Story
Neurovascular coupling
Last updated
In one pass The brain keeps almost no energy in reserve, yet it never stops spending energy.
Educational content, not medical advice. See a clinician about your own case.
Story path
Chapter 1
Blood follows nerve firing
The brain keeps almost no energy in reserve, yet it never stops spending energy. Every time a neuron fires, ion pumps in its membrane burn (the cell's energy currency) to push the scrambled ions back into place. So sugar and oxygen have to be delivered as they are used. The body does not turn up the blood supply to the whole head. It opens the vessels next to the small patch that is working right now and lets blood flow catch up. This is called neurovascular coupling.
The signal to open the vessels is mostly sent ahead of need. As a synapse releases glutamate (the most common excitatory signal between neurons), nearby neurons and glial cells (brain cells that do not fire and instead look after neurons) start releasing molecules that relax the vessel wall. They do not wait for fuel to run low. Blood flow rises, bringing sugar and oxygen and carrying away waste such as carbon dioxide.
So the bright patch on a brain-scan color map is that surge of blood, not the firing itself. The electrical event is over within milliseconds; the blood arrives a beat later. A bright patch means the supply has caught up with the activity, not that the brain is congested or injured.
The signal to open the vessels is mostly sent ahead of need. As a synapse releases glutamate (the most common excitatory signal between neurons), nearby neurons and glial cells (brain cells that do not fire and instead look after neurons) start releasing molecules that relax the vessel wall. They do not wait for fuel to run low. Blood flow rises, bringing sugar and oxygen and carrying away waste such as carbon dioxide.
So the bright patch on a brain-scan color map is that surge of blood, not the firing itself. The electrical event is over within milliseconds; the blood arrives a beat later. A bright patch means the supply has caught up with the activity, not that the brain is congested or injured.
Mechanism · Why blood goes only to the working patch
If the brain were a sponge soaking in blood, it would not matter which cluster of neurons was firing; sugar would seep in anyway. That is not what happens. Blood flow follows activity. By this mechanism, when you concentrate on a sum, your whole head does not get more blood at once; the supply is turned up next to the pathway doing the calculation, and it falls back when that pathway goes quiet. A 2017 review by Iadecola attributes this tight matching to the brain's lack of energy reserves: oxygen and sugar have to arrive on time, and only where the work is.How do the vessels know where the work is? The old explanation was negative feedback: neurons use up their fuel first, oxygen or glucose drops or carbon dioxide rises, and only then are the vessels forced open. A 2010 review by Attwell and colleagues argues that the main mechanism is feedforward instead: the signal goes out along with the neural activity, without waiting for a shortfall. Glutamate makes neurons produce nitric oxide () and makes astrocytes (glial cells that wrap their processes around blood vessels) produce prostaglandins and related molecules (local signals made from fatty acids in the cell membrane), and nearby small arteries relax. In animal experiments, blocking the enzymes that make these molecules sharply shrinks the rise in blood flow while leaving the neurons' energy use almost unchanged.
Fuel signals are not useless. Adenosine, left over when is broken down, does contribute to the rise in blood flow. But a local rise in carbon dioxide does not look like the trigger; the extra blood flow carries it away quickly. Iadecola's review describes such metabolic signals as a possible second layer that follows and fine-tunes the response.
One point is still open: blood flow rises by more than oxygen use does. Attwell's review notes that blocking most of the extra blood flow with drugs barely changes the rise in oxygen consumption, so normal activity does not seem to need that much extra delivery. Why the brain sends it anyway is still only a hypothesis. That surplus of blood is exactly what lets a brain scan see activity.
Most of these mechanisms were worked out in animal brain slices, retinas and live animals; testing them one by one in the human brain is hard. Read this as a model that makes sense and has animal experiments behind it.
Chapter 2
Which cells widen the vessels
Neurons do not open or close blood vessels themselves. Their neighbors do the work.
Astrocytes are glial cells that look after neurons. Each one sends out long, thin processes whose ends wrap around blood vessels. When a neuron releases glutamate, the astrocyte picks it up first and then releases molecules that relax the vessel (and sometimes tighten it). Pericytes sit on the walls of capillaries and can narrow or relax them, which means flow may be set not only by the small arteries upstream but also at the capillary level. The endothelial cells lining the vessel pass the local signal back upstream along the vessel, so that larger vessels join in.
Iadecola calls neurons, glia, pericytes and vessels together the neurovascular unit: not neighbors each doing their own job, but parts of one device that matches blood supply to demand. How much of the work falls to astrocytes and how much to pericytes, the same review notes, is still much debated.
Astrocytes are glial cells that look after neurons. Each one sends out long, thin processes whose ends wrap around blood vessels. When a neuron releases glutamate, the astrocyte picks it up first and then releases molecules that relax the vessel (and sometimes tighten it). Pericytes sit on the walls of capillaries and can narrow or relax them, which means flow may be set not only by the small arteries upstream but also at the capillary level. The endothelial cells lining the vessel pass the local signal back upstream along the vessel, so that larger vessels join in.
Iadecola calls neurons, glia, pericytes and vessels together the neurovascular unit: not neighbors each doing their own job, but parts of one device that matches blood supply to demand. How much of the work falls to astrocytes and how much to pericytes, the same review notes, is still much debated.
Myth · Not the same chain as migraine
A migraine attack also involves a chain of nerves and vessels. Endings of the trigeminal nerve (the nerve that carries sensation from the face and the meninges) release CGRP (calcitonin gene-related peptide, a potent vessel-widening signal); vessels in the meninges widen, local inflammation follows, and the pain signal travels on into the brain. In that chain, the widening of vessels looks more like a result than the cause. That is a pain chain of nerves and vessels.Coupling is a different job: a neuron at work in the cerebral cortex asking the small vessels right next to it for sugar and oxygen. It happens inside brain tissue, not in the meninges that wrap the brain, and the main known messengers are nitric oxide and prostaglandin-type molecules triggered by glutamate, not CGRP. Two sets of vessels and two sets of messengers: do not merge them into one sentence that "blood vessels in the brain dilate."
Chapter 3
Brain scans show blood, not signals
The BOLD signal that functional MRI (fMRI) most often uses does not measure the neurons' electricity. It measures blood. It is sensitive to deoxyhemoglobin, hemoglobin that has given up its oxygen. When neurons get to work, the fresh blood sent to them often exceeds the oxygen they use at that moment, so local deoxyhemoglobin actually falls, the signal gets stronger, and the scanner paints the patch bright.
So that color map is a blood-supply map, and inferring neural activity from it takes one more step. The electrical event has already happened while the blood is still on its way. The scanner is not making a mistake; the lag is part of coupling itself.
So that color map is a blood-supply map, and inferring neural activity from it takes one more step. The electrical event has already happened while the blood is still on its way. The scanner is not making a mistake; the lag is part of coupling itself.
Mechanism · What a bright patch actually tracks
BOLD stands for blood-oxygen-level dependent. The scanner reads a magnetic resonance signal given off by hydrogen nuclei in the body, and deoxyhemoglobin is paramagnetic: it makes the local magnetic field less uniform and weakens that signal (Attwell 2010). If blood flow did not rise, the oxygen the neurons burn would raise deoxyhemoglobin and the patch should go darker. It goes brighter instead, because the blood flow that activity brings in delivers excess oxygenated blood. Roughly, brightness is the difference between two forces: the extra blood makes it brighter, and the oxygen the neurons use makes it darker.So which kind of activity lights up? Because blood flow is opened mainly by signals triggered by glutamate, not forced open by energy use, Attwell's review argues that BOLD mainly reflects the vascular signals triggered by synaptic activity, not how much energy the neurons spend. In the cortex, BOLD correlates slightly better with field potentials, which reflect synaptic currents, than with action potentials (the moment a neuron actually fires), though it still tracks action potentials well because the two run together in the cortex.
So keep two layers in mind when you read the picture. Bright means synaptic activity there drew in extra blood. It does not tell you directly how much more the neurons there fired, and it does not mean that brain tissue is straining or injured.
Chapter 4
When stiff vessels respond slowly
When small arteries carry too-high blood pressure for years, their walls thicken and stiffen. The tap is still there, but it turns stiffly: when a patch needs more blood, the vessels open late and not far enough, and the supply falls behind what that patch needs.
Blood flow has not stopped, so nothing happens suddenly the way it does in a stroke. A 2017 review by Iadecola describes a different outcome: when supply stays low, or cannot keep up with what the tissue needs, slow, cumulative damage builds up in vulnerable areas, and it often comes with cognitive decline.
So by this mechanism, keeping blood pressure under control also protects this matching. That step is reasoning, though; no randomized trial has tested it directly.
Blood flow has not stopped, so nothing happens suddenly the way it does in a stroke. A 2017 review by Iadecola describes a different outcome: when supply stays low, or cannot keep up with what the tissue needs, slow, cumulative damage builds up in vulnerable areas, and it often comes with cognitive decline.
So by this mechanism, keeping blood pressure under control also protects this matching. That step is reasoning, though; no randomized trial has tested it directly.
Mechanism · Why white matter feels it first
White matter, deep inside the brain, is made of bundled nerve fibers, each wrapped in an insulating layer called myelin, like a wiring district full of cables. It sits at the far end of the blood supply, and keeping myelin intact costs a lot of energy, so it is especially sensitive to a shortage of blood. By this mechanism, when supply falls behind, the first thing to suffer is how fast signals travel along those cables. The story on cognitive aging makes the point that what slows is white-matter insulation; one road into that point starts here.This chain mixes what has been observed with what is inferred, so keep them apart. In Iadecola's review, the evidence that high blood pressure impairs neurovascular coupling comes mainly from animal models of chronic hypertension. That small-vessel disease is a major vascular cause of cognitive impairment is something seen in patients. The review calls looking after brain vessels helps look after the brain an emerging idea: falling dementia rates in the population may partly reflect better cardiovascular health, but the review also says this hypothesis still needs testing in randomized trials.
The one thing to take away: among the reasons to keep blood pressure under control, the brain probably counts as one, but that part of the account has not been settled.
References · 2
- Attwell, D., Buchan, A. M., Charpak, S., Lauritzen, M., MacVicar, B. A., & Newman, E. A. (2010). Glial and neuronal control of brain blood flow. Nature, 468(7321), 232-243. 10.1038/nature09613
- Iadecola, C. (2017). The neurovascular unit coming of age: a journey through neurovascular coupling in health and disease. Neuron, 96(1), 17-42. 10.1016/j.neuron.2017.07.030
Written from the sources listed above. Checked sentence by sentence against them in . No physician reviews this site: it explains how the body works, and it is not medical advice. How we make content
Spotted an error, or something unclear? Email us. We read every message, check it against the sources, and fix what's wrong.