What Is Cortical Spreading Depression?

Cortical spreading depression is a slow-moving wave of intense electrical and chemical disruption that rolls across the brain’s surface, silencing neural activity in its wake. It travels at roughly two to five millimeters per minute, accompanied by massive shifts in the ions that neurons depend on to function and by swelling of brain cells.¹ First described in the 1940s by the Brazilian neurophysiologist Aristides Leão, the phenomenon has since been linked to migraine aura, stroke expansion, traumatic brain injury, and even the final electrical event in dying brains.

How the Wave Works

Under normal conditions, neurons maintain a careful balance of charged particles across their membranes. Potassium stays mostly inside cells, while sodium and calcium stay mostly outside. Cortical spreading depression (CSD) begins when something tips that balance past a threshold. A massive depolarization sweeps outward from the trigger point: potassium floods out of neurons into the surrounding space, while sodium and calcium rush in. This ionic avalanche is self-propagating, meaning that the disruption at one spot raises the extracellular potassium concentration enough to destabilize neighboring neurons, which then undergo the same collapse and pass it along.

Computational modeling has identified two parallel mechanisms that keep the wave moving. One depends on a type of sodium channel (called a persistent sodium channel) and the diffusion of potassium through the extracellular space. The other depends on glutamate, the brain’s primary excitatory chemical messenger, activating a receptor known as the NMDA receptor. These two pathways operate simultaneously and reinforce each other. Even when one pathway is blocked in simulations, the other can sustain the wave, though not as robustly.² This dual mechanism helps explain why CSD has proven so difficult to prevent pharmacologically: blocking one route still leaves the other active.

Astrocytes, the star-shaped support cells that outnumber neurons in many brain regions, play a central role. During CSD, astrocyte endfeet swell dramatically, driven by potassium influx and abnormal calcium signaling. That swelling squeezes the perivascular spaces through which the brain normally clears waste, slowing what researchers call the glymphatic system. The resulting buildup of inflammatory molecules can make the brain more susceptible to further waves of CSD, creating a vicious cycle.³ A range of metabolic factors also influence how easily CSD can be triggered and how long each wave lasts, including glucose levels, genetic markers, and the activity of enzymes involved in the inflammatory response.⁴ Low blood sugar, for instance, does not make CSD easier to trigger, but it significantly prolongs each depolarization event once one gets started.⁵

The Migraine Connection

If you have ever experienced a migraine with aura, that slowly expanding arc of shimmering light or temporary blind spot in your visual field almost certainly corresponds to a wave of CSD sweeping across your visual cortex. The hypothesis dates back to the 1940s, and decades of animal research have built substantial indirect support for it. Human studies in patients with brain injuries have confirmed that CSD does occur in the living human brain, strengthening the case that it underlies migraine aura as well.⁶ Mathematical models that simulate how CSD propagates across the folded surface of the cortex reproduce the visual hallucinations migraine sufferers describe with striking accuracy.⁷

But aura itself is painless. The headache that typically follows is a separate problem, and CSD appears to cause that too, through a chain of events that takes several minutes to unfold. In animal experiments, a single wave of CSD roughly doubled the firing rate of pain-sensing nerve fibers in the meninges, the protective membranes surrounding the brain. That heightened firing persisted for an average of about 37 minutes when it eventually returned to baseline, and in many trials it was still elevated when recording stopped after more than an hour.⁸ Subsequent work showed that the activation does not stop at those peripheral nerve endings. CSD also triggers neurons deeper in the pain pathway, in the brainstem’s trigeminal nucleus, which is the relay station for head and face pain signals heading toward conscious perception.⁹ This two-step activation, first peripheral then central, helps explain why migraine headache typically begins 20 to 60 minutes after aura starts and can last for hours even after the aura resolves.

CSD in Stroke and Brain Injury

Outside of migraine, CSD takes on a far more dangerous character. In healthy brain tissue, a single wave of CSD is usually reversible. Neurons recover within minutes, and no lasting damage occurs. But in tissue that is already stressed, as in the border zones around a stroke or a traumatic brain lesion, CSD waves become a mechanism of injury rather than just a symptom of it.

After a stroke cuts off blood flow to a region of brain, the tissue immediately surrounding the core of dead cells exists in a precarious state. This “penumbra” still has some blood supply but not enough to function normally. Spreading depolarizations ripple outward from the dying core into this fragile zone, and each wave places enormous metabolic demands on cells that are already starved for oxygen and glucose. The result is that each depolarization pushes the penumbra closer to irreversible damage, effectively expanding the stroke.¹⁰ These peri-infarct depolarizations are not rare events. They occur spontaneously and repeatedly in the hours and days after stroke and have been shown to worsen tissue outcomes in both animal models and human patients.¹¹

In traumatic brain injury, the picture is similar. A prospective observational study found that spreading depolarizations were associated with unfavorable outcomes even after accounting for standard prognostic variables like injury severity and patient age.¹² A later, larger study concluded that depolarizations are not merely a marker of how badly the brain was hurt. They are an independent mechanism of secondary injury, meaning they actively make things worse after the initial trauma.¹³

In subarachnoid hemorrhage, where bleeding occurs in the space surrounding the brain, CSD may be even more consequential. Delayed cerebral ischemia, a feared complication that can strike days after the initial bleed, has traditionally been blamed on vasospasm (narrowing of blood vessels). More recent evidence suggests that spreading depolarizations may be the primary culprits, occurring even in healthy-looking brain tissue without detectable vasospasm and triggering a cascade of inflammation, tiny blood clots, and vessel constriction.¹⁴

What CSD Does to Blood Flow

One of the reasons CSD can be either harmless or devastating lies in how it interacts with blood vessels. In a healthy brain, CSD triggers a surge of blood flow (hyperemia) that delivers the extra oxygen and glucose neurons need to restore their ionic gradients and recover. But in an injured brain, the opposite can happen. Instead of more blood, the wave triggers vasoconstriction and reduced flow, a pattern called inverse neurovascular coupling. Research in patients with severe brain trauma found that about a third of these pathological inverse responses occurred during periods when blood flow was already critically low, piling further oxygen deprivation on top of the metabolic crisis the depolarization wave was already causing.¹⁵ This flipped vascular response was seen specifically when recording electrodes sat in or near tissue that was actively deteriorating, suggesting that the local state of the tissue determines whether CSD is tolerable or catastrophic.

At the structural level, the damage can be strikingly rapid. Two-photon microscopy of neurons in the penumbra around an experimental stroke revealed that dendrites, the branch-like projections neurons use to receive signals, developed bead-like swellings within seconds of each passing depolarization wave. They recovered within a few minutes between waves, but eventually a single wave would cause terminal dendritic injury from which no recovery occurred.¹⁶ Each wave essentially rolls the dice on whether those synaptic connections survive.

Terminal Spreading Depolarization and the Dying Brain

Perhaps the most philosophically striking appearance of this phenomenon occurs at the moment of death. When the heart stops and blood flow ceases, the brain’s electrical activity falls silent within seconds as neurons can no longer fire. But the cells are still alive at this point, slowly depleting their last reserves of energy. Minutes later, a final wave of spreading depolarization sweeps across the cortex. In a study of nine patients who had monitoring electrodes in place when life support was withdrawn, this terminal spreading depolarization appeared between 13 and 266 seconds after electrical silence began.¹⁷

This terminal wave marks the point at which neurons completely lose their membrane potential and their internal chemistry becomes indistinguishable from the fluid outside. It is, in a sense, the cellular moment of death. Whether it is immediately irreversible depends on how quickly blood flow could hypothetically be restored. In the minutes before the terminal wave, the changes are still potentially reversible if circulation returns. After it, recovery becomes far less likely. This finding has implications for resuscitation science and for understanding what happens during near-death experiences, though research in that area remains early.

How Clinicians Detect Spreading Depolarizations

Detecting CSD in patients requires specialized equipment. The gold standard is subdural electrocorticography, in which electrode strips are placed directly on the brain’s surface during neurosurgery. Standard hospital EEG systems filter out the slow voltage shifts that define spreading depolarizations, so recordings must use direct current-coupled amplifiers that can capture the full signal bandwidth.¹⁸ In practice, this monitoring is performed at the bedside in neurointensive care units and can run continuously for days. One study using less-invasive electrode arrays in subarachnoid hemorrhage patients recorded 84 spreading depolarizations over a cumulative monitoring period of nearly 1,200 hours.¹⁹

A multinational research consortium called COSBID (Cooperative Study on Brain Injury Depolarizations) has published standardized recommendations for recording and interpreting these signals. Spreading depolarizations are increasingly treated as a causal biomarker of metabolic failure and excitotoxic injury, meaning they do not just predict bad outcomes but actually reflect the ongoing process that produces them.²⁰

The obvious limitation is that subdural electrodes require a craniotomy, restricting this monitoring to patients who already need brain surgery. Researchers are working on non-invasive alternatives. A recent proof-of-concept study demonstrated that scalp-based EEG with high spatial resolution and direct current coupling could detect spreading depolarizations without placing anything inside the skull.²¹ MRI-based approaches are also being explored, with spin-lock functional MRI showing promise for imaging the metabolic transitions associated with CSD.²² If non-invasive detection becomes reliable, it could open the door to monitoring for CSD in stroke patients, migraine patients, and others who would never undergo brain surgery solely for electrode placement.

Ketamine as a Pharmacological Brake

Because one of CSD’s two propagation mechanisms depends on NMDA receptors, drugs that block those receptors are natural candidates for treatment. Ketamine, widely used as an anesthetic and sedative in intensive care, has emerged as the most studied option. A prospective crossover trial in patients with acute brain injury found a strong dose-dependent effect: hours without ketamine or on low doses were associated with dramatically higher odds of spreading depolarizations compared to hours on higher doses. Above a threshold of roughly 1.15 mg/kg per hour, depolarizations became rare.²³

A retrospective study in subarachnoid hemorrhage patients found similar results with s-ketamine, the more potent mirror-image form of the molecule. Starting s-ketamine was associated with an 87% reduction in the odds of spreading depolarizations, and higher doses produced further suppression. Importantly, the drug continued to work over at least five days of continuous use without obvious tolerance developing.²⁴ These findings are encouraging but still preliminary. No large randomized trial has yet proven that suppressing spreading depolarizations with ketamine translates into better patient outcomes, though that is an active area of investigation.

CSD Beyond the Human Brain

Spreading depolarization is not unique to mammals or even to vertebrates. Insects experience the same fundamental phenomenon when environmental stresses such as oxygen deprivation overwhelm the ability of their central nervous system to maintain ionic balance. In locusts, for example, anoxia triggers spreading depolarization marked by sudden potassium loss, leading to the neural shutdown that underlies coma.²⁵ The fact that such distantly related nervous systems share this failure mode suggests that spreading depolarization is a fundamental property of excitable tissue rather than a quirk of cortical architecture.

This universality has practical research implications. The chicken retina has long served as a convenient experimental model because it can be kept alive in a dish for hours to days and its tissue is transparent enough that the wave of spreading depression can literally be seen with the naked eye as it passes through. The electrical characteristics of retinal spreading depression are identical to those measured in the cortex, but the optical signal makes experimentation far simpler, particularly for screening potential drugs.²⁶ For a phenomenon that is otherwise difficult to study without invasive brain electrodes, having a preparation where you can watch the wave move in real time is a genuine advantage for pharmacological screening and basic biophysics alike.