Panic

Abstract

Panic is a form of fear: a sudden, discrete surge of intense apprehension and bodily arousal that peaks within minutes. What distinguishes a panic attack from ordinary fear is its abruptness, its intensity, and the frequency with which it arrives without a proportionate external danger, so that the alarm seems to come from inside the body rather than from the world. Psychology studies panic as the engine of panic disorder, and three accounts dominate: a neuroanatomical account locating panic in a false firing of the brain's fear network, a cognitive account in which attacks grow from the catastrophic misinterpretation of benign bodily sensations, and a learning-theory account in which panic becomes conditioned to interoceptive cues. This article defines panic against fear and anxiety, surveys its bodily and brain basis, and reviews how panic is modeled, measured, and treated.

Keywords: panic, panic attack, fear network, interoceptive conditioning, catastrophic misinterpretation

Panic is fear at its most sudden and overwhelming. Where ordinary fear is the organized response to a danger that is recognized and present, a panic attack is a fear reaction that erupts with little warning, climbs to a crescendo of pounding heart, breathlessness, and dread within minutes, and often fastens on no outside threat at all (Bouton et al., 2001). The construct matters to cognitive psychology because it sits exactly where a normal, protective emotion tips into disorder: the same circuitry that makes fear adaptive, misfiring in the absence of danger, produces an experience so aversive that sufferers reorganize their lives to avoid it. Panic is also one of the best-characterized targets in clinical science, with well-validated cognitive, pharmacological, and behavioral treatments built directly on competing theories of what a panic attack is.

What Panic Is

A panic attack is a sudden episode of intense fear or discomfort that develops abruptly and reaches a peak within minutes, accompanied by a cluster of physical and cognitive symptoms: a racing or pounding heart, sweating, trembling, shortness of breath, chest pain, nausea, dizziness, tingling, chills or heat, feelings of unreality, and the thought that one is about to die, lose control, or go mad (Roy-Byrne et al., 2006). Panic is catalogued in the MeSH thesaurus as the single descriptor filed directly beneath fear, reflecting its status as fear's most acute and abrupt form. Panic disorder, by contrast, is the clinical condition defined by recurrent unexpected attacks together with persistent worry about further attacks or a maladaptive change in behavior to avoid them.

The field draws a sharp line between panic and its neighbour anxiety. A panic attack is an acute state built to deal with a danger experienced as already in progress; anxiety is an anticipatory state that prepares the person for a threat that has not yet arrived and may never come (Bouton et al., 2001). In panic disorder the two interlock: the attacks are the acute surges, while the anxious apprehension about having another attack—sometimes called fear of fear—is the anticipatory state that fills the spaces between them. This distinction matters clinically, because the earliest evidence that panic was a distinct phenomenon came from the observation that a drug could suppress the attacks while leaving the anticipatory anxiety untouched (Klein, 1964).

A further distinction is between unexpected and cued attacks. Unexpected (uncued) attacks arrive without any obvious trigger, sometimes even waking a person from sleep, and it is these that are the hallmark of panic disorder; cued attacks occur predictably on exposure to a feared situation and are common across the anxiety disorders (Craske et al., 2017). The apparent spontaneity of unexpected attacks—an alarm with no visible cause—is precisely the puzzle that each major theory of panic sets out to solve.

Panic in the Body and Brain

Panic is as much a bodily event as a mental one, and the body is where most attacks begin. The symptoms that define an attack—cardiac, respiratory, and vestibular—are the output of a massive, abrupt discharge of the sympathetic nervous system, the same mobilization that serves genuine fear. Respiratory dysregulation is especially characteristic: many patients hyperventilate, driving down the partial pressure of carbon dioxide in the blood, which itself produces dizziness, tingling, and breathlessness and can feed the sense that something is catastrophically wrong (Meuret et al., 2008).

The influential neuroanatomical hypothesis proposes that these responses are generated by a “fear network” in the brain, centered on the amygdala and involving its interactions with the hippocampus and medial prefrontal cortex (Gorman et al., 2000). On this account a panic attack is what happens when that network, whose job is to detect threat and orchestrate defense, fires in the absence of a real external danger. Projections from the amygdala to the hypothalamus and brainstem then drive the autonomic and behavioral storm that constitutes the attack, so the physiology of a panic attack closely mirrors the physiology of a conditioned fear response in animals.

Figure 1

A Panic Attack as a False Alarm of the Fear Network

A long-standing variant of the biological account is Klein's false-suffocation-alarm theory, which proposes that patients with panic disorder have a hypersensitive brainstem monitor that misreads signs of rising carbon dioxide—a suffocation signal—and triggers a smothering sensation and the urge to flee even when oxygen is adequate. The theory grew from the observation that people with panic disorder are unusually prone to attacks when they inhale carbon dioxide (Roy-Byrne et al., 2006).

The Cognitive Model

The dominant psychological account of panic is the cognitive model, which holds that a panic attack results from the catastrophic misinterpretation of benign bodily sensations (Clark, 1986). On this view the engine of panic is not the sensation itself but its interpretation: a person notices a normal fluctuation—a quickened heartbeat, a wave of dizziness, a tightness in the chest—and reads it as the sign of an imminent catastrophe, such as a heart attack, suffocation, or loss of sanity. That interpretation generates apprehension, which intensifies the bodily sensations, which are then taken as further confirmation of danger, so the process spirals upward into a full attack.

The vicious circle explains several otherwise puzzling features of panic. It accounts for why attacks seem to come from nowhere: the triggering sensation can be too subtle to notice consciously, so the attack appears spontaneous even though it began with an internal cue. It explains why panic disorder persists: avoidance of feared sensations and reliance on “safety behaviors” prevent the person from ever learning that the sensations are harmless, so the catastrophic beliefs are never disconfirmed. And it makes a sharp treatment prediction—that changing the misinterpretation should stop the attacks—which is the basis of cognitive therapy for panic (Clark et al., 1994). The cognitive model is complementary to the neuroanatomical one: catastrophic misinterpretation can be understood as cognitive activity in prefrontal and hippocampal regions that lowers the threshold at which the fear network fires.

Learning and the Return of Panic

Modern learning theory offers a third account that integrates much of the other two. On this view, exposure to an initial panic attack—whatever its cause—acts as an aversive event that conditions anxiety, and sometimes panic itself, to the interoceptive and external cues present at the time (Bouton et al., 2001). Because the most reliable cues present during an attack are the early bodily sensations of arousal, those sensations become conditioned stimuli: a slight increase in heart rate, once neutral, comes to elicit anxiety through interoceptive conditioning. This conditioned anxiety then potentiates the next panic, which conditions more anxiety, driving the spiral into panic disorder.

The learning account draws a careful functional distinction between anxiety and panic that maps onto the clinical picture. Anxiety is an anticipatory emotional state that prepares the individual for the next panic; panic is the state designed to deal with a traumatic event already in progress (Bouton et al., 2001). The presence of conditioned anxiety lowers the threshold for the next attack, which is why the apprehensive worry between attacks is not a mere byproduct but part of the mechanism that sustains the disorder.

Learning theory also explains one of the most clinically important facts about panic: its tendency to return. Extinction—the reduction of a conditioned response when the cue is presented without the aversive outcome—is new inhibitory learning layered over the original association rather than an erasure of it, so an extinguished fear can reappear with the passage of time, a change of context, or a stressful event. This is the conceptual foundation of exposure-based treatment and the reason relapse after successful treatment is understood as the return of an incompletely inhibited conditioned response rather than a failure of the original therapy.

Measuring and Treating Panic

Panic attacks are strikingly common, but panic disorder is not. In the US National Comorbidity Survey Replication, the lifetime prevalence of isolated panic attacks without agoraphobia was 22.7%, while panic disorder without agoraphobia affected 3.7% and panic disorder with agoraphobia 1.1% (Kessler et al., 2006). The cross-national World Mental Health Surveys, covering more than 140,000 respondents across 25 countries, found a lifetime prevalence of 13.2% for panic attacks but only 1.7% for panic disorder, with a median age of onset of 32 years; notably, people who had recurrent panic attacks, rather than a single one, were at markedly elevated risk for the later onset of a range of mental disorders, suggesting that recurrent panic serves as a generic marker of vulnerability to psychopathology (de Jonge et al., 2016). Anxiety disorders as a group, of which panic disorder is one, are the most prevalent class of mental disorders, affecting up to roughly a third of people over the lifespan (Bandelow & Michaelis, 2015).

Panic is measured both by structured diagnostic interview, which establishes whether attacks are recurrent and unexpected and whether the worry and avoidance criteria are met, and by severity scales that track the frequency and intensity of attacks and the associated anticipatory anxiety and avoidance over the course of treatment. Because respiratory dysregulation is central to many attacks, physiological measures such as end-tidal carbon dioxide can serve both as an index of the disorder and as a therapeutic target (Meuret et al., 2008).

Treatment follows directly from the theories. Pharmacologically, the first evidence that panic was a distinct syndrome was Klein's finding that imipramine blocked the panic attacks while leaving anticipatory anxiety largely intact (Klein, 1964); antidepressants, particularly those acting on the serotonin system, remain a mainstay. Psychologically, cognitive therapy targets the catastrophic misinterpretations, and in a direct comparison it was superior to both applied relaxation and imipramine at the end of acute treatment, with imipramine-treated patients more likely to relapse after the drug was withdrawn (Clark et al., 1994). The large multi-center trial comparing cognitive-behavioral therapy, imipramine, and their combination found both active treatments superior to placebo, broadly comparable to each other, and the combination offering only a limited additional advantage, while CBT proved the more durable after treatment ended (Barlow et al., 2000). A Cochrane network meta-analysis of 60 studies concluded that cognitive-behavioral therapy—by far the most extensively studied psychological treatment—was often superior to the alternatives, though the effect sizes were frequently small and the overall quality of evidence low (Pompoli et al., 2016).

Worked Example

Consider the interoceptive-conditioning account made concrete with the Rescorla-Wagner rule, the standard model of how associative strength accrues. On each pairing of a bodily cue (a skipped heartbeat) with an aversive panic surge, the conditioned anxiety V attached to that cue is updated by ΔV = α(λ − V), where λ is the maximum anxiety the aversive outcome can support (set to 100 while the surge is present) and α is a learning-rate parameter combining the salience and associability of the cue (set to 0.30 here).

Starting from V = 0 for a sensation that has never preceded an attack, the acquisition pairings unfold deterministically. The first pairing adds 0.30(100 − 0) = 30, giving V = 30. The second adds 0.30(100 − 30) = 21, giving V = 51. The increments shrink each pairing because the term (λ − V) — the anxiety still left to condition — contracts as V climbs toward its asymptote, which is why a few early attacks can establish a strong fear of the sensation and further attacks add progressively less.

Table 1. Conditioned anxiety V to an interoceptive cue across eight panic pairings (α = 0.30, λ = 100).

Now switch to extinction — the cue recurs but no attack follows, so λ = 0. The same rule runs in reverse from V = 94: the first extinction trial subtracts 0.30(0 − 94) = 28, giving V = 66; the next gives 46, then 32, then 23, then 16. The conditioned anxiety decays toward zero, but the model makes the clinical point explicit: extinction drives the measured anxiety down without restoring the pre-learning state, consistent with the finding that extinction is new inhibitory learning rather than erasure and that panic can therefore return (Bouton et al., 2001). The curve the interoceptive-conditioning demo draws is this very sequence.

Discussion

The study of panic carries three lessons for cognitive psychology. First, panic is the clearest case in the emotions of a protective system misfiring: the neuroanatomical hypothesis locates the attack in a false firing of the same amygdala-centered fear network that makes ordinary fear adaptive, which is why a single construct can be approached from neuroscience, cognition, and learning theory at once and why both medication and psychotherapy relieve it (Gorman et al., 2000). This convergence is also a caution: because the three accounts each capture part of the mechanism, treating them as rivals rather than components of one pathway has at times fragmented the field.

Second, panic shows how an internal event can masquerade as an external one. The cognitive and learning accounts agree that the trigger of an apparently spontaneous attack is usually a benign bodily sensation read as catastrophe, so the alarm that feels like it comes from the world in fact begins inside the body (Clark, 1986; Bouton et al., 2001). Third, the treatment evidence makes panic a model for translational success and its limits: cognitive-behavioral therapy derived directly from the misinterpretation model is the best-supported psychological treatment and the more durable after treatment ends, yet effect sizes are often modest and relapse common, keeping the return of panic a live problem (Barlow et al., 2000; Pompoli et al., 2016).

Current Directions

Contemporary work on panic is pulling the respiratory and interoceptive strands of the older theories into sharper, more mechanistic focus. Treatments that target the physiology directly—capnometry-assisted breathing training, which teaches patients to raise their end-tidal carbon dioxide and so reverse the hyperventilation that fuels symptoms—have shown clinical benefit and point toward panic as a disorder of respiratory as well as cognitive regulation (Meuret et al., 2008). More broadly, the field has moved toward integrative models in which genetic and early-experiential vulnerabilities, interoceptive sensitivity, conditioning history, and catastrophic appraisal are not competing explanations but components of a single developmental pathway into panic disorder (Craske et al., 2017).

A second current is the reappraisal of exposure therapy through the lens of inhibitory learning. Because extinction does not erase the original conditioned association, the aim of modern exposure is understood less as reducing fear during a session and more as building robust, retrievable inhibitory memories that compete with the fear memory across contexts, so as to minimize the return of panic after treatment (Craske et al., 2017). The epidemiological finding that recurrent panic attacks predict a wide range of later disorders has also reframed panic as a possible transdiagnostic marker, raising the question of whether early intervention on recurrent attacks might alter the trajectory of psychopathology more generally (de Jonge et al., 2016).

Common Misconceptions

“A panic attack means something is physically wrong with my heart or lungs.”

The cardiac and respiratory symptoms of panic are the output of a normal fear response firing without a real threat; the cognitive model holds that reading them as signs of physical catastrophe is itself part of what drives the attack (Clark, 1986).

“Everyone who has a panic attack has panic disorder.”

Isolated panic attacks are far more common than the disorder: lifetime attack prevalence runs around 13–23%, while panic disorder affects only a few percent of people (Kessler et al., 2006).

“Unexpected attacks have no cause.”

An attack can feel spontaneous because its trigger is an internal bodily cue too subtle to notice, not because it is genuinely uncaused (Clark, 1986).

“Panic and anxiety are the same thing.”

Panic is an acute surge dealing with a danger experienced as present; anxiety is the anticipatory state that prepares for a threat not yet arrived. In panic disorder the anxious “fear of fear” between attacks is distinct from the attacks themselves (Bouton et al., 2001).

Glossary

Agoraphobia.

Fear and avoidance of situations from which escape might be difficult or help unavailable should a panic attack occur, frequently complicating panic disorder.

Anticipatory anxiety.

The apprehensive state between attacks, directed at the possibility of a future attack; the “fear of fear” that sustains panic disorder.

Catastrophic misinterpretation.

The reading of a benign bodily sensation as a sign of imminent disaster, the central mechanism of the cognitive model of panic.

Cued attack.

A panic attack that occurs predictably on exposure to a feared situation or stimulus, in contrast to the unexpected attacks that define panic disorder.

Exposure therapy.

A behavioral treatment in which the patient confronts feared sensations or situations without avoidance, building inhibitory learning that competes with the conditioned fear.

Extinction.

The reduction of a conditioned response when the cue is presented without the aversive outcome; understood as new inhibitory learning layered over the original association rather than its erasure.

False-suffocation alarm.

A hypothesized hypersensitive brainstem monitor that misreads carbon-dioxide signals as suffocation, triggering panic; proposed to explain the sensitivity of patients to CO₂ inhalation.

Fear network.

The amygdala-centered brain circuit, interacting with the hippocampus and medial prefrontal cortex, whose false firing the neuroanatomical hypothesis identifies with a panic attack.

Hyperventilation.

Overbreathing that lowers the blood's carbon-dioxide level, producing dizziness, tingling, and breathlessness that can intensify a panic attack.

Interoceptive conditioning.

Pavlovian conditioning in which internal bodily sensations serve as the conditioned stimulus, so that early signs of arousal come to elicit anxiety or panic.

Panic attack.

An abrupt surge of intense fear or discomfort that peaks within minutes, with a defined cluster of physical and cognitive symptoms.

Panic disorder.

The clinical condition defined by recurrent unexpected panic attacks together with persistent worry about further attacks or maladaptive avoidance.

Safety behavior.

An action taken to avert a feared catastrophe during panic, such as sitting down or carrying medication, which prevents the catastrophic belief from being disconfirmed and so helps maintain the disorder.

Unexpected (uncued) attack.

A panic attack that arises without an identifiable trigger, the hallmark of panic disorder, as opposed to a cued attack provoked by a feared situation.

Key Researchers

David H. Barlow

. Professor Emeritus of Psychology and Psychiatry and founder of the Center for Anxiety and Related Disorders at Boston University; developed and validated cognitive-behavioral treatments for panic disorder and led the landmark multi-site trial comparing CBT, imipramine, and their combination. Google Scholar - Wikipedia

Mark E. Bouton

. Lawson Professor of Psychology at the University of Vermont; built the modern learning-theory account of panic disorder, showing how Pavlovian conditioning to interoceptive cues, context, and the return of extinguished fear explain the onset and relapse of panic. Google Scholar

David M. Clark

. Professor of Experimental Psychology at the University of Oxford; formulated the cognitive model of panic, in which attacks arise from the catastrophic misinterpretation of benign bodily sensations, and developed the cognitive therapy that follows from it. Google Scholar - Wikipedia

Michelle G. Craske

. Distinguished Professor of Psychology, Psychiatry and Biobehavioral Sciences at the University of California, Los Angeles; advanced the understanding and treatment of panic and anxiety disorders, including inhibitory-learning approaches to exposure therapy and large-scale syntheses of panic disorder. Google Scholar - Wikipedia

Jack M. Gorman

. Psychiatrist and researcher, formerly Professor of Psychiatry at Columbia University and the Icahn School of Medicine at Mount Sinai; proposed and revised the neuroanatomical hypothesis of panic disorder, locating panic in a fear network centered on the amygdala. Google Scholar

Donald F. Klein

(1928-2019). Late Professor of Psychiatry at Columbia University; delineated panic as a distinct, drug-responsive anxiety syndrome, showing that imipramine blocked panic attacks but not anticipatory anxiety, and proposed the false-suffocation-alarm theory. Wikipedia

Frequently Asked Questions

What is a panic attack?

A panic attack is an abrupt surge of intense fear or discomfort that peaks within minutes and is accompanied by physical symptoms such as a pounding heart, breathlessness, dizziness, and chest pain, together with thoughts of imminent catastrophe such as dying or losing control (Roy-Byrne et al., 2006).

What is the difference between a panic attack and panic disorder?

A panic attack is a single episode; panic disorder is the clinical condition defined by recurrent unexpected attacks plus persistent worry about further attacks or avoidance behavior. Isolated attacks are common, but only a minority of those who have them develop the disorder (Kessler et al., 2006).

What causes panic attacks?

Three accounts dominate: a false firing of an amygdala-centered fear network (Gorman et al., 2000), the catastrophic misinterpretation of benign bodily sensations (Clark, 1986), and conditioned anxiety to interoceptive cues following an initial attack (Bouton et al., 2001). These are complementary rather than mutually exclusive.

How is panic different from anxiety?

Panic is an acute state that deals with a danger experienced as present; anxiety is an anticipatory state that prepares for a threat not yet arrived. In panic disorder, the anxious “fear of fear” between attacks is distinct from the attacks themselves (Bouton et al., 2001).

Why do panic attacks seem to come from nowhere?

An unexpected attack can feel spontaneous because its trigger is an internal bodily sensation—a subtle change in heartbeat or breathing—too faint to notice consciously, which is then read as a threat and amplified into a full attack (Clark, 1986).

How common is panic disorder?

Lifetime prevalence of panic disorder is roughly 1–4% depending on the survey—about 3.7% without agoraphobia in the US National Comorbidity Survey Replication and 1.7% cross-nationally—whereas isolated panic attacks are far more common, around 13–23% over the lifespan (Kessler et al., 2006; de Jonge et al., 2016).

What treatments work for panic disorder?

Cognitive-behavioral therapy is the best-supported psychological treatment and is often superior to the alternatives, while antidepressants acting on the serotonin system are an effective medication option; a large trial found both superior to placebo and CBT the more durable after treatment ended (Barlow et al., 2000; Pompoli et al., 2016).

Does hyperventilation play a role in panic?

Yes. Many attacks involve hyperventilation, which lowers blood carbon dioxide and produces dizziness, tingling, and breathlessness that can intensify the attack; breathing treatments that raise end-tidal carbon dioxide have shown clinical benefit (Meuret et al., 2008).

References