What Is a Watershed Stroke and What Causes It?

A watershed stroke is an ischemic stroke that occurs in the border zones between the territories supplied by the brain’s major arteries, rather than in the core of any single artery’s territory. These border zones sit at the farthest reach of each artery’s blood supply, making them the first areas to suffer when blood flow drops or when small clots drift into hard-to-reach vessels.¹ The causes turn out to be more layered than a simple drop in blood pressure, involving an interplay between reduced flow and tiny emboli that most people never hear about.

Why the Brain Has Vulnerable Border Zones

The brain receives its blood from three major cerebral arteries on each side: the anterior, middle, and posterior cerebral arteries. Each one supplies a defined region of brain tissue, but at the outer edges of those regions, coverage thins out. The tissue sitting at the boundary between two arterial territories gets the tail end of supply from both arteries but is not the priority of either one. Think of two garden sprinklers placed side by side: the grass right at the edge of each sprinkler’s reach gets the least water. In the brain, these “last-meadow” strips of tissue are called watershed zones or border zones.²

This anatomy means that when blood flow through one or more arteries falls even modestly, the border zones are the first to be starved. It is the same principle behind why the tips of fingers and toes are the first body parts to suffer frostbite: they are at the ends of the supply line. Because these zones exist wherever two major arterial territories meet, watershed strokes can appear in multiple locations across both hemispheres simultaneously, which is unusual for other types of stroke.

Two Distinct Patterns

Clinicians recognize two main types of watershed infarct, and the distinction matters because each one points toward a somewhat different mechanism.²

External, or cortical, watershed infarcts occur along the brain’s surface at the junction between two major arterial territories. You can picture these as wedge-shaped zones running along the outer edge of the brain. Internal, or subcortical, watershed infarcts occur deeper in the brain’s white matter, in the border zone between the deep penetrating branches of the major arteries and the arteries that supply the cortical surface. On imaging, internal watershed infarcts sometimes appear as a string of small lesions lined up in a row through the white matter, a pattern researchers have called “rosary-like” or “string of pearls.”³

The difference is clinically meaningful. Internal watershed strokes are more strongly linked to a straightforward drop in blood flow, while cortical watershed strokes appear to have a more complicated mix of causes, including tiny clot fragments that travel into the distal vessels. When a patient has both types at once, severe hemodynamic compromise is usually the driving factor.

How Reduced Blood Flow and Tiny Emboli Work Together

For decades, the standard explanation for watershed strokes was purely hemodynamic: blood pressure drops, the border zones lose perfusion first, and tissue dies. That story is true but incomplete. Research has shown that small emboli, tiny clot fragments or bits of plaque debris, play a significant role as well, especially in cortical watershed strokes.³

A key concept that ties these two mechanisms together is “impaired washout.” Under normal conditions, the bloodstream has enough flow to flush small emboli through and out of the brain’s arterial tree before they can lodge and cause damage. But when perfusion is already reduced, perhaps because of a narrowed carotid artery, the blood moves more slowly and cannot clear those tiny fragments. The emboli get stuck in the most distal arterioles, which happen to be in the border zones. So reduced flow and embolism are not two competing explanations; they reinforce each other.⁴

Narrowing of arteries also creates the conditions for more emboli to form in the first place. Unstable plaque and damaged vessel walls promote clot formation, which then sheds fragments downstream. The same artery that is restricting flow is also generating the debris that gets trapped because flow is restricted. It is a vicious cycle that makes carotid artery disease one of the most important risk factors for watershed strokes.

Studies using transcranial Doppler ultrasound to detect microembolic signals and MRI perfusion imaging to measure blood flow have confirmed that both mechanisms are frequently present in the same patient. One study found that perfusion-based imaging picked up reduced blood flow in watershed stroke patients with high sensitivity, while Doppler detected embolic signals in a substantial proportion as well.⁵ In practice, this means that attributing a given watershed stroke entirely to low blood pressure or entirely to embolism is often an oversimplification.

Common Risk Factors and Triggers

Severe carotid artery stenosis is the risk factor most closely associated with watershed strokes. When one of the internal carotid arteries is significantly narrowed, blood supply to the entire downstream territory drops, and the border zones bear the brunt. Whether a patient with carotid narrowing actually suffers a watershed stroke depends partly on the anatomy of their backup circulation, specifically the circle of Willis, a ring of connecting arteries at the base of the brain that can reroute blood when one supply artery is compromised.

Not everyone has a complete circle of Willis. A study of patients with severe carotid stenosis found that those who had recently suffered a stroke were far less likely to have key connecting arteries present and functional. Patients with a recent stroke had any collateral pathway present in about two-thirds of cases, compared to roughly 85 percent of those without a recent stroke. Having those connecting arteries was associated with a substantially lower risk of ischemic stroke.⁶ In other words, your natural backup plumbing matters a lot, and many people lack the anatomy that would protect them.

Cardiac surgery is another well-known trigger. During open-heart procedures, blood pressure can fluctuate dramatically, and the brain may be exposed to periods of reduced flow. A study of patients who developed strokes after cardiac surgery found that those whose average blood pressure dropped by at least 10 mmHg during the operation were about four times more likely to develop bilateral watershed infarcts than other stroke patterns.⁷ Bilateral watershed infarcts, meaning both hemispheres are hit, are particularly telling because they suggest a global perfusion problem rather than a clot in a single artery.

Other situations that can precipitate watershed strokes include cardiac arrest, severe bleeding, septic shock, and any condition that causes prolonged low blood pressure. People with pre-existing carotid disease are at heightened risk during any of these events because their baseline perfusion to the border zones is already marginal.

What Symptoms Look Like

Watershed strokes produce symptoms that reflect the location of the affected border zones. Because the cortical border zones between the anterior and middle cerebral arteries overlap with areas that control the shoulder and upper arm, weakness in the proximal arms is a hallmark. In severe bilateral cases, this leads to a striking presentation called “man-in-the-barrel” syndrome, where the patient can move their hands and legs but cannot lift or control their upper arms, as if their torso is trapped in a barrel with their arms pinned at the sides.⁸

The syndrome looks alarming and unusual because most strokes cause weakness on one side of the body in a pattern that affects the arm, face, and leg together. Man-in-the-barrel syndrome spares the face and legs while hitting both arms at the shoulders, which can initially puzzle clinicians who are not thinking about watershed territory. Other symptoms can include visual field deficits when the posterior border zones are involved, or language difficulties if the left-hemisphere border zones are damaged in the dominant hemisphere.

One important clinical point: watershed strokes do not always arrive suddenly the way other strokes do. Because they are often tied to chronic low-flow states rather than a sudden arterial blockage, symptoms can develop gradually or fluctuate, sometimes worsening when blood pressure drops, such as when a patient stands up quickly, and improving when they lie down. This stuttering course can delay diagnosis.

How They Are Identified on Imaging

Watershed strokes have characteristic patterns on brain imaging that a trained radiologist can recognize. On CT or MRI, the infarcts follow the boundary lines between arterial territories rather than sitting within a single artery’s domain. Internal watershed infarcts show up as a chain of small lesions running through the deep white matter, while external watershed infarcts appear as wedge-shaped areas along the cortical surface.

Advanced MRI techniques can go further by measuring actual blood flow to these border zones before a stroke occurs. Multimodal MRI studies of patients with asymptomatic carotid stenosis have detected significant impairments in blood flow, vascular reactivity, and other hemodynamic measures specifically within the watershed areas. The most pronounced changes were found in the border zones on the side of the narrowed artery, even in patients who had not yet had symptoms.⁹ This kind of imaging could eventually help identify patients at highest risk before a stroke happens, though it is not yet a routine screening tool.

Blood Pressure Management Is Tricky

Managing blood pressure after a watershed stroke is more nuanced than after other stroke types. With most ischemic strokes, aggressively lowering very high blood pressure is standard practice to prevent further damage. But watershed strokes are, by definition, the result of insufficient blood flow to vulnerable areas. Lowering blood pressure too aggressively can worsen the ischemia by further reducing perfusion to the already-starved border zones.

A study examining blood pressure and neurological outcomes in watershed stroke patients found that very high baseline blood pressure, above roughly 180/100 mmHg, was associated with worsening neurological function. But the relationship was not linear: below that threshold, higher blood pressure was not clearly harmful. A modest rise in diastolic blood pressure of a few points on day three after the stroke was actually linked to a reduced risk of neurological decline.¹⁰ This suggests that for watershed stroke patients, permitting moderately elevated blood pressure in the acute period, rather than lowering it reflexively, could help maintain perfusion to the border zones.

In practice, clinicians treating watershed strokes often need to maintain or even augment blood pressure with fluids or medications, at least initially, which feels counterintuitive given that most stroke protocols push toward blood pressure reduction. Getting the balance wrong in either direction is risky. The underlying cause also needs to be addressed: if severe carotid stenosis is driving the low flow, procedures to open or bypass the narrowing may be necessary to prevent recurrence.

Long-Term Cognitive Effects

Watershed strokes, even small ones, can leave lasting marks on cognition. A large autopsy study found that people with multiple tiny infarcts specifically in the cortical watershed regions had measurably lower overall cognitive function compared to those with microinfarcts elsewhere. The deficits were particularly notable in working memory and visuospatial abilities, and these associations held up even after accounting for microinfarcts in other brain regions and other age-related brain pathology like Alzheimer’s plaques.¹¹

There is also evidence that watershed infarction from internal carotid artery occlusion can trigger progressive dementia and brain atrophy over time. A case report documented a patient whose watershed stroke from carotid occlusion was followed by slowly worsening dementia and progressive shrinkage of the affected brain tissue over the subsequent months, with a clear correlation between the degree of atrophy and the severity of cognitive decline.¹² While a single case cannot establish a general rule, the observation aligns with the broader finding that chronic low-flow states can cause ongoing damage beyond the initial stroke event.

These findings suggest that watershed strokes should not be dismissed as “minor” just because they may initially seem smaller or less dramatic than a large territorial stroke. The border zones they damage are involved in connecting different brain regions, and disruption of those connections can have outsized cognitive consequences.

Watershed Injury in Newborns

Watershed-pattern brain injury is not limited to adults with atherosclerosis. It is one of the two main patterns seen in newborns who experience oxygen deprivation around the time of birth. In neonates with mild to moderate brain injury from oxygen deprivation, the damage tends to concentrate in the deep white matter along the border zones between arterial territories, following the same watershed logic seen in adults.¹³ Only in severe cases does the injury spread to deeper structures like the basal ganglia and thalamus.

The pattern carries prognostic significance. A prospective study of term infants with brain injury from birth found that the severity of watershed-pattern damage on early MRI was specifically associated with lower verbal IQ when the children were tested at age four. When researchers accounted for both watershed and deep-structure injury patterns, only the watershed pattern maintained a significant link to verbal IQ.¹⁴ This makes early MRI after birth-related brain injury more than just diagnostic: it offers parents and clinicians a way to anticipate which developmental domains might need extra support.

Watershed Infarction in the Spinal Cord and the Eye

The brain is not the only organ with vulnerable border zones. The spinal cord has its own watershed areas, and they follow the same logic: regions where blood supply from different arterial feeders barely overlaps are the most susceptible to ischemia. The mid-thoracic spinal cord, roughly between the fourth and sixth thoracic vertebrae, is the best-known spinal watershed zone. This area receives the tail end of supply from both upper and lower arterial feeders, making it especially vulnerable during periods of low blood flow, such as aortic surgery.¹⁵ Less common watershed zones exist lower in the spine. When spinal cord watershed infarction occurs, it can cause sudden weakness in the legs, loss of sensation, or bowel and bladder dysfunction, depending on the level affected.

The retina has its own arterial supply with interarterial border zones, and it too can suffer watershed-type infarction. In the eye, this happens most often in the context of carotid artery disease or other conditions that reduce blood flow to the ophthalmic artery. Researchers have described a pattern of subtle retinal damage at the junctions between the retinal arteries’ territories, analogous to what happens in the brain.¹⁶ In one reported case, a patient with severe internal carotid stenosis developed a watershed stroke and then, a month later, experienced severe vision loss in the eye on the same side. Placing a stent to open the narrowed artery restored blood flow to the eye and improved vision immediately.¹⁷ The takeaway is that the same carotid disease fueling a watershed stroke in the brain can simultaneously threaten the eye, and visual symptoms should be taken seriously as a signal of the same underlying problem.

Ischemic disorders of the retina and optic nerve are a common cause of vision loss in people with atherosclerosis, reinforcing the idea that watershed vulnerability extends well beyond the brain itself.¹⁸ Wherever in the body you find border zones between arterial territories, you find tissue at heightened risk when perfusion drops.