuel — when blood stops, it starts dying within minutes.
So when a brain artery blocks, time becomes an organ. Someone did the arithmetic: in a typical large-vessel ischemic stroke, left untreated, about 1.9 million neurons die every minute, along with 14 billion synapses and 12 km of nerve fibers (Saver 2006). Converted to another scale, the ischemic brain ages roughly 3.6 years per hour.
So *time is brain* isn't a slogan, it's a sum: every minute you hesitate is subtracted from the speaking, walking, and face-recognizing of the rest of your life.
The brain isn't entirely without backup. Small bypass channels run between arteries — collaterals; when the main road blocks, some blood still squeezes around. And blood flow doesn't fall to zero in one step: it crosses two thresholds in turn. First it drops below the level where neurons stop firing (which is when that side of your body stops obeying you), and only lower down does it cross the level where cells actually die (Astrup 1981). People whose collaterals reach land between the two thresholds, and are salvageable; people whose don't fall straight through. That's why the same blocked artery leaves one person nearly fine and another instantly paralyzed on one side.
So when a brain artery blocks, time becomes an organ. Someone did the arithmetic: in a typical large-vessel ischemic stroke, left untreated, about 1.9 million neurons die every minute, along with 14 billion synapses and 12 km of nerve fibers (Saver 2006). Converted to another scale, the ischemic brain ages roughly 3.6 years per hour.
So *time is brain* isn't a slogan, it's a sum: every minute you hesitate is subtracted from the speaking, walking, and face-recognizing of the rest of your life.
The brain isn't entirely without backup. Small bypass channels run between arteries — collaterals; when the main road blocks, some blood still squeezes around. And blood flow doesn't fall to zero in one step: it crosses two thresholds in turn. First it drops below the level where neurons stop firing (which is when that side of your body stops obeying you), and only lower down does it cross the level where cells actually die (Astrup 1981). People whose collaterals reach land between the two thresholds, and are salvageable; people whose don't fall straight through. That's why the same blocked artery leaves one person nearly fine and another instantly paralyzed on one side.
The numbers in China
Stroke is the leading cause of death in China, with over 2 million new cases a year, and it costs more healthy life-years than any other disease (Wu 2019).The lifetime risk is starker. The Global Burden of Disease study computed, across 195 countries, the probability that an adult aged 25+ will have a stroke: the global average is about 24.9%, while China's is 39.3% — the highest in the world (41.1% in men, 36.7% in women) (GBD 2018).
A national door-to-door survey of 480,687 adults also gives the breakdown (Wang 2017):
Of incident strokes, 69.6% are ischemic23.8% are intracerebral hemorrhage4.4% are subarachnoid hemorrhageAmong stroke survivors, 88% had hypertension, 48% smoked, 44% drank
Hold on to that 23.8% for hemorrhage — the next scene uses it, and it changes what you should do at the scene.
wu-2019-stroke-chinagbd-2018-lifetime-risk-stroke
Chapter 2
Ischemic vs hemorrhagic · opposites
Ischemic vs hemorrhagic · opposites
The word *stroke* hides two diseases that run in opposite directions.
Ischemic: a clot blocks the artery and the tissue downstream loses its blood supply. Treating it means getting the blockage open — thrombolysis, thrombectomy.Hemorrhagic: a vessel has ruptured and blood is leaking into brain tissue. Treating it means getting the bleeding stopped.
Same event; one needs opening, the other needs closing. The drug that saves an ischemic stroke pours fuel on a hemorrhagic one.
The problem: the symptoms can be identical. One side of the face suddenly droops, one arm won't lift, speech slurs — from those signs alone you have no way at home to tell a block from a bleed. Only one thing can: a head CT at the hospital, which takes minutes.
This is why you don't self-medicate at the scene. The line every family passes around — *take an aspirin first* — makes sense for a heart attack, but transplanted onto stroke it turns dangerous: aspirin stops platelets clumping, and if this is a hemorrhagic stroke that is exactly the direction you don't want — you've slowed the clotting on a vessel that is actively leaking (AHA/ASA 2022 ICH guideline).
China's numbers raise the stakes. That 23.8% from the last scene: roughly one in four strokes here is a bleed. Gambling on an aspirin in the living room isn't a one-in-ten-thousand bet — it's one in four.
So the emergency-department order is fixed: imaging first, drugs second. Only once CT has ruled out hemorrhage will a doctor move toward thrombolysis or antiplatelets (AHA/ASA 2026). That step cannot be skipped in your living room.
Ischemic: a clot blocks the artery and the tissue downstream loses its blood supply. Treating it means getting the blockage open — thrombolysis, thrombectomy.Hemorrhagic: a vessel has ruptured and blood is leaking into brain tissue. Treating it means getting the bleeding stopped.
Same event; one needs opening, the other needs closing. The drug that saves an ischemic stroke pours fuel on a hemorrhagic one.
The problem: the symptoms can be identical. One side of the face suddenly droops, one arm won't lift, speech slurs — from those signs alone you have no way at home to tell a block from a bleed. Only one thing can: a head CT at the hospital, which takes minutes.
This is why you don't self-medicate at the scene. The line every family passes around — *take an aspirin first* — makes sense for a heart attack, but transplanted onto stroke it turns dangerous: aspirin stops platelets clumping, and if this is a hemorrhagic stroke that is exactly the direction you don't want — you've slowed the clotting on a vessel that is actively leaking (AHA/ASA 2022 ICH guideline).
China's numbers raise the stakes. That 23.8% from the last scene: roughly one in four strokes here is a bleed. Gambling on an aspirin in the living room isn't a one-in-ten-thousand bet — it's one in four.
So the emergency-department order is fixed: imaging first, drugs second. Only once CT has ruled out hemorrhage will a doctor move toward thrombolysis or antiplatelets (AHA/ASA 2026). That step cannot be skipped in your living room.
What to do and not do at the scene
If you suspect a stroke, what not to do:Don't give any medication — aspirin, blood-pressure pills, heart pills, whatever herbal preparation is in the cabinet. You can't tell a block from a bleed, so any of them could be the wrong one.Don't give water or food. Stroke often knocks out swallowing too, so anything you give can go into the lungs; hospitals themselves run a swallow screen first and only allow oral intake once it passes (AHA/ASA 2026).Don't hand out blood-pressure medication just because the reading is high. Acute-phase blood pressure has its own rules — leave it to the doctor.Don't pinch pressure points, don't bleed them, don't shake them, don't slap them.
What to do:
Call emergency services immediately.Note the time they were last seen normal — that single fact decides what treatments are open to them.Lay them on their side, head slightly raised, collar loosened.Find the medications they normally take and bring them along.
Why an ambulance rather than your own car: the ambulance can call ahead so the CT scanner and thrombolysis team are standing by, and the patient goes straight into the pathway. Driving yourself means starting again at the triage desk.
Chapter 3
Three roads into ischemic stroke
Three roads into ischemic stroke
The same blocked artery in the brain can have completely different stories upstream. Clinically, ischemic stroke is subtyped by cause, and the scheme that has been in use for thirty years is TOAST (Adams 1993). Its trunk is three roads.
Road one · large-artery atherosclerosis. In the wall of a carotid or large intracranial artery, a plaque builds up over years. One day the fibrous cap on its surface tears, platelets pile on and form a clot: either it blocks that artery on the spot, or fragments wash downstream and plug a finer branch (Libby 2011). The real upstream here is apolipoprotein B: One sits on every artery-clogging particle, so counting it counts the harmful particles directly.-containing lipoprotein particles (the ones read alongside LDL on your panel) burrowing into the artery wall — for how particle counts work, dive into the dyslipidemia island.
Road two · cardioembolism. In atrial fibrillation the left atrium stops contracting rhythmically and quivers instead. Blood stagnates in a cul-de-sac called the left atrial appendage and slowly congeals; one day a clot washes out and rides the carotid straight up into the brain. In non-rheumatic atrial fibrillation, about nine out of ten left-atrial thrombi sit in that appendage (Blackshear 1996). These emboli tend to be large and block large vessels, which is why cardioembolic strokes are usually more severe. For how atrial fibrillation gets cultivated in the first place, dive into the sleep-apnea island.
Road three · small-vessel disease. Years of high blood pressure slowly remodel the perforating arteries deep in the brain (only 40 to 200 microns across): the walls thicken, hyalinize, and undergo lipohyalinosis, the lumen narrows and finally closes, leaving a tiny lacunar infarct (Wardlaw 2013). The sly part is that each small infarct may cause no symptoms at all, but stacked up they become unsteady walking, slower thinking, cognitive decline. For how pressure remodels a vessel like this, dive into the hypertension island.
One symptom, three entirely different upstreams. This is why a doctor must pin down the cause — the AHA/ASA 2021 secondary-prevention guideline even sets a clock: after a first stroke or transient ischemic attack, the diagnostic work-up should start within 48 hours (Kleindorfer 2021). Because if the reason for the block differs, the prevention differs completely: large-artery atherosclerosis needs ApoB driven down, cardioembolism needs anticoagulation, small-vessel disease needs blood pressure down. The wrong key opens no door.
Road one · large-artery atherosclerosis. In the wall of a carotid or large intracranial artery, a plaque builds up over years. One day the fibrous cap on its surface tears, platelets pile on and form a clot: either it blocks that artery on the spot, or fragments wash downstream and plug a finer branch (Libby 2011). The real upstream here is apolipoprotein B: One sits on every artery-clogging particle, so counting it counts the harmful particles directly.-containing lipoprotein particles (the ones read alongside LDL on your panel) burrowing into the artery wall — for how particle counts work, dive into the dyslipidemia island.
Road two · cardioembolism. In atrial fibrillation the left atrium stops contracting rhythmically and quivers instead. Blood stagnates in a cul-de-sac called the left atrial appendage and slowly congeals; one day a clot washes out and rides the carotid straight up into the brain. In non-rheumatic atrial fibrillation, about nine out of ten left-atrial thrombi sit in that appendage (Blackshear 1996). These emboli tend to be large and block large vessels, which is why cardioembolic strokes are usually more severe. For how atrial fibrillation gets cultivated in the first place, dive into the sleep-apnea island.
Road three · small-vessel disease. Years of high blood pressure slowly remodel the perforating arteries deep in the brain (only 40 to 200 microns across): the walls thicken, hyalinize, and undergo lipohyalinosis, the lumen narrows and finally closes, leaving a tiny lacunar infarct (Wardlaw 2013). The sly part is that each small infarct may cause no symptoms at all, but stacked up they become unsteady walking, slower thinking, cognitive decline. For how pressure remodels a vessel like this, dive into the hypertension island.
One symptom, three entirely different upstreams. This is why a doctor must pin down the cause — the AHA/ASA 2021 secondary-prevention guideline even sets a clock: after a first stroke or transient ischemic attack, the diagnostic work-up should start within 48 hours (Kleindorfer 2021). Because if the reason for the block differs, the prevention differs completely: large-artery atherosclerosis needs ApoB driven down, cardioembolism needs anticoagulation, small-vessel disease needs blood pressure down. The wrong key opens no door.
Chapter 4
The penumbra · where the clock comes from
The penumbra · where the clock comes from
The moment an artery blocks, the brain tissue downstream does not die all at once.
The innermost part, farthest from any collateral, loses flow hardest and is genuinely dead within minutes — this is the infarct core. The ring around it, fed by a trickle sneaking in from neighboring collaterals, sits at an awkward middle value: not enough flow for neurons to work, but still enough to keep them from dying.
That middle zone has a name: the ischemic penumbra. The two thresholds from the last scene are its definition — the ring of cells stuck between stopping firing and dying is the penumbra (Astrup 1981).
The crucial part: the penumbra is still alive, just shut down. Restore the blood and those cells clock back in, and that side of your body may come back.
So thrombolysis and thrombectomy are never racing for the core — that's already gone. They're racing for the ring. And the penumbra doesn't wait: every minute, some of it slides from shut-down into dead, eaten away by the growing core.
That is where the treatment clock physically comes from. The time window isn't an administrative rule — it's how long that ring can hold out.
The innermost part, farthest from any collateral, loses flow hardest and is genuinely dead within minutes — this is the infarct core. The ring around it, fed by a trickle sneaking in from neighboring collaterals, sits at an awkward middle value: not enough flow for neurons to work, but still enough to keep them from dying.
That middle zone has a name: the ischemic penumbra. The two thresholds from the last scene are its definition — the ring of cells stuck between stopping firing and dying is the penumbra (Astrup 1981).
The crucial part: the penumbra is still alive, just shut down. Restore the blood and those cells clock back in, and that side of your body may come back.
So thrombolysis and thrombectomy are never racing for the core — that's already gone. They're racing for the ring. And the penumbra doesn't wait: every minute, some of it slides from shut-down into dead, eaten away by the growing core.
That is where the treatment clock physically comes from. The time window isn't an administrative rule — it's how long that ring can hold out.
The windows, and why they're stretching
The physiology of the penumbra grows directly into today's treatment windows:IV thrombolysis (a drug into the vein to dissolve the clot): the standard window is within 4.5 hours of onset (Hacke 2008, ECASS III). The earlier NINDS trial set 3 hours in 1995 (NINDS 1995); ECASS III later pushed it to 4.5.Endovascular thrombectomy (a catheter that pulls the clot out directly): for large-vessel occlusion the window stretches to 6 to 24 hours after onset (DAWN, Nogueira 2018; DEFUSE 3, Albers 2018).
Why that last one stretches so far is worth seeing clearly: DAWN and DEFUSE 3 selected patients not by the clock but by imaging — using perfusion scans to see how much penumbra is left. Someone with a big penumbra and a small core still benefits at 20 hours; someone whose penumbra burned out long ago gains nothing at 3 hours.
So the real rule isn't *how many hours* — it's *how much salvageable brain is left*. And on average that amount shrinks every minute: measured across 58,353 real cases, the earlier thrombolysis starts, the better the outcome, and the relationship is continuous (Saver 2013).
The window is still stretching. In the 2026 AHA/ASA acute ischemic stroke guideline, selected patients chosen by imaging can receive IV thrombolysis 4.5 to 9 hours after onset, and thrombectomy for basilar occlusion is given out to 24 hours (AHA/ASA 2026). The direction is consistent: select by imaging, not only by the clock.
But none of this is permission to wait. Every window above is a ceiling, not a target. Earlier reperfusion salvages more — and whether you qualify at all depends on whether the onset time can be stated. That's why the instruction at the scene, *note the time they were last seen normal*, matters so much.
hacke-2008-ecass3ninds-1995-tpanogueira-2018-dawnalbers-2018-defuse3saver-2013-time-to-treatment
Chapter 5
Recognize · call now
Recognize · call now
Stroke symptoms almost always share two features: they're sudden, and they're one-sided.
The internationally taught mnemonic is FAST:
F (face): ask them to smile — is one corner of the mouth drooping?A (arm): both arms out level — does one drift down?S (speech): ask them to say a sentence — is it slurred, absent, or are they unable to understand you?T (time): if any of these, note the time and call an ambulance immediately
The Chinese-language world has a version that's easier to remember, called Stroke 1-2-0 (Zhao 2017):
1 look at one face: is it asymmetric, is the mouth pulled to one side2 check two arms: held level, is one side weak0 listen to their speech: is it unclear
Any one of the three, dial 120 immediately. The clever part of the design is that the mnemonic and the number you must dial are the same thing.
Call emergency services right now if (don't drive them, don't wait for family to get off work, don't give medication first):
Sudden numbness or weakness of the face, arm, or leg on one sideSudden trouble speaking, or trouble understanding othersSudden trouble seeing in one or both eyesSudden dizziness, unsteady walking, loss of balanceA sudden headache that is the worst of their life (possible subarachnoid hemorrhage)
This page is education to understand the *why*; it does not replace a doctor. If any of the above appears, seek care immediately.
The internationally taught mnemonic is FAST:
F (face): ask them to smile — is one corner of the mouth drooping?A (arm): both arms out level — does one drift down?S (speech): ask them to say a sentence — is it slurred, absent, or are they unable to understand you?T (time): if any of these, note the time and call an ambulance immediately
The Chinese-language world has a version that's easier to remember, called Stroke 1-2-0 (Zhao 2017):
1 look at one face: is it asymmetric, is the mouth pulled to one side2 check two arms: held level, is one side weak0 listen to their speech: is it unclear
Any one of the three, dial 120 immediately. The clever part of the design is that the mnemonic and the number you must dial are the same thing.
Call emergency services right now if (don't drive them, don't wait for family to get off work, don't give medication first):
Sudden numbness or weakness of the face, arm, or leg on one sideSudden trouble speaking, or trouble understanding othersSudden trouble seeing in one or both eyesSudden dizziness, unsteady walking, loss of balanceA sudden headache that is the worst of their life (possible subarachnoid hemorrhage)
This page is education to understand the *why*; it does not replace a doctor. If any of the above appears, seek care immediately.
Wait and see · the most expensive decision
The most expensive decision at the scene of a stroke is let's wait and see if it gets better on its own.Why it's so expensive, the penumbra scene already worked out for you: every minute you wait, the salvageable ring shrinks.
And *wait and see* often appears to win — because stroke symptoms genuinely do sometimes resolve. Which brings us to the most dangerous situation of all.
Transient ischemic attack (TIA): symptoms identical to a stroke, but they disappear on their own within minutes to an hour. Many people exhale, decide it was nothing, and skip the hospital.
This is exactly backwards. A TIA is not a false alarm — it's the strongest warning you will ever get: the artery has already blocked once, and this time the clot happened to break up by itself. Next time it may not. And the most dangerous window after a TIA is precisely the first few days.
The good news is that the warning can be caught. The EXPRESS study in Oxford did something very plain: it moved TIA and minor-stroke patients from *waiting their turn in the clinic queue* to *assessed the same day, secondary prevention started the same day*. Early recurrent stroke risk fell by about 80% (Rothwell 2007). Same drugs — just given days sooner.
So hold on to this line: if the symptoms went away, that's more reason to go to hospital now — to the emergency department, not next week's clinic.
rothwell-2007-express
Chapter 6
Prevention is the real battlefield
Prevention is the real battlefield
Everything so far has been about minutes. But the real battlefield for stroke is the decades before it happens.
Start with a number that should steady you. INTERSTROKE ran a case-control study across 32 countries and concluded: ten modifiable factors together account for about 90.7% of stroke risk worldwide (91.5% for ischemic, 87.1% for intracerebral hemorrhage) (O'Donnell 2016).
The ten: hypertension (global PAR 47.9%, the largest single factor), physical inactivity, blood lipids, diet, waist-to-hip ratio, psychosocial factors, smoking, cardiac causes (atrial fibrillation and others), alcohol, diabetes.
In other words: stroke is not fate. Nine-tenths of it is written in things you can change.
China's own two large trials each pin down one end.
CHANCE (Wang 2013): 5,170 Chinese patients, within 24 hours of a minor stroke or high-risk transient ischemic attack (TIA), randomized to clopidogrel plus aspirin for 21 days then single-agent, versus aspirin alone throughout. Stroke recurrence at 90 days fell by about 32%, with no increase in bleeding. This Chinese result went on to be written into guidelines worldwide.
Note it does not contradict the *don't take aspirin yourself* line from scene two: the difference is after the CT, confirmed ischemic, prescribed by a doctor. Get the order right and the same drug turns from dangerous into life-saving.
SSaSS (Neal 2021): over twenty thousand rural Chinese residents (most with prior stroke, or aged 60+ with elevated blood pressure) swapped household cooking salt for a low-sodium salt of 75% sodium chloride plus 25% potassium chloride. Over about 5 years — stroke down 14%, cardiovascular events down 13%, all-cause death down 12%. From changing one bag of salt. (Not for people with chronic kidney disease or on potassium-sparing drugs — hyperkalemia risk; details on the hypertension island.)
Start with a number that should steady you. INTERSTROKE ran a case-control study across 32 countries and concluded: ten modifiable factors together account for about 90.7% of stroke risk worldwide (91.5% for ischemic, 87.1% for intracerebral hemorrhage) (O'Donnell 2016).
The ten: hypertension (global PAR 47.9%, the largest single factor), physical inactivity, blood lipids, diet, waist-to-hip ratio, psychosocial factors, smoking, cardiac causes (atrial fibrillation and others), alcohol, diabetes.
In other words: stroke is not fate. Nine-tenths of it is written in things you can change.
China's own two large trials each pin down one end.
CHANCE (Wang 2013): 5,170 Chinese patients, within 24 hours of a minor stroke or high-risk transient ischemic attack (TIA), randomized to clopidogrel plus aspirin for 21 days then single-agent, versus aspirin alone throughout. Stroke recurrence at 90 days fell by about 32%, with no increase in bleeding. This Chinese result went on to be written into guidelines worldwide.
Note it does not contradict the *don't take aspirin yourself* line from scene two: the difference is after the CT, confirmed ischemic, prescribed by a doctor. Get the order right and the same drug turns from dangerous into life-saving.
SSaSS (Neal 2021): over twenty thousand rural Chinese residents (most with prior stroke, or aged 60+ with elevated blood pressure) swapped household cooking salt for a low-sodium salt of 75% sodium chloride plus 25% potassium chloride. Over about 5 years — stroke down 14%, cardiovascular events down 13%, all-cause death down 12%. From changing one bag of salt. (Not for people with chronic kidney disease or on potassium-sparing drugs — hyperkalemia risk; details on the hypertension island.)
The skeleton of secondary prevention
The skeleton of the AHA/ASA 2021 secondary-prevention guideline (Kleindorfer 2021) is roughly four things:1. Pin down the cause first. Start the work-up within 48 hours of a first stroke or transient ischemic attack (TIA). Large-artery, cardioembolic, small-vessel — the drugs that follow are completely different.
2. Blood pressure. The largest single factor by global PAR (47.9%). But INTERSTROKE states plainly that the ranking varies by region — in China the highest PAR actually belongs to physical inactivity. So don't map the global leaderboard onto yourself: the robust finding is that the ten factors *together* explain about 90%, not that any one of them always ranks first. Mechanism, the DASH diet, and salt substitution all live on the hypertension island.
3. Lipids. After an ischemic stroke, apolipoprotein B: One sits on every artery-clogging particle, so counting it counts the harmful particles directly.-containing particles (the ones read alongside LDL on your panel) need to come down. In a statin meta-analysis of 170,000 people, each 1 mmol/L drop in LDL cut major vascular events by about a fifth (CTT 2010). The particle-count arithmetic is on the dyslipidemia island.
4. Antithrombotics. Here's a fork you must not blur: ischemic stroke from large-artery or small-vessel disease takes an antiplatelet; cardioembolism from atrial fibrillation takes an anticoagulant, not aspirin. The two are not interchangeable.
Add quitting smoking, moving, and glucose control (dive into the type-2-diabetes island), plus a check for sleep apnea — it feeds hypertension and atrial fibrillation at once, which means it feeds two of the three roads at once (dive into the sleep-apnea island).
On whether people who have never had a stroke should routinely take aspirin for primary prevention: that answer has changed in recent years, and it is no longer *everyone should*. Talk to a doctor; don't decide this yourself.
ctt-2010-statin-meta
Myths that are especially stubborn in Chinese
Seasonal IV drips to flush out your vessels? A blood vessel isn't a water pipe and plaque isn't limescale — you can't rinse it out. Plaque grows inside the vessel wall, not floating in the lumen (Libby 2011), so flushing the lumen can't anatomically reach it. No mainstream stroke-prevention guideline lists periodic infusions among its measures (Kleindorfer 2021). And the ginkgo preparations commonly used in those drips found no convincing evidence in the Cochrane review of acute ischemic stroke (10 trials, 792 people) (Zeng 2005). What you spend is money, time, and an unnecessary needle.Pricking the fingers to bleed at the first sign of stroke? This one spreads hardest on WeChat and is the most dangerous. Nothing supports it, and its real lethality is delay: the 20 minutes you spend pricking fingers at home is, by the arithmetic in scene one, about 38 million neurons (Saver 2006). This isn't *no harm in trying* — trying costs brain.
Nattokinase, ginkgo, fish oil dissolve clots? Be precise here, because the truth isn't simply *they all get digested*. The evidence for nattokinase reaches only surrogate markers: the 2023 meta-analysis (6 RCTs, 546 people) looked at blood pressure, lipids, and coagulation indices — not one endpoint was stroke or a cardiovascular event (Li 2023). Meaning it has never been shown to dissolve a real clot lodged in your brain, and certainly can't substitute for thrombolysis. Supplements work on a scale of months and years; stroke works on a scale of minutes. The two aren't in the same order of magnitude.
Symptoms vanished in a few minutes, so it's fine? As the last scene said: that's a transient ischemic attack (TIA), the strongest warning you will ever get — not *fine* (Rothwell 2007).
Young people don't have strokes? They do. The stroke death burden among Chinese aged 15 to 49 has never gone away, and among young men the death burden from ischemic stroke and subarachnoid hemorrhage is still rising (Wang 2024). Youth is not a talisman — hypertension, smoking, and atrial fibrillation work at any age.
The honest close: a person in China has about a 39.3% lifetime probability of stroke, the highest in the world (GBD 2018). That number sounds frightening, but read it alongside INTERSTROKE's nine-tenths and the conclusion is actually optimistic — the country with the highest risk is exactly the country with the most room to change it. Blood pressure, tobacco, atrial fibrillation, salt: these largely decide how much of that 39.3% lands on you.
And if it does happen, there's only one thing to do, already written above: look at one face, check two arms, listen to the speech, call 120 now.
zeng-2005-cochrane-ginkgoli-2023-nattokinase-metarothwell-2007-expresswang-2024-young-stroke-chinagbd-2018-lifetime-risk-stroke