What Happens in Your Brain During a Panic Attack

A panic attack is the brain’s threat system firing without a threat. Brainstem sensors, the amygdala and the insula help set off a surge of adrenaline that speeds the heart and breath, while the prefrontal cortex struggles to veto the alarm. The episode peaks within minutes, then fades, leaving exhaustion behind.

The people living through one rarely describe it in neurological terms. They describe a heart attack, a suffocation, a loss of sanity.

The clinical definition is an abrupt surge of intense fear or discomfort that peaks within minutes and includes at least four of 13 symptoms, a list that runs from pounding heart, sweating and trembling to chest pain, tingling, feelings of unreality and the fear of dying.

The experience is also more common than the label suggests. In the US National Comorbidity Survey Replication, 22.7 percent of respondents reported isolated panic attacks at some point in life, while 3.7 percent met criteria for panic disorder without agoraphobia and 1.1 percent for panic disorder with it.

A World Mental Health Surveys analysis across 25 countries put lifetime prevalence at 13.2 percent for panic attacks and 1.7 percent for panic disorder. The gap between those figures is the real story.

Many people have just one or two attacks in a lifetime, and the problem goes away. Panic disorder is a separate matter, defined less by the attack than by what follows it.

The Alarm Chain: How the Brain Builds a Panic Attack

Panic starts in a circuit, not a thought. The circuitry implicated in panic disorder includes the amygdala, thalamus, hippocampus, insula, locus coeruleus, periaqueductal gray, anterior cingulate cortex and prefrontal regions.

Animal research has been central to mapping it, because the defensive responses of animals closely resemble the symptoms of a human panic attack.

In a genuine emergency, heart rate and breathing accelerate so the body can fight or flee. Many of the same reactions occur in a panic attack, yet it is unknown why they occur when no obvious danger is present. That unanswered question drives most of the research.

The physical symptoms follow predictable biology. Adrenaline produces the pounding heart, sweating and trembling. Rapid breathing lowers carbon dioxide in the blood, which commonly produces the dizziness, lightheadedness and tingling in the hands and face.

Derealization, the sense that surroundings have turned flat or distant, remains among the least understood of the 13 symptoms.

The Amygdala Is Not the Whole Story

Popular accounts call the amygdala the brain’s fear center, and the imaging evidence gives that account some support. A 2025 systematic review highlighted amygdala activity in panic disorder patients, and also found higher activation in the hippocampus, brainstem and cingulate cortices.

A simple “panic amygdala” model has not survived, however. It lost ground after studies found that patients without an amygdala developed panic attacks.

The most striking evidence came from the University of Iowa, where Justin Feinstein, John Wemmie and colleagues studied three women with Urbach-Wiethe disease, a rare genetic condition that destroys the amygdala. One of them, known as S.M., showed only minimal fear when shown horror clips or exposed to large spiders and snakes.

The team then had the women inhale 35 percent carbon dioxide. The patients gasped for air, their heart rates shot up, and they tried to rip off their masks; only three of 12 healthy participants panicked, and none of the three patients had a history of panic attacks. The result suggests that fear triggered by external threats differs from fear triggered internally by CO2.

The Iowa team also proposed that an intact amygdala may normally inhibit panic. Three patients is a small sample, and the inhibition idea remains a hypothesis. It still changes the picture: the amygdala handles threats arriving through the senses, while signals from inside the body can reach the brainstem and insula and produce terror without it.

The Suffocation False Alarm Hypothesis

The most influential biological explanation of spontaneous panic comes from psychiatrist

. His 1993 theory held that a physiologic misinterpretation by a suffocation monitor misfires an evolved suffocation alarm system, producing sudden respiratory distress followed by brief hyperventilation, panic and the urge to flee. In this view, people with panic disorder have a lower threshold for detecting carbon dioxide.

Challenge studies lend it support. In one study of 51 patients, 65 percent had a panic attack in response to a CO2 procedure, and the association was stronger in patients with prominent respiratory symptoms.

Maurice Preter and Klein later extended the theory, proposing that episodic dysfunction in endogenous opioid regulation, which co-regulates breathing and social-affiliative behavior, explains the alarm’s failure. The theory was built partly to explain attacks that arrive in sleep or during relaxation, when no external trigger exists.

A counterargument comes from cognitive models associated with David Clark, which hold that panic grows from catastrophic misreading of bodily sensations. The two camps are less opposed than they appear.

One researcher noted that the findings from CO2 studies are consistent with an alternative psychological explanation. A workable synthesis is that biology supplies the spark and interpretation supplies the fuel.

Why Reason Arrives Late

The prefrontal cortex and anterior cingulate normally help assess whether a signal is a real danger. In panic, that regulation fails to arrive in time.

The prevailing fear-network model holds that regulatory dysfunction at any key site in the network may lead to the initiation of panic disorder symptoms. Gorman and colleagues went further, suggesting that different treatments affect different parts of the brain as well as different symptoms of the illness.

Readers should hold these findings loosely. The 2025 systematic review of fear circuitry identified only 13 human studies and 3 animal studies, and imaging samples in this field tend to be small. The broad outline is credible. The fine detail is still contested.

A Four-Stage Model of a Panic Attack

The research can be organized into a practical sequence. This is an explanatory framework, not an official clinical model, but it maps closely to where treatment intervenes: detect, amplify, interpret, avoid.

Detection comes first. A body signal, often a breathing change, a skipped heartbeat or a rise in carbon dioxide, registers as a threat. Sometimes there is a trigger such as stress or caffeine, and sometimes there is none.

Amplification follows within seconds. Adrenaline, faster breathing and muscle tension create more sensations, and each new sensation confirms the alarm. This loop explains why attacks escalate so quickly.

Interpretation decides the intensity. A racing heart read as exertion or excitement subsides. The same heart read as a cardiac emergency feeds the loop, and fear of dying joins the symptom list.

Avoidance is where an attack becomes a disorder. One of the worst features of panic attacks is the intense fear of having another one, and many people avoid situations where one might occur.

The formal threshold reflects this: panic disorder requires at least one attack to be followed by a month of worry about further attacks or their implications, or by maladaptive changes in behavior.

How Long Does a Panic Attack Last?

Symptoms develop abruptly and peak in less than 10 minutes. The peak is brief, but the aftermath is not. Fatigue and a worn-out feeling often follow once the attack subsides.

Time distortion during an attack makes it feel far longer than the clock suggests, which is part of why people report attacks lasting an hour.

Panic Attack or Heart Attack

The overlap in symptoms sends many first-time sufferers to the emergency department, and that is reasonable. Panic attack symptoms can resemble those of a heart attack, so evaluation by a medical provider is recommended when the cause is unclear.

Clinicians look at pattern. Panic-related chest pain tends to be sharp or stabbing, lasts only seconds, stays in one small area and occurs at rest; heart attack pain tends to travel to the arms, jaw or back, and is often brought on by exertion.

One emergency physician at Cedars-Sinai said the best predictor is the patient’s age and previous history of panic attacks. A 25-year-old with a long history of identical episodes presents differently from a 55-year-old with new chest pressure. Anyone with new or unfamiliar symptoms should treat them as a possible cardiac event until a clinician says otherwise.

The reverse error is also costly. Repeated emergency visits for panic symptoms are a recognized pattern, and frequent medical care for health concerns is among the complications linked to panic attacks. Once the heart has been cleared, continued testing tends to reinforce the belief that something is wrong.

Who Develops Panic Disorder

Past-year prevalence among US adults is 2.7 percent, higher for women (3.8 percent) than for men (1.6 percent), with a lifetime figure of 4.7 percent. Heredity is estimated at roughly 43 percent. Other risk factors include major life changes, smoking and excessive caffeine intake.

The disorder rarely travels alone. About 80.4 percent of people with lifetime panic disorder had another lifetime mental disorder. It also disables: 44.8 percent of adults with panic disorder in the past year had serious impairment.

What the Evidence Says About Treatment

The treatment literature is unusually consistent. Cognitive behavior therapy can be used alone or with antidepressants, and benzodiazepines are less effective than antidepressants and CBT. That finding contradicts a common assumption that fast-acting sedatives are the best answer to a fast-moving crisis.

Panic-focused CBT, as tested in a Dutch trial, included interoceptive exposure, cognitive interventions and exposure in vivo. Interoceptive exposure means deliberately producing feared sensations in a controlled setting, such as spinning in a chair or breathing through a narrow straw, until the body learns the sensations are survivable. It trains the detection and interpretation stages directly.

Medication has a different profile. Antidepressants generally take 4 to 6 weeks to become effective for panic disorder. In the Dutch trial’s one-year follow-up, no fall-off in gains was observed for either modality after treatment discontinuation, and SSRIs were associated with adverse events.

Cost points the same direction: CBT was associated with the lowest societal costs, and CBT and CBT plus an SSRI were more cost-effective than an SSRI alone.

Choosing a Therapist and Managing Cost

The practical questions are how to find panic-focused CBT and how to afford it. A clinician should be able to describe interoceptive exposure and give a session estimate. Most of the trials behind the meta-analyses used eight to 15 sessions of CBT. Prospective patients should ask whether insurance covers the sessions and whether a sliding scale exists.

Internet-delivered CBT is a legitimate lower-cost option, with caveats. A meta-analysis of 27 studies found large effects against a waitlist (Hedges g of 1.16 for panic symptoms). In a psychiatric care setting, guided internet CBT was as effective as group CBT and more cost-effective with respect to therapist time.

One open trial found that therapist time was reduced by a factor of three, though attrition was about twice that of face-to-face treatment, at 46 percent. Therapist guidance appears to matter: a network meta-analysis found unguided internet CBT was not superior to care as usual.

Access remains the larger problem. About 65 percent of US adults with panic disorder used some service in the past year, yet roughly one in four received minimally adequate treatment.

Mistakes That Keep Panic Going

Avoidance is the most common. Skipping the elevator, the highway or the gym removes the chance to learn that the sensations pass, and the list of avoided places grows. Safety behaviors do the same work in quieter form: carrying medication that is never taken, sitting near exits, checking pulse repeatedly.

Treating a benzodiazepine as the cure is the second. It may calm one attack while leaving the underlying fear of sensations intact, and the evidence ranks it below the alternatives.

Ignoring stimulants is the third. Caffeine and nicotine produce the same palpitations the alarm system is primed to read as danger.

The fourth is assuming one attack equals a disorder. A single episode warrants a medical check, not a lifetime of vigilance.

Persistent attacks, or a month of dread about the next one, justify a clinical evaluation. The alarm is real, the danger usually is not, and the system that raises the alarm responds to training.

What People Ask