What Actually Happens During a Panic Attack

You're standing in line at the store, or sitting at your desk, or lying in bed at night. Nothing dangerous is happening. But suddenly your heart is slamming, your chest feels tight, and your hands start tingling. This is your brain's alarm system going off by mistake. A small part of your brain that watches for threats decided something was wrong, even though it wasn't, and it hit the emergency button. Your body doesn't know the difference. It responds the same way it would if you were in actual physical danger.

Once that alarm fires, your body floods with adrenaline. That's the chemical that prepares you to run or fight. Your heart speeds up to pump blood to your muscles. Blood pulls away from your hands and feet toward your core, which is why your fingers go tingly or numb and your face might flush. Your stomach shuts down because digestion isn't a priority when your body thinks you're in danger. Every one of these sensations has a purpose in a real emergency. They just feel terrifying when there's no emergency to match them.

Here's what helps to know: a panic attack puts about the same stress on your heart as a brisk walk. It feels catastrophic, but your heart can handle it. And the attack doesn't last. It peaks around ten minutes, then your body's calming system gradually takes over. Adrenaline gets used up. Your breathing slows. Your heart settles. The attack ends on its own. You don't have to figure out how to stop it. Your body already knows how.

When the alarm fires and adrenaline hits, your breathing speeds up. That part makes sense. But what happens next is what turns a scary few minutes into something that feels unbearable. When you breathe too fast, you blow off too much carbon dioxide. Your blood chemistry shifts, and your brain gets slightly less blood flow. That's when the really frightening symptoms show up: dizziness, a feeling like the room is pulling away from you, tunnel vision, lightheadedness. These sensations don't feel like anxiety. They feel like something is medically wrong.

And that's exactly what your brain decides. You're already afraid. Now the room looks strange and your head is spinning. So your brain says: it must be worse than I thought. Fear goes up. Breathing gets faster. More carbon dioxide leaves. More dizziness. More fear. It's a loop, and each trip around makes it worse. This is why panic attacks escalate. The breathing change creates symptoms that look like new problems, and your brain treats each new symptom as proof that the danger is real.

But here's the thing about that loop: it breaks on its own. Your body won't let carbon dioxide stay depleted forever. As the adrenaline wears off, your breathing naturally starts slowing. Carbon dioxide comes back to normal levels. The dizziness clears. The strange feelings fade. This is one of the bravest truths to hold onto when you're in the middle of an attack: it can't go on forever. Your body's own chemistry puts a time limit on it. The wave crests, and then it comes back down.

One panic attack is frightening. But what often happens afterward is what really changes things. Your brain, which just went through an overwhelming false alarm, starts keeping watch. It begins paying attention to your heartbeat, your breathing, any little sensation that might mean another attack is coming. You climb a flight of stairs and your heart speeds up, like it normally does, but now that faster heartbeat feels dangerous. A moment of dizziness from getting up too fast becomes a warning sign. Sensations you never noticed before are suddenly on high alert.

This is why some people have panic attacks that seem to come out of nowhere. The trigger isn't always something you can see or point to. Sometimes it's a tiny internal shift: your heart skips a beat, your breathing changes slightly, a wave of warmth washes through your chest. You might not even notice it consciously. But your brain does. It learned during the last attack that those sensations mean danger, so it sounds the alarm again. The panic attack doesn't arrive out of thin air. It's your brain responding to something inside your body that it now reads as a threat.

And here's the part that genuinely matters: what your brain learned, it can unlearn. The pattern of fearing your own body's signals is a kind of conditioning, like any other learned response. Evidence-based approaches work specifically on this, helping people face the feared sensations gradually, in safe settings, so the brain can recalibrate. That takes real work and usually professional support. But it happens every day. Understanding what's happening during a panic attack is a real first step. It takes some of the mystery out of the experience. And the courage to go further, to gently face the sensations instead of running from them, is what changes the pattern over time.

A panic attack begins with a misfire in your brain's threat-detection system. There's a small structure called the amygdala whose entire job is scanning for danger. When it detects a threat, it sends emergency signals to the rest of the body. The problem: it can't tell the difference between physical danger and the feeling that something is wrong. So it fires the same alarm for a crowded elevator as it would for a car accident. Within seconds, your adrenal glands release adrenaline and your body shifts into full emergency mode. Your heart rate jumps. Muscles tense. Blood redirects toward your core.

The physical sensations follow a predictable pattern, and each one has a mechanical cause. Adrenaline makes your heart pound. Vasoconstriction, the narrowing of blood vessels in your extremities, is why your hands go numb and tingly. Your chest muscles tighten to drive faster breathing. Your stomach shuts down because the body deprioritizes digestion during perceived emergencies. People who've had panic attacks often describe feeling certain they were having a heart attack. That fear makes perfect sense. The sensations are almost identical to what you'd expect from a cardiac event. But they're being produced by adrenaline, not by heart damage.

The reassuring part is that your heart can handle this. Researchers who monitored people during naturally occurring panic attacks found heart rate increases that were real but within the range your cardiovascular system manages during moderate exercise. Individual attacks don't injure heart tissue. And they don't last. The typical attack peaks in about ten minutes, then gradually resolves over twenty to thirty as adrenaline is metabolized and the body's calming system takes over. That said, repeated panic over many years can contribute to cardiovascular strain through chronic stress. A single attack is a fire drill. It's the drills that never stop that warrant attention.

When the initial adrenaline surge hits, breathing naturally speeds up. But faster breathing does something most people don't expect: it blows off too much carbon dioxide from the blood. Carbon dioxide isn't just a waste gas. It helps regulate blood acidity and blood vessel diameter. When CO2 drops, your blood becomes slightly more alkaline and the blood vessels in your brain constrict, reducing cerebral blood flow. That's when the second wave of symptoms hits: dizziness, lightheadedness, a feeling that the world looks flat or distant, tunnel vision. These feel nothing like anxiety. They feel neurological.

And your brain reads them that way. You're already in a state of high fear. Now new symptoms have appeared that don't match anything you were expecting. Your brain does what any threat system would do with alarming new data: it escalates. Fear rises. Breathing speeds up further. CO2 drops again. More dizziness. More strangeness. The feedback loop is what turns a ten-second alarm into a five- or ten-minute episode. Each trip around the loop adds symptoms that confirm the fear driving it.

But the loop can't sustain itself indefinitely. Your body has regulatory systems that pull breathing back toward baseline even when the alarm is still running. As adrenaline metabolizes, the brainstem's respiratory center recalibrates. CO2 levels normalize. Blood pH returns to range. The parasympathetic nervous system, your body's recovery mode, overrides the emergency state. This is one of the bravest pieces of knowledge to carry into a panic attack: it has a ceiling. The chemistry that drives it also limits it. The wave crests, and the body brings itself home.

A single panic attack is a crisis in the moment. What transforms it into a recurring problem is what the brain does with the memory. After that first overwhelming experience, the brain starts hypervigilant monitoring of body sensations. Every heartbeat fluctuation, every moment of lightheadedness, every warm flush becomes a potential signal. The staircase that raises your heart rate normally now feels ominous. Standing up too fast and getting dizzy becomes a warning instead of a nuisance. The brain has learned a new association: these sensations mean danger.

This explains attacks that seem to arrive without any trigger. Some panic attacks are clearly connected to specific situations. But others appear to come from nowhere. Research suggests these aren't truly random. The trigger is internal: a slight change in heart rhythm, a shift in breathing pattern, a subtle temperature fluctuation. These signals operate below conscious awareness, but the brain detects them and fires the full alarm anyway. The person experiences a panic attack with no visible cause, which makes it scarier. But the cause is there, happening inside the body at a level the conscious mind doesn't track.

The genuinely good news is that conditioning works in both directions. If your brain learned to associate normal body sensations with danger, it can learn to associate them with safety again. Targeted approaches work on this by gradually exposing people to the feared sensations in controlled, safe settings. A slightly elevated heart rate stops predicting catastrophe. Dizziness stops meaning disaster. This takes sustained effort, usually with professional guidance. But it happens. Understanding what your body does during a panic attack genuinely helps; it shrinks the catastrophic story your brain tells. And the courage to face those sensations rather than avoid them is what rewrites the pattern.

Gorman and colleagues mapped the neuroanatomy of panic in a model that remains influential. The amygdala sits at the center, receiving threat-relevant input and projecting to brainstem nuclei that each produce a specific symptom cluster. The locus coeruleus drives noradrenergic activation and heightened arousal. The parabrachial nucleus affects respiratory rate. The hypothalamus triggers the hormonal cascade through the HPA axis. The periaqueductal gray governs defensive behavioral responses. This distributed architecture explains why panic attacks produce such varied symptoms simultaneously. The racing heart, the breathing changes, the muscle tension, and the sense of dread aren't separate problems. They're outputs of a single alarm projecting through different neural pathways.

Charney's yohimbine challenge studies provided some of the earliest direct evidence for noradrenergic involvement. Yohimbine, which blocks alpha-2 adrenergic autoreceptors and increases norepinephrine release, provoked panic attacks in patients with panic disorder at significantly higher rates than in healthy controls. This vulnerability to noradrenergic surges appears to be a biological characteristic of panic-prone individuals, not simply a product of psychological expectation. Ehlers and colleagues corroborated the real-world cardiovascular impact with ambulatory monitoring, documenting heart rate increases of 8 to 33 bpm during naturally occurring panic episodes.

The cardiac safety question deserves careful treatment. Individual panic attacks produce hemodynamic stress comparable to moderate physical exercise, with no evidence of acute myocardial damage. But chronic panic disorder has been linked to modestly elevated cardiovascular risk over decades, likely mediated by sustained cortisol elevation, autonomic dysregulation, reduced physical activity from avoidance behaviors, and higher rates of smoking in panic populations. The accurate framing: any individual attack is a fire drill your body can handle. The concern is when the drills never stop and chronic stress pathways accumulate.

The respiratory component of panic has been studied with particular rigor because it's both measurable and experimentally reproducible. During panic, increased breathing rate and tidal volume blow off carbon dioxide faster than the body produces it. This hypocapnia shifts blood pH toward alkalosis and triggers mild cerebral vasoconstriction. The resulting reduction in cerebral blood flow produces the symptoms that often frighten people most: dizziness, depersonalization, derealization, visual disturbances, and paresthesias. Maddock and Carter demonstrated that voluntary hyperventilation in panic-prone participants reliably reproduced symptom profiles matching their naturally occurring attacks.

The feedback loop between breathing and fear is the central amplification mechanism. As hypocapnia-induced symptoms appear, the brain's threat evaluation system treats them as new evidence of danger. Sympathetic activation increases, breathing accelerates further, and CO2 drops again. Papp and colleagues found that panic disorder patients show lower resting end-tidal CO2 compared to controls, suggesting chronic mild hyperventilation that places them closer to the threshold where the loop engages. The margin between baseline and the tipping point is narrower, which helps explain why some patients escalate rapidly from the first sign of discomfort.

Klein proposed the suffocation false alarm theory as an alternative framework. A CO2-sensitive monitor in the brainstem, evolved to detect suffocation, is miscalibrated in panic disorder. Normal CO2 fluctuations trigger a suffocation alarm that initiates the respiratory and autonomic cascade. Gorman's CO2 inhalation studies, where 35% CO2 mixtures provoked panic at higher rates in patients than controls, provide evidence consistent with this model. The suffocation theory and the hyperventilation feedback model aren't mutually exclusive. One addresses initiation; the other addresses amplification. Both converge on a system where riding out the spiral, rather than fighting it, allows the body's own corrective mechanisms to resolve the episode.

Clark's cognitive model, published in 1986, reframed panic disorder around a single mechanism: catastrophic misinterpretation of bodily sensations. A person notices a normal physiological fluctuation, interprets it as dangerous, and the interpretation triggers anxiety that produces more sensations. The model explains both the onset and maintenance of panic disorder. But it has been criticized for overemphasizing conscious cognitive appraisal. Many panic attacks, particularly unexpected ones, occur without identifiable conscious misinterpretation. The person finds themselves in full panic with no mediating thought they can recall.

Bouton and colleagues addressed this gap with a learning-theory account grounded in interoceptive conditioning. Internal sensations present during a panic attack become conditioned stimuli through classical conditioning. After one or more panic experiences, these interoceptive cues can trigger a conditioned anxiety response without requiring conscious catastrophic thought. This explains unexpected panic: the trigger is a subtle physiological shift the brain has learned to associate with the full cascade. Reiss and McNally's concept of anxiety sensitivity captures the individual-difference dimension. People higher in anxiety sensitivity attend more closely to bodily sensations and react more fearfully, making the conditioning loop tighter and faster.

Barlow's triple vulnerability model integrates these threads. A generalized biological vulnerability creates the substrate. A generalized psychological vulnerability (learned sense that threats are uncontrollable) amplifies it. A specific psychological vulnerability (learned association between body sensations and danger) gives panic its particular form. This explains why not everyone who has a panic attack develops panic disorder. Treatment based on this model, particularly interoceptive exposure, targets the specific vulnerability directly: provoking feared sensations in safe contexts dissolves the conditioned association. Understanding the physiology strips panic of some of its power. Facing the sensations with courage strips it of the rest.

Gorman et al. (2000) proposed the most comprehensive neuroanatomical model of panic in the American Journal of Psychiatry. The amygdala receives cortical and thalamic threat-relevant input and projects to brainstem nuclei generating specific symptom clusters: the locus coeruleus (noradrenergic activation, heightened vigilance), the parabrachial nucleus (respiratory acceleration), the lateral hypothalamus (sympathetic activation via the HPA axis), and the periaqueductal gray (freezing and defensive behaviors). The model maps panic symptoms to discrete neural pathways rather than treating them as undifferentiated arousal, and accounts for the efficacy of both SSRIs (modulating serotonergic input to the amygdala) and CBT (strengthening prefrontal regulation of amygdala output).

Charney et al. (1984) provided direct pharmacological evidence for noradrenergic involvement by administering yohimbine, an alpha-2 adrenergic antagonist that increases norepinephrine firing from the locus coeruleus. Panic disorder patients experienced significantly more panic attacks, greater increases in plasma MHPG (a norepinephrine metabolite), and higher self-reported anxiety than healthy controls. This established that the noradrenergic system is functionally hypersensitive in panic-prone individuals. Ehlers et al. (1986), using 24-hour ambulatory ECG monitoring, documented heart rate increases of 8 to 33 bpm during naturally occurring attacks, confirming that subjective cardiac acceleration has an objective physiological correlate within normal exercise tolerance ranges.

Acute versus chronic cardiac risk requires separation. Individual attacks produce cardiovascular stress within moderate aerobic exercise range, with no evidence of acute myocardial damage. Celano et al. (2016) reviewed the longitudinal relationship between panic disorder and cardiovascular outcomes, finding modestly elevated coronary event risk over decades. Proposed mediators include chronic autonomic dysregulation (sustained sympathetic dominance reducing heart rate variability), elevated baseline cortisol, agoraphobic-related physical inactivity, and higher comorbid substance use. Craske et al. (2010), in their Annual Review of Clinical Psychology overview, emphasize these are population-level risk factors associated with the chronic disorder, not with individual episodes.

The respiratory physiology of panic is among the best-characterized mechanisms in the anxiety literature. Increased respiratory rate and tidal volume produce hypocapnia (reduced arterial PCO2), shifting blood pH toward respiratory alkalosis. Alkalosis causes cerebral vasoconstriction, reducing brain blood flow and producing dizziness, depersonalization, derealization, paresthesias, visual narrowing, and subjective dyspnea despite adequate oxygenation. Ley (1985) articulated the hyperventilation theory, arguing that respiratory alkalosis directly produces the majority of panic symptoms. Maddock and Carter (1991) confirmed this experimentally: voluntary hyperventilation in panic disorder patients reproduced symptom profiles closely matching their naturally occurring attacks, with symptom similarity ratings significantly higher than in controls.

The amplification mechanism operates through a positive feedback loop. Hypocapnia-induced symptoms are appraised as evidence of worsening danger, increasing sympathetic drive and respiratory rate, further depleting CO2. Papp et al. (1997) documented lower end-tidal CO2 at rest in panic disorder patients compared to controls, consistent with chronic mild hyperventilation. This lower baseline means the distance to symptomatic hypocapnia is shorter, explaining the rapid escalation many patients describe. The loop is the primary mechanism determining whether an amygdala-initiated alarm remains a brief spike or develops into a sustained episode.

Klein (1993) proposed the suffocation false alarm theory: a brainstem CO2 chemoreceptor evolved to detect asphyxia is calibrated too sensitively in panic disorder. Gorman et al. (1984) and replication by Griez and Schruers demonstrated that 35% CO2 inhalation provokes panic in patients at dramatically higher rates than controls. The hyperventilation model and suffocation alarm model address different aspects of the same system: Klein's explains initiation, the feedback model explains amplification. Both converge on a clinical truth worth carrying: the respiratory spiral is self-limiting. Brainstem chemoreceptors detect pH deviation and reset respiratory rate. Parasympathetic tone reasserts through vagal pathways. The courage embedded in this biology is that the body's corrective mechanisms hold a stronger hand than the alarm.

Clark (1986), published in Behaviour Research and Therapy, proposed that panic disorder is maintained by catastrophic misinterpretation of bodily sensations. A benign sensation is interpreted as dangerous, and the interpretation triggers anxiety that intensifies the sensation. The model is well-supported for expected attacks with identifiable cognitive triggers. Its limitation is reliance on conscious appraisal: a substantial proportion of unexpected attacks occur without reportable catastrophic thought preceding them, with patients describing panic arriving before any thought they can identify.

Bouton et al. (2001), in Psychological Review, addressed this with a conditioning model. Internal sensations present during an initial attack become conditioned stimuli through classical conditioning. Subsequent encounters with those interoceptive cues trigger conditioned anxiety without conscious appraisal, explaining unexpected attacks as responses to subliminal interoceptive triggers. Reiss and McNally (1985) provided the moderating variable: anxiety sensitivity, measured by the Anxiety Sensitivity Index. McNally (2002) demonstrated that anxiety sensitivity prospectively predicts panic symptoms after controlling for trait anxiety, establishing it as a specific risk factor rather than a proxy for general fearfulness.

Barlow (2002) integrated these threads in the triple vulnerability model: generalized biological vulnerability (heritable autonomic reactivity), generalized psychological vulnerability (learned uncontrollability), and specific psychological vulnerability (conditioned association between interoceptive cues and panic). The model explains differential outcomes: isolated attacks require all three vulnerabilities to develop into disorder. Craske and Barlow's (2007) Mastery of Your Anxiety and Panic protocol targets the specific vulnerability through interoceptive exposure, deliberately provoking feared sensations to weaken the conditioned association. Understanding the physiology described throughout this article is the mechanism that exposure targets. Facing it with courage, gradually and with professional support, turns that knowledge into lasting change.