How the Heartbeat in the Brain Shapes Perception

Every time your heart contracts, a pressure wave travels up through your arteries and into your skull, physically moving your brain tissue, pushing cerebrospinal fluid through tiny channels, and triggering electrical responses in your cortex. This is not a metaphor. Your brain expands and contracts with each heartbeat, and that rhythmic pulse does far more than deliver oxygen. It shapes how you detect a faint touch, how intensely you perceive a threat, and even when you choose to take in new information. The relationship between heartbeat and brain function is one of the more surprising stories in modern neuroscience, and it touches nearly every domain of cognition and clinical medicine.

Your Brain Physically Pulses With Each Heartbeat

When your heart beats, the surge of blood into the brain’s arteries causes brain tissue to expand slightly, then recoil as the pulse passes. Specialized MRI techniques can measure this motion directly. Using a method called Displacement Encoding with Stimulated Echoes (DENSE), researchers have found that the central brain expands with each arterial pulse to strain values on the order of a few ten-thousandths of its volume.¹ That sounds tiny, but it translates to a total brain volume change of roughly half a milliliter per heartbeat.² Grey matter deforms more than white matter, likely because grey matter is more densely packed with blood vessels and receives more of the pulse pressure.

The brain accommodates this rhythmic swelling without dangerous pressure spikes thanks to a clever hydraulic arrangement. As brain tissue expands, veins in the skull compress slightly and cerebrospinal fluid (CSF) shifts out of the cranial cavity and into the spinal canal. The volume of CSF displaced closely matches the volume of brain expansion, keeping intracranial pressure roughly constant across the cardiac cycle.³ If you could watch the brain’s ventricles in slow motion, you would see them flex inward and outward in a piston-like rhythm synchronized to the heart.

A technique called amplified MRI (aMRI) has made this motion visible in striking detail. By magnifying the tiny spatial changes in cardiac-gated brain scans, aMRI reveals deformations throughout the brain, with especially prominent movement in the brainstem, cerebellum, and spinal cord.⁴ A 3D version of the technique captures the complete tissue and fluid motion, showing how surrounding tissue squeezes the ventricles in a rhythmic cycle.⁵ Videos produced with aMRI are genuinely strange to watch: the brain looks like it is breathing.

Cardiac Pulsation as the Brain’s Fluid Pump

The pulsing motion of arteries inside the brain does more than push tissue around. It acts as a pump for cerebrospinal fluid in the narrow spaces that surround blood vessels, called perivascular spaces. Research in mice using two-photon imaging showed that the peak velocity of artery-wall expansion closely matched the peak velocity of nearby CSF flow, occurring about 35 milliseconds after the heartbeat’s electrical peak. That timing strongly suggests arterial pulsation is the main force driving perivascular CSF movement.⁶ The same study found that this fluid flow was reduced in hypertensive animals, offering one possible link between high blood pressure and brain health.

Human imaging has confirmed a similar relationship. When brain arteries dilate during the cardiac cycle, perivascular CSF flow speeds up both along the artery and perpendicular to its wall.⁷ This matters because perivascular fluid flow is thought to be part of the brain’s waste-clearance system, sometimes called the glymphatic system. The heartbeat, in other words, may serve as the engine that helps flush metabolic byproducts out of brain tissue.

Sleep plays an important role here. Research on sleep deprivation and sleep intensity found that homeostatic sleep mechanisms are closely linked to brain fluid dynamics and potentially also to waste clearance.⁸ During deep sleep, cardiac-driven pulsations interact with slower brain rhythms in ways that appear to enhance fluid movement. This is one reason sleep deprivation may not just leave you tired but could also impair the brain’s ability to clean itself.

The Brain’s Electrical Reply to Each Heartbeat

Beyond the physical pulse, your brain generates a distinct electrical response to every heartbeat. This signal, called the heartbeat-evoked potential (HEP), shows up on EEG recordings as a small voltage shift time-locked to each cardiac contraction.⁹ You are almost certainly not aware of it, but your brain is continuously registering the fact that your heart just beat.

Where does this signal come from? An intracranial EEG study, where electrodes were placed directly on the brain surface, found that the insula and surrounding operculum were the main sources of HEP activity, along with contributions from the amygdala and parts of the frontal and temporal cortex. The mechanism appears to involve phase resetting: the heartbeat signal does not add new electrical activity so much as it synchronizes the timing of ongoing brain oscillations, nudging them into alignment with the cardiac rhythm.¹⁰ The same study found that HEP changes in the insula reflected alterations in participants’ sense of self-identification during an out-of-body illusion experiment, hinting that cardiac signals feed into our basic sense of being a self in a body.

The HEP is not static. Paying attention to your heartbeat increases its amplitude, and this increase is strongest during exhalation. A study that had participants count their heartbeats found that the HEP grew larger specifically when people were attending to cardiac sensations during the exhale phase of breathing, with minimal change during inhalation.¹¹ Breathing and heartbeat awareness, then, are not independent channels. They interact in the brain’s processing of body signals.

How the Cardiac Cycle Filters What You Perceive

One of the more counterintuitive findings in this area is that your ability to detect faint sensory stimuli fluctuates across the cardiac cycle. The two relevant phases are systole (when the heart contracts and pressure rises) and diastole (when the heart relaxes and pressure falls). Across multiple studies, touch detection has consistently been found to be more sensitive during diastole. A study measuring somatosensory detection found that sensitivity was higher during diastole than systole.¹² Reaction times to sensory stimuli are also faster during diastole, and the brain’s ability to integrate multisensory inputs (particularly those involving touch) is enhanced in that relaxed phase of the heartbeat.¹³

Fear works differently. Fearful faces shown during systole were detected more easily and rated as more emotionally intense than the same faces shown during diastole. Brain imaging showed that the amygdala responded more strongly to fear stimuli timed to systole.¹⁴ This selective enhancement of threat detection during the contraction phase is thought to be driven by baroreceptors, pressure sensors in the arteries that fire most vigorously during systole. The baroreceptor signal appears to boost the brain’s processing of fear while simultaneously dampening processing of other types of stimuli, including physical pain.¹⁵

The pain-dampening effect has been confirmed directly. Cortical responses to painful laser stimuli are smaller during systole than during diastole, consistent with the idea that baroreceptor activation mid-heartbeat briefly suppresses nociceptive processing.¹⁶ A separate study found that pain intensity and unpleasantness ratings were both lower during diastole, and the accompanying brain potentials were reduced as well.¹⁷ The picture that emerges is that each heartbeat briefly reshuffles the brain’s sensory priorities: threat detection goes up, routine sensory processing and pain go down, then everything resets before the next beat.

Motor Readiness and the Cardiac Cycle

The heartbeat’s influence extends into the motor system. When researchers delivered transcranial magnetic stimulation (TMS) pulses to the motor cortex at different points in the cardiac cycle, the resulting muscle twitches were larger during systole than diastole. The brain’s cortical excitability, measured independently from the spinal cord’s contribution, was also higher during systole.¹⁸ Your motor system, in other words, is slightly more primed to act during each heartbeat. This makes intuitive sense alongside the threat-detection boost: if your brain is momentarily better at spotting danger during systole, it helps to also be fractionally readier to move.

A follow-up study using input-output curves to map corticospinal excitability in more detail found a somewhat different pattern, observing greater inhibition during systole at high stimulation intensities and lower overall motor gain. That study also found that the strength of each individual heartbeat influenced motor responses on a beat-by-beat basis.¹⁹ The apparent contradiction likely reflects the complexity of how baroreceptor signals propagate through different layers of the motor system, but the core finding is consistent: cardiac phase and motor readiness are linked.

Memory, Attention, and the Rhythm of Information Intake

Given that perception shifts across the cardiac cycle, it would be surprising if memory were completely untouched. Research on emotional face encoding found that memory for faces, both their identity and their expression, was better when the faces were presented early in the cardiac cycle (well before the baroreceptor burst) than when they were presented during peak baroreceptor firing.²⁰ This suggests that the same baroreceptor-mediated dampening that suppresses pain during systole also attenuates the brain’s ability to lay down new declarative memories.

Intriguingly, your brain seems to compensate for this automatically. A study tracking self-paced viewing of emotional pictures found that participants tended to initiate picture onsets during systole, not diastole, even though they had no conscious awareness of their cardiac timing.²¹ In that study, memory performance was not affected by whether a picture happened to land on systole or diastole, which the authors interpreted as evidence that the brain’s spontaneous timing of information sampling may itself be an adaptive strategy. You look at the world in a rhythm set partly by your heart, and your brain may have evolved to optimize that rhythm for perception and action even at the cost of some memory encoding.

Sensing Your Own Heart and What It Means for Mental Health

Some people are better than others at sensing their own heartbeat without taking their pulse. This ability, called interoceptive accuracy, is reflected in the amplitude of the HEP. People who score higher on heartbeat-counting tasks have larger HEPs, particularly at right-central scalp locations roughly 250 to 350 milliseconds after each heartbeat.²² Structural brain imaging adds further detail: heartbeat-counting accuracy correlates with grey matter volume in the frontal pole, while heartbeat-discrimination ability tracks with volume in the orbitofrontal cortex and posterior insula.²³

The insula, which is central to both HEP generation and interoceptive awareness, also shows up consistently in studies linking heartbeat awareness to anxiety. An fMRI study found that the left anterior insula, supplementary motor area, and middle cingulate gyrus were jointly activated during both cardiac interoception and anxiety-related attention tasks.²⁴ The overlap between brain regions processing heartbeat awareness and those processing anxious arousal may help explain why people with anxiety disorders often report heightened awareness of their own heartbeat.

In panic disorder, the HEP appears disrupted. Patients with panic disorder showed reduced HEP amplitudes at rest compared to healthy controls, alongside elevated beta-band activity suggesting cortical hyperarousal. Despite (or perhaps because of) these neural differences, panic disorder patients reported higher interoceptive attention scores on questionnaires.²⁵ In generalized anxiety disorder, a study using a drug that mimics adrenaline to increase heart rate found that patients had abnormally large HEP changes in response to the cardiac stimulation, and even during placebo infusions their baseline HEP amplitudes were higher than those of healthy controls.²⁶ The picture is not yet simple enough for clinical use, but it increasingly suggests that how the brain processes each heartbeat may be a measurable feature of anxiety disorders.

Epilepsy and the Risk of Sudden Death

The heart-brain connection has grave implications in epilepsy. Sudden unexpected death in epilepsy (SUDEP) is the leading cause of premature death among people with the condition, and evidence points to seizure-induced cardiovascular and respiratory failure as the immediate mechanism. Resting-state brain imaging has found altered connectivity between brain regions involved in cardiorespiratory control in people who later died of SUDEP, with insula connectivity differences inversely related to the time between the brain scan and death.²⁷ In other words, the more disrupted the insula’s connections were, the sooner the patient died. That raises the possibility that insula connectivity could serve as a biomarker for SUDEP risk.

More broadly, patients with refractory epilepsy show fundamentally altered heart-brain coupling even between seizures. In a study comparing epilepsy patients to healthy controls, heart-rate variability was positively correlated with cortical oscillatory power across all standard EEG frequency bands in epilepsy patients, the opposite of the pattern seen in controls. The patients also had higher HEP amplitudes.²⁸ During seizures themselves, abnormal cross-frequency coupling between respiratory and cardiac rhythms at elevated frequencies has been identified in SUDEP cases, suggesting that during fatal events, disordered breathing may drive dangerous cardiac acceleration through brainstem-mediated pathways.²⁹

When Stiff Arteries Deliver Too Much Pulse

If the heartbeat pulse in the brain is a normal and necessary feature, what happens when that pulse gets too strong? As arteries stiffen with age, the cushioning effect that normally softens the cardiac pressure wave is lost, allowing more of the raw pulsatile force to reach the brain’s delicate small vessels. Data from the Reykjavik Study, a large population-based cohort of older adults, found that marked aortic stiffening was associated with reduced buffering of the pulse wave at the carotid arteries, transmission of excessive flow pulsatility into the brain, microvascular structural damage, and lower cognitive performance across multiple domains.³⁰

This is not only a problem of old age. In the Framingham Heart Study, higher arterial stiffness in young adults was already associated with reduced white matter integrity in the corpus callosum and corona radiata, along with lower grey matter density in the thalamus.³¹ The ARIC Study, another large cohort, found that people in the highest quartile of arterial stiffness had greater white matter lesion burden, smaller total brain volumes, smaller volumes in Alzheimer disease signature regions, and lower scores on tests of executive function and general cognition.³² The implication is straightforward: a healthy heartbeat pulse in the brain requires healthy, elastic arteries to buffer it. When that buffer erodes, each heartbeat becomes a tiny hammer blow to the brain’s microvasculature.

The Heartbeat as Neuroimaging Noise and Signal

For brain researchers who use fMRI, the cardiac pulse has long been a nuisance. The blood-oxygen-level-dependent (BOLD) signal that fMRI measures changes not only with neural activity but also with cardiac and respiratory rhythms. In both human and primate studies, these cardiac artifacts appear in predictable brain regions and can distort estimates of which brain areas are functionally connected to each other. In macaque monkeys under anesthesia, where there are no task-related neural confounds, removing cardiac and respiratory artifacts reduced false correlations in the brain’s so-called resting-state network.³³ Any fMRI study that does not account for the cardiac cycle risks misattributing heartbeat-driven blood flow changes to neural communication.

But the same signal that is noise in one analysis becomes data in another. When fMRI scans are deliberately aligned to the cardiac pulse rather than treated as continuous streams, they reveal cardiac-gated brain responses that track the pulsatile nature of blood flow in real time.³⁴ Combined with techniques like aMRI, which amplifies the sub-voxel tissue motions that standard scans discard, these cardiac-aligned methods are opening up a new window on brain biomechanics.³⁵ What used to be an artifact to scrub from data is increasingly treated as a physiological signal worth studying on its own terms, one that could eventually help detect conditions like hydrocephalus, Chiari malformation, or small-vessel disease by revealing how abnormally the brain pulses.